@@ -1,1257 +0,0 @@
-[Changes for 1.007 - Oct 22, 2012]
- - Hopefully fix "pp -C ..." for any modules that assume an
- actual tree of files, e.g. looking for all installed modules Foo::Bar::*;
- call _extract_inc even when $ENV{PAR_CLEAN} is true
- - update to Module::Install 1.06
-
-[Changes for 1.006 - Oct 15, 2012]
- - Fix RT #78633: PAR::import ignores url => $repo_client_object
- applied patch from KENO, thanks!
- - Fix RT #73491: cache directory naming problem
- In PAR::SetupTemp::_get_par_user_tempdir (actually _find_username)
- we try to sanitize username (so that there are no problematic
- characters in the name of the per-user cache directory), but the
- strategy fails miserably for some charset encodings.
- E.g. for EUC-KR or CP949 (as in the bug report) we may produce an
- illegal sequence of bytes; in other cases we may cause collisions
- (different usernames mapping to the same directory name).
- Fix the problem once and for all by encoding the username
- (bytewise) as 2 hex digits.
-
-[Changes for 1.005 - Dec 2, 2011]
- - bump Perl version requirement to 5.8.1 (Schwern: The End Of 5.6 Is Nigh!)
- - run all tests using a nonce PAR_TMPDIR (otherwise CPAN Testers
- goes crazy as top level /tmp/par-USER directories (or similar)
- from previous tests may now be considered "unsafe")
-
-[Changes for 1.004 - Nov 30, 2011]
- - back out r1241: it causes errors in PAR::Packer's test suite
- - change "unsafe directory" error message to match the wording
- used by PAR::Packer
- - remove "debian" sub directory: it isn't released to CPAN and
- Debian will supply its own anyway
- - remove some cruft from MANIFEST.SKIP
-
-[Changes for 1.003 - Nov 28, 2011]
- - RT #69560/CVE-2011-4114: PAR packed files are extracted to unsafe
- and predictable temporary directories
- (Note: this bug was originally reported against PAR::Packer, but
- it applies to PAR as well)
- - create parent of cache directory (i.e. /tmp/par-USER) with mode 0700
- - if it already exists, make sure that (and bail out if not)
- - it's not a symlink
- - it's mode 0700
- - it's owned by USER
- - Fix a problem packing XML::LibXSLT on Windows (see the thread starting
- with http://www.nntp.perl.org/group/perl.par/2011/02/msg4919.html)
- - Die (with a hopefully useful message) if any error is encountered
- during an Archive::Zip extract operation
-
-[Changes for 1.002 - Jul 25, 2010]
- - Fixes to VERSIONs of PAR::Setup*.
- - No change in behaviour since 1.001
-
-[Changes for 1.001 - Jul 25, 2010]
- - RT #57399: extract everything (including DLLs) in File::ShareDir
- directories. Module::ScanDeps classifies everything in
- File::ShareDir directories as "data", but PAR uses it's own
- heuristics what to extract from a .par.
- - Previous release was missing META.yml.
- - Upgrade Module::Install to 1.00
-
-[Changes for 1.000 - Apr 10, 2010]
- - Fix defined(%hash) deprecation warning in PAR::Heavy
-
-[Changes for 0.994 - Jul 23, 2009]
- - Fix for the PAR::Heavy fix to the INC priority handling.
-
-[Changes for 0.993 - Jul 19, 2009]
- - The priority (fallback=>0) repositories should look at @PAR_INC
- for the loaded PARs instead of @PAR_INC_LAST.
- - Don't reload from a downloaded .par file after installing it
- via "upgrade".
- - Band-aid fix for the loading priority of shared librares from
- PAR files: Try PAR's first, the local stuff, then fallback-PARs.
- - Initial support for running external perl scripts from a packaged
- interpreter.
-
-[Changes for 0.992 - Apr 5, 2009]
- - Support for non-fallback repositories.
-
-[Changes for 0.991 - Mar 10, 2009]
- - Promote previous release to stable release.
-
-[Changes for 0.989_01 - Mar 2, 2009]
-* Bug fixes, etc.
- - Track the locations of all archives that have been extracted to
- $ENV{PAR_TEMP}/inc in this run.
-
-[Changes for 0.988 - Mar 1, 2009]
- - Promote previous release to stable release.
-
-[Changes for 0.987_01 - Feb 20, 2009]
-* New features
- - Better cleanup of the $TMPDIR/par-$USER/temp-$$ directories
- that are typically used as caches in the "use PAR 'foo.par'" use case.
-
-* Bug fixes, etc.
- - Very slightly more careful handling of PAR-specific environment
- variables.
- - Cleanup of PAR::SetupTemp
-
-[Changes for 0.986 - Feb 19, 2009]
- - Promote to stable release.
-
-[Changes for 0.985_01 - Feb 2, 2009]
-* New features
- - Support for the brand new static dependency resolution
- of PAR::Repository::Client 0.23.
-
-* Bug fixes, etc.
- - Fix issue with running scripts from repositories:
- The INC hooks used to be set up too late for this to
- automatically pick up dependencies.
-
-[Changes for 0.984 - Jan 25, 2009]
-* New features
- - Implemented the auto-upgrading option for loading and installing
- from PAR::Repositories.
-
-* Bug fixes, etc.
- - Fix issue with PAR::Repository::Client development versions.
-
-[Changes for 0.983 - Sep 12, 2008]
-* Dependencies
- - Require AutoLoader 5.67 which contains a PAR-related
- bug-fix.
- - Require PAR::Dist 0.32.
-
-* Internal changes
- - The full extraction process _extract_inc
- (which is triggered when a non--clean pp packaged
- executable is run) can now be forced to do the
- extraction (instead of doing if !-d).
- - That same extraction routine now accepts Archive::Zip
- handles or file names.
- - When, during the full extraction, the extracted paths
- are to be added to @INC, we now make sure they're not
- in @INC yet.
-
-[Changes for 0.982 - Aug 10, 2008]
-* New features
- - Moved the routines that setup the PAR_TEMP environment variable
- to a separate module in the distribution so it's possible to
- use PAR::Repository::Client without loading all of PAR.
- - Same for the function that sets up PAR_PROGNAME.
-
-* Bug fixes, etc.
- - Upgrade to Module::Install 0.77
- - Fix for running scripts from repositories.
-
-[Changes for 0.980 - May 22, 2008]
-* Bug fixes, etc.
- - The function PAR::get_filehandle() that was introduced in 0.979
- is really broken because Archive::Zip is broken. Turns out calling
- Archive::Zip::Member->fh() returns a file handle to the zip file,
- not a virtual/tied/whatever file handle to the member file.
- Therefore, the get_filehandle() function has been removed again until
- we work around this issue.
- - Upgrade to Module::Install 0.73
-
-[Changes for 0.979 - May 13, 2008]
-* New features
- - New function PAR::get_filehandle() returns a file handle for a file
- in any open .par files. Similar to read_file().
-
-[Changes for 0.977 - Oct 19, 2007]
-* Bug fixes, etc.
- - HPUX doesn't like shared libraries being unlinked while in use. So don't
- try to do this even in clean mode. (Workaround will be in par.pl)
- (Scott Stanton)
- - Fix DLL extraction file name matching in PAR::Heavy.
- - Save two system calls per DLL during DLL extraction in PAR::Heavy.
-
-[Changes for 0.976 - Jul 29, 2007]
-* New features
- - Use Archive::Unzip::Burst for unpacking binary executables if
- available. (This yields a significant startup speed-up.)
-
-* Bug fixes, etc.
- - Removed the auto_install feature from Makefile.PL. auto_install
- is conceptually broken.
-
-[Changes for 0.973 - Feb 3, 2007]
-* New features
- - A new option for the "use PAR { ... };" use case:
- "no_shlib_unpack" signals that for this particular .par file,
- shared libraries that were added with the --lib option should
- not be extracted. This allows users to do their own cache
- handling for these libraries.
- - PAR no longer unpacks *all* shared libraries by default but only
- those in the shlib/ directory (i.e. added with --lib).
- The shared libraries in auto/ should be picked up by the DynaLoader
- hack.
- - If available, the prefork.pm module will be used to declare
- run-time dependencies for better memory use in forking
- environments.
- - PAR now uses a caching mechanism to speed up the extraction process.
- This should particularly impact users of the "use PAR {file =>...}"
- form.
-
-* Bug fixes, etc.
- - Applied an optimization of the unpacking process on case insensitive
- file systems.
-
-[Changes for 0.972 - Jan 16, 2007]
-* Bug fixes, etc.
- - Removed PAR::AutoLoaderFix again. It wasn't working as expected
- all the time.
- - To fix the problem AutoLoaderFix was supposed to fix, we now
- require AutoLoader 5.62 or newer which was just recently released
- to CPAN. (Previously only available from blead perl.)
-
-[Changes for 0.971 - Jan 12, 2007]
-* Bug fixes, etc.
- - Fixed typo in the POD. (Jerrad Pierce)
- - Included fix for a bug in AutoLoader.pm as shipped with all
- perl versions up to and including 5.8.8 as PAR::AutoLoaderFix.
- This cures a problem of endless looping when the %INC entry of a
- module doesn't point to a file of the same name.
- This may happen during "use PAR 'foo.par'".
-
-[Changes for 0.970 - Dec 3, 2006]
-* This release introduces some rather radical changes, so read carefully:
-* All PAR::Packer related logic has been moved to a separate distribution,
- PAR-Packer. This includes pp, parl and all packaging tools. This way,
- PAR becomes a pure-Perl distribution that can be most easily installed by
- users of software which requires PAR.
- Developers who want to use the PAR packager, pp, need to install the
- PAR-Packer distribution from CPAN.
-
-[Changes for 0.961 - Nov 23, 2006]
-* Bug fixes, etc.
- - PAR::StrippedPARL::Base->write_parl() failed to work if the @INC
- directories contained spaces in 0.960. (Steven Mackenzie)
- - Much improved documentation of the environment variables
- (Glenn Linderman)
- - Fix for a spaces-in-pathname problem on Windows for
- t/30-current_exec.t. (Malcolm Nooning)
-
-[Changes for 0.960 - Nov 21, 2006]
-* Bug fixes, etc.
- - myldr/Makefile.PL fix: Clean up myldr/usernamefrompwuid.h.
- - Silence warning in myldr/internals.c.
- - Silence warnings seen on Irix from myldr/env.c.
- - Skip most tests in 10-parl-generation.t if there is no parl.
- - Skip loading ActiveState Perl's "sitecustomize.pl" in par.pl.
- - Load modules via require and other files via do.
- - The parl-regeneration-for-every-pp-call addition of the 0.958 release
- should now also work for static perls.
-
-* New features
- - Adressing RT ticket #6612: Now using getpwuid() to determine the
- user name if supported by the OS.
-
-[Changes for 0.959 - Nov 12, 2006]
-* This is just a hotfix release because 0.958 lacked META.yml. One day, I will
- switch from Module::Install to Module::Build...
-
-[Changes for 0.958 - Oct 25, 2006]
-* Bug fixes, etc.
- - myldr/Makefile.PL fix: make static.o depend on mktmpdir.c, my_perl.c,
- my_par.c. (Roderich Schupp)
- - Modules included with the -M option to pp were previously scanned
- for dependencies but not mapped through the
- %Module::ScanDeps::Preload hash for custom dependencies. That's
- fixed now.
- - $ENV{PAR_RUN} isn't set by PAR::Packer any more because nothing in
- the PAR sources uses it. $ENV{PAR_RUN} is no longer used by PAR
- at all.
- - Unified the environment variables which are looked at for
- finding the system's temporary directory.
-
-* New features
- - During the build process, PAR appends stripped down copies of parl
- (and parldyn if applicable) to the data classes
- PAR::StrippedPARL::Static and ::Dynamic. These
- parls-without-embedded-modules are used for packaging so the formerly
- embedded modules are now packaged from the packaging system. (Instead
- of stemming from the system where PAR/parl was built.)
- - The "use PAR { repository => $url };" syntax now also supports
- the use of user-constructed PAR::Repository::Client objects instead
- of an URL.
- - The -F (module code filter) option now supports selective filtering
- of modules. The syntax is "-F FILTER=REGEX" or - as before -
- "-F FILTER". The regular expression is applied to the *file name*,
- of the module inside the PAR (e.g. Foo/Bar.pm). This behaviour was
- chosen over matching against the module name (e.g. Foo::Bar) because
- the filters can be applied to module-like and script files as well
- (.pl, .al, etc.).
- - Updated PAR/FAQ.pod with the new FAQ's from the PAR wiki.
- - Added a POD file PAR/Environment.pod which is intended to become
- an index of all environment variables PAR uses of its own or
- recognizes from its users. Still mostly a stub.
-
-[Changes for 0.957 - Oct 24, 2006]
-* Bug fixes, etc.
- - Fix executable PARs top properly detect embedded scripts
- named the same as the executable. (Jesse Vincent)
- - Comment out the call to par_current_exec_proc (in the C loader)
- which breaks the use of symlinks to pp-ed executables when not
- called with a path. (I.e. using a search in $PATH).
-
-[Changes for 0.956 - Oct 3, 2006]
-* This is another hotfix release. Fixed a mindless bug introduced in 0.955.
-
-[Changes for 0.955 - Oct 3, 2006]
-* Bug fixes, etc.
- - 0.952 introduced removal of system module search paths if -B is in
- effect. This resulted in some valid PAR-related paths being removed
- as well. Fixed. Upgrading from 0.952 and 0.954 is suggested.
- - Changed the use of hard-coded '/' as path-separator to using File::Spec.
-
-[Changes for 0.954 - Sep 26, 2006]
-* This release is equivalent to 0.953. The 0.953 CPAN upload is
- broken!
-
-[Changes for 0.953 - Sep 18, 2006]
-* Bug fixes, etc.
- - Added optional POD tests.
- - Modified -B so that if -B is in effect, all entries are stripped
- out of @INC except for the PAR hooks. This happens right before
- the script contained in the pp-ed binary is executed.
-
-[Changes for 0.952 - Aug 22, 2006]
-
-* New features
- - Added the "install" option to the PAR loading syntax.
- If specified, the contents of the PAR distribution are
- permanently installed.
- This requires PAR::Repository::Client 0.04.
-
-* Bug fixes, etc.
- - Fixed broken META.yml in 0.951.
-
-[Changes for 0.951 - Aug 12, 2006]
-(This includes any changes up to 0.950.)
-
-* New features
-
- - Introduced new PAR loading syntax and semantics:
- use PAR { file => 'path/to/par/or/URL' };
- ==> equivalent to "use PAR 'path/to/par/or/URL';"
- - Introduced the 'fallback' option: (default = 0)
- use PAR { file => 'foo.par', fallback => 1 };
- ==> Loads modules from the PAR file only if loading
- them from @INC did not succeed.
- - Introduced the 'run' option which executes a script
- in a PAR archive just like
- perl -MPAR foo.par script.pl
- - If PAR::Repository::Client is installed, you can add a
- repository of .par distributions to your library search path
- as follows:
- use PAR { repository => 'http://foo' };
- - Of course, 'run' also works with repositories:
- use PAR { repository => 'http://foo', run => 'my_app' };
- (This searches the repository for any distributions that have a
- my_app script.)
- --> For details on repositories, have a look at the
- PAR::Repository::Client module.
-
-- Bug fixes, etc.
-
- - Commented a couple of the routines in PAR.pm. (Yay!)
- - New test script for the new fallback loading feature.
- - Fixed a bug in the Spreadsheet::ParseExcel handling in
- PatchContent.pm.
-
-[Changes for 0.942 - Jul 22, 2006]
-
-* Bug fixes, etc.
-
- - Better support for diagnostics.pm (in conjunction with
- Module::ScanDeps 0.62.)
- - Now requiring Module::ScanDeps 0.62.
-
-[Changes for 0.941 - Jun 20, 2006]
-(No, PAR isn't stagnating. It's just that 1.00 would draw close if
-we continued with 0.01 increases.)
-
-* Bug fixes, etc.
-
- - Version 0.94 of PAR would use the same cache area for all pp-ed
- applications due to a faulty hotfix for Digest::SHA. This applies
- to PAR 0.94 only. Think of 0.941 being PAR 0.94 done right.
-
-[Changes for 0.94 - Jun 1, 2006]
-
-* New Features
-
- - Added support for reading options to pp from a file using a
- '@filename' argument to pp: pp -o foo --gui @filename foo.pl
-
-* Bug fixes, etc.
-
- - Workaround for a bug in Digest::SHA 5.38 and 5.39 that would prevent
- PAR from being built.
- - Fixed details in the 2-pp.t test file.
- - Now recognizes text files that aren't picked up by the -T operator
- but by the "file" tool.
- - Applied Roderich Schupp's patch to 30-current_exec.t to fix a
- path issue.
- - Now requiring Module::ScanDeps 0.60 which fixes a couple of bugs
- which might be observed as PAR bugs.
- - Now working well with Spreadsheet::ParseExcel which uses an invalid
- POD section to comment out a code block. This wasn't recognized by
- PAR::Filter::PodStrip as POD and hence partly left in...
- - If the output directory doesn't exist, we create it now and output
- a meaningful error message if that failed.
-
-[Changes for 0.93 - May 19, 2006]
-
-* New Features
-
- - Added support for PAR_TMPDIR (PAR_GLOBAL_TMPDIR) so that the
- temp directory can be controlled for just the PAR file bits.
- (Leolo)
- - Added par_current_exec_proc() which finds the file of the current
- executable in /proc, if possible. (Leolo)
- - Added par_current_exec() which finds he file of the current
- executable, if possible on this OS. (Leolo)
- - par_findprog() now uses par_current_exec() if possible.
-
-* Bug Fixes, etc.
-
- - Upgraded to Module::Install 0.62+ (Audrey Tang, Steffen Mueller)
- - Document a strange interaction with chdir() and relative
- paths. (Chris Dolan)
- - Documented the bits that make up PAR_TEMP. (Leolo)
- - Fixed the call to par_findprog. path (aka val) was set to
- tmpdir. (Leolo)
- - Documented the CACHE name at the end of a self-executing
- PAR. (Leolo)
- - myldr/Makefile.PL now generates some dependencies for
- main.c (Leolo)
- - Applied patch from RT ticket. (tsee)
- https://rt.cpan.org/Ticket/Display.html?id=13959
- - Applied Ivan Kudryavtsev's patch that fixes a couple of calls
- to PAR subroutines in PatchContent filtered code. (tsee)
-
-[Changes for 0.92 - February 22, 2006]
-
-* Bug Fixes
-
- - Now requiring Module::ScanDeps 0.56 which handles autouse
- correctly.
- - Now shipping with a correct SIGNATURE. (Which was broken for
- 0.91.)
-
-[Changes for 0.91 - February 13, 2006]
-
-* Bug Fixes
-
- - Applied Alan Stewart's patch which fixes @ARGV pollution in
- daughter programs. See also
- http://www.nntp.perl.org/group/perl.par/2152
- - Now mentioning the ENV var "PAR_VERBATIM" in the documentation.
- See also http://www.nntp.perl.org/group/perl.par/2196
- - Applied Malcolm Nooning's fix for the test suite. We used to get
- failed tests on Windows because of spaces in path names.
- - Applied Roderich Schupp's and Malcolm Nooning's patches to
- the test suite fixing problems with Cygwin.
- - Applied Vincent Ladeuil's patch to PAR::Filter::Bleach to return a
- true value for modules that loaded okay.
- - Changed 'PAR_BASE' in the Makefile.PL to 'SMUELLER'.
-
-[Changes for 0.90 - November 25, 2005]
-
-* Bug Fixes
-
- - When compiling with static libperl, myldr/ may fail "make"
- due to sha1.c not generated properly.
-
- - Pod stripping could fail on __DATA__ sections for files
- with CRLF line endings.
-
- - The documentation erroneously referred to the PAR_TEMP
- environment variable, whereas it should be PAR_GLOBAL_TEMP.
-
- - Compilation fixes for MinGW/MSYS.
-
-[Changes for 0.89 - June 10, 2005]
-
-* Bug Fixes
-
- - Stop static.c from pulling in Perl header files, otherwise
- parl.exe ends up depending on the Perl DLL on Win32 when
- Perl is built without PERL_IMPLICIT_SYS.
-
- - With *nix and File::Path 1.06, par.pl's avoidance of loading
- Cwd.pm caused syntax errors.
-
-[Changes for 0.88 - June 7, 2005]
-
-* Bug Fixes
-
- - Extracted .pl files should be loadable via the coderef-in-@INC too,
- just like .pm files and autosplit files. This makes PAR work with
- Perl 5.8.7 on Win32.
-
- - Fix the build with GCC 4.0.
-
- - If $ENV{PWD} is not defined, fallback to use `pwd` to obtain the
- working directory for invoking.
-
-[Changes for 0.87 - January 31, 2005]
-
-* Bug Fixes
-
- - On Win32, some versions of File::Spec::Win32 contains explicit
- "use Cwd;" lines, which renders parl.exe unusable.
-
- - Executable made by "pp" may fail when invoked as "./a.out" or
- "../a.out", due to incorrect PWD handling logic.
-
-[Changes for 0.86 - December 11, 2004]
-
-* New Features
-
- - New "pp -z" (--compress) option to set compression level (0-9).
-
- - New "pp -T" (--tempcache) option to override the per-executable
- directory name; it defaults to a hash of the executable, computed at
- compile time. This makes startup much faster for large executables.
-
- - The hash algorithm described above now prefers Digest::SHA if installed,
- otherwise Digest::SHA1, then fallbacks to Digest::MD5.
-
- - Functionality of "pp -X" is now extended: if the argument after -X is a
- zip or par file, files in it are excluded from the produced executable,
- and the executable will "use" the zip/par instead. For multiple -X args,
- successive args are only "use"d if they contain additional unique files.
