Add HasStatus instance for Headers (that defers StatusOf to underlying value) (#1649)

* Add HasStatus instance for Headers (that defers StatusOf to underlying value)

* changelog.d/1649

.github/FUNDING.yml

@@ -0,0 +1,12 @@
+# These are supported funding model platforms
+
+github: [arianvp]
+patreon: # Replace with a single Patreon username
+open_collective: # Replace with a single Open Collective username
+ko_fi: # Replace with a single Ko-fi username
+tidelift: # Replace with a single Tidelift platform-name/package-name e.g., npm/babel
+community_bridge: # Replace with a single Community Bridge project-name e.g., cloud-foundry
+liberapay: # Replace with a single Liberapay username
+issuehunt: # Replace with a single IssueHunt username
+otechie: # Replace with a single Otechie username
+custom: https://github.com/haskell-servant/haskell-servant.github.io/blob/hakyll/consultancies.md

.github/run-ghcjs-tests.sh

@@ -0,0 +1,14 @@
+#!/usr/bin/env bash
+#
+# cabal v2-test does not work with GHCJS
+# See: https://github.com/haskell/cabal/issues/6175
+#
+# This invokes cabal-plan to figure out test binaries, and invokes them with node.
+
+cabal-plan list-bins '*:test:*' | while read -r line
+do
+    testpkg=$(echo "$line" | perl -pe 's/:.*//')
+    testexe=$(echo "$line" | awk '{ print $2 }')
+    echo "testing $textexe in package $textpkg"
+    (cd "$testpkg" && node "$testexe".jsexe/all.js)
+done

.github/workflows/master.yml

@@ -0,0 +1,148 @@
+name: CI
+
+# Trigger the workflow on push or pull request, but only for the master branch
+on:
+  pull_request:
+  push:
+    branches: [master]
+
+jobs:
+  cabal:
+    name: ${{ matrix.os }} / ghc ${{ matrix.ghc }}
+    runs-on: ${{ matrix.os }}
+    strategy:
+      matrix:
+        os: [ubuntu-latest]
+        cabal: ["3.6"]
+        ghc:
+          - "8.6.5"
+          - "8.8.4"
+          - "8.10.7"
+          - "9.0.2"
+          - "9.2.2"
+          - "9.4.2"
+
+    steps:
+    - uses: actions/checkout@v2
+
+    - uses: haskell/actions/setup@v1
+      id: setup-haskell-cabal
+      name: Setup Haskell
+      with:
+        ghc-version: ${{ matrix.ghc }}
+        cabal-version: ${{ matrix.cabal }}
+
+    - name: Freeze
+      run: |
+        cabal configure --enable-tests --enable-benchmarks --test-show-details=direct
+        cabal freeze
+
+    - uses: actions/cache@v2.1.3
+      name: Cache ~/.cabal/store and dist-newstyle
+      with:
+        path: |
+          ${{ steps.setup-haskell-cabal.outputs.cabal-store }}
+          dist-newstyle
+        key: ${{ runner.os }}-${{ matrix.ghc }}-${{ hashFiles('cabal.project.freeze') }}
+        restore-keys: |
+          ${{ runner.os }}-${{ matrix.ghc }}-
+
+    - name: Configure
+      run: |
+        cabal install --ignore-project -j2 doctest --constraint='doctest ^>=0.20'
+
+    - name: Build
+      run: |
+        cabal build all
+
+    - name: Test
+      run: |
+        cabal test all
+
+    - name: Run doctests
+      run: |
+        # Necessary for doctest to be found in $PATH
+        export PATH="$HOME/.cabal/bin:$PATH"
+
+        DOCTEST="cabal repl --with-ghc=doctest --ghc-options=-w"
+        (cd servant && eval $DOCTEST)
+        (cd servant-client && eval $DOCTEST)
+        (cd servant-client-core && eval $DOCTEST)
+        (cd servant-http-streams && eval $DOCTEST)
+        (cd servant-docs && eval $DOCTEST)
+        (cd servant-foreign && eval $DOCTEST)
+        (cd servant-server && eval $DOCTEST)
+        (cd servant-machines && eval $DOCTEST)
+        (cd servant-conduit && eval $DOCTEST)
+        (cd servant-pipes && eval $DOCTEST)
+
+  # stack:
+  #   name: stack / ghc ${{ matrix.ghc }}
+  #   runs-on: ubuntu-latest
+  #   strategy:
+  #     matrix:
+  #       stack: ["2.7.5"]
+  #       ghc: ["8.10.7"]
+
+  #   steps:
+  #   - uses: actions/checkout@v2
+
+  #   - uses: haskell/actions/setup@v1
+  #     name: Setup Haskell Stack
+  #     with:
+  #       ghc-version: ${{ matrix.ghc }}
+  #       stack-version: ${{ matrix.stack }}
+
+  #   - uses: actions/cache@v2.1.3
+  #     name: Cache ~/.stack
+  #     with:
+  #       path: ~/.stack
+  #       key: ${{ runner.os }}-${{ matrix.ghc }}-stack
+
+  #   - name: Install dependencies
+  #     run: |
+  #       stack build --system-ghc --test --bench --no-run-tests --no-run-benchmarks --only-dependencies
+
+  #   - name: Build
+  #     run: |
+  #       stack build --system-ghc --test --bench --no-run-tests --no-run-benchmarks
+
+  #   - name: Test
+  #     run: |
+  #       stack test --system-ghc
+
+  ghcjs:
+    name: ubuntu-latest / ghcjs 8.6
+    runs-on: "ubuntu-latest"
+
+    steps:
+    - uses: actions/checkout@v2
+    - uses: cachix/install-nix-action@v13
+      with:
+        extra_nix_config: |
+          trusted-public-keys = ryantrinkle.com-1:JJiAKaRv9mWgpVAz8dwewnZe0AzzEAzPkagE9SP5NWI=1aba6f367982bd6dd78ec2fda75ab246a62d32c5 cache.nixos.org-1:6NCHdD59X431o0gWypbMrAURkbJ16ZPMQFGspcDShjY=
+          substituters = https://nixcache.reflex-frp.org https://cache.nixos.org/
+    - name: Setup
+      run: |
+        # Override cabal.project with the lightweight GHCJS one
+        cp cabal.ghcjs.project cabal.project
+        cat cabal.project
+        nix-shell ghcjs.nix --run "cabal v2-update && cabal v2-freeze"
+
+    - uses: actions/cache@v2.1.3
+      name: Cache ~/.cabal/store and dist-newstyle
+      with:
+        path: |
+          ~/.cabal/store
+          dist-newstyle
+        key: ${{ runner.os }}-ghcjs8.6-${{ hashFiles('cabal.project.freeze') }}
+        restore-keys: |
+          ${{ runner.os }}-ghcjs8.6-
+
+    - name: Build
+      run: |
+        nix-shell ghcjs.nix --run "cabal v2-build --ghcjs --enable-tests --enable-benchmarks all"
+
+    - name: Tests
+      run: |
+        nix-shell ghcjs.nix --run ".github/run-ghcjs-tests.sh"

.gitignore

@@ -1,5 +1,6 @@
 **/*/dist
-dist-newstyle
+*~
+dist-*
 .ghc.environment.*
 /bin
 /lib
@@ -29,3 +30,11 @@ doc/_build
 doc/venv
 doc/tutorial/static/api.js
 doc/tutorial/static/jq.js
+shell.nix
+.hspec-failures
+
+# nix
+result*
+
+# local versions of things
+servant-multipart

.travis.yml

@@ -1,171 +0,0 @@
-# This Travis job script has been generated by a script via
-#
-#   runghc make_travis_yml_2.hs '--config=cabal.make-travis-yml' '--output=.travis.yml' '--max-backjumps=10000' 'cabal.project'
-#
-# For more information, see https://github.com/hvr/multi-ghc-travis
-#
-language: c
-sudo: false
-
-git:
-  submodules: false  # whether to recursively clone submodules
-
-branches:
-  only:
-    - master
-    - release-0.12
-
-cache:
-  directories:
-    - $HOME/.cabal/packages
-    - $HOME/.cabal/store
-
-before_cache:
-  - rm -fv $HOME/.cabal/packages/hackage.haskell.org/build-reports.log
-  # remove files that are regenerated by 'cabal update'
-  - rm -fv $HOME/.cabal/packages/hackage.haskell.org/00-index.*
-  - rm -fv $HOME/.cabal/packages/hackage.haskell.org/*.json
-  - rm -fv $HOME/.cabal/packages/hackage.haskell.org/01-index.cache
-  - rm -fv $HOME/.cabal/packages/hackage.haskell.org/01-index.tar
-  - rm -fv $HOME/.cabal/packages/hackage.haskell.org/01-index.tar.idx
-
-  - rm -rfv $HOME/.cabal/packages/head.hackage
-
-matrix:
-  include:
-    - compiler: "ghc-7.8.4"
-    # env: TEST=--disable-tests BENCH=--disable-benchmarks
-      addons: {apt: {packages: [ghc-ppa-tools,cabal-install-head,ghc-7.8.4], sources: [hvr-ghc]}}
-    - compiler: "ghc-7.10.3"
-    # env: TEST=--disable-tests BENCH=--disable-benchmarks
-      addons: {apt: {packages: [ghc-ppa-tools,cabal-install-head,ghc-7.10.3], sources: [hvr-ghc]}}
-    - compiler: "ghc-8.0.2"
-    # env: TEST=--disable-tests BENCH=--disable-benchmarks
-      addons: {apt: {packages: [ghc-ppa-tools,cabal-install-head,ghc-8.0.2], sources: [hvr-ghc]}}
-    - compiler: "ghc-8.2.2"
-    # env: TEST=--disable-tests BENCH=--disable-benchmarks
-      addons: {apt: {packages: [ghc-ppa-tools,cabal-install-head,ghc-8.2.2], sources: [hvr-ghc]}}
-
-before_install:
-  - HC=${CC}
-  - HCPKG=${HC/ghc/ghc-pkg}
-  - unset CC
-  - ROOTDIR=$(pwd)
-  - mkdir -p $HOME/.local/bin
-  - "PATH=/opt/ghc/bin:/opt/ghc-ppa-tools/bin:$HOME/local/bin:$PATH"
-  - HCNUMVER=$(( $(${HC} --numeric-version|sed -E 's/([0-9]+)\.([0-9]+)\.([0-9]+).*/\1 * 10000 + \2 * 100 + \3/') ))
-  - echo $HCNUMVER
-
-install:
-  - cabal --version
-  - echo "$(${HC} --version) [$(${HC} --print-project-git-commit-id 2> /dev/null || echo '?')]"
-  - BENCH=${BENCH---enable-benchmarks}
-  - TEST=${TEST---enable-tests}
-  - HADDOCK=${HADDOCK-true}
-  - INSTALLED=${INSTALLED-true}
-  - GHCHEAD=${GHCHEAD-false}
-  - CABALNEWBUILDOPTS=--max-backjumps=10000
-  - travis_retry cabal update -v
-  - "sed -i.bak 's/^jobs:/-- jobs:/' ${HOME}/.cabal/config"
-  - rm -fv cabal.project cabal.project.local
-  - "if [ $HCNUMVER -ge 70800 ]; then sed -i.bak 's/-- ghc-options:.*/ghc-options: -j2/' ${HOME}/.cabal/config; fi"
-  - grep -Ev -- '^\s*--' ${HOME}/.cabal/config | grep -Ev '^\s*$'
-  - "printf 'packages: \"servant\" \"servant-client\" \"servant-client-core\" \"servant-docs\" \"servant-foreign\" \"servant-server\" \"doc/tutorial\" \"doc/cookbook/db-postgres-pool\" \"doc/cookbook/jwt-and-basic-auth\" \"doc/cookbook/db-sqlite-simple\" \"doc/cookbook/basic-auth\" \"doc/cookbook/https\" \"doc/cookbook/structuring-apis\" \"doc/cookbook/using-custom-monad\" \"doc/cookbook/file-upload\"\\n' > cabal.project"
-  - "echo 'constraints: foundation >=0.0.14,memory <0.14.12 || >0.14.12' >> cabal.project"
-  - "echo 'allow-newer: servant-js:servant,servant-js:servant-foreign,servant-auth-server:http-types,servant-multipart:lens,servant-multipart:resourcet,servant-multipart:servant,servant-multipart:servant-server,servant-auth-server:servant-server' >> cabal.project"
-  - cat cabal.project
-  - if [ -f "servant/configure.ac" ]; then
-      (cd "servant" && autoreconf -i);
-    fi
-  - if [ -f "servant-client/configure.ac" ]; then
-      (cd "servant-client" && autoreconf -i);
-    fi
-  - if [ -f "servant-client-core/configure.ac" ]; then
-      (cd "servant-client-core" && autoreconf -i);
-    fi
-  - if [ -f "servant-docs/configure.ac" ]; then
-      (cd "servant-docs" && autoreconf -i);
-    fi
-  - if [ -f "servant-foreign/configure.ac" ]; then
-      (cd "servant-foreign" && autoreconf -i);
-    fi
-  - if [ -f "servant-server/configure.ac" ]; then
-      (cd "servant-server" && autoreconf -i);
-    fi
-  - if [ -f "doc/tutorial/configure.ac" ]; then
-      (cd "doc/tutorial" && autoreconf -i);
-    fi
-  - if [ -f "doc/cookbook/db-postgres-pool/configure.ac" ]; then
-      (cd "doc/cookbook/db-postgres-pool" && autoreconf -i);
-    fi
-  - if [ -f "doc/cookbook/jwt-and-basic-auth/configure.ac" ]; then
-      (cd "doc/cookbook/jwt-and-basic-auth" && autoreconf -i);
-    fi
-  - if [ -f "doc/cookbook/db-sqlite-simple/configure.ac" ]; then
-      (cd "doc/cookbook/db-sqlite-simple" && autoreconf -i);
-    fi
-  - if [ -f "doc/cookbook/basic-auth/configure.ac" ]; then
-      (cd "doc/cookbook/basic-auth" && autoreconf -i);
-    fi
-  - if [ -f "doc/cookbook/https/configure.ac" ]; then
-      (cd "doc/cookbook/https" && autoreconf -i);
-    fi
-  - if [ -f "doc/cookbook/structuring-apis/configure.ac" ]; then
-      (cd "doc/cookbook/structuring-apis" && autoreconf -i);
-    fi
-  - if [ -f "doc/cookbook/using-custom-monad/configure.ac" ]; then
-      (cd "doc/cookbook/using-custom-monad" && autoreconf -i);
-    fi
-  - if [ -f "doc/cookbook/file-upload/configure.ac" ]; then
-      (cd "doc/cookbook/file-upload" && autoreconf -i);
-    fi
-  - rm -f cabal.project.freeze
-  - rm -rf .ghc.environment.* "servant"/dist "servant-client"/dist "servant-client-core"/dist "servant-docs"/dist "servant-foreign"/dist "servant-server"/dist "doc/tutorial"/dist "doc/cookbook/db-postgres-pool"/dist "doc/cookbook/jwt-and-basic-auth"/dist "doc/cookbook/db-sqlite-simple"/dist "doc/cookbook/basic-auth"/dist "doc/cookbook/https"/dist "doc/cookbook/structuring-apis"/dist "doc/cookbook/using-custom-monad"/dist "doc/cookbook/file-upload"/dist
-  - DISTDIR=$(mktemp -d /tmp/dist-test.XXXX)
-
-# Here starts the actual work to be performed for the package under test;
-# any command which exits with a non-zero exit code causes the build to fail.
-script:
-  # test that source-distributions can be generated
-  - echo Packaging... && echo -en 'travis_fold:start:sdist\\r'
-  - (cd "servant" && cabal sdist)
-  - (cd "servant-client" && cabal sdist)
-  - (cd "servant-client-core" && cabal sdist)
-  - (cd "servant-docs" && cabal sdist)
-  - (cd "servant-foreign" && cabal sdist)
-  - (cd "servant-server" && cabal sdist)
-  - (cd "doc/tutorial" && cabal sdist)
-  - (cd "doc/cookbook/db-postgres-pool" && cabal sdist)
-  - (cd "doc/cookbook/jwt-and-basic-auth" && cabal sdist)
-  - (cd "doc/cookbook/db-sqlite-simple" && cabal sdist)
-  - (cd "doc/cookbook/basic-auth" && cabal sdist)
-  - (cd "doc/cookbook/https" && cabal sdist)
-  - (cd "doc/cookbook/structuring-apis" && cabal sdist)
-  - (cd "doc/cookbook/using-custom-monad" && cabal sdist)
-  - (cd "doc/cookbook/file-upload" && cabal sdist)
-  - echo -en 'travis_fold:end:sdist\\r'
-  - echo Unpacking... && echo -en 'travis_fold:start:unpack\\r'
-  - mv "servant"/dist/servant-*.tar.gz "servant-client"/dist/servant-client-*.tar.gz "servant-client-core"/dist/servant-client-core-*.tar.gz "servant-docs"/dist/servant-docs-*.tar.gz "servant-foreign"/dist/servant-foreign-*.tar.gz "servant-server"/dist/servant-server-*.tar.gz "doc/tutorial"/dist/tutorial-*.tar.gz "doc/cookbook/db-postgres-pool"/dist/cookbook-db-postgres-pool-*.tar.gz "doc/cookbook/jwt-and-basic-auth"/dist/cookbook-jwt-and-basic-auth-*.tar.gz "doc/cookbook/db-sqlite-simple"/dist/cookbook-db-sqlite-simple-*.tar.gz "doc/cookbook/basic-auth"/dist/cookbook-basic-auth-*.tar.gz "doc/cookbook/https"/dist/cookbook-https-*.tar.gz "doc/cookbook/structuring-apis"/dist/cookbook-structuring-apis-*.tar.gz "doc/cookbook/using-custom-monad"/dist/cookbook-using-custom-monad-*.tar.gz "doc/cookbook/file-upload"/dist/cookbook-file-upload-*.tar.gz ${DISTDIR}/
-  - cd ${DISTDIR} || false
-  - find . -maxdepth 1 -name '*.tar.gz' -exec tar -xvf '{}' \;
-  - "printf 'packages: servant-*/*.cabal servant-client-*/*.cabal servant-client-core-*/*.cabal servant-docs-*/*.cabal servant-foreign-*/*.cabal servant-server-*/*.cabal tutorial-*/*.cabal cookbook-db-postgres-pool-*/*.cabal cookbook-jwt-and-basic-auth-*/*.cabal cookbook-db-sqlite-simple-*/*.cabal cookbook-basic-auth-*/*.cabal cookbook-https-*/*.cabal cookbook-structuring-apis-*/*.cabal cookbook-using-custom-monad-*/*.cabal cookbook-file-upload-*/*.cabal\\n' > cabal.project"
-  - "echo 'constraints: foundation >=0.0.14,memory <0.14.12 || >0.14.12' >> cabal.project"
-  - "echo 'allow-newer: servant-js:servant,servant-js:servant-foreign,servant-auth-server:http-types,servant-multipart:lens,servant-multipart:resourcet,servant-multipart:servant,servant-multipart:servant-server,servant-auth-server:servant-server' >> cabal.project"
-  - cat cabal.project
-  - echo -en 'travis_fold:end:unpack\\r'
-
-
-  - echo Building with tests and benchmarks... && echo -en 'travis_fold:start:build-everything\\r'
-  # build & run tests, build benchmarks
-  - cabal new-build -w ${HC} ${TEST} ${BENCH} ${CABALNEWBUILDOPTS} all
-  - echo -en 'travis_fold:end:build-everything\\r'
-  - if [ "x$TEST" = "x--enable-tests" ]; then cabal new-test -w ${HC} ${TEST} ${BENCH} ${CABALNEWBUILDOPTS} all; fi
-
-  - echo Haddock... && echo -en 'travis_fold:start:haddock\\r'
-  # haddock
-  - rm -rf ./dist-newstyle
-  - if $HADDOCK; then cabal new-haddock -w ${HC} ${TEST} ${BENCH} ${CABALNEWBUILDOPTS} all; else echo "Skipping haddock generation";fi
-
-  - echo -en 'travis_fold:end:haddock\\r'
-# REGENDATA ["--config=cabal.make-travis-yml","--output=.travis.yml","--max-backjumps=10000","cabal.project"]
-# EOF

CONTRIBUTING.md

@@ -21,6 +21,7 @@ Or `nix`:
 ./scripts/generate-nix-files.sh   # Get up-to-date shell.nix files
 ```
 
+To build the docs, see `doc/README.md`.
 
 ## General
 
@@ -34,8 +35,34 @@ Some things we like:
 Though we aren't sticklers for style, the `.stylish-haskell.yaml` and `HLint.hs`
 files in the repository provide a good baseline for consistency.
 
-Please include a description of the changes in your PR in the `CHANGELOG.md` of
-the packages you've changed. And of course, write tests!
+**Important**: please do not modify the versions of the servant packages you are sending patches for.
+
+## Changelog entries
+
+We experiment with using [changelog-d tool](https://github.com/phadej/changelog-d) to assemble changelogs.
+You are not required to install it.
+
+In each PR please add a file to `changelog.d` directory named after issue you are solving or the pull request itself (in a separate commit after you know the pull request number). For example
+
+```cabal
+synopsis: One sentence summary of the change.
+prs: #1219
+issues: #1028
+
+description: {
+
+A longer description. Small changes don't need this.
+Bigger ones definitely do, for example we try to include migration hints
+for breaking changes.
+
+However if you don't know what to write, that's ok too.
+
+By the way, the braces around are omitted when the file is parsed.
+They can be used so the field doesn't need to be indented, which is handy
+for prose.
+
+}
+```
 
 ## PR process
 
@@ -52,8 +79,10 @@ not been a timely response to a PR, you can ping the Maintainers group (with
 We encourage people to experiment with new combinators and instances - it is
 one of the most powerful ways of using `servant`, and a wonderful way of
 getting to know it better. If you do write a new combinator, we would love to
-know about it! Either hop on #servant on freenode and let us know, or open an
-issue with the `news` tag (which we will close when we read it).
+know about it! Either hop on
+[#haskell-servant on libera.chat](https://web.libera.chat/#haskell-servant) and
+let us know, or open an issue with the `news` tag (which we will close when we
+read it).
 
 As for adding them to the main repo: maintaining combinators can be expensive,
 since official combinators must have instances for all classes (and new classes
@@ -76,7 +105,7 @@ the `news` label if you make a new package so we can know about it!
 
 ## Release policy
 
-We are currently moving to a more aggresive release policy, so that you can get
+We are currently moving to a more aggressive release policy, so that you can get
 what you contribute from Hackage fairly soon. However, note that prior to major
 releases it may take some time in between releases.
 

README.md

@@ -4,37 +4,18 @@
 
 ## Getting Started
 
-We have a [tutorial](http://haskell-servant.readthedocs.org/en/stable/tutorial/index.html) that
+We have a [tutorial](http://docs.servant.dev/en/stable/tutorial/index.html) that
 introduces the core features of servant. After this article, you should be able
 to write your first servant webservices, learning the rest from the haddocks'
 examples.
 
-The central documentation can be found [here](http://haskell-servant.readthedocs.org/).
+The core documentation can be found [here](http://docs.servant.dev/).
 Other blog posts, videos and slides can be found on the
-[website](http://haskell-servant.github.io/).
+[website](http://www.servant.dev/).
 
-If you need help, drop by the IRC channel (#servant on freenode) or [mailing
+If you need help, drop by the IRC channel (#haskell-servant on libera.chat) or [mailing
 list](https://groups.google.com/forum/#!forum/haskell-servant).
 
-## Version history
-
-This table lists the versions of some `servant-` libraries at the point of
-release of `servant` package.
-
-|                     | **0.10** | **0.11** | **0.12** |
-| ------------------- | -------- |----------|----------|
-| servant             | 0.10     | 0.11     | 0.12     |
-| servant-blaze       | 0.7.1    | ?        | ?        |
-| servant-cassava     | 0.7      | ?        | ?        |
-| servant-client      | 0.10     | 0.11     | 0.12     |
-| servant-docs        | 0.10     | 0.11     | 0.11.1   |
-| servant-foreign     | 0.10     | 0.10.0.1 | 0.10.2   |
-| servant-js          | 0.9.1    | ?        | ?        |
-| servant-lucid       | 0.7.1    | ?        | ?        |
-| servant-mock        | 0.8.1.1  | ?        | ?        |
-| servant-server      | 0.10     | 0.11     | 0.12     |
-| servant-swagger     | 1.1.2.1  | ?        | ?        |
-
 ## Contributing
 
 See `CONTRIBUTING.md`
@@ -43,7 +24,7 @@ See `CONTRIBUTING.md`
 
 - Update changelog and bump versions in `master`
     - `git log --oneline v0.12.. | grep 'Merge pull request'` is a good starting point (use correct previous release tag)
-- Create a release branch, e.g. `release-0.13`, and *protect it* from accidental force pushes.
+- Create a release branch, e.g. `release-0.13`
     - Release branch is useful for backporting fixes from `master`
 - Smoke test in [`servant-universe`](https://github.com/phadej/servant-universe)
     - `git submodule foreach git checkout master` and `git submodule foreach git pull` to get newest of everything.
@@ -51,7 +32,7 @@ See `CONTRIBUTING.md`
         - It's a good idea to separate these steps, as tests often pass, if they compile :)
     - See `cabal.project` to selectively `allow-newer`
     - If some packages are broken, on your discretisation there are two options:
-        - Fix them and make PRs: it's good idea to test against older `servant` version too.
+        - Fix them and make PRs: it's a good idea to test against older `servant` version too.
         - Temporarily comment out broken package
     - If you make a commit for `servant-universe`, you can use it as submodule in private projects to test even more
 - When ripples are cleared out:
@@ -59,22 +40,32 @@ See `CONTRIBUTING.md`
     - `git push --tags`
     - `cabal sdist` and `cabal upload`
 
-## travis
+## TechEmpower framework benchmarks
 
-`.travis.yml` is generated using `make-travis-yml` tool, in
-[multi-ghc-travis](https://github.com/haskell-hvr/multi-ghc-travis) repository.
+We develop and maintain the servant TFB entry in https://github.com/haskell-servant/FrameworkBenchmarks/
 
-To regenerate the script use (*note:* atm you need to comment `doc/cookbook/` packages).
+To verify (i.e. compile and test that it works)
 
-```
-runghc ~/Documents/other-haskell/multi-ghc-travis/make_travis_yml_2.hs regenerate
+```sh
+./tfb --mode verify --test servant servant-beam servant-psql-simple --type json plaintext db fortune
 ```
 
-In case Travis jobs fail due failing build of dependency, you can temporarily
-add `constraints` to the `cabal.project`, and regenerate the `.travis.yml`.
-For example, the following will disallow single `troublemaker-13.37` package version:
+To compare with  `warp`
 
+```sh
+./tfb --mode benchmark --test warp servant servant-beam servant-psql-simple --type json plaintext db fortune
 ```
-constraints:
-  troublemaker <13.37 && > 13.37
+
+To compare with `reitit` (Clojure framework)
+
+```sh
+./tfb --mode benchmark --test reitit reitit-async reitit-jdbc servant servant-beam servant-psql-simple --type json plaintext db fortune
 ```
+
+You can see the visualised results at https://www.techempower.com/benchmarks/#section=test
+
+## Nix
+
+A developer shell.nix file is provided in the `nix` directory
+
+See  [nix/README.md](nix/README.md)

cabal.ghcjs.project

@@ -0,0 +1,14 @@
+-- Using https://launchpad.net/~hvr/+archive/ubuntu/ghcjs
+
+packages:
+  servant/
+  servant-client/
+  servant-client-core/
+
+-- we need to tell cabal we are using GHCJS
+compiler: ghcjs
+tests: True
+
+-- Constraints so that reflex-platform provided packages are selected.
+constraints: attoparsec == 0.13.2.2
+constraints: hashable == 1.3.0.0

cabal.make-travis-yml

@@ -1,15 +0,0 @@
-folds: all-but-test
-branches: master release-0.12
-
--- We have inplace packages (servant-js) so we skip installing dependencies in a separate step
-install-dependencies-step: False
-
--- this speed-ups the build a little, but we have to check these for release
-no-tests-no-benchmarks: False
-build-with-installed-step: False
-
--- Don't run cabal check, as cookbook examples won't pass it
-cabal-check: False
-
--- ghc-options: -j2
-jobs: :2

cabal.project

@@ -1,23 +1,54 @@
-packages: servant/
+packages:
+  servant/
+  servant-auth/servant-auth
+  servant-auth/servant-auth-client
+  servant-auth/servant-auth-docs
+  servant-auth/servant-auth-server
+  servant-auth/servant-auth-swagger
+
   servant-client/
   servant-client-core/
+  servant-http-streams/
   servant-docs/
   servant-foreign/
   servant-server/
+  servant-swagger/
   doc/tutorial/
-  doc/cookbook/*/*.cabal
 
-allow-newer:
-  servant-js:servant,
-  servant-js:servant-foreign,
-  servant-auth-server:http-types,
-  servant-multipart:lens,
-  servant-multipart:resourcet,
-  servant-multipart:servant,
-  servant-multipart:servant-server,
-  servant-auth-server:servant-server
+-- servant streaming
+packages:
+  servant-machines/
+  servant-conduit/
+  servant-pipes/
 
