Nicholas Clark [Mon, 24 Jan 2011 15:47:50 +0000 (15:47 +0000)]
Convert Storable's store.t to Test::More
The non-obvious reason for the test count increasing from 20 to 21 is that the
file used to read:
print "not " unless open(OUT, 'store');
binmode OUT;
$r = fd_retrieve(::OUT);
print "not " unless defined $r;
print "ok 12\n";
Nicholas Clark [Mon, 24 Jan 2011 15:33:52 +0000 (15:33 +0000)]
Convert I18N::Colate's test to Test::More.
Nicholas Clark [Mon, 24 Jan 2011 11:06:50 +0000 (11:06 +0000)]
Merge perl_keyword.pl into regen/keywords.pl, to generate keywords.[ch]
Prepend + or - in front of all the keyword names in __DATA__ to mark weak and
strong keywords, needed for keyword.c
As keywords.c needs Devel::Tokenizer::C 0.05, not a core module (and not a
common module either) we can no longer run it as part of regen.pl. So store
the sha256 of the source in the generated files, and use this in to check that
they are not stale (in t/porting/regen.t)
Nicholas Clark [Mon, 24 Jan 2011 10:30:51 +0000 (10:30 +0000)]
Break out the generated function Perl_keywords() into keywords.c, a new file.
As it and Perl_yylex() both need FEATURE_IS_ENABLED, feature_is_enabled() is
no longer static, and the two macro definitions move from toke.c to perl.h
Previously, one had to cut and paste the output of perl_keywords.pl into the
middle of toke.c, and it was not clear that it was generated code.
Nicholas Clark [Mon, 24 Jan 2011 09:43:21 +0000 (09:43 +0000)]
Propagate 2 edits made in Perl_keywords() back to perl_keywords.pl
7918f24d20384771 added PERL_ARGS_ASSERT_KEYWORD and
9b38784150c0aca5 combined
Perl_ckwarn_d() and Perl_warner() into Perl_ck_warner_d().
Nicholas Clark [Mon, 24 Jan 2011 08:11:44 +0000 (08:11 +0000)]
regen/mk_PL_charclass.pl can test that l1_char_class_tab.h is up to date.
Nicholas Clark [Mon, 24 Jan 2011 08:00:47 +0000 (08:00 +0000)]
Move mk_PL_charclass.pl from Porting/ to regen/
Nicholas Clark [Mon, 24 Jan 2011 06:28:19 +0000 (06:28 +0000)]
Convert mk_PL_charclass.pl to use regen_lib.pl
Change it to read CaseFolding.txt from lib/unicore, instead of the file
installed with perl, so that it can run with an uninstalled perl.
Add "read only" editor blocks to l1_char_class_tab.h
Nicholas Clark [Sun, 23 Jan 2011 21:51:23 +0000 (21:51 +0000)]
Move the non-generated parts of l1_char_class_tab.h out into handy.h
Now the contents of l1_char_class_tab.h is only the output of
Porting/mk_PL_charclass.pl
Craig A. Berry [Mon, 24 Jan 2011 00:33:01 +0000 (18:33 -0600)]
Override _fixin_replace_shebang on VMS so MM->fixin works.
Script generation doesn't look like it's ever worked on VMS, but
this hasn't been that big a deal because all the utilities
distributed with the core have their own generation facilities
under utils/. However, Nicholas wants to put that directory on
a diet and have the scripts be generated by the modules that own
them.
So we have to get script generation via MM->fixin working, and we
do that here by prepending the same $Config{startperl} incantation
to the shebang line that we've been using in core for eons.
Tracks upstream commit:
https://github.com/Perl-Toolchain-Gang/ExtUtils-MakeMaker/commit/
14ff930dd862286d6dafe4228b3a72415b5df9a5
Craig A. Berry [Mon, 24 Jan 2011 00:30:56 +0000 (18:30 -0600)]
Split shebang line fix-up from file processing in MM->fixin().
By putting the shebang line gyrations on their own in the new
_fixin_replace_shebang method, we can override that part without
dragging along the file processing code.
