Hi all,
>>>>> Arash Esbati <[email protected]> writes:
> Andreas Matthias <[email protected]> writes:
>> The following regexp find LuaTeX's TeX-warnings:
>>
>> "\\(?:LaTeX\\|Class\\|Package\\|pdfTeX\\|\\*\\)? ?[-A-Za-z0-9]*
>> ?[Ww]arning\\(?: +([- A-Za-z0-9]+)\\)?:"
> This is probably a little safer in case we have non-ASCII chars in the log:
> "\\(?:LaTeX\\|Class\\|Package\\|pdfTeX\\|\\*\\)? ?[-[:alnum:]]*
> ?[Ww]arning\\(?: +([- [:alnum:]]+)\\)?:"
> And I think we can reduce this to:
> "\\(?:LaTeX\\|Class\\|Package\\|pdfTeX\\|\\*\\)?[- [:alnum:]]*[Ww]arning\\(?:
> +([^)]+)\\)?:"
Unfortunately, it seems that we have to reconsider the implementation.
`LaTeX-warning-regexp' is used in `TeX-warning', where we read
----------------------------------------------------------------------
(context (cond ((string-match LaTeX-warnings-regexp warning)
;; The warnings matching `LaTeX-warnings-regexp' are
;; emitted by \GenericWarning macro, or macros based on
;; it (\ClassWarning, \PackageWarning, etc). After
;; such warnings there is an empty line, just look for
;; it to find the end.
(beginning-of-line)
(while (null (eolp))
(forward-line 1))
(buffer-substring context-start (progn (end-of-line)
(point))))
----------------------------------------------------------------------
This "empty line" doesn't exist for the current pdfTeX warning, so this
code returns wrong result for our current tentative work around. :-(
Non-trivial work is needed here. I don't have a good idea yet.
Regards,
Ikumi Keita
#StandWithUkraine #StopWarInUkraine
#Gaza #StopMassiveKilling #CeasefireNOW