-constraints:
-  -- see https://github.com/haskell-infra/hackage-trustees/issues/119
-  foundation >=0.0.14,
-  memory <0.14.12 || >0.14.12
+-- servant GHCJS
+-- packages:
+--   servant-jsaddle/
+
+-- Cookbooks
+packages:
+  doc/cookbook/basic-auth
+  doc/cookbook/curl-mock
+  doc/cookbook/custom-errors
+  doc/cookbook/basic-streaming
+  doc/cookbook/db-postgres-pool
+  doc/cookbook/db-sqlite-simple
+  doc/cookbook/file-upload
+  doc/cookbook/generic
+  doc/cookbook/hoist-server-with-context
+  doc/cookbook/https
+  doc/cookbook/jwt-and-basic-auth
+  doc/cookbook/pagination
+  -- doc/cookbook/sentry
+  -- Commented out because servant-quickcheck currently doesn't build.
+  -- doc/cookbook/testing
+  doc/cookbook/uverb
+  doc/cookbook/structuring-apis
+  doc/cookbook/using-custom-monad
+  doc/cookbook/using-free-client
+  -- doc/cookbook/open-id-connect
+  doc/cookbook/managed-resource
+
+tests: True
+optimization: False
+-- reorder-goals: True

changelog.d/1432

@@ -0,0 +1,9 @@
+synopsis: Fixes encoding of URL parameters in servant-client
+prs: #1432
+issues: #1418
+description: {
+Some applications use query parameters to pass arbitrary (non-unicode) binary
+data.  This change modifies how servant-client handles query parameters, so
+that application developers can use `ToHttpApiData` to marshal binary data into
+query parameters.
+}

changelog.d/1469

@@ -0,0 +1,11 @@
+synopsis: Derive HasClient good response status from Verb status
+prs: #1469
+description: {
+`HasClient` instances for the `Verb` datatype use `runRequest` in
+`clientWithRoute` definitions.
+This means that a request performed with `runClientM` will be successful if and
+only if the endpoint specify a response status code >=200 and <300.
+This change replaces `runRequest` with `runRequestAcceptStatus` in `Verb`
+instances for the `HasClient` class, deriving the good response status from
+the `Verb` status.
+}

changelog.d/1477

@@ -0,0 +1,9 @@
+synopsis: Enable FlexibleContexts in Servant.API.ContentTypes
+prs: #1477
+
+description: {
+
+Starting with GHC 9.2, UndecidableInstances no longer implies FlexibleContexts.
+Add this extension where it's needed to make compilation succeed.
+
+}

changelog.d/1529

@@ -0,0 +1,10 @@
+synopsis: Fix performRequest in servant-client-ghcjs
+prs: #1529
+
+description: {
+
+performRequest function in servant-client-ghcjs was not compatible with the
+latest RunClient typeclass. Added the acceptStatus parameter and fixed the
+functionality to match what servant-client provides.
+
+}

changelog.d/1556

@@ -0,0 +1,81 @@
+synopsis: Display capture hints in router layout
+prs: #1556
+
+description: {
+
+This PR enhances the `Servant.Server.layout` function, which produces a textual description of the routing layout of an API. More precisely, it changes `<capture>` blocks, so that they display the name and type of the variable being captured instead.
+
+Example:
+
+For the following API
+```haskell
+type API =
+       "a" :> "d" :> Get '[JSON] NoContent
+  :<|> "b" :> Capture "x" Int :> Get '[JSON] Bool
+  :<|> "a" :> "e" :> Get '[JSON] Int
+```
+
+we previously got the following output:
+
+```
+/
+├─ a/
+│  ├─ d/
+│  │  └─•
+│  └─ e/
+│     └─•
+└─ b/
+   └─ <capture>/
+      ├─•
+      ┆
+      └─•
+```
+
+now we get:
+
+```
+/
+├─ a/
+│  ├─ d/
+│  │  └─•
+│  └─ e/
+│     └─•
+└─ b/
+   └─ <x::Int>/
+      ├─•
+      ┆
+      └─•
+```
+
+This change is achieved by the introduction of a CaptureHint type, which is passed as an extra argument to the CaptureRouter and CaptureAllRouter constructors for the Router' type.
+CaptureHint values are then used in routerLayout, to display the name and type of captured values, instead of just `<capture>` previously.
+
+N.B.:
+Because the choice smart constructor for routers can aggregate Capture combinators with different capture hints, the Capture*Router constructors actually take a list of CaptureHint, instead of a single one.
+
+This PR also introduces Spec tests for the routerLayout function.
+
+Warning:
+This change is potentially breaking, because it adds the constraint `Typeable a` to all types that are to be captured. Because all types are typeable since GHC 7.10, this is not as bad as it sounds ; it only break expressions where `a` is quantified in an expression with `Capture a`.
+In those cases, the fix is easy: it suffices to add `Typeable a` to the left-hand side of the quantification constraint.
+
+For instance, the following code will no longer compile:
+```haskell
+type MyAPI a = Capture "foo" a :> Get '[JSON] ()
+
+myServer :: forall a. Server (MyAPI a)
+myServer = const $ return ()
+
+myApi :: forall a. Proxy (MyAPI a)
+myApi = Proxy
+
+app :: forall a. (FromHttpApiData a) => Application
+app = serve (myApi @a) (myServer @a)
+```
+
+Indeed, `app` should be replaced with:
+```haskell
+app :: forall a. (FromHttpApiData a, Typeable a) => Application
+app = serve (myApi @a) (myServer @a)
+```
+}

changelog.d/1569

@@ -0,0 +1,13 @@
+synopsis: Encode captures using toEncodedUrlPiece
+prs: #1569
+issues: #1511
+
+description: {
+The `servant-client` library now makes direct use of `toEncodedUrlPiece` from `ToHttpApiData`
+to encode captured values when building the request path. It gives user freedom to implement
+URL-encoding however they need.
+
+Previous behavior was to use `toUrlPiece` and URL-encode its output using `toEncodedUrlPiece`
+from the `Text` instance of `ToHttpApiData`. The issue with this approach is that
+`ToHttpApiData Text` is overly zealous and also encodes characters, such as `*`, which are perfectly valid in a URL.
+}

changelog.d/1573

@@ -0,0 +1,2 @@
+synopsis: Add API docs for ServerT
+prs: #1573

changelog.d/1580

@@ -0,0 +1,12 @@
+synopsis: Allow IO in validationKeys
+prs: #1580
+issues: #1579
+
+description: {
+
+Currently validationKeys are a fixed JWKSet. This does not work with OIDC
+providers such as AWS Cognito or Okta, which regularly fetching jwks_uri to
+discover new and expired keys.
+
+This change alters the type of validationKeys from JWKSet to IO JWKSet.
+}

changelog.d/1589

@@ -0,0 +1,2 @@
+synopsis: Only include question mark for nonempty query strings
+prs: 1589

changelog.d/1595

@@ -0,0 +1,2 @@
+synopsis: Run ClientEnv's makeClientRequest in IO.
+prs: #1595

changelog.d/1606

@@ -0,0 +1,10 @@
+synopsis: Handle Cookies correctly for RunStreamingClient
+prs: #1606
+issues: #1605
+
+description: {
+
+Makes performWithStreamingRequest take into consideration the
+CookieJar, which it previously didn't.
+
+}

changelog.d/1638

@@ -0,0 +1,2 @@
+synopsis: Add Functor instance to AuthHandler.
+prs: #1638

changelog.d/1649

@@ -0,0 +1,8 @@
+synopsis: Add HasStatus instance for Headers (that defers StatusOf to underlying value)
+prs: #1649
+
+description: {
+
+Adds a new HasStatus (Headers hs a) instance (StatusOf (Headers hs a) = StatusOf a)
+
+}

changelog.d/config

@@ -0,0 +1,2 @@
+organization: haskell-servant
+repository:   servant

changelog.d/servant-docs-curl

@@ -0,0 +1,16 @@
+synopsis: Add sample cURL requests to generated documentation
+prs: #1401
+
+description: {
+
+Add sample cURL requests to generated documentation.
+
+Those supplying changes to the Request `header` field manually using
+lenses will need to add a sample bytestring value.
+
+`headers <>~ ["unicorn"]`
+
+becomes
+
+`headers <>~ [("unicorn", "sample value")]`
+}

default.nix

@@ -0,0 +1,38 @@
+with (builtins.fromJSON (builtins.readFile ./nix/nixpkgs.json));
+{
+  pkgs ? import (builtins.fetchTarball {
+    url = "https://github.com/NixOS/nixpkgs/archive/${rev}.tar.gz";
+    inherit sha256;
+  }) {}
+, compiler ? "ghc883"
+}:
+let
+  overrides = self: super: {
+    servant              = self.callCabal2nix "servant"              ./servant              {};
+    servant-docs         = self.callCabal2nix "servant-docs"         ./servant-docs         {};
+    servant-pipes        = self.callCabal2nix "servant-pipes"        ./servant-pipes        {};
+    servant-server       = self.callCabal2nix "servant-server"       ./servant-server       {};
+    servant-client       = self.callCabal2nix "servant-client"       ./servant-client       {};
+    servant-foreign      = self.callCabal2nix "servant-foreign"      ./servant-foreign      {};
+    servant-conduit      = self.callCabal2nix "servant-conduit"      ./servant-conduit      {};
+    servant-machines     = self.callCabal2nix "servant-machines"     ./servant-machines     {};
+    servant-client-core  = self.callCabal2nix "servant-client-core"  ./servant-client-core  {};
+    servant-http-streams = self.callCabal2nix "servant-http-streams" ./servant-http-streams {};
+  };
+  hPkgs = pkgs.haskell.packages.${compiler}.override { inherit overrides; };
+in
+  with hPkgs;
+  {
+    inherit
+      servant
+      servant-client
+      servant-client-core
+      servant-conduit
+      servant-docs
+      servant-foreign
+      servant-http-streams
+      servant-machines
+      servant-pipes
+      servant-server;
+  }
+

doc/Makefile

@@ -1,216 +1,22 @@
-# Makefile for Sphinx documentation
+# Minimal makefile for Sphinx documentation
 #
 
 # You can set these variables from the command line.
 SPHINXOPTS    =
 SPHINXBUILD   = sphinx-build
-PAPER         =
+SPHINXPROJ    = Servant
+SOURCEDIR     = .
 BUILDDIR      = _build
 
-# User-friendly check for sphinx-build
-ifeq ($(shell which $(SPHINXBUILD) >/dev/null 2>&1; echo $$?), 1)
-$(error The '$(SPHINXBUILD)' command was not found. Make sure you have Sphinx installed, then set the SPHINXBUILD environment variable to point to the full path of the '$(SPHINXBUILD)' executable. Alternatively you can add the directory with the executable to your PATH. If you don't have Sphinx installed, grab it from http://sphinx-doc.org/)
-endif
-
-# Internal variables.
-PAPEROPT_a4     = -D latex_paper_size=a4
-PAPEROPT_letter = -D latex_paper_size=letter
-ALLSPHINXOPTS   = -d $(BUILDDIR)/doctrees $(PAPEROPT_$(PAPER)) $(SPHINXOPTS) .
-# the i18n builder cannot share the environment and doctrees with the others
-I18NSPHINXOPTS  = $(PAPEROPT_$(PAPER)) $(SPHINXOPTS) .
-
-.PHONY: help
+# Put it first so that "make" without argument is like "make help".
 help:
-	@echo "Please use \`make <target>' where <target> is one of"
-	@echo "  html       to make standalone HTML files"
-	@echo "  dirhtml    to make HTML files named index.html in directories"
-	@echo "  singlehtml to make a single large HTML file"
-	@echo "  pickle     to make pickle files"
-	@echo "  json       to make JSON files"
-	@echo "  htmlhelp   to make HTML files and a HTML help project"
-	@echo "  qthelp     to make HTML files and a qthelp project"
-	@echo "  applehelp  to make an Apple Help Book"
-	@echo "  devhelp    to make HTML files and a Devhelp project"
-	@echo "  epub       to make an epub"
-	@echo "  latex      to make LaTeX files, you can set PAPER=a4 or PAPER=letter"
-	@echo "  latexpdf   to make LaTeX files and run them through pdflatex"
-	@echo "  latexpdfja to make LaTeX files and run them through platex/dvipdfmx"
-	@echo "  text       to make text files"
-	@echo "  man        to make manual pages"
-	@echo "  texinfo    to make Texinfo files"
-	@echo "  info       to make Texinfo files and run them through makeinfo"
-	@echo "  gettext    to make PO message catalogs"
-	@echo "  changes    to make an overview of all changed/added/deprecated items"
-	@echo "  xml        to make Docutils-native XML files"
-	@echo "  pseudoxml  to make pseudoxml-XML files for display purposes"
-	@echo "  linkcheck  to check all external links for integrity"
-	@echo "  doctest    to run all doctests embedded in the documentation (if enabled)"
-	@echo "  coverage   to run coverage check of the documentation (if enabled)"
+	@if [ ! -d venv ]; then echo "WARNING: There is no venv directory, did you forget to 'virtualenv venv'. Check README.md."; fi
+	@if [ ! "z$$(which $(SPHINXBUILD))" = "z$$(pwd)/venv/bin/sphinx-build" ]; then echo "WARNING: Did you forgot to 'source venv/bin/activate'"; fi
+	@$(SPHINXBUILD) -M help "$(SOURCEDIR)" "$(BUILDDIR)" $(SPHINXOPTS) $(O)
 
-.PHONY: clean
-clean:
-	rm -rf $(BUILDDIR)/*
+.PHONY: help Makefile
 
-.PHONY: html
-html:
-	$(SPHINXBUILD) -b html $(ALLSPHINXOPTS) $(BUILDDIR)/html
-	@echo
-	@echo "Build finished. The HTML pages are in $(BUILDDIR)/html."
-
-.PHONY: dirhtml
-dirhtml:
-	$(SPHINXBUILD) -b dirhtml $(ALLSPHINXOPTS) $(BUILDDIR)/dirhtml
-	@echo
-	@echo "Build finished. The HTML pages are in $(BUILDDIR)/dirhtml."
-
-.PHONY: singlehtml
-singlehtml:
-	$(SPHINXBUILD) -b singlehtml $(ALLSPHINXOPTS) $(BUILDDIR)/singlehtml
-	@echo
-	@echo "Build finished. The HTML page is in $(BUILDDIR)/singlehtml."
-
-.PHONY: pickle
-pickle:
-	$(SPHINXBUILD) -b pickle $(ALLSPHINXOPTS) $(BUILDDIR)/pickle
-	@echo
-	@echo "Build finished; now you can process the pickle files."
-
-.PHONY: json
-json:
-	$(SPHINXBUILD) -b json $(ALLSPHINXOPTS) $(BUILDDIR)/json
-	@echo
-	@echo "Build finished; now you can process the JSON files."
-
-.PHONY: htmlhelp
-htmlhelp:
-	$(SPHINXBUILD) -b htmlhelp $(ALLSPHINXOPTS) $(BUILDDIR)/htmlhelp
-	@echo
-	@echo "Build finished; now you can run HTML Help Workshop with the" \
-	      ".hhp project file in $(BUILDDIR)/htmlhelp."
-
-.PHONY: qthelp
-qthelp:
-	$(SPHINXBUILD) -b qthelp $(ALLSPHINXOPTS) $(BUILDDIR)/qthelp
-	@echo
-	@echo "Build finished; now you can run "qcollectiongenerator" with the" \
-	      ".qhcp project file in $(BUILDDIR)/qthelp, like this:"
-	@echo "# qcollectiongenerator $(BUILDDIR)/qthelp/generics-eot.qhcp"
-	@echo "To view the help file:"
-	@echo "# assistant -collectionFile $(BUILDDIR)/qthelp/generics-eot.qhc"
-
-.PHONY: applehelp
-applehelp:
-	$(SPHINXBUILD) -b applehelp $(ALLSPHINXOPTS) $(BUILDDIR)/applehelp
-	@echo
-	@echo "Build finished. The help book is in $(BUILDDIR)/applehelp."
-	@echo "N.B. You won't be able to view it unless you put it in" \
-	      "~/Library/Documentation/Help or install it in your application" \
-	      "bundle."
-
-.PHONY: devhelp
-devhelp:
-	$(SPHINXBUILD) -b devhelp $(ALLSPHINXOPTS) $(BUILDDIR)/devhelp
-	@echo
-	@echo "Build finished."
-	@echo "To view the help file:"
-	@echo "# mkdir -p $$HOME/.local/share/devhelp/generics-eot"
-	@echo "# ln -s $(BUILDDIR)/devhelp $$HOME/.local/share/devhelp/generics-eot"
-	@echo "# devhelp"
-
-.PHONY: epub
-epub:
-	$(SPHINXBUILD) -b epub $(ALLSPHINXOPTS) $(BUILDDIR)/epub
-	@echo
-	@echo "Build finished. The epub file is in $(BUILDDIR)/epub."
-
-.PHONY: latex
-latex:
-	$(SPHINXBUILD) -b latex $(ALLSPHINXOPTS) $(BUILDDIR)/latex
-	@echo
-	@echo "Build finished; the LaTeX files are in $(BUILDDIR)/latex."
-	@echo "Run \`make' in that directory to run these through (pdf)latex" \
-	      "(use \`make latexpdf' here to do that automatically)."
-
-.PHONY: latexpdf
-latexpdf:
-	$(SPHINXBUILD) -b latex $(ALLSPHINXOPTS) $(BUILDDIR)/latex
-	@echo "Running LaTeX files through pdflatex..."
-	$(MAKE) -C $(BUILDDIR)/latex all-pdf
-	@echo "pdflatex finished; the PDF files are in $(BUILDDIR)/latex."
-
-.PHONY: latexpdfja
-latexpdfja:
-	$(SPHINXBUILD) -b latex $(ALLSPHINXOPTS) $(BUILDDIR)/latex
-	@echo "Running LaTeX files through platex and dvipdfmx..."
-	$(MAKE) -C $(BUILDDIR)/latex all-pdf-ja
-	@echo "pdflatex finished; the PDF files are in $(BUILDDIR)/latex."
-
-.PHONY: text
-text:
-	$(SPHINXBUILD) -b text $(ALLSPHINXOPTS) $(BUILDDIR)/text
-	@echo
-	@echo "Build finished. The text files are in $(BUILDDIR)/text."
-
-.PHONY: man
-man:
-	$(SPHINXBUILD) -b man $(ALLSPHINXOPTS) $(BUILDDIR)/man
-	@echo
-	@echo "Build finished. The manual pages are in $(BUILDDIR)/man."
-
-.PHONY: texinfo
-texinfo:
-	$(SPHINXBUILD) -b texinfo $(ALLSPHINXOPTS) $(BUILDDIR)/texinfo
-	@echo
-	@echo "Build finished. The Texinfo files are in $(BUILDDIR)/texinfo."
-	@echo "Run \`make' in that directory to run these through makeinfo" \
-	      "(use \`make info' here to do that automatically)."
-
-.PHONY: info
-info:
-	$(SPHINXBUILD) -b texinfo $(ALLSPHINXOPTS) $(BUILDDIR)/texinfo
-	@echo "Running Texinfo files through makeinfo..."
-	make -C $(BUILDDIR)/texinfo info
-	@echo "makeinfo finished; the Info files are in $(BUILDDIR)/texinfo."
-
-.PHONY: gettext
-gettext:
-	$(SPHINXBUILD) -b gettext $(I18NSPHINXOPTS) $(BUILDDIR)/locale
-	@echo
-	@echo "Build finished. The message catalogs are in $(BUILDDIR)/locale."
-
-.PHONY: changes
-changes:
-	$(SPHINXBUILD) -b changes $(ALLSPHINXOPTS) $(BUILDDIR)/changes
-	@echo
-	@echo "The overview file is in $(BUILDDIR)/changes."
-
-.PHONY: linkcheck
-linkcheck:
-	$(SPHINXBUILD) -b linkcheck $(ALLSPHINXOPTS) $(BUILDDIR)/linkcheck
-	@echo
-	@echo "Link check complete; look for any errors in the above output " \
-	      "or in $(BUILDDIR)/linkcheck/output.txt."
-
-.PHONY: doctest
-doctest:
-	$(SPHINXBUILD) -b doctest $(ALLSPHINXOPTS) $(BUILDDIR)/doctest
-	@echo "Testing of doctests in the sources finished, look at the " \
-	      "results in $(BUILDDIR)/doctest/output.txt."
-
-.PHONY: coverage
-coverage:
-	$(SPHINXBUILD) -b coverage $(ALLSPHINXOPTS) $(BUILDDIR)/coverage
-	@echo "Testing of coverage in the sources finished, look at the " \
-	      "results in $(BUILDDIR)/coverage/python.txt."
-
-.PHONY: xml
-xml:
-	$(SPHINXBUILD) -b xml $(ALLSPHINXOPTS) $(BUILDDIR)/xml
-	@echo
-	@echo "Build finished. The XML files are in $(BUILDDIR)/xml."
-
-.PHONY: pseudoxml
-pseudoxml:
-	$(SPHINXBUILD) -b pseudoxml $(ALLSPHINXOPTS) $(BUILDDIR)/pseudoxml
-	@echo
-	@echo "Build finished. The pseudo-XML files are in $(BUILDDIR)/pseudoxml."
+# Catch-all target: route all unknown targets to Sphinx using the new
+# "make mode" option.  $(O) is meant as a shortcut for $(SPHINXOPTS).
+%: Makefile
+	@$(SPHINXBUILD) -M $@ "$(SOURCEDIR)" "$(BUILDDIR)" $(SPHINXOPTS) $(O)

doc/README.md

@@ -0,0 +1,8 @@
+To build the docs locally:
+
+    $ virtualenv venv
+    $ . ./venv/bin/activate
+    $ pip install -r requirements.txt
+    $ make html
+
+Docs will be built in _build/html/index.html .

doc/building-the-docs

@@ -1,8 +0,0 @@
-To build the docs locally:
-
-$ virtualenv venv
-$ . ./venv/bin/activate
-$ pip install -r requirements.txt
-$ make html
-
-Docs will be built in _build/html/index.html .

doc/conf.py

@@ -1,7 +1,7 @@
 # -*- coding: utf-8 -*-
 #
-# servant documentation build configuration file, created by
-# sphinx-quickstart on Mon Nov 23 13:24:36 2015.
+# Servant documentation build configuration file, created by
+# sphinx-quickstart on Fri Jul  6 11:38:51 2018.
 #
 # This file is execfile()d with the current directory set to its
 # containing dir.
@@ -12,20 +12,21 @@
 # All configuration values have a default; values that are commented out
 # serve to show the default.
 
-import sys
-import os
-import shlex
-from recommonmark.parser import CommonMarkParser
-
 # If extensions (or modules to document with autodoc) are in another directory,
 # add these directories to sys.path here. If the directory is relative to the
 # documentation root, use os.path.abspath to make it absolute, like shown here.
-#sys.path.insert(0, os.path.abspath('.'))
+#
+# import os
+# import sys
+# sys.path.insert(0, os.path.abspath('.'))
+from recommonmark.parser import CommonMarkParser
+
 
 # -- General configuration ------------------------------------------------
 
 # If your documentation needs a minimal Sphinx version, state it here.
-#needs_sphinx = '1.0'
+#
+# needs_sphinx = '1.0'
 
 # Add any Sphinx extension module names here, as strings. They can be
 # extensions coming with Sphinx (named 'sphinx.ext.*') or your custom
@@ -37,17 +38,15 @@ templates_path = ['_templates']
 
 # The suffix(es) of source filenames.
 # You can specify multiple suffix as a list of string:
-source_suffix = ['.md', '.rst', '.lhs']
-
-# The encoding of source files.
-#source_encoding = 'utf-8-sig'
+#
+source_suffix = ['.rst', '.md', '.lhs']
 
 # The master toctree document.
 master_doc = 'index'
 
 # General information about the project.
-project = u'servant'
-copyright = u'2016, Servant Contributors'
+project = u'Servant'
+copyright = u'2022, Servant Contributors'
 author = u'Servant Contributors'
 
 # The version info for the project you're documenting, acts as replacement for
@@ -55,9 +54,9 @@ author = u'Servant Contributors'
 # built documents.
 #
 # The short X.Y version.
-# version = 'latest'
+# version = u'0.14'
 # The full version, including alpha/beta/rc tags.
-# release = 'latest'
+# release = u'0.14'
 
 # The language for content autogenerated by Sphinx. Refer to documentation
 # for a list of supported languages.
@@ -66,45 +65,14 @@ author = u'Servant Contributors'
 # Usually you set "language" from the command line for these cases.
 language = None
 
-# There are two options for replacing |today|: either, you set today to some
-# non-false value, then it is used:
-#today = ''
-# Else, today_fmt is used as the format for a strftime call.
-#today_fmt = '%B %d, %Y'
-
 # List of patterns, relative to source directory, that match files and
 # directories to ignore when looking for source files.
-exclude_patterns = ['_build', 'venv']
-
-# The reST default role (used for this markup: `text`) to use for all
-# documents.
-#default_role = None
-
-# If true, '()' will be appended to :func: etc. cross-reference text.
-#add_function_parentheses = True
-
-# If true, the current module name will be prepended to all description
-# unit titles (such as .. function::).
-#add_module_names = True
-
-# If true, sectionauthor and moduleauthor directives will be shown in the
-# output. They are ignored by default.
-#show_authors = False
+# This patterns also effect to html_static_path and html_extra_path
+exclude_patterns = ['_build', 'Thumbs.db', '.DS_Store']
 
 # The name of the Pygments (syntax highlighting) style to use.
 pygments_style = 'sphinx'
 
-def setup(app):
-    from sphinx.highlighting import lexers
-    from pygments.lexers import HaskellLexer
-    lexers['haskell ignore'] = HaskellLexer(stripnl=False)
-
-# A list of ignored prefixes for module index sorting.
-#modindex_common_prefix = []
-
-# If true, keep warnings as "system message" paragraphs in the built documents.
-#keep_warnings = False
-
 # If true, `todo` and `todoList` produce output, else they produce nothing.
 todo_include_todos = False
 
@@ -113,157 +81,77 @@ todo_include_todos = False
 
 # The theme to use for HTML and HTML Help pages.  See the documentation for
 # a list of builtin themes.
+#
 html_theme = 'sphinx_rtd_theme'
 
 # Theme options are theme-specific and customize the look and feel of a theme
 # further.  For a list of options available for each theme, see the
 # documentation.
-#html_theme_options = {}
-
-# Add any paths that contain custom themes here, relative to this directory.
-#html_theme_path = []
-
-# The name for this set of Sphinx documents.  If None, it defaults to
-# "<project> v<release> documentation".
-#html_title = None
-
-# A shorter title for the navigation bar.  Default is the same as html_title.
-#html_short_title = None
-
-# The name of an image file (relative to this directory) to place at the top
-# of the sidebar.
-#html_logo = None
-
-# The name of an image file (within the static path) to use as favicon of the
-# docs.  This file should be a Windows icon file (.ico) being 16x16 or 32x32
-# pixels large.
-#html_favicon = None
+#
+# html_theme_options = {}
 
 # Add any paths that contain custom static files (such as style sheets) here,
 # relative to this directory. They are copied after the builtin static files,
 # so a file named "default.css" will overwrite the builtin "default.css".
 html_static_path = ['_static']
 
-# Add any extra paths that contain custom files (such as robots.txt or
-# .htaccess) here, relative to this directory. These files are copied
-# directly to the root of the documentation.
-#html_extra_path = []
+# Custom sidebar templates, must be a dictionary that maps document names
+# to template names.
+#
+# This is required for the alabaster theme
+# refs: http://alabaster.readthedocs.io/en/latest/installation.html#sidebars
+html_sidebars = {
+    '**': [
+        'relations.html',  # needs 'show_related': True theme option to display
+        'searchbox.html',
+    ]
+}
 
-# If not '', a 'Last updated on:' timestamp is inserted at every page bottom,
-# using the given strftime format.
-#html_last_updated_fmt = '%b %d, %Y'
 
-# If true, SmartyPants will be used to convert quotes and dashes to
-# typographically correct entities.
-#html_use_smartypants = True
-
-# Custom sidebar templates, maps document names to template names.
-#html_sidebars = {}
-
-# Additional templates that should be rendered to pages, maps page names to
-# template names.
-#html_additional_pages = {}
-
-# If false, no module index is generated.
-#html_domain_indices = True
-
-# If false, no index is generated.
-#html_use_index = True
-
-# If true, the index is split into individual pages for each letter.
-#html_split_index = False
-
-# If true, links to the reST sources are added to the pages.
-#html_show_sourcelink = True
-
-# If true, "Created using Sphinx" is shown in the HTML footer. Default is True.
-#html_show_sphinx = True
-
-# If true, "(C) Copyright ..." is shown in the HTML footer. Default is True.
-#html_show_copyright = True
-
-# If true, an OpenSearch description file will be output, and all pages will
-# contain a <link> tag referring to it.  The value of this option must be the
-# base URL from which the finished HTML is served.
-#html_use_opensearch = ''
-
-# This is the file name suffix for HTML files (e.g. ".xhtml").
-#html_file_suffix = None
-
-# Language to be used for generating the HTML full-text search index.
-# Sphinx supports the following languages:
-#   'da', 'de', 'en', 'es', 'fi', 'fr', 'hu', 'it', 'ja'
-#   'nl', 'no', 'pt', 'ro', 'ru', 'sv', 'tr'
-#html_search_language = 'en'
-
-# A dictionary with options for the search language support, empty by default.
-# Now only 'ja' uses this config value
-#html_search_options = {'type': 'default'}
-
-# The name of a javascript file (relative to the configuration directory) that
-# implements a search results scorer. If empty, the default will be used.
-#html_search_scorer = 'scorer.js'
+# -- Options for HTMLHelp output ------------------------------------------
 
 # Output file base name for HTML help builder.
-htmlhelp_basename = 'servantdoc'
+htmlhelp_basename = 'Servantdoc'
+
 
 # -- Options for LaTeX output ---------------------------------------------
 
 latex_elements = {
-# The paper size ('letterpaper' or 'a4paper').
-#'papersize': 'letterpaper',
+    # The paper size ('letterpaper' or 'a4paper').
+    #
+    # 'papersize': 'letterpaper',
 
-# The font size ('10pt', '11pt' or '12pt').
-#'pointsize': '10pt',
+    # The font size ('10pt', '11pt' or '12pt').
+    #
+    # 'pointsize': '10pt',
 
-# Additional stuff for the LaTeX preamble.
-#'preamble': '',
+    # Additional stuff for the LaTeX preamble.
+    #
+    # 'preamble': '',
 
-# Latex figure (float) alignment
-#'figure_align': 'htbp',
+    # Latex figure (float) alignment
+    #
+    # 'figure_align': 'htbp',
 }
 
 # Grouping the document tree into LaTeX files. List of tuples
 # (source start file, target name, title,
 #  author, documentclass [howto, manual, or own class]).
 latex_documents = [
-    (master_doc, 'servant.tex', u'servant Documentation',
+    (master_doc, 'Servant.tex', u'Servant Documentation',
      u'Servant Contributors', 'manual'),
 ]
 
-# The name of an image file (relative to this directory) to place at the top of
-# the title page.
-#latex_logo = None
-
-# For "manual" documents, if this is true, then toplevel headings are parts,
-# not chapters.
-#latex_use_parts = False
-
-# If true, show page references after internal links.
-#latex_show_pagerefs = False
-
-# If true, show URL addresses after external links.
-#latex_show_urls = False
-
-# Documents to append as an appendix to all manuals.
-#latex_appendices = []
-
-# If false, no module index is generated.
-#latex_domain_indices = True
-
 
 # -- Options for manual page output ---------------------------------------
 
 # One entry per manual page. List of tuples
 # (source start file, name, description, authors, manual section).
 man_pages = [
-    (master_doc, 'servant', u'servant Documentation',
+    (master_doc, 'servant', u'Servant Documentation',
      [author], 1)
 ]
 
-# If true, show URL addresses after external links.
-#man_show_urls = False
-
 
 # -- Options for Texinfo output -------------------------------------------
 
@@ -271,24 +159,13 @@ man_pages = [
 # (source start file, target name, title, author,
 #  dir menu entry, description, category)
 texinfo_documents = [
-  (master_doc, 'servant', u'servant Documentation',
-   author, 'servant', 'One line description of project.',
-   'Miscellaneous'),
+    (master_doc, 'Servant', u'Servant Documentation',
+     author, 'Servant', 'One line description of project.',
+     'Miscellaneous'),
 ]
 
