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

Docx Reader: Parse soft, no-break hyphen elements

Browse source
This commit is contained in:
Ophir Lifshitz 2015-10-04 06:07:23 -04:00
parent 6d0e586b2d
commit 0b899ce7ef
2 changed files with 7 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
src/Text/Pandoc/Readers/Docx.hs
View file

@ -206,11 +206,15 @@ runElemToInlines :: RunElem -> Inlines
runElemToInlines (TextRun s) = text s
runElemToInlines (LnBrk) = linebreak
runElemToInlines (Tab) = space
runElemToInlines (SoftHyphen) = text "\xad"
runElemToInlines (NoBreakHyphen) = text "\x2011"
runElemToString :: RunElem -> String
runElemToString (TextRun s) = s
runElemToString (LnBrk) = ['\n']
runElemToString (Tab) = ['\t']
runElemToString (SoftHyphen) = ['\xad']
runElemToString (NoBreakHyphen) = ['\x2011']
runToString :: Run -> String
runToString (Run _ runElems) = concatMap runElemToString runElems

4
src/Text/Pandoc/Readers/Docx/Parse.hs
View file

@ -208,7 +208,7 @@ data Run = Run RunStyle [RunElem]
| InlineDrawing FilePath B.ByteString
deriving Show
data RunElem = TextRun String | LnBrk | Tab
data RunElem = TextRun String | LnBrk | Tab | SoftHyphen | NoBreakHyphen
deriving Show
data VertAlign = BaseLn | SupScrpt | SubScrpt
@ -877,6 +877,8 @@ elemToRunElem ns element
map (\x -> fromMaybe x . getUnicode f . lowerFromPrivate $ x) str
| isElem ns "w" "br" element = return LnBrk
| isElem ns "w" "tab" element = return Tab
| isElem ns "w" "softHyphen" element = return SoftHyphen
| isElem ns "w" "noBreakHyphen" element = return NoBreakHyphen
| isElem ns "w" "sym" element = return (getSymChar ns element)
| otherwise = throwError WrongElem
where