Localize

Utility to convert csv data into android, ios or web string resource formats.

Motivation

A lot of the times we build applications that spans across countries and regions, and we want to add local languages (i.e localization) translation in our application. Nowadays, there are tools to manage localization, which easily integrate with CI CD pipelines to automatically update content, even so they sometimes provide resource files for platform specific format. Sometimes, they dont, and one of the most common formats you can get data is csv. Hence, this tool is written focusing csv file as input, which is described in detail later. This tool can be used to automate updating the string resources from same source, without manually updating the file each time.

Sneak Peek

Build

• Install Go.
• Goto working directory.
• To run the utility go run localize.go -file=strings.csv -platform=ios
• To build an executable, go build localize and you can redistribute.

Usage

• Prepare a csv file with 1st column as key, and subsequent columns as locales e.g.

key	en	es	it
welcome_text	welcome	bienvenidos	benvenuto
bye_text	goodbye	adiós	arrivederci

• run ./localize -file=<PATH_TO_CSV> -platform=<IOS_ANDROID_WEB> -debug=<TRUE_FALSE>
• -file: Absolute path to csv file
• -platform: platform for which string resource will be generated. Options: ios,android and web (anyone at a time).
• -debug: Show logs while creating resources, by default it is false. Options: true or false
• Currently, it will generate resource files in the same folder where the localize command is executed.

Resource files

Android

If you chose -platfrom=android, the generated files will be strings.xml, since we are targeting multiple locales, this tool will generate strings_en.xml for en locale, and so on. Here is what this generated file looks like when above mentioned csv is used:

<resources>
<string name="welcome_text">welcome</string>
<string name="bye_text">goodbye</string>
</resources>

Specifically for android, you have to specify keys without spaces.

iOS

Similarly, if you chose -platfrom=ios, the generated files will be Localized.strings, since we are targeting multiple locales, this tool will generate Localized_en.strings for en locale, and so on. Here is what this generated file looks like when above mentioned csv is used:

"welcome_text" = "welcome";
"bye_text" = "goodbye";

Web

And finally, if you chose -platfrom=web (assuming you are working on a typescript framework), the generated files will be strings.ts, since we are targeting multiple locales, this tool will generate strings_en.ts for en locale, and so on. Here is what this generated file looks like when above mentioned csv is used:

const LOCALIZED_STRINGS = {
welcome_text: "welcome",
bye_text: "goodbye",
}

export default LOCALIZED_STRINGS;

Customization

You can always customize the way this tool work, following are the files where you will find relevant logic:

• reader.go: Reads data from .csv file and returns a 2d array.
• parser.go: Gets 2d array data as input and returns a list of Record object
• writer.go: Writes to file for each Record entry, and also responsible for opening and closing data specific to platform.

Explanation

• Once the data is ready to be written in the form of list (slice) of Record, the code creates separate channels for each locale.
• This helps writing in separate files asynchronously. So for 3 locales, 3 channels will be created, with 3 goroutines.
• Once goroutines are started, they know the platform and locale, so the code creates platform + locale specific files, and populate some pretext if needed by that specific platform.
• Then the writer waits for the entry via channel that was passed to it at the time of execution of goroutine.
• A loop is started to iterate over all entries in Record list, calling locale specific channel to send entry to be written.
• This is acheived by creating a map, where locale is key, and channel instance is value. So calling channelMap[locale] will tell us where to write.
• Since channel instance is locale specific and goroutine is also locale specifice, an entry sent to en channel will be received by a goroutine that was started for en writing, since goroutine is receiving on this channel, entry is written.
• In the end, when loop is completing, CloseFile calls are made which writes closing statement (if needed by platform) to platform + locale specific file.

Releases

Find releases for your specific platforms in releases section. .zip file contains executables for 64-bit (marked as amd64) and 32-bit (marked as 386) versions. arm is target release for Apple M1 Macbooks. .exe for Microsoft Windows.

Get Involved

Feel free to create PRs and report issues or submit feedback. Just remember to be polite and respectful :)