Tissevert/pandoc
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 1

Adds Asciidoctor sprcific writer and tests

Browse source
This commit is contained in:
TG 2019-02-05 19:12:48 +01:00 committed by John MacFarlane
parent 8fdeae2872
commit 568b25d33a
9 changed files with 794 additions and 16 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

9
MANUAL.txt
View file

@ -279,7 +279,7 @@ General options {.options}
: Specify output format. *FORMAT* can be:
::: {#output-formats}
- `asciidoc` ([AsciiDoc])
- `asciidoc` ([AsciiDoc]) or `asciidoctor` ([AsciiDoctor])
- `beamer` ([LaTeX beamer][`beamer`] slide show)
- `commonmark` ([CommonMark] Markdown)
- `context` ([ConTeXt])
@ -465,6 +465,7 @@ General options {.options}
[GNU Texinfo]: http://www.gnu.org/software/texinfo/
[Emacs Org mode]: http://orgmode.org
[AsciiDoc]: http://www.methods.co.nz/asciidoc/
[AsciiDoctor]: https://asciidoctor.org/
[DZSlides]: http://paulrouget.com/dzslides/
[Word docx]: https://en.wikipedia.org/wiki/Office_Open_XML
[PDF]: https://www.adobe.com/pdf/
@ -3513,7 +3514,11 @@ reStructuredText
~ It will be rendered using an [interpreted text role `:math:`].
AsciiDoc
~ It will be rendered as `latexmath:[...]` when inline and `[latexmath]++++....++++` when in display mode.
~ For AsciiDoc output format (`-t asciidoc`) it will appear verbatim
surrounded by `latexmath:[$...$]` (for inline math) or
`[latexmath]++++\[...\]+++` (for display math).
For AsciiDoctor output format (`-t asciidoctor`) the LaTex delimiters
(`$..$` and `\[..\]`) are omitted.
Texinfo
~ It will be rendered inside a `@math` command.

2
README.md
View file

@ -81,7 +81,7 @@ It can convert *to*
<div id="output-formats">
- `asciidoc` ([AsciiDoc](http://www.methods.co.nz/asciidoc/))
- `asciidoc` ([AsciiDoc](http://www.methods.co.nz/asciidoc/)) or `asciidoctor` ([AsciiDoctor](https://asciidoctor.org/))
- `beamer` ([LaTeX beamer](https://ctan.org/pkg/beamer) slide show)
- `commonmark` ([CommonMark](http://commonmark.org) Markdown)
- `context` ([ConTeXt](http://www.contextgarden.net/))

38
data/templates/default.asciidoctor Normal file
View file

@ -0,0 +1,38 @@
$if(titleblock)$
= $title$
$if(author)$
$for(author)$$author$$sep$; $endfor$
$endif$
$if(date)$
$date$
$endif$
$if(keywords)$
:keywords: $for(keywords)$$keywords$$sep$, $endfor$
$endif$
$if(lang)$
:lang: $lang$
$endif$
$if(toc)$
:toc:
$endif$
$endif$
$if(abstract)$
[abstract]
== Abstract
$abstract$
$endif$
$for(header-includes)$
$header-includes$
$endfor$
$for(include-before)$
$include-before$
$endfor$
$body$
$for(include-after)$
$include-after$
$endfor$

1
pandoc.cabal
View file

@ -70,6 +70,7 @@ data-files:
data/templates/default.revealjs
data/templates/default.dzslides
data/templates/default.asciidoc
data/templates/default.asciidoctor
data/templates/default.haddock
data/templates/default.textile
data/templates/default.org

2
src/Text/Pandoc/Writers.hs
View file

@ -37,6 +37,7 @@ module Text.Pandoc.Writers
Writer(..)
, writers
, writeAsciiDoc
, writeAsciiDoctor
, writeBeamer
, writeCommonMark
, writeConTeXt
@ -178,6 +179,7 @@ writers = [
,("rtf" , TextWriter writeRTF)
,("org" , TextWriter writeOrg)
,("asciidoc" , TextWriter writeAsciiDoc)
,("asciidoctor" , TextWriter writeAsciiDoctor)
,("haddock" , TextWriter writeHaddock)
,("commonmark" , TextWriter writeCommonMark)
,("gfm" , TextWriter writeCommonMark)