Tracks upstream commit:
https://github.com/Perl-Toolchain-Gang/ExtUtils-MakeMaker/commit/
33d7db29a2adb5bbb12fb2a783063245fe58dfc1
Nicholas Clark [Sun, 23 Jan 2011 21:29:12 +0000 (21:29 +0000)]
t/porting/regen.t can test that regcharclass.h is up to date.
As regen/regcharclass.pl uses regen_lib.pl and rename_if_different(), all that
is needed is to pass it --tap on the command line.
Nicholas Clark [Sun, 23 Jan 2011 21:08:12 +0000 (21:08 +0000)]
Move regencharclass.pl from Porting/ to regen/
Nicholas Clark [Sun, 23 Jan 2011 21:03:05 +0000 (21:03 +0000)]
Convert regcharclass.pl to use regen_lib.pl
This results in small changes to the formatting of the generated comments
in regcharclass.h
David Leadbeater [Sun, 23 Jan 2011 20:39:28 +0000 (12:39 -0800)]
[perl #82650] Ignore DTrace build product
Nicholas Clark [Sun, 23 Jan 2011 19:15:23 +0000 (19:15 +0000)]
Avoid blank lines in META.yml
Nicholas Clark [Sun, 23 Jan 2011 18:42:39 +0000 (18:42 +0000)]
t/porting/regen.t can test that META.yml is up to date.
As Porting/makemeta uses regen_lib.pl and rename_if_different(), all that is
needed is to pass it --tap on the command line.
t/porting/regen.t now tests that all files mentioned in the release managers
guide are up to date (and tests will fail if they are not), hence there is now
no need for the release manager to explicitly test that generated files are
not stale.
Nicholas Clark [Sun, 23 Jan 2011 18:29:20 +0000 (18:29 +0000)]
Store the SHA-256 of the source in files generated by regen_perly.pl
bison isn't available everywhere, so we can't simply re-run regen_perly.pl to
verify that perly.{act,h,tab} are up to date. So instead store the SHA-256 of
the input files, and extend t/porting/regen.t to check that the input files
haven't been changed subsequently.
Nicholas Clark [Sun, 23 Jan 2011 16:25:30 +0000 (16:25 +0000)]
Output "read only" editor blocks from regen_perly.pl
Use safer_open() and read_only_bottom_close_and_rename() from regen_lib.pl
Consistently use 3 argument open and lexical file handles.
A side effect of this change is that the generated files are no longer made
read-only on disk - if this is desirable, then probably better to change
regen_lib.pl so that all generated files are made read-only.
Chris 'BinGOs' Williams [Sun, 23 Jan 2011 16:51:54 +0000 (16:51 +0000)]
Tracked down some other places to make the Unicode-Collate changes
Chris 'BinGOs' Williams [Sun, 23 Jan 2011 16:48:32 +0000 (16:48 +0000)]
Update Unicode-Collate to CPAN version 0.72
Second attempt to integrate the XS version of
Unicode::Collate into core.
[DELTA]
0.72 Sat Jan 22 17:28:32 2011
- xs: fix mixing char* and U8*.
0.71 Tue Jan 18 22:29:44 2011
- t/loc_test.t should not fail without Unicode::Normalize.
0.70 Sun Jan 16 20:31:07 2011
- Now U::C::Locale->new will use the compiled DUCET via XS if available.
added some tests in t/loc_test.t.
0.69 Sat Jan 15 19:41:11 2011
- clarified about XSUB. revised INSTALL in README.
- xs: flag passed to utf8n_to_uvuni().
- doc and comments: [perl #81876] Fix typos by Peter J. Acklam.
David Golden [Sun, 23 Jan 2011 15:53:40 +0000 (10:53 -0500)]
Update ExtUtils::CBuilder in Maintainers.pl
David Golden [Sun, 23 Jan 2011 15:29:41 +0000 (10:29 -0500)]
bump ExtUtils::CBuilder version numbers
Nicholas Clark [Sun, 23 Jan 2011 11:15:14 +0000 (11:15 +0000)]
In regen scripts, print to explicit file handles instead of using select.
Also put explicit quotes on heredoc declarations to show whether they should
interpolate, merge some heredocs, and remove & from calls to &tab(...)
