[issue1957] win32text need be more smart encode/decode

[AlexRayne]

New submission from AlexRayne <AlexRaynePE196 at lavabit.com>:

Hallow.
i have an mercurial repo that is periodicaly synronise with svn-repo. 
because another developer works on linux we loose original lineendings on
many files and them have got an bad commited into svn-repo.
now i try to import them.
to supress this proplem some i set property eol-style=native on sources 
files in svn-repo, and updating from it provide me ones files with CRLF for 
me and LF for my linux-coworker, in svn repo this files are converted to LF 
endings, but this is not a problem for me. So is solved this problem in svn.

now i try to import changes from svn-repo to my loral hg-repo.
i try to use win32text extention by selecting extentions wich is used for 
dumbencode and thouse gonna fail, because some files was with LF ending, 
but most with CRLF.
thus, because dumbencode converts windows-linux ending regardless to wich 
sendstyle was in repository, i got conflicts on files wich not require 
convertion.

so here i asking for a wish - make dumbencode a bit smarter: look to
repository version file, check it - what lineendings it use, and convert
only if required. and same behaviour for decode.

----------
messages: 11275
nosy: AlexRayne
priority: wish
status: unread
title: win32text need be more smart encode/decode

____________________________________________________
Mercurial issue tracker <bugs at mercurial.selenic.com>
<http://mercurial.selenic.com/bts/issue1957>
____________________________________________________