diagnostic · one of 45 Pro translation file
i18n strings
How do I get every string on this page into a translation file?
Starting a translation means finding every string first, and the usual way is somebody reading the page and copying lines into a spreadsheet. This lists the visible text and the four attributes a screen reader or a tooltip reads - alt, aria-label, placeholder and title - deduplicated, each with a generated key, and marks the text an inline element cuts in half, since a translator cannot reorder half a sentence.
Reach it Press ⌘ + K on any tab and type i18n strings. Ctrl + K on Windows and Linux
in the panel · this one is Pro
What this actually is
every visible string as an i18next, .po or csv translation fileScoping a translation for a quote, handing a copywriter the whole page as a sheet, or starting to localise a site that was built with its copy hardcoded. Pseudoloc finds the strings that were missed; this one collects the ones that are there.
Asked about this diagnostic
3 of themWhat is a fragment?
Text that an inline element - a link, a bold word - splits into pieces. Each piece becomes its own string, and a translator who needs to reorder the sentence cannot. They are counted so they can be joined in the source first.
Are the keys good enough to commit?
They are stable enough to start from: the first five words of each string. Most teams rename them to describe where the string lives before merging.
Is the page's text sent anywhere?
No. The reading goes from the page to the panel and stops there. Nothing is requested and nothing is stored.
how to use i18n strings
- Open i18n strings on the page you want to translate.
- Read the count of strings, fragments and strings with numbers in them.
- Choose i18next JSON, gettext .po or CSV as the format.
- Copy the file, and rename the generated keys before committing it.
96 of these · no host permissions · three free audits