Bogus assumption prevents d2u/u2d/conv/etal working on mixed files.

[Chris Murray]

FWIW,
I would just like to chime in with one instance that I know of where you 
may have reason to expect to see files that have mixed EOL 
characters.   The scenario is this: Person develops a code under cygwin and 
is really careful to ensure all files use /n endings.  Person distributes 
said code to others.  One of those individuals uses Visual Studio to edit 
the source files for whatever reason.  They make changes, run a contextual 
diff to generate a patch file and submit that patch file to the 
developer.  The lines in the context diff that detail the changes now have 
the /r/n EOL chars for the modified lines instead of /r that the originals 
had.  The developer, before applying the patches will need to strip these 
characters to make sure the patch process works properly and doesn't 
introduce /r/n accidentally.  It would be nice not having to create a 
separate tool to do this.  (I wrote one in perl and would be happy to 
share, although it also strips out trailing whitespace and converts tabs to 
spaces - but that functionality could be removed rather easily)

My opinion is that having a d2u tool with a --force option (or perhaps 
"--mixed" is a bit more conceptually consistent) is a reasonable solution 
to all concerned.

Chris


--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/