From: Jakub Narebski Date: Tue, 18 Dec 2007 00:32:09 GMT Subject: Re: [PATCH 5/6] whitespace: more accurate initial-indent highlighting Message-ID: In-Reply-To: <7vwsrdd9wa.fsf@gitster.siamese.dyndns.org> Junio C Hamano writes: > Wincent Colaiuta writes: > > > Basically I would have proposed extracting out each type of whitespace > > error into an inline function in ws.c, where it could be used by both > > check_and_emit_line() in ws.c and apply_one_fragment() in builtin- > > apply.c. > > > > Unfortunately, mixing checking and emission phases makes this proposed > > refactoring a little bit ugly. > > The right refactoring would be what JBF hinted in his message, to record > and return a list of suspicious ranges from the checker function and > have the highlighter and the fixer make use of that list. > > Such a refactoring is still possible but I think it is beyond the scope > of pre 1.5.4 clean-up. By the way, does "trailing empty lines at the end of file" whitespace error get detected and hightlighted with refactored whitespace checking? -- Jakub Narebski Poland ShadeHawk on #git