Skip to content

Latest commit

 

History

History
193 lines (169 loc) · 10.9 KB

README.md

File metadata and controls

193 lines (169 loc) · 10.9 KB
Raw

android-gradle-localization-plugin

Maven Central Bintray Android Arsenal Build Status codecov

Gradle plugin for generating localized string resources

Overview

This plugin generates Android string resource XML files from CSV or XLS(X) file. Generation has to be invoked as additional gradle task. Java 1.8 is required.

Supported features

  • string arrays - see Arrays
  • plurals - see Plurals
  • non-translatable resources - translatable XML attribute
  • auto-escaping double quotes, apostrophes and newlines
  • auto-quoting leading and trailing spaces
  • syntax validation - duplicated, empty, invalid names detection
  • comments
  • formatted strings - formatted XML attribute
  • default locale specification - tools:locale
  • evaluate cell formulas

Applying plugin

Gradle 2.1+

In whichever build.gradle file.

plugins {
  id 'pl.droidsonroids.localization' version '1.0.19'
}

Note: exact version number must be specified, + cannot be used as wildcard.

All versions of Gradle

  1. Add dependency to the top-level build.gradle file.
 buildscript {
    repositories {
        mavenCentral()
        jcenter()
    }
    dependencies {
        classpath 'com.android.tools.build:gradle:3.5.3'
        classpath 'pl.droidsonroids.gradle.localization:android-gradle-localization-plugin:1.0.19'
    }
}

Note: mavenCentral() and/or jcenter() repository can be specified, + can be used as wildcard in version number.

  1. Apply plugin and add configuration to build.gradle of the application, eg:
apply plugin: 'pl.droidsonroids.localization'

Usage

Invoke localization gradle task. Task may be invoked from commandline or from Android Studio GUI.

  • from commandline: ./gradlew localization (or gradlew.bat localization on Windows)
  • from GUI: menu View->Tool Windows->Gradle and double click localization

Non existent folders will be created. WARNING existing XML files will be overwritten.

Example

The following CSV file:

name,default    ,pl       ,comment   ,translatable
file,File       ,"Plik"   ,file label,
app ,Application,,,false

will produce 2 XML files:

  • values/strings.xml:
<?xml version="1.0" encoding="utf-8"?>
<resources>
  <string name="file">File</string><!-- file label -->
  <string name="app" translatable="false">Application</string>
</resources>
  • values-pl/strings.xml:
<?xml version="1.0" encoding="utf-8"?>
<resources>
  <string name="file">Plik</string><!-- file label -->
</resources>

Configuration

Add localization extension in build.gradle of particular module.

localization {
        csvFile=file('translations.csv')
        OR
        csvFileURI='https://docs.google.com/spreadsheets/d/<key>/export?format=csv'
        OR
        csvGenerationCommand='/usr/bin/xlsx2csv translation.xlsx'
        OR
        xlsFile=file('translations.xlsx')
        OR
        xlsFileURI='https://docs.google.com/spreadsheets/d/<key>/export?format=xlsx'
}
  • csvFileURI and xlsFileURI can be any valid URI, not necessarily Google Docs' one
  • xlsFile and xlsFileURI accepts both XLSX and XLS files. If filename ends with xls file will be treated as XLS, XLSX otherwise

Sources, exactly one of them must be specified:

  • csvFile, xlsFile - CSV/XLS(X) file, Gradle's file() can be used to retrieve files by path relative to module location or absolute
  • csvFileURI, xlsFileURI - CSV/XLS(X) file URI
  • csvGenerationCommand - shell command which writes CSV as text to standard output. Command string should be specified like for Runtime#exec(). Standard error of the command is redirected to the standard error of the process executing Gradle, so it could be seen in the Gradle console.

Spreadsheet format:

  • defaultColumnName - default='default', name of the column which corresponds to default localization (values folder)
  • nameColumnIndex - default=unset (nameColumnName is taken into account), index of the column containing key names (source for the name XML attribute)
  • nameColumnName - default='name' (if nameColumnIndex is not present), name of the column containing key names (source for the name XML attribute)
  • translatableColumnName - default='translatable', name of the column containing translatable flags (source for the translatable XML attribute)
  • commentColumnName - default='comment', name of the column containing comments
  • formattedColumnName - default='formatted', name of the column formatted flags (source for the formatted XML attribute)

If both nameColumnIndex and nameColumnName are specified exception is thrown.