41
src/Text/Pandoc/Writers/AsciiDoc.hs
View file

@ -37,7 +37,7 @@ that it has omitted the construct.
AsciiDoc: <http://www.methods.co.nz/asciidoc/>
-}
module Text.Pandoc.Writers.AsciiDoc (writeAsciiDoc) where
module Text.Pandoc.Writers.AsciiDoc (writeAsciiDoc, writeAsciiDoctor) where
import Prelude
import Control.Monad.State.Strict
import Data.Char (isPunctuation, isSpace, toLower)
@ -61,18 +61,27 @@ data WriterState = WriterState { defListMarker :: String
, bulletListLevel :: Int
, intraword :: Bool
, autoIds :: Set.Set String
, asciidoctorVariant :: Bool
}
defaultWriterState :: WriterState
defaultWriterState = WriterState { defListMarker = "::"
, orderedListLevel = 0
, bulletListLevel = 0
, intraword = False
, autoIds = Set.empty
, asciidoctorVariant = False
}
-- | Convert Pandoc to AsciiDoc.
writeAsciiDoc :: PandocMonad m => WriterOptions -> Pandoc -> m Text
writeAsciiDoc opts document =
evalStateT (pandocToAsciiDoc opts document) WriterState{
defListMarker = "::"
, orderedListLevel = 0
, bulletListLevel = 0
, intraword = False
, autoIds = Set.empty
}
evalStateT (pandocToAsciiDoc opts document) defaultWriterState
-- | Convert Pandoc to AsciiDoctor compatible AsciiDoc.
writeAsciiDoctor :: PandocMonad m => WriterOptions -> Pandoc -> m Text
writeAsciiDoctor opts document =
evalStateT (pandocToAsciiDoc opts document) defaultWriterState{ asciidoctorVariant = True }
type ADW = StateT WriterState
@ -411,12 +420,20 @@ inlineToAsciiDoc opts (Quoted DoubleQuote lst) =
inlineToAsciiDoc _ (Code _ str) = return $
text "`" <> text (escapeStringUsing (backslashEscapes "`") str) <> "`"
inlineToAsciiDoc _ (Str str) = return $ text $ escapeString str
inlineToAsciiDoc _ (Math InlineMath str) =
return $ "latexmath:[$" <> text str <> "$]"
inlineToAsciiDoc _ (Math DisplayMath str) =
inlineToAsciiDoc _ (Math InlineMath str) = do
isAsciidoctor <- gets asciidoctorVariant
let content = if isAsciidoctor
then text str
else "$" <> text str <> "$"
return $ "latexmath:[" <> content <> "]"
inlineToAsciiDoc _ (Math DisplayMath str) = do
isAsciidoctor <- gets asciidoctorVariant
let content = if isAsciidoctor
then text str
else "\\[" <> text str <> "\\]"
return $
blankline <> "[latexmath]" $$ "++++" $$
"\\[" <> text str <> "\\]"
content
$$ "++++" $$ blankline
inlineToAsciiDoc _ il@(RawInline f s)
| f == "asciidoc" = return $ text s

2
test/Tests/Old.hs
View file

@ -159,7 +159,7 @@ tests pandocPath =
"tikiwiki-reader.tikiwiki" "tikiwiki-reader.native" ]
, testGroup "other writers" $ map (\f -> testGroup f $ writerTests' f)
[ "opendocument" , "context" , "texinfo", "icml", "tei"
, "man" , "plain" , "rtf", "org", "asciidoc", "zimwiki"
, "man" , "plain" , "rtf", "org", "asciidoc", "asciidoctor", "zimwiki"
]
, testGroup "writers-lang-and-dir"
[ test' "latex" ["-f", "native", "-t", "latex", "-s"]

67
test/tables.asciidoctor Normal file
View file

@ -0,0 +1,67 @@
Simple table with caption:
.Demonstration of simple table syntax.
[cols=">,<,^,",options="header",]
|===
|Right |Left |Center |Default
|12 |12 |12 |12
|123 |123 |123 |123
|1 |1 |1 |1
|===
Simple table without caption:
[cols=">,<,^,",options="header",]
|===
|Right |Left |Center |Default
|12 |12 |12 |12
|123 |123 |123 |123
|1 |1 |1 |1
|===
Simple table indented two spaces:
.Demonstration of simple table syntax.
[cols=">,<,^,",options="header",]
|===
|Right |Left |Center |Default
|12 |12 |12 |12
|123 |123 |123 |123
|1 |1 |1 |1
|===
Multiline table with caption:
.Here’s the caption. It may span multiple lines.
[width="80%",cols="^20%,<17%,>20%,<43%",options="header",]
|===
|Centered Header |Left Aligned |Right Aligned |Default aligned
|First |row |12.0 |Example of a row that spans multiple lines.
|Second |row |5.0 |Here’s another one. Note the blank line between rows.
|===
Multiline table without caption:
[width="80%",cols="^20%,<17%,>20%,<43%",options="header",]
|===
|Centered Header |Left Aligned |Right Aligned |Default aligned
|First |row |12.0 |Example of a row that spans multiple lines.
|Second |row |5.0 |Here’s another one. Note the blank line between rows.
|===
Table without column headers:
[cols=">,<,^,>",]
|===
|12 |12 |12 |12
|123 |123 |123 |123
|1 |1 |1 |1
|===
Multiline table without column headers:
[width="80%",cols="^20%,<17%,>20%,43%",]
|===
|First |row |12.0 |Example of a row that spans multiple lines.
|Second |row |5.0 |Here’s another one. Note the blank line between rows.
|===

