Re: [PATCH] Pass "-O xhtml" param to highlight instead of "-xhtml"
- From
- Adam Tkac <atkac@redhat.com>
- Date
- Jan 27, 2011, 12:03 UTC
- Message-ID
- <20110127120323.GA3416@traged.englab.brq.redhat.com>
- In-Reply-To
- <7vaainjq7f.fsf@alter.siamese.dyndns.org>
On Wed, Jan 26, 2011 at 01:43:48PM -0800, Junio C Hamano wrote:
Show 20 quoted lines
> Jakub Narebski <jnareb@gmail.com> writes: > > >> Current highlight utility doesn't recognize "--xhtml" parameter, it > >> recognizes only "-O xhtml" parameter. > >> > >> Reference: https://bugzilla.redhat.com/show_bug.cgi?id=672293 > > > > Shouldn't the above be in commit message? > > Actually, I prefer not having that "Reference" in the commit message > myself, but I do want to have some relevant details > missing from the proposed commit message but is in > that ticket (especially the comment #5) in the commit message. > > > In highlight 2.4.5 '-O' means "name of output directory", i.e. --outdir. > > There is no --out-format either. > > Thanks for digging about 2.4.5; in that case, unlike the redhat ticket > hinted, this change may not be backward compatible enough, as it seems > that the compatibility goes only back to 3.0.something.
Right you are, I digged too shallowly, thanks for your inspection.
Show 17 quoted lines
> > > > > WTF this backward incompatibile change in highlight... the only > > solution that would make it work both for old and for new versions is > > to rely on the fact that HTML is default output format, i.e. > > > > open $fd, quote_command(git_cmd(), "cat-file", "blob", $hash)." | ". > > quote_command($highlight_bin). > > - " --xhtml --fragment --syntax $syntax |" > > + " --fragment --syntax $syntax |" > > or die_error(500, "Couldn't open file or run syntax highlighter"); > > Adam, Jochen? How does the counterproposal look to you? > > Without knowing much about highlight nor using gitweb in general myself, > the above looks a bit more reasonable fix to the issue, if the default > format has been and will stay to be HTML.
The proposal looks like the best (and the only one) correct solution. I will send updated patch.
Regards, Adam
-- Adam Tkac, Red Hat, Inc.