From ef4efa5373a419edbb99355808ddc63d35ddef20 Mon Sep 17 00:00:00 2001
From: John MacFarlane <jgm@berkeley.edu>
Date: Thu, 19 Aug 2021 10:49:20 -0700
Subject: [PATCH] Improve docx reader's robustness in extracting images.
The docx reader made a couple assumptions about how docx
containers were laid out that were not always true, with
the result that some images in documents did not get
found/extracted.
Closes #7511.
---
src/Text/Pandoc/Readers/Docx/Parse.hs | 11 ++++++-----
1 file changed, 6 insertions(+), 5 deletions(-)
diff --git a/src/Text/Pandoc/Readers/Docx/Parse.hs b/src/Text/Pandoc/Readers/Docx/Parse.hs
index dbb16a821..eb048ab14 100644
--- a/src/Text/Pandoc/Readers/Docx/Parse.hs
+++ b/src/Text/Pandoc/Readers/Docx/Parse.hs
@@ -507,9 +507,7 @@ archiveToRelationships archive docXmlPath =
filePathIsMedia :: FilePath -> Bool
filePathIsMedia fp =
- let (dir, _) = splitFileName fp
- in
- (dir == "word/media/")
+ "media" `elem` splitPath (takeDirectory fp)
lookupLevel :: T.Text -> T.Text -> Numbering -> Maybe Level
lookupLevel numId ilvl (Numbering _ numbs absNumbs) = do
@@ -774,8 +772,11 @@ expandDrawingId s = do
target <- asks (fmap T.unpack . lookupRelationship location s . envRelationships)
case target of
Just filepath -> do
- bytes <- asks (lookup ("word/" ++ filepath) . envMedia)
- case bytes of
+ media <- asks envMedia
+ let filepath' = case filepath of
+ ('/':rest) -> rest
+ _ -> "word/" ++ filepath
+ case lookup filepath' media of
Just bs -> return (filepath, bs)
Nothing -> throwError DocxError
Nothing -> throwError DocxError