-
- - "pp -l" now searches for libraries in "." and PATH in Win32.
-
- - "pp -l" shared libraries are now added to %skip, so it will not
- be included in both shlib/ and lib/.
-
- - "pp -l" now chases symbolic links. For example, if "libsomelib.so"
- is a symlink to "libsomelib.so.1", which is another symlink to
- "libsomelib.so.1.2", pp now follows these symlinks and add the real
- file the par, rather than "libsomelib.so".
-
- - New contributed code in "contrib/stdio/": Useful Tk console for
- "pp -g" users.
-
- - New contributed tutorial documents, currently in "contrib/docs/",
- which will eventually be turned into POD documents.
-
- - Running "perl Makefile.PL" with $ENV{DEBUG} set to true now produces
- "parl" with debug symbols.
-
- - Remove Cwd.pm (and Cwd.so) from the bundled dependencies.
-
-* Bug Fixes
-
- - More robust probing for case-insensitive file systems.
-
- - PodStrip now attempts to match "standard" pod start before =cut,
- otherwise =cut gets removed by itself.
-
- - Win32 slashes are now normalized in privlib and archlib directories.
-
- - Don't extract shared libraries to inc/, since they were extracted
- in $PAR_TEMP already.
-
- - Don't re-extract shared libraries in subdirectories, since they
- are picked up by corresponding "use".
-
- - Tk now exits properly with a non-zero exit() value.
-
- - Fix libperl probing problem on Debian and Gentoo that manifests as a
- "libperl5.8.so not found" error during runtime.
-
- - gpp: Fixed typo in options with multiple filenames; cleaned up
- pp parameters.
-
- - When PAR_TEMP is set, shlib/ was not correctly added to the dynamic
- load path environment variables.
-
- - PAR now builds with Win32 VC++ without CVTRES.EXE available.
-
- - Detection of cl.exe, gcc.exe and cc.exe is now case-insensitive.
-
-[Changes for 0.85 - July 2, 2004]
-
-* New Features
-
- - New version of "gpp"; see contrib/gui_pp/gpp_readme.txt for details.
-
-* Bug Fixes
-
- - MANIFEST and META.yml were not properly updated by PAR::Packer.
-
- - Setting directory aliases with "pp -a"/"pp -A" was broken.
- Fixed, and tests were added for it.
-
- - Statically-built executables was needlessly extracting libperl
- each time it runs; now it is eliminated and hence much faster.
-
-[Changes for 0.83 - May 29, 2004]
-
-* New Features
-
- - Revamped PAR::FAQ and sychronized with par.perl.org.
-
- - In pp-generated programs, $0 is now set to the pathname leading
- to the invoked executable. Use $ENV{PAR_0} instead to get the
- filename that contains the main perl program.
-
- - Updated "contrib/gui_pp/gpp" to support PAR::Packer options.
-
-* Bug Fixes
-
- - Core XS modules, such as Data::Dumper, were skipped by "pp".
-
- - Fix t/2-pp.t for Cygwin by probing $Config{_exe} rather than uname().
-
- - Scripts made by "pp -P", when invoked as "perl scriptname",
- should not search for the same-named programs in PATH.
-
- - Correctly remove leading slash and drive letters from absolute
- filenames passed to "pp -a". Also normalized blackslahes to slashes.
-
- - The PP_OPTS environment variable was not recognized.
-
- - "pp -a dirname;diralias" was broken.
-
- - "pp -f" and "pp -F" were broken.
-
-[Changes for 0.82 - May 24, 2004]
-
-* New Features
-
- - New module PAR::Packer provides an OO interface to "pp"'s
- functionality; "pp" is now merely a thin wrapper for it.
-
- - New module App::Packer::PAR is a modified version of
- App::Packer, designed to work with PAR::Packer, and will
- hopefully be merged back to App::Packer.
-
- - The old, procedural "pp" is moved to contrib/; end-users
- should notice no changes in "pp"'s behaviour.
-
- - New options "pp -a" and "pp -A" (--addfile/--addlist) provides
- ways to include extra files and directories in the package.
-
- - The long option name for "pp -M" is changed from --add to
- --module. The old name is still recognized but no longer
- documented. Using "pp -M" to include non-library files
- is now deprecated; use "pp -a" instead.
-
- - par.pl and parl now writes messages to STDOUT, instead of
- STDERR. As a consequence, t/2-pp.t no longer prints extra
- warnings during "make test".
-
-* Bug Fixes
-
- - On Non-Win32 platforms, perl 5.8.0 and earlier versions produced
- pp-generated executables that immediately segfaults.
-
- - Running pp-generated executables with absolute pathname failed
- on statically-built perls.
-
- - Tests were failing due to a missing pipe_a_command.pm in MANIFEST.
-
- - Add the missing myldr/win32.coff for building on Cygwin/MinGW.
-
- - If the "perl" in path is different from the perl interpreter used
- for "make test", t/2-pp.t is known to fail and is now skipped.
-
- - Cygwin failed t/2-pp.t because "parl" is spelled as "parl.exe" there.
-
-[Changes for 0.81 - May 23, 2004]
-
-* New Features
-
- - Regained support for Win9x, Cygwin and MinGW.
-
- - PAR now supports 64-bit platforms, such as Tru64 and AIX.
-
- - Cygwin and MinGW can now build EXEs with icons, too; MinGW can update
- the icons, but Cygwin cannot.
-
- - Newly supported modules: Pod::Usage, DBIx::SearchBuilder,
- DBIx::ReportBuilder, SVK::Command, SVN::Core, and the ':encoding()'
- IO discipline.
-
-* Bug Fixes
-
- - On non-Win32 systems, invoking pp-generated executable from PATH
- did not work.
-
- - Standalone executables were clobbered by existing perl environments
- with an identical "auto/IO" libpath as the author's environment.
-
- - Standalone executables did not work on systems with an unset
- dynamic load path environment variable (eg. LD_LIBRARY_PATH).
-
- - "pp -p -o multi.par 1.pl 2.pl; parl multi.par 1.pl" now works.
-
- - $ENV{PATH} and $ENV{TEMP} were truncated at first path delimiter.
-
- - "pp -f Bleach" did not work for ActivePerl on Win32.
-
- - Windows 9x systems were generating invalid cache directory names.
-
- - $ENV{path} is also recognized as $ENV{PATH} for Win32.
-
-[Changes for 0.80 - March 17, 2004]
-
-* New Features
-
- - A comprehensive test suite for pp in contrib/automated_pp_test/.
- It is run as part of the "make test" process from t/2-pp.t.
-
- - Much better support for "pp -i" and "pp -N" (--icon/--info)
- using the Win32::Exe module. You may now use EXE and DLL as
- icon files.
-
- - If PAR_GLOBAL_CLEAN (-C, --clean) is not set, we now preemptively
- extracts files under the cache directory. That made POSIX.pm
- and other modules that depends on %INC pointing to real files
- work correctly.
-
- - Now uses SHA-1 to create temporary directories and files,
- instead of mtime.
-
- - Verbosity level is now 1..3, not 0..5; "pp -v" now takes
- an optional integer, so "pp -v input.pl" is no longer an error.
-
- - New flags "-vv" and "-vvv", as shorthands for "-v 2" and "-v 3".
-
- - The user-settable PAR_CLEAN and PAR_TEMP environment variables has
- been renamed to PAR_GLOBAL_CLEAN and PAR_GLOBAL_TEMP; the original
- variables are still accessible within the program. This is so that a
- pp-generated program can exec() or system() another one without
- crippling its environment variables.
-
- - File lookups are now case-insensitive on case-insensitive filesystems.
-
- - Another Tk-based GUI in contrib/gui_pp/; not installed by default.
-
- - OOified "pp" in contrib/object_oriented_pp/; not installed by default.
-
-* Bug Fixes
-
- - "pp -d" (--dependent) prevented "pp -C" (--clean) from working.
-
- - The "pp -m" (--multiarch) option was implemented incorrectly
- and thus broken.
-
- - Many documentation tweaks.
-
- - Previously, "pp -M" (--module) did not add the module itself,
- only its dependencies.
-
- - Suppress a bogus warning when $ENV{$Config{ldlibpthname}} is empty.
-
- - "parl -v" without Module::Signature installed could delete all
- files within the current directory. Oops.
-
- - On *nix systems, pp-generated executables erroneously linked to
- libperl even if "pp -d" (--dependent) is not set.
-
- - Spurious =cut directives in source files is now handled gracefully
- by PAR::Filter::PodStrip.
-
- - "pp -L" (--log) now logs all output messages to the log file,
- not just the ones printed by "pp" itself.
-
-
-[Changes for 0.79 - January 8, 2004]
-
-* Bug Fixes
-
- - Setting PAR_CLEAN had the reversed effect. Oops.
-
- - Dynamic libraries in cached directories was not detected
- properly, resulting in "permission denied" errors during
- certain race conditions.
-
-[Changes for 0.78 - January 7, 2004]
-
-* New Features
-
- - By default, executables generated by "pp" will now store
- extracted files in cache directories. You may override
- this by setting the PAR_CLEAN environment variable to "1",
- or generate executables using "pp -C".
-
- - New "pp -C" (--clean) option to make the generated executable
- clean up temporary directories after each run.
-
- - PAR_CLEARTEMP is renamed to PAR_CLEAN.
-
-* Bug Fixes
-
- - On Win32, temporary directories containing shared libraries
- was not being properly cleaned up.
-
- - If no suitable temporary directories are found, use the current
- directory (".") instead of the root directory ("/").
-
-[Changes for 0.77 - January 1, 2004]
-
-* New Features
-
- - New "pp -c" and "pp -x" (--compile/--execute) options run the
- script with "perl -c" to check for dependencies.
-
- - Also, the new "pp -n" (--noscan) command skips the default
- static scanning altogether.
-
- - Added support for "pp -c/-x/-n" to tkpp.
-
- - For dynamically-built perls, pp-generated .exe files will now
- appear in the process table with the same name as it was launched,
- instead of "par.exe".
-
- - New filter "Obfuscate", which uses B::Deobfuscate to strip away
- PODs and comments, as well as mangling variable names.
-
- - Merged tkpp 1.1 from Doug Gruber.
-
- - OS/2 is now supported.
-
- - External Zlib is no longer required to run pp-generated binaries.
-
-* Bug Fixes
-
- - Makefile.PL was failing if $Config{cc} contains spaces.
-
- - No longer needs setting "Windows 95 compatible mode" to run on WinXP.
-
- - On Win9x with Perl 5.6.1, "nmake" was failing due to extra "@[...]"
- symbols in Makefile. It should be fixed now.
-
- - The "bad signature" problem with newer Archive::Zip versions is fixed.
-
- - App::Packer::Backend::PAR was misplaced into App/Packer/PAR.
-
- - Signature tests were failing under new ExtUtils::MakeMaker versions.
-
- - ActiveState's PPM building machine was having problem with PAR;
- a ".pdb" entry in MANIFEST.SKIP is added to fix that.
-
- - Some self-built PAR instances on Windows were failing due to
- mismatching short and long pathnames.
-
-[Changes for 0.76 - October 28, 2003]
-
-* New Features
-
- - Input filters. "pp --filter Bleach" now obfuscates the incoming
- script with PAR::Filter::Bleach; "pp --modfilter Bleach" applies
- Bleach to all packed modules.
-
- - Two previously built-in filters, PodStrip and PatchContent,
- are refactored out as PAR::Filter subclasses.
-
- - Two new filters, Bleach and Bytecode, are added for source-hiding
- purporses.
-
- - New utility, "tkpp", provides a GUI frontend to "pp".
-
- - New option, "pp --perlscript", to generate stand-alone scripts.
-
- - The old "PAR::Intro" documentation has been replaced by two
- new ones: "PAR::Tutorial" and "PAR::FAQ".
-
- - Tk pixmap (.xpm) files can now be packed with "pp --add".
-
-* Bug Fixes
-
- - Perl 5.8.1 has an off-by-one bug that prevents "parl" to
- function properly. We have now provided a workaround;
- this bug should also be fixed in Perl 5.8.2.
-
- - Fixed https support for LWP via the new Module::ScapDeps.
-
-[Changes for 0.75 - September 21, 2003]
-
-* New Features
-
- - "pp -o file.exe file.par" now packs file.par into file.exe;
- this means you can hand-tweak PAR files generated by "pp -p"
- before packing it into an executable.
-
-* Bug Fixes
-
- - Packing multiple programs by "pp script1.pl script2.pl" was
- producing syntax errors; fixed.
-
- - "pp -M datafile" now works.
-
- - Exit code from pp-packed executables now properly propagates out.
-
- - Fixed "use base" detection, Math::BigInt support and spurious
- signature warnings, by updated versions of Module::ScapDeps
- and Module::Signature.
-
- - On Win32, the PE info headers no longer show PAR_XXXXXXXXXXX.
-
-[Changes for 0.74 - August 20, 2003]
-
-* New Features
-
- - pp now has a set of "PatchContent" rules, dealing with
- non-PAR-compatible modules: Tk, Tk::Widget, Win32::API::Type,
- Win32::SystemInfo, SQL::Parser, diagnostics. These rules may
- get refactored back to PAR.pm in the future.
-
- - New function, PAR::reload_libs(), to reload currently used
- libraries inside PAR files.
-
- - PAR.pm itself is now never packed into pp-generated files,
- to perserve interface compatibility and reduce bloat.
-
- - PAR.pm now handles "use PAR 'othercode.par'" called from
- program or modules inside PAR files, even recursively.
-
- - A new icon for Win32 that is hopefully prettier.
-
-* Bug Fixes
-
- - All data after __DATA__ are preserved for included libraries.
- This helps self-reading modules like Net::LDAP::Constants.
-
- - PAR::read_file() was broken. It now works.
-
- - "use PAR" inside pp-generated executables was failing with
- 'file too short' errors due the mishandling of seek/tell.
-
- - Occasional crashes on Win32 due to rmdir() called too early
- with DLLs still open is fixed; however, "pp -d" executables
- may still exhibit this problem.
-
- - "pp -X" used to only take full pathnames as arguments.
- It now also takes "Module::Name" and "Module/Name.pm".
-
- - Dynamically built Perl under Cygwin failed to build, because
- libperl.dll.a was not found.
-
- - Eliminated "callback called on exit" warnings, and the related
- "access violation" error on Win32.
-
-[Changes for 0.73 - August 6, 2003]
-
-* New Features
-
- - The PAR Homepage is now online at http://par.perl.org/.
- Documentations have been changed to link to it.
-
-* Bug Fixes
-
- - Tk applications can now properly access xpm/xbm files
- with Tk->findINC.
-
- - On Win32, pp-generated executables could not start from
- Explorer, if its path contains space characters. Fixed.
-
- - On Win32, pp-generated executables used to leave around an
- empty directory in $ENV{TEMP}. It is now properly rmdir'ed.
-
- - Some systems (notably OpenBSD and Debian) does not put their
- libperl.so in the default location, which breaks the build
- process; now searches inside $ENV{$Config{ldlibpthname}} and
- $Config{libpth} to find it.
-
-[Changes for 0.72 - August 2, 2003]
-
-* New Features
-
- - CHECK and INIT blocks in programs inside PAR are now supported.
-
-* Bug Fixes
-
- - Two debug statements were mistakenly left in the source,
- resulting in "trying to get rid of /tmp/par_priv.xxxx.tmp"
- messages.
-
- - Building on Linux with GCC 3.2.2 was failing due to massive
- heap required for my_perl.c. Fixed by splitting it into 3k
- chunks.
-
- - Depends on Module::ScanDeps 0.21; it supports utf8 on Perl 5.6.1
- and can significantly reduce executable file size by eliminating
- unneccessary shared libraries.
-
-[Changes for 0.71 - July 30, 2003]
-
-* Bug Fixes
-
- - A nasty data-loss bug has been uncovered immediately after the
- previous release; it only affects Windows platforms, and may
- cause all files to be erased under the current root (\) directory.
-
- - Building on Red Hat linux was failing, with error message that
- says "my_perl not declared". This has since been fixed.
-
-[Changes for 0.70 - July 29, 2003]
-
-* New Features
-
- - On machines with shared libperl, "pp" now makes truly stand-alone
- executables; the old behaviour is available with "pp --dependent".
-
- - Under Windows NT/2000/XP, "pp --icon=name.ico" now changes the icon
- for the generated executable; otherwise, a default "white camel"
- icon is used.
-
- - "use PAR 'http://example.com/foo.par'" now works, as does
- "perl -MPAR -Ihttp://example.com/foo.par".
-
- - PAR::Dist is now a mandatory prerequisite, which provides
- functions to turn any CPAN distribution into a PAR distribution,
- as well as to install, uninstall, sign and verify such files.
-
- - Integrated PAR::Dist into "par.pl" and "parl". For example,
- "parl -i Foo-0.01-i386-freebsd-5.8.0.par" installs a PAR
- distribution; "parl -v out.exe" verifies a digitally signed
- executable generated by "pp --sign".
-
- - A new option, "pp --multiarch", lets you generate PAR files
- that can work on several architectures.
-
- - "pp --sign" now adds digital signatures to generated executables
- and PAR files.
-
- - PAR files may now (recursively) contain other PAR files inside
- their par/ directories.
-
- - shlib/ and par/ directories inside PAR files can now contain
- architecture- and perl-version-specific subdirectories.
-
- - The "Cross-Platform Packaging and Deployment with PAR" tutorial
- is now online as http://www.autrijus.org/par-tutorial/.
-
-* Bug Fixes
-
- - MANIFEST.SKIP was broken on Win32.
-
- - C compilers that doesn't handle long line well can now
- compile PAR.
-
- - DLL files inside the same auto/ library as XS modules
- was not properly extracted and loaded. This specifically
- affects Win32.
-
- - Because parl's @INC is '.', pp-generated executables may
- miss IO.dll and other shared libraries since they couldn't
- be correctly found in @INC.
-
-[Changes for 0.69 - May 31, 2003]
-
-* New Features
-
- - Under Perl 5.8, "pp -p" now works with Apache::PAR.
- See http://aut.dyndns.org/par-tutorial/slide018.html for
- a simple example.
-
- - "pp -M filename" now adds "filename" to /, not /lib/,
- unless filename ends in (pm|ix|al). This makes it
- possible to bundle "web.conf" needed by Apache::PAR.
-
- - "pp -l" now searchs in system library paths, and
- appends "lib" / prepends ".$dl_ext" where necessary.
-
-* Bug Fixes
-
- - PAR segfaults on some Unix platforms due to a NULL pointer
- used in mktmpdir.c. Fixed.
-
- - "pp -o out.par -p -e '...'" now honors -o; previously
- it used "a.out.par" anyway.
-
- - Inhibited spurious uninitialized warnings under -w in the
- POD-stripping code.
-
- - Win32 did not properly cleans up PAR_TEMP directory, resulting
- in failure for executables that reused the same PID. Fixed.
-
-[Changes for 0.68 - May 26, 2003]
-
-* New Features
-
- - New 'pp -l' option to pack additional shared libraries (DLLs).
-
- - POD-stripped libraries inside PAR files now have #line directives
- inserted, so they report the original line numbers on failure.
-
- - PAR files generated by 'pp' now has a MANIFEST file that can be
- viewed by Gecko-based browsers with Javascript turned on, e.g.:
- jar:http://aut.dyndns.org/par/test.par!/MANIFEST
-
-* Bug Fixes
-
- - Each pp-executable instance now creates its own PAR_TEMP directory;
- this avoids permission errors when multiple users run the same binary.
- As a consequence, PAR_CLEARTEMP is now set to "1" by default.
-
- - Newer versions of shared Zlib library no longer causes "pp" to
- generate broken executables.
-
- - Fixed dynamic loading on Cygwin was failing due to missing +x mode.
-
- - Like "use lib", "use PAR 'name.par'" now unshift()s instead of push()es
- into @INC. Same applies for "par.pl -A" and "parl -A".
-
- - Fixed building on ActivePerl 626 and below due to a missing $Config{ld}.
-
-[Changes for 0.67 - April 1, 2003]
-
-* New Features
-
- - PAR now works on Cygwin and MinGW/MSYS.
-
- - Globbing support in PAR::import(): use PAR "/path/*.pm";
-
- - New license clarification messages added to POD and 'pp -V'.
-
- - All 'pp' options now has a short form (-o) and a long form (--output).
-
- - Revamped documentation for 'pp'.
-
- - New -g (--gui) flag for 'pp' to build console-less Win32 executables.
-
-* Bug Fixes
-
- - Building on Darwin Perl 5.6.0 was broken with 'cc -s'.
-
- - Building on 5.6.0 was broken due to bad 'base.pm'.
-
- - Win32 Tk::Widget autoloading was broken due to a binmode() bug.
-
- - IPC::Run was pod-stripped incorrectly. Fixed.
-
- - Depends on Module::ScanDeps 0.19, which supports utf8 and .ph files.
-
- - Better AutoInstall support, which uses 'sudo' where necessary.
-
-[Changes for 0.66 - March 20, 2003]
-
-* New Features
-
- - Adds PAR::Intro, a PODified version of the online presentation.
-
- - Adds App::Packer::Backend::PAR, a bridge between PAR and App::Packer.
-
- - Scripts and modules are now searched in "/" last, instead of first.
-
- - Experimental patch for packing shared libraries via "pp -l".
-
- - HTTP fetching of precompiled packages in addition to FTP.
-
-* Bug Fixes
-
- - Makefile.PL now downloads precompiled packages only if needed.
-
- - AutoInstall has been made to work for an easier installation.
-
- - The redundant "parl.exe.bat" is no longer created on Win32.
- - Pre-0.63 PARs used to leave broken .dll in TEMP; now they're cleaned.
-
- - "pp c:\something.pl" no longer treats c: as a relative path.
-
- - "pp -I dir" now searches 'dir' first, instead of last.
