11945ea · typeable/pandoc

git.iko.soy/typeable/pandoc

Docx reader: Only use LTR when it is overriding BiDi setting

The left-to-right direction setting in docx is used in the spec only for overriding an explicit right-to-left setting. We only process it when it happens in a paragraph set with BiDi. This is especially important for docs exported from Google Docs, which explicitly (and unnecessarily) set "rtl=0" for every paragraph. Closes: #5723

commit11945ea5ec0a9ed42b0c24f6f9ecff69e86784a1 authorJesse Rosenthal <jrosenthal@jhu.edu> dateSun 3 Nov 2019, 21:25 (7 years ago) parenta6612b3

3 files changed+14−2

compare with master

unifiedsplit

src/Text/Pandoc/Readers/Docx.hs+10−2 src/Text/Pandoc/Readers/Docx/Parse.hs+3 src/Text/Pandoc/Readers/Docx/Parse/Styles.hs+1

src/Text/Pandoc/Readers/Docx.hs+10−2

@@ -124,10 +124,11 @@ instance Default DState where

124124

125125data DEnv = DEnv { docxOptions :: ReaderOptions

126126 , docxInHeaderBlock :: Bool

127+ , docxInBidi :: Bool

127128 }

128129

129130instance Default DEnv where

130− def = DEnv def False

131+ def = DEnv def False False

131132

132133type DocxContext m = ReaderT DEnv (StateT DState m)

133134

@@ -290,7 +291,10 @@ runStyleToTransform rPr

290291 return $ spanWith ("",[],[("dir","rtl")]) . transform

291292 | Just False <- isRTL rPr = do

292293 transform <- runStyleToTransform rPr{isRTL = Nothing}

293− return $ spanWith ("",[],[("dir","ltr")]) . transform

294+ inBidi <- asks docxInBidi

295+ return $ if inBidi

296+ then spanWith ("",[],[("dir","ltr")]) . transform

297+ else transform

294298 | Just SupScrpt <- rVertAlign rPr = do

295299 transform <- runStyleToTransform rPr{rVertAlign = Nothing}

296300 return $ superscript . transform

@@ -561,6 +565,10 @@ normalizeToClassName = map go . fromStyleName

561565

562566bodyPartToBlocks :: PandocMonad m => BodyPart -> DocxContext m Blocks

563567bodyPartToBlocks (Paragraph pPr parparts)

568+ | Just True <- pBidi pPr = do

569+ let pPr' = pPr { pBidi = Nothing }

570+ local (\s -> s{ docxInBidi = True })

571+ (bodyPartToBlocks (Paragraph pPr' parparts))

564572 | isCodeDiv pPr = do

565573 transform <- parStyleToTransform pPr

566574 return $

src/Text/Pandoc/Readers/Docx/Parse.hs+3

@@ -224,6 +224,7 @@ data ParagraphStyle = ParagraphStyle { pStyle :: [ParStyle]

224224 , indentation :: Maybe ParIndentation

225225 , dropCap :: Bool

226226 , pChange :: Maybe TrackedChange

227+ , pBidi :: Maybe Bool

227228 }

228229 deriving Show

229230

@@ -232,6 +233,7 @@ defaultParagraphStyle = ParagraphStyle { pStyle = []

232233 , indentation = Nothing

233234 , dropCap = False

234235 , pChange = Nothing

236+ , pBidi = Just False

235237 }

236238

237239

@@ -998,6 +1000,7 @@ elemToParagraphStyle ns element sty

9981000 isElem ns "w" "moveFrom" e

9991001 ) >>=

10001002 getTrackedChange ns

1003+ , pBidi = checkOnOff ns pPr (elemName ns "w" "bidi")

10011004 }

10021005elemToParagraphStyle _ _ _ = defaultParagraphStyle

10031006

src/Text/Pandoc/Readers/Docx/Parse/Styles.hs+1

@@ -40,6 +40,7 @@ module Text.Pandoc.Readers.Docx.Parse.Styles (

4040 , getNumInfo

4141 , elemToRunStyle

4242 , defaultRunStyle

43+ , checkOnOff

4344 ) where

4445import Prelude

4546import Codec.Archive.Zip