-# Documents to append as an appendix to all manuals.
-#texinfo_appendices = []
-
-# If false, no module index is generated.
-#texinfo_domain_indices = True
-
-# How to display URL addresses: 'footnote', 'no', or 'inline'.
-#texinfo_show_urls = 'footnote'
-
-# If true, do not generate a @detailmenu in the "Top" node's menu.
-#texinfo_no_detailmenu = False
+# -- Markdown -------------------------------------------------------------
 
 source_parsers = {
-    '.md': CommonMarkParser,
     '.lhs': CommonMarkParser,
 }

doc/cookbook/basic-auth/basic-auth.cabal

@@ -1,14 +1,14 @@
+cabal-version:       2.2
 name:                cookbook-basic-auth
 version:             0.1
 synopsis:            Basic Authentication cookbook example
-homepage:            http://haskell-servant.readthedocs.org/
-license:             BSD3
+homepage:            http://docs.servant.dev/
+license:             BSD-3-Clause
 license-file:        ../../../servant/LICENSE
 author:              Servant Contributors
 maintainer:          haskell-servant-maintainers@googlegroups.com
 build-type:          Simple
-cabal-version:       >=1.10
-tested-with:         GHC==7.8.4, GHC==7.10.3, GHC==8.0.2, GHC==8.2.2
+tested-with:         GHC==8.6.5, GHC==8.8.3, GHC ==8.10.7
 
 executable cookbook-basic-auth
   main-is:             BasicAuth.lhs

doc/cookbook/basic-streaming/Streaming.lhs

@@ -0,0 +1,129 @@
+# Streaming out-of-the-box
+
+In other words, without streaming libraries.
+
+## Introduction
+
+- Servant supports streaming
+- Some basic usage doesn't require usage of streaming libraries,
+  like `conduit`, `pipes`, `machines` or `streaming`.
+  We have bindings for them though.
+- Similar example is bundled with each of our streaming library interop packages (see 
+[servant-pipes](https://github.com/haskell-servant/servant/blob/master/servant-pipes/example/Main.hs),
+[servant-conduit](https://github.com/haskell-servant/servant/blob/master/servant-conduit/example/Main.hs) and
+[servant-machines](https://github.com/haskell-servant/servant/blob/master/servant-machines/example/Main.hs))
+- `SourceT` doesn't have *Prelude* with handy combinators, so we have to write
+  things ourselves. (Note to self: `mapM` and `foldM` would be handy to have). 
+
+## Code
+
+```haskell
+{-# LANGUAGE BangPatterns  #-}
+{-# LANGUAGE DataKinds     #-}
+{-# LANGUAGE DeriveGeneric #-}
+{-# LANGUAGE TypeOperators #-}
+module Main (main) where
+
+import           Control.Concurrent
+                 (threadDelay)
+import           Control.Monad.IO.Class
+                 (MonadIO (..))
+import qualified Data.ByteString              as BS
+import           Data.Maybe
+                 (fromMaybe)
+import           Network.HTTP.Client
+                 (defaultManagerSettings, newManager)
+import           Network.Wai
+                 (Application)
+import           System.Environment
+                 (getArgs, lookupEnv)
+import           Text.Read
+                 (readMaybe)
+
+import           Servant
+import           Servant.Client.Streaming
+import qualified Servant.Types.SourceT as S
+
+import qualified Network.Wai.Handler.Warp     as Warp
+
+type FastAPI = "get" :> Capture "num" Int :> StreamGet NewlineFraming JSON (SourceIO Int)
+
+type API = FastAPI
+    :<|> "slow" :> Capture "num" Int :> StreamGet NewlineFraming JSON (SourceIO Int)
+    -- monad can be ResourceT IO too.
+    :<|> "readme" :> StreamGet NoFraming OctetStream (SourceIO BS.ByteString)
+    -- we can have streaming request body
+    :<|> "proxy"
+        :> StreamBody NoFraming OctetStream (SourceIO BS.ByteString)
+        :> StreamPost NoFraming OctetStream (SourceIO BS.ByteString)
+
+api :: Proxy API
+api = Proxy
+
+server :: Server API
+server = fast :<|> slow :<|> readme :<|> proxy where
+    fast n = liftIO $ do
+        putStrLn $ "/get/" ++ show n
+        return $ fastSource n
+
+    slow n = liftIO $ do
+        putStrLn $ "/slow/" ++ show n
+        return $ slowSource n
+
+    readme = liftIO $ do
+        putStrLn "/proxy"
+        return (S.readFile "README.md")
+
+    proxy c = liftIO $ do
+        putStrLn "/proxy"
+        return c
+
+    -- for some reason unfold leaks?
+    fastSource = S.fromStepT . mk where
+        mk m
+            | m < 0     = S.Stop
+            | otherwise = S.Yield m (mk (m - 1))
+
+    slowSource m = S.mapStepT delay (fastSource m) where
+        delay S.Stop        = S.Stop
+        delay (S.Error err) = S.Error err
+        delay (S.Skip s)    = S.Skip (delay s)
+        delay (S.Effect ms) = S.Effect (fmap delay ms)
+        delay (S.Yield x s) = S.Effect $
+            S.Yield x (delay s) <$ threadDelay 1000000
+
+app :: Application
+app = serve api server
+
+cli :: Client ClientM FastAPI
+cli :<|> _ :<|> _ :<|> _ = client api
+
+main :: IO ()
+main = do
+    args <- getArgs
+    case args of
+        ("server":_) -> do
+            putStrLn "Starting cookbook-basic-streaming at http://localhost:8000"
+            port <- fromMaybe 8000 . (>>= readMaybe) <$> lookupEnv "PORT"
+            Warp.run port app
+        ("client":ns:_) -> do
+            n <- maybe (fail $ "not a number: " ++ ns) pure $ readMaybe ns
+            mgr <- newManager defaultManagerSettings
+            burl <- parseBaseUrl "http://localhost:8000/"
+            withClientM (cli n) (mkClientEnv mgr burl) $ \me -> case me of
+                Left err  -> print err
+                Right src -> do
+                    x <- S.unSourceT src (go (0 :: Int))
+                    print x
+                  where
+                    go !acc S.Stop        = return acc
+                    go !acc (S.Error err) = print err >> return acc
+                    go !acc (S.Skip s)    = go acc s
+                    go !acc (S.Effect ms) = ms >>= go acc
+                    go !acc (S.Yield _ s) = go (acc + 1) s
+        _ -> do
+            putStrLn "Try:"
+            putStrLn "cabal new-run cookbook-basic-streaming server"
+            putStrLn "cabal new-run cookbook-basic-streaming client 10"
+            putStrLn "time curl -H 'Accept: application/json' localhost:8000/slow/5"
+```

doc/cookbook/basic-streaming/basic-streaming.cabal

@@ -0,0 +1,28 @@
+cabal-version:       2.2
+name:                cookbook-basic-streaming
+version:             2.1
+synopsis:            Streaming in servant without streaming libs
+homepage:            http://docs.servant.dev/
+license:             BSD-3-Clause
+license-file:        ../../../servant/LICENSE
+author:              Servant Contributors
+maintainer:          haskell-servant-maintainers@googlegroups.com
+build-type:          Simple
+tested-with:         GHC==8.6.5, GHC==8.8.3, GHC ==8.10.7
+
+executable cookbook-basic-streaming
+  main-is:             Streaming.lhs
+  build-tool-depends:  markdown-unlit:markdown-unlit
+  default-language:    Haskell2010
+  ghc-options:         -Wall -pgmL markdown-unlit -threaded -rtsopts
+
+  hs-source-dirs: .
+  build-depends:       base >= 4.8 && <5
+                     , aeson
+                     , bytestring
+                     , servant
+                     , servant-server
+                     , servant-client
+                     , http-client
+                     , wai
+                     , warp

doc/cookbook/cabal.project

@@ -1,12 +1,19 @@
 packages:
   basic-auth/
+  curl-mock/
+  db-mysql-basics/
   db-sqlite-simple/
   db-postgres-pool/
   using-custom-monad/
   jwt-and-basic-auth/
+  hoist-server-with-context/
   file-upload/
   structuring-apis/
   https/
+  pagination/
+  sentry/
+  testing/
+  open-id-connect/
   ../../servant
   ../../servant-server
   ../../servant-client-core

doc/cookbook/curl-mock/CurlMock.lhs

@@ -0,0 +1,205 @@
+# Generating mock curl calls
+
+In this example we will generate curl requests with mock post data from a servant API.
+This may be useful for testing and development purposes.
+Especially post requests with a request body are tedious to send manually.
+
+Also, we will learn how to use the servant-foreign library to generate stuff from servant APIs.
+
+
+Language extensions and imports:
+``` haskell
+{-# LANGUAGE DataKinds                  #-}
+{-# LANGUAGE DeriveGeneric              #-}
+{-# LANGUAGE FlexibleContexts           #-}
+{-# LANGUAGE FlexibleInstances          #-}
+{-# LANGUAGE GeneralizedNewtypeDeriving #-}
+{-# LANGUAGE MultiParamTypeClasses      #-}
+{-# LANGUAGE OverloadedStrings          #-}
+{-# LANGUAGE RankNTypes                 #-}
+{-# LANGUAGE RecordWildCards            #-}
+{-# LANGUAGE ScopedTypeVariables        #-}
+{-# LANGUAGE TypeOperators              #-}
+
+import           Control.Lens                      ((^.))
+import           Data.Aeson
+import           Data.Aeson.Text
+import           Data.Proxy                        (Proxy (Proxy))
+import           Data.Text                         (Text)
+import           Data.Text.Encoding                (decodeUtf8)
+import qualified Data.Text.IO                      as T.IO
+import qualified Data.Text.Lazy                    as LazyT
+import           GHC.Generics
+import           Servant                           ((:<|>), (:>), Get, JSON,
+                                                    Post, ReqBody)
+import           Servant.Foreign                   (Foreign, GenerateList,
+                                                    HasForeign, HasForeignType, Req,
+                                                    Segment, SegmentType (Cap, Static),
+                                                    argName, listFromAPI, path,
+                                                    reqBody, reqMethod, reqUrl, typeFor,
+                                                    unPathSegment, unSegment,)
+import           Test.QuickCheck.Arbitrary
+import           Test.QuickCheck.Arbitrary.Generic
+import           Test.QuickCheck.Gen               (generate)
+import qualified Data.Text                         as T
+
+```
+
+
+Let's define our API:
+
+``` haskell
+type UserAPI = "users" :> Get '[JSON] [User]
+          :<|> "new" :> "user" :> ReqBody '[JSON] User :> Post '[JSON] ()
+
+data User = User
+  { name :: String
+  , age :: Int
+  , email :: String
+  } deriving (Eq, Show, Generic)
+
+instance Arbitrary User where
+  arbitrary = genericArbitrary
+  shrink = genericShrink
+instance ToJSON User
+instance FromJSON User
+```
+
+Notice the `Arbitrary User` instance which we will later need to create mock data.
+
+Also, the obligatory servant boilerplate:
+
+``` haskell
+api :: Proxy UserAPI
+api = Proxy
+```
+
+
+## servant-foreign and the HasForeignType Class
+
+Servant-foreign allows us to look into the API we designed.
+The entry point is `listFromAPI` which takes three types and returns a list of endpoints:
+
+``` haskell ignore
+listFromAPI :: (HasForeign lang ftype api, GenerateList ftype (Foreign ftype api)) => Proxy lang -> Proxy ftype -> Proxy api -> [Req ftype]
+```
+
+This looks a bit confusing...
+[Here](https://hackage.haskell.org/package/servant-foreign/docs/Servant-Foreign.html#t:HasForeign) is the documentation for the `HasForeign` typeclass.
+We will not go into details here, but this allows us to create a value of type `ftype` for any type `a` in our API.
+
+In our case we want to create a mock of every type `a`.
+
+We create a new datatype that holds our mocked value. Well, not the mocked value itself. To mock it we need IO (random). So the promise of a mocked value after some IO is performed:
+
+``` haskell
+data NoLang
+
+data Mocked = Mocked (IO Text)
+```
+
+Now, we create an instance of `HasForeignType` for `NoLang` and `Mocked` for every `a` that implements `ToJSON` and `Arbitrary`:
+``` haskell
+instance (ToJSON a, Arbitrary a) => HasForeignType NoLang Mocked a where
+  typeFor _ _ _ =
+    Mocked (genText (Proxy :: Proxy a))
+```
+
+What does `genText` do? It generates an arbitrary value of type `a` and encodes it as text. (And does some lazy to non-lazy text transformation we do not care about):
+
+``` haskell
+genText :: (ToJSON a, Arbitrary a) => Proxy a -> IO Text
+genText p =
+  fmap (\v -> LazyT.toStrict $ encodeToLazyText v) (genArb p)
+
+genArb :: Arbitrary a => Proxy a -> IO a
+genArb _ =
+  generate arbitrary
+```
+
+### Generating curl calls for every endpoint
+
+Everything is prepared now and we can start generating some curl calls.
+
+``` haskell
+generateCurl :: (GenerateList Mocked (Foreign Mocked api), HasForeign NoLang Mocked api)
+  => Proxy api
+  -> Text
+  -> IO Text
+generateCurl p host =
+  fmap T.unlines body
+  where
+  body = mapM (generateEndpoint host)
+    $ listFromAPI (Proxy :: Proxy NoLang) (Proxy :: Proxy Mocked) p
+```
+
+First, `listFromAPI` gives us a list of `Req Mocked`. Each `Req` describes one endpoint from the API type.
+We generate a curl call for each of them using the following helper.
+
+``` haskell
+generateEndpoint :: Text -> Req Mocked -> IO Text
+generateEndpoint host req =
+  case maybeBody of
+    Just body ->
+      body >>= \b -> return $ T.intercalate " " [ "curl", "-X", method, "-d", "'" <> b <> "'"
+                                                , "-H 'Content-Type: application/json'", host <> "/" <> url ]
+    Nothing ->
+      return $ T.intercalate " " [ "curl", "-X", method, host <> "/" <> url ]
+  where
+    method = decodeUtf8 $ req ^. reqMethod
+
+    url = T.intercalate "/" $ map segment (req ^. reqUrl . path)
+
+    maybeBody = fmap (\(Mocked io) -> io) (req ^. reqBody)
+
+```
+`servant-foreign` offers a multitude of lenses to be used with `Req`-values.
+
+`reqMethod` gives us a straigthforward `Network.HTTP.Types.Method`, `reqUrl` the url part and so on.
+Just take a look at [the docs](https://hackage.haskell.org/package/servant-foreign/docs/Servant-Foreign.html).
+
+But how do we get our mocked json string? This seems to be a bit to short to be true:
+
+``` haskell ignore
+maybeBody = fmap (\(Mocked io) -> io) (req ^. reqBody)
+```
+
+But it is that simple!
+The docs say `reqBody` gives us a `Maybe f`. What is `f`, you ask? As defined in `generateCurl`, `f` is `Mocked` and contains a `IO Text`. How is this `Mocked` value created? The `HasForeignType::typeFor` does it!
+
+Of course only if the endpoint has a request body.
+
+
+Some (incomplete) code for url segments:
+``` haskell
+segment :: Segment Mocked -> Text
+segment seg =
+  case unSegment seg of
+    Static p ->
+      unPathSegment p
+
+    Cap arg ->
+      -- Left as exercise for the reader: Mock args in the url
+      unPathSegment $ arg ^. argName
+```
+
+And now, lets hook it all up in our main function:
+
+``` haskell
+main :: IO ()
+main =
+  generateCurl api "localhost:8081" >>= T.IO.putStrLn
+```
+
+Done:
+
+``` curl
+curl -X GET localhost:8081/users
+curl -X POST -d '{"email":"wV򝣀_b򆎘:z񁊞򲙲!(3DM V","age":10,"name":"=|W"}' -H 'Content-Type: application/json' localhost:8081/new/user
+
+```
+
+This is of course no complete curl call mock generator, many things including path arguments are missing.
+But it correctly generates mock calls for simple POST requests.
+
+Also, we now know how to use `HasForeignType` and `listFromAPI` to generate anything we want.

doc/cookbook/curl-mock/curl-mock.cabal

@@ -0,0 +1,29 @@
+cabal-version:       2.2
+name:                cookbook-curl-mock
+version:             0.1
+synopsis:            Generate curl mock requests cookbook example
+homepage:            http://docs.servant.dev
+license:             BSD-3-Clause
+license-file:        ../../../servant/LICENSE
+author:              Servant Contributors
+maintainer:          haskell-servant-maintainers@googlegroups.com
+build-type:          Simple
+tested-with:         GHC==8.6.5, GHC==8.8.3, GHC ==8.10.7
+
+executable cookbock-curl-mock
+  if impl(ghc >= 9.2)
+    -- generic-arbitrary is incompatible
+    buildable: False
+  main-is:             CurlMock.lhs
+  build-depends:       base == 4.*
+                     , aeson
+                     , lens
+                     , text
+                     , servant
+                     , servant-server
+                     , servant-foreign
+                     , QuickCheck
+                     , generic-arbitrary
+  default-language:    Haskell2010
+  ghc-options:         -Wall -pgmL markdown-unlit
+  build-tool-depends: markdown-unlit:markdown-unlit

doc/cookbook/custom-errors/CustomErrors.lhs

@@ -0,0 +1,189 @@
+# Customizing errors from Servant
+
+Servant handles a lot of parsing and validation of the input request. When it can't parse something: query
+parameters, URL parts or request body, it will return appropriate HTTP codes like 400 Bad Request.
+
+These responses will contain the error message in their body without any formatting. However, it is often
+desirable to be able to provide custom formatting for these error messages, for example, to wrap them in JSON.
+
+Recently Servant got a way to add such formatting. This Cookbook chapter demonstrates how to use it.
+
+Extensions and imports:
+```haskell
+{-# LANGUAGE DataKinds         #-}
+{-# LANGUAGE DeriveGeneric     #-}
+{-# LANGUAGE OverloadedStrings #-}
+{-# LANGUAGE PolyKinds         #-}
+{-# LANGUAGE TypeFamilies      #-}
+{-# LANGUAGE TypeOperators     #-}
+
+import           Data.Aeson
+import           Data.Proxy
+import           Data.Text
+import           GHC.Generics
+import           Network.Wai
+import           Network.Wai.Handler.Warp
+
+import           Servant
+
+import           Data.String.Conversions
+                 (cs)
+import           Servant.API.ContentTypes
+```
+
+The API (from `greet.hs` example in Servant sources):
+
+```haskell
+-- | A greet message data type
+newtype Greet = Greet { _msg :: Text }
+  deriving (Generic, Show)
+
+instance FromJSON Greet
+instance ToJSON Greet
+
+-- API specification
+type TestApi =
+       -- GET /hello/:name?capital={true, false}  returns a Greet as JSON
+       "hello" :> Capture "name" Text :> QueryParam "capital" Bool :> Get '[JSON] Greet
+
+       -- POST /greet with a Greet as JSON in the request body,
+       --             returns a Greet as JSON
+  :<|> "greet" :> ReqBody '[JSON] Greet :> Post '[JSON] Greet
+
+       -- DELETE /greet/:greetid
+  :<|> "greet" :> Capture "greetid" Text :> Delete '[JSON] NoContent
+
+testApi :: Proxy TestApi
+testApi = Proxy
+
+-- Server-side handlers.
+--
+-- There's one handler per endpoint, which, just like in the type
+-- that represents the API, are glued together using :<|>.
+--
+-- Each handler runs in the 'Handler' monad.
+server :: Server TestApi
+server = helloH :<|> postGreetH :<|> deleteGreetH
+
+  where helloH name Nothing = helloH name (Just False)
+        helloH name (Just False) = return . Greet $ "Hello, " <> name
+        helloH name (Just True) = return . Greet . toUpper $ "Hello, " <> name
+
+        postGreetH greet = return greet
+
+        deleteGreetH _ = return NoContent
+```
+
+## Error formatters
+
+`servant-server` provides an `ErrorFormatter` type to specify how the error message will be
+formatted. A formatter is just a function accepting three parameters:
+
+- `TypeRep` from `Data.Typeable`: this is a runtime representation of the type of the combinator
+  (like `Capture` or `ReqBody`) that generated the error. It can be used to display its name (with
+  `show`) or even dynamically dispatch on the combinator type. See the docs for `Data.Typeable` and
+  `Type.Reflection` modules.
+- `Request`: full information for the request that led to the error.
+- `String`: specific error message from the combinator.
+
+The formatter is expected to produce a `ServerError` which will be returned from the handler.
+
+Additionally, there is `NotFoundErrorFormatter`, which accepts only `Request` and can customize the
+error in case when no route can be matched (HTTP 404).
+
+Let's make two formatters. First one will wrap our error in a JSON:
+
+```json
+{
+  "error": "ERROR MESSAGE",
+  "combinator": "NAME OF THE COMBINATOR"
+}
+```
+
+Additionally, this formatter will examine the `Accept` header of the request and generate JSON
+message only if client can accept it.
+
+```haskell
+customFormatter :: ErrorFormatter
+customFormatter tr req err =
+  let
+    -- aeson Value which will be sent to the client
+    value = object ["combinator" .= show tr, "error" .= err]
+    -- Accept header of the request
+    accH = getAcceptHeader req
+  in
+  -- handleAcceptH is Servant's function that checks whether the client can accept a
+  -- certain message type.
+  -- In this case we call it with "Proxy '[JSON]" argument, meaning that we want to return a JSON.
+  case handleAcceptH (Proxy :: Proxy '[JSON]) accH value of
+    -- If client can't handle JSON, we just return the body the old way
+    Nothing -> err400 { errBody = cs err }
+    -- Otherwise, we return the JSON formatted body and set the "Content-Type" header.
+    Just (ctypeH, body) -> err400
+      { errBody = body
+      , errHeaders = [("Content-Type", cs ctypeH)]
+      }
+
+notFoundFormatter :: NotFoundErrorFormatter
+notFoundFormatter req =
+  err404 { errBody = cs $ "Not found path: " <> rawPathInfo req }
+```
+
+If you don't need to react to the `Accept` header, you can just unconditionally return the JSON like
+this (with `encode` from `Data.Aeson`):
+
+```
+err400
+  { errBody = encode body
+  , errHeaders = [("Content-Type", "application/json")]
+  }
+```
+
+## Passing formatters to Servant
+
+Servant uses the Context to configure formatters. You only need to add a value of type
+`ErrorFormatters` to your context. This is a record with the following fields:
+
+- `bodyParserErrorFormatter :: ErrorFormatter`
+- `urlParseErrorFormatter :: ErrorFormatter`
+- `headerParseErrorFormatter :: ErrorFormatter`
+- `notFoundErrorFormatter :: NotFoundErrorFormatter`
+
+Default formatters are exported as `defaultErrorFormatters`, so you can use record update syntax to
+set the only ones you need:
+
+```haskell
+customFormatters :: ErrorFormatters
+customFormatters = defaultErrorFormatters
+  { bodyParserErrorFormatter = customFormatter
+  , notFoundErrorFormatter = notFoundFormatter
+  }
+```
+
+And at last, use `serveWithContext` to run your server as usual:
+
+```haskell
+app :: Application
+app = serveWithContext testApi (customFormatters :. EmptyContext) server
+
+main :: IO ()
+main = run 8000 app
+```
+
+Now if we try to request something with a wrong body, we will get a nice error:
+
+```
+$ http -j POST localhost:8000/greet 'foo=bar'
+HTTP/1.1 400 Bad Request
+Content-Type: application/json;charset=utf-8
+Date: Fri, 17 Jul 2020 13:34:18 GMT
+Server: Warp/3.3.12
+Transfer-Encoding: chunked
+
+{
+    "combinator": "ReqBody'",
+    "error": "Error in $: parsing Main.Greet(Greet) failed, key \"_msg\" not found"
+}
+```
+
+Notice the `Content-Type` header set by our combinator.

doc/cookbook/custom-errors/custom-errors.cabal

@@ -0,0 +1,25 @@
+cabal-version:       2.2
+name:                cookbook-custom-errors
+version:             0.1
+synopsis:            Return custom error messages from combinators
+homepage:            http://docs.servant.dev
+license:             BSD-3-Clause
+license-file:        ../../../servant/LICENSE
+author:              Servant Contributors
+maintainer:          haskell-servant-maintainers@googlegroups.com
+build-type:          Simple
+tested-with:         GHC==8.6.5, GHC==8.8.3, GHC ==8.10.7
+
+executable cookbook-custom-errors
+  main-is:             CustomErrors.lhs
+  build-depends:       base == 4.*
+                     , aeson
+                     , servant
+                     , servant-server
+                     , string-conversions
+                     , text
+                     , wai
+                     , warp
+  default-language:    Haskell2010
+  ghc-options:         -Wall -pgmL markdown-unlit
+  build-tool-depends: markdown-unlit:markdown-unlit

doc/cookbook/db-mysql-basics/MysqlBasics.lhs

@@ -0,0 +1,236 @@
+# Overview
+
+This doc will walk through a single-module implementation of a servant API connecting to a MySQL database. It will also include some basic CRUD operations.
+
+Once you can wrap your head around this implementation, understanding more complex features like resource pools would be beneficial next steps.
+
+The only *prerequisite* is that you have a MySQL database open on port 3306 of your machine. Docker is an easy way to manage this.
+
+## Setup
+
+- The mysql database should be up and running on 127.0.0.1:3306
+
+- Our API will be exposed on localhost:8080
+
+## REST actions available
+
+*Get all people*
+
+```
+/people    GET
+```
+
+*Get person by ID*
+
+```
+/people/:id    GET
+```
+
+*Insert a new person*
+
+```
+/people    POST
+
+{
+  "name": "NewName",
+  "age": 24
+}
+```
+
+*Delete a person*
+
+```
+/people/:id    DELETE
+```
+
+## Other notes
+
+At the time of writing this issue may occur when building your project:
+
+```
+setup: Missing dependencies on foreign libraries:
+* Missing (or bad) C libraries: ssl, crypto
+```
+
+If using stack, this can be fixed by adding the following lines to your `stack.yaml`:
+
+```
+extra-include-dirs:
+- /usr/local/opt/openssl/include
+extra-lib-dirs:
+- /usr/local/opt/openssl/lib
+```
+
+Or for cabal, running your builds with these configurations passed as options.
+
+## Implementation: Main.hs
+
+Let's jump in:
+
+```haskell
+{-# LANGUAGE DataKinds                  #-}
+{-# LANGUAGE DeriveGeneric              #-}
+{-# LANGUAGE FlexibleInstances          #-}
+{-# LANGUAGE GADTs                      #-}
+{-# LANGUAGE GeneralizedNewtypeDeriving #-}
+{-# LANGUAGE MultiParamTypeClasses      #-}
+{-# LANGUAGE QuasiQuotes                #-}
+{-# LANGUAGE TemplateHaskell            #-}
+{-# LANGUAGE TypeFamilies               #-}
+{-# LANGUAGE TypeOperators              #-}
+module Lib where
+
+import           Control.Monad.IO.Class       (liftIO)
+import           Control.Monad.Logger         (NoLoggingT (..))
+import           Control.Monad.Trans.Reader   (runReaderT)
+import           Control.Monad.Trans.Resource (ResourceT, runResourceT)
+import           Data.Aeson as JSON
+import           Data.Int                     (Int64 (..))
+import           Data.Text                    (Text)
+import qualified Data.Text                    as T
+import qualified Data.Text.IO                 as T
+import           Database.Persist
+import           Database.Persist.MySQL       (ConnectInfo (..),
+                                               SqlBackend (..),
+                                               defaultConnectInfo, fromSqlKey, runMigration,
+                                               runSqlPool, toSqlKey, withMySQLConn)
+import           Database.Persist.Sql         (SqlPersistT, runSqlConn)
+import           Database.Persist.TH          (mkMigrate, mkPersist,
+                                               persistLowerCase, share,
+                                               sqlSettings)
+import           Database.Persist.Types       (PersistValue(PersistInt64))
+import           Servant                      (Handler, throwError)
+
+import           GHC.Generics
+import           Network.Wai
+import           Network.Wai.Handler.Warp
+import           Servant
+import           Servant.API
+import           System.Environment           (getArgs)
+
+share [mkPersist sqlSettings, mkMigrate "migrateAll"] [persistLowerCase|
+Person json
+    Id   Int Primary Unique
+    name Text
+    age  Text
+    deriving Eq Show Generic
+|]
+
+type Api =
+       "person" :> Get '[JSON] [Person]
+  :<|> "person" :> Capture "id" Int :> Get '[JSON] Person
+  :<|> "person" :> Capture "id" Int :> Delete '[JSON] ()
+  :<|> "person" :> ReqBody '[JSON] Person :> Post '[JSON] Person
+
+apiProxy :: Proxy Api
+apiProxy = Proxy
+
+app :: Application
+app = serve apiProxy server
+
+-- Run a database operation, and lift the result into a Handler.
+-- This minimises usage of IO operations in other functions
+runDB :: SqlPersistT (ResourceT (NoLoggingT IO)) a -> Handler a
+runDB a = liftIO $ runNoLoggingT $ runResourceT $ withMySQLConn connInfo $ runSqlConn a
+
+-- Change these out to suit your local setup
+connInfo :: ConnectInfo
+connInfo = defaultConnectInfo { connectHost = "127.0.0.1", connectUser = "root", connectPassword = "abcd", connectDatabase = "test-database" }
+
+doMigration :: IO ()
+doMigration = runNoLoggingT $ runResourceT $ withMySQLConn connInfo $ runReaderT $ runMigration migrateAll
+
+server :: Server Api
+server = 
+  personGET :<|> 
+  personGETById :<|> 
+  personDELETE :<|>
+  personPOST
+    where
+        personGET = selectPersons
+        personGETById id = selectPersonById id
+        personDELETE id = deletePerson id
+        personPOST personJson = createPerson personJson
+
+selectPersons :: Handler [Person]
+selectPersons = do
+  personList <- runDB $ selectList [] []
+  return $ map (\(Entity _ u) -> u) personList
+
+selectPersonById :: Int -> Handler Person
+selectPersonById id = do
+  sqlResult <- runDB $ get $ PersonKey id
+  case sqlResult of
+    Just person -> return person
+    Nothing -> throwError err404 { errBody = JSON.encode "Person with ID not found." }
+
+createPerson :: Person -> Handler Person
+createPerson person = do
+  attemptCreate <- runDB $ insert person
+  case attemptCreate of
+    PersonKey k -> return person
+    _           -> throwError err503 { errBody = JSON.encode "Could not create Person." }
+
+deletePerson :: Int -> Handler ()
+deletePerson id = do runDB $ delete $ PersonKey id
+
+startApp :: IO ()
+startApp = do
+    args <- getArgs
+    let arg1 = if not (null args) then Just (head args) else Nothing
+    case arg1 of
+        Just "migrate" -> doMigration
+        _              -> run 8080 app
+```
+
+## Sample requests
+
+Assuming that you have the db running and have first run `stack exec run migrate`, the following sample requests will test your API:
+
+*Create a person*
+
+```bash
+curl -X POST \
+  http://localhost:8080/person \
+  -H 'Accept: */*' \
+  -H 'Accept-Encoding: gzip, deflate' \
+  -H 'Cache-Control: no-cache' \
+  -H 'Connection: keep-alive' \
+  -H 'Content-Length: 62' \
+  -H 'Content-Type: application/json' \
+  -H 'Host: localhost:8080' \
+  -H 'cache-control: no-cache' \
+  -d '{
+	"name": "Jake",
+	"age": "25"
+}'
+```
+
+*Get all persons*
+
+```bash
+curl -X GET \
+  http://localhost:8080/person \
+  -H 'Accept: */*' \
+  -H 'Accept-Encoding: gzip, deflate' \
+  -H 'Cache-Control: no-cache' \
+  -H 'Connection: keep-alive' \
+  -H 'Content-Length: 33' \
+  -H 'Content-Type: application/json' \
+  -H 'Host: localhost:8080' \
+  -H 'cache-control: no-cache'
+```
+
+*Get person by ID*
+
+```bash
+curl -X GET \
+  http://localhost:8080/person/1 \
+  -H 'Accept: */*' \
+  -H 'Accept-Encoding: gzip, deflate' \
+  -H 'Cache-Control: no-cache' \
+  -H 'Connection: keep-alive' \
+  -H 'Content-Type: application/json' \
+  -H 'Host: localhost:8080' \
+  -H 'cache-control: no-cache'
+```

doc/cookbook/db-mysql-basics/mysql-basics.cabal

@@ -0,0 +1,40 @@
+cabal-version:       2.2
+name:                mysql-basics
+version:             0.1.0.0
+synopsis:            Simple MySQL API cookbook example
+homepage:            http://docs.servant.dev/
+license:             BSD-3-Clause
+license-file:        ../../../servant/LICENSE
+author:              Servant Contributors
+maintainer:          haskell-servant-maintainers@googlegroups.com
+build-type:          Simple
+
+executable run
+  hs-source-dirs:      .
+  main-is:             MysqlBasics.hs
+  ghc-options:         -threaded -rtsopts -with-rtsopts=-N
+  build-depends:       aeson
+                     , base
+                     , bytestring
+                     , http-client
+                     , monad-logger
+                     , mysql-simple
+                     , persistent
+                     , persistent-mysql
+                     , persistent-template
+                     , resource-pool
+                     , resourcet
+                     , servant
+                     , servant-client
+                     , servant-server
+                     , text
+                     , transformers
+                     , wai
+                     , warp
+  default-language:    Haskell2010
+  ghc-options:         -Wall -pgmL markdown-unlit
+  build-tool-depends: markdown-unlit:markdown-unlit
+
+source-repository head
+  type:     git
+  location: https://github.com/githubuser/mysql-basics

doc/cookbook/db-postgres-pool/db-postgres-pool.cabal

@@ -1,14 +1,14 @@
+cabal-version:       2.2
 name:                cookbook-db-postgres-pool
 version:             0.1
 synopsis:            Simple PostgreSQL connection pool cookbook example
-homepage:            http://haskell-servant.readthedocs.org/
-license:             BSD3
+homepage:            http://docs.servant.dev/
+license:             BSD-3-Clause
 license-file:        ../../../servant/LICENSE
 author:              Servant Contributors
 maintainer:          haskell-servant-maintainers@googlegroups.com
 build-type:          Simple
-cabal-version:       >=1.10
-tested-with:         GHC==7.8.4, GHC==7.10.3, GHC==8.0.2, GHC==8.2.2
+tested-with:         GHC==8.6.5, GHC==8.8.3, GHC ==8.10.7
 
 executable cookbook-db-postgres-pool
   main-is:             PostgresPool.lhs

doc/cookbook/db-sqlite-simple/db-sqlite-simple.cabal

@@ -1,14 +1,14 @@
+cabal-version:       2.2
 name:                cookbook-db-sqlite-simple
 version:             0.1
 synopsis:            Simple SQLite DB cookbook example
-homepage:            http://haskell-servant.readthedocs.org/
-license:             BSD3
+homepage:            http://docs.servant.dev/
+license:             BSD-3-Clause
 license-file:        ../../../servant/LICENSE
 author:              Servant Contributors
 maintainer:          haskell-servant-maintainers@googlegroups.com
 build-type:          Simple
-cabal-version:       >=1.10
-tested-with:         GHC==7.8.4, GHC==7.10.3, GHC==8.0.2, GHC==8.2.2
+tested-with:         GHC==8.6.5, GHC==8.8.3, GHC ==8.10.7
 
 executable cookbook-db-sqlite-simple
   main-is:             DBConnection.lhs
@@ -23,7 +23,7 @@ executable cookbook-db-sqlite-simple
                      , http-types >= 0.12
                      , markdown-unlit >= 0.4
                      , http-client >= 0.5
-                     , sqlite-simple >= 0.4
+                     , sqlite-simple >= 0.4.5.0
                      , transformers
   default-language:    Haskell2010
   ghc-options:         -Wall -pgmL markdown-unlit

doc/cookbook/file-upload/FileUpload.lhs

@@ -17,7 +17,7 @@ import Control.Exception
 import Control.Monad
 import Control.Monad.IO.Class
 import Data.Text.Encoding (encodeUtf8)
-import Network (withSocketsDo)
+import Network.Socket (withSocketsDo)
 import Network.HTTP.Client hiding (Proxy)
 import Network.HTTP.Client.MultipartFormData
 import Network.Wai.Handler.Warp
@@ -90,8 +90,8 @@ startServer = run 8080 (serve api upload)
 
