diff options
| author | John MacFarlane <jgm@berkeley.edu> | 2014-01-01 09:22:37 -0800 |
|---|---|---|
| committer | John MacFarlane <jgm@berkeley.edu> | 2014-01-01 09:22:37 -0800 |
| commit | f3ee82373b4ad8e955db12d3c2c2159a2bea53a0 (patch) | |
| tree | 47e96d389717b60c5aa465aa359ecab7c0107bdf | |
| parent | d6ec6cf9cf5731977fa0f476cabef4786edd5665 (diff) | |
HTML reader: Parse name/content pairs from meta tags as metadata.
Closes #1106.
| -rw-r--r-- | src/Text/Pandoc/Readers/HTML.hs | 11 | ||||
| -rw-r--r-- | tests/html-reader.native | 2 |
2 files changed, 11 insertions, 2 deletions
diff --git a/src/Text/Pandoc/Readers/HTML.hs b/src/Text/Pandoc/Readers/HTML.hs index 4b44a3a21..506fe7770 100644 --- a/src/Text/Pandoc/Readers/HTML.hs +++ b/src/Text/Pandoc/Readers/HTML.hs @@ -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 diff --git a/tests/html-reader.native b/tests/html-reader.native index 794512426..e80905729 100644 --- a/tests/html-reader.native +++ b/tests/html-reader.native @@ -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"] |