[libnumber-phone-perl] annotated tag upstream/3.0003 created (now 9ca3239)

gregor herrmann gregoa at debian.org
Mon Sep 8 20:31:10 UTC 2014


This is an automated email from the git hooks/post-receive script.

gregoa pushed a change to annotated tag upstream/3.0003
in repository libnumber-phone-perl.

        at  9ca3239   (tag)
   tagging  8ecb5ac2932abe5df6a0d4845d66721c49339dcc (commit)
 tagged by  gregor herrmann
        on  Mon Sep 8 22:18:43 2014 +0200

- Log -----------------------------------------------------------------
Upstream version 3.0003

David Cantrell (178):
      Tidying up after import from CVS
      Fix Number::Phone::UK constructor bug - was returning 0 instead of undef for invalid numbers
      UK updates for Number::Phone. 1.7002 release
      Bugfix, see RT 48581
      Number::Phone 1.7003 release
      1.7004 release - UK geographics are not necessarily fixed lines
      Bugfix for Number::Phone->new('CC', '123456789'), bug reported by mstevens
      Add 5+4 format, eg 016977 2222
      Fix test for dodgy 4+4 format (OFCOM fixed their data)
      No longer need to be able to parse 4+4
      Bugfix for N::P->new('CC', '+NN12345')
      location of OFCOM data files has changed
      minimal test for RT61177
      notes
      Make unknown countries return something more useful than undef
      Doco, fix other tests what we broke with this incompatible change
      Add more incompatibility warnings in doco
      Test for 074 numbers
      Updated UK data
      Tidied up Mayotte; ETNS no longer exists; Ellipso is dead
      Ellipso is *really* dead
      no more dialling GI as part of ES
      Fix reverse for YT
      Rationalise +7
      DO now has 849
      Update CA
      Kosovo, US, Western Sahara
      Netherlands Antilles; UK data
      N::P::UK parser made a bit more robust (see RT 65557).
      N::P->new comments
      N::P::UK parser (and data builder) fixed for 01768 88 mixed length stupidity
      remove pointless 'use UNIVERSAL qw(isa)'
      Merge pull request #1 from dolmen/fix/subclassing
      Correct doco
      Bump version, update data
      Add tests to distro
      Add South Sudan (SS, +211)
      New UK allocations
      See also libphonenumber on code.google
      move the new SEE ALSO to the right place. Oops.
      Fix broken formats in OFCOM's data
      refactoring; de-tabbification
      Lazily populate is_* in cache for UK numbers
      Update to modern licencing, preparations for using libphonenumber's data
      Have Makefile.PL auto-build stuff as necessary;
      MANIFEST updates; notes in CHANGELOG
      Pull in area code names for the NANP from libphonenumber
      Acknowledgement for libphonenumber in README
      Details on how libphonenumber's data are used
      Kill obsolete comment
      Excise mathoms; note about IOM mobiles
      Added stuff to README at Ian Galpin's suggestion
      Fix nasty bug in formatting unallocated numbers instantiated as N::P::UK->new('0...')
      CHANGELOG
      Re-write t/01_uk_data.t to use Test::More
      Test::More-ify t/00_superclass_and_uk_different_method_invocations.t
      betterify build-data.sh to only rebuild UK data if necessary
      Sint Maarten moved from +599 5 to +1 721
      SX doesn't have AS's numbering plan!
      Add lib/Number/Phone/NANP/SX.pm to MANIFEST
      Improve UK number formatting
      Isle of Man (IM) now kinda-sorta treated as a country
      Fix cut n paste error
      kill spurious comment
      Merge pull request #5 from g1smd/master
      fix IM namespace, related tests, UK data builder, country code calculator
      Add StubCountry::* based on libphonenumber data
      more Weird Little Islands
      Tests
      Split tests out from t/01_uk_data so they can be shared with something that tests the auto-generated stubs
      Start fiddling with tests for stubbed UK
      numerous test fixes for things libphonenumber lacks. still broken though.
      test for russian mobiles (bug reported by bzero on irc)
      almost all tests pass; added areaname to stubs
      correct regex checks in StubCountry.pm
      re-enable the hard cases :-(
      voip and uan numbers in stubs
      libphonenumber changed its formatting rule for UK mobiles
      deprecation warnings for stupid ways of calling methods
      correct deprecation messages; suppress them in tests
      auto-rebuild in 'make dist'
      move libphonenumber's UK data into t/lib
      doco for deprecations, other improvements
      README updates
      add a .gitignore
      bump versions
      check KZ, not MOCKing
      release date
      bugfix: excise extraneous \ from stubs' area names
      add repository to META.yml
      trivial doco fix
      confirmed 00 and 0 prefixes for South Sudan
      betterify doco; add operator_ported method; add tests and doco for dial_to method
      implementation of dial_to
      mark dial_to() as being experimental
      make warnings fatal in tests
      re-factor dial_to tests
      Merge branch 'master' of git at github.com:DrHyde/perl-modules-Number-Phone
      more tests for oddities in UK scheme
      just a version bump to get new data to the CPAN
      Merge branch '2.0-update-1'
      fix for https://github.com/DrHyde/perl-modules-Number-Phone/issues/7
      prepare for 2.0002 release
      merge of 7-fix-stub-building
      point bug reporters to github, not rt.cpan
      don't die trying to parse Inmarsat etc numbers, build a REALLY dumb stub instead
      prepare for 2.1 release
      merge from 8-no-stub-for-inmarsat
      fix POD bug
      remove embedded copy of DBM::Deep 0.9something, use latest CPAN version
      update CHANGELOG prior to merge into master
      Merge branch 'remove-embedded-dbm-deep'
      link to problem in CHANGELOG; de-tabbify
      UK intra-country dialling as 0xxxx xxxxxx always works, so use that; test UK <-> offshore dependencies
      clarify/correct dial_to doco
      bump version numbers
      update changelog
      Change to use curl instead of wget, because Fink is broken
      Test to make sure we don't load broken third-party Number::Phone::CC modules
      make that test fail even if Number::Phone::JP isn't installed
      detect broken N::P::CC, use stub instead
      bump version; make curl preserve timestamp when downloading
      bump version; add notes re MA/EH and RU/KZ
      stupid telcos changing their names and breaking my tests
      add tests for too-short unallocated numbers
      add slightly more rigourous length checks for 02, 07 and 011 numbers; also cache invalidity better
      bump versions in preparation for release
      noise from re-building data files
      doco updates
      don't duplicate all the area names into all the NANP stubs, instead use what we've already got in N::P::NANP::Data
      Merge pull request #16 from theory/libphonenumber-us
      add --force option to build-data.sh; check both geocoding and metadata files when deciding whether NANP is up to date; metadata file got renamed upstream (only matters for devs on case-sensitive filesystems)
      be bug-compatible with libphonenumber
      add note about recent changes to CHANGELOG
      keep up with random changes in the META.yml spec. Yay stability
      fix tyop
      giganto test-refactoring, to make sure that when we add/edit tests in the future they get used in all the right places
      no longer any special-case the GB stub - we can use the AWESOME POWAH of subclassing
      huh, that phone number isn't supposed to exist
      use libphonenumber data for matching NPAs to countries instead of hard-coding NANP data
      half these numbers are apparently not valid anyway
      libphonenumber doesn't recognise 555 COs
      update CHANGELOG
      +883 is generic international networks, not Inmarsat
      test that different bits of +883 get recognised correctly
      point bug reporters at github issues
      libphonenumber is too strict, disallowing eg 555, so just extract fixed/mobile data, hardcode country data
      update lists of CA/US area codes
      tidying
      bah, we can't use this after all - libphonenumber's regexs for fixed-line/mobile both match some US/CA area codes
      moan about bad data
      bump version
      remove mathoms
      expose the list of NANP area codes for each country
      teach N::P::NANP::* about is_mobile etc
      make the way stubs figure out fixed_line/geographic a lot less dumb
      clearer explanation
      building stubs now depends on NANP data
      bump last-updated date
      move shell nonsense into perl script, in preparation for turning build-data.sh into a Makefile
      bump/fix versions
      may need to re-run Makefile.PL after building data files; get rid of warning that was just for debuggery
      prepare for next release
      prepare for release
      add test for N::P::UK->new('0843...')->format()
      format numbers more consistently
      can't instantiate a ::UK object when mocking
      update change log
      some data changes affect the tests
      need to open more fancy-like because of newlines within fields in OFCOM data; opportunity to refactor a little bit
      tidying
      more data changes broke the tests
      update changelog
      Get rid of UK database optimisation step. It's unnecesary.
      bump versions, complete changelog
      this was published in June, not July
      bugfix for N::P->new('CC', '0...') when using stubs, where 0 represents national dialing prefix
      bump version for release

