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

HTML reader: Parse name/content pairs from meta tags as metadata.

Browse source 
Closes #1106.
This commit is contained in:
John MacFarlane 2014-01-01 09:22:37 -08:00
parent d6ec6cf9cf
commit f3ee82373b
2 changed files with 11 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

11
src/Text/Pandoc/Readers/HTML.hs
View file

@ -76,9 +76,18 @@ pBody :: TagParser [Block]
pBody = pInTags "body" block
pHead :: TagParser [Block]
pHead = pInTags "head" $ pTitle <|> ([] <$ pAnyTag)
pHead = pInTags "head" $ pTitle <|> pMetaTag <|> ([] <$ pAnyTag)
where pTitle = pInTags "title" inline >>= setTitle . normalizeSpaces
setTitle t = [] <$ (updateState $ B.setMeta "title" (B.fromList t))
pMetaTag = do
mt <- pSatisfy (~== TagOpen "meta" [])
let name = fromAttrib "name" mt
if null name
then return []
else do
let content = fromAttrib "content" mt
updateState $ B.setMeta name (B.text content)
return []
block :: TagParser [Block]
block = choice

2
tests/html-reader.native
View file

@ -1,4 +1,4 @@
Pandoc (Meta {unMeta = fromList [("title",MetaInlines [Str "Pandoc",Space,Str "Test",Space,Str "Suite"])]})
Pandoc (Meta {unMeta = fromList [("generator",MetaInlines [Str "pandoc"]),("title",MetaInlines [Str "Pandoc",Space,Str "Test",Space,Str "Suite"])]})
[Para [Str "This",Space,Str "is",Space,Str "a",Space,Str "set",Space,Str "of",Space,Str "tests",Space,Str "for",Space,Str "pandoc",Str ".",Space,Str "Most",Space,Str "of",Space,Str "them",Space,Str "are",Space,Str "adapted",Space,Str "from",Space,Str "John",Space,Str "Gruber",Str "'",Str "s",Space,Str "markdown",Space,Str "test",Space,Str "suite",Str "."]
,HorizontalRule
,Header 1 ("",[],[]) [Str "Headers"]