Re: Why does etags ask me if I want to keep the old TAGS - I do not want to load a new

[Francesco Potorti`]

Sorry for the delay of this answer.

>I work on a software project with the source files in quite some
>directories. I create my TAGS table in
>k:/ERADIUM_kalle/LD_swmodules_005 and work in another. The first time
>I use "M-.", I am asked for the tags table to use. After responding to
>this I am taken to the definition.
>
>If I then use "M-." to get to another definition, I am asked "Keep
>current list of tags tables also?".
>If I have the *Message*-buffer in another window, I can now see this:
>k:/ERADIUM_kalle/LD_swmodules_005/TAGS and
>k:/ERADIUM_kalle/LD_SwModules_005/TAGS are the same file [2 times]

Notice how the second file name's capitalisation is different from the
first one: "SwModules" versus "swmodules".  This is probably where the
problem comes from.

>Since I haven't been able to find a way to produce TAGS-tables on
>windows 2000 recursively, I use "ctags -e" with ctags from
>"http://ctags.sourceforge.net/";. Thes has worked flawlessly for me for
>me for years. The problems have started rather recently.

Do you mean that the problem was introduced with Emacs 22?

Alternatively, it could be that ctags has changed recently, and the file
names it stores in the TAGS file have different capitalisation than
before.  Can you possibly check if this is the case?

One problem for me to reproduce the bug is that it is probably specific
to the Windows platform, which I use only rarely and only as a user, not
as a developer.

As a side note: you can use etags for generating TAGS files recursively
if you have the program "find", like this:
     find . -name "*.[chCH]" -print | etags -

NOTE TO DEVELOPERS:
I plan to add file recursion capability to etags, but I could not find
the time until now.  Any help for implementing this is appreciated: even
a pointer to portable code to recursively read file names would help a
lot.