Apache OpenOffice (AOO) Bugzilla – Issue 70630
wrong encoding in LaTeX export
Last modified: 2013-07-30 02:39:42 UTC
When I export a german .odt document to LaTeX the following laTeX preamble entries are created: \usepackage[ascii]{inputenc} \usepackage[T1]{fontenc} \usepackage[ngerman]{babel} The first line is not correct as the encoding "ascii" is chosen although babel is correctly loaded with the parameter for german. The wrong encoding leads to LaTeX errors for every umlaut character (ä, ö, ü) in the document. The correct encoding for german documents is \usepackage[latin1]{inputenc} or \usepackage[cp1252]{inputenc} OOo should recognize the correct encoding for every document language like it currently already does this for the babel package.
Reassigned to JW
次の添付ファイルを作成しました。 (id=40150) Writer file containing Japanese characters.
次の添付ファイルを作成しました。 (id=40151) wrong export of Japanese of former document (latex.odt)
次の添付ファイルを作成しました。 (id=40152) possible correct export in ISO-2022-JP
hi, this does also happens for documents containing Japanese. example files are attached.
IMHO, uwestoehr and maho addresses different aspects. For the uwestoehr's point, the author, Henrik Just, states in his web page that Writer2LaTeX supports latin1. http://www.hj-gym.dk/~hj/writer2latex/index3.html#features Discussion for the solution would be mainly about how to tell a type of language to the Writer2LaTeX filter. A code fragment can be found in xmerge/source/writer2latex/source/writer2latex/latex/style/I18n.java For Japanese, the uwestoehr's point might be (but, i am not sure): \usepackage[cp932]{inputenc} for Windows variants Japanese \usepackage[eucjp]{inputenc} for UNIX variants where $LANG=ja or ja_JP.eucJP or ja_JP.eucjp \usepackage[sjis]{inputenc} for UNIX variants where $LANG=ja_JP.PCK (PC Kanji code) \usepackage[utf8]{inputenc} for UNIX variants where $LANG=ja_JP.UTF-8 or ja_JP.utf8 \usepackage[iso-2022-jp]{inputenc} for some UNIX variants using JIS encoding (*1) *1: Japanese Industrial Standard For the maho's point, according to the web page, current Writer2LaTeX does not supports any Asian languages. It would be a sort of enhancement. For the phenomenon that a Japanese character gets converted into [xxxx?] notation, which maho addresses, I have made a patch that hides the problem. http://www.tora-japan.com/ooo/download/filters/Writer2LaTeX_by_Henrik_Just/OOo_2.0.4/
Yes, these are actually two independent rfe's * Support for automatic choice of inputenc based on document language (Currently ascii is chosen as default encoding to ensure that the user can read the LaTeX file with any text editor.) * Support for CJK languages (Currently only latin-cyrillic-greek scripts are supported)
henrik i reassign this issue to you. maho please open a new issue for your request.
Any progress with this bug?
Reset assignee on issues not touched by assignee in more than 1000 days.