Nicholas Clark [Sun, 23 Jan 2011 10:51:29 +0000 (10:51 +0000)]
Move the duplicated subroutine tab() into regen_lib.pl
Nicholas Clark [Sun, 23 Jan 2011 10:38:58 +0000 (10:38 +0000)]
Change close_and_rename() to read_only_bottom_close_and_rename()
All users of close_and_rename() were printing out the appropriate "ex: set ro:"
string to the file handle immediately before closing it. So move that into
the common function and rename it to reflect what it now does. [Except
overload.pl, which should have been, given that it calls read_only_top()]
Print a newline above the "ex: set ro:" line. This removes many newlines from
the regen scripts, but does add newlines to a couple of generated files.
Nicholas Clark [Sun, 23 Jan 2011 10:07:52 +0000 (10:07 +0000)]
In regen/*.pl, refactor the repeated code for close and rename if different.
Pass the final file name as an optional second argument of safer_open() and
store it with the file handle. Add a function close_and_rename() which closes
the file handle, then retrieves the final name, and renames the temporary file
if the two differ.
Nicholas Clark [Sun, 23 Jan 2011 09:39:21 +0000 (09:39 +0000)]
In regen_lib.pl, when opening, first unlink a previous file of that name.
Previously regcomp.pl had an explicit unlink for the file it creates, but none
of the other regen/*.pl scripts did. Also, remove commented out code relating
to unlink.
Nicholas Clark [Sun, 23 Jan 2011 09:15:58 +0000 (09:15 +0000)]
In regen_lib.pl use a meaningful key for storing the file's name.
regen_lib.pl is actually using the hash slot of the typeglob, not the scalar
slot. Typeglob syntax is particularly confusing and prone to errors, although
in this case it doesn't make any behaviour difference.
H.Merijn Brand [Sun, 23 Jan 2011 09:17:39 +0000 (10:17 +0100)]
From: Benny Siegert <bsiegert@gmail.com>
Date: Thu, 20 Jan 2011 19:27:04 +0100
Subject: [PATCH] MirBSD: fix rpath for shared libperl
If you build perl with a shared libperl.so on MirBSD (the default config),
it will work up to the installation; however, once installed, it will be
unable to find libperl. Treat rpath handling like in the other BSD
dialects.
Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
Nicholas Clark [Sun, 23 Jan 2011 07:59:25 +0000 (07:59 +0000)]
opcode.pl hasn't generated files named *-old since 2008, so don't delete them.
Commit
b6b9a09997c80269 removed the last vestiges of the code that generated
any files named *-old, but the code to clean them up remained. It serves no
purpose.
Nicholas Clark [Sun, 23 Jan 2011 07:35:47 +0000 (07:35 +0000)]
regen/overload.pl should use rename_if_different() for lib/overload/numbers.pm
It was the only file generated by any of the regen/*.pl scripts which was
always written out directly, instead of first creating a temporary file, and
only moving that if successfully generated (and different).
Nicholas Clark [Sun, 23 Jan 2011 00:09:24 +0000 (00:09 +0000)]
Move all the generated file header printing into read_only_top()
Previously all the scripts in regen/ had code to generate header comments
(buffer-read-only, "do not edit this file", and optionally regeneration
script, regeneration data, copyright years and filename).
This change results in some minor reformatting of header blocks, and
standardises the copyright line as "Larry Wall and others".
Jesse Vincent [Sat, 22 Jan 2011 23:58:39 +0000 (18:58 -0500)]
Update TOC for perl51310delta
Jesse Vincent [Sat, 22 Jan 2011 23:57:46 +0000 (18:57 -0500)]
perldelta for 5.13.10
Nicholas Clark [Sat, 22 Jan 2011 23:29:59 +0000 (23:29 +0000)]
In embed.pl, move the logic for perlapi.c's Tolkien quote out from do_not_edit()
Previously do_not_edit() has a special case for this based on the filename
parameter. Now the code to add the comment is next to the relevant call to
do_not_edit().
