[MERGE] Don't generate .html files for .txt's that are transcluded

[Eric Siegerman]

On Thu, 2009-03-26 at 15:05 -0500, John Arbash Meinel wrote:
> Which means that just writing a .txt file in doc/developers/* means
that
> it will get generated into .html, unless you explicitly exclude it.

Indeed.  That seemed an important feature to preserve; it's why I
opted for an exclude list in the first place.

(I considered generating the exclude list automatically, by
grepping for "..include" directives or something, but felt that
(a) that would be overengineering the thing, and (b) whatever
rule(s) I came up with, exceptions might some day be needed, at
which point the list generation would start to get pretty hairy.
So KISS won out.)


> So while I don't feel there is any harm in generating the .html (is
> there a reason you feel it should be omitted?)

Well, the existence of those redundant .html files isn't a huge
problem, but it seems a bit untidy, especially since they aren't
reachable from the index.  And it can be confusing.  (It
certainly confused me when I set out to index all the unreachable
stuff at Ian's request, and found that a bunch of the words were
in fact reachable even though the files containing them weren't :-/)

Besides, as you wrote in another thread:
> I'll mention that I just added "improved_chk_index.txt" and forgot to
> add a link to it. It might be nice if we had a non-human check for
this.
> But since there is the possibility of "include", etc, it seems
difficult
> to do.

If the files that shouldn't be indexed simply don't exist, one
big barrier to automatic checking goes away :-)

  - Eric