834da53 · typeable/pandoc

git.iko.soy/typeable/pandoc

Add `rebase_relative_paths` extension.

- Add manual entry for (non-default) extension `rebase_relative_paths`. - Add constructor `Ext_rebase_relative_paths` to `Extensions` in Text.Pandoc.Extensions [API change]. When enabled, this extension rewrites relative image and link paths by prepending the (relative) directory of the containing file. - Make Markdown reader sensitive to the new extension. - Add tests for #3752. Closes #3752. NB. currently the extension applies to markdown and associated readers but not commonmark/gfm.

commit834da53058069fe50da510fa86e0807a7ff7868f authorJohn MacFarlane <jgm@berkeley.edu> dateMon 24 May 2021, 10:17 (5 years ago) parent0226d23

9 files changed+111−9

compare with master

unifiedsplit

MANUAL.txt+26−2 pandoc.cabal+4 src/Text/Pandoc/Extensions.hs+3 src/Text/Pandoc/Readers/Markdown.hs+29−7 test/command/3752.md+35 test/command/chap1/spider.png test/command/chap1/text.md+11 test/command/chap2/spider.png test/command/chap2/text.md+3

MANUAL.txt+26−2

@@ -3755,7 +3755,7 @@ definition:

37553755Note that space between items in a definition list is required.

37563756(A variant that loosens this requirement, but disallows "lazy"

37573757hard wrapping, can be activated with `compact_definition_lists`: see

3758−[Non-pandoc extensions], below.)

3758+[Non-default extensions], below.)

37593759

37603760[^3]: I have been influenced by the suggestions of [David

37613761 Wheeler](https://justatheory.com/2009/02/modest-markdown-proposal/).

@@ -5051,13 +5051,37 @@ author-in-text style inside notes when using a note style.

50515051[finding and editing styles]: https://citationstyles.org/authors/

50525052[CSL locale files]: https://github.com/citation-style-language/locales

50535053

5054−## Non-pandoc extensions

5054+## Non-default extensions

50555055

50565056The following Markdown syntax extensions are not enabled by default

50575057in pandoc, but may be enabled by adding `+EXTENSION` to the format

50585058name, where `EXTENSION` is the name of the extension. Thus, for

50595059example, `markdown+hard_line_breaks` is Markdown with hard line breaks.

50605060

5061+#### Extension: `rebase_relative_paths` ####

5062+

5063+Rewrite relative paths for Markdown links and images, depending

5064+on the path of the file containing the link or image link. For

5065+each link or image, pandoc will compute the directory of the

5066+containing file, relative to the working directory, and prepend

5067+the resulting path to the link or image path.

5068+

5069+The use of this extension is best understood by example.

5070+Suppose you have a a subdirectory for each chapter of a book,

5071+`chap1`, `chap2`, `chap3`. Each contains a file `text.md` and a

5072+number of images used in the chapter. You would like to have

5073+`![image](spider.jpg)` in `chap1/text.md` refer to

5074+`chap1/spider.jpg` and `![image](spider.jpg)` in `chap2/text.md`

5075+refer to `chap2/spider.jpg`. To do this, use

5076+

5077+ pandoc chap*/*.md -f markdown+rebase_relative_paths

5078+

5079+Without this extension, you would have to use

5080+`![image](chap1/spider.jpg)` in `chap1/text.md` and

5081+`![image](chap2/spider.jpg)` in `chap2/text.md`. Links with

5082+relative paths will be rewritten in the same way as images.