-
- - "pp" was broken on Perl 5.6.0 due to => stringification bugs.
-
- - Support for Tk::Widget autoloading has been added.
-
- - "parl" was not stripped if "gcc" was invoked as "cc"; fixed.
-
- - On a machine with multiple "parl"s, "pp" now uses the correct one.
-
- - File::Temp was missing as a dependency.
-
-* Known Issues
-
- - Cygwin support is still broken.
-
- - PAR does not include utf8_heavy.pl nor unicore/* for scripts that
- has "use utf8;". This has since been fixed by Module::ScanDeps 0.18.
-
-[Changes for 0.65 - March 9, 2003]
-
-This release comes with several significant improvements:
-
-* Automatic binary installation
-
- Previously, users without a C compiler cannot build the 'parl'
- executable, and is therefore unable to create self-contained
- binaries using 'pp'.
-
- Now, if there is a binary package available for that architecture
- under my CPAN directory, the Makefile.PL script will automatically
- fetch it, unpack into blib/, and the installation will continue as
- normal, resulting in a fully-functional 'pp'.
-
- This feature is part of the soon-to-be-released Module::Install
- framework; it will greatly benefit all CPAN authors with non-pure-perl
- distributions.
-
-* POD stripping
-
- Packages generated with 'pp' will now strip POD sections from all
- packed dependencies (your own scripts is unaffected); all binary
- executables will save at least 276732 bytes, with additional ~20%
- saving in additional packed dependencies.
-
- You can turn off this feature with the PAR_VERBATIM environment
- variable.
-
-* XS Incompatibility solved
-
- Because 'pp'-generated executables includes some fixed version of
- shared libraries (IO, Zlib, etc), they used to break when the
- target system has different version of shared libraries.
-
- Now PAR::Heavy intercepts DynaLoader::dl_expandspec to always prefer
- the library inside the PAR file, so this issue is resolved.
-
-* 5.6.1 Reclaimed
-
- Thanks to Sisyphus and various others, building on Perl 5.6.1
- (with its old ExtUtils::MakeMaker and lack of PTHREAD_ATFORK)
- now works again.
-
-[Changes for 0.64 - March 2, 2003]
-
-* New Features
-
- - The t/0-signature.t test is re-enabled for people using my
- Module::Signature to verify the module's OpenPGP signature.
-
- - This release is the first distribution on CPAN to use the
- Module::Install framework, which is a stand-alone, extensible
- drop-in replacement for ExtUtils::MakeMaker that needs no
- extra action/prerequisites for end users.
-
-* Bug Fixes
-
- - Dynamic loading on Win32 was broken, due to a binmode() bug
- reported by Bill Atkins, D. Menzel and others.
-
- - Building on Win32 in directory names that contain spaces
- did not work.
-
-[Changes for 0.63 - February 6, 2003]
-
-* Bug Fixes
-
- - The 'parl' binary (which replaces the old 'par' or 'par.exe')
- didn't work properly when bundling perl modules for self-
- contained executables, rendering 'pp' useless on machines
- without core perl.
@@ -0,0 +1,1277 @@
+1.008 2015-01-24
+
+ - Fix #101662: Prevent shared libs from being cached in memory on AIX
+ applied patch from Thoke@northpeak.org, thanks.
+ Note: this is one half of the fix, the other is in PAR::Packer
+
+ - Fix #86178: dll files not extracted to shlib subfolder when using -l or to other folder when using -a
+ extract all files from the zip, including all DLLs
+
+ - Fix #86014: [PATCH] POD fix
+ applied Debian patch from gregor herrmann <gregoa@debian.org>, thanks!
+ while we're at it fix two bad L<> hyperlinks reported by podchecker
+
+ - Fix RT #86650: typo fixes
+
+ - reformat ChangeLog file according to CPAN::Changes::Spec
+ and rename it to Changes
+
+ - Update to Module::Build 1.14
+
+1.007 2012-10-22
+ - Hopefully fix "pp -C ..." for any modules that assume an
+ actual tree of files, e.g. looking for all installed modules Foo::Bar::*;
+ call _extract_inc even when $ENV{PAR_CLEAN} is true
+ - update to Module::Install 1.06
+
+1.006 2012-10-15
+ - Fix RT #78633: PAR::import ignores url => $repo_client_object
+ applied patch from KENO, thanks!
+ - Fix RT #73491: cache directory naming problem
+ In PAR::SetupTemp::_get_par_user_tempdir (actually _find_username)
+ we try to sanitize username (so that there are no problematic
+ characters in the name of the per-user cache directory), but the
+ strategy fails miserably for some charset encodings.
+ E.g. for EUC-KR or CP949 (as in the bug report) we may produce an
+ illegal sequence of bytes; in other cases we may cause collisions
+ (different usernames mapping to the same directory name).
+ Fix the problem once and for all by encoding the username
+ (bytewise) as 2 hex digits.
+
+1.005 2011-12-02
+ - bump Perl version requirement to 5.8.1 (Schwern: The End Of 5.6 Is Nigh!)
+ - run all tests using a nonce PAR_TMPDIR (otherwise CPAN Testers
+ goes crazy as top level /tmp/par-USER directories (or similar)
+ from previous tests may now be considered "unsafe")
+
+1.004 2011-11-30
+ - back out r1241: it causes errors in PAR::Packer's test suite
+ - change "unsafe directory" error message to match the wording
+ used by PAR::Packer
+ - remove "debian" sub directory: it isn't released to CPAN and
+ Debian will supply its own anyway
+ - remove some cruft from MANIFEST.SKIP
+
+1.003 2011-11-28
+ - RT #69560/CVE-2011-4114: PAR packed files are extracted to unsafe
+ and predictable temporary directories
+ (Note: this bug was originally reported against PAR::Packer, but
+ it applies to PAR as well)
+ - create parent of cache directory (i.e. /tmp/par-USER) with mode 0700
+ - if it already exists, make sure that (and bail out if not)
+ - it's not a symlink
+ - it's mode 0700
+ - it's owned by USER
+ - Fix a problem packing XML::LibXSLT on Windows (see the thread starting
+ with http://www.nntp.perl.org/group/perl.par/2011/02/msg4919.html)
+ - Die (with a hopefully useful message) if any error is encountered
+ during an Archive::Zip extract operation
+
+1.002 2010-07-25
+ - Fixes to VERSIONs of PAR::Setup*.
+ - No change in behaviour since 1.001
+
+1.001 2010-07-25
+ - RT #57399: extract everything (including DLLs) in File::ShareDir
+ directories. Module::ScanDeps classifies everything in
+ File::ShareDir directories as "data", but PAR uses it's own
+ heuristics what to extract from a .par.
+ - Previous release was missing META.yml.
+ - Upgrade Module::Install to 1.00
+
+1.000 2010-04-10
+ - Fix defined(%hash) deprecation warning in PAR::Heavy
+
+0.994 2009-07-23
+ - Fix for the PAR::Heavy fix to the INC priority handling.
+
+0.993 2009-07-19
+ - The priority (fallback=>0) repositories should look at @PAR_INC
+ for the loaded PARs instead of @PAR_INC_LAST.
+ - Don't reload from a downloaded .par file after installing it
+ via "upgrade".
+ - Band-aid fix for the loading priority of shared librares from
+ PAR files: Try PAR's first, the local stuff, then fallback-PARs.
+ - Initial support for running external perl scripts from a packaged
+ interpreter.
+
+0.992 2009-04-05
+ - Support for non-fallback repositories.
+
+0.991 2009-03-10
+ - Promote previous release to stable release.
+
+0.989_01 2009-03-02
+ [Bug fixes, etc.]
+ - Track the locations of all archives that have been extracted to
+ $ENV{PAR_TEMP}/inc in this run.
+
+0.988 2009-03-01
+ - Promote previous release to stable release.
+
+0.987_01 2009-02-20
+ [New features]
+ - Better cleanup of the $TMPDIR/par-$USER/temp-$$ directories
+ that are typically used as caches in the "use PAR 'foo.par'" use case.
+
+ [Bug fixes, etc.]
+ - Very slightly more careful handling of PAR-specific environment
+ variables.
+ - Cleanup of PAR::SetupTemp
+
+0.986 2009-02-19
+ - Promote to stable release.
+
+0.985_01 2009-02-02
+ [New features]
+ - Support for the brand new static dependency resolution
+ of PAR::Repository::Client 0.23.
+
+ [Bug fixes, etc.]
+ - Fix issue with running scripts from repositories:
+ The INC hooks used to be set up too late for this to
+ automatically pick up dependencies.
+
+0.984 2009-01-25
+ [New features]
+ - Implemented the auto-upgrading option for loading and installing
+ from PAR::Repositories.
+
+ [Bug fixes, etc.]
+ - Fix issue with PAR::Repository::Client development versions.
+
+0.983 2008-09-12
+ [Dependencies]
+ - Require AutoLoader 5.67 which contains a PAR-related
+ bug-fix.
+ - Require PAR::Dist 0.32.
+
+ [Internal changes]
+ - The full extraction process _extract_inc
+ (which is triggered when a non--clean pp packaged
+ executable is run) can now be forced to do the
+ extraction (instead of doing if !-d).
+ - That same extraction routine now accepts Archive::Zip
+ handles or file names.
+ - When, during the full extraction, the extracted paths
+ are to be added to @INC, we now make sure they're not
+ in @INC yet.
+
+0.982 2008-08-10
+ [New features]
+ - Moved the routines that setup the PAR_TEMP environment variable
+ to a separate module in the distribution so it's possible to
+ use PAR::Repository::Client without loading all of PAR.
+ - Same for the function that sets up PAR_PROGNAME.
+
+ [Bug fixes, etc.]
+ - Upgrade to Module::Install 0.77
+ - Fix for running scripts from repositories.
+
+0.980 2008-05-22
+ [Bug fixes, etc.]
+ - The function PAR::get_filehandle() that was introduced in 0.979
+ is really broken because Archive::Zip is broken. Turns out calling
+ Archive::Zip::Member->fh() returns a file handle to the zip file,
+ not a virtual/tied/whatever file handle to the member file.
+ Therefore, the get_filehandle() function has been removed again until
+ we work around this issue.
+ - Upgrade to Module::Install 0.73
+
+0.979 2008-05-13
+ [New features]
+ - New function PAR::get_filehandle() returns a file handle for a file
+ in any open .par files. Similar to read_file().
+
+0.977 2007-10-19
+ [Bug fixes, etc.]
+ - HPUX doesn't like shared libraries being unlinked while in use. So don't
+ try to do this even in clean mode. (Workaround will be in par.pl)
+ (Scott Stanton)
+ - Fix DLL extraction file name matching in PAR::Heavy.
+ - Save two system calls per DLL during DLL extraction in PAR::Heavy.
+
+0.976 2007-07-29
+ [New features]
+ - Use Archive::Unzip::Burst for unpacking binary executables if
+ available. (This yields a significant startup speed-up.)
+
+ [Bug fixes, etc.]
+ - Removed the auto_install feature from Makefile.PL. auto_install
+ is conceptually broken.
+
+0.973 2007-02-03
+ [New features]
+ - A new option for the "use PAR { ... };" use case:
+ "no_shlib_unpack" signals that for this particular .par file,
+ shared libraries that were added with the --lib option should
+ not be extracted. This allows users to do their own cache
+ handling for these libraries.
+ - PAR no longer unpacks *all* shared libraries by default but only
+ those in the shlib/ directory (i.e. added with --lib).
+ The shared libraries in auto/ should be picked up by the DynaLoader
+ hack.
+ - If available, the prefork.pm module will be used to declare
+ run-time dependencies for better memory use in forking
+ environments.
+ - PAR now uses a caching mechanism to speed up the extraction process.
+ This should particularly impact users of the "use PAR {file =>...}"
+ form.
+
+ [Bug fixes, etc.]
+ - Applied an optimization of the unpacking process on case insensitive
+ file systems.
+
+0.972 2007-01-16
+ [Bug fixes, etc.]
+ - Removed PAR::AutoLoaderFix again. It wasn't working as expected
+ all the time.
+ - To fix the problem AutoLoaderFix was supposed to fix, we now
+ require AutoLoader 5.62 or newer which was just recently released
+ to CPAN. (Previously only available from blead perl.)
+
+0.971 2007-01-12
+ [Bug fixes, etc.]
+ - Fixed typo in the POD. (Jerrad Pierce)
+ - Included fix for a bug in AutoLoader.pm as shipped with all
+ perl versions up to and including 5.8.8 as PAR::AutoLoaderFix.
+ This cures a problem of endless looping when the %INC entry of a
+ module doesn't point to a file of the same name.
+ This may happen during "use PAR 'foo.par'".
+
+0.970 2006-12-03
+ [This release introduces some rather radical changes, so read carefully:]
+ [All PAR::Packer related logic has been moved to a separate distribution,]
+ PAR-Packer. This includes pp, parl and all packaging tools. This way,
+ PAR becomes a pure-Perl distribution that can be most easily installed by
+ users of software which requires PAR.
+ Developers who want to use the PAR packager, pp, need to install the
+ PAR-Packer distribution from CPAN.
+
+0.961 2006-11-23
+ [Bug fixes, etc.]
+ - PAR::StrippedPARL::Base->write_parl() failed to work if the @INC
+ directories contained spaces in 0.960. (Steven Mackenzie)
+ - Much improved documentation of the environment variables
+ (Glenn Linderman)
+ - Fix for a spaces-in-pathname problem on Windows for
+ t/30-current_exec.t. (Malcolm Nooning)
+
+0.960 2006-11-21
+ [Bug fixes, etc.]
+ - myldr/Makefile.PL fix: Clean up myldr/usernamefrompwuid.h.
+ - Silence warning in myldr/internals.c.
+ - Silence warnings seen on Irix from myldr/env.c.
+ - Skip most tests in 10-parl-generation.t if there is no parl.
+ - Skip loading ActiveState Perl's "sitecustomize.pl" in par.pl.
+ - Load modules via require and other files via do.
+ - The parl-regeneration-for-every-pp-call addition of the 0.958 release
+ should now also work for static perls.
+
+ [New features]
+ - Adressing RT ticket #6612: Now using getpwuid() to determine the
+ user name if supported by the OS.
+
+0.959 2006-11-12
+ [This is just a hotfix release because 0.958 lacked META.yml. One day, I will]
+ switch from Module::Install to Module::Build...
+
+0.958 2006-10-25
+ [Bug fixes, etc.]
+ - myldr/Makefile.PL fix: make static.o depend on mktmpdir.c, my_perl.c,
+ my_par.c. (Roderich Schupp)
+ - Modules included with the -M option to pp were previously scanned
+ for dependencies but not mapped through the
+ %Module::ScanDeps::Preload hash for custom dependencies. That's
+ fixed now.
+ - $ENV{PAR_RUN} isn't set by PAR::Packer any more because nothing in
+ the PAR sources uses it. $ENV{PAR_RUN} is no longer used by PAR
+ at all.
+ - Unified the environment variables which are looked at for
+ finding the system's temporary directory.
+
+ [New features]
+ - During the build process, PAR appends stripped down copies of parl
+ (and parldyn if applicable) to the data classes
+ PAR::StrippedPARL::Static and ::Dynamic. These
+ parls-without-embedded-modules are used for packaging so the formerly
+ embedded modules are now packaged from the packaging system. (Instead
+ of stemming from the system where PAR/parl was built.)
+ - The "use PAR { repository => $url };" syntax now also supports
+ the use of user-constructed PAR::Repository::Client objects instead
+ of an URL.
+ - The -F (module code filter) option now supports selective filtering
+ of modules. The syntax is "-F FILTER=REGEX" or - as before -
+ "-F FILTER". The regular expression is applied to the *file name*,
+ of the module inside the PAR (e.g. Foo/Bar.pm). This behaviour was
+ chosen over matching against the module name (e.g. Foo::Bar) because
+ the filters can be applied to module-like and script files as well
+ (.pl, .al, etc.).
+ - Updated PAR/FAQ.pod with the new FAQ's from the PAR wiki.
+ - Added a POD file PAR/Environment.pod which is intended to become
+ an index of all environment variables PAR uses of its own or
+ recognizes from its users. Still mostly a stub.
+
+0.957 2006-10-24
+ [Bug fixes, etc.]
+ - Fix executable PARs top properly detect embedded scripts
+ named the same as the executable. (Jesse Vincent)
+ - Comment out the call to par_current_exec_proc (in the C loader)
+ which breaks the use of symlinks to pp-ed executables when not
+ called with a path. (I.e. using a search in $PATH).
+
+0.956 2006-10-03
+ [Bug fixes, etc.]
+ - This is another hotfix release. Fixed a mindless bug introduced in 0.955.
+
+0.955 2006-10-03
+ [Bug fixes, etc.]
+ - 0.952 introduced removal of system module search paths if -B is in
+ effect. This resulted in some valid PAR-related paths being removed
+ as well. Fixed. Upgrading from 0.952 and 0.954 is suggested.
+ - Changed the use of hard-coded '/' as path-separator to using File::Spec.
+
+0.954 2006-09-26
+ - This release is equivalent to 0.953. The 0.953 CPAN upload is broken!
+
+0.953 2006-09-18
+ [Bug fixes, etc.]
+ - Added optional POD tests.
+ - Modified -B so that if -B is in effect, all entries are stripped
+ out of @INC except for the PAR hooks. This happens right before
+ the script contained in the pp-ed binary is executed.
+
+0.952 2006-08-22
+
+ [New features]
+ - Added the "install" option to the PAR loading syntax.
+ If specified, the contents of the PAR distribution are
+ permanently installed.
+ This requires PAR::Repository::Client 0.04.
+
+ [Bug fixes, etc.]
+ - Fixed broken META.yml in 0.951.
+
+0.951 2006-08-12
+ (This includes any changes up to 0.950.)
+
+ [New features]
+
+ - Introduced new PAR loading syntax and semantics:
+ use PAR { file => 'path/to/par/or/URL' };
+ ==> equivalent to "use PAR 'path/to/par/or/URL';"
+ - Introduced the 'fallback' option: (default = 0)
+ use PAR { file => 'foo.par', fallback => 1 };
+ ==> Loads modules from the PAR file only if loading
+ them from @INC did not succeed.
+ - Introduced the 'run' option which executes a script
+ in a PAR archive just like
+ perl -MPAR foo.par script.pl
+ - If PAR::Repository::Client is installed, you can add a
+ repository of .par distributions to your library search path
+ as follows:
+ use PAR { repository => 'http://foo' };
+ - Of course, 'run' also works with repositories:
+ use PAR { repository => 'http://foo', run => 'my_app' };
+ (This searches the repository for any distributions that have a
+ my_app script.)
+ --> For details on repositories, have a look at the
+ PAR::Repository::Client module.
+
+- Bug fixes, etc.
+
+ - Commented a couple of the routines in PAR.pm. (Yay!)
+ - New test script for the new fallback loading feature.
+ - Fixed a bug in the Spreadsheet::ParseExcel handling in
+ PatchContent.pm.
+
+0.942 2006-07-22
+
+ [Bug fixes, etc.]
+
+ - Better support for diagnostics.pm (in conjunction with
+ Module::ScanDeps 0.62.)
+ - Now requiring Module::ScanDeps 0.62.
+
+0.941 2006-06-20
+ (No, PAR isn't stagnating. It's just that 1.00 would draw close if
+ we continued with 0.01 increases.)
+
+ [Bug fixes, etc.]
+
+ - Version 0.94 of PAR would use the same cache area for all pp-ed
+ applications due to a faulty hotfix for Digest::SHA. This applies
+ to PAR 0.94 only. Think of 0.941 being PAR 0.94 done right.
+
+0.94 2006-06-01
+
+ [New Features]
+
+ - Added support for reading options to pp from a file using a
+ '@filename' argument to pp: pp -o foo --gui @filename foo.pl
+
+ [Bug fixes, etc.]
+
+ - Workaround for a bug in Digest::SHA 5.38 and 5.39 that would prevent
+ PAR from being built.
+ - Fixed details in the 2-pp.t test file.
+ - Now recognizes text files that aren't picked up by the -T operator
+ but by the "file" tool.
+ - Applied Roderich Schupp's patch to 30-current_exec.t to fix a
+ path issue.
+ - Now requiring Module::ScanDeps 0.60 which fixes a couple of bugs
+ which might be observed as PAR bugs.
+ - Now working well with Spreadsheet::ParseExcel which uses an invalid
+ POD section to comment out a code block. This wasn't recognized by
+ PAR::Filter::PodStrip as POD and hence partly left in...
+ - If the output directory doesn't exist, we create it now and output
+ a meaningful error message if that failed.
+
+0.93 2006-05-19
+
+ [New Features]
+
+ - Added support for PAR_TMPDIR (PAR_GLOBAL_TMPDIR) so that the
+ temp directory can be controlled for just the PAR file bits.
+ (Leolo)
+ - Added par_current_exec_proc() which finds the file of the current
+ executable in /proc, if possible. (Leolo)
+ - Added par_current_exec() which finds he file of the current
+ executable, if possible on this OS. (Leolo)
+ - par_findprog() now uses par_current_exec() if possible.
+
+ [Bug Fixes, etc.]
+
+ - Upgraded to Module::Install 0.62+ (Audrey Tang, Steffen Mueller)
+ - Document a strange interaction with chdir() and relative
+ paths. (Chris Dolan)
+ - Documented the bits that make up PAR_TEMP. (Leolo)
+ - Fixed the call to par_findprog. path (aka val) was set to
+ tmpdir. (Leolo)
+ - Documented the CACHE name at the end of a self-executing
+ PAR. (Leolo)
+ - myldr/Makefile.PL now generates some dependencies for
+ main.c (Leolo)
+ - Applied patch from RT ticket. (tsee)
+ https://rt.cpan.org/Ticket/Display.html?id=13959
+ - Applied Ivan Kudryavtsev's patch that fixes a couple of calls
+ to PAR subroutines in PatchContent filtered code. (tsee)
+
+0.92 2006-02-22
+
+ [Bug Fixes]
+
+ - Now requiring Module::ScanDeps 0.56 which handles autouse
+ correctly.
