Am 25.02.20 um 16:27 schrieb Felix Schumacher: > Am 25.02.20 um 16:24 schrieb Christopher Schultz: >> Felix, >> >> On 2/25/20 04:22, Felix Schumacher wrote: >>> Am 24.02.2020 22:13, schrieb Christopher Schultz: Mark, >>> On 2/24/20 15:46, Mark Thomas wrote: >>>>>> On 24/02/2020 20:31, Christopher Schultz wrote: >>>>>>> Mark, >>>>>>> Why not use the full commit id instead of a prefix? >>>>>>> Couldn't some future commit conflict with some arbitrary >>>>>>> prefix? Or do I not know what the hell I'm talking about? >>>>>> No, you are spot on. The reason I used the prefix is that I >>>>>> was transcribing the commit IDs by hand so the prefixes were >>>>>> easier. We probably should use the full ID or at least a >>>>>> longer prefix. >>> I wonder if we could script this: grab a prefix, find the (one and >>> only one) commit in the repo and expand it. If there is a conflict >>> (or more than one match), emit an error and continue without >>> changing the commit id. >>> When I do this kind of thing for $work, I like to write scripts >>> that emit sed scripts. So you process e.g. security-8.html as >>> input, but emit something like this as output: >>> s/(\b)69c5608(\b)/\169c56080fb3355507e1b55d014ec0ee6767a6150\2/g >>> ... >>> You get a script that can be inspected, re-used and, even better, >>> it doesn't directly modify the input files. So you can even do >>> something like this: >>> $ expand-git-commit-ids.pl security-*.xml > expand.sed >>> $ sed -i .bak -f expand.sed security-*.xml >>> And then if you find other files where the same kind of thing needs >>> to be done, you can re-use the expand.sed script, or even (pun >>> intended) expand the sed script if necessary. >>>> I would use the full hash in the xml and use a substring-function >>>> in the xsl to shorten the hash for readability. >>>> No need for sed here :) >> This was to initially fetch the full hashes. In the XML, now, they are >> already shortened. > See my other mail :)
Changed with r1874502. Felix >> -chris >> >> --------------------------------------------------------------------- >> To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org >> For additional commands, e-mail: dev-h...@tomcat.apache.org >> > > --------------------------------------------------------------------- > To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org > For additional commands, e-mail: dev-h...@tomcat.apache.org > --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