Nicholas Clark [Sat, 22 Jan 2011 22:19:45 +0000 (22:19 +0000)]
"Don't repeat yourself" in regen/opcode.pl
Nicholas Clark [Sat, 22 Jan 2011 20:15:05 +0000 (20:15 +0000)]
Remove Mac OS classic special case code from Cwd's cwd.t
Nicholas Clark [Sat, 22 Jan 2011 19:56:41 +0000 (19:56 +0000)]
Convert File::Spec's remaining tests to Test::More from Test.
Nicholas Clark [Sat, 22 Jan 2011 19:17:36 +0000 (19:17 +0000)]
Remove Mac OS classic only tests from Cwd's Spec.t
As MacOS was the only platform with platform-only tests, we can also remove the
entire skip-if-not-this-platform infrastructure. Also, as tryfunc() is only
called from one place, inline it.
Nicholas Clark [Sat, 22 Jan 2011 18:18:27 +0000 (18:18 +0000)]
Convert I18N::LangTags to Test::More from Test.
Nicholas Clark [Sat, 22 Jan 2011 17:34:28 +0000 (17:34 +0000)]
Remove duplicate test for $ENV{LANG} in I18N::LangTags test 80_all_env.t
Nicholas Clark [Sat, 22 Jan 2011 16:21:35 +0000 (16:21 +0000)]
Convert Net::Ping's tests to use Test::More from Test.
Ævar Arnfjörð Bjarmason [Sat, 22 Jan 2011 14:11:39 +0000 (14:11 +0000)]
Changes: correct mis-indentation introduced in
dc7c5368d1
Michael Stevens corrected the link to the 5.10.0 tarball in
dc7c5368d173865841aa41764f828d34e663e0bc, but changed four spaces to a
tab in the process.
Since Changes does not otherwise use tabs for indenting convert this
back to four spaces.
Signed-off-by: Ævar Arnfjörð Bjarmason <avar@cpan.org>
Florian Ragwitz [Sat, 22 Jan 2011 08:03:20 +0000 (09:03 +0100)]
Stop EU::CBuilder's tests from failing in parallel
It used to use the same paths for temporary files in all tests. This blew up
randomly when the tests were run in parallel.
Chris 'BinGOs' Williams [Sat, 22 Jan 2011 13:08:05 +0000 (13:08 +0000)]
Update Changes and bump Module::CoreList to 2.44 for a CPAN release
Ricardo Signes [Sat, 22 Jan 2011 12:39:52 +0000 (07:39 -0500)]
fix 5.12.3 release date
Ricardo Signes [Sat, 22 Jan 2011 12:36:22 +0000 (07:36 -0500)]
add 5.12.3 hsitory to perlhist
Nicholas Clark [Sat, 22 Jan 2011 12:37:53 +0000 (12:37 +0000)]
Convert buildtoc to lexical file handles and 3-arg open.
In the refactoring, I don't see any reason to retain void context explicit
close of lexical filehandles opened for reading.
Nicholas Clark [Sat, 22 Jan 2011 12:21:45 +0000 (12:21 +0000)]
Extend pod/buildtoc --test to validate that regenerated files are up to date.
Nicholas Clark [Sat, 22 Jan 2011 12:05:22 +0000 (12:05 +0000)]
In buildtoc, only have entries in %Build for targets being built.
Also, use the file names as values. Effectively now it is a slice of %Targets,
containing just things to be built.
Previously it held all targets (and the phony 'all' target), and had
defined/undefined values for whether they should be built.
Nicholas Clark [Sat, 22 Jan 2011 11:35:12 +0000 (11:35 +0000)]
In buildtoc, only "find all the modules" if rebuilding pod/perltoc.pod
The results of the expensive scan aren't needed for any of the other targets.
Nicholas Clark [Sat, 22 Jan 2011 11:25:34 +0000 (11:25 +0000)]
In buildtoc, rename &output_perltoc to &do_toc, and integrate its invocation.
There is still some special casing, as all other targets modify an existing
file, but with this change more code is shared.
Nicholas Clark [Sat, 22 Jan 2011 11:15:14 +0000 (11:15 +0000)]
In buildtoc, pull the sanity check for NUL bytes to the top level.
It's needed in 3 of the do_* routines, and causes no problems for the others.