+ - Now shipping with a correct SIGNATURE. (Which was broken for
+ 0.91.)
+
+0.91 2006-02-13
+
+ [Bug Fixes]
+
+ - Applied Alan Stewart's patch which fixes @ARGV pollution in
+ daughter programs. See also
+ http://www.nntp.perl.org/group/perl.par/2152
+ - Now mentioning the ENV var "PAR_VERBATIM" in the documentation.
+ See also http://www.nntp.perl.org/group/perl.par/2196
+ - Applied Malcolm Nooning's fix for the test suite. We used to get
+ failed tests on Windows because of spaces in path names.
+ - Applied Roderich Schupp's and Malcolm Nooning's patches to
+ the test suite fixing problems with Cygwin.
+ - Applied Vincent Ladeuil's patch to PAR::Filter::Bleach to return a
+ true value for modules that loaded okay.
+ - Changed 'PAR_BASE' in the Makefile.PL to 'SMUELLER'.
+
+0.90 2005-11-25
+
+ [Bug Fixes]
+
+ - When compiling with static libperl, myldr/ may fail "make"
+ due to sha1.c not generated properly.
+
+ - Pod stripping could fail on __DATA__ sections for files
+ with CRLF line endings.
+
+ - The documentation erroneously referred to the PAR_TEMP
+ environment variable, whereas it should be PAR_GLOBAL_TEMP.
+
+ - Compilation fixes for MinGW/MSYS.
+
+0.89 2005-06-10
+
+ [Bug Fixes]
+
+ - Stop static.c from pulling in Perl header files, otherwise
+ parl.exe ends up depending on the Perl DLL on Win32 when
+ Perl is built without PERL_IMPLICIT_SYS.
+
+ - With *nix and File::Path 1.06, par.pl's avoidance of loading
+ Cwd.pm caused syntax errors.
+
+0.88 2005-06-07
+
+ [Bug Fixes]
+
+ - Extracted .pl files should be loadable via the coderef-in-@INC too,
+ just like .pm files and autosplit files. This makes PAR work with
+ Perl 5.8.7 on Win32.
+
+ - Fix the build with GCC 4.0.
+
+ - If $ENV{PWD} is not defined, fallback to use `pwd` to obtain the
+ working directory for invoking.
+
+0.87 2005-01-31
+
+ [Bug Fixes]
+
+ - On Win32, some versions of File::Spec::Win32 contains explicit
+ "use Cwd;" lines, which renders parl.exe unusable.
+
+ - Executable made by "pp" may fail when invoked as "./a.out" or
+ "../a.out", due to incorrect PWD handling logic.
+
+0.86 2004-12-11
+
+ [New Features]
+
+ - New "pp -z" (--compress) option to set compression level (0-9).
+
+ - New "pp -T" (--tempcache) option to override the per-executable
+ directory name; it defaults to a hash of the executable, computed at
+ compile time. This makes startup much faster for large executables.
+
+ - The hash algorithm described above now prefers Digest::SHA if installed,
+ otherwise Digest::SHA1, then fallbacks to Digest::MD5.
+
+ - Functionality of "pp -X" is now extended: if the argument after -X is a
+ zip or par file, files in it are excluded from the produced executable,
+ and the executable will "use" the zip/par instead. For multiple -X args,
+ successive args are only "use"d if they contain additional unique files.
+
+ - "pp -l" now searches for libraries in "." and PATH in Win32.
+
+ - "pp -l" shared libraries are now added to %skip, so it will not
+ be included in both shlib/ and lib/.
+
+ - "pp -l" now chases symbolic links. For example, if "libsomelib.so"
+ is a symlink to "libsomelib.so.1", which is another symlink to
+ "libsomelib.so.1.2", pp now follows these symlinks and add the real
+ file the par, rather than "libsomelib.so".
+
+ - New contributed code in "contrib/stdio/": Useful Tk console for
+ "pp -g" users.
+
+ - New contributed tutorial documents, currently in "contrib/docs/",
+ which will eventually be turned into POD documents.
+
+ - Running "perl Makefile.PL" with $ENV{DEBUG} set to true now produces
+ "parl" with debug symbols.
+
+ - Remove Cwd.pm (and Cwd.so) from the bundled dependencies.
+
+ [Bug Fixes]
+
+ - More robust probing for case-insensitive file systems.
+
+ - PodStrip now attempts to match "standard" pod start before =cut,
+ otherwise =cut gets removed by itself.
+
+ - Win32 slashes are now normalized in privlib and archlib directories.
+
+ - Don't extract shared libraries to inc/, since they were extracted
+ in $PAR_TEMP already.
+
+ - Don't re-extract shared libraries in subdirectories, since they
+ are picked up by corresponding "use".
+
+ - Tk now exits properly with a non-zero exit() value.
+
+ - Fix libperl probing problem on Debian and Gentoo that manifests as a
+ "libperl5.8.so not found" error during runtime.
+
+ - gpp: Fixed typo in options with multiple filenames; cleaned up
+ pp parameters.
+
+ - When PAR_TEMP is set, shlib/ was not correctly added to the dynamic
+ load path environment variables.
+
+ - PAR now builds with Win32 VC++ without CVTRES.EXE available.
+
+ - Detection of cl.exe, gcc.exe and cc.exe is now case-insensitive.
+
+0.85 2004-07-02
+
+ [New Features]
+
+ - New version of "gpp"; see contrib/gui_pp/gpp_readme.txt for details.
+
+ [Bug Fixes]
+
+ - MANIFEST and META.yml were not properly updated by PAR::Packer.
+
+ - Setting directory aliases with "pp -a"/"pp -A" was broken.
+ Fixed, and tests were added for it.
+
+ - Statically-built executables was needlessly extracting libperl
+ each time it runs; now it is eliminated and hence much faster.
+
+0.83 2004-05-29
+
+ [New Features]
+
+ - Revamped PAR::FAQ and sychronized with par.perl.org.
+
+ - In pp-generated programs, $0 is now set to the pathname leading
+ to the invoked executable. Use $ENV{PAR_0} instead to get the
+ filename that contains the main perl program.
+
+ - Updated "contrib/gui_pp/gpp" to support PAR::Packer options.
+
+ [Bug Fixes]
+
+ - Core XS modules, such as Data::Dumper, were skipped by "pp".
+
+ - Fix t/2-pp.t for Cygwin by probing $Config{_exe} rather than uname().
+
+ - Scripts made by "pp -P", when invoked as "perl scriptname",
+ should not search for the same-named programs in PATH.
+
+ - Correctly remove leading slash and drive letters from absolute
+ filenames passed to "pp -a". Also normalized blackslahes to slashes.
+
+ - The PP_OPTS environment variable was not recognized.
+
+ - "pp -a dirname;diralias" was broken.
+
+ - "pp -f" and "pp -F" were broken.
+
+0.82 2004-05-24
+
+ [New Features]
+
+ - New module PAR::Packer provides an OO interface to "pp"'s
+ functionality; "pp" is now merely a thin wrapper for it.
+
+ - New module App::Packer::PAR is a modified version of
+ App::Packer, designed to work with PAR::Packer, and will
+ hopefully be merged back to App::Packer.
+
+ - The old, procedural "pp" is moved to contrib/; end-users
+ should notice no changes in "pp"'s behaviour.
+
+ - New options "pp -a" and "pp -A" (--addfile/--addlist) provides
+ ways to include extra files and directories in the package.
+
+ - The long option name for "pp -M" is changed from --add to
+ --module. The old name is still recognized but no longer
+ documented. Using "pp -M" to include non-library files
+ is now deprecated; use "pp -a" instead.
+
+ - par.pl and parl now writes messages to STDOUT, instead of
+ STDERR. As a consequence, t/2-pp.t no longer prints extra
+ warnings during "make test".
+
+ [Bug Fixes]
+
+ - On Non-Win32 platforms, perl 5.8.0 and earlier versions produced
+ pp-generated executables that immediately segfaults.
+
+ - Running pp-generated executables with absolute pathname failed
+ on statically-built perls.
+
+ - Tests were failing due to a missing pipe_a_command.pm in MANIFEST.
+
+ - Add the missing myldr/win32.coff for building on Cygwin/MinGW.
+
+ - If the "perl" in path is different from the perl interpreter used
+ for "make test", t/2-pp.t is known to fail and is now skipped.
+
+ - Cygwin failed t/2-pp.t because "parl" is spelled as "parl.exe" there.
+
+0.81 2004-05-23
+
+ [New Features]
+
+ - Regained support for Win9x, Cygwin and MinGW.
+
+ - PAR now supports 64-bit platforms, such as Tru64 and AIX.
+
+ - Cygwin and MinGW can now build EXEs with icons, too; MinGW can update
+ the icons, but Cygwin cannot.
+
+ - Newly supported modules: Pod::Usage, DBIx::SearchBuilder,
+ DBIx::ReportBuilder, SVK::Command, SVN::Core, and the ':encoding()'
+ IO discipline.
+
+ [Bug Fixes]
+
+ - On non-Win32 systems, invoking pp-generated executable from PATH
+ did not work.
+
+ - Standalone executables were clobbered by existing perl environments
+ with an identical "auto/IO" libpath as the author's environment.
+
+ - Standalone executables did not work on systems with an unset
+ dynamic load path environment variable (eg. LD_LIBRARY_PATH).
+
+ - "pp -p -o multi.par 1.pl 2.pl; parl multi.par 1.pl" now works.
+
+ - $ENV{PATH} and $ENV{TEMP} were truncated at first path delimiter.
+
+ - "pp -f Bleach" did not work for ActivePerl on Win32.
+
+ - Windows 9x systems were generating invalid cache directory names.
+
+ - $ENV{path} is also recognized as $ENV{PATH} for Win32.
+
+0.80 2004-03-17
+
+ [New Features]
+
+ - A comprehensive test suite for pp in contrib/automated_pp_test/.
+ It is run as part of the "make test" process from t/2-pp.t.
+
+ - Much better support for "pp -i" and "pp -N" (--icon/--info)
+ using the Win32::Exe module. You may now use EXE and DLL as
+ icon files.
+
+ - If PAR_GLOBAL_CLEAN (-C, --clean) is not set, we now preemptively
+ extracts files under the cache directory. That made POSIX.pm
+ and other modules that depends on %INC pointing to real files
+ work correctly.
+
+ - Now uses SHA-1 to create temporary directories and files,
+ instead of mtime.
+
+ - Verbosity level is now 1..3, not 0..5; "pp -v" now takes
+ an optional integer, so "pp -v input.pl" is no longer an error.
+
+ - New flags "-vv" and "-vvv", as shorthands for "-v 2" and "-v 3".
+
+ - The user-settable PAR_CLEAN and PAR_TEMP environment variables has
+ been renamed to PAR_GLOBAL_CLEAN and PAR_GLOBAL_TEMP; the original
+ variables are still accessible within the program. This is so that a
+ pp-generated program can exec() or system() another one without
+ crippling its environment variables.
+
+ - File lookups are now case-insensitive on case-insensitive filesystems.
+
+ - Another Tk-based GUI in contrib/gui_pp/; not installed by default.
+
+ - OOified "pp" in contrib/object_oriented_pp/; not installed by default.
+
+ [Bug Fixes]
+
+ - "pp -d" (--dependent) prevented "pp -C" (--clean) from working.
+
+ - The "pp -m" (--multiarch) option was implemented incorrectly
+ and thus broken.
+
+ - Many documentation tweaks.
+
+ - Previously, "pp -M" (--module) did not add the module itself,
+ only its dependencies.
+
+ - Suppress a bogus warning when $ENV{$Config{ldlibpthname}} is empty.
+
+ - "parl -v" without Module::Signature installed could delete all
+ files within the current directory. Oops.
+
+ - On *nix systems, pp-generated executables erroneously linked to
+ libperl even if "pp -d" (--dependent) is not set.
+
+ - Spurious =cut directives in source files is now handled gracefully
+ by PAR::Filter::PodStrip.
+
+ - "pp -L" (--log) now logs all output messages to the log file,
+ not just the ones printed by "pp" itself.
+
+
+0.79 2004-01-08
+
+ [Bug Fixes]
+
+ - Setting PAR_CLEAN had the reversed effect. Oops.
+
+ - Dynamic libraries in cached directories was not detected
+ properly, resulting in "permission denied" errors during
+ certain race conditions.
+
+0.78 2004-01-07
+
+ [New Features]
+
+ - By default, executables generated by "pp" will now store
+ extracted files in cache directories. You may override
+ this by setting the PAR_CLEAN environment variable to "1",
+ or generate executables using "pp -C".
+
+ - New "pp -C" (--clean) option to make the generated executable
+ clean up temporary directories after each run.
+
+ - PAR_CLEARTEMP is renamed to PAR_CLEAN.
+
+ [Bug Fixes]
+
+ - On Win32, temporary directories containing shared libraries
+ was not being properly cleaned up.
+
+ - If no suitable temporary directories are found, use the current
+ directory (".") instead of the root directory ("/").
+
+0.77 2004-01-01
+
+ [New Features]
+
+ - New "pp -c" and "pp -x" (--compile/--execute) options run the
+ script with "perl -c" to check for dependencies.
+
+ - Also, the new "pp -n" (--noscan) command skips the default
+ static scanning altogether.
+
+ - Added support for "pp -c/-x/-n" to tkpp.
+
+ - For dynamically-built perls, pp-generated .exe files will now
+ appear in the process table with the same name as it was launched,
+ instead of "par.exe".
+
+ - New filter "Obfuscate", which uses B::Deobfuscate to strip away
+ PODs and comments, as well as mangling variable names.
+
+ - Merged tkpp 1.1 from Doug Gruber.
+
+ - OS/2 is now supported.
+
+ - External Zlib is no longer required to run pp-generated binaries.
+
+ [Bug Fixes]
+
+ - Makefile.PL was failing if $Config{cc} contains spaces.
+
+ - No longer needs setting "Windows 95 compatible mode" to run on WinXP.
+
+ - On Win9x with Perl 5.6.1, "nmake" was failing due to extra "@[...]"
+ symbols in Makefile. It should be fixed now.
+
+ - The "bad signature" problem with newer Archive::Zip versions is fixed.
+
+ - App::Packer::Backend::PAR was misplaced into App/Packer/PAR.
+
+ - Signature tests were failing under new ExtUtils::MakeMaker versions.
+
+ - ActiveState's PPM building machine was having problem with PAR;
+ a ".pdb" entry in MANIFEST.SKIP is added to fix that.
+
+ - Some self-built PAR instances on Windows were failing due to
+ mismatching short and long pathnames.
+
+0.76 2003-10-28
+
+ [New Features]
+
+ - Input filters. "pp --filter Bleach" now obfuscates the incoming
+ script with PAR::Filter::Bleach; "pp --modfilter Bleach" applies
+ Bleach to all packed modules.
+
+ - Two previously built-in filters, PodStrip and PatchContent,
+ are refactored out as PAR::Filter subclasses.
+
+ - Two new filters, Bleach and Bytecode, are added for source-hiding
+ purporses.
+
+ - New utility, "tkpp", provides a GUI frontend to "pp".
+
+ - New option, "pp --perlscript", to generate stand-alone scripts.
+
+ - The old "PAR::Intro" documentation has been replaced by two
+ new ones: "PAR::Tutorial" and "PAR::FAQ".
+
+ - Tk pixmap (.xpm) files can now be packed with "pp --add".
+
+ [Bug Fixes]
+
+ - Perl 5.8.1 has an off-by-one bug that prevents "parl" to
+ function properly. We have now provided a workaround;
+ this bug should also be fixed in Perl 5.8.2.
+
+ - Fixed https support for LWP via the new Module::ScapDeps.
+
+0.75 2003-09-21
+
+ [New Features]
+
+ - "pp -o file.exe file.par" now packs file.par into file.exe;
+ this means you can hand-tweak PAR files generated by "pp -p"
+ before packing it into an executable.
+
+ [Bug Fixes]
+
+ - Packing multiple programs by "pp script1.pl script2.pl" was
+ producing syntax errors; fixed.
+
+ - "pp -M datafile" now works.
+
+ - Exit code from pp-packed executables now properly propagates out.
+
+ - Fixed "use base" detection, Math::BigInt support and spurious
+ signature warnings, by updated versions of Module::ScapDeps
+ and Module::Signature.
+
+ - On Win32, the PE info headers no longer show PAR_XXXXXXXXXXX.
+
+0.74 2003-08-20
+
+ [New Features]
+
+ - pp now has a set of "PatchContent" rules, dealing with
+ non-PAR-compatible modules: Tk, Tk::Widget, Win32::API::Type,
+ Win32::SystemInfo, SQL::Parser, diagnostics. These rules may
+ get refactored back to PAR.pm in the future.
+
+ - New function, PAR::reload_libs(), to reload currently used
+ libraries inside PAR files.
+
+ - PAR.pm itself is now never packed into pp-generated files,
+ to perserve interface compatibility and reduce bloat.
+
+ - PAR.pm now handles "use PAR 'othercode.par'" called from
+ program or modules inside PAR files, even recursively.
+
+ - A new icon for Win32 that is hopefully prettier.
+
+ [Bug Fixes]
+
+ - All data after __DATA__ are preserved for included libraries.
+ This helps self-reading modules like Net::LDAP::Constants.
+
+ - PAR::read_file() was broken. It now works.
+
+ - "use PAR" inside pp-generated executables was failing with
+ 'file too short' errors due the mishandling of seek/tell.
+
+ - Occasional crashes on Win32 due to rmdir() called too early
+ with DLLs still open is fixed; however, "pp -d" executables
+ may still exhibit this problem.
+
+ - "pp -X" used to only take full pathnames as arguments.
+ It now also takes "Module::Name" and "Module/Name.pm".
+
+ - Dynamically built Perl under Cygwin failed to build, because
+ libperl.dll.a was not found.
+
+ - Eliminated "callback called on exit" warnings, and the related
+ "access violation" error on Win32.
+
+0.73 2003-08-06
+
+ [New Features]
+
+ - The PAR Homepage is now online at http://par.perl.org/.
+ Documentations have been changed to link to it.
+
+ [Bug Fixes]
+
+ - Tk applications can now properly access xpm/xbm files
+ with Tk->findINC.
+
+ - On Win32, pp-generated executables could not start from
+ Explorer, if its path contains space characters. Fixed.
+
+ - On Win32, pp-generated executables used to leave around an
+ empty directory in $ENV{TEMP}. It is now properly rmdir'ed.
+
+ - Some systems (notably OpenBSD and Debian) does not put their
+ libperl.so in the default location, which breaks the build
+ process; now searches inside $ENV{$Config{ldlibpthname}} and
+ $Config{libpth} to find it.
+
+0.72 2003-08-02
+
+ [New Features]
+
+ - CHECK and INIT blocks in programs inside PAR are now supported.
+
+ [Bug Fixes]
+
+ - Two debug statements were mistakenly left in the source,
+ resulting in "trying to get rid of /tmp/par_priv.xxxx.tmp"
+ messages.
+
+ - Building on Linux with GCC 3.2.2 was failing due to massive
+ heap required for my_perl.c. Fixed by splitting it into 3k
+ chunks.
+
+ - Depends on Module::ScanDeps 0.21; it supports utf8 on Perl 5.6.1
+ and can significantly reduce executable file size by eliminating
+ unneccessary shared libraries.
+
+0.71 2003-07-30
+
+ [Bug Fixes]
+
+ - A nasty data-loss bug has been uncovered immediately after the
+ previous release; it only affects Windows platforms, and may
+ cause all files to be erased under the current root (\) directory.
+
+ - Building on Red Hat linux was failing, with error message that
+ says "my_perl not declared". This has since been fixed.
+
+0.70 2003-07-29
+
+ [New Features]
+
+ - On machines with shared libperl, "pp" now makes truly stand-alone
+ executables; the old behaviour is available with "pp --dependent".
+
+ - Under Windows NT/2000/XP, "pp --icon=name.ico" now changes the icon
+ for the generated executable; otherwise, a default "white camel"
+ icon is used.
+
+ - "use PAR 'http://example.com/foo.par'" now works, as does
+ "perl -MPAR -Ihttp://example.com/foo.par".
+
+ - PAR::Dist is now a mandatory prerequisite, which provides
+ functions to turn any CPAN distribution into a PAR distribution,
+ as well as to install, uninstall, sign and verify such files.
+
+ - Integrated PAR::Dist into "par.pl" and "parl". For example,
+ "parl -i Foo-0.01-i386-freebsd-5.8.0.par" installs a PAR
+ distribution; "parl -v out.exe" verifies a digitally signed
+ executable generated by "pp --sign".
+
+ - A new option, "pp --multiarch", lets you generate PAR files
+ that can work on several architectures.
+
+ - "pp --sign" now adds digital signatures to generated executables
+ and PAR files.
+
+ - PAR files may now (recursively) contain other PAR files inside
+ their par/ directories.
+
+ - shlib/ and par/ directories inside PAR files can now contain
+ architecture- and perl-version-specific subdirectories.
+
+ - The "Cross-Platform Packaging and Deployment with PAR" tutorial
+ is now online as http://www.autrijus.org/par-tutorial/.
+
+ [Bug Fixes]
+
+ - MANIFEST.SKIP was broken on Win32.
+
+ - C compilers that doesn't handle long line well can now
+ compile PAR.
+
+ - DLL files inside the same auto/ library as XS modules
+ was not properly extracted and loaded. This specifically
+ affects Win32.
+
+ - Because parl's @INC is '.', pp-generated executables may
+ miss IO.dll and other shared libraries since they couldn't
+ be correctly found in @INC.
