On 12 Jan 2010, at 13:58, Andrew Hill wrote:
[...] Anyway, it's the 0x0D character (carriage return) — I suspect my browser or the clip-board decided to clean up the line endings of the copied/ pasted text by converting CR to LF before...
So it's picking up the CR char because my .tex file has CRLF line endings, and I guess the ruby script is finding ends of lines by LF character only. To test this, I converted the document to LF line endings and the original ruby script worked fine (as did the one with the .strip addition). With the CRLF line endings, the .strip (or better yet a removal of just trailing CR) is required. Alternatively, can ruby deal with CRLF line endings on its own?
For completeness I tried saving my document with CR line endings, and the error message was basically a re-paste of my entire document as the filename (i.e. Ruby is not parsing the CR chars as line endings, only LFs). Though I doubt anyone really uses CR line endings...
We should be able to make it robust with respect to line endings, for now I added an issue about it: http://github.com/textmate/latex.tmbundle/issues/1