 Finally, a main function that brings up our server and
 sends some test request with `http-client` (and not
-servant-client this time, has servant-multipart does not
-yet have support for client generation.
+servant-client this time, as servant-multipart does not
+yet have support for client generation).
 
 ``` haskell
 main :: IO ()
@@ -126,7 +126,7 @@ Content of "README.md"
 
 ## Getting Started
 
-We have a [tutorial](http://haskell-servant.readthedocs.org/en/stable/tutorial/index.html) that
+We have a [tutorial](http://docs.servant.dev/en/stable/tutorial/index.html) that
 introduces the core features of servant. After this article, you should be able
 to write your first servant webservices, learning the rest from the haddocks'
 examples.

doc/cookbook/file-upload/file-upload.cabal

@@ -1,14 +1,14 @@
+cabal-version:       2.2
 name:                cookbook-file-upload
 version:             0.1
 synopsis:            File upload cookbook example
-homepage:            http://haskell-servant.readthedocs.org/
-license:             BSD3
+homepage:            http://docs.servant.dev/
+license:             BSD-3-Clause
 license-file:        ../../../servant/LICENSE
 author:              Servant Contributors
 maintainer:          haskell-servant-maintainers@googlegroups.com
 build-type:          Simple
-cabal-version:       >=1.10
-tested-with:         GHC==7.8.4, GHC==7.10.3, GHC==8.0.2, GHC==8.2.2
+tested-with:         GHC==8.6.5, GHC==8.8.3, GHC ==8.10.7
 
 executable cookbook-file-upload
   main-is:             FileUpload.lhs

doc/cookbook/generic/Generic.lhs

@@ -0,0 +1,141 @@
+# Using generics
+
+```haskell
+{-# LANGUAGE DataKinds     #-}
+{-# LANGUAGE DeriveGeneric #-}
+{-# LANGUAGE RankNTypes    #-}
+{-# LANGUAGE TypeOperators #-}
+module Main (main, api, getLink, routesLinks, cliGet) where
+
+import Control.Exception          (throwIO)
+import Control.Monad.Trans.Reader (ReaderT, runReaderT)
+import Data.Proxy                 (Proxy (..))
+import Network.Wai.Handler.Warp   (run)
+import System.Environment         (getArgs)
+
+import Servant
+import Servant.Client
+
+import Servant.API.Generic
+import Servant.Client.Generic
+import Servant.Server.Generic
+```
+
+The usage is simple, if you only need a collection of routes.
+First you define a record with field types prefixed by a parameter `route`:
+
+```haskell
+data Routes route = Routes
+    { _get :: route :- Capture "id" Int :> Get '[JSON] String
+    , _put :: route :- ReqBody '[JSON] Int :> Put '[JSON] Bool
+    }
+  deriving (Generic)
+```
+
+Then we'll use this data type to define API, links, server and client.
+
+## API
+
+You can get a `Proxy` of the API using `genericApi`:
+
+```haskell
+api :: Proxy (ToServantApi Routes)
+api = genericApi (Proxy :: Proxy Routes)
+```
+
+It's recommended to use `genericApi` function, as then you'll get
+better error message, for example if you forget to `derive Generic`.
+
+## Links
+
+The clear advantage of record-based generics approach, is that
+we can get safe links very conveniently. We don't need to define endpoint types,
+as field accessors work as proxies:
+
+```haskell
+getLink :: Int -> Link
+getLink = fieldLink _get
+```
+
+We can also get all links at once, as a record:
+
+```haskell
+routesLinks :: Routes (AsLink Link)
+routesLinks = allFieldLinks
+```
+
+## Client
+
+Even more power starts to show when we generate a record of client functions.
+Here we use `genericClientHoist` function, which lets us simultaneously
+hoist the monad, in this case from `ClientM` to `IO`.
+
+```haskell
+cliRoutes :: Routes (AsClientT IO)
+cliRoutes = genericClientHoist
+    (\x -> runClientM x env >>= either throwIO return)
+  where
+    env = error "undefined environment"
+
+cliGet :: Int -> IO String
+cliGet = _get cliRoutes
+```
+
+## Server
+
+Finally, probably the most handy usage: we can convert record of handlers into
+the server implementation:
+
+```haskell
+record :: Routes AsServer
+record = Routes
+    { _get = return . show
+    , _put = return . odd
+    }
+
+app :: Application
+app = genericServe record
+
+main :: IO ()
+main = do
+    args <- getArgs
+    case args of
+        ("run":_) -> do
+            putStrLn "Starting cookbook-generic at http://localhost:8000"
+            run 8000 app
+        -- see this cookbook below for custom-monad explanation
+        ("run-custom-monad":_) -> do
+            putStrLn "Starting cookbook-generic with a custom monad at http://localhost:8000"
+            run 8000 (appMyMonad AppCustomState)
+        _ -> putStrLn "To run, pass 'run' argument: cabal new-run cookbook-generic run"
+```
+
+## Using generics together with a custom monad
+
+If your app uses a custom monad, here's how you can combine it with
+generics.
+
+```haskell
+data AppCustomState =
+  AppCustomState
+
+type AppM = ReaderT AppCustomState Handler
+
+apiMyMonad :: Proxy (ToServantApi Routes)
+apiMyMonad = genericApi (Proxy :: Proxy Routes)
+
+getRouteMyMonad :: Int -> AppM String
+getRouteMyMonad = return . show
+
+putRouteMyMonad :: Int -> AppM Bool
+putRouteMyMonad = return . odd
+
+recordMyMonad :: Routes (AsServerT AppM)
+recordMyMonad = Routes {_get = getRouteMyMonad, _put = putRouteMyMonad}
+
+-- natural transformation
+nt :: AppCustomState -> AppM a -> Handler a
+nt s x = runReaderT x s
+
+appMyMonad :: AppCustomState -> Application
+appMyMonad state = genericServeT (nt state) recordMyMonad

doc/cookbook/generic/generic.cabal

@@ -0,0 +1,25 @@
+cabal-version:       2.2
+name:                cookbook-generic
+version:             0.1
+synopsis:            Using custom monad to pass a state between handlers
+homepage:            http://docs.servant.dev/
+license:             BSD-3-Clause
+license-file:        ../../../servant/LICENSE
+author:              Servant Contributors
+maintainer:          haskell-servant-maintainers@googlegroups.com
+build-type:          Simple
+tested-with:         GHC==8.6.5, GHC==8.8.3, GHC ==8.10.7
+
+executable cookbook-using-custom-monad
+  main-is:             Generic.lhs
+  build-depends:       base == 4.*
+                     , servant
+                     , servant-client
+                     , servant-client-core
+                     , servant-server
+                     , base-compat
+                     , warp >= 3.2
+                     , transformers >= 0.3
+  default-language:    Haskell2010
+  ghc-options:         -Wall -pgmL markdown-unlit
+  build-tool-depends: markdown-unlit:markdown-unlit >= 0.4

doc/cookbook/hoist-server-with-context/HoistServerWithContext.lhs

@@ -0,0 +1,413 @@
+# Hoist Server With Context for Custom Monads
+
+In this example we'll combine some of the patterns we've seen in other examples
+in order to demonstrate using a custom monad with Servant's `Context` and the function
+`hoistServerWithContext`.
+
+`hoistServerWithContext` is a pattern you may encounter if you are trying to use a library such as
+[servant-auth-server](https://hackage.haskell.org/package/servant-auth-server) along
+with your own custom monad.
+
+In this example, our custom monad will be based on the commonly used `ReaderT env IO a` stack.
+We'll create an `AppCtx` to represent our `env` and include some logging utilities as well as
+other variables we'd like to have available.
+
+In addition, in order to demonstrate a custom `Context`, we'll also include authentication in
+our example. As noted previously (in [jwt-and-basic-auth](../jwt-and-basic-auth/JWTAndBasicAuth.lhs)),
+while basic authentication comes with Servant itself,
+[servant-auth](https://hackage.haskell.org/package/servant-auth) and
+[servant-auth-server](https://hackage.haskell.org/package/servant-auth-server)
+packages are needed for JWT-based authentication.
+
+Finally, we're going to use [fast-logger](http://hackage.haskell.org/package/fast-logger)
+for our logging example below.
+
+This recipe uses the following ingredients:
+
+```haskell
+{-# LANGUAGE DataKinds #-}
+{-# LANGUAGE DeriveGeneric #-}
+{-# LANGUAGE OverloadedStrings #-}
+{-# LANGUAGE TypeFamilies #-}
+{-# LANGUAGE TypeOperators #-}
+
+import Prelude ()
+import Prelude.Compat
+
+import Control.Monad.IO.Class (liftIO)
+import Control.Monad.Reader
+import Data.Aeson
+import Data.Default
+import Data.Proxy
+import Data.Text
+import Data.Time.Clock ( UTCTime, getCurrentTime )
+import GHC.Generics
+import Network.Wai (Middleware)
+import Network.Wai.Handler.Warp as Warp
+import Network.Wai.Middleware.RequestLogger
+import Network.Wai.Middleware.RequestLogger.JSON
+import Servant as S
+import Servant.Auth as SA
+import Servant.Auth.Server as SAS
+import System.Log.FastLogger                      ( ToLogStr(..)
+                                                  , LoggerSet
+                                                  , defaultBufSize
+                                                  , newStdoutLoggerSet
+                                                  , flushLogStr
+                                                  , pushLogStrLn )
+
+
+port :: Int
+port = 3001
+```
+
+## Custom Monad
+
+Let's say we'd like to create a custom monad based on `ReaderT env` in order to hold
+access to a config object as well as some logging utilities.
+
+With that, we could define an `AppCtx` and `AppM` like this:
+
+```haskell
+type AppM = ReaderT AppCtx Handler
+
+data AppCtx = AppCtx {
+  _getConfig   :: SiteConfig
+  , _getLogger :: LoggerSet
+  }
+
+data SiteConfig = SiteConfig {
+  environment     :: !Text
+  , version       :: !Text
+  , adminUsername :: !Text
+  , adminPasswd   :: !Text
+} deriving (Generic, Show)
+```
+
+This `SiteConfig` is a simple example: it refers to our deployment environment as well as an
+application version. For instance, we may do something different based on the environment our app is
+deployed into. When emitting log messages, we may want to include information about
+the deployed version of our application.
+
+In addition, we're going to identify a single admin user in our config and use
+that definition to authenticate requests inside our handlers. This is not too
+flexible (and probably not too secure...), but it works as a simple example.
+
+## Logging
+
+A common contemporary pattern is to emit log messages as JSON for later ingestion
+into a database like Elasticsearch.
+
+To emit JSON log messages, we'll create a `LogMessage` object and make it so we can turn it
+into a JSON-encoded `LogStr` (a type from `fast-logger`).
+
+```haskell
+data LogMessage = LogMessage {
+  message        :: !Text
+  , timestamp    :: !UTCTime
+  , level        :: !Text
+  , lversion     :: !Text
+  , lenvironment :: !Text
+} deriving (Eq, Show, Generic)
+
+instance FromJSON LogMessage
+instance ToJSON LogMessage where
+  toEncoding = genericToEncoding defaultOptions
+
+instance ToLogStr LogMessage where
+  toLogStr = toLogStr . encode
+```
+
+Eventually, when we'd like to emit a log message inside one of our Handlers, it'll look like this:
+
+```haskell
+sampleHandler :: AppM LogMessage
+sampleHandler = do
+    config <- asks _getConfig
+    logset <- asks _getLogger
+
+    tstamp <- liftIO getCurrentTime
+    let logMsg = LogMessage { message = "let's do some logging!"
+                            , timestamp = tstamp
+                            , level = "info"
+                            , lversion = version config
+                            , lenvironment = environment config
+                            }
+    -- emit log message
+    liftIO $ pushLogStrLn logset $ toLogStr logMsg
+    -- return handler result (for simplicity, result is also a LogMessage)
+    pure logMsg
+```
+
+## Authentication
+
+To demonstrate the other part of this recipe, we are going to use a simple
+representation of a user, someone who may have access to an admin section of our site:
+
+```haskell
+data AdminUser = AdminUser { name :: Text }
+   deriving (Eq, Show, Read, Generic)
+```
+
+The following instances are needed for JWT:
+
+```haskell
+instance ToJSON AdminUser
+instance FromJSON AdminUser
+instance SAS.ToJWT AdminUser
+instance SAS.FromJWT AdminUser
+```
+
+## API
+
+Now we can define our API.
+
+We'll have an `admin` endpoint and a `login` endpoint that takes a `LoginForm`:
+
+```haskell
+type AdminApi =
+  "admin" :> Get '[JSON] LogMessage
+
+type LoginApi =
+  "login"
+      :> ReqBody '[JSON] LoginForm
+      :> Post '[JSON] (Headers '[ Header "Set-Cookie" SetCookie, Header "Set-Cookie" SetCookie] LogMessage)
+
+data LoginForm = LoginForm {
+  username :: Text
+  , password :: Text
+} deriving (Eq, Show, Generic)
+
+instance ToJSON LoginForm
+instance FromJSON LoginForm
+```
+
+We can combine both APIs into one like so:
+
+```haskell
+type AdminAndLogin auths = (SAS.Auth auths AdminUser :> AdminApi) :<|> LoginApi
+```
+
+## Server
+
+When we define our server, we'll have to define handlers for the `AdminApi` and the `LoginApi` and
+we'll have to supply `JWTSettings` and `CookieSettings` so our `login` handler can authenticate users:
+
+```haskell
+adminServer :: SAS.CookieSettings -> SAS.JWTSettings -> ServerT (AdminAndLogin auths) AppM
+adminServer cs jwts = adminHandler :<|> loginHandler cs jwts
+```
+
+The `admin` route should receive an authenticated `AdminUser` as an argument
+or it should return a `401`:
+
+```haskell
+adminHandler :: AuthResult AdminUser -> AppM LogMessage
+adminHandler (SAS.Authenticated adminUser) = do
+    config <- asks _getConfig
+    logset <- asks _getLogger
+
+    tstamp <- liftIO getCurrentTime
+    let logMsg = LogMessage { message = "Admin User accessing admin: " <> name adminUser
+                            , timestamp = tstamp
+                            , level = "info"
+                            , lversion = version config
+                            , lenvironment = environment config
+                            }
+    -- emit log message
+    liftIO $ pushLogStrLn logset $ toLogStr logMsg
+    -- return handler result (for simplicity, result is a LogMessage)
+    pure logMsg
+adminHandler _ = throwError err401
+```
+
+By contrast, the `login` handler is waiting for a `POST` with a login form.
+
+If login is successful, it will set session cookies and return a value.
+
+Here we're going to include lots of log messages:
+
+```haskell
+loginHandler :: CookieSettings
+             -> JWTSettings
+             -> LoginForm
+             -> AppM (Headers '[ Header "Set-Cookie" SetCookie, Header "Set-Cookie" SetCookie] LogMessage)
+loginHandler cookieSettings jwtSettings form = do
+  config     <- asks _getConfig
+  logset     <- asks _getLogger
+
+  tstamp <- liftIO getCurrentTime
+  let logMsg = LogMessage { message = "AdminUser login attempt failed!"
+                          , timestamp = tstamp
+                          , level = "info"
+                          , lversion = version config
+                          , lenvironment = environment config
+                          }
+  case validateLogin config form of
+    Nothing -> do
+      liftIO $ pushLogStrLn logset $ toLogStr logMsg
+      throwError err401
+    Just usr -> do
+      mApplyCookies <- liftIO $ SAS.acceptLogin cookieSettings jwtSettings usr
+      case mApplyCookies of
+        Nothing           -> do
+          liftIO $ pushLogStrLn logset $ toLogStr logMsg
+          throwError err401
+        Just applyCookies -> do
+          let successMsg = logMsg{message = "AdminUser successfully authenticated!"}
+          liftIO $ pushLogStrLn logset $ toLogStr successMsg
+          pure $ applyCookies successMsg
+loginHandler _ _ _ = throwError err401
+
+validateLogin :: SiteConfig -> LoginForm -> Maybe AdminUser
+validateLogin config (LoginForm uname passwd ) =
+  if (uname == adminUsername config) && (passwd == adminPasswd config)
+    then Just $ AdminUser uname
+    else Nothing
+```
+
+## `serveWithContext` and `hoistServerWithContext`
+
+In order to build a working server, we'll need to `hoist` our custom monad
+into Servant's Handler monad. We'll also need to pass in the proper context to ensure
+authentication will work.
+
+This will require both `serveWithContext` and `hoistServerWithContext`.
+
+Let's define the function which will create our `Application`:
+
+```haskell
+adminLoginApi :: Proxy (AdminAndLogin '[JWT])
+adminLoginApi = Proxy
+
+mkApp :: Context '[SAS.CookieSettings, SAS.JWTSettings] -> CookieSettings -> JWTSettings -> AppCtx -> Application
+mkApp cfg cs jwts ctx =
+  serveWithContext adminLoginApi cfg $
+    hoistServerWithContext adminLoginApi (Proxy :: Proxy '[SAS.CookieSettings, SAS.JWTSettings])
+      (flip runReaderT ctx) (adminServer cs jwts)
+```
+
+One footnote: because we'd like our logs to be in JSON form, we'll also create a `Middleware` object
+so that `Warp` *also* will emit logs as JSON. This will ensure *all* logs are emitted as JSON:
+
+```haskell
+jsonRequestLogger :: IO Middleware
+jsonRequestLogger =
+  mkRequestLogger $ def { outputFormat = CustomOutputFormatWithDetails formatAsJSON }
+```
+
+We now have all the pieces we need to serve our application inside a `main` function:
+
+```haskell
+main :: IO ()
+main = do
+  -- typically, we'd create our config from environment variables
+  -- but we're going to just make one here
+  let config = SiteConfig "dev" "1.0.0" "admin" "secretPassword"
+
+  warpLogger <- jsonRequestLogger
+  appLogger <- newStdoutLoggerSet defaultBufSize
+
+  tstamp <- getCurrentTime
+  myKey <- generateKey
+
+  let lgmsg = LogMessage {
+    message = "My app starting up!"
+    , timestamp = tstamp
+    , level = "info"
+    , lversion = version config
+    , lenvironment = environment config
+  }
+  pushLogStrLn appLogger (toLogStr lgmsg) >> flushLogStr appLogger
+
+  let ctx = AppCtx config appLogger
+
+      warpSettings = Warp.defaultSettings
+      portSettings = Warp.setPort port warpSettings
+      settings = Warp.setTimeout 55 portSettings
+      jwtCfg = defaultJWTSettings myKey
+      cookieCfg = if environment config == "dev"
+                  then defaultCookieSettings{cookieIsSecure=SAS.NotSecure}
+                  else defaultCookieSettings
+      cfg = cookieCfg :. jwtCfg :. EmptyContext
+
+  Warp.runSettings settings $ warpLogger $ mkApp cfg cookieCfg jwtCfg ctx
+```
+
+
+## Usage
+
+Now we can run it and try it out with `curl`. In one terminal, let's run our application
+and see what our log output looks like:
+
+```$ ./cookbook-hoist-server-with-context
+{"message":"My app starting up!","timestamp":"2018-10-04T00:33:12.482568Z","level":"info","lversion":"1.0.0","lenvironment":"dev"}
+```
+
+In another terminal, let's ensure that it fails with `err401` if
+we're not authenticated:
+
+```
+$ curl -v 'http://localhost:3001/admin'
+…
+< HTTP/1.1 401 Unauthorized
+```
+
+```
+$ curl -v -XPOST 'http://localhost:3001/login' \
+  -H "Content-Type:application/json" \
+  -d '{"username": "bad", "password": "wrong"}'
+…
+< HTTP/1.1 401 Unauthorized
+```
+
+And in the other terminal with our log messages (from our JSON `Middleware`):
+
+```
+{"time":"03/Oct/2018:17:35:56 -0700","response":{"status":401,"size":null,"body":""},"request":{"httpVersion":"1.1","path":"/admin","size":0,"body":"","durationMs":0.22,"remoteHost":{"hostAddress":"127.0.0.1","port":51029},"headers":[["Host","localhost:3001"],["User-Agent","curl/7.60.0"],["Accept","*/*"]],"queryString":[],"method":"GET"}}
+```
+
+Now let's see that authentication works, and that we get JWTs:
+
+```
+$ curl -v -XPOST 'http://localhost:3001/login' \
+  -H "Content-Type:application/json" \
+  -d '{"username": "admin", "password": "secretPassword"}'
+…
+< HTTP/1.1 200 OK
+...
+< Server: Warp/3.2.25
+< Content-Type: application/json;charset=utf-8
+< Set-Cookie: JWT-Cookie=eyJhbGciOiJIUzUxMiJ9.eyJkYXQiOnsibmFtZSI6ImFkbWluIn19.SIoRcABKSO4mXnRifzqPWlHJUhVwuy32Qon7s1E_c3vHOsLXdXyX4V4eXOw9tMFoeIqgsXMZucqoFb36vAdKwQ; Path=/; HttpOnly; SameSite=Lax
+< Set-Cookie: XSRF-TOKEN=y5PmrYHX3ywFUCwGRQqHh1TDheTLiQpwRQB3FFRd8N4=; Path=/
+...
+{"message":"AdminUser succesfully authenticated!","timestamp":"2018-10-04T00:37:44.455441Z","level":"info","lversion":"1.0.0","lenvironment":"dev"}
+```
+
+And in the other terminal with our log messages (note that logging out passwords is insecure...):
+
+```
+{"message":"AdminUser succesfully authenticated!","timestamp":"2018-10-04T00:37:44.455441Z","level":"info","lversion":"1.0.0","lenvironment":"dev"}
+{"time":"03/Oct/2018:17:37:44 -0700","response":{"status":200,"size":null,"body":null},"request":{"httpVersion":"1.1","path":"/login","size":51,"body":"{\"username\": \"admin\", \"password\": \"secretPassword\"}","durationMs":0.23,"remoteHost":{"hostAddress":"127.0.0.1","port":51044},"headers":[["Host","localhost:3001"],["User-Agent","curl/7.60.0"],["Accept","*/*"],["Content-Type","application/json"],["Content-Length","51"]],"queryString":[],"method":"POST"}}
+```
+
+Finally, let's make sure we can access a protected resource with our tokens:
+
+```
+$ export jwt=eyJhbGciOiJIUzUxMiJ9.eyJkYXQiOnsibmFtZSI6ImFkbWluIn19.SIoRcABKSO4mXnRifzqPWlHJUhVwuy32Qon7s1E_c3vHOsLXdXyX4V4eXOw9tMFoeIqgsXMZucqoFb36vAdKwQ
+$ curl -v \
+  -H "Authorization: Bearer $jwt" \
+  'http://localhost:3001/admin'
+…
+< HTTP/1.1 200 OK
+{"message":"Admin User accessing admin: admin","timestamp":"2018-10-04T00:58:07.216605Z","level":"info","lversion":"1.0.0","lenvironment":"dev"}
+```
+
+And we should see this message logged-out as well:
+
+```
+{"message":"Admin User accessing admin: admin","timestamp":"2018-10-04T00:58:07.216605Z","level":"info","lversion":"1.0.0","lenvironment":"dev"}
+```
+
+This program is available as a cabal project
+[here](https://github.com/haskell-servant/servant/tree/master/doc/cookbook/hoist-server-with-context).

doc/cookbook/hoist-server-with-context/hoist-server-with-context.cabal

@@ -0,0 +1,37 @@
+cabal-version:       2.2
+name:                cookbook-hoist-server-with-context
+version:             0.0.1
+synopsis:            JWT and basic access authentication with a Custom Monad cookbook example
+description:         Using servant-auth to support both JWT-based and basic
+                     authentication.
+homepage:            http://docs.servant.dev/
+license:             BSD-3-Clause
+license-file:        ../../../servant/LICENSE
+author:              Servant Contributors
+maintainer:          haskell-servant-maintainers@googlegroups.com
+category:            Servant
+build-type:          Simple
+tested-with:         GHC==8.6.5, GHC==8.8.3, GHC ==8.10.7
+
+executable cookbook-hoist-server-with-context
+  main-is:             HoistServerWithContext.lhs
+  build-depends:       base == 4.*
+                     , base-compat
+                     , text >= 1.2
+                     , aeson >= 1.2
+                     , data-default
+                     , fast-logger
+                     , servant
+                     , servant-server
+                     , servant-auth >= 0.3.2
+                     , servant-auth-server >= 0.4.4.0
+                     , time
+                     , warp >= 3.2
+                     , wai >= 3.2
+                     , wai-extra
+                     , http-types >= 0.12
+                     , bytestring >= 0.10.4
+                     , mtl
+  default-language:    Haskell2010
+  ghc-options:         -Wall -pgmL markdown-unlit
+  build-tool-depends:  markdown-unlit:markdown-unlit

doc/cookbook/https/Https.lhs

@@ -34,16 +34,16 @@ app = serve api server
 ```
 
 It's now time to actually run the `Application`.
-The [`warp-tls`](https://hackage.haskell.org/package/warp-tls-3.2.4/docs/Network-Wai-Handler-WarpTLS.html)
+The [`warp-tls`](https://hackage.haskell.org/package/warp-tls/docs/Network-Wai-Handler-WarpTLS.html)
 package provides two functions for running an `Application`, called
-[`runTLS`](https://hackage.haskell.org/package/warp-tls-3.2.4/docs/Network-Wai-Handler-WarpTLS.html#v:runTLS)
-and [`runTLSSocket`](https://hackage.haskell.org/package/warp-tls-3.2.4/docs/Network-Wai-Handler-WarpTLS.html#v:runTLSSocket).
+[`runTLS`](https://hackage.haskell.org/package/warp-tls/docs/Network-Wai-Handler-WarpTLS.html#v:runTLS)
+and [`runTLSSocket`](https://hackage.haskell.org/package/warp-tls/docs/Network-Wai-Handler-WarpTLS.html#v:runTLSSocket).
 We will be using the first one.
 
 It takes two arguments,
-[the TLS settings](https://hackage.haskell.org/package/warp-tls-3.2.4/docs/Network-Wai-Handler-WarpTLS.html#t:TLSSettings)
+[the TLS settings](https://hackage.haskell.org/package/warp-tls/docs/Network-Wai-Handler-WarpTLS.html#t:TLSSettings)
 (certificates, keys, ciphers, etc)
-and [the warp settings](https://hackage.haskell.org/package/warp-3.2.12/docs/Network-Wai-Handler-Warp-Internal.html#t:Settings)
+and [the warp settings](https://hackage.haskell.org/package/warp/docs/Network-Wai-Handler-Warp-Internal.html#t:Settings)
 (port, logger, etc).
 
 We will be using very simple settings for this example but you are of

doc/cookbook/https/https.cabal

@@ -1,14 +1,14 @@
+cabal-version:       2.2
 name:                cookbook-https
 version:             0.1
 synopsis:            HTTPS cookbook example
-homepage:            http://haskell-servant.readthedocs.org/
-license:             BSD3
+homepage:            http://docs.servant.dev/
+license:             BSD-3-Clause
 license-file:        ../../../servant/LICENSE
 author:              Servant Contributors
 maintainer:          haskell-servant-maintainers@googlegroups.com
 build-type:          Simple
-cabal-version:       >=1.10
-tested-with:         GHC==7.8.4, GHC==7.10.3, GHC==8.0.2, GHC==8.2.2
+tested-with:         GHC==8.6.5, GHC==8.8.3, GHC ==8.10.7
 
 executable cookbook-https
   main-is:             Https.lhs
@@ -17,7 +17,7 @@ executable cookbook-https
                      , servant-server
                      , wai >= 3.2
                      , warp >= 3.2
-                     , warp-tls >= 3.2
+                     , warp-tls >= 3.2.9
                      , markdown-unlit >= 0.4
   default-language:    Haskell2010
   ghc-options:         -Wall -pgmL markdown-unlit

doc/cookbook/index.rst

@@ -6,8 +6,8 @@ how to solve many common problems with servant. If you're
 interested in contributing examples of your own, feel free
 to open an issue or a pull request on
 `our github repository <https://github.com/haskell-servant/servant>`_
-or even to just get in touch with us on the **#servant** IRC channel
-on freenode or on
+or even to just get in touch with us on the `**#haskell-servant** IRC channel
+on libera.chat <https://web.libera.chat/#haskell-servant>_ or on
 `the mailing list <https://groups.google.com/forum/#!forum/haskell-servant>`_.
 
 The scope is very wide. Simple and fancy authentication schemes,
@@ -18,10 +18,23 @@ you name it!
   :maxdepth: 1
 
   structuring-apis/StructuringApis.lhs
+  generic/Generic.lhs
   https/Https.lhs
+  db-mysql-basics/MysqlBasics.lhs
   db-sqlite-simple/DBConnection.lhs
   db-postgres-pool/PostgresPool.lhs
   using-custom-monad/UsingCustomMonad.lhs
+  using-free-client/UsingFreeClient.lhs
+  custom-errors/CustomErrors.lhs
+  uverb/UVerb.lhs
   basic-auth/BasicAuth.lhs
+  basic-streaming/Streaming.lhs
   jwt-and-basic-auth/JWTAndBasicAuth.lhs
+  hoist-server-with-context/HoistServerWithContext.lhs
   file-upload/FileUpload.lhs
+  pagination/Pagination.lhs
+  curl-mock/CurlMock.lhs
+  sentry/Sentry.lhs
+  testing/Testing.lhs
+  open-id-connect/OpenIdConnect.lhs
+  managed-resource/ManagedResource.lhs

doc/cookbook/jwt-and-basic-auth/jwt-and-basic-auth.cabal

@@ -1,22 +1,19 @@
+cabal-version:       2.2
 name:                cookbook-jwt-and-basic-auth
 version:             0.0.1
 synopsis:            JWT and basic access authentication cookbook example
 description:         Using servant-auth to support both JWT-based and basic
                      authentication.
-homepage:            http://haskell-servant.readthedocs.org/
-license:             BSD3
+homepage:            http://docs.servant.dev/
+license:             BSD-3-Clause
 license-file:        ../../../servant/LICENSE
 author:              Servant Contributors
 maintainer:          haskell-servant-maintainers@googlegroups.com
 category:            Servant
 build-type:          Simple
-cabal-version:       >=1.10
-tested-with:         GHC==7.8.4, GHC==7.10.3, GHC==8.0.2, GHC==8.2.2
+tested-with:         GHC==8.6.5, GHC==8.8.3, GHC ==8.10.7
 
 executable cookbook-jwt-and-basic-auth
-  if !impl(ghc >= 7.10)
-    buildable: False
-
   main-is:             JWTAndBasicAuth.lhs
   build-depends:       base == 4.*
                      , text >= 1.2
@@ -25,7 +22,7 @@ executable cookbook-jwt-and-basic-auth
                      , servant
                      , servant-client
                      , servant-server
-                     , servant-auth ==0.3.*
+                     , servant-auth == 0.4.*
                      , servant-auth-server >= 0.3.1.0
                      , warp >= 3.2
                      , wai >= 3.2

doc/cookbook/managed-resource/ManagedResource.lhs

@@ -0,0 +1,114 @@
+# Request-lifetime Managed Resources
+
+Let's see how we can write a handle that uses a resource managed by Servant. The resource is created automatically by Servant when the server recieves a request, and the resource is automatically destroyed when the server is finished handling a request.
+
+As usual, we start with a little bit of throat clearing.
+
+
+``` haskell
+{-# LANGUAGE DataKinds #-}
+{-# LANGUAGE TypeOperators #-}
+import Control.Concurrent
+import Control.Exception (bracket, throwIO)
+import Control.Monad.IO.Class
+import Control.Monad.Trans.Resource
+import Data.Acquire
+import Network.HTTP.Client (newManager, defaultManagerSettings)
+import Network.Wai.Handler.Warp
+import Servant
+import Servant.Client
+import System.IO
+```
+
+Here we define an API type that uses the `WithResource` combinator. The server handler for an endpoint with a `WithResource res` component will receive a value of that type as an argument.
+
+``` haskell
+type API = WithResource Handle :> ReqBody '[PlainText] String :> Post '[JSON] NoContent
+
+api :: Proxy API
+api = Proxy
+```
+
+But this resource value has to come from somewhere. Servant obtains the value using an Acquire provided in the context. The Acquire knows how to both create and destroy resources of a particular type.
+
+``` haskell
+appContext :: Context '[Acquire Handle]
+appContext = acquireHandle :. EmptyContext
+
+acquireHandle :: Acquire Handle
+acquireHandle = mkAcquire newHandle closeHandle
+
+newHandle :: IO Handle
+newHandle = do
+  putStrLn "opening file"
+  h <- openFile "test.txt" AppendMode
+  putStrLn "opened file"
+  return h
+
+closeHandle :: Handle -> IO ()
+closeHandle h = do
+  putStrLn "closing file"
+  hClose h
+  putStrLn "closed file"
+```
+
+Now we create the handler which will use this resource. This handler will write the request message to the System.IO.Handle which was provided to us. In some situations the handler will succeed, but in some in will fail. In either case, Servant will clean up the resource for us.
+
+``` haskell
+server :: Server API
+server = writeToFile
+
+  where writeToFile :: (ReleaseKey, Handle) -> String -> Handler NoContent
+        writeToFile (_, h) msg = case msg of
+          "illegal" -> error "wait, that's illegal!"
+          legalMsg -> liftIO $ do
+            putStrLn "writing file"
+            hPutStrLn h legalMsg
+            putStrLn "wrote file"
+            return NoContent
+```
+
+Finally we run the server in the background while we post messages to it.
+
+``` haskell
+runApp :: IO ()
+runApp = run 8080 (serveWithContext api appContext $ server)
+
+postMsg :: String -> ClientM NoContent
+postMsg = client api
+
+main :: IO ()
+main = do
+  mgr <- newManager defaultManagerSettings
+  bracket (forkIO $ runApp) killThread $ \_ -> do
+    ms <- flip runClientM (mkClientEnv mgr (BaseUrl Http "localhost" 8080 "")) $ do
+      liftIO $ putStrLn "sending hello message"
+      _ <- postMsg "hello"
+      liftIO $ putStrLn "sending illegal message"
+      _ <- postMsg "illegal"
+      liftIO $ putStrLn "done"
+    print ms
+```
+
+This program prints
+
+```
+sending hello message
+opening file
+opened file
+writing file
+wrote file
+closing file
+closed file
+sending illegal message
+opening file
+opened file
+closing file
+closed file
+wait, that's illegal!
+CallStack (from HasCallStack):
+  error, called at ManagedResource.lhs:63:24 in main:Main
+Left (FailureResponse (Request {requestPath = (BaseUrl {baseUrlScheme = Http, baseUrlHost = "localhost", baseUrlPort = 8080, baseUrlPath = ""},""), requestQueryString = fromList [], requestBody = Just ((),text/plain;charset=utf-8), requestAccept = fromList [], requestHeaders = fromList [], requestHttpVersion = HTTP/1.1, requestMethod = "POST"}) (Response {responseStatusCode = Status {statusCode = 500, statusMessage = "Internal Server Error"}, responseHeaders = fromList [("Transfer-Encoding","chunked"),("Date","Thu, 24 Nov 2022 21:04:47 GMT"),("Server","Warp/3.3.23"),("Content-Type","text/plain; charset=utf-8")], responseHttpVersion = HTTP/1.1, responseBody = "Something went wrong"}))
+```
+
+and appends to a file called `test.txt`. We can see from the output that when a legal message is sent, the file is opened, written to, and closed. We can also see that when an illegal message is sent, the file is opened but not written to. Crucially, it is still closed even though the handler threw an exception.