+
+0.69 2003-05-31
+
+ [New Features]
+
+ - Under Perl 5.8, "pp -p" now works with Apache::PAR.
+ See http://aut.dyndns.org/par-tutorial/slide018.html for
+ a simple example.
+
+ - "pp -M filename" now adds "filename" to /, not /lib/,
+ unless filename ends in (pm|ix|al). This makes it
+ possible to bundle "web.conf" needed by Apache::PAR.
+
+ - "pp -l" now searchs in system library paths, and
+ appends "lib" / prepends ".$dl_ext" where necessary.
+
+ [Bug Fixes]
+
+ - PAR segfaults on some Unix platforms due to a NULL pointer
+ used in mktmpdir.c. Fixed.
+
+ - "pp -o out.par -p -e '...'" now honors -o; previously
+ it used "a.out.par" anyway.
+
+ - Inhibited spurious uninitialized warnings under -w in the
+ POD-stripping code.
+
+ - Win32 did not properly cleans up PAR_TEMP directory, resulting
+ in failure for executables that reused the same PID. Fixed.
+
+0.68 2003-05-26
+
+ [New Features]
+
+ - New 'pp -l' option to pack additional shared libraries (DLLs).
+
+ - POD-stripped libraries inside PAR files now have #line directives
+ inserted, so they report the original line numbers on failure.
+
+ - PAR files generated by 'pp' now has a MANIFEST file that can be
+ viewed by Gecko-based browsers with Javascript turned on, e.g.:
+ jar:http://aut.dyndns.org/par/test.par!/MANIFEST
+
+ [Bug Fixes]
+
+ - Each pp-executable instance now creates its own PAR_TEMP directory;
+ this avoids permission errors when multiple users run the same binary.
+ As a consequence, PAR_CLEARTEMP is now set to "1" by default.
+
+ - Newer versions of shared Zlib library no longer causes "pp" to
+ generate broken executables.
+
+ - Fixed dynamic loading on Cygwin was failing due to missing +x mode.
+
+ - Like "use lib", "use PAR 'name.par'" now unshift()s instead of push()es
+ into @INC. Same applies for "par.pl -A" and "parl -A".
+
+ - Fixed building on ActivePerl 626 and below due to a missing $Config{ld}.
+
+0.67 2003-04-01
+
+ [New Features]
+
+ - PAR now works on Cygwin and MinGW/MSYS.
+
+ - Globbing support in PAR::import(): use PAR "/path/*.pm";
+
+ - New license clarification messages added to POD and 'pp -V'.
+
+ - All 'pp' options now has a short form (-o) and a long form (--output).
+
+ - Revamped documentation for 'pp'.
+
+ - New -g (--gui) flag for 'pp' to build console-less Win32 executables.
+
+ [Bug Fixes]
+
+ - Building on Darwin Perl 5.6.0 was broken with 'cc -s'.
+
+ - Building on 5.6.0 was broken due to bad 'base.pm'.
+
+ - Win32 Tk::Widget autoloading was broken due to a binmode() bug.
+
+ - IPC::Run was pod-stripped incorrectly. Fixed.
+
+ - Depends on Module::ScanDeps 0.19, which supports utf8 and .ph files.
+
+ - Better AutoInstall support, which uses 'sudo' where necessary.
+
+0.66 2003-03-20
+
+ [New Features]
+
+ - Adds PAR::Intro, a PODified version of the online presentation.
+
+ - Adds App::Packer::Backend::PAR, a bridge between PAR and App::Packer.
+
+ - Scripts and modules are now searched in "/" last, instead of first.
+
+ - Experimental patch for packing shared libraries via "pp -l".
+
+ - HTTP fetching of precompiled packages in addition to FTP.
+
+ [Bug Fixes]
+
+ - Makefile.PL now downloads precompiled packages only if needed.
+
+ - AutoInstall has been made to work for an easier installation.
+
+ - The redundant "parl.exe.bat" is no longer created on Win32.
+ - Pre-0.63 PARs used to leave broken .dll in TEMP; now they're cleaned.
+
+ - "pp c:\something.pl" no longer treats c: as a relative path.
+
+ - "pp -I dir" now searches 'dir' first, instead of last.
+
+ - "pp" was broken on Perl 5.6.0 due to => stringification bugs.
+
+ - Support for Tk::Widget autoloading has been added.
+
+ - "parl" was not stripped if "gcc" was invoked as "cc"; fixed.
+
+ - On a machine with multiple "parl"s, "pp" now uses the correct one.
+
+ - File::Temp was missing as a dependency.
+
+ [Known Issues]
+
+ - Cygwin support is still broken.
+
+ - PAR does not include utf8_heavy.pl nor unicore/* for scripts that
+ has "use utf8;". This has since been fixed by Module::ScanDeps 0.18.
+
+0.65 2003-03-09
+
+This release comes with several significant improvements:
+
+ [Automatic binary installation]
+
+ Previously, users without a C compiler cannot build the 'parl'
+ executable, and is therefore unable to create self-contained
+ binaries using 'pp'.
+
+ Now, if there is a binary package available for that architecture
+ under my CPAN directory, the Makefile.PL script will automatically
+ fetch it, unpack into blib/, and the installation will continue as
+ normal, resulting in a fully-functional 'pp'.
+
+ This feature is part of the soon-to-be-released Module::Install
+ framework; it will greatly benefit all CPAN authors with non-pure-perl
+ distributions.
+
+ [POD stripping]
+
+ Packages generated with 'pp' will now strip POD sections from all
+ packed dependencies (your own scripts is unaffected); all binary
+ executables will save at least 276732 bytes, with additional ~20%
+ saving in additional packed dependencies.
+
+ You can turn off this feature with the PAR_VERBATIM environment
+ variable.
+
+ [XS Incompatibility solved]
+
+ Because 'pp'-generated executables includes some fixed version of
+ shared libraries (IO, Zlib, etc), they used to break when the
+ target system has different version of shared libraries.
+
+ Now PAR::Heavy intercepts DynaLoader::dl_expandspec to always prefer
+ the library inside the PAR file, so this issue is resolved.
+
+ [5.6.1 Reclaimed]
+
+ Thanks to Sisyphus and various others, building on Perl 5.6.1
+ (with its old ExtUtils::MakeMaker and lack of PTHREAD_ATFORK)
+ now works again.
+
+0.64 2003-03-02
+
+ [New Features]
+
+ - The t/0-signature.t test is re-enabled for people using my
+ Module::Signature to verify the module's OpenPGP signature.
+
+ - This release is the first distribution on CPAN to use the
+ Module::Install framework, which is a stand-alone, extensible
+ drop-in replacement for ExtUtils::MakeMaker that needs no
+ extra action/prerequisites for end users.
+
+ [Bug Fixes]
+
+ - Dynamic loading on Win32 was broken, due to a binmode() bug
+ reported by Bill Atkins, D. Menzel and others.
+
+ - Building on Win32 in directory names that contain spaces
+ did not work.
+
+0.63 2003-02-06
+
+ [Bug Fixes]
+
+ - The 'parl' binary (which replaces the old 'par' or 'par.exe')
+ didn't work properly when bundling perl modules for self-
+ contained executables, rendering 'pp' useless on machines
+ without core perl.
@@ -1,19 +1,13 @@
AUTHORS
-ChangeLog
+Changes
inc/Module/Install.pm
inc/Module/Install/Base.pm
inc/Module/Install/Can.pm
inc/Module/Install/Fetch.pm
-inc/Module/Install/Include.pm
inc/Module/Install/Makefile.pm
inc/Module/Install/Metadata.pm
inc/Module/Install/Win32.pm
inc/Module/Install/WriteAll.pm
-inc/parent.pm
-inc/PerlIO.pm
-inc/Test/Builder.pm
-inc/Test/Builder/Module.pm
-inc/Test/More.pm
lib/PAR.pm
lib/PAR/Environment.pod
lib/PAR/FAQ.pod
@@ -24,7 +18,6 @@ lib/PAR/Tutorial.pod
Makefile.PL
MANIFEST This list of files
MANIFEST.SKIP
-META.json
META.yml
README
t/00-pod.t
@@ -1,46 +0,0 @@
-{
- "abstract" : "Perl Archive Tookit",
- "author" : [
- "Audrey Tang <cpan@audreyt.org>"
- ],
- "dynamic_config" : 0,
- "generated_by" : "ExtUtils::MakeMaker version 6.6302, CPAN::Meta::Converter version 2.120921",
- "license" : [
- "perl_5"
- ],
- "meta-spec" : {
- "url" : "http://search.cpan.org/perldoc?CPAN::Meta::Spec",
- "version" : "2"
- },
- "name" : "PAR",
- "no_index" : {
- "directory" : [
- "t",
- "inc"
- ]
- },
- "prereqs" : {
- "build" : {
- "requires" : {
- "ExtUtils::MakeMaker" : "6.59"
- }
- },
- "configure" : {
- "requires" : {
- "ExtUtils::MakeMaker" : "0"
- }
- },
- "runtime" : {
- "requires" : {
- "Archive::Zip" : "1",
- "AutoLoader" : "5.66_02",
- "Compress::Zlib" : "1.3",
- "File::Temp" : "0.05",
- "PAR::Dist" : "0.32",
- "perl" : "5.008001"
- }
- }
- },
- "release_status" : "stable",
- "version" : "1.007"
-}
@@ -5,9 +5,10 @@ author:
build_requires:
ExtUtils::MakeMaker: 6.59
configure_requires:
- ExtUtils::MakeMaker: 0
-dynamic_config: 0
-generated_by: 'ExtUtils::MakeMaker version 6.6302, CPAN::Meta::Converter version 2.120921'
+ ExtUtils::MakeMaker: 6.59
+distribution_type: module
+dynamic_config: 1
+generated_by: 'Module::Install version 1.14'
license: perl
meta-spec:
url: http://module-build.sourceforge.net/META-spec-v1.4.html
@@ -15,13 +16,30 @@ meta-spec:
name: PAR
no_index:
directory:
- - t
- inc
+ - t
+provides:
+ PAR:
+ file: lib/PAR.pm
+ version: '1.008'
+ PAR::Heavy:
+ file: lib/PAR/Heavy.pm
+ version: '0.12'
+ PAR::SetupProgname:
+ file: lib/PAR/SetupProgname.pm
+ version: '1.002'
+ PAR::SetupTemp:
+ file: lib/PAR/SetupTemp.pm
+ version: '1.002'
+recommends:
+ Digest: Module::Signature
requires:
Archive::Zip: 1
AutoLoader: 5.66_02
Compress::Zlib: 1.3
File::Temp: 0.05
PAR::Dist: 0.32
- perl: 5.008001
-version: 1.007
+ perl: 5.8.1
+resources:
+ license: http://dev.perl.org/licenses/
+version: '1.008'
@@ -29,7 +29,6 @@ if (can_use('Crypt::OpenPGP') or can_run('gpg')) {
'Module::Signature';
}
-include_deps 'Test::More';
no_index directory => 'contrib';
auto_provides;
@@ -4,7 +4,7 @@ package Module::Install::Base;
use strict 'vars';
use vars qw{$VERSION};
BEGIN {
- $VERSION = '1.06';
+ $VERSION = '1.14';
}
# Suspend handler for "redefined" warnings
@@ -8,7 +8,7 @@ use Module::Install::Base ();
use vars qw{$VERSION @ISA $ISCORE};
BEGIN {
- $VERSION = '1.06';
+ $VERSION = '1.14';
@ISA = 'Module::Install::Base';
$ISCORE = 1;
}
@@ -6,7 +6,7 @@ use Module::Install::Base ();
use vars qw{$VERSION @ISA $ISCORE};
BEGIN {
- $VERSION = '1.06';
+ $VERSION = '1.14';
@ISA = 'Module::Install::Base';
$ISCORE = 1;
}
@@ -1,34 +0,0 @@
-#line 1
-package Module::Install::Include;
-
-use strict;
-use Module::Install::Base ();
-
-use vars qw{$VERSION @ISA $ISCORE};
-BEGIN {
- $VERSION = '1.06';
- @ISA = 'Module::Install::Base';
- $ISCORE = 1;
-}
-
-sub include {
- shift()->admin->include(@_);
-}
-
-sub include_deps {
- shift()->admin->include_deps(@_);
-}
-
-sub auto_include {
- shift()->admin->auto_include(@_);
-}
-
-sub auto_include_deps {
- shift()->admin->auto_include_deps(@_);
-}
-
-sub auto_include_dependent_dists {
- shift()->admin->auto_include_dependent_dists(@_);
-}
-
-1;
@@ -8,7 +8,7 @@ use Fcntl qw/:flock :seek/;
use vars qw{$VERSION @ISA $ISCORE};
BEGIN {
- $VERSION = '1.06';
+ $VERSION = '1.14';
@ISA = 'Module::Install::Base';
$ISCORE = 1;
}
@@ -133,7 +133,7 @@ sub makemaker_args {
return $args;
}
-# For mm args that take multiple space-seperated args,
+# For mm args that take multiple space-separated args,
# append an argument to the current list.
sub makemaker_append {
my $self = shift;
@@ -6,7 +6,7 @@ use Module::Install::Base ();
use vars qw{$VERSION @ISA $ISCORE};
BEGIN {
- $VERSION = '1.06';
+ $VERSION = '1.14';
@ISA = 'Module::Install::Base';
$ISCORE = 1;
}
@@ -347,7 +347,7 @@ sub name_from {
^ \s*
package \s*
([\w:]+)
- \s* ;
+ [\s|;]*
/ixms
) {
my ($name, $module_name) = ($1, $1);
@@ -705,7 +705,7 @@ sub _write_mymeta_data {
my @yaml = Parse::CPAN::Meta::LoadFile('META.yml');
my $meta = $yaml[0];
- # Overwrite the non-configure dependency hashs
+ # Overwrite the non-configure dependency hashes
delete $meta->{requires};
delete $meta->{build_requires};
delete $meta->{recommends};
@@ -6,7 +6,7 @@ use Module::Install::Base ();
use vars qw{$VERSION @ISA $ISCORE};
BEGIN {
- $VERSION = '1.06';
+ $VERSION = '1.14';
@ISA = 'Module::Install::Base';
$ISCORE = 1;
}
@@ -6,7 +6,7 @@ use Module::Install::Base ();
use vars qw{$VERSION @ISA $ISCORE};
BEGIN {
- $VERSION = '1.06';
+ $VERSION = '1.14';
@ISA = qw{Module::Install::Base};
$ISCORE = 1;
}
@@ -17,7 +17,7 @@ package Module::Install;
# 3. The ./inc/ version of Module::Install loads
# }
-use 5.005;
+use 5.006;
use strict 'vars';
use Cwd ();
use File::Find ();
@@ -31,7 +31,7 @@ BEGIN {
# This is not enforced yet, but will be some time in the next few
# releases once we can make sure it won't clash with custom
# Module::Install extensions.
- $VERSION = '1.06';
+ $VERSION = '1.14';
# Storage for the pseudo-singleton
$MAIN = undef;
@@ -156,10 +156,10 @@ END_DIE
sub autoload {
my $self = shift;
my $who = $self->_caller;
- my $cwd = Cwd::cwd();
+ my $cwd = Cwd::getcwd();
my $sym = "${who}::AUTOLOAD";
$sym->{$cwd} = sub {
- my $pwd = Cwd::cwd();
+ my $pwd = Cwd::getcwd();
if ( my $code = $sym->{$pwd} ) {
# Delegate back to parent dirs
goto &$code unless $cwd eq $pwd;
@@ -239,7 +239,7 @@ sub new {
# ignore the prefix on extension modules built from top level.
my $base_path = Cwd::abs_path($FindBin::Bin);
- unless ( Cwd::abs_path(Cwd::cwd()) eq $base_path ) {
+ unless ( Cwd::abs_path(Cwd::getcwd()) eq $base_path ) {
delete $args{prefix};
}
return $args{_self} if $args{_self};
@@ -338,7 +338,7 @@ sub find_extensions {
if ( $subpath eq lc($subpath) || $subpath eq uc($subpath) ) {
my $content = Module::Install::_read($subpath . '.pm');
my $in_pod = 0;
- foreach ( split //, $content ) {
+ foreach ( split /\n/, $content ) {
$in_pod = 1 if /^=\w/;
$in_pod = 0 if /^=cut/;
next if ($in_pod || /^=cut/); # skip pod text
@@ -378,6 +378,7 @@ eval( $] >= 5.006 ? <<'END_NEW' : <<'END_OLD' ); die $@ if $@;
sub _read {
local *FH;
open( FH, '<', $_[0] ) or die "open($_[0]): $!";
+ binmode FH;
my $string = do { local $/; <FH> };
close FH or die "close($_[0]): $!";
return $string;
@@ -386,6 +387,7 @@ END_NEW
sub _read {
local *FH;
open( FH, "< $_[0]" ) or die "open($_[0]): $!";
+ binmode FH;
my $string = do { local $/; <FH> };
close FH or die "close($_[0]): $!";
return $string;
@@ -416,6 +418,7 @@ eval( $] >= 5.006 ? <<'END_NEW' : <<'END_OLD' ); die $@ if $@;
sub _write {
local *FH;
open( FH, '>', $_[0] ) or die "open($_[0]): $!";
+ binmode FH;
foreach ( 1 .. $#_ ) {
print FH $_[$_] or die "print($_[0]): $!";
}
@@ -425,6 +428,7 @@ END_NEW
sub _write {
local *FH;
open( FH, "> $_[0]" ) or die "open($_[0]): $!";
+ binmode FH;
foreach ( 1 .. $#_ ) {
print FH $_[$_] or die "print($_[0]): $!";
}
@@ -434,7 +438,7 @@ END_OLD
# _version is for processing module versions (eg, 1.03_05) not
# Perl versions (eg, 5.8.1).
-sub _version ($) {
+sub _version {
my $s = shift || 0;
my $d =()= $s =~ /(\.)/g;
if ( $d >= 2 ) {
@@ -450,12 +454,12 @@ sub _version ($) {
return $l + 0;
}
-sub _cmp ($$) {
+sub _cmp {
_version($_[1]) <=> _version($_[2]);
}
# Cloned from Params::Util::_CLASS
-sub _CLASS ($) {
+sub _CLASS {
(
defined $_[0]
and
@@ -1,33 +0,0 @@
-#line 1
-package PerlIO;
-
-our $VERSION = '1.07';
-
-# Map layer name to package that defines it
-our %alias;
-
-sub import
-{
- my $class = shift;
- while (@_)
- {
- my $layer = shift;
- if (exists $alias{$layer})
- {
- $layer = $alias{$layer}
- }
- else
- {
- $layer = "${class}::$layer";
- }
- eval "require $layer";
- warn $@ if $@;
- }
-}
-
-sub F_UTF8 () { 0x8000 }
-
-1;
-__END__
-
-#line 332
@@ -1,73 +0,0 @@
-#line 1
-package Test::Builder::Module;
-
-use strict;
-
-use Test::Builder;
-
-require Exporter;
-our @ISA = qw(Exporter);
-
-our $VERSION = '0.98';
-$VERSION = eval $VERSION; ## no critic (BuiltinFunctions::ProhibitStringyEval)
-
-
-#line 74
-
-sub import {
- my($class) = shift;
-
- # Don't run all this when loading ourself.
- return 1 if $class eq 'Test::Builder::Module';
-
- my $test = $class->builder;
-
- my $caller = caller;
-
- $test->exported_to($caller);
-
- $class->import_extra( \@_ );
- my(@imports) = $class->_strip_imports( \@_ );
-
- $test->plan(@_);
-
- $class->export_to_level( 1, $class, @imports );
-}
-
-sub _strip_imports {
- my $class = shift;
- my $list = shift;
-
- my @imports = ();
- my @other = ();
- my $idx = 0;
- while( $idx <= $#{$list} ) {
- my $item = $list->[$idx];
-
- if( defined $item and $item eq 'import' ) {
- push @imports, @{ $list->[ $idx + 1 ] };
- $idx++;
- }
- else {
- push @other, $item;
- }
-
- $idx++;
- }
-
- @$list = @other;
-
- return @imports;
-}
-
-#line 137
-
-sub import_extra { }
-
-#line 167
-
-sub builder {
- return Test::Builder->new;
-}
-
-1;
@@ -1,1633 +0,0 @@
-#line 1
-package Test::Builder;
-
-use 5.006;
-use strict;
-use warnings;
-
-our $VERSION = '0.98';
-$VERSION = eval $VERSION; ## no critic (BuiltinFunctions::ProhibitStringyEval)
-
-BEGIN {
- if( $] < 5.008 ) {
- require Test::Builder::IO::Scalar;
- }
-}
-
-
-# Make Test::Builder thread-safe for ithreads.
-BEGIN {
- use Config;
- # Load threads::shared when threads are turned on.
- # 5.8.0's threads are so busted we no longer support them.
- if( $] >= 5.008001 && $Config{useithreads} && $INC{'threads.pm'} ) {
- require threads::shared;
-
- # Hack around YET ANOTHER threads::shared bug. It would
- # occasionally forget the contents of the variable when sharing it.
- # So we first copy the data, then share, then put our copy back.
- *share = sub (\[$@%]) {
- my $type = ref $_[0];
- my $data;
-
- if( $type eq 'HASH' ) {
- %$data = %{ $_[0] };
- }
- elsif( $type eq 'ARRAY' ) {
- @$data = @{ $_[0] };
- }
- elsif( $type eq 'SCALAR' ) {
- $$data = ${ $_[0] };
- }
- else {
- die( "Unknown type: " . $type );
- }
-
- $_[0] = &threads::shared::share( $_[0] );
-
- if( $type eq 'HASH' ) {
- %{ $_[0] } = %$data;
- }
- elsif( $type eq 'ARRAY' ) {
- @{ $_[0] } = @$data;
- }
- elsif( $type eq 'SCALAR' ) {
- ${ $_[0] } = $$data;
- }
- else {
- die( "Unknown type: " . $type );
- }
-
- return $_[0];
- };
- }
- # 5.8.0's threads::shared is busted when threads are off
- # and earlier Perls just don't have that module at all.