The following options turn off some character escaping and substitutions, can be useful if you have something already escaped in source:

  • escapeApostrophes - default=true, if set to false apostrophes (') won't be escaped
  • escapeQuotes - default=true, if set to false double quotes (") won't be escaped
  • escapeNewLines - default=true, if set to false newline characters won't be escaped
  • convertTripleDotsToHorizontalEllipsis - default=true, if set to false triple dots (...) won't be converted to ellipsis entity &#8230;
  • escapeSlashes - default=true, if set to false slashes (\) won't be escaped
  • normalizationForm - default=Normalizer.Form.NFC if set to null Unicode normalization won't be performed, see javadoc of Normalizer for more details

(X)HTML tags escaping

  • tagEscapingStrategy - default=IF_TAGS_ABSENT, defines X(H)TML tag brackets (< and >) escaping strategy possible values:
  • ALWAYS - brackets are always escaped. E.g. "<" in source becomes "&lt;" in output XML
  • NEVER - brackets are never escaped. E.g. "<" in source is passed without change to output XML
  • IF_TAGS_ABSENT - Brackets aren't escaped if text contains tags or CDATA section. E.g. <b>bold</b> will be passed without change, but "if x<4 then…" becomes "if x&lt;4 then…".
  • tagEscapingStrategyColumnName - default=unset (no column), name of the column containing non-default tag escaping strategy, if cell is non-empty then strategy defined there is used instead of global one

CSV format:

XLS(X) format:

  • sheetName - default=<name of the first sheet>, name of the sheet to be processed, only one can be specified, ignored if useAllSheets is set to true
  • useAllSheets - default=false, if set to true all sheets are processed and sheetName is ignored
  • evaluateFormulas - default=false, if set to true evaluates formulas in cells

Advanced options:

  • ignorableColumns - default=[], columns from that list will be ignored during parsing. List should contain column names e.g. ['Section', 'Notes']. Columns containing only empty cells are always ignored.
  • allowNonTranslatableTranslation - default=false, if set to true resources marked non-translatable but translated are permitted
  • allowEmptyTranslations - default=false, if set to true then empty values are permitted
  • handleEmptyTranslationsAsDefault - default=false, if set to true empty values do not result in entries in non-default languages, i.e. no empty XML entries for non-default languages are created. If set to true then allowEmptyTranslations is ignored for all but default language
  • outputFileName - default=strings.xml, XML file name (with extension) which should be generated as an output
  • outputIndent - default= (two spaces), character(s) used to indent each line in output XML files
  • skipInvalidName - default=false, if set to true then rows with invalid key names will be ignored instead of throwing an exception
  • skipDuplicatedName - default=false, if set to true then rows with duplicated key names will be ignored instead of throwing an exception. First rows with given key will be taken into account.
  • defaultLocaleQualifier - language (eg. es) and optionally region (eg. es_US) ISO codes of default translations. Default=null(unset) which effectively means English en, if set then value will be placed in tools:locale XML attribute. See Tools Attributes for more information.

Migration from versions < 1.0.19

Since version 1.0.19 completely empty (effectively empty in case of XLS(X)) rows and columns are ignored. Moreover if qualifier (usually language code) is empty exception is thrown. Previously behavior in such cases was undefined.

Migration from versions < 1.0.13:

Obsolete, non-scoped localization plugin id is no longer supported. The only valid id is pl.droidsonroids.localization.

Migration from versions < 1.0.7:

Versions older than 1.0.7 provided escapeBoundarySpaces option, which defaulted to true. Currently strings are always escaped when corresponding parsed cell contains leading or trailing spaces, but such spaces are stripped by default CSV strategy. So effectively strings are trimmed by default. If you want to include mentioned spaces in output set appropriate csvStrategy.

License

MIT License
See LICENSE file.