doc/cookbook/managed-resource/managed-resource.cabal

@@ -0,0 +1,30 @@
+cabal-version:       2.2
+name:                cookbook-managed-resource
+version:             0.1
+synopsis:            Simple managed resource cookbook example
+homepage:            http://docs.servant.dev/
+license:             BSD-3-Clause
+license-file:        ../../../servant/LICENSE
+author:              Servant Contributors
+maintainer:          haskell-servant-maintainers@googlegroups.com
+build-type:          Simple
+tested-with:         GHC==9.4.2
+
+executable cookbook-managed-resource
+  main-is:             ManagedResource.lhs
+  build-depends:       base == 4.*
+                     , text >= 1.2
+                     , aeson >= 1.2
+                     , servant
+                     , servant-client
+                     , servant-server
+                     , warp >= 3.2
+                     , wai >= 3.2
+                     , http-types >= 0.12
+                     , markdown-unlit >= 0.4
+                     , http-client >= 0.5
+                     , transformers
+                     , resourcet
+  default-language:    Haskell2010
+  ghc-options:         -Wall -pgmL markdown-unlit
+  build-tool-depends: markdown-unlit:markdown-unlit

doc/cookbook/open-id-connect/OpenIdConnect.cabal

@@ -0,0 +1,45 @@
+cabal-version:  2.2
+name:           open-id-connect
+version:        0.1
+synopsis:       OpenId Connect with Servant example
+homepage:       http://haskell-servant.readthedocs.org/
+license:        BSD-3-Clause
+license-file:   ../../../servant/LICENSE
+author:         Servant Contributors
+maintainer:     haskell-servant-maintainers@googlegroups.com
+build-type:     Simple
+tested-with:    GHC==8.6.5
+
+executable cookbook-openidconnect
+  main-is: OpenIdConnect.lhs
+  build-depends: base ==4.*
+    , aeson
+    , aeson-pretty
+    , binary
+    , blaze-html
+    , blaze-markup
+    , bytestring
+    , case-insensitive
+    , cereal
+    , containers
+    , generic-lens
+    , http-client
+    , http-client-tls
+    , http-types
+    , jose-jwt
+    , lens
+    , lens-aeson
+    , oidc-client
+    , protolude
+    , random
+    , servant
+    , servant-blaze
+    , servant-server
+    , text
+    , time
+    , vector
+    , wai
+    , warp >= 3.2
+  default-language: Haskell2010
+  ghc-options: -Wall -Wcompat -Wincomplete-uni-patterns -Wredundant-constraints -Wnoncanonical-monad-instances -pgmL markdown-unlit
+  build-tool-depends: markdown-unlit:markdown-unlit >= 0.4

doc/cookbook/open-id-connect/OpenIdConnect.lhs

@@ -0,0 +1,472 @@
+[OpenID Connect](https://openid.net/connect/)
+=============================================
+
+Use OpenID Connect to authenticate your users.
+This example use google OIDC provider.
+It was made for a working with single page application where
+some login token would be saved in the user agent local storage.
+
+Workflow:
+
+1.  user is presented with a login button,
+2.  when the user clicks on the button it is redirected to the OIDC
+    provider,
+3.  the user login in the OIDC provider,
+4.  the OIDC provider will redirect the user and provide a `code`,
+5.  the server will use this code to make a POST to the OIDC provider
+    and will get back authentication infos,
+6.  The user will get display an HTML page that will save a secret
+    identifying him in the local storage, then it will be redirected to
+    /.
+
+Let's put the imports behind us:
+
+``` haskell
+{-# LANGUAGE DataKinds             #-}
+{-# LANGUAGE DeriveGeneric         #-}
+{-# LANGUAGE DuplicateRecordFields #-}
+{-# LANGUAGE FlexibleContexts      #-}
+{-# LANGUAGE FlexibleInstances     #-}
+{-# LANGUAGE MultiParamTypeClasses #-}
+{-# LANGUAGE NoImplicitPrelude     #-}
+{-# LANGUAGE OverloadedLists       #-}
+{-# LANGUAGE OverloadedStrings     #-}
+{-# LANGUAGE PartialTypeSignatures #-}
+{-# LANGUAGE PolyKinds             #-}
+{-# LANGUAGE RankNTypes            #-}
+{-# LANGUAGE RecordWildCards       #-}
+{-# LANGUAGE ScopedTypeVariables   #-}
+{-# LANGUAGE TypeFamilies          #-}
+{-# LANGUAGE TypeOperators         #-}
+{-# LANGUAGE TypeSynonymInstances  #-}
+
+module Main where
+
+import           Protolude
+
+import           Data.Aeson
+                 (FromJSON (..), (.:))
+import qualified Data.Aeson                       as JSON
+import qualified Data.Aeson.Types                 as AeT
+import qualified Data.ByteString.Lazy             as LBS
+import qualified Data.List                        as List
+import qualified Data.Text                        as Text
+import           Jose.Jwt
+                 (Jwt (..), decodeClaims)
+import           Network.HTTP.Client
+                 (Manager, newManager)
+import           Network.HTTP.Client.TLS
+                 (tlsManagerSettings)
+import           Network.Wai.Handler.Warp
+                 (run)
+import           Servant
+import           Servant.HTML.Blaze
+                 (HTML)
+import qualified System.Random                    as Random
+import           Text.Blaze
+                 (ToMarkup (..))
+import qualified Text.Blaze.Html                  as H
+import           Text.Blaze.Html5
+                 ((!))
+import qualified Text.Blaze.Html5                 as H
+import qualified Text.Blaze.Html5.Attributes      as HA
+import           Text.Blaze.Renderer.Utf8
+                 (renderMarkup)
+import qualified Web.OIDC.Client                  as O
+```
+
+You'll need to create a new OpenID Connect client in an OpenID Provider.
+This example was tested with Google.
+
+You can find a list of public OIDC provider here:
+https://connect2id.com/products/nimbus-oauth-openid-connect-sdk/openid-connect-providers
+
+I copied some here:
+
+- Google: https://developers.google.com/identity/protocols/OpenIDConnect
+  more precisely: https://console.developers.google.com/apis/credentials
+- Microsoft: https://docs.microsoft.com/en-us/previous-versions/azure/dn645541(v=azure.100)
+- Yahoo: https://developer.yahoo.com/oauth2/guide/openid_connect/
+- PayPal: https://developer.paypal.com/docs/integration/direct/identity/log-in-with-paypal/
+
+During the configuration you'll need to provide a redirect uri.
+The redirect_uri should correspond to the uri user will be redirected to
+after a successful login into the OpenID provider.
+
+So during your test, you should certainly just use `http://localhost:3000/login/cb`.
+In general you should use your own domain name.
+
+You'll then be given a `client_id` and a `client_password`.
+Fill those values in here:
+
+``` haskell
+oidcConf :: OIDCConf
+oidcConf = OIDCConf { redirectUri = "http://localhost:3000/login/cb"
+                    , clientId = "xxxxxxxxxxxx-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx.apps.googleusercontent.com"
+                    , clientPassword = "************************" }
+```
+
+Then we declare our main server:
+
+``` haskell
+main :: IO ()
+main = do
+  oidcEnv <- initOIDC oidcConf
+  run 3000 (app oidcEnv)
+
+type API = IdentityRoutes Customer
+         :<|> Get '[HTML] Homepage
+
+api :: Proxy API
+api = Proxy
+
+server :: OIDCEnv -> Server API
+server oidcEnv = serveOIDC oidcEnv handleOIDCLogin
+               :<|> return Homepage
+
+-- | Then main app
+app :: OIDCEnv -> Application
+app oidcEnv = serve api (server oidcEnv)
+```
+
+OIDC
+----
+
+That part try to separate concern, and certainly in a real world
+application that should be in its distinct module.
+
+``` haskell
+-- * OIDC
+
+data OIDCConf =
+  OIDCConf { redirectUri    :: ByteString
+           , clientId       :: ByteString
+           , clientPassword :: ByteString
+           } deriving (Show, Eq)
+```
+
+First we need to initialize OIDC.
+A short explanation about it:
+
+- to complete the workflow we need to make a POST request to the OIDC provider.
+  So we need to create an http manager to make those call properly.
+- Then in order to prevent replay attack, each time an user wants to login we
+  should provide a random string called the `state`. When the user is
+  redirected to the `redirect_uri`, the OIDC provider should provide the same
+ `state` along a `code` parameter.
+
+``` haskell
+initOIDC :: OIDCConf -> IO OIDCEnv
+initOIDC OIDCConf{..} = do
+  mgr  <- newManager tlsManagerSettings
+  prov <- O.discover "https://accounts.google.com" mgr
+  let oidc     = O.setCredentials clientId clientPassword redirectUri (O.newOIDC prov)
+  return OIDCEnv { oidc = oidc
+                 , mgr = mgr
+                 , genState = genRandomBS
+                 , prov = prov
+                 , redirectUri = redirectUri
+                 , clientId = clientId
+                 , clientPassword = clientPassword
+                 }
+
+data OIDCEnv = OIDCEnv { oidc           :: O.OIDC
+                       , mgr            :: Manager
+                       , genState       :: IO ByteString
+                       , prov           :: O.Provider
+                       , redirectUri    :: ByteString
+                       , clientId       :: ByteString
+                       , clientPassword :: ByteString
+                       }
+```
+
+The `IdentityRoutes` are two endpoints:
+
+- an endpoint to redirect the users to the OIDC Provider,
+- another one the user will be redirected to from the OIDC Provider.
+
+``` haskell
+type IdentityRoutes a =
+  "login" :> ( -- redirect User to the OpenID Provider
+              Get '[JSON] NoContent
+               -- render the page that will save the user creds in the user-agent
+              :<|> "cb" :> QueryParam "error" Text
+                        :> QueryParam "code" Text
+                        :> Get '[HTML] User)
+
+-- | gen a 302 redirect helper
+redirects :: (StringConv s ByteString) => s -> Handler ()
+redirects url = throwError err302 { errHeaders = [("Location",toS url)]}
+```
+
+That function will generate the URL to redirect the users to when
+they'll click on the login link: `https://yourdomain/login`.
+
+``` haskell
+genOIDCURL :: OIDCEnv -> IO ByteString
+genOIDCURL OIDCEnv{..} = do
+  st <- genState -- generate a random string
+  let oidcCreds = O.setCredentials clientId clientPassword redirectUri (O.newOIDC prov)
+  loc <- O.getAuthenticationRequestUrl oidcCreds [O.openId, O.email, O.profile] (Just st) []
+  return (show loc)
+
+handleLogin :: OIDCEnv -> Handler NoContent
+handleLogin oidcenv = do
+  loc <- liftIO (genOIDCURL oidcenv)
+  redirects loc
+  return NoContent
+```
+
+The `AuthInfo` is about the infos we can grab from OIDC provider.
+
+To be more precise, the user should come with a `code` (a token) and
+POSTing that code to the correct OIDC provider endpoint should return a JSON
+object. One of the fields should be named `id_token` which should be a
+JWT containing all the information we need. Depending on the scopes we
+asked we might get more information.
+
+``` haskell
+-- | @AuthInfo@
+data AuthInfo = AuthInfo { email         :: Text
+                         , emailVerified :: Bool
+                         , name :: Text } deriving (Eq, Show, Generic)
+
+instance FromJSON AuthInfo where
+  parseJSON (JSON.Object v) = do
+    email   :: Text <- v .: "email"
+    email_verified   :: Bool <- v .: "email_verified"
+    name :: Text <- v .: "name"
+    return $ AuthInfo (toS email) email_verified (toS name)
+  parseJSON invalid    = AeT.typeMismatch "Coord" invalid
+instance JSON.ToJSON AuthInfo where
+  toJSON (AuthInfo e ev n) =
+    JSON.object [ "email"   JSON..= (toS e :: Text)
+                , "email_verified"  JSON..= ev
+                , "name" JSON..= (toS n :: Text)
+                ]
+
+type LoginHandler = AuthInfo -> IO (Either Text User)
+```
+
+The `handleLoggedIn` is that part that will retrieve the information from
+the user once he is redirected from the OIDC Provider after login.
+
+If the user is redirected to the `redirect_uri` but with an `error` query
+parameter then it means something went wrong.
+If there is no error query param but a `code` query param it means the user
+successfully logged in. From there we need to make a request to the token
+endpoint of the OIDC provider. It's a POST that should contain the code
+as well as the client id and secret.
+Making this HTTP POST is the responsibility of `requestTokens`.
+
+From there we extract the `claims` of the JWT contained in one of the value
+of the JSON returned by the POST HTTP Request.
+
+``` haskell
+data User = User { userId          :: Text
+                 , userSecret      :: Text
+                 , localStorageKey :: Text
+                 , redirectUrl     :: Maybe Text
+                 } deriving (Show,Eq,Ord)
+
+handleLoggedIn :: OIDCEnv
+               -> LoginHandler -- ^ handle successful id
+               -> Maybe Text -- ^ error
+               -> Maybe Text -- ^ code
+               -> Handler User
+handleLoggedIn oidcenv handleSuccessfulId err mcode =
+  case err of
+    Just errorMsg -> forbidden errorMsg
+    Nothing -> case mcode of
+      Just oauthCode -> do
+        tokens <- liftIO $ O.requestTokens (oidc oidcenv) (toS oauthCode) (mgr oidcenv)
+        putText . show . O.claims . O.idToken $ tokens
+        let jwt = toS . unJwt . O.jwt . O.idToken $ tokens
+            eAuthInfo = decodeClaims jwt :: Either O.JwtError (O.JwtHeader,AuthInfo)
+        case eAuthInfo of
+          Left jwtErr -> forbidden $ "JWT decode/check problem: " <> show jwtErr
+          Right (_,authInfo) ->
+            if emailVerified authInfo
+              then do
+                user <- liftIO $ handleSuccessfulId authInfo
+                either forbidden return user
+              else forbidden "Please verify your email"
+      Nothing -> do
+        liftIO $ putText "No code param"
+        forbidden "no code parameter given"
+```
+
+When you render a User with blaze-html, it will generate a page with a js
+that will put a secret for that user in the local storage. And it will
+redirect the user to /.
+
+``` haskell
+instance ToMarkup User where
+  toMarkup User{..} = H.docTypeHtml $ do
+    H.head $
+      H.title "Logged In"
+    H.body $ do
+      H.h1 "Logged In"
+      H.p (H.toHtml ("Successful login with id " <> userId))
+      H.script (H.toHtml ("localStorage.setItem('" <> localStorageKey <> "','" <> userSecret <> "');"
+                          <> "localStorage.setItem('user-id','" <> userId <> "');"
+                         <> "window.location='" <> fromMaybe "/" redirectUrl <> "';" -- redirect the user to /
+                         ));
+
+serveOIDC :: OIDCEnv -> LoginHandler -> Server (IdentityRoutes a)
+serveOIDC oidcenv loginHandler =
+  handleLogin oidcenv :<|> handleLoggedIn oidcenv loginHandler
+
+-- * Auth
+type APIKey = ByteString
+type Account = Text.Text
+type Conf = [(APIKey,Account)]
+data Customer = Customer {
+  account :: Account
+  , apiKey :: APIKey
+  , mail :: Maybe Text
+  , fullname :: Maybe Text
+  }
+```
+
+Here is the code that displays the homepage.
+It should contain a link to the `/login` URL.
+When the user clicks on this link it will be redirected to Google login page
+with some generated information.
+
+The page also displays the content of the local storage.
+And in particular the items `api-key` and `user-id`.
+Those items should be set after a successful login when the user is redirected to
+`/login/cb`.
+
+The logic used generally is to use that api-key to uniquely identify an user.
+Another option would have been to set a cookie.
+
+``` haskell
+data Homepage = Homepage
+
+instance ToMarkup Homepage where
+  toMarkup Homepage = H.docTypeHtml $ do
+    H.head $ do
+      H.title "OpenID Connect Servant Example"
+      H.style (H.toHtml ("body { font-family: monospace; font-size: 18px; }" :: Text.Text))
+    H.body $ do
+      H.h1 "OpenID Connect Servant Example"
+      H.div $
+        H.a ! HA.href "/login" $ "Click here to login"
+      H.ul $ do
+        H.li $ do
+          H.span "API Key in Local storage: "
+          H.script (H.toHtml ("document.write(localStorage.getItem('api-key'));" :: Text.Text))
+        H.li $ do
+          H.span "User ID in Local storage: "
+          H.script (H.toHtml ("document.write(localStorage.getItem('user-id'));" :: Text.Text))
+```
+
+We need some helpers to generate random string for generating state and API Keys.
+
+``` haskell
+-- | generate a random ByteString, not necessarily extremely good randomness
+-- still the password will be long enough to be very difficult to crack
+genRandomBS :: IO ByteString
+genRandomBS = do
+  g <- Random.newStdGen
+  Random.randomRs (0, n) g & take 42 & fmap toChar & readable 0 & toS & return
+  where
+    n = length letters - 1
+    toChar i = letters List.!! i
+    letters = ['A'..'Z'] <> ['0'..'9'] <> ['a'..'z']
+    readable :: Int -> [Char] -> [Char]
+    readable _ [] = []
+    readable i str =
+      let blocksize = case n of
+            0 -> 8
+            1 -> 4
+            2 -> 4
+            3 -> 4
+            _ -> 12
+          block = take blocksize str
+          rest = drop blocksize str
+      in if List.null rest
+         then str
+         else block <> "-" <> readable (i+1) rest
+
+customerFromAuthInfo :: AuthInfo -> IO Customer
+customerFromAuthInfo authinfo = do
+  apikey <- genRandomBS
+  return Customer { account = toS (email authinfo)
+                  , apiKey = apikey
+                  , mail = Just (toS (email authinfo))
+                  , fullname = Just (toS (name authinfo))
+                  }
+
+handleOIDCLogin :: LoginHandler
+handleOIDCLogin authInfo = do
+  custInfo <- customerFromAuthInfo authInfo
+  if emailVerified authInfo
+    then return . Right . customerToUser $ custInfo
+    else return (Left "You emails is not verified by your provider. Please verify your email.")
+  where
+    customerToUser :: Customer -> User
+    customerToUser c =
+      User { userId = toS (account c)
+           , userSecret = toS (apiKey c)
+           , redirectUrl = Nothing
+           , localStorageKey = "api-key"
+           }
+```
+
+`Error` helpers
+---------------
+
+``` haskell
+data Err = Err { errTitle :: Text
+               , errMsg :: Text }
+
+instance ToMarkup Err where
+  toMarkup Err{..} = H.docTypeHtml $ do
+    H.head $ do
+      H.title "Error"
+    H.body $ do
+      H.h1 (H.a ! HA.href "/" $ "Home")
+      H.h2 (H.toHtml errTitle)
+      H.p (H.toHtml errMsg)
+
+format :: ToMarkup a => a -> LBS.ByteString
+format err = toMarkup err & renderMarkup
+
+appToErr :: ServerError -> Text -> ServerError
+appToErr x msg = x
+  { errBody = toS $ format (Err (toS (errReasonPhrase x)) msg)
+  , errHeaders =  [("Content-Type","text/html")]}
+
+unauthorized :: (MonadError ServerError m) => Text -> m a
+unauthorized = throwError . unauthorizedErr
+
+unauthorizedErr :: Text -> ServerError
+unauthorizedErr = appToErr err401
+
+forbidden :: (MonadError ServerError m) => Text -> m a
+forbidden = throwError . forbiddenErr
+
+forbiddenErr :: Text -> ServerError
+forbiddenErr = appToErr err403
+
+notFound :: ( MonadError ServerError m) => Text -> m a
+notFound = throwError . notFoundErr
+
+notFoundErr :: Text -> ServerError
+notFoundErr = appToErr err404
+
+preconditionFailed :: ( MonadError ServerError m) => Text -> m a
+preconditionFailed = throwError . preconditionFailedErr
+
+preconditionFailedErr :: Text -> ServerError
+preconditionFailedErr = appToErr err412
+
+serverError :: ( MonadError ServerError m) => Text -> m a
+serverError = throwError . serverErrorErr
+
+serverErrorErr :: Text -> ServerError
+serverErrorErr = appToErr err500
+```

doc/cookbook/pagination/Pagination.lhs

@@ -0,0 +1,250 @@
+# Pagination
+
+## Overview
+
+Let's see an approach to typed pagination with *Servant* using [servant-pagination](https://hackage.haskell.org/package/servant-pagination).
+
+This module offers opinionated helpers to declare a type-safe and a flexible pagination
+mechanism for Servant APIs. This design, inspired by [Heroku's API](https://devcenter.heroku.com/articles/platform-api-reference#ranges),
+provides a small framework to communicate about a possible pagination feature of an endpoint,
+enabling a client to consume the API in different fashions (pagination with offset / limit,
+endless scroll using last referenced resources, ascending and descending ordering, etc.)
+
+Therefore, client can provide a `Range` header with their request with the following format:
+
+- `Range: <field> [<value>][; offset <o>][; limit <l>][; order <asc|desc>]`
+
+For example: `Range: createdAt 2017-01-15T23:14:67.000Z; offset 5; order desc` indicates that
+the client is willing to retrieve the next batch of document in descending order that were
+created after the fifteenth of January, skipping the first 5.
+
+As a response, the server may return the list of corresponding documents, and augment the
+response with 3 headers:
+
+- `Accept-Ranges`: A comma-separated list of fields upon which a range can be defined
+- `Content-Range`: Actual range corresponding to the content being returned
+- `Next-Range`: Indicate what should be the next `Range` header in order to retrieve the next range
+
+For example:
+
+- `Accept-Ranges: createdAt, modifiedAt`
+- `Content-Range: createdAt 2017-01-15T23:14:51.000Z..2017-02-18T06:10:23.000Z`
+- `Next-Range: createdAt 2017-02-19T12:56:28.000Z; offset 0; limit 100; order desc`
+
+
+## Getting Started
+
+Code-wise the integration is quite seamless and unobtrusive. `servant-pagination` provides a
+`Ranges (fields :: [Symbol]) (resource :: *) -> *` data-type for declaring available ranges
+on a group of _fields_ and a target _resource_. To each combination (resource + field) is
+associated a given type `RangeType (resource :: *) (field :: Symbol) -> *` as described by
+the type-family in the `HasPagination` type-class.
+
+So, let's start with some imports and extensions to get this out of the way:
+
+``` haskell
+{-# LANGUAGE DataKinds             #-}
+{-# LANGUAGE DeriveGeneric         #-}
+{-# LANGUAGE FlexibleInstances     #-}
+{-# LANGUAGE MultiParamTypeClasses #-}
+{-# LANGUAGE TypeApplications      #-}
+{-# LANGUAGE TypeFamilies          #-}
+{-# LANGUAGE TypeOperators         #-}
+
+import           Data.Aeson
+                 (ToJSON, genericToJSON)
+import           Data.Maybe
+                 (fromMaybe)
+import           Data.Proxy
+                 (Proxy (..))
+import           GHC.Generics
+                 (Generic)
+import           Servant
+                 ((:>), GetPartialContent, Handler, Header, Headers, JSON, Server, addHeader)
+import           Servant.Pagination
+                 (HasPagination (..), PageHeaders, Range (..), Ranges, RangeOptions(..),
+                  applyRange, extractRange, returnRange)
+
+import qualified Data.Aeson               as Aeson
+import qualified Network.Wai.Handler.Warp as Warp
+import qualified Servant
+import qualified Servant.Pagination       as Pagination
+```
+
+
+#### Declaring the Resource
+
+Servant APIs are rather resource-oriented, and so is `servant-pagination`. This
+guide shows a basic example working with `JSON` (as you could tell from the
+import list already). To make the world a <span style='text-decoration:
+line-through'>better</span> colored place, let's create an API to retrieve
+colors -- with pagination.
+
+``` haskell
+data Color = Color
+  { name :: String
+  , rgb  :: [Int]
+  , hex  :: String
+  } deriving (Eq, Show, Generic)
+
+instance ToJSON Color where
+  toJSON =
+    genericToJSON Aeson.defaultOptions
+
+colors :: [Color]
+colors =
+  [ Color "Black" [0, 0, 0] "#000000"
+  , Color "Blue" [0, 0, 255] "#0000ff"
+  , Color "Green" [0, 128, 0] "#008000"
+  , Color "Grey" [128, 128, 128] "#808080"
+  , Color "Purple" [128, 0, 128] "#800080"
+  , Color "Red" [255, 0, 0] "#ff0000"
+  , Color "Yellow" [255, 255, 0] "#ffff00"
+  ]
+```
+
+#### Declaring the Ranges
+
+Now that we have defined our _resource_ (a.k.a `Color`), we are ready to declare a new `Range`
+that will operate on a "name" field (genuinely named after the `name` fields from the `Color`
+record).
+For that, we need to tell `servant-pagination` two things:
+
+- What is the type of the corresponding `Range` values
+- How do we get one of these values from our resource
+
+This is done via defining an instance of `HasPagination` as follows:
+
+``` haskell
+instance HasPagination Color "name" where
+  type RangeType Color "name" = String
+  getFieldValue _ = name
+  -- getRangeOptions :: Proxy "name" -> Proxy Color -> RangeOptions
+  -- getDefaultRange :: Proxy Color  -> Range "name" String
+
+defaultRange :: Range "name" String
+defaultRange =
+  getDefaultRange (Proxy @Color)
+```
+
+Note that `getFieldValue :: Proxy "name" -> Color -> String` is the minimal complete definition
+of the class. Yet, you can define `getRangeOptions` to provide different parsing options (see
+the last section of this guide). In the meantime, we've also defined a `defaultRange` as it will
+come in handy when defining our handler.
+
+#### API
+
+Good, we have a resource, we have a `Range` working on that resource, we can now declare our
+API using other Servant combinators we already know:
+
+``` haskell
+type API =
+  "colors"
+    :> Header "Range" (Ranges '["name"] Color)
+    :> GetPartialContent '[JSON] (Headers MyHeaders [Color])
+
+type MyHeaders =
+  Header "Total-Count" Int ': PageHeaders '["name"] Color
+```
+
+`PageHeaders` is a type alias provided by the library to declare the necessary response headers
+we mentioned in introduction. Expanding the alias boils down to the following:
+
+``` haskell
+-- type MyHeaders =
+--  '[ Header "Total-Count"   Int
+--   , Header "Accept-Ranges" (AcceptRanges '["name"])
+--   , Header "Content-Range" (ContentRange '["name"] Color)
+--   , Header "Next-Range"    (Ranges '["name"] Color)
+--   ]
+```
+
+As a result, we will need to provide all those headers with the response in our handler. Worry
+not, _servant-pagination_ provides an easy way to lift a collection of resources into such handler.
+
+#### Server
+
+Time to connect the last bits by defining the server implementation of our colorful API. The `Ranges`
+type we've defined above (tied to the `Range` HTTP header) indicates the server to parse any `Range`
+header, looking for the format defined in introduction with fields and target types we have just declared.
+If no such header is provided, we will end up receiving `Nothing`. Otherwise, it will be possible
+to _extract_ a `Range` from our `Ranges`.
+
+``` haskell
+server :: Server API
+server = handler
+  where
+    handler :: Maybe (Ranges '["name"] Color) -> Handler (Headers MyHeaders [Color])
+    handler mrange = do
+      let range =
+            fromMaybe defaultRange (mrange >>= extractRange)
+
+      addHeader (length colors) <$> returnRange range (applyRange range colors)
+
+main :: IO ()
+main =
+  Warp.run 1442 $ Servant.serve (Proxy @API) server
+```
+
+Let's try it out using different ranges to observe the server's behavior. As a reminder, here's
+the format we defined, where `<field>` here can only be `name` and `<value>` must parse to a `String`:
+
+- `Range: <field> [<value>][; offset <o>][; limit <l>][; order <asc|desc>]`
+
+Beside the target field, everything is pretty much optional in the `Range` HTTP header. Missing parts
+are deduced from the `RangeOptions` that are part of the `HasPagination` instance. Therefore, all
+following examples are valid requests to send to our server:
+
+- 1 - `curl http://localhost:1442/colors -vH 'Range: name'`
+- 2 - `curl http://localhost:1442/colors -vH 'Range: name; limit 2'`
+- 3 - `curl http://localhost:1442/colors -vH 'Range: name Green; order asc; offset 1'`
+
+Considering the following default options:
+
+- `defaultRangeLimit: 100`
+- `defaultRangeOffset: 0`
+- `defaultRangeOrder: RangeDesc`
+
+The previous ranges reads as follows:
+
+- 1 - The first 100 colors, ordered by descending names
+- 2 - The first 2 colors, ordered by descending names
+- 3 - The 100 colors after `Green` (not included), ordered by ascending names.
+
+
+## Going Forward
+
+####  Multiple Ranges
+
+Note that in the simple above scenario, there's no ambiguity with `extractRange` and `returnRange`
+because there's only one possible `Range` defined on our resource. Yet, as you've most probably
+noticed, the `Ranges` combinator accepts a list of fields, each of which must declare a `HasPagination`
+instance. Doing so will make the other helper functions more ambiguous and type annotations are
+highly likely to be needed.
+
+
+``` haskell
+instance HasPagination Color "hex" where
+  type RangeType Color "hex" = String
+  getFieldValue _ = hex
+
+-- to then define: Ranges '["name", "hex"] Color
+```
+
+
+#### Parsing Options
+
+By default, `servant-pagination` provides an implementation of `getRangeOptions` for each
+`HasPagination` instance. However, this can be overridden when defining the instance to provide
+your own options. These options come into play when a `Range` header is received and isn't fully
+specified (`limit`, `offset`, `order` are all optional) to provide default fallback values for those.
+
+For instance, let's say we wanted to change the default limit to `5` in a new range on
+`"rgb"`, we could tweak the corresponding `HasPagination` instance as follows:
+
+``` haskell
+instance HasPagination Color "rgb" where
+  type RangeType Color "rgb" = Int
+  getFieldValue _ = sum . rgb
+  getRangeOptions _ _ = Pagination.defaultOptions { defaultRangeLimit = 5 }
+```

doc/cookbook/pagination/pagination.cabal

@@ -0,0 +1,23 @@
+cabal-version:       2.2
+name:                cookbook-pagination
+version:             2.1
+synopsis:            Pagination with Servant example
+homepage:            http://docs.servant.dev/
+license:             BSD-3-Clause
+license-file:        ../../../servant/LICENSE
+author:              Servant Contributors
+maintainer:          haskell-servant-maintainers@googlegroups.com
+build-type:          Simple
+tested-with:         GHC==8.6.5, GHC==8.8.3, GHC ==8.10.7
+
+executable cookbook-pagination
+  main-is:             Pagination.lhs
+  build-tool-depends:  markdown-unlit:markdown-unlit
+  default-language:    Haskell2010
+  ghc-options:         -Wall -pgmL markdown-unlit
+  build-depends:       base >= 4.9 && <5
+                     , aeson
+                     , servant
+                     , servant-server
+                     , servant-pagination >= 2.1.0 && < 3.0.0
+                     , warp

doc/cookbook/sentry/Sentry.lhs

@@ -0,0 +1,116 @@
+# Error logging with Sentry
+
+In this recipe we will use [Sentry](https://sentry.io) to collect the runtime exceptions generated by our application. We will use the [raven-haskell](https://hackage.haskell.org/package/raven-haskell) package, which is a client for a Sentry event server. Mind that this package is not present on [Stackage](https://www.stackage.org/), so if we are using [Stack](https://docs.haskellstack.org) we’ll need to add it to our `extra-deps` section in the `stack.yaml` file.
+
+To exemplify this we will need the following imports
+
+```haskell
+{-# LANGUAGE DataKinds     #-}
+{-# LANGUAGE TypeOperators #-}
+
+import           Control.Exception                      (Exception,
+                                                         SomeException, throw)
+import           Data.ByteString.Char8                  (unpack)
+import           Network.Wai                            (Request, rawPathInfo,
+                                                         requestHeaderHost)
+import           Network.Wai.Handler.Warp               (defaultOnException,
+                                                         defaultSettings,
+                                                         runSettings,
+                                                         setOnException,
+                                                         setPort)
+import           Servant
+import           System.Log.Raven                       (initRaven, register,
+                                                         silentFallback)
+import           System.Log.Raven.Transport.HttpConduit (sendRecord)
+import           System.Log.Raven.Types                 (SentryLevel (Error),
+                                                         SentryRecord (..))
+```
+
+Just for the sake of the example we will use the following API which will throw an exception
+
+```haskell
+type API = "break" :> Get '[JSON] ()
+
+data MyException = MyException deriving (Show)
+
+instance Exception MyException
+
+server = breakHandler
+  where breakHandler :: Handler ()
+        breakHandler = do
+          throw MyException
+          return ()
+```
+
+First thing we need to do if we want to intercept and log this exception, we need to look in the section of our code where we run the `warp` application, and instead of using the simple `run` function from `warp`, we use the `runSettings` functions which allows to customise the handling of requests
+
+```haskell
+main :: IO ()
+main =
+  let
+    settings =
+      setPort 8080 $
+      setOnException sentryOnException $
+      defaultSettings
+  in
+    runSettings settings $ serve (Proxy :: Proxy API) server
+```
+
+The definition of the `sentryOnException` function could look as follows
+
+```haskell
+sentryOnException :: Maybe Request -> SomeException -> IO ()
+sentryOnException mRequest exception = do
+  sentryService <- initRaven
+    "https://username:password@senty.host/id"
+    id
+    sendRecord
+    silentFallback
+  register
+    sentryService
+    "myLogger"
+    Error
+    (formatMessage mRequest exception)
+    (recordUpdate mRequest exception)
+  defaultOnException mRequest exception
+```
+
+It does three things. First it initializes the service which will communicate with Sentry. The parameters it receives are:
+
+- the Sentry `DSN`, which is obtained when creating a new project on Sentry
+- a default way to update sentry fields, where we use the identity function
+- an event transport, which generally would be `sendRecord`, an HTTPS capable transport which uses http-conduit
+- a fallback handler, which we choose to be `silentFallback` since later we are logging to the console anyway.
+
+In the second step it actually sends our message to Sentry with the `register` function. Its arguments are:
+
+- the configured Sentry service which we just created
+- the name of the logger
+- the error level (see [SentryLevel](https://hackage.haskell.org/package/raven-haskell/docs/System-Log-Raven-Types.html#t:SentryLevel) for the possible options)
+- the message we want to send
+- an update function to handle the specific `SentryRecord`
+
+Eventually it just delegates the error handling to the default warp mechanism.
+
+The function `formatMessage` simply uses the request and the exception to return a string with the error message.
+
+```haskell
+formatMessage :: Maybe Request -> SomeException -> String
+formatMessage Nothing exception        = "Exception before request could be parsed: " ++ show exception
+formatMessage (Just request) exception = "Exception " ++ show exception ++ " while handling request " ++ show request
+```
+
+The only piece left now is the `recordUpdate` function which allows to decorate with other [attributes](https://docs.sentry.io/clientdev/attributes/) the default `SentryRecord`.
+
+```haskell
+recordUpdate :: Maybe Request -> SomeException -> SentryRecord -> SentryRecord
+recordUpdate Nothing exception record        = record
+recordUpdate (Just request) exception record = record
+  { srCulprit = Just $ unpack $ rawPathInfo request
+  , srServerName = fmap unpack $ requestHeaderHost request
+  }
+```
+
+In this examples we set the raw path as the culprit and we use the `Host` header to populate the server name field.
+
+You can try to run this code using the `cookbook-sentry` executable. You should obtain a `MyException` error in the console and, if you provided a valid Sentry DSN, you should also find your error in the Sentry interface.

