Localization file formats

Weblate supports a wide range of translation formats. Each format is slightly different and provides a different set of capabilities.

Hint

When choosing a file format for your application, it’s better to stick some well established format in the toolkit/platform you use. This way your translators can additionally use whatever tools they are used to, and will more likely contribute to your project.

Automatic detection

Weblate tries to detect file format during Adding translation projects and components. The detection might be wrong for different variants of the same serialization format (JSON, YAML, properties) or file encoding, so please verify that File format is correct before creating the component.

Translation types capabilities

Please refer to the documentation page of each individual file format for information about which features are supported in that format.

Bilingual and monolingual formats

Both monolingual and bilingual formats are supported. Bilingual formats store two languages in single file—source and translation (typical examples are GNU gettext PO (Portable Object), XLIFF 1.1 and 1.2 or Apple iOS strings). On the other side, monolingual formats identify the string by ID, and each language file contains only the mapping of those to any given language (typically Android string resources). Some file formats are used in both variants, see the detailed description below.

For correct use of monolingual files, Weblate requires access to a file containing complete list of strings to translate with their source—this file is called Monolingual base language file within Weblate, though the naming might vary in your paradigm.

Additionally this workflow can be extended by utilizing Intermediate language file to include strings provided by developers, but not to be used as is in the final strings.

String states

Many file formats only differentiate “Untranslated” and “Translated” strings. With some formats it is possible to store more fine-grained state information, such as “Needs editing” or “Approved”.

Source string description

Source string descriptions can be used to pass additional info about the string to translate.

Several formats have native support for providing additional info to translators (for example XLIFF 1.1 and 1.2, GNU gettext PO (Portable Object), WebExtension JSON, CSV files, Excel Open XML, Qt Linguist .ts, go-i18n JSON files, gotext JSON files, ARB File, .NET resource files (RESX, RESW)). Many other formats extract closest comment as source string description.

Explanation

The Explanation on strings can be stored and parsed from a few file formats.

Currently supported only in TermBase eXchange format.

Source string location

Location of a string in source code might help proficient translators figure out how the string is used.

This information is typically available in bilingual formats where strings are extracted from the source code using tools. For example GNU gettext PO (Portable Object) and Qt Linguist .ts.

Translation flags

Translation flags allow customizing Weblate behavior. Some formats support defining those in the translation file (you can always define them in the Weblate interface, see Customizing behavior using flags).

This feature is modelled on flags in GNU gettext PO (Portable Object).

Additionally, for all XML based format, the flags are extracted from the non-standard attribute weblate-flags. Additionally max-length:N is supported through the maxwidth attribute as defined in the XLIFF standard, see Specifying translation flags.

Context

Context is used to differentiate identical strings in a bilingual format used in different scopes (for example Sun can be used as an abbreviated name of the day “Sunday” or as the name of our closest star).

For monolingual formats the string identifier (often called key) can serve the same purpose and additional context is not necessary.

Pluralized strings

Plurals are necessary to properly localize strings with variable count. The rules depend on a target language and many formats follow CLDR specification for that.

Hint

Pluralizing strings need proper support from the application framework as well. Choose native format of your platform such as GNU gettext PO (Portable Object), Android string resources or Stringsdict format.

Read-only strings

Read-only strings from translation files will be included, but can not be edited in Weblate. This feature is natively supported by few formats (XLIFF 1.1 and 1.2 and Android string resources), but can be emulated in others by adding a read-only flag, see Customizing behavior using flags.

Supporting other formats

Most formats supported by translate-toolkit which support serializing can be easily supported, but they did not (yet) received any testing. In most cases, an additional thin layer is needed in Weblate to hide differences in behavior of different storages.

To add support for a new format, the preferred approach is to first implement support for it in the translate-toolkit.

File format parameters

File format parameters provide a way to configure settings related to the file format. They are configured at component level and allow you to customize how file parsing and serialization are handled.

List of file format parameters

Parameter name

File formats

Label

Help text

csv_encoding

  • csv

  • csv-multi

File encoding

Encoding used for CSV files

Available choices:

auto

Auto-detect

utf-8

UTF-8

csv_simple_encoding

  • csv-simple

File encoding

Encoding used for simple CSV files

Available choices:

auto

Auto-detect

utf-8

UTF-8

iso-8859-1

ISO-8859-1

flatxml_key_name

  • flatxml

FlatXML key name

flatxml_root_name

  • flatxml

FlatXML Root name

flatxml_value_name

  • flatxml

FlatXML value name

gwt_encoding

  • gwt

File encoding

Encoding used for GWT Properties files

Available choices:

utf-8

UTF-8

iso-8859-1

ISO-8859-1

json_indent

  • json

  • json-nested

  • webextension

  • i18next

  • i18nextv4

  • arb

  • go-i18n-json

  • go-i18n-json-v2

  • formatjs

  • gotext

JSON indentation

json_indent_style

  • json

  • json-nested

  • webextension

  • i18next

  • i18nextv4

  • arb

  • go-i18n-json

  • go-i18n-json-v2

  • formatjs

  • gotext

JSON indentation style

Available choices:

spaces

Spaces

tabs

Tabs

json_sort_keys

  • json

  • json-nested

  • webextension

  • i18next

  • i18nextv4

  • arb

  • go-i18n-json

  • go-i18n-json-v2

  • formatjs

  • gotext

Sort JSON keys

json_use_compact_separators

  • json

  • json-nested

  • webextension

  • i18next

  • i18nextv4

  • arb

  • go-i18n-json

  • go-i18n-json-v2

  • formatjs

  • gotext

Avoid spaces after separators

merge_duplicates

  • markdown

  • html

  • txt

  • dokuwiki

  • mediawiki

  • asciidoc

Deduplicate identical strings

Consolidates identical source strings into a single translation unit. Prevents translation loss during file restructuring or table reordering by removing position-dependent context.

po_fuzzy_matching

  • po

Use fuzzy matching

po_keep_previous

  • po

Keep previous msgids of translated strings

po_line_wrap

  • po

  • po-mono

Long lines wrapping

By default, gettext wraps lines at 77 characters and at newlines. With the --no-wrap parameter, wrapping is only done at newlines.

Available choices:

77

Wrap lines at 77 characters and at newlines (xgettext default)

65535

Only wrap lines at newlines (like xgettext --no-wrap)

-1

No line wrapping

po_no_location

  • po

Do not include location information in the file

properties_encoding

  • properties

  • xwiki-page-properties

File encoding

Encoding used for Java Properties files

Available choices:

iso-8859-1

ISO-8859-1

utf-8

UTF-8

utf-16

UTF-16

strings_encoding

  • strings

File encoding

Encoding used for iOS strings files

Available choices:

utf-8

UTF-8

utf-16

UTF-16

xml_closing_tags

  • ts

  • plainxliff

  • xliff

  • xliff2

  • xliff2-placeables

  • poxliff

  • apple-xliff

  • resx

  • aresource

  • moko-resource

  • cmp-resource

  • tbx

Include closing tag for blank XML tags

yaml_indent

  • yaml

  • ruby-yaml

YAML indentation

yaml_line_break

  • yaml

  • ruby-yaml

Line breaks

Available choices:

dos

DOS (\r\n)

unix

UNIX (\n)

mac

MAC (\r)

yaml_line_wrap

  • yaml

  • ruby-yaml

Long lines wrapping

Available choices:

80

Wrap lines at 80 chars

100

Wrap lines at 100 chars

120

Wrap lines at 120 chars

180

Wrap lines at 180 chars

65535

No line wrapping