------------------------------------------------------------------- Sun Aug 10 08:50:39 UTC 2014 - lazy.kent@opensuse.org - Update to 0.15. * Summary of tag changes: + Added: - c-format-string-argument-type-mismatch - c-format-string-error - c-format-string-excess-arguments - c-format-string-non-portable-conversion - c-format-string-redundant-flag - partially-translated-message * Fix calculation of current time. * Check for errors in C format strings. * Check for inconsistent leading/trailing newlines in * translations even when encoding in broken. * Check for conflict markers in translations even when encoding is broken. * Check for incomplete translations in messages with plural forms. * When emitting unusual-character-in-translation, don't output the translated string (which is likely to contain mojibake), but only the message identifier. * Suggest using “-0000” (rather than “+0000”) to indicate that the offset to local time is unknown. This convention is inspired by RFC 3339 §4.3 and RFC 5322 §3.3. * Improve the test suite. ------------------------------------------------------------------- Mon Jul 21 11:58:41 UTC 2014 - lazy.kent@opensuse.org - Update to 0.14. * Summary of tag changes: + Added: - translation-in-template * Check for PO template files containing translated messages. * Check for duplicate messages, for problems with message flags, and for empty files even when encoding is broken or unknown. * Check for inconsistent leading/trailing newlines between msgid and msgid_plural even when encoding in unknown or broken. * Improve the test suite. ------------------------------------------------------------------- Thu Jul 10 13:47:08 UTC 2014 - lazy.kent@opensuse.org - Update to 0.3.15. * Use HTTPS URLs when they are available, in documentation and code. * Don't complain about conflict markers in fuzzy messages. ------------------------------------------------------------------- Thu Mar 20 16:57:19 UTC 2014 - lazy.kent@opensuse.org - Update to 0.13.4. * Regenerate the timezone information file using tzdata 2014a. * Recognize “8859-n” (without the “ISO-” prefix) as encoding names. * Recognize *.local, *.in-addr.arpa and *.ip6.arpa as special domain names. * Fix compatibility with polib 1.0.4. * Improve the test suite. - Update keyring. ------------------------------------------------------------------- Fri Jan 24 13:45:28 UTC 2014 - lazy.kent@opensuse.org - Update to 0.13.3. * Fix incorrect plural expression co-domain evaluation. * Fix incorrect precedence of the boolean not operator in plural expressions. * Forbid unary plus and unary minus in plural expressions. * Improve the test suite. ------------------------------------------------------------------- Thu Jan 2 17:27:05 UTC 2014 - lazy.kent@opensuse.org - Update to 0.13.2. * Reject early encodings that are not compatible with ASCII. * Improve the test suite. ------------------------------------------------------------------- Mon Dec 9 18:17:17 UTC 2013 - lazy.kent@opensuse.org - Update to 0.13.1. * Fix test failures with Python 3.4. * Fix stripping delay annotations from terminfo capabilities. * Improve the test suite. ------------------------------------------------------------------- Sat Sep 14 08:15:04 UTC 2013 - lazy.kent@opensuse.org - Update to 0.13. * Summary of tag changes: + Added: - conflicting-message-flags - duplicate-message-flag - invalid-range-flag - range-flag-without-plural-string - redundant-message-flag - unknown-message-flag * Check for duplicate, conflicting, redundant, or unknown message flags. * Strip leading and trailing spaces from flag lines. * Be verbose when checking for messages with empty msgid with source code references. * Reduce duplicate-flag-for-header-entry severity to minor. * Check for zero width no-break space (U+FEFF) in translations and header entries. * Improve the test suite. * Work around a trailing comment parsing bug in polib. https://bitbucket.org/izi/polib/issue/51 ------------------------------------------------------------------- Tue Jul 30 09:02:10 UTC 2013 - lazy.kent@opensuse.org - Update to 0.12. * Summary of tag changes: + Added: - boilerplate-in-date * Don't emit empty-file for MO files with only system-dependent messages, or with minor revision greater than 1. * Don't emit no-package-name-in-project-id-version if the package name consists of only non-ASCII letters. * Fix parsing some atypical PO comments. * Parse more date formats. * Check for xgettext boilerplate in dates. * Strip trailing whitespace when parsing header entry. * Allow only tabs and spaces between “nplurals=…” and “plural=…”. * Remove Bihari (codes “bh”, “bih”) from the data files; it's not a single language, but a language collection. * Implement 8-bit encodings without iconv(3) or iconv(1). * Add “SEE ALSO” section to the manual page. * Improve the test suite. * Improve the makefile. ------------------------------------------------------------------- Mon Jun 24 19:28:08 UTC 2013 - lazy.kent@opensuse.org - Update to 0.11.1. * Fix the MO file parser: + Detect encoding by inspecting only the first message with empty msgid. + Fix compatibility with Python 3.3. * Use the custom MO file encoding detection method rather than the provided by polib. - Changes in 0.11. * Summary of tag changes: + Added: - distant-header-entry - duplicate-flag-for-header-entry - duplicate-header-entry - duplicate-header-field-x-poedit - empty-msgid-message-with-plural-forms - empty-msgid-message-with-source-code-references - unexpected-flag-for-header-entry - unusual-character-in-header-entry * Fix letter codes for tags with severity important. * Reduce severity of arithmetic-error-in-unused-plural-forms, codomain-error-in-unused-plural-forms, and incorrect-unused-plural-forms to normal. * Implement custom header parser for PO files, and custom MO file parser. * Check for duplicate header entries. * Check for unusually located header entries. * Overhaul handling of duplicates and stray lines in the header entry. + Emit duplicate-header-field-x-poedit (instead of the generic duplicate-header-field) for duplicate X-Poedit-Language and X-Poedit-Country headers fields. * Work around a flag parsing bug in polib. * Check for duplicate and unexpected flags for header entries. * Check for unusual characters in header entries. * Check for messages with empty msgid (header entries?) with source code references or plural forms. * Fix some false-positive language-disparity when PO basename does not designate translation language. * Fix the no-report-msgid-bugs-to-header-field description. * Fix a few typos in the tag descriptions. * Improve the test suite. + Add new tests. + Use a dedicated nose plugin for blackbox tests. ------------------------------------------------------------------- Sat Jun 15 20:23:08 UTC 2013 - lazy.kent@opensuse.org - Update to 0.10. * Summary of tag changes: + Added: - boilerplate-in-content-type - conflict-marker-in-header-entry - conflict-marker-in-translation - duplicate-header-field-content-transfer-encoding - duplicate-header-field-content-type - duplicate-header-field-date - duplicate-header-field-language - duplicate-header-field-language-team - duplicate-header-field-last-translator - duplicate-header-field-mime-version - duplicate-header-field-plural-forms - duplicate-header-field-project-id-version - duplicate-header-field-report-msgid-bugs-to - fuzzy-header-entry * Check for boilerplate (“charset=CHARSET”) in the Content-Type header field. * Check header field name syntax. * Overhaul duplicate header field detection. + Emit duplicate-header-field only for non-standard fields. Downgrade duplicate-header-field to minor/wild-guess. + Emit duplicate-header-field-$NAME for standard fields. + Don't trust values of standard header fields if duplicates exist. * Check for conflict markers (“#-#-#-#-# … #-#-#-#-#”). * Check for fuzzy header entries. * Fix a typo in the language-team-equal-to-last-translator description. * Post-process the manual page, so that it can be more easily translated by po4a. * If iconv(3) is available in the C standard library, use it to implement encodings that are not implemented in the Python standard library. * Don't pass -s to iconv(1); it makes GNU iconv quieten errors, and other implementations don't have this option at all. * Improve the test suite: + Add new tests. + Make exception messages raised when a subprocess fails more readable. + Make it possible to use a custom Python interpreter for “make test”. ------------------------------------------------------------------- Sun Jun 9 18:58:54 UTC 2013 - lazy.kent@opensuse.org - Update to 0.9.2. * When emitting broken-encoding, don't output the whole file, but only the undecodable bytes in a small context. ------------------------------------------------------------------- Sat Jun 1 12:50:44 UTC 2013 - lazy.kent@opensuse.org - Correct source URL. - Correct build dependencies. ------------------------------------------------------------------- Sat May 25 19:58:18 UTC 2013 - lazy.kent@opensuse.org - Update to 0.9.1. * Brown paper bag release. * Don't complain about leading/trailing newlines in fuzzy messages. * Improve the test suite. - Changes in 0.9. * Summary of tag changes: + Added: - inconsistent-leading-newlines - inconsistent-trailing-newlines * Check for inconsistent leading/trailing newlines in messages. * Check for unusual characters also in plural translations. * Add information about version and date to the manual page. * Fix stripping delay annotations from terminfo capabilities. - Changes in 0.8.3. * Improve the test suite. + Skip some tests when run with (pseudo-)root privileges. * Add “test” target to Makefile. * Recognize “PROJECT VERSION” as boilerplate in the Project-Id-Version header field. - Drop i18nspector-0.8.2-run_tests.patch (fixed upstream). ------------------------------------------------------------------- Wed May 1 08:04:52 UTC 2013 - lazy.kent@opensuse.org - Run test suite (add i18nspector-0.8.2-run_tests.patch: add "test" target to Makefile; BuildRequires: python3-curses, python3-nose, python3-polib). ------------------------------------------------------------------- Thu Apr 18 07:41:49 UTC 2013 - lazy.kent@opensuse.org - Update to 0.8.2. * Make it possible to declare that a language has more than one correct Plural-Forms. * Add plural forms information for the following languages: Belarusian, Bosnian, Croatian, Hungarian, Russian, Serbian, Turkish, Ukrainian. * Improve the test suite. ------------------------------------------------------------------- Thu Feb 28 21:51:43 UTC 2013 - lazy.kent@opensuse.org - Update to 0.8.1. * Improve the documentation. * Remove an incorrect assertion in the plural expression parser. ------------------------------------------------------------------- Sun Feb 17 18:41:41 UTC 2013 - lazy.kent@opensuse.org - Requires python3 for openSUSE < 12.3. ------------------------------------------------------------------- Wed Feb 6 16:20:54 UTC 2013 - lazy.kent@opensuse.org - Initial package created - 0.8.