- else {
- *share = sub { return $_[0] };
- *lock = sub { 0 };
- }
-}
-
-#line 117
-
-our $Test = Test::Builder->new;
-
-sub new {
- my($class) = shift;
- $Test ||= $class->create;
- return $Test;
-}
-
-#line 139
-
-sub create {
- my $class = shift;
-
- my $self = bless {}, $class;
- $self->reset;
-
- return $self;
-}
-
-#line 168
-
-sub child {
- my( $self, $name ) = @_;
-
- if( $self->{Child_Name} ) {
- $self->croak("You already have a child named ($self->{Child_Name}) running");
- }
-
- my $parent_in_todo = $self->in_todo;
-
- # Clear $TODO for the child.
- my $orig_TODO = $self->find_TODO(undef, 1, undef);
-
- my $child = bless {}, ref $self;
- $child->reset;
-
- # Add to our indentation
- $child->_indent( $self->_indent . ' ' );
-
- $child->{$_} = $self->{$_} foreach qw{Out_FH Todo_FH Fail_FH};
- if ($parent_in_todo) {
- $child->{Fail_FH} = $self->{Todo_FH};
- }
-
- # This will be reset in finalize. We do this here lest one child failure
- # cause all children to fail.
- $child->{Child_Error} = $?;
- $? = 0;
- $child->{Parent} = $self;
- $child->{Parent_TODO} = $orig_TODO;
- $child->{Name} = $name || "Child of " . $self->name;
- $self->{Child_Name} = $child->name;
- return $child;
-}
-
-
-#line 211
-
-sub subtest {
- my $self = shift;
- my($name, $subtests) = @_;
-
- if ('CODE' ne ref $subtests) {
- $self->croak("subtest()'s second argument must be a code ref");
- }
-
- # Turn the child into the parent so anyone who has stored a copy of
- # the Test::Builder singleton will get the child.
- my($error, $child, %parent);
- {
- # child() calls reset() which sets $Level to 1, so we localize
- # $Level first to limit the scope of the reset to the subtest.
- local $Test::Builder::Level = $Test::Builder::Level + 1;
-
- $child = $self->child($name);
- %parent = %$self;
- %$self = %$child;
-
- my $run_the_subtests = sub {
- $subtests->();
- $self->done_testing unless $self->_plan_handled;
- 1;
- };
-
- if( !eval { $run_the_subtests->() } ) {
- $error = $@;
- }
- }
-
- # Restore the parent and the copied child.
- %$child = %$self;
- %$self = %parent;
-
- # Restore the parent's $TODO
- $self->find_TODO(undef, 1, $child->{Parent_TODO});
-
- # Die *after* we restore the parent.
- die $error if $error and !eval { $error->isa('Test::Builder::Exception') };
-
- local $Test::Builder::Level = $Test::Builder::Level + 1;
- return $child->finalize;
-}
-
-#line 281
-
-sub _plan_handled {
- my $self = shift;
- return $self->{Have_Plan} || $self->{No_Plan} || $self->{Skip_All};
-}
-
-
-#line 306
-
-sub finalize {
- my $self = shift;
-
- return unless $self->parent;
- if( $self->{Child_Name} ) {
- $self->croak("Can't call finalize() with child ($self->{Child_Name}) active");
- }
-
- local $? = 0; # don't fail if $subtests happened to set $? nonzero
- $self->_ending;
-
- # XXX This will only be necessary for TAP envelopes (we think)
- #$self->_print( $self->is_passing ? "PASS\n" : "FAIL\n" );
-
- local $Test::Builder::Level = $Test::Builder::Level + 1;
- my $ok = 1;
- $self->parent->{Child_Name} = undef;
- if ( $self->{Skip_All} ) {
- $self->parent->skip($self->{Skip_All});
- }
- elsif ( not @{ $self->{Test_Results} } ) {
- $self->parent->ok( 0, sprintf q[No tests run for subtest "%s"], $self->name );
- }
- else {
- $self->parent->ok( $self->is_passing, $self->name );
- }
- $? = $self->{Child_Error};
- delete $self->{Parent};
-
- return $self->is_passing;
-}
-
-sub _indent {
- my $self = shift;
-
- if( @_ ) {
- $self->{Indent} = shift;
- }
-
- return $self->{Indent};
-}
-
-#line 359
-
-sub parent { shift->{Parent} }
-
-#line 371
-
-sub name { shift->{Name} }
-
-sub DESTROY {
- my $self = shift;
- if ( $self->parent and $$ == $self->{Original_Pid} ) {
- my $name = $self->name;
- $self->diag(<<"FAIL");
-Child ($name) exited without calling finalize()
-FAIL
- $self->parent->{In_Destroy} = 1;
- $self->parent->ok(0, $name);
- }
-}
-
-#line 395
-
-our $Level;
-
-sub reset { ## no critic (Subroutines::ProhibitBuiltinHomonyms)
- my($self) = @_;
-
- # We leave this a global because it has to be localized and localizing
- # hash keys is just asking for pain. Also, it was documented.
- $Level = 1;
-
- $self->{Name} = $0;
- $self->is_passing(1);
- $self->{Ending} = 0;
- $self->{Have_Plan} = 0;
- $self->{No_Plan} = 0;
- $self->{Have_Output_Plan} = 0;
- $self->{Done_Testing} = 0;
-
- $self->{Original_Pid} = $$;
- $self->{Child_Name} = undef;
- $self->{Indent} ||= '';
-
- share( $self->{Curr_Test} );
- $self->{Curr_Test} = 0;
- $self->{Test_Results} = &share( [] );
-
- $self->{Exported_To} = undef;
- $self->{Expected_Tests} = 0;
-
- $self->{Skip_All} = 0;
-
- $self->{Use_Nums} = 1;
-
- $self->{No_Header} = 0;
- $self->{No_Ending} = 0;
-
- $self->{Todo} = undef;
- $self->{Todo_Stack} = [];
- $self->{Start_Todo} = 0;
- $self->{Opened_Testhandles} = 0;
-
- $self->_dup_stdhandles;
-
- return;
-}
-
-#line 474
-
-my %plan_cmds = (
- no_plan => \&no_plan,
- skip_all => \&skip_all,
- tests => \&_plan_tests,
-);
-
-sub plan {
- my( $self, $cmd, $arg ) = @_;
-
- return unless $cmd;
-
- local $Level = $Level + 1;
-
- $self->croak("You tried to plan twice") if $self->{Have_Plan};
-
- if( my $method = $plan_cmds{$cmd} ) {
- local $Level = $Level + 1;
- $self->$method($arg);
- }
- else {
- my @args = grep { defined } ( $cmd, $arg );
- $self->croak("plan() doesn't understand @args");
- }
-
- return 1;
-}
-
-
-sub _plan_tests {
- my($self, $arg) = @_;
-
- if($arg) {
- local $Level = $Level + 1;
- return $self->expected_tests($arg);
- }
- elsif( !defined $arg ) {
- $self->croak("Got an undefined number of tests");
- }
- else {
- $self->croak("You said to run 0 tests");
- }
-
- return;
-}
-
-#line 529
-
-sub expected_tests {
- my $self = shift;
- my($max) = @_;
-
- if(@_) {
- $self->croak("Number of tests must be a positive integer. You gave it '$max'")
- unless $max =~ /^\+?\d+$/;
-
- $self->{Expected_Tests} = $max;
- $self->{Have_Plan} = 1;
-
- $self->_output_plan($max) unless $self->no_header;
- }
- return $self->{Expected_Tests};
-}
-
-#line 553
-
-sub no_plan {
- my($self, $arg) = @_;
-
- $self->carp("no_plan takes no arguments") if $arg;
-
- $self->{No_Plan} = 1;
- $self->{Have_Plan} = 1;
-
- return 1;
-}
-
-#line 586
-
-sub _output_plan {
- my($self, $max, $directive, $reason) = @_;
-
- $self->carp("The plan was already output") if $self->{Have_Output_Plan};
-
- my $plan = "1..$max";
- $plan .= " # $directive" if defined $directive;
- $plan .= " $reason" if defined $reason;
-
- $self->_print("$plan\n");
-
- $self->{Have_Output_Plan} = 1;
-
- return;
-}
-
-
-#line 638
-
-sub done_testing {
- my($self, $num_tests) = @_;
-
- # If done_testing() specified the number of tests, shut off no_plan.
- if( defined $num_tests ) {
- $self->{No_Plan} = 0;
- }
- else {
- $num_tests = $self->current_test;
- }
-
- if( $self->{Done_Testing} ) {
- my($file, $line) = @{$self->{Done_Testing}}[1,2];
- $self->ok(0, "done_testing() was already called at $file line $line");
- return;
- }
-
- $self->{Done_Testing} = [caller];
-
- if( $self->expected_tests && $num_tests != $self->expected_tests ) {
- $self->ok(0, "planned to run @{[ $self->expected_tests ]} ".
- "but done_testing() expects $num_tests");
- }
- else {
- $self->{Expected_Tests} = $num_tests;
- }
-
- $self->_output_plan($num_tests) unless $self->{Have_Output_Plan};
-
- $self->{Have_Plan} = 1;
-
- # The wrong number of tests were run
- $self->is_passing(0) if $self->{Expected_Tests} != $self->{Curr_Test};
-
- # No tests were run
- $self->is_passing(0) if $self->{Curr_Test} == 0;
-
- return 1;
-}
-
-
-#line 689
-
-sub has_plan {
- my $self = shift;
-
- return( $self->{Expected_Tests} ) if $self->{Expected_Tests};
- return('no_plan') if $self->{No_Plan};
- return(undef);
-}
-
-#line 706
-
-sub skip_all {
- my( $self, $reason ) = @_;
-
- $self->{Skip_All} = $self->parent ? $reason : 1;
-
- $self->_output_plan(0, "SKIP", $reason) unless $self->no_header;
- if ( $self->parent ) {
- die bless {} => 'Test::Builder::Exception';
- }
- exit(0);
-}
-
-#line 731
-
-sub exported_to {
- my( $self, $pack ) = @_;
-
- if( defined $pack ) {
- $self->{Exported_To} = $pack;
- }
- return $self->{Exported_To};
-}
-
-#line 761
-
-sub ok {
- my( $self, $test, $name ) = @_;
-
- if ( $self->{Child_Name} and not $self->{In_Destroy} ) {
- $name = 'unnamed test' unless defined $name;
- $self->is_passing(0);
- $self->croak("Cannot run test ($name) with active children");
- }
- # $test might contain an object which we don't want to accidentally
- # store, so we turn it into a boolean.
- $test = $test ? 1 : 0;
-
- lock $self->{Curr_Test};
- $self->{Curr_Test}++;
-
- # In case $name is a string overloaded object, force it to stringify.
- $self->_unoverload_str( \$name );
-
- $self->diag(<<"ERR") if defined $name and $name =~ /^[\d\s]+$/;
- You named your test '$name'. You shouldn't use numbers for your test names.
- Very confusing.
-ERR
-
- # Capture the value of $TODO for the rest of this ok() call
- # so it can more easily be found by other routines.
- my $todo = $self->todo();
- my $in_todo = $self->in_todo;
- local $self->{Todo} = $todo if $in_todo;
-
- $self->_unoverload_str( \$todo );
-
- my $out;
- my $result = &share( {} );
-
- unless($test) {
- $out .= "not ";
- @$result{ 'ok', 'actual_ok' } = ( ( $self->in_todo ? 1 : 0 ), 0 );
- }
- else {
- @$result{ 'ok', 'actual_ok' } = ( 1, $test );
- }
-
- $out .= "ok";
- $out .= " $self->{Curr_Test}" if $self->use_numbers;
-
- if( defined $name ) {
- $name =~ s|#|\\#|g; # # in a name can confuse Test::Harness.
- $out .= " - $name";
- $result->{name} = $name;
- }
- else {
- $result->{name} = '';
- }
-
- if( $self->in_todo ) {
- $out .= " # TODO $todo";
- $result->{reason} = $todo;
- $result->{type} = 'todo';
- }
- else {
- $result->{reason} = '';
- $result->{type} = '';
- }
-
- $self->{Test_Results}[ $self->{Curr_Test} - 1 ] = $result;
- $out .= "\n";
-
- $self->_print($out);
-
- unless($test) {
- my $msg = $self->in_todo ? "Failed (TODO)" : "Failed";
- $self->_print_to_fh( $self->_diag_fh, "\n" ) if $ENV{HARNESS_ACTIVE};
-
- my( undef, $file, $line ) = $self->caller;
- if( defined $name ) {
- $self->diag(qq[ $msg test '$name'\n]);
- $self->diag(qq[ at $file line $line.\n]);
- }
- else {
- $self->diag(qq[ $msg test at $file line $line.\n]);
- }
- }
-
- $self->is_passing(0) unless $test || $self->in_todo;
-
- # Check that we haven't violated the plan
- $self->_check_is_passing_plan();
-
- return $test ? 1 : 0;
-}
-
-
-# Check that we haven't yet violated the plan and set
-# is_passing() accordingly
-sub _check_is_passing_plan {
- my $self = shift;
-
- my $plan = $self->has_plan;
- return unless defined $plan; # no plan yet defined
- return unless $plan !~ /\D/; # no numeric plan
- $self->is_passing(0) if $plan < $self->{Curr_Test};
-}
-
-
-sub _unoverload {
- my $self = shift;
- my $type = shift;
-
- $self->_try(sub { require overload; }, die_on_fail => 1);
-
- foreach my $thing (@_) {
- if( $self->_is_object($$thing) ) {
- if( my $string_meth = overload::Method( $$thing, $type ) ) {
- $$thing = $$thing->$string_meth();
- }
- }
- }
-
- return;
-}
-
-sub _is_object {
- my( $self, $thing ) = @_;
-
- return $self->_try( sub { ref $thing && $thing->isa('UNIVERSAL') } ) ? 1 : 0;
-}
-
-sub _unoverload_str {
- my $self = shift;
-
- return $self->_unoverload( q[""], @_ );
-}
-
-sub _unoverload_num {
- my $self = shift;
-
- $self->_unoverload( '0+', @_ );
-
- for my $val (@_) {
- next unless $self->_is_dualvar($$val);
- $$val = $$val + 0;
- }
-
- return;
-}
-
-# This is a hack to detect a dualvar such as $!
-sub _is_dualvar {
- my( $self, $val ) = @_;
-
- # Objects are not dualvars.
- return 0 if ref $val;
-
- no warnings 'numeric';
- my $numval = $val + 0;
- return $numval != 0 and $numval ne $val ? 1 : 0;
-}
-
-#line 939
-
-sub is_eq {
- my( $self, $got, $expect, $name ) = @_;
- local $Level = $Level + 1;
-
- if( !defined $got || !defined $expect ) {
- # undef only matches undef and nothing else
- my $test = !defined $got && !defined $expect;
-
- $self->ok( $test, $name );
- $self->_is_diag( $got, 'eq', $expect ) unless $test;
- return $test;
- }
-
- return $self->cmp_ok( $got, 'eq', $expect, $name );
-}
-
-sub is_num {
- my( $self, $got, $expect, $name ) = @_;
- local $Level = $Level + 1;
-
- if( !defined $got || !defined $expect ) {
- # undef only matches undef and nothing else
- my $test = !defined $got && !defined $expect;
-
- $self->ok( $test, $name );
- $self->_is_diag( $got, '==', $expect ) unless $test;
- return $test;
- }
-
- return $self->cmp_ok( $got, '==', $expect, $name );
-}
-
-sub _diag_fmt {
- my( $self, $type, $val ) = @_;
-
- if( defined $$val ) {
- if( $type eq 'eq' or $type eq 'ne' ) {
- # quote and force string context
- $$val = "'$$val'";
- }
- else {
- # force numeric context
- $self->_unoverload_num($val);
- }
- }
- else {
- $$val = 'undef';
- }
-
- return;
-}
-
-sub _is_diag {
- my( $self, $got, $type, $expect ) = @_;
-
- $self->_diag_fmt( $type, $_ ) for \$got, \$expect;
-
- local $Level = $Level + 1;
- return $self->diag(<<"DIAGNOSTIC");
- got: $got
- expected: $expect
-DIAGNOSTIC
-
-}
-
-sub _isnt_diag {
- my( $self, $got, $type ) = @_;
-
- $self->_diag_fmt( $type, \$got );
-
- local $Level = $Level + 1;
- return $self->diag(<<"DIAGNOSTIC");
- got: $got
- expected: anything else
-DIAGNOSTIC
-}
-
-#line 1032
-
-sub isnt_eq {
- my( $self, $got, $dont_expect, $name ) = @_;
- local $Level = $Level + 1;
-
- if( !defined $got || !defined $dont_expect ) {
- # undef only matches undef and nothing else
- my $test = defined $got || defined $dont_expect;
-
- $self->ok( $test, $name );
- $self->_isnt_diag( $got, 'ne' ) unless $test;
- return $test;
- }
-
- return $self->cmp_ok( $got, 'ne', $dont_expect, $name );
-}
-
-sub isnt_num {
- my( $self, $got, $dont_expect, $name ) = @_;
- local $Level = $Level + 1;
-
- if( !defined $got || !defined $dont_expect ) {
- # undef only matches undef and nothing else
- my $test = defined $got || defined $dont_expect;
-
- $self->ok( $test, $name );
- $self->_isnt_diag( $got, '!=' ) unless $test;
- return $test;
- }
-
- return $self->cmp_ok( $got, '!=', $dont_expect, $name );
-}
-
-#line 1081
-
-sub like {
- my( $self, $this, $regex, $name ) = @_;
-
- local $Level = $Level + 1;
- return $self->_regex_ok( $this, $regex, '=~', $name );
-}
-
-sub unlike {
- my( $self, $this, $regex, $name ) = @_;
-
- local $Level = $Level + 1;
- return $self->_regex_ok( $this, $regex, '!~', $name );
-}
-
-#line 1105
-
-my %numeric_cmps = map { ( $_, 1 ) } ( "<", "<=", ">", ">=", "==", "!=", "<=>" );
-
-sub cmp_ok {
- my( $self, $got, $type, $expect, $name ) = @_;
-
- my $test;
- my $error;
- {
- ## no critic (BuiltinFunctions::ProhibitStringyEval)
-
- local( $@, $!, $SIG{__DIE__} ); # isolate eval
-
- my($pack, $file, $line) = $self->caller();
-
- # This is so that warnings come out at the caller's level
- $test = eval qq[
-#line $line "(eval in cmp_ok) $file"
-\$got $type \$expect;
-];
- $error = $@;
- }
- local $Level = $Level + 1;
- my $ok = $self->ok( $test, $name );
-
- # Treat overloaded objects as numbers if we're asked to do a
- # numeric comparison.
- my $unoverload
- = $numeric_cmps{$type}
- ? '_unoverload_num'
- : '_unoverload_str';
-
- $self->diag(<<"END") if $error;
-An error occurred while using $type:
-------------------------------------
-$error
-------------------------------------
-END
-
- unless($ok) {
- $self->$unoverload( \$got, \$expect );
-
- if( $type =~ /^(eq|==)$/ ) {
- $self->_is_diag( $got, $type, $expect );
- }
- elsif( $type =~ /^(ne|!=)$/ ) {
- $self->_isnt_diag( $got, $type );
- }
- else {
- $self->_cmp_diag( $got, $type, $expect );
- }
- }
- return $ok;
-}
-
-sub _cmp_diag {
- my( $self, $got, $type, $expect ) = @_;
-
- $got = defined $got ? "'$got'" : 'undef';
- $expect = defined $expect ? "'$expect'" : 'undef';
-
- local $Level = $Level + 1;
- return $self->diag(<<"DIAGNOSTIC");
- $got
- $type
- $expect
-DIAGNOSTIC
-}
-
-sub _caller_context {
- my $self = shift;
-
- my( $pack, $file, $line ) = $self->caller(1);
-
- my $code = '';
- $code .= "#line $line $file\n" if defined $file and defined $line;
-
- return $code;
-}
-
-#line 1205
-
-sub BAIL_OUT {
- my( $self, $reason ) = @_;
-
- $self->{Bailed_Out} = 1;
- $self->_print("Bail out! $reason");
- exit 255;
-}
-
-#line 1218
-
-{
- no warnings 'once';
- *BAILOUT = \&BAIL_OUT;
-}
-
-#line 1232
-
-sub skip {
- my( $self, $why ) = @_;
- $why ||= '';
- $self->_unoverload_str( \$why );
-
- lock( $self->{Curr_Test} );
- $self->{Curr_Test}++;
-
- $self->{Test_Results}[ $self->{Curr_Test} - 1 ] = &share(
- {
- 'ok' => 1,
- actual_ok => 1,
- name => '',
- type => 'skip',
- reason => $why,
- }
- );
-
- my $out = "ok";
- $out .= " $self->{Curr_Test}" if $self->use_numbers;
- $out .= " # skip";
- $out .= " $why" if length $why;
- $out .= "\n";
-
- $self->_print($out);
-
- return 1;
-}
-
-#line 1273
-
-sub todo_skip {
- my( $self, $why ) = @_;
- $why ||= '';
-
- lock( $self->{Curr_Test} );
- $self->{Curr_Test}++;
-
- $self->{Test_Results}[ $self->{Curr_Test} - 1 ] = &share(
- {
- 'ok' => 1,
- actual_ok => 0,
- name => '',
- type => 'todo_skip',
- reason => $why,
- }
- );
-
- my $out = "not ok";
- $out .= " $self->{Curr_Test}" if $self->use_numbers;
- $out .= " # TODO & SKIP $why\n";
-
- $self->_print($out);
-
- return 1;
-}
-
-#line 1353
-
-sub maybe_regex {
- my( $self, $regex ) = @_;
- my $usable_regex = undef;
-
- return $usable_regex unless defined $regex;
-
- my( $re, $opts );
-
- # Check for qr/foo/
- if( _is_qr($regex) ) {
- $usable_regex = $regex;
- }
- # Check for '/foo/' or 'm,foo,'
- elsif(( $re, $opts ) = $regex =~ m{^ /(.*)/ (\w*) $ }sx or
- ( undef, $re, $opts ) = $regex =~ m,^ m([^\w\s]) (.+) \1 (\w*) $,sx
- )
- {
- $usable_regex = length $opts ? "(?$opts)$re" : $re;
- }
-
- return $usable_regex;
-}
-
-sub _is_qr {
- my $regex = shift;
-
- # is_regexp() checks for regexes in a robust manner, say if they're
- # blessed.