5083+*This option currently only affects Markdown input.*

5084+

50615085#### Extension: `attributes` ####

50625086

50635087Allows attributes to be attached to any inline or block-level

pandoc.cabal+4

@@ -214,6 +214,10 @@ extra-source-files:

214214 test/command/C.txt

215215 test/command/D.txt

216216 test/command/01.csv

217+ test/command/chap1/spider.png

218+ test/command/chap2/spider.png

219+ test/command/chap1/text.md

220+ test/command/chap2/text.md

217221 test/command/defaults1.yaml

218222 test/command/defaults2.yaml

219223 test/command/defaults3.yaml

src/Text/Pandoc/Extensions.hs+3

@@ -136,6 +136,8 @@ data Extension =

136136 | Ext_raw_html -- ^ Allow raw HTML

137137 | Ext_raw_tex -- ^ Allow raw TeX (other than math)

138138 | Ext_raw_markdown -- ^ Parse markdown in ipynb as raw markdown

139+ | Ext_rebase_relative_paths -- ^ Rebase relative image and link paths,

140+ -- relative to directory of containing file

139141 | Ext_shortcut_reference_links -- ^ Shortcut reference links

140142 | Ext_simple_tables -- ^ Pandoc-style simple tables

141143 | Ext_smart -- ^ "Smart" quotes, apostrophes, ellipses, dashes

@@ -462,6 +464,7 @@ getAllExtensions f = universalExtensions <> getAll f

462464 , Ext_gutenberg

463465 , Ext_smart

464466 , Ext_literate_haskell

467+ , Ext_rebase_relative_paths

465468 ]

466469 getAll "markdown_strict" = allMarkdownExtensions

467470 getAll "markdown_phpextra" = allMarkdownExtensions

src/Text/Pandoc/Readers/Markdown.hs+29−7

@@ -29,7 +29,7 @@ import qualified Data.Set as Set

2929import Data.Text (Text)

3030import qualified Data.Text as T

3131import qualified Data.ByteString.Lazy as BL

32−import System.FilePath (addExtension, takeExtension)

32+import System.FilePath (addExtension, takeExtension, isAbsolute, takeDirectory)

3333import Text.HTML.TagSoup hiding (Row)

3434import Text.Pandoc.Builder (Blocks, Inlines)

3535import qualified Text.Pandoc.Builder as B

@@ -1836,9 +1836,12 @@ regLink :: PandocMonad m

18361836 -> MarkdownParser m (F Inlines)

18371837regLink constructor lab = try $ do

18381838 (src, tit) <- source

1839+ rebase <- option False (True <$ guardEnabled Ext_rebase_relative_paths)

1840+ pos <- getPosition

1841+ let src' = if rebase then rebasePath pos src else src

18391842 attr <- option nullAttr $

18401843 guardEnabled Ext_link_attributes >> attributes

1841− return $ constructor attr src tit <$> lab

1844+ return $ constructor attr src' tit <$> lab

18421845

18431846-- a link like [this][ref] or [this][] or [this]

18441847referenceLink :: PandocMonad m

@@ -1854,6 +1857,8 @@ referenceLink constructor (lab, raw) = do

18541857 return (mempty, "")))

18551858 <|>

18561859 try ((guardDisabled Ext_spaced_reference_links <|> spnl) >> reference)

1860+ rebase <- option False (True <$ guardEnabled Ext_rebase_relative_paths)

1861+ pos <- getPosition

18571862 when (raw' == "") $ guardEnabled Ext_shortcut_reference_links

18581863 let labIsRef = raw' == "" || raw' == "[]"

18591864 let key = toKey $ if labIsRef then raw else raw'

@@ -1878,7 +1883,9 @@ referenceLink constructor (lab, raw) = do

18781883 Just ((src, tit), _) -> constructor nullAttr src tit <$> lab

18791884 Nothing -> makeFallback

18801885 else makeFallback

1881− Just ((src,tit), attr) -> constructor attr src tit <$> lab

1886+ Just ((src,tit), attr) ->

1887+ let src' = if rebase then rebasePath pos src else src

1888+ in constructor attr src' tit <$> lab

18821889

18831890dropBrackets :: Text -> Text

18841891dropBrackets = dropRB . dropLB

@@ -1911,15 +1918,30 @@ autoLink = try $ do

19111918 return $ return $ B.linkWith attr (src <> escapeURI extra) ""

19121919 (B.str $ orig <> extra)

19131920

1921+-- | Rebase a relative path, by adding the (relative) directory

1922+-- of the containing source position. Absolute links and URLs

1923+-- are untouched.