Nicholas Clark [Sat, 22 Jan 2011 11:05:08 +0000 (11:05 +0000)]
In buildtoc, convert all the do_* functions to taking 2 scalars.
Previously they took a list of lines from the file, and all except
do_manifest() immediately joined it back to a single scalar. Instead, slurp
the file in whole, and do_manifest() split it out.
Also, change the code that invokes do_* to accept only a scalar as a return,
now that all functions return a scalar, rather than (potentially) a list to
join.
Nicholas Clark [Sat, 22 Jan 2011 10:51:37 +0000 (10:51 +0000)]
In buildtoc, convert do_manifest() to returning a scalar.
Previously it (alone of the do_* functions) returned a list for the regenerated
file's contents.
Nicholas Clark [Sat, 22 Jan 2011 09:34:11 +0000 (09:34 +0000)]
Re-add 5.12.3's release date to Module::Corelist.
This was missed by change
57e52dbefae98516, which reverted
01bc7817c9b86c12,
because 5.12.3's release date was removed as an inadvertent side effect of
e0698539c3e1cd22, the change to add 5.13.9
Nicholas Clark [Sat, 22 Jan 2011 09:15:39 +0000 (09:15 +0000)]
Add perl5123delta.pod to MANIFEST and pod.lst, after
f1d4662fac7469d5.
re-run pod/buildtoc to propagate the new file to the relevant platform
makefiles.
Ricardo Signes [Sat, 22 Jan 2011 03:13:50 +0000 (22:13 -0500)]
pick perl5123delta.pod from maint-5.12
Ricardo Signes [Sat, 22 Jan 2011 03:12:37 +0000 (22:12 -0500)]
re-add 5.12.3 versions
Ricardo Signes [Sat, 22 Jan 2011 03:09:53 +0000 (22:09 -0500)]
fix bizarre date typo in 5.12.2 epigraph
Ricardo Signes [Sat, 22 Jan 2011 03:09:26 +0000 (22:09 -0500)]
add v5.12.3 epigram
Michael Stevens [Thu, 20 Jan 2011 13:49:18 +0000 (13:49 +0000)]
[perl #82526] Fix kldp links.
Michael Stevens [Mon, 17 Jan 2011 18:23:15 +0000 (18:23 +0000)]
Fix an invalid =item warning.
Craig A. Berry [Fri, 21 Jan 2011 21:55:32 +0000 (15:55 -0600)]
Automate current perldelta entry in vms/descrip_mms.template.
pod/buildtoc knows which one is current, so we might as well
substitute it in one place and be done with it.
David Golden [Fri, 21 Jan 2011 22:01:33 +0000 (17:01 -0500)]
Updated CPAN.pm to CPAN version 1.94_64
[DELTA]
2011-01-20 David Golden <dagolden@cpan.org>
* release 1.94_64
* remove 'use_file_homedir' config option and fix #62986 using
a more robust method. Original config directories will be found
even if File::HomeDir is installed
* streamline configuration intro text
* add missing documentation for 'atexit' and local::lib bootstrap
Chip Salzenberg [Fri, 21 Jan 2011 20:54:08 +0000 (12:54 -0800)]
clear up unused var warning in prev patch
David Golden [Fri, 21 Jan 2011 20:44:52 +0000 (15:44 -0500)]
Fix up MANIFEST and update Maintainers.pl for ExtUtils::CBuilder
David Golden [Fri, 21 Jan 2011 20:43:53 +0000 (15:43 -0500)]
Update Maintainers.pl for ExtUtils-CBuilder-0.280201
David Golden [Fri, 21 Jan 2011 20:38:19 +0000 (15:38 -0500)]
remove Changes file from dist/ExtUtils-CBuilder
Chip Salzenberg [Fri, 21 Jan 2011 20:41:51 +0000 (12:41 -0800)]
Since xmldump_packsubs does not actually output XML, don't call it from xmldump_all.
Craig A. Berry [Fri, 21 Jan 2011 19:12:13 +0000 (13:12 -0600)]
Don't export PerlIORaw_open if it doesn't exist anymore.
Follow-up to
ecfd064986aef03b9b7d8b56da83ae37ff435721.
