Re: Copyright years update in etags test suite

[Eli Zaretskii]

> Cc: address@hidden
> From: Paul Eggert <address@hidden>
> Date: Sat, 30 Jan 2016 10:18:29 -0800
> 
> Eli Zaretskii wrote:
> > I fixed this, but I wonder whether we really need to do these updates
> > in the etags test suite.  The source files there are not maintained in
> > any way, so perhaps they could be skipped?
> 
> Although I suppose skipping could be arranged in the procedure that updates 
> copyrights, this would be a change to copyright-notice policy which would 
> probably require going through channels and getting it approved.

I understand.

> > If we do have to update the Copyright there, then the annual update
> > procedure should include the corresponding update to the
> > expected-results files.
> 
> How about another idea: fix tests so that they are robust in the presence of 
> changes to copyright notices. After all, tests are supposed to be testing 
> features, not copyrights. I don't know which tests were failing and why, so I 
> don't know how much work this would be.

All of the etags tests failed, because the TAGS file records both line
numbers and character positions of tags.  The latter changed for tags
from any test file that has an FSF Copyright notice; the former
changed for tags from a few test files that inserted an extra line
(because the Copyright notice became too long for a single line).

Fixing the tests for such effects would mean somehow parsing the TAGS
files, extracting the numbers, and correcting them for the inserted
text, which sounds like a PITA for something that happens once a year.
Is there an easier way?