- return re::is_regexp($regex) if defined &re::is_regexp;
- return ref $regex eq 'Regexp';
-}
-
-sub _regex_ok {
- my( $self, $this, $regex, $cmp, $name ) = @_;
-
- my $ok = 0;
- my $usable_regex = $self->maybe_regex($regex);
- unless( defined $usable_regex ) {
- local $Level = $Level + 1;
- $ok = $self->ok( 0, $name );
- $self->diag(" '$regex' doesn't look much like a regex to me.");
- return $ok;
- }
-
- {
- ## no critic (BuiltinFunctions::ProhibitStringyEval)
-
- my $test;
- my $context = $self->_caller_context;
-
- local( $@, $!, $SIG{__DIE__} ); # isolate eval
-
- $test = eval $context . q{$test = $this =~ /$usable_regex/ ? 1 : 0};
-
- $test = !$test if $cmp eq '!~';
-
- local $Level = $Level + 1;
- $ok = $self->ok( $test, $name );
- }
-
- unless($ok) {
- $this = defined $this ? "'$this'" : 'undef';
- my $match = $cmp eq '=~' ? "doesn't match" : "matches";
-
- local $Level = $Level + 1;
- $self->diag( sprintf <<'DIAGNOSTIC', $this, $match, $regex );
- %s
- %13s '%s'
-DIAGNOSTIC
-
- }
-
- return $ok;
-}
-
-# I'm not ready to publish this. It doesn't deal with array return
-# values from the code or context.
-
-#line 1449
-
-sub _try {
- my( $self, $code, %opts ) = @_;
-
- my $error;
- my $return;
- {
- local $!; # eval can mess up $!
- local $@; # don't set $@ in the test
- local $SIG{__DIE__}; # don't trip an outside DIE handler.
- $return = eval { $code->() };
- $error = $@;
- }
-
- die $error if $error and $opts{die_on_fail};
-
- return wantarray ? ( $return, $error ) : $return;
-}
-
-#line 1478
-
-sub is_fh {
- my $self = shift;
- my $maybe_fh = shift;
- return 0 unless defined $maybe_fh;
-
- return 1 if ref $maybe_fh eq 'GLOB'; # its a glob ref
- return 1 if ref \$maybe_fh eq 'GLOB'; # its a glob
-
- return eval { $maybe_fh->isa("IO::Handle") } ||
- eval { tied($maybe_fh)->can('TIEHANDLE') };
-}
-
-#line 1521
-
-sub level {
- my( $self, $level ) = @_;
-
- if( defined $level ) {
- $Level = $level;
- }
- return $Level;
-}
-
-#line 1553
-
-sub use_numbers {
- my( $self, $use_nums ) = @_;
-
- if( defined $use_nums ) {
- $self->{Use_Nums} = $use_nums;
- }
- return $self->{Use_Nums};
-}
-
-#line 1586
-
-foreach my $attribute (qw(No_Header No_Ending No_Diag)) {
- my $method = lc $attribute;
-
- my $code = sub {
- my( $self, $no ) = @_;
-
- if( defined $no ) {
- $self->{$attribute} = $no;
- }
- return $self->{$attribute};
- };
-
- no strict 'refs'; ## no critic
- *{ __PACKAGE__ . '::' . $method } = $code;
-}
-
-#line 1639
-
-sub diag {
- my $self = shift;
-
- $self->_print_comment( $self->_diag_fh, @_ );
-}
-
-#line 1654
-
-sub note {
- my $self = shift;
-
- $self->_print_comment( $self->output, @_ );
-}
-
-sub _diag_fh {
- my $self = shift;
-
- local $Level = $Level + 1;
- return $self->in_todo ? $self->todo_output : $self->failure_output;
-}
-
-sub _print_comment {
- my( $self, $fh, @msgs ) = @_;
-
- return if $self->no_diag;
- return unless @msgs;
-
- # Prevent printing headers when compiling (i.e. -c)
- return if $^C;
-
- # Smash args together like print does.
- # Convert undef to 'undef' so its readable.
- my $msg = join '', map { defined($_) ? $_ : 'undef' } @msgs;
-
- # Escape the beginning, _print will take care of the rest.
- $msg =~ s/^/# /;
-
- local $Level = $Level + 1;
- $self->_print_to_fh( $fh, $msg );
-
- return 0;
-}
-
-#line 1704
-
-sub explain {
- my $self = shift;
-
- return map {
- ref $_
- ? do {
- $self->_try(sub { require Data::Dumper }, die_on_fail => 1);
-
- my $dumper = Data::Dumper->new( [$_] );
- $dumper->Indent(1)->Terse(1);
- $dumper->Sortkeys(1) if $dumper->can("Sortkeys");
- $dumper->Dump;
- }
- : $_
- } @_;
-}
-
-#line 1733
-
-sub _print {
- my $self = shift;
- return $self->_print_to_fh( $self->output, @_ );
-}
-
-sub _print_to_fh {
- my( $self, $fh, @msgs ) = @_;
-
- # Prevent printing headers when only compiling. Mostly for when
- # tests are deparsed with B::Deparse
- return if $^C;
-
- my $msg = join '', @msgs;
- my $indent = $self->_indent;
-
- local( $\, $", $, ) = ( undef, ' ', '' );
-
- # Escape each line after the first with a # so we don't
- # confuse Test::Harness.
- $msg =~ s{\n(?!\z)}{\n$indent# }sg;
-
- # Stick a newline on the end if it needs it.
- $msg .= "\n" unless $msg =~ /\n\z/;
-
- return print $fh $indent, $msg;
-}
-
-#line 1793
-
-sub output {
- my( $self, $fh ) = @_;
-
- if( defined $fh ) {
- $self->{Out_FH} = $self->_new_fh($fh);
- }
- return $self->{Out_FH};
-}
-
-sub failure_output {
- my( $self, $fh ) = @_;
-
- if( defined $fh ) {
- $self->{Fail_FH} = $self->_new_fh($fh);
- }
- return $self->{Fail_FH};
-}
-
-sub todo_output {
- my( $self, $fh ) = @_;
-
- if( defined $fh ) {
- $self->{Todo_FH} = $self->_new_fh($fh);
- }
- return $self->{Todo_FH};
-}
-
-sub _new_fh {
- my $self = shift;
- my($file_or_fh) = shift;
-
- my $fh;
- if( $self->is_fh($file_or_fh) ) {
- $fh = $file_or_fh;
- }
- elsif( ref $file_or_fh eq 'SCALAR' ) {
- # Scalar refs as filehandles was added in 5.8.
- if( $] >= 5.008 ) {
- open $fh, ">>", $file_or_fh
- or $self->croak("Can't open scalar ref $file_or_fh: $!");
- }
- # Emulate scalar ref filehandles with a tie.
- else {
- $fh = Test::Builder::IO::Scalar->new($file_or_fh)
- or $self->croak("Can't tie scalar ref $file_or_fh");
- }
- }
- else {
- open $fh, ">", $file_or_fh
- or $self->croak("Can't open test output log $file_or_fh: $!");
- _autoflush($fh);
- }
-
- return $fh;
-}
-
-sub _autoflush {
- my($fh) = shift;
- my $old_fh = select $fh;
- $| = 1;
- select $old_fh;
-
- return;
-}
-
-my( $Testout, $Testerr );
-
-sub _dup_stdhandles {
- my $self = shift;
-
- $self->_open_testhandles;
-
- # Set everything to unbuffered else plain prints to STDOUT will
- # come out in the wrong order from our own prints.
- _autoflush($Testout);
- _autoflush( \*STDOUT );
- _autoflush($Testerr);
- _autoflush( \*STDERR );
-
- $self->reset_outputs;
-
- return;
-}
-
-sub _open_testhandles {
- my $self = shift;
-
- return if $self->{Opened_Testhandles};
-
- # We dup STDOUT and STDERR so people can change them in their
- # test suites while still getting normal test output.
- open( $Testout, ">&STDOUT" ) or die "Can't dup STDOUT: $!";
- open( $Testerr, ">&STDERR" ) or die "Can't dup STDERR: $!";
-
- $self->_copy_io_layers( \*STDOUT, $Testout );
- $self->_copy_io_layers( \*STDERR, $Testerr );
-
- $self->{Opened_Testhandles} = 1;
-
- return;
-}
-
-sub _copy_io_layers {
- my( $self, $src, $dst ) = @_;
-
- $self->_try(
- sub {
- require PerlIO;
- my @src_layers = PerlIO::get_layers($src);
-
- _apply_layers($dst, @src_layers) if @src_layers;
- }
- );
-
- return;
-}
-
-sub _apply_layers {
- my ($fh, @layers) = @_;
- my %seen;
- my @unique = grep { $_ ne 'unix' and !$seen{$_}++ } @layers;
- binmode($fh, join(":", "", "raw", @unique));
-}
-
-
-#line 1926
-
-sub reset_outputs {
- my $self = shift;
-
- $self->output ($Testout);
- $self->failure_output($Testerr);
- $self->todo_output ($Testout);
-
- return;
-}
-
-#line 1952
-
-sub _message_at_caller {
- my $self = shift;
-
- local $Level = $Level + 1;
- my( $pack, $file, $line ) = $self->caller;
- return join( "", @_ ) . " at $file line $line.\n";
-}
-
-sub carp {
- my $self = shift;
- return warn $self->_message_at_caller(@_);
-}
-
-sub croak {
- my $self = shift;
- return die $self->_message_at_caller(@_);
-}
-
-
-#line 1992
-
-sub current_test {
- my( $self, $num ) = @_;
-
- lock( $self->{Curr_Test} );
- if( defined $num ) {
- $self->{Curr_Test} = $num;
-
- # If the test counter is being pushed forward fill in the details.
- my $test_results = $self->{Test_Results};
- if( $num > @$test_results ) {
- my $start = @$test_results ? @$test_results : 0;
- for( $start .. $num - 1 ) {
- $test_results->[$_] = &share(
- {
- 'ok' => 1,
- actual_ok => undef,
- reason => 'incrementing test number',
- type => 'unknown',
- name => undef
- }
- );
- }
- }
- # If backward, wipe history. Its their funeral.
- elsif( $num < @$test_results ) {
- $#{$test_results} = $num - 1;
- }
- }
- return $self->{Curr_Test};
-}
-
-#line 2040
-
-sub is_passing {
- my $self = shift;
-
- if( @_ ) {
- $self->{Is_Passing} = shift;
- }
-
- return $self->{Is_Passing};
-}
-
-
-#line 2062
-
-sub summary {
- my($self) = shift;
-
- return map { $_->{'ok'} } @{ $self->{Test_Results} };
-}
-
-#line 2117
-
-sub details {
- my $self = shift;
- return @{ $self->{Test_Results} };
-}
-
-#line 2146
-
-sub todo {
- my( $self, $pack ) = @_;
-
- return $self->{Todo} if defined $self->{Todo};
-
- local $Level = $Level + 1;
- my $todo = $self->find_TODO($pack);
- return $todo if defined $todo;
-
- return '';
-}
-
-#line 2173
-
-sub find_TODO {
- my( $self, $pack, $set, $new_value ) = @_;
-
- $pack = $pack || $self->caller(1) || $self->exported_to;
- return unless $pack;
-
- no strict 'refs'; ## no critic
- my $old_value = ${ $pack . '::TODO' };
- $set and ${ $pack . '::TODO' } = $new_value;
- return $old_value;
-}
-
-#line 2193
-
-sub in_todo {
- my $self = shift;
-
- local $Level = $Level + 1;
- return( defined $self->{Todo} || $self->find_TODO ) ? 1 : 0;
-}
-
-#line 2243
-
-sub todo_start {
- my $self = shift;
- my $message = @_ ? shift : '';
-
- $self->{Start_Todo}++;
- if( $self->in_todo ) {
- push @{ $self->{Todo_Stack} } => $self->todo;
- }
- $self->{Todo} = $message;
-
- return;
-}
-
-#line 2265
-
-sub todo_end {
- my $self = shift;
-
- if( !$self->{Start_Todo} ) {
- $self->croak('todo_end() called without todo_start()');
- }
-
- $self->{Start_Todo}--;
-
- if( $self->{Start_Todo} && @{ $self->{Todo_Stack} } ) {
- $self->{Todo} = pop @{ $self->{Todo_Stack} };
- }
- else {
- delete $self->{Todo};
- }
-
- return;
-}
-
-#line 2298
-
-sub caller { ## no critic (Subroutines::ProhibitBuiltinHomonyms)
- my( $self, $height ) = @_;
- $height ||= 0;
-
- my $level = $self->level + $height + 1;
- my @caller;
- do {
- @caller = CORE::caller( $level );
- $level--;
- } until @caller;
- return wantarray ? @caller : $caller[0];
-}
-
-#line 2315
-
-#line 2329
-
-#'#
-sub _sanity_check {
- my $self = shift;
-
- $self->_whoa( $self->{Curr_Test} < 0, 'Says here you ran a negative number of tests!' );
- $self->_whoa( $self->{Curr_Test} != @{ $self->{Test_Results} },
- 'Somehow you got a different number of results than tests ran!' );
-
- return;
-}
-
-#line 2350
-
-sub _whoa {
- my( $self, $check, $desc ) = @_;
- if($check) {
- local $Level = $Level + 1;
- $self->croak(<<"WHOA");
-WHOA! $desc
-This should never happen! Please contact the author immediately!
-WHOA
- }
-
- return;
-}
-
-#line 2374
-
-sub _my_exit {
- $? = $_[0]; ## no critic (Variables::RequireLocalizedPunctuationVars)
-
- return 1;
-}
-
-#line 2386
-
-sub _ending {
- my $self = shift;
- return if $self->no_ending;
- return if $self->{Ending}++;
-
- my $real_exit_code = $?;
-
- # Don't bother with an ending if this is a forked copy. Only the parent
- # should do the ending.
- if( $self->{Original_Pid} != $$ ) {
- return;
- }
-
- # Ran tests but never declared a plan or hit done_testing
- if( !$self->{Have_Plan} and $self->{Curr_Test} ) {
- $self->is_passing(0);
- $self->diag("Tests were run but no plan was declared and done_testing() was not seen.");
- }
-
- # Exit if plan() was never called. This is so "require Test::Simple"
- # doesn't puke.
- if( !$self->{Have_Plan} ) {
- return;
- }
-
- # Don't do an ending if we bailed out.
- if( $self->{Bailed_Out} ) {
- $self->is_passing(0);
- return;
- }
- # Figure out if we passed or failed and print helpful messages.
- my $test_results = $self->{Test_Results};
- if(@$test_results) {
- # The plan? We have no plan.
- if( $self->{No_Plan} ) {
- $self->_output_plan($self->{Curr_Test}) unless $self->no_header;
- $self->{Expected_Tests} = $self->{Curr_Test};
- }
-
- # Auto-extended arrays and elements which aren't explicitly
- # filled in with a shared reference will puke under 5.8.0
- # ithreads. So we have to fill them in by hand. :(
- my $empty_result = &share( {} );
- for my $idx ( 0 .. $self->{Expected_Tests} - 1 ) {
- $test_results->[$idx] = $empty_result
- unless defined $test_results->[$idx];
- }
-
- my $num_failed = grep !$_->{'ok'}, @{$test_results}[ 0 .. $self->{Curr_Test} - 1 ];
-
- my $num_extra = $self->{Curr_Test} - $self->{Expected_Tests};
-
- if( $num_extra != 0 ) {
- my $s = $self->{Expected_Tests} == 1 ? '' : 's';
- $self->diag(<<"FAIL");
-Looks like you planned $self->{Expected_Tests} test$s but ran $self->{Curr_Test}.
-FAIL
- $self->is_passing(0);
- }
-
- if($num_failed) {
- my $num_tests = $self->{Curr_Test};
- my $s = $num_failed == 1 ? '' : 's';
-
- my $qualifier = $num_extra == 0 ? '' : ' run';
-
- $self->diag(<<"FAIL");
-Looks like you failed $num_failed test$s of $num_tests$qualifier.
-FAIL
- $self->is_passing(0);
- }
-
- if($real_exit_code) {
- $self->diag(<<"FAIL");
-Looks like your test exited with $real_exit_code just after $self->{Curr_Test}.
-FAIL
- $self->is_passing(0);
- _my_exit($real_exit_code) && return;
- }
-
- my $exit_code;
- if($num_failed) {
- $exit_code = $num_failed <= 254 ? $num_failed : 254;
- }
- elsif( $num_extra != 0 ) {
- $exit_code = 255;
- }
- else {
- $exit_code = 0;
- }
-
- _my_exit($exit_code) && return;
- }
- elsif( $self->{Skip_All} ) {
- _my_exit(0) && return;
- }
- elsif($real_exit_code) {
- $self->diag(<<"FAIL");
-Looks like your test exited with $real_exit_code before it could output anything.
-FAIL
- $self->is_passing(0);
- _my_exit($real_exit_code) && return;
- }
- else {
- $self->diag("No tests run!\n");
- $self->is_passing(0);
- _my_exit(255) && return;
- }
-
- $self->is_passing(0);
- $self->_whoa( 1, "We fell off the end of _ending()" );
-}
-
-END {
- $Test->_ending if defined $Test;
-}
-
-#line 2574
-
-1;
-
@@ -1,762 +0,0 @@
-#line 1
-package Test::More;
-
-use 5.006;
-use strict;
-use warnings;
-
-#---- perlcritic exemptions. ----#
-
-# We use a lot of subroutine prototypes
-## no critic (Subroutines::ProhibitSubroutinePrototypes)
-
-# Can't use Carp because it might cause use_ok() to accidentally succeed
-# even though the module being used forgot to use Carp. Yes, this
-# actually happened.
-sub _carp {
- my( $file, $line ) = ( caller(1) )[ 1, 2 ];
- return warn @_, " at $file line $line\n";
-}
-
-our $VERSION = '0.98';
-$VERSION = eval $VERSION; ## no critic (BuiltinFunctions::ProhibitStringyEval)
-
-use Test::Builder::Module;
-our @ISA = qw(Test::Builder::Module);
-our @EXPORT = qw(ok use_ok require_ok
- is isnt like unlike is_deeply
- cmp_ok
- skip todo todo_skip
- pass fail
- eq_array eq_hash eq_set
- $TODO
- plan
- done_testing
- can_ok isa_ok new_ok
- diag note explain
- subtest
- BAIL_OUT
-);
-
-#line 164
-
-sub plan {
- my $tb = Test::More->builder;
-
- return $tb->plan(@_);
-}
-
-# This implements "use Test::More 'no_diag'" but the behavior is
-# deprecated.
-sub import_extra {
- my $class = shift;
- my $list = shift;
-
- my @other = ();
- my $idx = 0;
- while( $idx <= $#{$list} ) {
- my $item = $list->[$idx];
-
- if( defined $item and $item eq 'no_diag' ) {
- $class->builder->no_diag(1);
- }
- else {
- push @other, $item;
- }
-
- $idx++;
- }
-
- @$list = @other;
-
- return;
-}
-
-#line 217
-
-sub done_testing {
- my $tb = Test::More->builder;
- $tb->done_testing(@_);
-}
-
-#line 289
-
-sub ok ($;$) {
- my( $test, $name ) = @_;
- my $tb = Test::More->builder;
-
- return $tb->ok( $test, $name );
-}
-
-#line 372
-
-sub is ($$;$) {
- my $tb = Test::More->builder;
-
- return $tb->is_eq(@_);
-}
-
-sub isnt ($$;$) {
- my $tb = Test::More->builder;
-
- return $tb->isnt_eq(@_);
-}
-
-*isn't = \&isnt;
-
-#line 416
-
-sub like ($$;$) {
- my $tb = Test::More->builder;
-
- return $tb->like(@_);
-}
-
-#line 431
-
-sub unlike ($$;$) {
- my $tb = Test::More->builder;
-
- return $tb->unlike(@_);
-}
-
-#line 476
-
-sub cmp_ok($$$;$) {
- my $tb = Test::More->builder;
-
- return $tb->cmp_ok(@_);
-}
-
-#line 511
-
-sub can_ok ($@) {
- my( $proto, @methods ) = @_;
- my $class = ref $proto || $proto;
- my $tb = Test::More->builder;
-
- unless($class) {
- my $ok = $tb->ok( 0, "->can(...)" );
- $tb->diag(' can_ok() called with empty class or reference');
- return $ok;
- }
-
- unless(@methods) {
- my $ok = $tb->ok( 0, "$class->can(...)" );
- $tb->diag(' can_ok() called with no methods');
- return $ok;
- }
-
- my @nok = ();
- foreach my $method (@methods) {
- $tb->_try( sub { $proto->can($method) } ) or push @nok, $method;
- }
-
- my $name = (@methods == 1) ? "$class->can('$methods[0]')" :
- "$class->can(...)" ;
-
- my $ok = $tb->ok( !@nok, $name );
-
- $tb->diag( map " $class->can('$_') failed\n", @nok );
-
- return $ok;
-}
-
-#line 577
-
-sub isa_ok ($$;$) {
- my( $object, $class, $obj_name ) = @_;
- my $tb = Test::More->builder;
-
- my $diag;
-
- if( !defined $object ) {
- $obj_name = 'The thing' unless defined $obj_name;
- $diag = "$obj_name isn't defined";
- }
- else {
- my $whatami = ref $object ? 'object' : 'class';
- # We can't use UNIVERSAL::isa because we want to honor isa() overrides
- my( $rslt, $error ) = $tb->_try( sub { $object->isa($class) } );
- if($error) {
- if( $error =~ /^Can't call method "isa" on unblessed reference/ ) {
- # Its an unblessed reference
- $obj_name = 'The reference' unless defined $obj_name;
- if( !UNIVERSAL::isa( $object, $class ) ) {
- my $ref = ref $object;
- $diag = "$obj_name isn't a '$class' it's a '$ref'";
- }
- }
- elsif( $error =~ /Can't call method "isa" without a package/ ) {
- # It's something that can't even be a class
- $obj_name = 'The thing' unless defined $obj_name;
- $diag = "$obj_name isn't a class or reference";
- }
- else {
- die <<WHOA;
-WHOA! I tried to call ->isa on your $whatami and got some weird error.
