p4.pm #2

# Change User Description Committed
#2 28901 marc_tooley ""
#1 5835 marc_tooley VCP branch to prep for incoming bugfixes for the Source/svn.pm module.
//guest/perforce_software/revml/lib/VCP/Utils/p4.pm
#67 5085 Barrie Slaymaker - Prevent $_ from getting clobbered, esp.
in sdbm.pm
#66 4523 Barrie Slaymaker - Don't autodetect P4::Client by default, require setting VCPP4CLIENT=1
#65 4478 Barrie Slaymaker - "client not changed" message ignored (thanks clkao)
#64 4470 Barrie Slaymaker - VCP::Utils::p4 is more test friendly
#63 4414 Barrie Slaymaker - Indentation corrected
#62 4067 Barrie Slaymaker - VCP::Source::p4 now tells the user what's taking so long while
  it's requesting initial data sets from the repository
- VCP::Source::p4 no longer loads the branch specs from the source,
  they're no longer used.
#61 4021 Barrie Slaymaker - Remove all phashes and all base & fields pragmas
- Work around SWASHGET error
#60 4019 Barrie Slaymaker - typo fixed
#59 4006 Barrie Slaymaker - VCP will now use the P4::Client module if it's installed.
#58 3973 Barrie Slaymaker - p4d launching no longer require IPC::Run on Unix
#57 3970 Barrie Slaymaker - VCP::Source handles rev queing, uses disk to reduce RAM
- Lots of other fixes
#56 3900 Barrie Slaymaker - p4-like specs (such as VCP's config files) may now contain
  underscores in their field names ("csv_trace:").
#55 3893 Barrie Slaymaker - p4d is now launched in the proper directory
#54 3833 Barrie Slaymaker - Minor debugging improvements
#53 3799 Barrie Slaymaker - Empty strings values emitted by p4 set are parsed
#52 3736 Barrie Slaymaker - Source and dest specs are now quoted in .vcp files as needed
- The server and client paths are now quoted in the p4 client
  vcp creates
#51 3649 Barrie Slaymaker - Stray DB::single = 1s removed
#50 3572 John Fetkovich added y/n question to accept default of user_id
#49 3570 John Fetkovich change p4_get_settings (wrapper around 'p4 set' command)
to remove parenthesized things at the end of a line
which indicate where a value came from (such as a config file)
#48 3569 Barrie Slaymaker - Work around bug caused by p4's using the long pathname when
  $ENV{PWD} is not set.
#47 3508 John Fetkovich added p4_get_settings sub to read P4...
env vars/config settings
#46 3480 John Fetkovich add friendly error message if p4_spec has no leading '/'.
#45 3432 Barrie Slaymaker - "p4 print" to "p4 sync"
#44 3422 Barrie Slaymaker - Factor some common code in to VCP::Utils::p4.
- Add error when a p4 filespec ends in a '/' or '\'
- Clean up revml2p4 test scripts
#43 3402 Barrie Slaymaker - now passes all tests using the p4 api library.
  (still not default, set env var VCPP4API=1)
- foo->p4 handles branch-but-no-change case when
  --change-branch-rev-1 is passed.
- sources & dests can now provide their own command
  execution routine in place of shelling out to an
  external command (as in call the p4api library
  instead of running the p4 command).
#42 3382 John Fetkovich Moved setting of repo_id to 'sub init'
       moved defaulting of repo_server to P4PORT env var to 'sub init'
#41 3377 John Fetkovich bug fixes for weird input
#40 3284 John Fetkovich 'sub new' constructor in Source and Dest p4.pm fixed so parse_p4_repo_spec
       only called when a $spec is provided to the constructor.  parse_p4_repo_spec
       now also sets the repo_id.  parse_repo_spec (TODO item) no longer returns
       a hash value of the values parsed, it only sets fields in $self.  Fixed a few
       places where that return hash was used.
#39 3173 Barrie Slaymaker Add partial support for P4 API to test with.
#38 3167 Barrie Slaymaker Add profiling report that details various chunks of time
       taken.
#37 3155 Barrie Slaymaker Convert to logging using VCP::Logger to reduce stdout/err spew.
       Simplify & speed up debugging quite a bit.