doc/cookbook/sentry/sentry.cabal

@@ -0,0 +1,24 @@
+cabal-version:       2.2
+name:                cookbook-sentry
+version:             0.1
+synopsis:            Collecting runtime exceptions using Sentry
+homepage:            http://docs.servant.dev/
+license:             BSD-3-Clause
+license-file:        ../../../servant/LICENSE
+author:              Servant Contributors
+maintainer:          haskell-servant-maintainers@googlegroups.com
+build-type:          Simple
+tested-with:         GHC==8.6.5, GHC==8.8.3, GHC ==8.10.7
+
+executable cookbook-sentry
+  main-is:             Sentry.lhs
+  build-depends:       base == 4.*
+                     , bytestring
+                     , markdown-unlit >= 0.4
+                     , raven-haskell >= 0.1.2
+                     , servant-server
+                     , warp >= 3.2
+                     , wai >= 3.2
+  default-language:    Haskell2010
+  ghc-options:         -Wall -pgmL markdown-unlit
+  build-tool-depends: markdown-unlit:markdown-unlit

doc/cookbook/structuring-apis/StructuringApis.lhs

@@ -144,7 +144,7 @@ simpleAPIServer
   :: m [a]
   -> (i -> m a)
   -> (a -> m NoContent)
-  -> Server (SimpleAPI name a i) m
+  -> ServerT (SimpleAPI name a i) m
 simpleAPIServer listAs getA postA =
   listAs :<|> getA :<|> postA
 

doc/cookbook/structuring-apis/structuring-apis.cabal

@@ -1,14 +1,14 @@
+cabal-version:       2.2
 name:                cookbook-structuring-apis
 version:             0.1
 synopsis:            Example that shows how APIs can be structured
-homepage:            http://haskell-servant.readthedocs.org/
-license:             BSD3
+homepage:            http://docs.servant.dev/
+license:             BSD-3-Clause
 license-file:        ../../../servant/LICENSE
 author:              Servant Contributors
 maintainer:          haskell-servant-maintainers@googlegroups.com
 build-type:          Simple
-cabal-version:       >=1.10
-tested-with:         GHC==7.8.4, GHC==7.10.3, GHC==8.0.2, GHC==8.2.2
+tested-with:         GHC==8.6.5, GHC==8.8.3, GHC ==8.10.7
 
 executable cookbook-structuring-apis
   main-is:             StructuringApis.lhs

doc/cookbook/testing/Testing.lhs

@@ -0,0 +1,506 @@
+# How To Test Servant Applications
+
+Even with a nicely structured API that passes Haskell's strict type checker,
+it's a good idea to write some tests for your application.
+
+In this recipe we'll work through some common testing strategies and provide
+examples of utlizing these testing strategies in order to test Servant
+applications.
+
+## Testing strategies
+
+There are many testing strategies you may wish to employ when testing your
+Servant application, but included below are three common testing patterns:
+
+- We'll use `servant-client` to derive client functions and then send valid
+requests to our API, running in another thread. This is great for testing
+that our **business logic** is correctly implemented with only valid HTTP
+requests.
+
+- We'll also use `hspec-wai` to make **arbitrary HTTP requests**, in order to
+test how our application may respond to invalid or otherwise unexpected
+requests.
+
+- Finally, we can also use `servant-quickcheck` for **whole-API tests**, in order
+to assert that our entire application conforms to **best practices**.
+
+## Useful Libraries
+
+The following libraries will often come in handy when we decide to test our
+Servant applications:
+
+- [hspec](https://hspec.github.io/)
+- [hspec-wai](http://hackage.haskell.org/package/hspec-wai)
+- [QuickCheck](http://hackage.haskell.org/package/QuickCheck)
+- [servant-quickcheck](https://hackage.haskell.org/package/servant-quickcheck)
+
+## Imports and Our Testing Module
+
+This recipe starts with the following ingredients:
+
+```haskell
+{-# LANGUAGE OverloadedStrings, TypeFamilies, DataKinds,
+  DeriveGeneric, TypeOperators #-}
+import           Prelude ()
+import           Prelude.Compat
+
+import qualified Control.Concurrent               as C
+import           Control.Concurrent.MVar
+import           Control.Exception                (bracket)
+import           Control.Lens              hiding (Context)
+import           Data.Aeson
+import           Data.Aeson.Lens
+import qualified Data.HashMap.Strict              as HM
+import           Data.Text                        (Text, unpack)
+import           GHC.Generics
+import           Network.HTTP.Client       hiding (Proxy)
+import           Network.HTTP.Types
+import           Network.Wai
+import qualified Network.Wai.Handler.Warp         as Warp
+
+import           Servant
+import           Servant.Client
+import           Servant.Server
+import           Servant.QuickCheck
+import           Servant.QuickCheck.Internal (serverDoesntSatisfy)
+
+import           Test.Hspec
+import           Test.Hspec.Wai
+import           Test.Hspec.Wai.Matcher
+```
+
+We're going to produce different `Spec`s that represent different
+aspects of our application, and we'll ask `hspec` to run all of our different
+`Spec`s. This is a common organizational method for testing modules:
+
+```haskell
+spec :: Spec
+spec = do
+  businessLogicSpec
+  thirdPartyResourcesSpec
+  servantQuickcheckSpec
+```
+
+Often, codebases will use `hspec`'s
+[autodiscover pragma](http://hspec.github.io/hspec-discover.html)
+to find all testing modules and `Spec`s inside, but we're going to
+explicitly make a `main` function to run our tests because we have only one
+`spec` defined above:
+
+```haskell
+main :: IO ()
+main = hspec spec
+```
+
+## Testing Your Business Logic
+
+Let's say we have an API that looks something like this:
+
+```haskell
+data User = User {
+  name :: Text
+  , user_id :: Integer
+  } deriving (Eq, Show, Generic)
+
+instance FromJSON User
+instance ToJSON User
+
+type UserApi =
+  -- One endpoint: create a user
+  "user" :> Capture "userId" Integer :> Post '[JSON] User
+```
+
+A real server would likely use a database to store, retrieve, and validate
+users, but we're going to do something really simple merely to have something
+to test. With that said, here's a sample handler, server, and `Application`
+for the endpoint described above:
+
+```haskell
+userApp :: Application
+userApp = serve (Proxy :: Proxy UserApi) userServer
+
+userServer :: Server UserApi
+userServer = createUser
+
+createUser :: Integer -> Handler User
+createUser userId = do
+  if userId > 5000
+    then pure $ User { name = "some user", user_id = userId }
+    else throwError $ err400 { errBody = "userId is too small" }
+```
+
+### Strategy 1: Spin Up a Server, Create a Client, Make Some Requests
+
+One of the benefits of Servant's type-level DSL for describing APIs is that
+once you have provided a type-level description of your API, you can create
+clients, documentation, or other tools for it somewhat magically.
+
+In this case, we'd like to *test* our server, so we can use `servant-client`
+to create a client, after which we'll run our server, and then make requests
+of it and see how it responds.
+
+Let's write some tests:
+
+```haskell
+withUserApp :: (Warp.Port -> IO ()) -> IO ()
+withUserApp action =
+  -- testWithApplication makes sure the action is executed after the server has
+  -- started and is being properly shutdown.
+  Warp.testWithApplication (pure userApp) action
+
+
+businessLogicSpec :: Spec
+businessLogicSpec =
+  -- `around` will start our Server before the tests and turn it off after
+  around withUserApp $ do
+    -- create a test client function
+    let createUser = client (Proxy :: Proxy UserApi)
+    -- create a servant-client ClientEnv
+    baseUrl <- runIO $ parseBaseUrl "http://localhost"
+    manager <- runIO $ newManager defaultManagerSettings
+    let clientEnv port = mkClientEnv manager (baseUrl { baseUrlPort = port })
+
+    -- testing scenarios start here
+    describe "POST /user" $ do
+      it "should create a user with a high enough ID" $ \port -> do
+        result <- runClientM (createUser 50001) (clientEnv port)
+        result `shouldBe` (Right $ User { name = "some user", user_id = 50001})
+      it "will it fail with a too-small ID?" $ \port -> do
+        result <- runClientM (createUser 4999) (clientEnv port)
+        result `shouldBe` (Right $ User { name = "some user", user_id = 50001})
+```
+
+### Running These Tests
+
+Let's run our tests and see what happens:
+
+```
+$ cabal new-test all
+POST /user
+  should create a user with a high enough ID
+  should fail with a too-small ID FAILED [1]
+
+  Failures:
+
+  Testing.lhs:129:7:
+  1) POST /user should fail with a too-small ID
+       expected: Right (User {name = "some user", user_id = 50001})
+        but got: Left (FailureResponse (Response {responseStatusCode = Status {statusCode = 400, statusMessage = "Bad Request"}, responseHeaders = fromList [("Transfer-Encoding","chunked"),("Date","Fri, 12 Oct 2018 04:36:22 GMT"),("Server","Warp/3.2.25")], responseHttpVersion = HTTP/1.1, responseBody = "userId is too small"}))
+
+  To rerun use: --match "/POST /user/should fail with a too-small ID/"
+```
+
+Hmm. One passed and one failed! It looks like I *was* expecting a success
+response in the second test, but I actually got a failure. We should fix that,
+but first I'd like to introduce `hspec-wai`, which will give us different
+mechanisms for making requests of our application and validating the responses
+we get. We're also going to spin up a fake Elasticsearch server, so that our
+server can think it's talking to a real database.
+
+
+## *Mocking* 3rd Party Resources
+
+Often our web applications will need to make their own web
+requests to other 3rd-party applications. These requests provide a lot
+of opportunity for failure and so we'd like to test that the right
+messages and failure values (in addition to success values) are returned
+from our application.
+
+### Define the 3rd-Party Resource
+
+With Servant's type-level API definitions, assuming you've already defined the
+API you want to mock, it's relatively trivial to create a simple server for
+the purposes of running tests. For instance, consider an API server that needs
+to get data out of Elasticsearch. Let's first define the Elasticsearch server
+and client using Servant API descriptions:
+
+```haskell
+type SearchAPI =
+  -- We're using Aeson's Generic JSON `Value` to make things easier on
+  -- ourselves. We're also representing only one Elasticsearch endpoint:
+  -- get item by id
+  "myIndex" :> "myDocType" :> Capture "docId" Integer :> Get '[JSON] Value
+
+-- Here's our Servant Client function
+getDocument = client (Proxy :: Proxy SearchAPI)
+
+-- We can use these helpers when we want to make requests
+-- using our client function
+clientEnv :: Text -> Text -> IO ClientEnv
+clientEnv esHost esPort = do
+  baseUrl <- parseBaseUrl $ unpack $ esHost <> ":" <> esPort
+  manager <- newManager defaultManagerSettings
+  pure $ mkClientEnv manager baseUrl
+
+runSearchClient :: Text -> Text -> ClientM a -> IO (Either ClientError a)
+runSearchClient esHost esPort = (clientEnv esHost esPort >>=) . runClientM
+```
+
+### Servant Server Example Using this 3rd-Party Resource
+
+So we've got an Elasticsearch server and a client to talk to it. Let's now
+build a simple app server that uses this client to retrieve documents. This
+is somewhat contrived, but hopefully it illustrates the typical three-tier
+application architecture.
+
+One note: we're also going to take advantage of `lens-aeson` here, which may
+look a bit foreign. The gist of it is that we're going to traverse a JSON
+`Value` from Elasticsearch and try to extract some  kind of document to
+return.
+
+Imagine, then, that this is our real server implementation:
+
+```haskell
+type DocApi =
+  "docs" :> Capture "docId" Integer :> Get '[JSON] Value
+
+docsApp :: Text -> Text -> Application
+docsApp esHost esPort = serve (Proxy :: Proxy DocApi) $ docServer esHost esPort
+
+docServer :: Text -> Text -> Server DocApi
+docServer esHost esPort = getDocById esHost esPort
+
+-- Our Handler tries to get a doc from Elasticsearch and then tries to parse
+-- it. Unfortunately, there's a lot of opportunity for failure in these
+-- actions
+getDocById :: Text -> Text -> Integer -> Handler Value
+getDocById esHost esPort docId = do
+  -- Our Servant Client function returns Either ClientError Value here:
+  docRes <- liftIO $ runSearchClient esHost esPort (getDocument docId)
+  case docRes of
+    Left err -> throwError $ err404 { errBody = "Failed looking up content" }
+    Right value -> do
+      -- we'll either fail to parse our document or we'll return it
+      case value ^? _Object . ix "_source" of
+        Nothing -> throwError $ err400 { errBody = "Failed parsing content" }
+        Just obj -> pure obj
+```
+
+### Testing Our Backend
+
+So the above represents our application and is close to a server we may
+actually deploy. How then shall we test this application?
+
+Ideally, we'd like it to make requests of a *real* Elasticsearch server, but
+we certainly don't want our tests to trigger requests to a live, production
+database. In addition, we don't want to depend on our real Elasticsearch
+server having specific, consistent results for us to test against, because
+that would make our tests flaky (and flaky tests are sometimes described as
+worse than not having tests at all).
+
+One solution to this is to create a trivial Elasticsearch server as part of
+our testing code. We can do this relatively easily because we already have
+an API definition for it above. With a *real* server, we can then let our own
+application make requests of it and we'll simulate different scenarios in
+order to make sure our application responds the way we expect it to.
+
+Let's start with some helpers which will allow us to run a testing version
+of our Elasticsearch server in another thread:
+
+```haskell
+-- | We'll run the Elasticsearch server so we can test behaviors
+withElasticsearch :: IO () -> IO ()
+withElasticsearch action =
+  bracket (liftIO $ C.forkIO $ Warp.run 9999 esTestApp)
+    C.killThread
+    (const action)
+
+esTestApp :: Application
+esTestApp = serve (Proxy :: Proxy SearchAPI) esTestServer
+
+esTestServer :: Server SearchAPI
+esTestServer = getESDocument
+
+-- This is the *mock* handler we're going to use. We create it
+-- here specifically to trigger different behavior in our tests.
+getESDocument :: Integer -> Handler Value
+getESDocument docId
+  -- arbitrary things we can use in our tests to simulate failure:
+  -- we want to trigger different code paths.
+  | docId > 1000 = throwError err500
+  | docId > 500 = pure . Object $ HM.fromList [("bad", String "data")]
+  | otherwise = pure $ Object $ HM.fromList [("_source", Object $ HM.fromList [("a", String "b")])]
+```
+
+Now, we should be ready to write some tests.
+
+In this case, we're going to use `hspec-wai`, which will give us a simple way
+to run our application, make requests, and make assertions against the
+responses we receive.
+
+Hopefully, this will simplify our testing code:
+
+```haskell
+thirdPartyResourcesSpec :: Spec
+thirdPartyResourcesSpec = around_ withElasticsearch $ do
+  -- we call `with` from `hspec-wai` and pass *real* `Application`
+  with (pure $ docsApp "localhost" "9999") $ do
+    describe "GET /docs" $ do
+      it "should be able to get a document" $
+        -- `get` is a function from hspec-wai`.
+        get "/docs/1" `shouldRespondWith` 200
+      it "should be able to handle connection failures" $
+        get "/docs/1001" `shouldRespondWith` 404
+      it "should be able to handle parsing failures" $
+        get "/docs/501" `shouldRespondWith` 400
+      it "should be able to handle odd HTTP requests" $
+        -- we can also make all kinds of arbitrary custom requests to see how
+        -- our server responds using the `request` function:
+        -- request :: Method -> ByteString -> [Header]
+        --   -> LB.ByteString -> WaiSession SResponse
+        request methodPost "/docs/501" [] "{" `shouldRespondWith` 405
+      it "we can also do more with the Response using hspec-wai's matchers" $
+        -- see also `MatchHeader` and JSON-matching tools as well...
+        get "/docs/1" `shouldRespondWith` 200 { matchBody = MatchBody bodyMatcher }
+
+bodyMatcher :: [Network.HTTP.Types.Header] -> Body -> Maybe String
+bodyMatcher _ body = case (decode body :: Maybe Value) of
+  -- success in this case means we return `Nothing`
+  Just val | val == (Object $ HM.fromList [("a", String "b")]) -> Nothing
+  _ -> Just "This is how we represent failure: this message will be printed"
+```
+
+Out of the box, `hspec-wai` provides a lot of useful tools for us to run tests
+against our application. What happens when we run these tests?
+
+```
+$ cabal new-test all
+...
+
+GET /docs
+  should be able to get a document
+  should be able to handle connection failures
+  should be able to handle parsing failures
+  should be able to handle odd HTTP requests
+  we can also do more with the Response using hspec-wai's matchers
+```
+
+Fortunately, they all passed! Let's move to another strategy: whole-API
+testing.
+
+## Servant Quickcheck
+
+[`servant-quickcheck`](https://github.com/haskell-servant/servant-quickcheck)
+is a project that allows users to write tests for whole Servant APIs using
+quickcheck-style property-checking mechanisms.
+
+`servant-quickcheck` is great for asserting API-wide rules, such as "no
+endpoint throws a 500" or "all 301 status codes also come with a Location
+header". The project even comes with a number of predicates that reference
+the [RFCs they originate from](https://github.com/haskell-servant/servant-quickcheck/blob/master/src/Servant/QuickCheck/Internal/Predicates.hs).
+
+In other words, it's one way to assert that your APIs conform to specs and
+best practices.
+
+### Quickcheckable API
+
+Let's make an API and a server to demonstrate how to use `servant-quickcheck`:
+
+```haskell
+type API = ReqBody '[JSON] String :> Post '[JSON] String
+      :<|> Get '[JSON] Int
+      :<|> BasicAuth "some-realm" () :> Get '[JSON] ()
+
+api :: Proxy API
+api = Proxy
+
+server :: IO (Server API)
+server = do
+    mvar <- newMVar ""
+    return $ (\x -> liftIO $ swapMVar mvar x)
+        :<|> (liftIO $ readMVar mvar >>= return . length)
+        :<|> (const $ return ())
+```
+
+### Using `servant-quickcheck`
+
+Let's build some tests for our API using `servant-quickcheck`.
+
+Similar to the above examples, we're going to create `Spec`s, but in this
+case, we'll rely on a number of predicates available from `servant-quickcheck`
+to see if our API server conforms to best practices:
+
+```haskell
+-- Let's set some QuickCheck values
+args :: Args
+args = defaultArgs { maxSuccess = 500 }
+
+-- Here's a Servant Context object we'll use
+ctx :: Context '[BasicAuthCheck ()]
+ctx = BasicAuthCheck (const . return $ NoSuchUser) :. EmptyContext
+
+
+servantQuickcheckSpec :: Spec
+servantQuickcheckSpec = describe "" $ do
+  it "API demonstrates best practices" $
+    -- `withServerServer` and `withServantServerAndContext` come from `servant-quickcheck`
+    withServantServerAndContext api ctx server $ \burl ->
+      -- `serverSatisfies` and the predicates also come from `servant-quickcheck`
+      serverSatisfies api burl args (unauthorizedContainsWWWAuthenticate
+                                 <%> not500
+                                 <%> onlyJsonObjects  -- this one isn't true!
+                                 <%> mempty)
+
+  it "API doesn't have these things implemented yet" $
+    withServantServerAndContext api ctx server $ \burl -> do
+      serverDoesntSatisfy api burl args (getsHaveCacheControlHeader
+                                    <%> notAllowedContainsAllowHeader
+                                    <%> mempty)
+```
+
+Let's see what happens when we run these tests:
+
+
+```
+  API demonstrates best practices FAILED [2]
++++ OK, passed 500 tests.
+  API doesn't have these things implemented yet
+
+  src/Servant/QuickCheck/Internal/QuickCheck.hs:143:11:
+  2) Main[339:25] API demonstrates best practices
+       Failed:
+       Just Predicate failed
+            Predicate: onlyJsonObjects
+
+            Response:
+                 Status code: 200
+                 Headers:  "Transfer-Encoding": "chunked"
+                           "Date": "Fri, 12 Oct 2018 04:36:22 GMT"
+                           "Server": "Warp/3.2.25"
+                           "Content-Type": "application/json;charset=utf-8"
+                 Body: ""
+
+  To rerun use: --match "/Main[339:25]/API demonstrates best practices/"
+
+Randomized with seed 1046277487
+
+Finished in 0.4306 seconds
+```
+
+Hmm. It looks like we *thought* our API only returned JSON objects, which is a
+best practice, but in fact, we *did* have an endpoint that returned an empty
+body, which you can see in the printed response above: `Body: ""`. We should
+consider revising our API to only return top-level JSON Objects in the future!
+
+### Other Cool Things
+
+`servant-quickcheck` also has a cool mechanism where you can compare two API
+servers to demonstrate that they respond identically to requests. This may be
+useful if you are planning to rewrite one API in another language or with
+another web framework. You have to specify whether you're looking for
+`jsonEquality` vs regular `ByteString` equality, though.
+
+## Conclusion
+
+There are lots of techniques for testing and we only covered a few here.
+
+Useful libraries such as `hspec-wai` have ways of running Wai `Application`s
+and sending requests to them, while Servant's type-level DSL for defining APIs
+allows us to more easily mock out servers and to derive clients, which will
+only craft valid requests.
+
+Lastly, if you want a broad overview of where your application fits in with
+regard to best practices, consider using `servant-quickcheck`.
+
+This program is available as a cabal project
+[here](https://github.com/haskell-servant/servant/tree/master/doc/cookbook/testing).

doc/cookbook/testing/testing.cabal

@@ -0,0 +1,38 @@
+cabal-version:       2.2
+name:                cookbook-testing
+version:             0.0.1
+synopsis:            Common testing patterns in Servant apps
+description:         This recipe includes various strategies for writing tests for Servant.
+homepage:            http://docs.servant.dev/
+license:             BSD-3-Clause
+license-file:        ../../../servant/LICENSE
+author:              Servant Contributors
+maintainer:          haskell-servant-maintainers@googlegroups.com
+category:            Servant
+build-type:          Simple
+tested-with:         GHC==8.6.5, GHC==8.8.3, GHC ==8.10.7
+
+executable cookbook-testing
+  main-is:             Testing.lhs
+  build-depends:       base == 4.*
+                     , base-compat
+                     , text >= 1.2
+                     , aeson >= 1.2
+                     , lens-aeson
+                     , lens
+                     , servant
+                     , servant-client
+                     , servant-server
+                     , servant-quickcheck >= 0.0.10
+                     , http-client
+                     , http-types >= 0.12
+                     , hspec
+                     , hspec-wai
+                     , QuickCheck
+                     , unordered-containers
+                     , warp >= 3.2
+                     , wai >= 3.2
+                     , wai-extra
+  default-language:    Haskell2010
+  ghc-options:         -Wall -pgmL markdown-unlit
+  build-tool-depends:  markdown-unlit:markdown-unlit

doc/cookbook/using-custom-monad/UsingCustomMonad.lhs

@@ -1,6 +1,6 @@
 # Using a custom monad
 
-In this section we will create and API for a book shelf without any backing DB storage.
+In this section we will create an API for a book shelf without any backing DB storage.
 We will keep state in memory and share it between requests using `Reader` monad and `STM`.
 