1924+rebasePath :: SourcePos -> Text -> Text

1925+rebasePath pos path = do

1926+ let fp = sourceName pos

1927+ in if isAbsolute (T.unpack path) || isURI path

1928+ then path

1929+ else

1930+ case takeDirectory fp of

1931+ "" -> path

1932+ "." -> path

1933+ d -> T.pack d <> "/" <> path

1934+

19141935image :: PandocMonad m => MarkdownParser m (F Inlines)

19151936image = try $ do

19161937 char '!'

19171938 (lab,raw) <- reference

19181939 defaultExt <- getOption readerDefaultImageExtension

1919− let constructor attr' src = case takeExtension (T.unpack src) of

1920− "" -> B.imageWith attr' (T.pack $ addExtension (T.unpack src)

1921− $ T.unpack defaultExt)

1922− _ -> B.imageWith attr' src

1940+ let constructor attr' src =

1941+ case takeExtension (T.unpack src) of

1942+ "" -> B.imageWith attr' (T.pack $ addExtension (T.unpack src)

1943+ $ T.unpack defaultExt)

1944+ _ -> B.imageWith attr' src

19231945 regLink constructor lab <|> referenceLink constructor (lab,raw)

19241946

19251947note :: PandocMonad m => MarkdownParser m (F Inlines)

test/command/3752.mdnew file+35

@@ -0,0 +1,35 @@

1+```

2+% pandoc command/chap1/text.md command/chap2/text.md -f markdown+rebase_relative_paths --verbose -t docx | pandoc -f docx -t plain

3+^D

4+[INFO] Loaded command/chap1/spider.png from ./command/chap1/spider.png

5+[INFO] Loaded command/chap1/../../lalune.jpg from ./command/chap1/../../lalune.jpg

6+[INFO] Loaded command/chap2/spider.png from ./command/chap2/spider.png

7+Chapter one

9+A spider: [spider]

10+

11+The moon: [moon]

12+

13+Link to spider picture.

14+

15+URL left alone: manual.

16+

17+Absolute path left alone: absolute.

18+

19+Chapter two

20+

21+A spider: [spider]

22+```

23+

24+```

25+% pandoc command/chap1/text.md command/chap2/text.md -f markdown+rebase_relative_paths -t html

26+^D

27+<h1 id="chapter-one">Chapter one</h1>

28+<p>A spider: <img src="command/chap1/spider.png" alt="spider" /></p>

29+<p>The moon: <img src="command/chap1/../../lalune.jpg" alt="moon" /></p>

30+<p>Link to <a href="command/chap1/spider.png">spider picture</a>.</p>

31+<p>URL left alone: <a href="https://pandoc.org/MANUAL.html">manual</a>.</p>

32+<p>Absolute path left alone: <a href="/foo/bar/baz.png">absolute</a>.</p>

33+<h1 id="chapter-two">Chapter two</h1>

34+<p>A spider: <img src="command/chap2/spider.png" alt="spider" /></p>

35+```

test/command/chap1/spider.pngnew file

Binary file not shown.

test/command/chap1/text.mdnew file+11

@@ -0,0 +1,11 @@

1+# Chapter one

3+A spider: ![spider](spider.png)

5+The moon: ![moon](../../lalune.jpg)

7+Link to [spider picture](spider.png).

9+URL left alone: [manual](https://pandoc.org/MANUAL.html).

10+

11+Absolute path left alone: [absolute](/foo/bar/baz.png).

test/command/chap2/spider.pngnew file

Binary file not shown.

test/command/chap2/text.mdnew file+3

@@ -0,0 +1,3 @@

1+# Chapter two

3+A spider: ![spider](spider.png)