648
test/writer.asciidoctor Normal file
View file

@ -0,0 +1,648 @@
= Pandoc Test Suite
John MacFarlane; Anonymous
July 17, 2006
This is a set of tests for pandoc. Most of them are adapted from John Gruber’s
markdown test suite.
'''''
== Headers
=== Level 2 with an link:/url[embedded link]
==== Level 3 with _emphasis_
===== Level 4
====== Level 5
== Level 1
=== Level 2 with _emphasis_
==== Level 3
with no blank line
=== Level 2
with no blank line
'''''
== Paragraphs
Here’s a regular paragraph.
In Markdown 1.0.0 and earlier. Version 8. This line turns into a list item.
Because a hard-wrapped line in the middle of a paragraph looked like a list
item.
Here’s one with a bullet. * criminey.
There should be a hard line break +
here.
'''''
== Block Quotes
E-mail style:
____
This is a block quote. It is pretty short.
____
____
--
Code in a block quote:
....
sub status {
print "working";
}
....
A list:
[arabic]
. item one
. item two
Nested block quotes:
____
nested
____
____
nested
____
--
____
This should not be a block quote: 2 > 1.
And a following paragraph.
'''''
== Code Blocks
Code:
....
---- (should be four hyphens)
sub status {
print "working";
}
this code block is indented by one tab
....
And:
....
this code block is indented by two tabs
These should not be escaped: \$ \\ \> \[ \{
....
'''''
== Lists
=== Unordered
Asterisks tight:
* asterisk 1
* asterisk 2
* asterisk 3
Asterisks loose:
* asterisk 1
* asterisk 2
* asterisk 3
Pluses tight:
* Plus 1
* Plus 2
* Plus 3
Pluses loose:
* Plus 1
* Plus 2
* Plus 3
Minuses tight:
* Minus 1
* Minus 2
* Minus 3
Minuses loose:
* Minus 1
* Minus 2
* Minus 3
=== Ordered
Tight:
[arabic]
. First
. Second
. Third
and:
[arabic]
. One
. Two
. Three
Loose using tabs:
[arabic]
. First
. Second
. Third
and using spaces:
[arabic]
. One
. Two
. Three
Multiple paragraphs:
[arabic]
. Item 1, graf one.
+
Item 1. graf two. The quick brown fox jumped over the lazy dog’s back.
. Item 2.
. Item 3.
=== Nested
* Tab
** Tab
*** Tab
Here’s another:
[arabic]
. First
. Second:
* Fee
* Fie
* Foe
. Third
Same thing but with paragraphs:
[arabic]
. First
. Second:
* Fee
* Fie
* Foe
. Third
=== Tabs and spaces
* this is a list item indented with tabs
* this is a list item indented with spaces
** this is an example list item indented with tabs
** this is an example list item indented with spaces
=== Fancy list markers
[arabic, start=2]
. begins with 2
. and now 3
+
with a continuation
[lowerroman, start=4]
.. sublist with roman numerals, starting with 4
.. more items
[upperalpha]
... a subsublist
... a subsublist
Nesting:
[upperalpha]
. Upper Alpha
[upperroman]
.. Upper Roman.
[arabic, start=6]
... Decimal start with 6
[loweralpha, start=3]
.... Lower alpha with paren
Autonumbering:
. Autonumber.
. More.
.. Nested.
Should not be a list item:
M.A. 2007
B. Williams
'''''
== Definition Lists
Tight using spaces:
apple::
red fruit
orange::
orange fruit
banana::
yellow fruit
Tight using tabs:
apple::
red fruit
orange::
orange fruit
banana::
yellow fruit
Loose:
apple::
red fruit
orange::
orange fruit
banana::
yellow fruit
Multiple blocks with italics:
_apple_::
red fruit
+
contains seeds, crisp, pleasant to taste
_orange_::
orange fruit
+
....
{ orange code block }
....
+
____
orange block quote
____
Multiple definitions, tight:
apple::
red fruit
+
computer
orange::
orange fruit
+
bank
Multiple definitions, loose:
apple::
red fruit
+
computer
orange::
orange fruit
+
bank
Blank line after term, indented marker, alternate markers:
apple::
red fruit
+
computer
orange::
orange fruit
+
[arabic]
. sublist
. sublist
== HTML Blocks
Simple block on one line:
foo
And nested without indentation:
foo
bar
Interpreted markdown in a table:
This is _emphasized_
And this is *strong*
Here’s a simple block:
foo
This should be a code block, though:
....
<div>
foo
</div>
....
As should this:
....
<div>foo</div>
....
Now, nested:
foo
This should just be an HTML comment:
Multiline:
Code block:
....
<!-- Comment -->
....
