msguniq

msguniq [OPTION]... [INPUTFILE]...

-o FILE, --output-file=FILE
    Write output to the specified file instead of standard output.

-s, --sort-output
    Sort the output by msgid entries, which can make the file easier to review.

-u, --unique
    Output only unique messages; remove all duplicate occurrences, keeping only one instance of each msgid.

-z, --no-location
    Do not write #: filename:line comment lines (source code references).

-N, --add-location
    Write #: filename:line comment lines (this is the default behavior).

--keep-first
    When consolidating duplicates, keep the comments from the first occurrence found.

--keep-last
    When consolidating duplicates, keep the comments from the last occurrence found.

-v, --verbose
    Increase verbosity, showing progress messages and information about duplicates found.

--help
    Display a help message and exit.

--version
    Display version information and exit.

msguniq is a utility from the GNU gettext tools suite. Its primary purpose is to find and unify duplicate message definitions within GNU gettext Portable Object (PO) files. PO files are used in internationalization (i18n) and localization (l10n) to store translatable strings and their translations.

When multiple source code locations use the same translatable string, xgettext might create multiple entries for the same msgid in a PO file. msguniq processes these files, consolidating identical msgid entries into a single entry, combining their source code location references (comments). This helps in cleaning up PO files, reducing redundancy, and ensuring that translators only need to translate each unique string once, while still maintaining all references to its occurrences in the code. It is often used as part of a localization workflow to prepare PO files for translation or for merging updates.

msguniq primarily focuses on consolidating msgid entries. It relies on the msgid for uniqueness. If msgid entries are identical but have different msgctxt (message context), they are still treated as distinct by msguniq by default, which is correct behavior for gettext. It does not perform translation or fuzzy matching; its scope is strictly on de-duplication of message IDs within a single PO file.

msguniq reads PO files specified as command-line arguments. If no files are given, it reads from standard input. The processed output is written to standard output by default, unless the --output-file option is used to specify an output file. This allows for piping its output to other gettext utilities or redirecting it to update the original PO file.

In a typical localization workflow, msguniq is often used after xgettext has extracted strings from source code (which might produce duplicates) and before sending the PO file to translators or before merging it with existing translations using msgmerge. By unifying duplicates, it ensures a cleaner, more efficient translation process.

msguniq is an integral part of the GNU gettext project, which was started in 1990 by Sun Microsystems and later adopted and extended by the GNU Project. Its development has been continuous, providing robust tools for internationalization and localization. msguniq was created to address the common problem of duplicate strings in PO files that can arise from different source code locations referring to the same translatable string, streamlining the translation workflow by ensuring each unique message is handled only once.

gettext(1), xgettext(1), msgmerge(1), msgfmt(1), msginit(1)