Associated feature: T290781: Add an image: captions
When the user is past the caption onboarding and is in the caption step, we want to record events for these actions:
- Impression of landing on this step
- Selecting the "back" arrow to return to the image inspector
- Tapping into the caption box to focus there (only applicable if we do not bring the user's focus there automatically).
- Selecting to view image details, closing that dialog, and selecting the link to Commons in that dialog. This is already specified in T294669: Image instrumentation: image inspector.
- Selecting the question mark to re-open the caption onboarding, and closing that dialog. This is already specified in T294670: Image instrumentation: captions onboarding dialog.
- Each time the caption validation is triggered. We should record which validation rule is triggered. In Iteration 1, there will only be one rule, which we can refer to as "too short". See T293161: Add an image: minimal caption validation.
For each event, we want to include:
- Which article they are on
- Which image is being suggested
- The source for the suggestion (i.e. crosswiki, wikidata, commons category)