Community technical support mailing list was retired 2010 and replaced with a professional technical support team. For assistance please contact: Pre-sales Technical support via email to sales@march-hare.com.
Tony, I'm currently rewriting TortoiseCVS's parser for "cvs log", and I think it's not 100% foolproof. Let's look at an example: <--snip--> RCS file: /Test/Module 4/A2.txt,v Working file: A2.txt ... description: This is my file ---------------------------- revision 1.2 locked by: hhonisch; date: 2004/02/18 15:29:14; author: hhonisch; state: Exp; lines: +2 -2 My second revision ---------------------------- revision 1.1 locked by: hhonisch; date: 2004/02/07 10:01:35; author: hhonisch; state: Exp; My first revision ============================================================================ = <--snip--> CVS uses 28 '-' as delimiter for multiline comments, or 77 '=' for the end of the log. Assume that a user enters exactly those delimiters in a revision comment or for the file description: <--snip--> RCS file: /Test/Module 4/A2.txt,v Working file: A2.txt ... description: This is my file ---------------------------- I like dashes ---------------------------- revision 1.2 locked by: hhonisch; date: 2004/02/18 15:29:14; author: hhonisch; state: Exp; lines: +2 -2 My second revision ---------------------------- I like dashes ---------------------------- revision 1.1 locked by: hhonisch; date: 2004/02/07 10:01:35; author: hhonisch; state: Exp; My first revision ============================================================================ = I also like equals ============================================================================ = <--snip--> Do you think that CVS scould escape those delimiters if they show up in the file description or the revision comment? Or am I too paranoid here? -Hartmut