Mosè Giordano <[email protected]> writes:

>> Why two alists and not just one with (FILE MD5 CHANGED) entries?
>
> Uhm, I think that would be a bit awkward: `TeX-run-TeX' updates md5
> hash without touching the changed status, instead `TeX-LaTeX-sentinel'
> and `TeX-run-index' update the changed status without touching the
> hash.

I don't see why this would be awkward.  But another alternative was not
to have the notion of CHANGED at all but to store entries of the form
(FILE CURRENT-MD5 PREV-MD5).  Then only `TeX-run-TeX' has to

  (setcdr entry (list new-md5 (cadr entry)))

and nowhere else has the alist (or another one) to be modified.

> That's why I kept those alists separated, but I can try to merge them
> anyway.

I actually don't have a strong opinion on this so feel free to keep it
the way it is.

>> Storing them buffer-locally in the master file's buffer would also
>> work, no?
>
> Yes, but there is the same shortcoming as storing the variable locally
> in idx file: the buffer shouldn't be killed, and as far as I know we
> don't require master file's buffer to be always alive.

That's right.  Then I don't know anything better.

Bye,
Tassilo

_______________________________________________
auctex-devel mailing list
[email protected]
https://lists.gnu.org/mailman/listinfo/auctex-devel

Reply via email to