Related discussion: https://wordpress.slack.com/archives/C5UNMSU4R/p1519137264000547

The current npm run gettext-strings script will generate a languages/gutenberg.pot file using a custom Babel plugin, but only includes strings discovered in the JavaScript source, not those in PHP. The PHP strings can be generated by i18n tools for PHP.

Furthermore, it appears that the automatic plugin integration for Gutenberg + GlotPress will ignore the custom generated POT file and scan PHP files for strings, thereby ignoring the JavaScript strings altogether.

There are a few options to proceed here:

• We can generate a file which includes the JavaScript strings written as PHP equivalent function calls, which are detected by plugin integration automatically. This is a bit of a hack, but one which has served both Yoast and WordPress.com Calypso.
• We can work to improve GlotPress interoperability with projects with JavaScript localization, and include in our build step both JS and PHP string extraction with merge

In either case, we also want to ensure that, if possible, we only load strings into the browser which are relevant to JavaScript usage.