-Here's the error.
-$error
-WHOA
- }
- }
- else {
- $obj_name = "The $whatami" unless defined $obj_name;
- if( !$rslt ) {
- my $ref = ref $object;
- $diag = "$obj_name isn't a '$class' it's a '$ref'";
- }
- }
- }
-
- my $name = "$obj_name isa $class";
- my $ok;
- if($diag) {
- $ok = $tb->ok( 0, $name );
- $tb->diag(" $diag\n");
- }
- else {
- $ok = $tb->ok( 1, $name );
- }
-
- return $ok;
-}
-
-#line 656
-
-sub new_ok {
- my $tb = Test::More->builder;
- $tb->croak("new_ok() must be given at least a class") unless @_;
-
- my( $class, $args, $object_name ) = @_;
-
- $args ||= [];
- $object_name = "The object" unless defined $object_name;
-
- my $obj;
- my( $success, $error ) = $tb->_try( sub { $obj = $class->new(@$args); 1 } );
- if($success) {
- local $Test::Builder::Level = $Test::Builder::Level + 1;
- isa_ok $obj, $class, $object_name;
- }
- else {
- $tb->ok( 0, "new() died" );
- $tb->diag(" Error was: $error");
- }
-
- return $obj;
-}
-
-#line 741
-
-sub subtest {
- my ($name, $subtests) = @_;
-
- my $tb = Test::More->builder;
- return $tb->subtest(@_);
-}
-
-#line 765
-
-sub pass (;$) {
- my $tb = Test::More->builder;
-
- return $tb->ok( 1, @_ );
-}
-
-sub fail (;$) {
- my $tb = Test::More->builder;
-
- return $tb->ok( 0, @_ );
-}
-
-#line 833
-
-sub use_ok ($;@) {
- my( $module, @imports ) = @_;
- @imports = () unless @imports;
- my $tb = Test::More->builder;
-
- my( $pack, $filename, $line ) = caller;
-
- my $code;
- if( @imports == 1 and $imports[0] =~ /^\d+(?:\.\d+)?$/ ) {
- # probably a version check. Perl needs to see the bare number
- # for it to work with non-Exporter based modules.
- $code = <<USE;
-package $pack;
-use $module $imports[0];
-1;
-USE
- }
- else {
- $code = <<USE;
-package $pack;
-use $module \@{\$args[0]};
-1;
-USE
- }
-
- my( $eval_result, $eval_error ) = _eval( $code, \@imports );
- my $ok = $tb->ok( $eval_result, "use $module;" );
-
- unless($ok) {
- chomp $eval_error;
- $@ =~ s{^BEGIN failed--compilation aborted at .*$}
- {BEGIN failed--compilation aborted at $filename line $line.}m;
- $tb->diag(<<DIAGNOSTIC);
- Tried to use '$module'.
- Error: $eval_error
-DIAGNOSTIC
-
- }
-
- return $ok;
-}
-
-sub _eval {
- my( $code, @args ) = @_;
-
- # Work around oddities surrounding resetting of $@ by immediately
- # storing it.
- my( $sigdie, $eval_result, $eval_error );
- {
- local( $@, $!, $SIG{__DIE__} ); # isolate eval
- $eval_result = eval $code; ## no critic (BuiltinFunctions::ProhibitStringyEval)
- $eval_error = $@;
- $sigdie = $SIG{__DIE__} || undef;
- }
- # make sure that $code got a chance to set $SIG{__DIE__}
- $SIG{__DIE__} = $sigdie if defined $sigdie;
-
- return( $eval_result, $eval_error );
-}
-
-#line 902
-
-sub require_ok ($) {
- my($module) = shift;
- my $tb = Test::More->builder;
-
- my $pack = caller;
-
- # Try to determine if we've been given a module name or file.
- # Module names must be barewords, files not.
- $module = qq['$module'] unless _is_module_name($module);
-
- my $code = <<REQUIRE;
-package $pack;
-require $module;
-1;
-REQUIRE
-
- my( $eval_result, $eval_error ) = _eval($code);
- my $ok = $tb->ok( $eval_result, "require $module;" );
-
- unless($ok) {
- chomp $eval_error;
- $tb->diag(<<DIAGNOSTIC);
- Tried to require '$module'.
- Error: $eval_error
-DIAGNOSTIC
-
- }
-
- return $ok;
-}
-
-sub _is_module_name {
- my $module = shift;
-
- # Module names start with a letter.
- # End with an alphanumeric.
- # The rest is an alphanumeric or ::
- $module =~ s/\b::\b//g;
-
- return $module =~ /^[a-zA-Z]\w*$/ ? 1 : 0;
-}
-
-#line 979
-
-our( @Data_Stack, %Refs_Seen );
-my $DNE = bless [], 'Does::Not::Exist';
-
-sub _dne {
- return ref $_[0] eq ref $DNE;
-}
-
-## no critic (Subroutines::RequireArgUnpacking)
-sub is_deeply {
- my $tb = Test::More->builder;
-
- unless( @_ == 2 or @_ == 3 ) {
- my $msg = <<'WARNING';
-is_deeply() takes two or three args, you gave %d.
-This usually means you passed an array or hash instead
-of a reference to it
-WARNING
- chop $msg; # clip off newline so carp() will put in line/file
-
- _carp sprintf $msg, scalar @_;
-
- return $tb->ok(0);
- }
-
- my( $got, $expected, $name ) = @_;
-
- $tb->_unoverload_str( \$expected, \$got );
-
- my $ok;
- if( !ref $got and !ref $expected ) { # neither is a reference
- $ok = $tb->is_eq( $got, $expected, $name );
- }
- elsif( !ref $got xor !ref $expected ) { # one's a reference, one isn't
- $ok = $tb->ok( 0, $name );
- $tb->diag( _format_stack({ vals => [ $got, $expected ] }) );
- }
- else { # both references
- local @Data_Stack = ();
- if( _deep_check( $got, $expected ) ) {
- $ok = $tb->ok( 1, $name );
- }
- else {
- $ok = $tb->ok( 0, $name );
- $tb->diag( _format_stack(@Data_Stack) );
- }
- }
-
- return $ok;
-}
-
-sub _format_stack {
- my(@Stack) = @_;
-
- my $var = '$FOO';
- my $did_arrow = 0;
- foreach my $entry (@Stack) {
- my $type = $entry->{type} || '';
- my $idx = $entry->{'idx'};
- if( $type eq 'HASH' ) {
- $var .= "->" unless $did_arrow++;
- $var .= "{$idx}";
- }
- elsif( $type eq 'ARRAY' ) {
- $var .= "->" unless $did_arrow++;
- $var .= "[$idx]";
- }
- elsif( $type eq 'REF' ) {
- $var = "\${$var}";
- }
- }
-
- my @vals = @{ $Stack[-1]{vals} }[ 0, 1 ];
- my @vars = ();
- ( $vars[0] = $var ) =~ s/\$FOO/ \$got/;
- ( $vars[1] = $var ) =~ s/\$FOO/\$expected/;
-
- my $out = "Structures begin differing at:\n";
- foreach my $idx ( 0 .. $#vals ) {
- my $val = $vals[$idx];
- $vals[$idx]
- = !defined $val ? 'undef'
- : _dne($val) ? "Does not exist"
- : ref $val ? "$val"
- : "'$val'";
- }
-
- $out .= "$vars[0] = $vals[0]\n";
- $out .= "$vars[1] = $vals[1]\n";
-
- $out =~ s/^/ /msg;
- return $out;
-}
-
-sub _type {
- my $thing = shift;
-
- return '' if !ref $thing;
-
- for my $type (qw(Regexp ARRAY HASH REF SCALAR GLOB CODE)) {
- return $type if UNIVERSAL::isa( $thing, $type );
- }
-
- return '';
-}
-
-#line 1139
-
-sub diag {
- return Test::More->builder->diag(@_);
-}
-
-sub note {
- return Test::More->builder->note(@_);
-}
-
-#line 1165
-
-sub explain {
- return Test::More->builder->explain(@_);
-}
-
-#line 1231
-
-## no critic (Subroutines::RequireFinalReturn)
-sub skip {
- my( $why, $how_many ) = @_;
- my $tb = Test::More->builder;
-
- unless( defined $how_many ) {
- # $how_many can only be avoided when no_plan is in use.
- _carp "skip() needs to know \$how_many tests are in the block"
- unless $tb->has_plan eq 'no_plan';
- $how_many = 1;
- }
-
- if( defined $how_many and $how_many =~ /\D/ ) {
- _carp
- "skip() was passed a non-numeric number of tests. Did you get the arguments backwards?";
- $how_many = 1;
- }
-
- for( 1 .. $how_many ) {
- $tb->skip($why);
- }
-
- no warnings 'exiting';
- last SKIP;
-}
-
-#line 1315
-
-sub todo_skip {
- my( $why, $how_many ) = @_;
- my $tb = Test::More->builder;
-
- unless( defined $how_many ) {
- # $how_many can only be avoided when no_plan is in use.
- _carp "todo_skip() needs to know \$how_many tests are in the block"
- unless $tb->has_plan eq 'no_plan';
- $how_many = 1;
- }
-
- for( 1 .. $how_many ) {
- $tb->todo_skip($why);
- }
-
- no warnings 'exiting';
- last TODO;
-}
-
-#line 1370
-
-sub BAIL_OUT {
- my $reason = shift;
- my $tb = Test::More->builder;
-
- $tb->BAIL_OUT($reason);
-}
-
-#line 1409
-
-#'#
-sub eq_array {
- local @Data_Stack = ();
- _deep_check(@_);
-}
-
-sub _eq_array {
- my( $a1, $a2 ) = @_;
-
- if( grep _type($_) ne 'ARRAY', $a1, $a2 ) {
- warn "eq_array passed a non-array ref";
- return 0;
- }
-
- return 1 if $a1 eq $a2;
-
- my $ok = 1;
- my $max = $#$a1 > $#$a2 ? $#$a1 : $#$a2;
- for( 0 .. $max ) {
- my $e1 = $_ > $#$a1 ? $DNE : $a1->[$_];
- my $e2 = $_ > $#$a2 ? $DNE : $a2->[$_];
-
- next if _equal_nonrefs($e1, $e2);
-
- push @Data_Stack, { type => 'ARRAY', idx => $_, vals => [ $e1, $e2 ] };
- $ok = _deep_check( $e1, $e2 );
- pop @Data_Stack if $ok;
-
- last unless $ok;
- }
-
- return $ok;
-}
-
-sub _equal_nonrefs {
- my( $e1, $e2 ) = @_;
-
- return if ref $e1 or ref $e2;
-
- if ( defined $e1 ) {
- return 1 if defined $e2 and $e1 eq $e2;
- }
- else {
- return 1 if !defined $e2;
- }
-
- return;
-}
-
-sub _deep_check {
- my( $e1, $e2 ) = @_;
- my $tb = Test::More->builder;
-
- my $ok = 0;
-
- # Effectively turn %Refs_Seen into a stack. This avoids picking up
- # the same referenced used twice (such as [\$a, \$a]) to be considered
- # circular.
- local %Refs_Seen = %Refs_Seen;
-
- {
- $tb->_unoverload_str( \$e1, \$e2 );
-
- # Either they're both references or both not.
- my $same_ref = !( !ref $e1 xor !ref $e2 );
- my $not_ref = ( !ref $e1 and !ref $e2 );
-
- if( defined $e1 xor defined $e2 ) {
- $ok = 0;
- }
- elsif( !defined $e1 and !defined $e2 ) {
- # Shortcut if they're both undefined.
- $ok = 1;
- }
- elsif( _dne($e1) xor _dne($e2) ) {
- $ok = 0;
- }
- elsif( $same_ref and( $e1 eq $e2 ) ) {
- $ok = 1;
- }
- elsif($not_ref) {
- push @Data_Stack, { type => '', vals => [ $e1, $e2 ] };
- $ok = 0;
- }
- else {
- if( $Refs_Seen{$e1} ) {
- return $Refs_Seen{$e1} eq $e2;
- }
- else {
- $Refs_Seen{$e1} = "$e2";
- }
-
- my $type = _type($e1);
- $type = 'DIFFERENT' unless _type($e2) eq $type;
-
- if( $type eq 'DIFFERENT' ) {
- push @Data_Stack, { type => $type, vals => [ $e1, $e2 ] };
- $ok = 0;
- }
- elsif( $type eq 'ARRAY' ) {
- $ok = _eq_array( $e1, $e2 );
- }
- elsif( $type eq 'HASH' ) {
- $ok = _eq_hash( $e1, $e2 );
- }
- elsif( $type eq 'REF' ) {
- push @Data_Stack, { type => $type, vals => [ $e1, $e2 ] };
- $ok = _deep_check( $$e1, $$e2 );
- pop @Data_Stack if $ok;
- }
- elsif( $type eq 'SCALAR' ) {
- push @Data_Stack, { type => 'REF', vals => [ $e1, $e2 ] };
- $ok = _deep_check( $$e1, $$e2 );
- pop @Data_Stack if $ok;
- }
- elsif($type) {
- push @Data_Stack, { type => $type, vals => [ $e1, $e2 ] };
- $ok = 0;
- }
- else {
- _whoa( 1, "No type in _deep_check" );
- }
- }
- }
-
- return $ok;
-}
-
-sub _whoa {
- my( $check, $desc ) = @_;
- if($check) {
- die <<"WHOA";
-WHOA! $desc
-This should never happen! Please contact the author immediately!
-WHOA
- }
-}
-
-#line 1556
-
-sub eq_hash {
- local @Data_Stack = ();
- return _deep_check(@_);
-}
-
-sub _eq_hash {
- my( $a1, $a2 ) = @_;
-
- if( grep _type($_) ne 'HASH', $a1, $a2 ) {
- warn "eq_hash passed a non-hash ref";
- return 0;
- }
-
- return 1 if $a1 eq $a2;
-
- my $ok = 1;
- my $bigger = keys %$a1 > keys %$a2 ? $a1 : $a2;
- foreach my $k ( keys %$bigger ) {
- my $e1 = exists $a1->{$k} ? $a1->{$k} : $DNE;
- my $e2 = exists $a2->{$k} ? $a2->{$k} : $DNE;
-
- next if _equal_nonrefs($e1, $e2);
-
- push @Data_Stack, { type => 'HASH', idx => $k, vals => [ $e1, $e2 ] };
- $ok = _deep_check( $e1, $e2 );
- pop @Data_Stack if $ok;
-
- last unless $ok;
- }
-
- return $ok;
-}
-
-#line 1615
-
-sub eq_set {
- my( $a1, $a2 ) = @_;
- return 0 unless @$a1 == @$a2;
-
- no warnings 'uninitialized';
-
- # It really doesn't matter how we sort them, as long as both arrays are
- # sorted with the same algorithm.
- #
- # Ensure that references are not accidentally treated the same as a
- # string containing the reference.
- #
- # Have to inline the sort routine due to a threading/sort bug.
- # See [rt.cpan.org 6782]
- #
- # I don't know how references would be sorted so we just don't sort
- # them. This means eq_set doesn't really work with refs.
- return eq_array(
- [ grep( ref, @$a1 ), sort( grep( !ref, @$a1 ) ) ],
- [ grep( ref, @$a2 ), sort( grep( !ref, @$a2 ) ) ],
- );
-}
-
-#line 1817
-
-1;
@@ -1,37 +0,0 @@
-#line 1
-package parent;
-use strict;
-use vars qw($VERSION);
-$VERSION = '0.225';
-
-sub import {
- my $class = shift;
-
- my $inheritor = caller(0);
-
- if ( @_ and $_[0] eq '-norequire' ) {
- shift @_;
- } else {
- for ( my @filename = @_ ) {
- if ( $_ eq $inheritor ) {
- warn "Class '$inheritor' tried to inherit from itself\n";
- };
-
- s{::|'}{/}g;
- require "$_.pm"; # dies if the file is not found
- }
- }
-
- {
- no strict 'refs';
- push @{"$inheritor\::ISA"}, @_;
- };
-};
-
-"All your base are belong to us"
-
-__END__
-
-=encoding utf8
-
-#line 136
@@ -1,3 +1,5 @@
+=encoding utf8
+
=head1 NAME
PAR::FAQ - Frequently Asked Questions about PAR
@@ -29,7 +31,7 @@ and wait for the script to download and extract PAR and its dependencies. If you
=head3 ppm
-If you are using ActivePerl from ActiveState, you can use the 'ppm' program that comes with the ActiveState Perl. Instructions can be found below. PAR is availlable from various PPM repositories and some packages are compatible with some versions of ActivePerl and not with others. There is an incomplete PAR-Win32 Binary Compatibility List at L<http://par.wikia.com/wiki/PAR_PPM_Compatibility_List> There are at least three relevant PPM repositories: The default ActiveState one, the bribes repository which is used in the example below and Randy Kobes' repository at L<http://theoryx5.uwinnipeg.ca/ppms/>.
+If you are using ActivePerl from ActiveState, you can use the 'ppm' program that comes with the ActiveState Perl. Instructions can be found below. PAR is available from various PPM repositories and some packages are compatible with some versions of ActivePerl and not with others. There is an incomplete PAR-Win32 Binary Compatibility List at L<http://par.wikia.com/wiki/PAR_PPM_Compatibility_List> There are at least three relevant PPM repositories: The default ActiveState one, the bribes repository which is used in the example below and Randy Kobes' repository at L<http://theoryx5.uwinnipeg.ca/ppms/>.
C:\> ppm3
# activestate was out of date compared to this one
@@ -218,7 +220,7 @@ When searching for an answer to this myself, I found many references to using "e
=head2 How can I change the icon of the generated .exe file under Windows?
-There is another not-completely-documented switch that only works on windows, –icon MyIcon.ico. So just use this:
+There is another not-completely-documented switch that only works on windows, --icon MyIcon.ico. So just use this:
% pp --icon "c:\path to\MyIcon.ico" -o file.exe file.pl.
@@ -226,7 +228,7 @@ There is another not-completely-documented switch that only works on windows,
=head2 The command line parameters (@ARGV) of a pp-ed binary called from another pp-ed binary are missing or broken. What the...?
-This was a bug in releases up to and incuding PAR-0.90. Please upgrade to PAR 0.91 or later and the problem will go away.
+This was a bug in releases up to and including PAR-0.90. Please upgrade to PAR 0.91 or later and the problem will go away.
=head2 I want to include a pp-ed binary in an RPM package. How can I make this work?
@@ -256,7 +258,7 @@ After quite some time, it seems like the smart folks at Redhat found the culprit
this won't work.
Much praise to all who helped track this down! The discussion can be found in the following RT tickets:
-L<http://rt.cpan.org/Public/Bug/Display.html?id=18536 #18536> and L<http://rt.cpan.org/Public/Bug/Display.html?id=19609 #19609>.
+L<http://rt.cpan.org/Public/Bug/Display.html?id=18536|#18536> and L<http://rt.cpan.org/Public/Bug/Display.html?id=19609|#19609>.
-- Steffen Mueller, 22 July 2006
@@ -76,7 +76,7 @@ sub _bootstrap {
}
my $member;
- # First, try to find things in the peferentially loaded PARs:
+ # First, try to find things in the preferentially loaded PARs:
$member = PAR::_find_par_internals([@PAR::PAR_INC], undef, $file, 1)
if defined &PAR::_find_par_internals;
@@ -158,7 +158,7 @@ sub _dl_extract {
binmode($fh);
$member->extractToFileHandle($fh);
close $fh;
- chmod 0755, $filename;
+ chmod 0750, $filename;
}
return $filename;
@@ -1,5 +1,5 @@
package PAR;
-$PAR::VERSION = '1.007';
+$PAR::VERSION = '1.008';
use 5.006;
use strict;
@@ -722,14 +722,9 @@ sub _extract_inc {
for ( $zip->memberNames() ) {
s{^/}{};
-
- # Skip DLLs (these will be handled by the dynaloader hook)
- # except for those placed in File::ShareDir directories.
- next if (m{\.\Q$dlext\E[^/]*$} && !m{^lib/auto/share/(dist|module)/});
-
my $outfile = File::Spec->catfile($inc, $_);
next if -e $outfile and not -w _;
- $zip->extractMember($_, "$inc/" . $_);
+ $zip->extractMember($_, $outfile);
}
}
@@ -793,7 +788,7 @@ sub find_par {
return $rv if defined $rv or not @PriorityRepositoryObjects;
- # the repositories that are prefered over locally installed modules
+ # the repositories that are preferred over locally installed modules
my $module = $args[1];
$module =~ s/\.pm$//;
$module =~ s/\//::/g;