New feature for Gettext users: automatic string grouping

I just pushed a new feature to Web Translate It, and removed another one.

RIP Filters

Filters was a feature I introduced 6 months ago. It was a way to group strings by very specific criteria. For example, you could create a filter to group all the strings having the word bacon in their source text, or all the strings having their key starting by users.new. Then the filter would list all the strings matching these criteria. It was a very powerful feature, but it was too much work for the users to setup these filters manually.

Usage statistics show that this filter feature is not are not what users need: out of more than 1,6 million translations on Web Translate It, only 13 filters were created.

Automatic String Grouping

This new feature replaces the filters, and only applies to Gettext users for now. I will extend it to other file formats in the upcoming weeks —at least for Ruby yml, based on the string’s flattened key names.

To use this feature on your project (using the Gettext .pot/.po language file format), all you have to do is to upload a new version of your master language file (categories are only created during the language file import).

Web Translate It now uses the string location metadata that is included for each strings in pretty much all .po/.pot files.

What it means is that the translation interface is now capable of grouping your strings by the file name they were extracted from.

Select the category you want to filter by, and it will display only the strings that belong to that category. It makes it much easier to translate, because it restricts the size of the list of strings.

And, icing on the cake, you will see all the categories a string belong to on the String History page.

If you click on a category, it will bring you back to the translation interface, filtered by that category.

I hope you will find this feature useful. Thank your for using Web Translate It.