vocab tests: Gapfill format

In the vocabulary tests, allow for a "gapfill"/Cloze format:

I am h___y (_content_).

presented as a sentence + borderless input field with a precise number of expected letters appearing to the test taker (±similar to typing test, but with _ _ _ rather than correct response).

given in CSV:

id, sentence/context, #letters, answer, other-acceptable-answers
1, "I am h<>y (_content_).", 3, app, ""

=> using a special tag (e.g., <> or anything that would work) to indicate the location of the input field within the sentence

Bonus:

• possible to add a small picture (here, ~ as 1/i.jpg) just above/below the input field to provide help

added To do label

assigned to @bridubois

changed milestone to %0.1: Basic test & chat ready

changed the description

mentioned in commit 30897626

added To review label and removed To do label

The current implementation of this feature allow to use any amount of gaps in a text. A CSV entry for gaps looks like that:

3,I am h<app>y (content) to be <here> (ici).,-1,,,,,,,,

In my opinion, it was easier to implement, and prevent mistakes compared to the previous way of encode them.

Excellent! I agree that this is much less error-prone.

Is it possible to use basic Markdown (_content_) in the text? (If not, not a big deal.)

added 1 deleted label

closed

Observation: you may want to use a content/rude words filter to govern letter extraction from the word. For example, "I am h___y" could be 'horny' (rude), but "I am h__p_" has to be 'happy'.

mentioned in issue #123 (closed)