Run regression tests for lilypond-book (issue 2223).

Run regression tests for lilypond-book (issue 2223).

Re-enable running the regtests in input/regression/lilypond-book.
Fix various typos, clarify what each input file tests for and
how the output should look, and fix the following problems:

collated-files.tely:
  Was linking to the input files rather than the output files
  generated by lilypond-book. The input files were thus not
  tested individually as it should but as included files.
  Fixed in the local GNUmakefile by an explicit rule for
  collated-files.list which tracks the output files.

suffix-texi:
  Failed, was not being preprocessed by lilypond-book but simply
  copied to the output folder. Fixed in the local GNUmakefile by
  adding a specific make rule for it.

tex-twocolumn:
  Failed, image had full-page width.
  Fixed in the input file by moving \twocolumn to the preamble.
  Column width detection in the body of the tex document is not
  yet implemented.

texinfo-language-detection:
  Suspicious, had a weird @lydoctitle popping up.
  Fixed in the input file by defining the @lydoctitle macro.

[Julien Rioux, 2012-01-23 20:01:50]

Please review.

[Graham Percival, 2012-01-25 22:12:36]

LGTM

[Graham Percival, 2012-02-07 20:05:49]

http://codereview.appspot.com/5569045/diff/3001/input/regression/lilypond-boo...
File input/regression/lilypond-book/GNUmakefile (right):

http://codereview.appspot.com/5569045/diff/3001/input/regression/lilypond-boo...
input/regression/lilypond-book/GNUmakefile:19: OUT_TEXI_FILES +=
${TEXI_FILES:%.texi=$(outdir)/%.info}
how are the OUT_X_Y definitions used?  Is there a generic rule for OUT_X ? 
other than that, I can't see how they'd have any effect?

[Julien Rioux, 2012-02-07 20:20:35]

On 2012/02/07 20:05:49, Graham Percival wrote:
>
http://codereview.appspot.com/5569045/diff/3001/input/regression/lilypond-boo...
> File input/regression/lilypond-book/GNUmakefile (right):
> 
>
http://codereview.appspot.com/5569045/diff/3001/input/regression/lilypond-boo...
> input/regression/lilypond-book/GNUmakefile:19: OUT_TEXI_FILES +=
> ${TEXI_FILES:%.texi=$(outdir)/%.info}
> how are the OUT_X_Y definitions used?  Is there a generic rule for OUT_X ? 
> other than that, I can't see how they'd have any effect?

Yes those rules are in one of two dozen files in the make folder, I forgot which
one, but they are effective. They test that the files generated by -book are
compilable to .info and .pdf (as well as .html)

Cheers,
Julien

[hanwenn, 2020-02-22 22:54:40]

https://codereview.appspot.com/5569045/diff/3001/input/regression/lilypond-bo...
File input/regression/lilypond-book/GNUmakefile (right):

https://codereview.appspot.com/5569045/diff/3001/input/regression/lilypond-bo...
input/regression/lilypond-book/GNUmakefile:36: # Prevent parallel lilypond-book
instances for this subdir
I know this is a long time ago, but can you remember why this was added?

[Julien Rioux, 2020-02-25 05:16:48]

https://codereview.appspot.com/5569045/diff/3001/input/regression/lilypond-bo...
File input/regression/lilypond-book/GNUmakefile (right):

https://codereview.appspot.com/5569045/diff/3001/input/regression/lilypond-bo...
input/regression/lilypond-book/GNUmakefile:36: # Prevent parallel lilypond-book
instances for this subdir
On 2020/02/22 22:54:40, hanwenn wrote:
> I know this is a long time ago, but can you remember why this was added?

As far as I can remember, lilypond-book was not free of concurrency issues with
its use of a cache for snippets. Therefore the simplest to avoid any issue with
make -j was to disable concurrent calls to lilypond-book.