Craig A. Berry [Fri, 21 Jan 2011 19:08:05 +0000 (13:08 -0600)]
PerlIO_push returns NULL, not -1, on failure.
Follow-up to
81fe74fb3f93457df8e864c91546ac6b860657fc.
David Mitchell [Fri, 21 Jan 2011 17:23:51 +0000 (17:23 +0000)]
RT #75870 perldata.pod tied hash in scalar context
update inaccurate description.
Karl Williamson [Fri, 21 Jan 2011 16:55:28 +0000 (09:55 -0700)]
Add x-references in pods to new /a, etc.
Nicholas Clark [Fri, 21 Jan 2011 16:58:28 +0000 (16:58 +0000)]
In buildtoc, eliminate a substitution which mangles nested package names.
It's not clear what the intent of s/(\w+)::\1/$1/; was, but it converts
"autodie::exception" to "autodiexception" (etc), which is clearly buggy.
The regexp has been in pod/buildtoc since it was added in
cb1a09d0194fed9b.
Nicholas Clark [Fri, 21 Jan 2011 16:29:18 +0000 (16:29 +0000)]
In buildtoc, move the local $/ = ''; to the scope of the file it relates to.
Nicholas Clark [Fri, 21 Jan 2011 16:17:30 +0000 (16:17 +0000)]
In buildtoc, refactor the flow logic in podset() to avoid most uses of next.
Father Chrysostomos [Fri, 21 Jan 2011 16:26:50 +0000 (08:26 -0800)]
[perl #81750] Perl 5.12: undef-as-hashref bug
The addition of the boolkeys op type in commit
867fa1e2d did not
account for the fact that rv2hv (%{}) can sometimes return undef
(%$undef with strict refs turned off).
When the boolkeys op is created (and the rv2hv becomes its kid), the
rv2hv is flagged with OPf_REF, meaning that it must return a hash, not
the contents.
Perl_softrefxv in pp.c checks for that flag. If it is set, it dies
with ‘Can't use an undefined value as a HASH reference’ for unde-
fined values.
This commit changes it to make an exception if rv2hv->op_next is a
boolkeys op. It also changes pp_boolkeys to account for undef.
Nicholas Clark [Fri, 21 Jan 2011 15:06:19 +0000 (15:06 +0000)]
In buildtoc, eliminate the now vestigial sub output.
Nicholas Clark [Fri, 21 Jan 2011 15:00:19 +0000 (15:00 +0000)]
In buildtoc, call Text::Wrap::wrap() just before output.
Previously output() wrapped as it went along. This change renders output()
vestigial.
Nicholas Clark [Fri, 21 Jan 2011 14:43:25 +0000 (14:43 +0000)]
In buildtoc, remove whitespace only lines just before output.
Previously output() eliminated whitespace only lines. Whilst this change itself
adds code, it makes the logic clearer, which will permit more simplification.
Nicholas Clark [Fri, 21 Jan 2011 14:33:46 +0000 (14:33 +0000)]
In buildtoc, fold multiple newlines just before output.
Previously output() tracked newlines, and ignored more than 2 in a row. That
used more code, and a "global" variable for state.
Nicholas Clark [Fri, 21 Jan 2011 14:15:34 +0000 (14:15 +0000)]
In buildtoc, accumulate all perltoc.pod content before outputting it to disk.
David Mitchell [Fri, 21 Jan 2011 10:25:03 +0000 (10:25 +0000)]
add GvCV_set() and GvGP_set() macros.
and make GvCV() and GvGP() rvalue-only.
This it to allow a future commit to eliminate some backref magic between
GV and CVs, which will require complete control over assignment to the
gp_cv slot.
David Mitchell [Thu, 20 Jan 2011 16:30:55 +0000 (16:30 +0000)]
XS-APItest/t/caller.t: mark two passing tests
There are four TODO tests in caller.t, relating to the hints hash returned
under the debugger. Since
20439bc77dfeec46d94a15cf108446039e26c995, two
of these tests has started to pass. I don't understand this area enough
to know whether that commit *should* have made them pass, but it
seems fairly consistent. So un-TODO just those two.
This means that under the debugger, we now expect to get a hints hash, but
that its contents are still wrong.