Just plain comment, with trailing spaces on the line:
Code:
....
<hr />
....
Hr’s:
'''''
== Inline Markup
This is _emphasized_, and so _is this_.
This is *strong*, and so *is this*.
An _link:/url[emphasized link]_.
*_This is strong and em._*
So is *_this_* word.
*_This is strong and em._*
So is *_this_* word.
This is code: `>`, `$`, `\`, `\$`, `<html>`.
[line-through]*This is _strikeout_.*
Superscripts: a^bc^d a^_hello_^ a^hello there^.
Subscripts: H~2~O, H~23~O, H~many of them~O.
These should not be superscripts or subscripts, because of the unescaped
spaces: a^b c^d, a~b c~d.
'''''
== Smart quotes, ellipses, dashes
``Hello,'' said the spider. ```Shelob' is my name.''
`A', `B', and `C' are letters.
`Oak,' `elm,' and `beech' are names of trees. So is `pine.'
`He said, ``I want to go.''' Were you alive in the 70’s?
Here is some quoted ``code`' and a ``http://example.com/?foo=1&bar=2[quoted
link]''.
Some dashes: one—two — three—four — five.
Dashes between numbers: 5–7, 255–66, 1987–1999.
Ellipses…and…and….
'''''
== LaTeX
*
* latexmath:[2+2=4]
* latexmath:[x \in y]
* latexmath:[\alpha \wedge \omega]
* latexmath:[223]
* latexmath:[p]-Tree
* Here’s some display math:
[latexmath]
++++
\frac{d}{dx}f(x)=\lim_{h\to 0}\frac{f(x+h)-f(x)}{h}
++++
* Here’s one that has a line break in it:
latexmath:[\alpha + \omega \times x^2].
These shouldn’t be math:
* To get the famous equation, write `$e = mc^2$`.
* $22,000 is a _lot_ of money. So is $34,000. (It worked if ``lot'' is
emphasized.)
* Shoes ($20) and socks ($5).
* Escaped `$`: $73 _this should be emphasized_ 23$.
Here’s a LaTeX table:
'''''
== Special Characters
Here is some unicode:
* I hat: Î
* o umlaut: ö
* section: §
* set membership: ∈
* copyright: ©
AT&T has an ampersand in their name.
AT&T is another way to write it.
This & that.
4 < 5.
6 > 5.
Backslash: \
Backtick: `
Asterisk: *
Underscore: _
Left brace: \{
Right brace: }
Left bracket: [
Right bracket: ]
Left paren: (
Right paren: )
Greater-than: >
Hash: #
Period: .
Bang: !
Plus: +
Minus: -
'''''
== Links
=== Explicit
Just a link:/url/[URL].
link:/url/[URL and title].
link:/url/[URL and title].
link:/url/[URL and title].
link:/url/[URL and title]
link:/url/[URL and title]
link:/url/with_underscore[with_underscore]
mailto:nobody@nowhere.net[Email link]
link:[Empty].
=== Reference
Foo link:/url/[bar].
With link:/url/[embedded [brackets]].
link:/url/[b] by itself should be a link.
Indented link:/url[once].
Indented link:/url[twice].
Indented link:/url[thrice].
This should [not][] be a link.
....
[not]: /url
....
Foo link:/url/[bar].
Foo link:/url/[biz].
=== With ampersands
Here’s a http://example.com/?foo=1&bar=2[link with an ampersand in the URL].
Here’s a link with an amersand in the link text: http://att.com/[AT&T].
Here’s an link:/script?foo=1&bar=2[inline link].
Here’s an link:/script?foo=1&bar=2[inline link in pointy braces].
=== Autolinks
With an ampersand: http://example.com/?foo=1&bar=2
* In a list?
* http://example.com/
* It should.
An e-mail address: nobody@nowhere.net
____
Blockquoted: http://example.com/
____
Auto-links should not occur here: `<http://example.com/>`
....
or here: <http://example.com/>
....
'''''
== Images
From ``Voyage dans la Lune'' by Georges Melies (1902):
image:lalune.jpg[lalune,title="Voyage dans la Lune"]
Here is a movie image:movie.jpg[movie] icon.
'''''
== Footnotes
Here is a footnote reference,footnote:[Here is the footnote. It can go
anywhere after the footnote reference. It need not be placed at the end of the
document.] and another.[multiblock footnote omitted] This should _not_ be a
footnote reference, because it contains a space.[^my note] Here is an inline
note.footnote:[This is _easier_ to type. Inline notes may contain
http://google.com[links] and `]` verbatim characters, as well as [bracketed
text].]
____
Notes can go in quotes.footnote:[In quote.]
____
[arabic]
. And in list items.footnote:[In list.]
This paragraph should not be part of the note, as it is not indented.