This Gradle plugin adds tasks to upload and download translation files from crowdin.com
crowdinDownload downloads the latest translations and crowdinUpload uploads a source file.
Uses API v1.
Add this to your build.gradle:
plugins {
id "com.mendhak.gradlecrowdin" version "0.1.1"
}
For pre-Gradle-2.1, follow the instructions here
For download, the plugin invokes an export of your project's translations and then downloads the translations in a zip file. It then unzips and copies to the given destination.
crowdinDownload{
apiKey = "31727f222f203349979cf710a471b767"
destination = "$projectDir/src/main/res"
projectId = 'my-test-project'
//Optional
//Android projects
renameMapping = [
from: '^(.*)/(.*)$',
to: /values-\1\/\2/
]
//Optional
excludePattern = '**/*.txt'
}
In some cases you may want to modify the structure as it's copied over.
For example, Android projects would use values-es instead of es. You can use the renameMapping option to do this.
You may also want to exclude certain files, you can do this using excludePattern.
Crowdin gives ru files as well as pt-BR files. You can get them both with two tasks.
task crowdin1(group:"crowdin", type:com.mendhak.gradlecrowdin.DownloadTranslationsTask){
apiKey = CROWDIN_API_KEY
destination = "$projectDir/src/main/res"
projectId = 'gpslogger-for-android'
//Android projects
renameMapping = [
from: '^([^-]*)/strings.xml$',
to : /values-\1\/strings.xml/
]
excludePattern = '**/*.txt'
}
task crowdin2(group:"crowdin", type:com.mendhak.gradlecrowdin.DownloadTranslationsTask, dependsOn:"crowdin1"){
apiKey = CROWDIN_API_KEY
destination = "$projectDir/src/main/res"
projectId = 'gpslogger-for-android'
//Android projects
renameMapping = [
from: '^([^-]*)-([^-]*)/strings.xml$',
to : /values-\1-r\2\/strings.xml/
]
excludePattern = '**/*.txt'
}
And then call it like so
task getallTranslations(group:"crowdin", dependsOn:"crowdin2") { }
As an alternative you can go to the file settings on crowdin and set the resulting file name to something similar to /app/src/main/res/values-%android_code%/strings.xml. Crowdin will then automatically generate the appropriate file structure inside the downloaded zip file. In the example file name you would have to also adjust the destination to "$projectDir".
Point this at your source file, such as values/strings.xml and the task will update it on crowdin.
crowdinUpload {
apiKey = "31727f222f203349979cf710a471b767"
projectId = 'my-test-project'
files = [
[ name: 'strings.xml', source: 'app/src/main/res/values/strings.xml' ]
]
}
Note that the file must exist on crowdin, this simply acts as an updater. New files are not created. The files attribute expects a list of files where the name entry referrs to the file name used on crowdin and the source entry to the corresponding file in your project. This also allows you to batch upload strings.xml files from separate modules.
Either use the combination of project identifier(projectId) and project key(apiKey) or combination of project identifier(projectId), username(username), and account key(accountKey) to authorize yourself. If they are all present in the configuration file, the second option is prioritized.
crowdinDownload {
projectId = PROJECT_ID
//First option
apiKey = CROWDIN_API_KEY
//Second option, prioritized
username = USERNAME
accountKey = ACCOUNT_KEY
//...
}
You can also use the version management features of crowdin. This allows you to upload your files separately for each of your release and feature branches. For more details visit the official documentation.
The following snippet gets the currently checked out branch automatically
buildscript {
repositories {
jcenter()
}
dependencies {
classpath 'org.ajoberstar:grgit:1.9.0'
}
}
ext {
git = org.ajoberstar.grgit.Grgit.open()
gitBranchName = git.branch.getCurrent().name
}
The branch name needs to be set for the crowdinDownload and crowdinUpload tasks.
crowdinDownload {
...
branch = gitBranchName
...
}
crowdinUpload {
...
branch = gitBranchName
...
}
If the branch does not exist yet on crowdin it will be created automatically. Keep in mind that the characters \ / : * ? " < > | are not allowed in crowdin branch names and will be replaced with underscores by this crowdin plugin. Since a branch on crowdin is basically a special folder all translation files will be copied to the newly created branch. Therefore if you have set a title or the 'resulting file' field on crowdin and you want to keep them also for your newly created branches you need to specify the title and/or translation fields as well.
crowdinUpload {
...
files = [
[
title : 'App', // Artifact name shown to the translators (optional)
name : 'strings.xml', // File name used on crowdin
source : 'app/src/main/res/values/strings.xml', // File that should be translated
translation: 'app/src/main/res/values-%android_code%/strings.xml' // Path to store the translation (optional)
]
]
...
}
In IDEA, the task should appear under the category crowdin as shown here
To build, run the uploadArchives task under upload. This doesn't actually upload it anywhere, it just copies it to the 'repo' folder.
You can then reference it in another project locally, modify the gradle file to look like so:
buildscript {
repositories {
mavenCentral()
maven {
url uri('/home/mendhak/Code/gradle-crowdin-plugin/plugin/repo')
}
}
dependencies {
classpath 'com.mendhak.gradlecrowdin:crowdin-plugin:0.1.1'
}
}
apply plugin: 'com.mendhak.gradlecrowdin'
I then publish to Maven using the ./gradlew bintrayUpload task, and to the Gradle using ./gradlew publishPlugins
This is running on an old version of Gradle. I tried upgrading and nothing would build or run; the errors were cryptic and the stack traces were useless. I had to stick to OpenJDK 1.8, Gradle 2.6.