       Provide more verbose information in logs.
       Print to STDERR progress reports to keep users from wondering
       what's going on.
       Breaks test; halfway through upgrading run3() to an inline
       function for speed and for VCP specific features.
#36 3098 Barrie Slaymaker Convert all length p4 command line calls to us p4 -x -.
       All hail p4 -x -.
#35 3079 Barrie Slaymaker Remove old commented-out code
#34 3014 Barrie Slaymaker minor warnings cleanups
#33 2967 Barrie Slaymaker Fix p4d launch retry and port selection logic
#32 2953 John Fetkovich p4d was still not attempting a re-launch if a launch failed
       due to the TCP port apparently being taken.  Probably fixed now.
#31 2933 John Fetkovich Added calls to empty()
#30 2932 John Fetkovich Fixed launch_p4d so if it fails to find an open port,
       it will retry.
#29 2862 Barrie Slaymaker Swallow p4d output to stdout so it won't interfere with emitting
       revml (or any other output) to stdout.
#28 2719 Barrie Slaymaker Improve error reporting, allow run_safely's in_dir to be
       relative to the default command_chdir setting
#27 2704 Barrie Slaymaker export underscorify_name
#26 2692 Barrie Slaymaker Centralize escaping of p4 label (etc) names.
NOTE: I have not been able to find a description of the set of legal p4
names (namelength, character set, etc).  This is purely a first attempt,
if you have details on this, please let me know.
#25 2689 Barrie Slaymaker Prevent kill_all_vcp_p4ds() from swallowing $?
#24 2664 Barrie Slaymaker Try IPC::Run3 to see if it speeds up p4 calls
#23 2641 Barrie Slaymaker Add --run-p4d option to VCP::{Source,Dest}::p4.
       Implement port hunting and p4d up & ready detection for
       vcp-launched p4ds.
#22 2639 John Fetkovich Added profiling to be made active when VCPPROFILE environment
       variable turned on.  writes profile info to filename defined
       in VCPPROFILE.  Put some profiling statements (activated at
       compile time) in vcp and p4.pm.
#21 2589 John Fetkovich Split 90p4.t into 90revml2p4_0.t, 90revml2p4_1.t, 91p42revml.t, 95p42cvs.t.
       Added some utilities to the library files listed.
#20 2350 Barrie Slaymaker Get p4d to be killed after the last p4 cleanup commands,
       generalize the p4d launching & destruction system because
       we'll need it in VCP::Source::p4 too to take checkpoints.
#19 2318 Barrie Slaymaker enable end of line comments
#18 2303 Barrie Slaymaker Allow field values to be one letter in p4 forms
#17 2293 Barrie Slaymaker Update CHANGES, TODO, improve .vcp files, add --init-cvs
#16 2283 Barrie Slaymaker fix error message
#15 2282 Barrie Slaymaker Add in support for *.vcp files
#14 2278 Barrie Slaymaker Reduce use()ed modules quite a bit.
 old copy-n-paste
       leftovers, don't want to slow down VCP.pm which now
       needs the p4 forms i/o.
#13 2051 Barrie Slaymaker Enable p4_branch_spec to be carried through revml->revml.
#12 2042 Barrie Slaymaker Basic source::p4 branching support
#11 1855 Barrie Slaymaker Major VSS checkin.
 Works on Win32
#10 1852 Barrie Slaymaker force delete clients, so that later test runs won't trip over them
#9 1851 Barrie Slaymaker No need to die() in an END block.
#8 1841 Barrie Slaymaker Remove .exe if Win32 hack now that IPC::Run is better.
#7 1816 Barrie Slaymaker hack around the p4 client -o Options: hibit corruption problem
#6 1809 Barrie Slaymaker VCP::Patch should ignore lineends
#5 1358 Barrie Slaymaker Win32 changes
#4 703 Barrie Slaymaker VCP::Source::p4 now uses VCP::Utils::p4::parse_p4_repo_spec()
#3 702 Barrie Slaymaker Doc tweak
#2 694 Barrie Slaymaker cleanup up warning about undef var in VCP::Utils::p4.
#1 692 Barrie Slaymaker Add VCP::Utils::p4 and use it to get VCP::Dest::p4 to create it's
own client view as needed.