David E. Wheeler (12):
      Build all country stubs from libphonenumber.
      Update UK tests for new data.
      Extract country/number sorting into its own method.
      Add Number::Phone::Lib.
      Add another UK phone number.
      Update MANIFEST.
      Format numbers without leading digits pattern.
      Update MANIFEST.
      Detect libphonenumber territories that inherit US formatting.
      Add failing toll-free test for #17.
      Ha! Fix find-and-replace-o.
      Make GG, IM, and JE inherit UK formats.

DrHyde (1):
      OFCOM publish no format/length data for +44 843. This makes the format() function less stupid when that happens.

Michael Prokop (1):
      Import original source of Number-Phone 3.0001

Olivier Mengué (2):
      A testcase for ->country in subclasses
      Fix ->country for subclasses of a country base class

Sawyer X (1):
      merge by DrHyde of https://github.com/xsawyerx/perl-modules-Number-Phone/commit/137bf1a53

drhyde (23):
      Initial revision
      Now contains an old version of DBM::Deep, cos 1.0 broke everything
      Meh, broken version number in previous commit
      Updated docs to mumble about embedded DBM::Deep
      Mayotte and Komoros have split
      *** empty log message ***
      Removed bogus META.yml files
      Mayotte/Reunion/Comoros updates
      Removed docs for N::P::UK::DBM::Deep and replaced with a note
      *** empty log message ***
      Loads of tidying/rationalisation of tests
      Hlagh
      *** empty log message ***
      Minor fiddling in build-data.realwork
      Updated UK data;
      Hlagh
      New UK allocations;
      TODOs
      subclasses no longer need to register themselves;
      add dependency on Test::More
      Fix formatting bug in UK; new UK allocations; fix test that relied on a telco that changed its name
      Added country-to-code and national/international prefix data to N::P::C, thanks to Michael Schout
      Kill warnings about GE being deprecated;

g1smd (2):
      Change format list order.
      URL fixes in README.

gregor herrmann (1):
      Imported Upstream version 3.0003

-----------------------------------------------------------------------

No new revisions were added by this update.

-- 
Alioth's /usr/local/bin/git-commit-notice on /srv/git.debian.org/git/pkg-perl/packages/libnumber-phone-perl.git



More information about the Pkg-perl-cvs-commits mailing list