LilyIssues / Issues / #855 fix the command index

Google Importer - 2010-10-22

Originally posted by: percival.music.ca@gmail.com

If we want to keep both \foo and foo in the index, then convert-ly needs to be adjusted. See issue 788.

Owner: ---

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Google Importer - 2010-10-24

Originally posted by: v.villenave

Here's what I suggested in comment 4 of issue 788:

- where appropriate, we *only* index \foo entries. no more plain "foo"
- in texi2html.init, we rewrite the @printindex command so that \foo is ordered alphabetically just as if the backslash wasn't there
- therefore the index will no longer have this huge list under "\" (which is totally useless), but under the letter "A", for instance,
   annotate-spacing
   \autochange
   auto-first-page-number
   \accepts
etc.
will all be listed.

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Google Importer - 2010-10-24

Originally posted by: v.villenave

Oh, that's *exactly* what Mark said in his e-mail. Well, clever me :)

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Google Importer - 2010-12-21

Originally posted by: markpole...@gmail.com

I agree with Graham's comment #5 in issue 788 - modifying
lilypond-texi2html.init is not the right solution since it
doesn't address the info/PDF docs.

Regarding this:
@funindex \foo
@funindex foo
We can take a formal vote if necessary, but does it looks
like most of us already favor keeping only the "\foo"
version?

Also, is anybody even remotely interested in tackling the
possibility of ignoring "\" during the sorting of the index?
annotate-spacing
\autochange
auto-first-page-number
\accepts
I haven't heard a peep from anyone, and this has been around
for long enough.

********************

"Graham Percival" wrote:
>> Questions:
>> 1) Everything marked with a @funindex in the docs ends
>>    up in *both* NR indices. How and why do these items
>>    end up in appendix F?
>
> Because @funindex calls both @findex and @kindex, and we
> merge a copy of @findex with @kindex.
>
>> 2) Why do we need @funindex? Why don't we just use
>>    these:
>>      @findex
>>      @printindex fn
>
> Because then anything in appendix E won't appear in
> appendix F, and (at some point in our past) there was a
> feeling that commands should be printed in both indices.

Ugh. Who here wants commands printed in both indices?
It's so overstuffed already. A simple note with a
@ref{LilyPond command index} at the top of the general
index would be so much better, IMO. For me, the presence
of the commands in the general index is nothing but a
nuisance. The general index in PDF is 20 pages long, and
almost half of the entries are duplicated from the command
index.

In fact, I'd prefer to see three indices:

"Command index", syntactic items not starting with a letter:
@findex !
@findex \abs-fontsize
...
@printindex fn

"Variable index", syntactic items starting with a letter:
@vindex alignAboveContext
@vindex beamExceptions
...
@printindex vr

"General index", concepts *only*:
@cindex beaming in cadenzas
@cindex page breaks
...
@printindex cp

Does anyone else think this would be nice? As it stands
it's too easy to miss something like "beamExceptions" in
the command index since it's not betweem "\beam" and
"\bendAfter". And two @ref's at the top of each index (to
the other indices) would be easy to spot. And a decent
sed script could do most of the work.

Is now a bad time to resurrect this?
- Mark

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Google Importer - 2010-12-21

Originally posted by: v.villenave

I'm not sure that having a "variable" index would make much sense to most users. I think separating more clearly the "General concept" index from the "LilyPond index" would be welcome (i.e. "what you want to achieve" vs "what you have to type to achieve it").

Re. the backslash-alphabetic-order problem, I have been desperately trying to achieve it (I suspect a simple one-line regex would do the trick), but without any luck so far. Even more so when you consider that it should be handled by both texi2html and texi2pdf.

I agree that anything that comes with a backslash *should* be indexed with its backslash. If we ever get around this alphabetic order problem, that is.

... And, yes, I suspect it's a terrible time to revive this issue. But I'm glad I'm not the only one seeing this as a painful splinter in LilyPond's foot :-)

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Google Importer - 2010-12-22

Originally posted by: percival.music.ca@gmail.com

"is now a bad time"? on a personal note, definitely, since I have a bad fever, am in pain, and have less patience than normal. On a project-wide note, I think that any new initative should be balanced against the time+effort it would detract from the release-oriented-ness. I'm not saying that we should *never* try anything new until 2.14 is out (after all, the lilybuntu 3 stuff definitely isn't release-oriented, and has definitely postponed 2.14!), but that's the consideration I would be making.

re \foo vs. foo: at least one person complained most voiciferously about the suggestion of not including "foo". He never knew if the thing he was looking up was a \text or whether it was just text. IIRC that happened near the end of GDP -- please look up that discussion and give an argument about why he was wrong.

re ignore \ in the sorting: this won't be hard. Just go and talk to the texinfo people; I'm sure they'd be willing to give you pointers to the relevant part of your code. Say, 5 hours for a semi-competent C programmer with no previous experience with their code base?
admittedly, you might need to do the same thing for texi2html... it kind-of depends how far the texi2html+texinfo integration has gone. Of course, this would require either waiting until the next stable texinfo release before we can use it, or requiring the use of a cvs-version texinfo in order to build our docs. I don't favor the latter.

re commands in both indices: somebody definitely wanted this. I didn't add it for giggles; it must have been a significant number of users, or one use whom I respected. Try looking at when that was added, then look at the mailing list archives for a couple of months prior to that. Or maybe you can find it directly just with a few searches.

Finally, I don't recommend making any major changes to the index organization without pretending to ask for opinions on -user. That way, when people complain (and somebody *will* complain), you can point back at that discussion and say "see? people thought it was a good idea". of course, that assumes that the discussion doesn't go complete against you; when that happens, you need to wait a few months and the rephrase the question so that you get the answer you want. (woot! I'm playing Quebec referendum politics!) or just change your mind, of course.

Oh, you could just add a third index, and make them:
command-index
concept-index
unified-index
I wouldn't consider that to be a "major" change, and I can't see anybody seriously complaining about that.

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Google Importer - 2010-12-24

Originally posted by: tdanielsmusic

Just to place my opinion on record, I would prefer a single unified index with the leading backslash ignored in sorting, using differing typefaces to distinguish commands, concepts and other syntactic items. Why? Well, if I want to find something about beams that I know exists but can't remember the word for it I want to see everything beginning with "beam" in one index, rather than having to look in two or three. I'd also be happy to see all the @cindex duplications of @funindex entries removed right away.

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Google Importer - 2011-06-18

Originally posted by: pkx1...@gmail.com

I am probably way out of my depth here, but I came across this (after reading Valentin's comment about how to combine the \foo so that it sits between 'fon' and 'fop'.

http://www.cs.utah.edu/dept/old/texinfo/texinfo/texi_13.html#SEC123

I cannot see any specific mention about trailing '\' but as I say, I am a bit out of my depth here. But could this be a relatively simple thing with @syncodeindex function?

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Google Importer - 2011-06-18

Originally posted by: percival.music.ca@gmail.com

Please don't link to old texinfo manuals. That page is from 3 March 1994 !
@syscodeindex has nothing to do with this.

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Google Importer - 2012-06-24

Originally posted by: colingh...@gmail.com

User Federico Bruni sent in a related bug report here:

http://lists.gnu.org/archive/html/bug-lilypond/2012-06/msg00146.html

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

fix the command index

Issue Tracker for LilyPond

Searches

Help

#855 fix the command index

Related

Discussion