Nicholas Clark [Fri, 21 Jan 2011 13:36:05 +0000 (13:36 +0000)]
In buildtoc, eliminate nl() by adding an explicit "\n" at its two use points.
Nicholas Clark [Fri, 21 Jan 2011 11:41:42 +0000 (11:41 +0000)]
In buildtoc, the module de-duplication logic applies equally to pragmata.
Nicholas Clark [Fri, 21 Jan 2011 11:31:26 +0000 (11:31 +0000)]
In buildtoc, inline path2modname() into its only caller.
Nicholas Clark [Fri, 21 Jan 2011 10:51:53 +0000 (10:51 +0000)]
No need for buildtoc to scan ext/, now that it runs after the build process.
Since
7eb47696deb9a734, buildtoc has been run as part of the build process,
hence it has all modules copied to lib/ from ext/ (dist/ and cpan/), where
its scan finds them.
Previously it was run by the committers as and when (and sometimes not often
enough), and the generated pod/perltoc.pod checked in. Hence it might be run
on a clean checkout, and so couldn't assume that everything would be in lib/
This scan of lib/ is why buildtoc can still pick up all the modules, despite
their migration to cpan/ and dist/, which it has no idea about.
Also, scanning lib/ rather than ext/ means that Pod for modules that are not
being installed on this platform won't be included, which makes for a more
accurate perltoc.pod
Nicholas Clark [Fri, 21 Jan 2011 10:49:01 +0000 (10:49 +0000)]
Teach buildtoc that - in filenames in ext/ represents :: in the package name.
Previously it was treating - as a literal - in the name, and hence was
considering that files found in ext/ and in lib/ were different packages.
This has been a problem since the restructuring of ext/
Nicholas Clark [Fri, 21 Jan 2011 10:19:12 +0000 (10:19 +0000)]
Remove commented out code from buildtoc
Also remove code that would die if the filename matched TUT, which seems never
to have served any useful purpose.
Robin Barker [Fri, 21 Jan 2011 09:45:33 +0000 (09:45 +0000)]
Nicholas Clark [Fri, 21 Jan 2011 09:29:23 +0000 (09:29 +0000)]
Remove training whitespace in Pod which can confuse pod/buildtoc
Some whitespace is being interpreted as "not a blank line" causing it (and
possibly other parsers) to assume that subsequent text is part of the preceding
=head or =item
Nicholas Clark [Fri, 21 Jan 2011 09:28:46 +0000 (09:28 +0000)]
Remove training whitespace in Pod which can confuse pod/buildtoc
Some whitespace is being interpreted as "not a blank line" causing it (and
possibly other parsers) to assume that subsequent text is part of the preceding
=head or =item
Nicholas Clark [Fri, 21 Jan 2011 09:26:24 +0000 (09:26 +0000)]
Remove training whitespace in Pod which can confuse pod/buildtoc
Some whitespace is being interpreted as "not a blank line" causing it (and
possibly other parsers) to assume that subsequent text is part of the preceding
=head or =item
Nicholas Clark [Thu, 20 Jan 2011 17:12:50 +0000 (17:12 +0000)]
In buildtoc, remove a superfluous length in output()
If \S matches, then length must be non-zero, and therefore true.
This dates to commit
8ebc5c0145d2e355, when output() was added.
Nicholas Clark [Thu, 20 Jan 2011 16:54:17 +0000 (16:54 +0000)]
In buildtoc, remove a superfluous "\n " passed to output()
This is actually a hangover from the original code which used print directly,
left in as part of the refactoring that added sub output in
8ebc5c0145d2e355.
The machinations of output() mean that this "\n " generate no actual output,
hence the output of buildtoc is unchanged.
Nicholas Clark [Wed, 19 Jan 2011 12:01:44 +0000 (12:01 +0000)]
Convert threads-shared/t/disabled.t from Test to Test::More.
Curiously it was the only test of Threads::Shared that used Test.
Nicholas Clark [Wed, 19 Jan 2011 11:36:30 +0000 (11:36 +0000)]
Convert lib/Pod/t/utils.t from Test to Test::More
Curiously this test file is not part of the Pod-Parser distribution.