 We start with a pretty standard set of imports and definition of the model:
@@ -115,3 +115,6 @@ Running cookbook-using-custom-monad...
 [Book "To Kill a Mockingbird",Book "Harry Potter and the Order of the Phoenix"]
 [Book "The Picture of Dorian Gray",Book "To Kill a Mockingbird",Book "Harry Potter and the Order of the Phoenix"]
 ```
+
+To use `Raw` endpoints, look at the
+[servant-rawm](http://hackage.haskell.org/package/servant-rawm) package.

doc/cookbook/using-custom-monad/using-custom-monad.cabal

@@ -1,14 +1,14 @@
+cabal-version:       2.2
 name:                cookbook-using-custom-monad
 version:             0.1
 synopsis:            Using custom monad to pass a state between handlers
-homepage:            http://haskell-servant.readthedocs.org/
-license:             BSD3
+homepage:            http://docs.servant.dev/
+license:             BSD-3-Clause
 license-file:        ../../../servant/LICENSE
 author:              Servant Contributors
 maintainer:          haskell-servant-maintainers@googlegroups.com
 build-type:          Simple
-cabal-version:       >=1.10
-tested-with:         GHC==7.8.4, GHC==7.10.3, GHC==8.0.2, GHC==8.2.2
+tested-with:         GHC==8.6.5, GHC==8.8.3, GHC ==8.10.7
 
 executable cookbook-using-custom-monad
   main-is:             UsingCustomMonad.lhs

doc/cookbook/using-free-client/UsingFreeClient.lhs

@@ -0,0 +1,192 @@
+# Inspecting, debugging, simulating clients and more
+
+or simply put: _a practical introduction to `Servant.Client.Free`_.
+
+Someone asked on IRC how one could access the intermediate Requests (resp. Responses)
+produced (resp. received) by client functions derived using servant-client.
+My response to such inquiries is: to extend `servant-client` in an ad-hoc way (e.g for testing or debugging
+purposes), use `Servant.Client.Free`. This recipe shows how.
+
+First the module header, but this time We'll comment the imports.
+
+```haskell
+{-# LANGUAGE DataKinds     #-}
+{-# LANGUAGE TypeOperators #-}
+module Main (main) where
+```
+
+We will primarily use `Servant.Client.Free`, it doesn't re-export anything
+from `free` package, so we need to import it as well.
+
+```haskell
+import Control.Monad.Free
+import Servant.Client.Free
+```
+
+Also we'll use `servant-client` internals, which uses `http-client`,
+so let's import them *qualified*
+
+```haskell
+import qualified Servant.Client.Internal.HttpClient as I
+import qualified Network.HTTP.Client                as HTTP
+```
+
+The rest of the imports are for a server we implement here for completeness.
+
+```haskell
+import Servant
+import Network.Wai.Handler.Warp (run)
+import System.Environment       (getArgs)
+```
+
+## API & Main
+
+We'll work with a very simple API:
+
+```haskell
+type API = "square" :> Capture "n" Int :> Get '[JSON] Int
+
+api :: Proxy API
+api = Proxy
+```
+
+Next we implement a `main`. If passed `"server"` it will run `server`, if passed
+`"client"` it will run a `test` function (to be defined next). This should be
+pretty straightforward:
+
+```haskell
+main :: IO ()
+main = do
+    args <- getArgs
+    case args of
+        ("server":_) -> do
+            putStrLn "Starting cookbook-using-free-client at http://localhost:8000"
+            run 8000 $ serve api $ \n -> return (n * n)
+        ("client":_) ->
+            test
+        _ -> do
+            putStrLn "Try:"
+            putStrLn "cabal new-run cookbook-using-free-client server"
+            putStrLn "cabal new-run cookbook-using-free-client client"
+```
+
+## Test
+
+In the client part, we will use a `Servant.Client.Free` client.
+Because we have a single endpoint API, we'll get a single client function,
+running in the `Free ClientF` (free) monad:
+
+```haskell
+getSquare :: Int -> Free ClientF Int
+getSquare = client api
+```
+
+Such clients are "client functions without a backend", so to speak,
+or where the backend has been abstracted out. To be more precise, `ClientF` is a functor that
+precisely represents the operations servant-client-core needs from an http client backend.
+So if we are to emulate one or augment what such a backend does, it will be by interpreting
+all those operations, the way we want to. This also means we get access to the requests and
+responses and can do anything we want with them, right when they are produced or consumed,
+respectively.
+
+Next, we can write our small test. We'll pass a value to `getSquare` and inspect
+the `Free` structure. The first three possibilities are self-explanatory:
+
+```haskell
+test :: IO ()
+test = case getSquare 42 of
+    Pure n ->
+        putStrLn $ "ERROR: got pure result: " ++ show n
+    Free (Throw err) ->
+        putStrLn $ "ERROR: got error right away: " ++ show err
+```
+
+We are interested in `RunRequest`, that's what client should block on:
+
+```haskell
+    Free (RunRequest req k) -> do
+```
+
+Then we need to prepare the context, get HTTP (connection) `Manager`
+and `BaseUrl`:
+
+```haskell
+        burl <- parseBaseUrl "http://localhost:8000"
+        mgr <- HTTP.newManager HTTP.defaultManagerSettings
+```
+
+Now we can use `servant-client`'s internals to convert servant's `Request`
+to http-client's `Request`, and we can inspect it:
+
+```haskell
+        req' <- I.defaultMakeClientRequest burl req
+        putStrLn $ "Making request: " ++ show req'
+```
+
+`servant-client`'s request does a little more, but this is good enough for
+our demo. We get back an http-client `Response` which we can also inspect.
+
+```haskell
+        res' <- HTTP.httpLbs req' mgr
+        putStrLn $ "Got response: " ++ show res'
+```
+
+And we continue by turning http-client's `Response` into servant's `Response`,
+and calling the continuation. We should get a `Pure` value.
+
+```haskell
+        let res = I.clientResponseToResponse id res'
+
+        case k res of
+            Pure n ->
+                putStrLn $ "Expected 1764, got " ++ show n
+            _ ->
+                putStrLn "ERROR: didn't get a response"
+```
+
+So that's it. Using `Free` we can evaluate servant clients step-by-step, and
+validate that the client functions or the HTTP client backend does what we expect
+(e.g by printing requests/responses on the fly). In fact, using `Servant.Client.Free`
+is a little simpler than defining a custom `RunClient` instance, especially
+for those cases where it is handy to have the full sequence of client calls
+and responses available for us to inspect, since `RunClient` only gives us
+access to one `Request` or `Response` at a time.
+
+On the other hand, a "batch collection" of requests and/or responses can be achieved
+with both free clients and a custom `RunClient` instance rather easily, for example
+by using a `Writer [(Request, Response)]` monad.
+
+Here is an example of running our small `test` against a running server:
+
+```
+Making request: Request {
+  host                 = "localhost"
+  port                 = 8000
+  secure               = False
+  requestHeaders       = [("Accept","application/json;charset=utf-8,application/json")]
+  path                 = "/square/42"
+  queryString          = ""
+  method               = "GET"
+  proxy                = Nothing
+  rawBody              = False
+  redirectCount        = 10
+  responseTimeout      = ResponseTimeoutDefault
+  requestVersion       = HTTP/1.1
+}
+
+Got response: Response
+  { responseStatus = Status {statusCode = 200, statusMessage = "OK"}
+  , responseVersion = HTTP/1.1
+  , responseHeaders =
+    [ ("Transfer-Encoding","chunked")
+    , ("Date","Thu, 05 Jul 2018 21:12:41 GMT")
+    , ("Server","Warp/3.2.22")
+    , ("Content-Type","application/json;charset=utf-8")
+    ]
+  , responseBody = "1764"
+  , responseCookieJar = CJ {expose = []}
+  , responseClose' = ResponseClose
+  }
+
+Expected 1764, got 1764
+```

doc/cookbook/using-free-client/using-free-client.cabal

@@ -0,0 +1,26 @@
+cabal-version:       2.2
+name:                cookbook-using-free-client
+version:             0.1
+synopsis:            Using Free client
+homepage:            http://docs.servant.dev/
+license:             BSD-3-Clause
+license-file:        ../../../servant/LICENSE
+author:              Servant Contributors
+maintainer:          haskell-servant-maintainers@googlegroups.com
+build-type:          Simple
+tested-with:         GHC==8.6.5, GHC==8.8.3, GHC ==8.10.7
+
+executable cookbook-using-free-client
+  main-is:             UsingFreeClient.lhs
+  build-depends:       base >= 4.9 && <5
+                     , free
+                     , servant
+                     , servant-client
+                     , http-client
+                     , servant-client-core
+                     , base-compat
+                     , servant-server
+                     , warp >= 3.2
+  default-language:    Haskell2010
+  ghc-options:         -Wall -pgmL markdown-unlit
+  build-tool-depends: markdown-unlit:markdown-unlit >= 0.4

doc/cookbook/uverb/UVerb.lhs

@@ -0,0 +1,223 @@
+# Listing alternative responses and exceptions in your API types
+
+Servant allows you to talk about the exceptions you throw in your API
+types.  This is not limited to actual exceptions, you can write
+handlers that respond with arbitrary open unions of types.
+
+## Compatibility
+
+:warning: This cookbook is compatible with GHC 8.6.1 or higher :warning:
+
+## Preliminaries
+
+```haskell
+{-# LANGUAGE ConstraintKinds #-}
+{-# LANGUAGE DataKinds #-}
+{-# LANGUAGE DeriveGeneric #-}
+{-# LANGUAGE DerivingStrategies #-}
+{-# LANGUAGE DeriveAnyClass #-}
+{-# LANGUAGE DerivingVia #-}
+{-# LANGUAGE FlexibleContexts #-}
+{-# LANGUAGE FlexibleInstances #-}
+{-# LANGUAGE GeneralizedNewtypeDeriving #-}
+{-# LANGUAGE InstanceSigs #-}
+{-# LANGUAGE MultiParamTypeClasses #-}
+{-# LANGUAGE OverloadedStrings #-}
+{-# LANGUAGE ScopedTypeVariables #-}
+{-# LANGUAGE StandaloneDeriving #-}
+{-# LANGUAGE TypeApplications #-}
+{-# LANGUAGE TypeFamilies #-}
+{-# LANGUAGE TypeOperators #-}
+{-# LANGUAGE UndecidableInstances #-}
+{-# OPTIONS_GHC -Wall -Wno-orphans #-}
+
+import Control.Concurrent (threadDelay)
+import Control.Concurrent.Async (async)
+import Control.Monad (when)
+import Control.Monad.Except (ExceptT (..), MonadError (..), MonadTrans (..), runExceptT)
+import Data.Aeson (FromJSON (..), ToJSON (..))
+import Data.Aeson.Encode.Pretty (encodePretty)
+import Data.String.Conversions (cs)
+import Data.Swagger (ToSchema)
+import Data.Typeable (Proxy (Proxy))
+import qualified GHC.Generics as GHC
+import qualified Network.HTTP.Client as Client
+import qualified Network.Wai.Handler.Warp as Warp
+import Servant.API
+import Servant.Client
+import Servant.Server
+import Servant.Swagger
+```
+
+## The API
+
+This looks like a `Verb`-based routing table, except that `UVerb` has
+no status, and carries a list of response types rather than a single
+one.  Each entry in the list carries its own response code.
+
+```haskell
+type API =
+         "fisx"  :> Capture "bool" Bool
+         :> UVerb 'GET '[JSON] '[FisxUser, WithStatus 303 String]
+    :<|> "arian"
+         :> UVerb 'GET '[JSON] '[WithStatus 201 ArianUser]
+```
+
+Here are the details:
+
+```haskell
+data FisxUser = FisxUser {name :: String}
+  deriving (Eq, Show, GHC.Generic)
+
+instance ToJSON FisxUser
+instance FromJSON FisxUser
+instance ToSchema FisxUser
+
+-- | 'HasStatus' allows us to can get around 'WithStatus' if we want
+-- to, and associate the status code with our resource types directly.
+--
+-- (To avoid orphan instances and make it more explicit what's in the
+-- API and what isn't, we could even introduce a newtype 'Resource'
+-- that wraps all the types we're using in our routing table, and then
+-- define lots of 'HasStatus' instances for @Resource This@ and
+-- @Resource That@.)
+instance HasStatus FisxUser where
+  type StatusOf FisxUser = 203
+
+data ArianUser = ArianUser
+  deriving (Eq, Show, GHC.Generic)
+
+instance ToJSON ArianUser
+instance FromJSON ArianUser
+instance ToSchema ArianUser
+```
+
+## Server, Client, Swagger
+
+You can just respond with any of the elements of the union in handlers.
+
+```haskell
+fisx :: Bool -> Handler (Union '[FisxUser, WithStatus 303 String])
+fisx True = respond (FisxUser "fisx")
+fisx False = respond (WithStatus @303 ("still fisx" :: String))
+
+arian :: Handler (Union '[WithStatus 201 ArianUser])
+arian = respond (WithStatus @201 ArianUser)
+```
+
+You can create client functions like you're used to:
+
+```
+fisxClient :: Bool -> ClientM (Union '[FisxUser, WithStatus 303 String])
+arianClient :: ClientM (Union '[WithStatus 201 ArianUser])
+(fisxClient :<|> arianClient) = client (Proxy @API)
+```
+
+...  and that's basically it!  Here are a few sample commands that
+show you how the swagger docs look like and how you can handle the
+result unions in clients:
+
+```
+main :: IO ()
+main = do
+  putStrLn . cs . encodePretty $ toSwagger (Proxy @API)
+  _ <- async . Warp.run 8080 $ serve (Proxy @API) (fisx :<|> arian)
+  threadDelay 50000
+  mgr <- Client.newManager Client.defaultManagerSettings
+  let cenv = mkClientEnv mgr (BaseUrl Http "localhost" 8080 "")
+  result <- runClientM (fisxClient True) cenv
+  print $ foldMapUnion (Proxy @Show) show <$> result
+  print $ matchUnion @FisxUser <$> result
+  print $ matchUnion @(WithStatus 303 String) <$> result
+  pure ()
+```
+
+## Idiomatic exceptions
+
+Since `UVerb` (probably) will mostly be used for error-like responses, it may be desirable to be able to early abort handler, like with current servant one would use `throwError` with `ServerError`.
+
+```haskell
+newtype UVerbT xs m a = UVerbT { unUVerbT :: ExceptT (Union xs) m a }
+  deriving (Functor, Applicative, Monad, MonadTrans)
+
+-- | Deliberately hide 'ExceptT's 'MonadError' instance to be able to use
+-- underlying monad's instance.
+instance MonadError e m => MonadError e (UVerbT xs m) where
+  throwError = lift . throwError
+  catchError (UVerbT act) h = UVerbT $ ExceptT $
+    runExceptT act `catchError` (runExceptT . unUVerbT . h)
+
+-- | This combinator runs 'UVerbT'. It applies 'respond' internally, so the handler
+-- may use the usual 'return'.
+runUVerbT :: (Monad m, HasStatus x, IsMember x xs) => UVerbT xs m x -> m (Union xs)
+runUVerbT (UVerbT act) = either id id <$> runExceptT (act >>= respond)
+
+-- | Short-circuit 'UVerbT' computation returning one of the response types.
+throwUVerb :: (Monad m, HasStatus x, IsMember x xs) => x -> UVerbT xs m a
+throwUVerb = UVerbT . ExceptT . fmap Left . respond
+```
+
+Example usage:
+
+```haskell
+data Foo = Foo Int Int Int
+  deriving (Show, Eq, GHC.Generic, ToJSON)
+  deriving HasStatus via WithStatus 200 Foo
+
+data Bar = Bar
+  deriving (Show, Eq, GHC.Generic)
+
+instance ToJSON Bar
+
+h :: Handler (Union '[Foo, WithStatus 400 Bar])
+h = runUVerbT $ do
+  when ({- something bad -} True) $
+    throwUVerb $ WithStatus @400 Bar
+
+  when ({- really bad -} False) $
+    throwError $ err500
+
+  -- a lot of code here...
+
+  return $ Foo 1 2 3
+```
+
+## Related Work
+
+There is the [issue from
+2017](https://github.com/haskell-servant/servant/issues/841) that was
+resolved by the introduction of `UVerb`, with a long discussion on
+alternative designs.
+
+[servant-checked-exceptions](https://hackage.haskell.org/package/servant-checked-exceptions)
+is a good solution to the problem, but it restricts the user to JSON
+and a very specific envelop encoding for the union type, which is
+often not acceptable.  (One good reason for this design choice is that
+it makes writing clients easier, where you need to get to the union
+type from one representative, and you don't want to run several
+parsers in the hope that the ones that should will always error out so
+you can try until the right one returns a value.)
+
+[servant-exceptions](https://github.com/ch1bo/servant-exceptions) is
+another shot at the problem.  It is inspired by
+servant-checked-exceptions, so it may be worth taking a closer look.
+The README claims that
+[cardano-sl](https://github.com/input-output-hk/cardano-sl) also has
+some code for generalized error handling.
+
+In an earier version of the `UVerb` implementation, we have used some
+code from
+[world-peace](https://hackage.haskell.org/package/world-peace), but
+that package itself wasn't flexible enough, and we had to use
+[sop-core](https://hackage.haskell.org/package/sop-core) to implement
+the `HasServer` instance.
+
+Here is a blog post we found on the subject:
+https://lukwagoallan.com/posts/unifying-servant-server-error-responses
+
+(If you have anything else, please add it here or let us know.)
+
+```haskell
+main :: IO ()
+main = return ()
+```

doc/cookbook/uverb/uverb.cabal

@@ -0,0 +1,35 @@
+cabal-version:       2.2
+name:                cookbook-uverb
+version:             0.0.1
+synopsis:            How to use the 'UVerb' type.
+description:         Listing alternative responses and exceptions in your API types.
+homepage:            http://docs.servant.dev/
+license:             BSD-3-Clause
+license-file:        ../../../servant/LICENSE
+author:              Servant Contributors
+maintainer:          haskell-servant-maintainers@googlegroups.com
+category:            Servant
+build-type:          Simple
+tested-with:         GHC==8.6.5, GHC==8.8.4, GHC==8.10.7
+
+executable cookbook-uverb
+  main-is:             UVerb.lhs
+  build-depends:       base == 4.*
+                     , aeson >= 1.2
+                     , aeson-pretty >= 0.8.8
+                     , async
+                     , http-client
+                     , mtl
+                     , servant
+                     , servant-client
+                     , servant-server
+                     , servant-swagger
+                     , string-conversions
+                     , swagger2
+                     , wai
+                     , warp
+  if impl(ghc >= 9)
+      buildable: False
+  default-language: Haskell2010
+  ghc-options:         -Wall -pgmL markdown-unlit
+  build-tool-depends:  markdown-unlit:markdown-unlit

doc/examples.md

@@ -6,6 +6,11 @@
     including a test-suite using [**hspec**](http://hspec.github.io/) and
     **servant-client**.
 
+- **[servant-examples](https://github.com/sras/servant-examples)**:
+
+    Similar to [the cookbook](https://docs.servant.dev/en/latest/cookbook/index.html) but
+    with no explanations, for developers who just want to look at code examples to find out how to do X or Y
+    with servant.
 
 - **[stack-templates](https://github.com/commercialhaskell/stack-templates)**
 
@@ -32,7 +37,7 @@
 
     An example consisting of
 
-    - a backend in uses `haskell-servant`
+    - a backend that uses `servant`
     - a frontend written in [PureScript](http://www.purescript.org/) using
       [servant-purescript](https://github.com/eskimor/servant-purescript) to generate
       an API wrapper in PureScript to interface the web API with
@@ -43,3 +48,13 @@
     An example for a web server written with **servant-server** and
     [persistent](https://www.stackage.org/package/persistent) for writing data
     into a database.
+
+- **[full-example-servant-elm-auth-yeshql-postgresql](https://github.com/aRkadeFR/FlashCard)**:
+
+    A full open source website written with **servant-server**, yeshql, postgresql and elm 0.19.
+
+
+- [`import Servant` github search](https://github.com/search?q=%22import+Servant%22+language%3AHaskell&type=Code)
+
+    It has thousands of results and can be a good way to see how people use servant in their projects or even to discover
+    servant-related libraries.

doc/index.rst

@@ -3,22 +3,24 @@ servant – A Type-Level Web DSL
 
 .. image:: https://raw.githubusercontent.com/haskell-servant/servant/master/servant.png
 
-**servant** is a set of packages for declaring web APIs at the type-level and
-then using those API specifications to:
+**servant** is a set of Haskell libraries for writing *type-safe* web
+applications but also *deriving* clients (in Haskell and other languages) or
+generating documentation for them, and more.
 
-- write servers (this part of **servant** can be considered a web framework),
-- obtain client functions (in haskell),
-- generate client functions for other programming languages,
-- generate documentation for your web applications
-- and more...
+This is achieved by taking as input a description of the web API
+as a Haskell type. Servant is then able to check that your server-side request
+handlers indeed implement your web API faithfully, or to automatically derive
+Haskell functions that can hit a web application that implements this API,
+generate a Swagger description or code for client functions in some other
+languages directly.
 
-All in a type-safe manner.
+If you would like to learn more, click the tutorial link below.
 
 .. toctree::
   :maxdepth: 2
 
-  introduction.rst
   tutorial/index.rst
   cookbook/index.rst
   examples.md
   links.rst
+  principles.rst

doc/links.rst

@@ -3,7 +3,7 @@ Helpful Links
 -------------
 
 - the central documentation (this site):
-  `haskell-servant.readthedocs.org <http://haskell-servant.readthedocs.org/>`_
+  `docs.servant.dev <http://docs.servant.dev/>`_
 
 - the github repo:
   `github.com/haskell-servant/servant <https://github.com/haskell-servant/servant>`_
@@ -12,13 +12,13 @@ Helpful Links
   `https://github.com/haskell-servant/servant/issues <https://github.com/haskell-servant/servant/issues>`_
 
 - the irc channel:
-  ``#servant`` on freenode
+  `#haskell-servant on libera.chat <https://web.libera.chat/#haskell-servant>`_
 
 - the mailing list:
   `groups.google.com/forum/#!forum/haskell-servant <https://groups.google.com/forum/#!forum/haskell-servant>`_
 
 - blog posts and videos and slides of some talks on servant:
-  `haskell-servant.github.io <http://haskell-servant.github.io>`_
+  `www.servant.dev <http://www.servant.dev>`_
 
 - the servant packages on hackage:
 

doc/principles.rst

@@ -1,5 +1,5 @@
-Introduction
-------------
+Principles
+----------
 
 **servant** has the following guiding principles:
 

doc/requirements.txt

@@ -1,25 +1,4 @@
-alabaster==0.7.7
-argh==0.26.1
-Babel==2.2.0
-backports-abc==0.4
-backports.ssl-match-hostname==3.5.0.1
-certifi==2015.11.20.1
-CommonMark==0.5.4
-docutils==0.12
-Jinja2==2.8
-livereload==2.4.1
-MarkupSafe==0.23
-pathtools==0.1.2
-Pygments==2.2.0
-pytz==2015.7
-PyYAML==3.11
-recommonmark==0.4.0
-singledispatch==3.4.0.3
-six==1.10.0
-snowballstemmer==1.2.1
-Sphinx==1.3.6
-sphinx-autobuild==0.5.2
-sphinx-rtd-theme==0.1.9
-tornado==4.3
-watchdog==0.8.3
-wheel==0.26.0
+recommonmark==0.5.0
+Sphinx==1.8.4
+sphinx_rtd_theme>=0.4.2
+jinja2<3.1.0

doc/tutorial/ApiType.lhs

@@ -129,16 +129,23 @@ type UserAPI4 = "users" :> Get '[JSON] [User]
 
 ### `StreamGet` and `StreamPost`
 
+*Note*: Streaming has changed considerably in `servant-0.15`.
+
 The `StreamGet` and `StreamPost` combinators are defined in terms of the more general `Stream`
 
 ``` haskell ignore
-data Stream (method :: k1) (framing :: *) (contentType :: *) a
-type StreamGet  = Stream 'GET
-type StreamPost = Stream 'POST
+data Stream (method :: k1) (status :: Nat) (framing :: *) (contentType :: *) (a :: *)
+
+type StreamGet  = Stream 'GET 200
+type StreamPost = Stream 'POST 200
 ```
 
-These describe endpoints that return a stream of values rather than just a single value. They not only take a single content type as a parameter, but also a framing strategy -- this specifies how the individual results are delineated from one another in the stream. The two standard strategies given with Servant are `NewlineFraming` and `NetstringFraming`, but others can be written to match other protocols.
-
+These describe endpoints that return a stream of values rather than just a
+single value. They not only take a single content type as a parameter, but also
+a framing strategy -- this specifies how the individual results are delineated
+from one another in the stream. The three standard strategies given with
+Servant are `NewlineFraming`, `NetstringFraming` and `NoFraming`, but others
+can be written to match other protocols.
 
 ### `Capture`
 
@@ -170,13 +177,12 @@ type UserAPI5 = "user" :> Capture "userid" Integer :> Get '[JSON] User
                 -- except that we explicitly say that "userid"
                 -- must be an integer
 
-           :<|> "user" :> Capture "userid" Integer :> DeleteNoContent '[JSON] NoContent
+           :<|> "user" :> Capture "userid" Integer :> DeleteNoContent
                 -- equivalent to 'DELETE /user/:userid'
 ```
 
-In the second case, `DeleteNoContent` specifies a 204 response code,
-`JSON` specifies the content types on which the handler will match,
-and `NoContent` says that the response will always be empty.
+In the second case, `DeleteNoContent` specifies a 204 response code
+and that the response will always be empty.
 
 ### `QueryParam`, `QueryParams`, `QueryFlag`
 
@@ -383,3 +389,30 @@ One example for this is if you want to serve a directory of static files along
 with the rest of your API. But you can plug in everything that is an
 `Application`, e.g. a whole web application written in any of the web
 frameworks that support `wai`.
+
+Be mindful! The `servant-server`'s router works by pattern-matching the
+different routes that are composed using `:<|>`. `Raw`, as an escape hatch,
+matches any route that hasn't been matched by previous patterns. Therefore,
+any subsequent route will be silently ignored.
+
+``` haskell
+type UserAPI14 = Raw
+            :<|> "users" :> Get '[JSON] [User]
+                 -- In this situation, the /users endpoint
+                 -- will not be reachable because the Raw
+                 -- endpoint matches requests before
+```
+A simple way to avoid this pitfall is to either use `Raw` as the last
+definition, or to always have it under a static path.
+
+``` haskell
+type UserAPI15 = "files" :> Raw
+                 -- The raw endpoint is under the /files
+                 -- static path, so it won't match /users.
+            :<|> "users" :> Get '[JSON] [User]
+
+type UserAPI16 = "users" :> Get '[JSON] [User]
+            :<|> Raw
+                 -- The Raw endpoint is matched last, so
+                 -- it won't overlap another endpoint.
+```

doc/tutorial/Authentication.lhs

@@ -47,7 +47,6 @@ module Authentication where
 import Data.Aeson                       (ToJSON)
 import Data.ByteString                  (ByteString)
 import Data.Map                         (Map, fromList)
-import Data.Monoid                      ((<>))
 import qualified Data.Map            as Map
 import Data.Proxy                       (Proxy (Proxy))
 import Data.Text                        (Text)
@@ -108,7 +107,7 @@ API with "private." Additionally, the private parts of our API use the
 realm for this authentication is `"foo-realm"`).
 
 Unfortunately we're not done. When someone makes a request to our `"private"`
-API, we're going to need to provide to servant the logic for validifying
+API, we're going to need to provide to servant the logic for validating
 usernames and passwords. This adds a certain conceptual wrinkle in servant's
 design that we'll briefly discuss. If you want the **TL;DR**: we supply a lookup
 function to servant's new `Context` primitive.
@@ -133,7 +132,7 @@ combinator. Using `Context`, we can supply a function of type
 handler. This will allow the handler to check authentication and return a `User`
 to downstream handlers if successful.
 
-In practice we wrap `BasicAuthData -> Handler` into a slightly
+In practice we wrap `BasicAuthData -> Handler User` into a slightly
 different function to better capture the semantics of basic authentication:
 
 ``` haskell ignore
@@ -260,7 +259,7 @@ this.
 
 Let's implement a trivial authentication scheme. We will protect our API by
 looking for a cookie named `"servant-auth-cookie"`. This cookie's value will
-contain a key from which we can lookup a `Account`.
+contain a key from which we can lookup an `Account`.
 
 ```haskell
 -- | An account type that we "fetch from the database" after
@@ -274,7 +273,7 @@ database = fromList [ ("key1", Account "Anne Briggs")
                     , ("key3", Account "Ghédalia Tazartès")
                     ]
 
--- | A method that, when given a password, will return a Account.
+-- | A method that, when given a password, will return an Account.
 -- This is our bespoke (and bad) authentication logic.
 lookupAccount :: ByteString -> Handler Account
 lookupAccount key = case Map.lookup key database of
@@ -318,7 +317,7 @@ genAuthAPI = Proxy
 
 Now we need to bring everything together for the server. We have the
 `AuthHandler Request Account` value and an `AuthProtected` endpoint. To bind these
-together, we need to provide a [Type Family](https://downloads.haskell.org/~ghc/latest/docs/html/users_guide/type-families.html)
+together, we need to provide a [Type Family](https://downloads.haskell.org/~ghc/8.8.1/docs/html/users_guide/glasgow_exts.html#type-families)
 instance that tells the `HasServer` instance that our `Context` will supply a
 `Account` (via `AuthHandler Request Account`) and that downstream combinators will
 have access to this `Account` value (or an error will be thrown if authentication
@@ -346,7 +345,7 @@ genAuthServerContext = authHandler :. EmptyContext
 
 -- | Our API, where we provide all the author-supplied handlers for each end
 -- point. Note that 'privateDataFunc' is a function that takes 'Account' as an
--- argument. We dont' worry about the authentication instrumentation here,
+-- argument. We don't worry about the authentication instrumentation here,
 -- that is taken care of by supplying context
 genAuthServer :: Server AuthGenAPI
 genAuthServer =
@@ -368,10 +367,10 @@ genAuthMain = run 8080 (serveWithContext genAuthAPI genAuthServerContext genAuth
 $ curl -XGET localhost:8080/private
 Missing auth header
 
-$ curl -XGET localhost:8080/private -H "servant-auth-cookie: key3"
+$ curl -XGET localhost:8080/private -H "Cookie: servant-auth-cookie=key3"
 [{"ssshhh":"this is a secret: Ghédalia Tazartès"}]
 
-$ curl -XGET localhost:8080/private -H "servant-auth-cookie: bad-key"
+$ curl -XGET localhost:8080/private -H "Cookie: servant-auth-cookie=bad-key"
 Invalid Cookie
 
 $ curl -XGET localhost:8080/public
@@ -385,11 +384,11 @@ Creating a generalized, ad-hoc authentication scheme was fairly straight
 forward:
 
 1. use the `AuthProtect` combinator to protect your API.
-2. choose a application-specific data type used by your server when
+2. choose an application-specific data type used by your server when
 authentication is successful (in our case this was `Account`).
 3. Create a value of `AuthHandler Request Account` which encapsulates the
 authentication logic (`Request -> Handler Account`). This function
-will be executed everytime a request matches a protected route.
+will be executed every time a request matches a protected route.
 4. Provide an instance of the `AuthServerData` type family, specifying your
 application-specific data type returned when authentication is successful (in
 our case this was `Account`).

doc/tutorial/Client.lhs

@@ -1,9 +1,9 @@
 # Querying an API
 
-While defining handlers that [serve an API](Server.lhs) has a lot to it, querying an API is simpler: we do not care about what happens inside the webserver, we just need to know how to talk to it and get a response back. That said, we usually have to write the querying functions by hand because the structure of the API isn't a first class citizen and can't be inspected to generate the client-side functions.
+While defining handlers that [serve an API](Server.html) has a lot to it, querying an API is simpler: we do not care about what happens inside the webserver, we just need to know how to talk to it and get a response back. That said, we usually have to write the querying functions by hand because the structure of the API isn't a first class citizen and can't be inspected to generate the client-side functions.
 
 **servant** however has a way to inspect APIs, because APIs are just Haskell types and (GHC) Haskell lets us do quite a few things with types. In the same way that we look at an API type to deduce the types the handlers should have, we can inspect the structure of the API to *derive* Haskell functions that take one argument for each occurrence of `Capture`, `ReqBody`, `QueryParam`
-and friends (see [the tutorial introduction](ApiType.lhs) for an overview). By *derive*, we mean that there's no code generation involved - the functions are defined just by the structure of the API type.
+and friends (see [the tutorial introduction](ApiType.html) for an overview). By *derive*, we mean that there's no code generation involved - the functions are defined just by the structure of the API type.
 
 The source for this tutorial section is a literate Haskell file, so first we need to have some language extensions and imports:
 
@@ -20,6 +20,9 @@ import GHC.Generics
 import Network.HTTP.Client (newManager, defaultManagerSettings)
 import Servant.API
 import Servant.Client
+import Servant.Types.SourceT (foreach)
+
+import qualified Servant.Client.Streaming as S
 ```
 
 Also, we need examples for some domain specific data types:
@@ -155,46 +158,100 @@ Email {from = "great@company.com", to = "alp@foo.com", subject = "Hey Alp, we mi
 
 The types of the arguments for the functions are the same as for (server-side) request handlers.
 
+## Changing the monad the client functions live in
+
+Just like `hoistServer` allows us to change the monad in which request handlers
+of a web application live, we also have `hoistClient` for changing the monad
+in which _client functions_ live. Consider the following trivial API:
+
+``` haskell
+type HoistClientAPI = Get '[JSON] Int :<|> Capture "n" Int :> Post '[JSON] Int
+
+hoistClientAPI :: Proxy HoistClientAPI
+hoistClientAPI = Proxy
+```
+
+We already know how to derive client functions for this API, and as we have
+seen above they all return results in the `ClientM` monad when using `servant-client`.
+However, `ClientM` is rarely (or never) the actual monad we need to use the client
+functions in. Sometimes we need to run them in IO, sometimes in a custom monad
+stack. `hoistClient` is a very simple solution to the problem of "changing" the monad
+the clients run in.
+
+``` haskell ignore
+hoistClient
+  :: HasClient ClientM api   -- we need a valid API
+  => Proxy api               -- a Proxy to the API type
+  -> (forall a. m a -> n a)  -- a "monad conversion function" (natural transformation)
+  -> Client m api            -- clients in the source monad
+  -> Client n api            -- result: clients in the target monad
+```
+
+The "conversion function" argument above, just like the ones given to `hoistServer`, must
+be able to turn an `m a` into an `n a` for any choice of type `a`.
+
+Let's see this in action on our example. We first derive our client functions as usual,
+with all of them returning a result in `ClientM`.
+
+``` haskell
+getIntClientM :: ClientM Int
+postIntClientM :: Int -> ClientM Int
+getIntClientM :<|> postIntClientM = client hoistClientAPI
+```
+
+And we finally decide that we want the handlers to run in IO instead, by
+"post-applying" `runClientM` to a fixed client environment.
+
+``` haskell
+-- our conversion function has type: forall a. ClientM a -> IO a
+-- the result has type:
+-- Client IO HoistClientAPI = IO Int :<|> (Int -> IO Int)
+getClients :: ClientEnv -> Client IO HoistClientAPI
+getClients clientEnv
+  = hoistClient hoistClientAPI
+                ( fmap (either (error . show) id)
+                . flip runClientM clientEnv
+                )
+                (client hoistClientAPI)
+```
+
 ## Querying Streaming APIs.
 
 Consider the following streaming API type:
 
 ``` haskell
-type StreamAPI = "positionStream" :> StreamGet NewlineFraming JSON (ResultStream Position)
+type StreamAPI = "positionStream" :> StreamGet NewlineFraming JSON (SourceIO Position)
 ```
 
-Note that when we declared an API to serve, we specified a `StreamGenerator` as a producer of streams. Now we specify our result type as a `ResultStream`. With types that can be used both ways, if appropriate adaptors are written (in the form of `ToStreamGenerator` and `BuildFromStream` instances), then this asymmetry isn't necessary. Otherwise, if you want to share the same API across clients and servers, you can parameterize it like so:
-
-``` haskell ignore
-type StreamAPI f = "positionStream" :> StreamGet NewlineFraming JSON (f Position)
-type ServerStreamAPI = StreamAPI StreamGenerator
-type ClientStreamAPI = StreamAPI ResultStream
-```
+Note that we use the same `SourceIO` type as on the server-side 
+(this is different from `servant-0.14`).
+However, we have to use different client, `Servant.Client.Streaming`,
+which can stream (but has different API).
 
 In any case, here's how we write a function to query our API:
 
-``` haskell
+```haskell
 streamAPI :: Proxy StreamAPI
 streamAPI = Proxy
 
-posStream :: ClientM (ResultStream Position)
-
-posStream = client streamAPI
+posStream :: S.ClientM (SourceIO Position)
+posStream = S.client streamAPI
 ```
 
-And here's how to just print out all elements from a `ResultStream`, to give some idea of how to work with them.
+We'll get back a `SourceIO Position`. Instead of `runClientM`, the streaming
+client provides `withClientM`: the underlying HTTP connection is open only
+until the inner functions exits.  Inside the block we can e.g just print out
+all elements from a `SourceIO`, to give some idea of how to work with them.
 
 ``` haskell
-printResultStream :: Show a => ResultStream a -> IO ()
-printResultStream (ResultStream k) = k $ \getResult ->
-       let loop = do
-            r <- getResult
-            case r of
-                Nothing -> return ()
-                Just x -> print x >> loop
-       in loop
+printSourceIO :: Show a => ClientEnv -> S.ClientM (SourceIO a) -> IO ()
+printSourceIO env c = S.withClientM c env $ \e -> case e of
+    Left err -> putStrLn $ "Error: " ++ show err
+    Right rs -> foreach fail print rs
 ```
 
-The stream is parsed and provided incrementally. So the above loop prints out each result as soon as it is received on the stream, rather than waiting until they are all available to print them at once.
+The stream is parsed and provided incrementally. So the above loop prints out
+each result as soon as it is received on the stream, rather than waiting until
+they are all available to print them at once.
 
 You now know how to use **servant-client**!

doc/tutorial/Docs.lhs

@@ -77,7 +77,7 @@ instance ToSample HelloMessage where
     [ ("When a value is provided for 'name'", HelloMessage "Hello, Alp")
     , ("When 'name' is not specified", HelloMessage "Hello, anonymous coward")
     ]
-    -- mutliple examples to display this time
+    -- multiple examples to display this time
 
 ci :: ClientInfo
 ci = ClientInfo "Alp" "alp@foo.com" 26 ["haskell", "mathematics"]
@@ -108,7 +108,7 @@ apiDocs = docs exampleAPI
 markdown :: API -> String
 ```
 
-That lets us see what our API docs look down in markdown, by looking at `markdown apiDocs`.
+That lets us see what our API docs look like in markdown, by looking at `markdown apiDocs`.
 
