On Sun, Sep 08, 2019 at 12:56:44PM -0400, Scott Kostyshak wrote: > 1. Would it be more efficient to use "prefixIs()" instead of > "contains()"? Are you concerned that for some older versions there is > white space at the beginning of the line? Since this is a new feature,
I'm not sure what exact line of code you are refering to, but in some cases the same err message get prepended with different strings (like "Natbib Warning:") or similar. Given that I am aware of both prefixIs/contains routines I find it more likely that contains() was actually needed :) > 2. I think we can extend this to work with biblatex. For example, > currently if there is a citation of multiple keys, if one of the keys is > undefined, there is no error. There is no instance of "There were > undefined citations" in the log file. Instead, the following is in the > .log file: > > Package biblatex Warning: The following entry could not be found > (biblatex) in the database: > (biblatex) abc123 > (biblatex) Please verify the spelling and rerun > (biblatex) LaTeX afterwards. > > The attached patch fixes this partly. It gives an error, but it does not > specify the name of the missing key in the log file. If you think you > can extend the patch to also report the missing key in the log file, > please go ahead (I'm not interested in spending more time on this at the > moment). Otherwise, should I at least commit this attached patch? I can > change the prefixIs() to contains() if you prefer. I never worked with biblatex, so feel free to update the code any way you see fit. If you want me to check it I would need example file(s) how to reproduce your scenario. In any case it's unlikely I will have time for any coding in the forthcoming month... Pavel