 ````````` text
 ## GET /hello

doc/tutorial/Javascript.lhs

@@ -228,13 +228,13 @@ data CommonGeneratorOptions = CommonGeneratorOptions
   {
     -- | function generating function names
     functionNameBuilder :: FunctionName -> Text
-    -- | name used when a user want to send the request body (to let you redefine it)
+    -- | name used when a user wants to send the request body (to let you redefine it)
   , requestBody :: Text
     -- | name of the callback parameter when the request was successful
   , successCallback :: Text
     -- | name of the callback parameter when the request reported an error
   , errorCallback :: Text
-    -- | namespace on which we define the js function (empty mean local var)
+    -- | namespace on which we define the js function (empty means local var)
   , moduleName :: Text
     -- | a prefix that should be prepended to the URL in the generated JS
   , urlPrefix :: Text
@@ -477,7 +477,7 @@ data AngularOptions = AngularOptions
   }
 ```
 
-# Custom function name builder
+## Custom function name builder
 
 Servant comes with three name builders included:
 
@@ -518,4 +518,3 @@ var get_books = function(q, onSuccess, onError)
 }
 
 ```
-

doc/tutorial/Server.lhs

@@ -29,7 +29,7 @@ import Prelude.Compat
 
 import Control.Monad.Except
 import Control.Monad.Reader
-import Data.Aeson.Compat
+import Data.Aeson
 import Data.Aeson.Types
 import Data.Attoparsec.ByteString
 import Data.ByteString (ByteString)
@@ -46,6 +46,7 @@ import Servant
 import System.Directory
 import Text.Blaze
 import Text.Blaze.Html.Renderer.Utf8
+import Servant.Types.SourceT (source)
 import qualified Data.Aeson.Parser
 import qualified Text.Blaze.Html
 ```
@@ -95,12 +96,6 @@ users1 =
   ]
 ```
 
-Let's also write our API type.
-
-``` haskell ignore
-type UserAPI1 = "users" :> Get '[JSON] [User]
-```
-
 We can now take care of writing the actual webservice that will handle requests
 to such an API. This one will be very simple, being reduced to just a single
 endpoint. The type of the web application is determined by the API type,
@@ -188,7 +183,7 @@ users2 = [isaac, albert]
 
 Now, just like we separate the various endpoints in `UserAPI` with `:<|>`, we
 are going to separate the handlers with `:<|>` too! They must be provided in
-the same order as in in the API type.
+the same order as in the API type.
 
 ``` haskell
 server2 :: Server UserAPI2
@@ -318,8 +313,8 @@ For reference, here's a list of some combinators from **servant**:
 ## The `FromHttpApiData`/`ToHttpApiData` classes
 
 Wait... How does **servant** know how to decode the `Int`s from the URL? Or how
-to decode a `ClientInfo` value from the request body? This is what this and the
-following two sections address.
+to decode a `ClientInfo` value from the request body? The following three sections will
+help us answer these questions.
 
 `Capture`s and `QueryParam`s are represented by some textual value in URLs.
 `Header`s are similarly represented by a pair of a header name and a
@@ -604,7 +599,7 @@ $ curl -H 'Accept: text/html' http://localhost:8081/persons
 
 ## The `Handler` monad
 
-At the heart of the handlers is the monad they run in, namely a newtype `Handler` around `ExceptT ServantErr IO`
+At the heart of the handlers is the monad they run in, namely a newtype `Handler` around `ExceptT ServerError IO`
 ([haddock documentation for `ExceptT`](http://hackage.haskell.org/package/mtl-2.2.1/docs/Control-Monad-Except.html#t:ExceptT)).
 One might wonder: why this monad? The answer is that it is the
 simplest monad with the following properties:
@@ -622,12 +617,12 @@ newtype ExceptT e m a = ExceptT (m (Either e a))
 ```
 
 In short, this means that a handler of type `Handler a` is simply
-equivalent to a computation of type `IO (Either ServantErr a)`, that is, an IO
+equivalent to a computation of type `IO (Either ServerError a)`, that is, an IO
 action that either returns an error or a result.
 
-The module [`Control.Monad.Except`](https://hackage.haskell.org/package/mtl-2.2.1/docs/Control-Monad-Except.html#t:ExceptT)
+The module [`Control.Monad.Except`](https://hackage.haskell.org/package/mtl/docs/Control-Monad-Except.html#t:ExceptT)
 from which `ExceptT` comes is worth looking at.
-Perhaps most importantly, `ExceptT` and `Handler` are an instances of `MonadError`, so
+Perhaps most importantly, `ExceptT` and `Handler` are instances of `MonadError`, so
 `throwError` can be used to return an error from your handler (whereas `return`
         is enough to return a success).
 
@@ -637,9 +632,9 @@ kind and abort early. The next two sections cover how to do just that.
 
 ### Performing IO
 
-Another important instances from the list above are `MonadIO m => MonadIO
-(ExceptT e m)`, and therefore also `MonadIO Handler` as there is `MonadIO IO` instance..
-[`MonadIO`](http://hackage.haskell.org/package/transformers-0.4.3.0/docs/Control-Monad-IO-Class.html)
+Other important instances from the list above are `MonadIO m => MonadIO
+(ExceptT e m)`, and therefore also `MonadIO Handler` as there is a `MonadIO IO` instance.
+[`MonadIO`](http://hackage.haskell.org/package/base/docs/Control-Monad-IO-Class.html#t:MonadIO)
 is a class from the **transformers** package defined as:
 
 ``` haskell ignore
@@ -665,16 +660,16 @@ server5 = do
   return (FileContent filecontent)
 ```
 
-### Failing, through `ServantErr`
+### Failing, through `ServerError`
 
 If you want to explicitly fail at providing the result promised by an endpoint
 using the appropriate HTTP status code (not found, unauthorized, etc) and some
 error message, all you have to do is use the `throwError` function mentioned above
-and provide it with the appropriate value of type `ServantErr`, which is
+and provide it with the appropriate value of type `ServerError`, which is
 defined as:
 
 ``` haskell ignore
-data ServantErr = ServantErr
+data ServerError = ServerError
     { errHTTPCode     :: Int
     , errReasonPhrase :: String
     , errBody         :: ByteString -- lazy bytestring
@@ -690,7 +685,7 @@ use record update syntax:
 failingHandler :: Handler ()
 failingHandler = throwError myerr
 
-  where myerr :: ServantErr
+  where myerr :: ServerError
         myerr = err503 { errBody = "Sorry dear user." }
 ```
 
@@ -721,7 +716,7 @@ $ curl --verbose http://localhost:8081/myfile.txt
 >
 < HTTP/1.1 404 Not Found
 [snip]
-myfile.txt just isnt there, please leave this server alone.
+myfile.txt just isn't there, please leave this server alone.
 
 $ echo Hello > myfile.txt
 
@@ -823,7 +818,7 @@ If it doesn't exist, the handler will fail with a `404` status code.
 
 `serveDirectoryWebApp` uses some standard settings that fit the use case of
 serving static files for most web apps. You can find out about the other
-options in the documentation of the `Servant.Utils.StaticFiles` module.
+options in the documentation of the `Servant.Server.StaticFiles` module.
 
 ## Nested APIs
 
@@ -835,7 +830,7 @@ type UserAPI3 = -- view the user with given userid, in JSON
                 Capture "userid" Int :> Get '[JSON] User
 
            :<|> -- delete the user with given userid. empty response
-                Capture "userid" Int :> DeleteNoContent '[JSON] NoContent
+                Capture "userid" Int :> DeleteNoContent
 ```
 
 We can instead factor out the `userid`:
@@ -843,7 +838,7 @@ We can instead factor out the `userid`:
 ``` haskell
 type UserAPI4 = Capture "userid" Int :>
   (    Get '[JSON] User
-  :<|> DeleteNoContent '[JSON] NoContent
+  :<|> DeleteNoContent
   )
 ```
 
@@ -901,13 +896,13 @@ type API1 = "users" :>
 -- we factor out the Request Body
 type API2 = ReqBody '[JSON] User :>
   (    Get '[JSON] User -- just display the same user back, don't register it
-  :<|> PostNoContent '[JSON] NoContent  -- register the user. empty response
+  :<|> PostNoContent -- register the user. empty response
   )
 
 -- we factor out a Header
 type API3 = Header "Authorization" Token :>
   (    Get '[JSON] SecretData -- get some secret data, if authorized
-  :<|> ReqBody '[JSON] SecretData :> PostNoContent '[JSON] NoContent -- add some secret data, if authorized
+  :<|> ReqBody '[JSON] SecretData :> PostNoContent -- add some secret data, if authorized
   )
 
 newtype Token = Token ByteString
@@ -920,11 +915,11 @@ API type only at the end.
 ``` haskell
 type UsersAPI =
        Get '[JSON] [User] -- list users
-  :<|> ReqBody '[JSON] User :> PostNoContent '[JSON] NoContent -- add a user
+  :<|> ReqBody '[JSON] User :> PostNoContent -- add a user
   :<|> Capture "userid" Int :>
          ( Get '[JSON] User -- view a user
-      :<|> ReqBody '[JSON] User :> PutNoContent '[JSON] NoContent -- update a user
-      :<|> DeleteNoContent '[JSON] NoContent -- delete a user
+      :<|> ReqBody '[JSON] User :> PutNoContent -- update a user
+      :<|> DeleteNoContent -- delete a user
          )
 
 usersServer :: Server UsersAPI
@@ -953,11 +948,11 @@ usersServer = getUsers :<|> newUser :<|> userOperations
 ``` haskell
 type ProductsAPI =
        Get '[JSON] [Product] -- list products
-  :<|> ReqBody '[JSON] Product :> PostNoContent '[JSON] NoContent -- add a product
+  :<|> ReqBody '[JSON] Product :> PostNoContent -- add a product
   :<|> Capture "productid" Int :>
          ( Get '[JSON] Product -- view a product
-      :<|> ReqBody '[JSON] Product :> PutNoContent '[JSON] NoContent -- update a product
-      :<|> DeleteNoContent '[JSON] NoContent -- delete a product
+      :<|> ReqBody '[JSON] Product :> PutNoContent -- update a product
+      :<|> DeleteNoContent -- delete a product
          )
 
 data Product = Product { productId :: Int }
@@ -1001,11 +996,11 @@ abstract that away:
 -- indexed by values of type 'i'
 type APIFor a i =
        Get '[JSON] [a] -- list 'a's
-  :<|> ReqBody '[JSON] a :> PostNoContent '[JSON] NoContent -- add an 'a'
+  :<|> ReqBody '[JSON] a :> PostNoContent -- add an 'a'
   :<|> Capture "id" i :>
          ( Get '[JSON] a -- view an 'a' given its "identifier" of type 'i'
-      :<|> ReqBody '[JSON] a :> PutNoContent '[JSON] NoContent -- update an 'a'
-      :<|> DeleteNoContent '[JSON] NoContent -- delete an 'a'
+      :<|> ReqBody '[JSON] a :> PutNoContent -- update an 'a'
+      :<|> DeleteNoContent -- delete an 'a'
          )
 
 -- Build the appropriate 'Server'
@@ -1133,14 +1128,14 @@ This is the webservice in action:
 ``` bash
 $ curl http://localhost:8081/a
 1797
-$ curl http://localhost:8081/b
-"hi"
+$ curl http://localhost:8081/b -X GET -d '42.0' -H 'Content-Type: application/json'
+true
 ```
 
 ### An arrow is a reader too.
 
 In previous versions of `servant` we had an `enter` to do what `hoistServer`
-does now. `enter` had a ambitious design goals, but was problematic in practice.
+does now. `enter` had an ambitious design goals, but was problematic in practice.
 
 One problematic situation was when the source monad was `(->) r`, yet it's
 handy in practice, because `(->) r` is isomorphic to `Reader r`.
@@ -1166,24 +1161,37 @@ app5 = serve readerAPI (hoistServer readerAPI funToHandler funServerT)
 
 ## Streaming endpoints
 
-We can create endpoints that don't just give back a single result, but give back a *stream* of results, served one at a time. Stream endpoints only provide a single content type, and also specify what framing strategy is used to delineate the results. To serve these results, we need to give back a stream producer. Adapters can be written to `Pipes`, `Conduit` and the like, or written directly as `StreamGenerator`s. StreamGenerators are IO-based continuations that are handed two functions -- the first to write the first result back, and the second to write all subsequent results back. (This is to allow handling of situations where the entire stream is prefixed by a header, or where a boundary is written between elements, but not prior to the first element). The API of a streaming endpoint needs to explicitly specify which sort of generator it produces. Note that the generator itself is returned by a `Handler` action, so that additional IO may be done in the creation of one.
+We can create endpoints that don't just give back a single result, but give
+back a *stream* of results, served one at a time. Stream endpoints only provide
+a single content type, and also specify what framing strategy is used to
+delineate the results. To serve these results, we need to give back a stream
+producer. Adapters can be written to *Pipes*, *Conduit* and the like, or
+written directly as `SourceIO`s. SourceIO builds upon servant's own `SourceT`
+stream type (it's simpler than *Pipes* or *Conduit*).
+The API of a streaming endpoint needs to explicitly specify which sort of
+generator it produces. Note that the generator itself is returned by a
+`Handler` action, so that additional IO may be done in the creation of one.
 
 ``` haskell
-type StreamAPI = "userStream" :> StreamGet NewlineFraming JSON (StreamGenerator User)
+type StreamAPI = "userStream" :> StreamGet NewlineFraming JSON (SourceIO User)
 streamAPI :: Proxy StreamAPI
 streamAPI = Proxy
 
-streamUsers :: StreamGenerator User
-streamUsers = StreamGenerator $ \sendFirst sendRest -> do
-                       sendFirst isaac
-                       sendRest  albert
-                       sendRest  albert
+streamUsers :: SourceIO User
+streamUsers = source [isaac, albert, albert]
 
 app6 :: Application
 app6 = serve streamAPI (return streamUsers)
 ```
 
-This simple application returns a stream of `User` values encoded in JSON format, with each value separated by a newline. In this case, the stream will consist of the value of `isaac`, followed by the value of `albert`, then the value of `albert` a third time. Importantly, the stream is written back as results are produced, rather than all at once. This means first that results are delivered when they are available, and second, that if an exception interrupts production of the full stream, nonetheless partial results have already been written back.
+This simple application returns a stream of `User` values encoded in JSON
+format, with each value separated by a newline. In this case, the stream will
+consist of the value of `isaac`, followed by the value of `albert`, then the
+value of `albert` a second time. Importantly, the stream is written back as
+results are produced, rather than all at once. This means first that results
+are delivered when they are available, and second, that if an exception
+interrupts production of the full stream, nonetheless partial results have
+already been written back.
 
 ## Conclusion
 

doc/tutorial/index.rst

@@ -3,13 +3,13 @@ Tutorial
 
 This is an introductory tutorial to **servant**. Whilst browsing is fine, it makes more sense if you read the sections in order, or at least read the first section before anything else.
 
-(Any comments, issues or feedback about the tutorial can be submitted
-to `servant's issue tracker <http://github.com/haskell-servant/servant/issues>`_.)
-
+Any comments, issues or feedback about the tutorial can be submitted
+to `servant's issue tracker <http://github.com/haskell-servant/servant/issues>`_.
 
 .. toctree::
   :maxdepth: 1
 
+  install.rst
   ApiType.lhs
   Server.lhs
   Client.lhs

doc/tutorial/install.rst

@@ -0,0 +1,68 @@
+Install
+========
+
+cabal-install
+--------
+
+The whole tutorial is a `cabal <https://cabal.readthedocs.io/en/latest/>`_
+project and can be built locally as follows:
+
+.. code-block:: bash
+
+   $ git clone https://github.com/haskell-servant/servant.git
+   $ cd servant
+   # build
+   $ cabal new-build tutorial
+   # load in ghci to play with it
+   $ cabal new-repl tutorial
+
+stack
+--------
+
+The servant `stack <https://docs.haskellstack.org/en/stable/README/>`_ template includes the working tutorial. To initialize this template, run:
+
+.. code-block:: bash
+
+   $ stack new myproj servant
+   $ cd myproj
+   # build
+   $ stack build
+   # start server
+   $ stack exec myproj-exe
+
+The code can be found in the `*.lhs` files under `doc/tutorial/` in the
+repository. Feel free to edit it while you're reading this documentation and
+see the effect of your changes.
+
+nix
+--------
+
+`Nix <https://nixos.org/nix/>`_ users should feel free to take a look at
+the `nix/shell.nix` file in the repository and use it to provision a suitable
+environment to build and run the examples.
+
+Note for Ubuntu users
+--------
+
+Ubuntu's packages for `ghc`, `cabal`, and `stack` are years out of date.
+If the instructions above fail for you,
+try replacing the Ubuntu packages with up-to-date versions.
+First remove the installed versions:
+
+.. code-block:: bash
+
+   # remove the obsolete versions
+   $ sudo apt remove ghc haskell-stack cabal-install
+
+Then install fresh versions of the Haskell toolchain
+using the `ghcup <https://www.haskell.org/ghcup/install/>`_ installer.
+
+As of February 2022, one easy way to do this is by running a bootstrap script:
+
+.. code-block:: bash
+
+   $ curl --proto '=https' --tlsv1.2 -sSf https://get-ghcup.haskell.org | sh
+
+The script is interactive and will prompt you for details about what
+you want installed and where.  To install manually,
+see `the detailed instructions <https://www.haskell.org/ghcup/install/#manual-install>`_.

doc/tutorial/test/Spec.hs

@@ -1 +1,11 @@
-{-# OPTIONS_GHC -F -pgmF hspec-discover #-}
+module Main where
+
+import qualified JavascriptSpec
+
+import Test.Hspec (Spec, hspec, describe)
+
+main :: IO ()
+main = hspec spec
+
+spec :: Spec
+spec = describe "Javascript" JavascriptSpec.spec

doc/tutorial/tinc.yaml

@@ -1,13 +0,0 @@
-dependencies:
-  - name: servant
-    path: ../../servant
-  - name: servant-server
-    path: ../../servant-server
-  - name: servant-client
-    path: ../../servant-client
-  - name: servant-js
-    path: ../../servant-js
-  - name: servant-docs
-    path: ../../servant-docs
-  - name: servant-foreign
-    path: ../../servant-foreign

doc/tutorial/tutorial.cabal

@@ -1,22 +1,20 @@
+cabal-version:       2.2
 name:                tutorial
 version:             0.10
 synopsis:            The servant tutorial
 description:
   The servant tutorial can be found at
-  <http://haskell-servant.readthedocs.org/>
-homepage:            http://haskell-servant.readthedocs.org/
+  <http://docs.servant.dev/>
+homepage:            http://docs.servant.dev/
 category:            Servant, Documentation
-license:             BSD3
+license:             BSD-3-Clause
 license-file:        LICENSE
 author:              Servant Contributors
 maintainer:          haskell-servant-maintainers@googlegroups.com
 build-type:          Simple
-cabal-version:       >=1.10
 tested-with:
-  GHC==7.8.4
-  GHC==7.10.3
-  GHC==8.0.2
-  GHC==8.2.2
+  GHC==8.6.5
+  GHC==8.8.3, GHC ==8.10.7
 extra-source-files:
   static/index.html
   static/ui.js
@@ -34,9 +32,8 @@ library
   -- Packages `servant` depends on.
   -- We don't need to specify bounds here as this package is never released.
   build-depends:
-      base >= 4.7 && <4.11
+      base >= 4.7 && <5
     , aeson
-    , aeson-compat
     , attoparsec
     , base-compat
     , bytestring
@@ -66,16 +63,15 @@ library
       blaze-html   >= 0.9.0.1 && < 0.10
     , blaze-markup >= 0.8.0.0 && < 0.9
     , cookie       >= 0.4.3   && < 0.5
-    , js-jquery    >= 3.2.1   && < 3.3
-    , lucid        >= 2.9.9   && < 2.10
-    , mtl-compat   >= 0.2.1   && < 0.3
-    , random       >= 1.1     && < 1.2
+    , js-jquery    >= 3.3.1   && < 3.4
+    , lucid        >= 2.9.11  && < 2.12
+    , random       >= 1.1     && < 1.3
     , servant-js   >= 0.9     && < 0.10
-    , time         >= 1.4.2   && < 1.9
+    , time         >= 1.6.0.1 && < 1.13
 
   -- For legacy tools, we need to specify build-depends too
-  build-depends: markdown-unlit >= 0.4.1 && <0.5
-  build-tool-depends: markdown-unlit:markdown-unlit >= 0.4.1 && <0.5
+  build-depends: markdown-unlit >= 0.5.0 && <0.6
+  build-tool-depends: markdown-unlit:markdown-unlit >= 0.5.0 && <0.6
 
 test-suite spec
   type: exitcode-stdio-1.0
@@ -84,8 +80,6 @@ test-suite spec
   hs-source-dirs: test
   main-is: Spec.hs
   other-modules: JavascriptSpec
-  build-tool-depends:
-    hspec-discover:hspec-discover
   build-depends: base
                , tutorial
                , hspec

ghcjs.nix

@@ -0,0 +1,22 @@
+let reflex-platform = import (builtins.fetchTarball
+      { name = "reflex-platform";
+        url = "https://github.com/reflex-frp/reflex-platform/archive/1aba6f367982bd6dd78ec2fda75ab246a62d32c5.tar.gz";
+      }) {};
+    pkgs = import ./nix/nixpkgs.nix; in
+
+pkgs.stdenv.mkDerivation {
+  name = "ghcjs-shell";
+  buildInputs =
+    [ (reflex-platform.ghcjs.ghcWithPackages (p: with p; [
+        attoparsec
+        hashable
+      ]))
+      pkgs.cabal-install
+      pkgs.gmp
+      pkgs.haskellPackages.cabal-plan
+      pkgs.haskellPackages.hspec-discover
+      pkgs.nodejs
+      pkgs.perl
+      pkgs.zlib
+    ];
+}

nix/README.md

@@ -21,3 +21,21 @@ a particular ghc version, e.g:
 ``` sh
 $ nix-shell nix/shell.nix --argstr compiler ghcHEAD
 ```
+
+**Possible GHC versions**
+-   `ghc865Binary`
+-   `ghc884`
+-   `ghc8104` - default
+-   `ghc901`
+
+### Cabal users
+
+GHC version can be chosen via the nix-shell parameter
+
+`cabal build all`
+
+### Stack version
+
+Since the ghc version is set by the LTS version, it is preferable to use the `ghc8104` version parameter for the nix-shell.
+
+`stack --no-nix --system-ghc <command>`

nix/nixpkgs.json

@@ -0,0 +1,4 @@
+{
+  "rev" : "05f0934825c2a0750d4888c4735f9420c906b388",
+  "sha256" : "1g8c2w0661qn89ajp44znmwfmghbbiygvdzq0rzlvlpdiz28v6gy"
+}

nix/nixpkgs.nix

@@ -0,0 +1,4 @@
+import (builtins.fetchTarball {
+  url = "https://github.com/NixOS/nixpkgs/archive/refs/tags/21.05.tar.gz";
+  sha256 = "sha256:1ckzhh24mgz6jd1xhfgx0i9mijk6xjqxwsshnvq789xsavrmsc36";
+}) {}

nix/shell.nix

@@ -1,21 +1,20 @@
-{ pkgs ? import <nixpkgs> {}
-, compiler ? "ghc821"
+{ compiler ? "ghc8104"
 , tutorial ? false
+, pkgs ? import ./nixpkgs.nix
 }:
+  
+  with pkgs;
 
-with pkgs;
-
-let
-  ghc = haskell.packages.${compiler}.ghcWithPackages (_: []);
-  docstuffs = python3.withPackages (ps: with ps; [ recommonmark sphinx sphinx_rtd_theme ]);
-in
-
-stdenv.mkDerivation {
-    name = "servant-dev";
-    buildInputs = [ ghc zlib python3 wget ]
-      ++ (if tutorial then [docstuffs postgresql] else []);
-    shellHook = ''
-      eval $(grep export ${ghc}/bin/ghc)
-      export LD_LIBRARY_PATH="${zlib}/lib";
-    '';
-}
+  let
+    ghc = haskell.packages.${compiler}.ghcWithPackages (_: []);
+    docstuffs = python3.withPackages (ps: with ps; [ recommonmark sphinx sphinx_rtd_theme ]);
+  in
+  stdenv.mkDerivation {
+      name = "servant-dev";
+      buildInputs = [ ghc zlib python3 wget cabal-install postgresql openssl stack haskellPackages.hspec-discover ]
+        ++ (if tutorial then [docstuffs postgresql] else []);
+      shellHook = ''
+        eval $(grep export ${ghc}/bin/ghc)
+        export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:"${zlib}/lib";
+      '';
+  }

servant-auth/README.md

@@ -0,0 +1 @@
+servant-auth-server/README.lhs

servant-auth/servant-auth-client/.ghci

@@ -0,0 +1 @@
+:set -isrc -itest -idoctest/ghci-wrapper/src

servant-auth/servant-auth-client/CHANGELOG.md

@@ -0,0 +1,26 @@
+# Changelog
+
+All notable changes to this project will be documented in this file.
+
+The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/)
+and this project adheres to [PVP Versioning](https://pvp.haskell.org/).
+
+## [Unreleased]
+
+## [0.4.1.0] - 2020-10-06
+
+- Support generic Bearer token auth 
+
+## [0.4.0.0] - 2019-03-08
+
+## Changed
+
+- #145 Support servant-0.16 in tests @domenkozar
+- #145 Drop GHC 7.10 support @domenkozar
+
+## [0.3.3.0] - 2018-06-18
+
+### Added
+- Support for GHC 8.4 by @phadej
+- Support for servant-0.14 by @phadej
+- Changelog by @domenkozar

servant-auth/servant-auth-client/LICENSE

@@ -0,0 +1,31 @@
+Copyright Julian K. Arni (c) 2015
+
+All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:
+
+    * Redistributions of source code must retain the above copyright
+      notice, this list of conditions and the following disclaimer.
+
+    * Redistributions in binary form must reproduce the above
+      copyright notice, this list of conditions and the following
+      disclaimer in the documentation and/or other materials provided
+      with the distribution.
+
+    * Neither the name of Julian K. Arni nor the names of other
+      contributors may be used to endorse or promote products derived
+      from this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+

servant-auth/servant-auth-client/Setup.hs

@@ -0,0 +1,2 @@
+import Distribution.Simple
+main = defaultMain

servant-auth/servant-auth-client/servant-auth-client.cabal

@@ -0,0 +1,80 @@
+cabal-version:  2.2
+name:           servant-auth-client
+version:        0.4.1.0
+synopsis:       servant-client/servant-auth compatibility
+description:    This package provides instances that allow generating clients from
+                <https://hackage.haskell.org/package/servant servant>
+                APIs that use
+                <https://hackage.haskell.org/package/servant-auth servant-auth's> @Auth@ combinator.
+                .
+                For a quick overview of the usage, see the <https://github.com/haskell-servant/servant/tree/master/servant-auth#readme README>.
+category:       Web, Servant, Authentication
+homepage:       https://github.com/haskell-servant/servant/tree/master/servant-auth#readme
+bug-reports:    https://github.com/haskell-servant/servant/issues
+author:         Julian K. Arni
+maintainer:     jkarni@gmail.com
+copyright:      (c) Julian K. Arni
+license:        BSD-3-Clause
+license-file:   LICENSE
+tested-with:    GHC ==8.6.5 || ==8.8.4 || ==8.10.4 || ==9.0.1
+build-type:     Simple
+extra-source-files:
+    CHANGELOG.md
+
+source-repository head
+  type: git
+  location: https://github.com/haskell-servant/servant
+
+library
+  hs-source-dirs:
+      src
+  default-extensions: ConstraintKinds DataKinds DefaultSignatures DeriveFoldable DeriveFunctor DeriveGeneric DeriveTraversable FlexibleContexts FlexibleInstances FunctionalDependencies GADTs KindSignatures MultiParamTypeClasses OverloadedStrings RankNTypes ScopedTypeVariables TypeFamilies TypeOperators
+  ghc-options: -Wall
+  build-depends:
+      base                >= 4.10     && < 4.18
+    , bytestring          >= 0.10.6.0 && < 0.12
+    , containers          >= 0.5.6.2  && < 0.7
+    , servant-auth        == 0.4.*
+    , servant             >= 0.13     && < 0.20
+    , servant-client-core >= 0.13     && < 0.20
+
+  exposed-modules:
+      Servant.Auth.Client
+      Servant.Auth.Client.Internal
+  default-language: Haskell2010
+
+test-suite spec
+  type: exitcode-stdio-1.0
+  main-is: Spec.hs
+  hs-source-dirs:
+      test
+  default-extensions: ConstraintKinds DataKinds DefaultSignatures DeriveFoldable DeriveFunctor DeriveGeneric DeriveTraversable FlexibleContexts FlexibleInstances FunctionalDependencies GADTs KindSignatures MultiParamTypeClasses OverloadedStrings RankNTypes ScopedTypeVariables TypeFamilies TypeOperators
+  ghc-options: -Wall
+  build-tool-depends: hspec-discover:hspec-discover >=2.5.5 && <2.10
+
+  -- dependencies with bounds inherited from the library stanza
+  build-depends:
+      base
+    , servant-client
+    , servant-auth
+    , servant
+    , servant-auth-client
+ 
+  -- test dependencies
+  build-depends:
+      hspec                >= 2.5.5    && < 2.10
+    , QuickCheck           >= 2.11.3   && < 2.15
+    , aeson                >= 1.3.1.1  && < 3
+    , bytestring           >= 0.10.6.0 && < 0.12
+    , http-client          >= 0.5.13.1 && < 0.8
+    , http-types           >= 0.12.2   && < 0.13
+    , servant-auth-server  >= 0.4.2.0  && < 0.5
+    , servant-server       >= 0.13     && < 0.20
+    , time                 >= 1.5.0.1  && < 1.13
+    , transformers         >= 0.4.2.0  && < 0.6
+    , wai                  >= 3.2.1.2  && < 3.3
+    , warp                 >= 3.2.25   && < 3.4
+    , jose                 >= 0.10     && < 0.11
+  other-modules:
+      Servant.Auth.ClientSpec
+  default-language: Haskell2010

servant-auth/servant-auth-client/src/Servant/Auth/Client.hs

@@ -0,0 +1,3 @@
+module Servant.Auth.Client (Token(..), Bearer) where
+
+import Servant.Auth.Client.Internal (Bearer, Token(..))

servant-auth/servant-auth-client/src/Servant/Auth/Client/Internal.hs

@@ -0,0 +1,64 @@
+{-# LANGUAGE CPP                        #-}
+{-# LANGUAGE GeneralizedNewtypeDeriving #-}
+{-# LANGUAGE UndecidableInstances       #-}
+{-# OPTIONS_GHC -fno-warn-orphans #-}
+#if __GLASGOW_HASKELL__ == 800
+{-# OPTIONS_GHC -fno-warn-redundant-constraints #-}
+#endif
+module Servant.Auth.Client.Internal where
+
+import qualified Data.ByteString    as BS
+import           Data.Monoid
+import           Data.Proxy         (Proxy (..))
+import           Data.String        (IsString)
+import           GHC.Exts           (Constraint)
+import           GHC.Generics       (Generic)
+import           Servant.API        ((:>))
+import           Servant.Auth
+
+import           Servant.Client.Core
+import           Data.Sequence ((<|))
+
+-- | A simple bearer token.
+newtype Token = Token { getToken :: BS.ByteString }
+  deriving (Eq, Show, Read, Generic, IsString)
+
+type family HasBearer xs :: Constraint where
+  HasBearer (Bearer ': xs) = ()
+  HasBearer (JWT ': xs) = ()
+  HasBearer (x ': xs)   = HasBearer xs
+  HasBearer '[]         = BearerAuthNotEnabled
+
+class BearerAuthNotEnabled
+
+-- | @'HasBearer' auths@ is nominally a redundant constraint, but ensures we're not
+-- trying to send a token to an API that doesn't accept them.
+instance (HasBearer auths, HasClient m api) => HasClient m (Auth auths a :> api) where
+  type Client m (Auth auths a :> api) = Token -> Client m api
+
+  clientWithRoute m _ req (Token token)
+    = clientWithRoute m (Proxy :: Proxy api)
+    $ req { requestHeaders = ("Authorization", headerVal) <| requestHeaders req  }
+      where
+        headerVal = "Bearer " <> token
+
+#if MIN_VERSION_servant_client_core(0,14,0)
+  hoistClientMonad pm _ nt cl = hoistClientMonad pm (Proxy :: Proxy api) nt . cl
+#endif
+
+
+-- * Authentication combinators
+
+-- | A Bearer token in the Authorization header:
+--
+--    @Authorization: Bearer <token>@
+--
+-- This can be any token recognized by the server, for example,
+-- a JSON Web Token (JWT).
+--
+-- Note that, since the exact way the token is validated is not specified,
+-- this combinator can only be used in the client. The server would not know
+-- how to validate it, while the client does not care.
+-- If you want to implement Bearer authentication in your server, you have to
+-- choose a specific combinator, such as 'JWT'.
+data Bearer