Output format for weeklyplanner

It turns out that the output format for the "semainier" is not optimal for our post-collection processing tools
is transformed in our processing tools in a
This format is problematic for data recovery in our tools because WEEKLYPLANNER is more a wrapper than a real variable.

---

Here is another proposition to store the data in JSON

- A DATES vector containing the dates of the days of the surveyed week. (format YYYY-MM-DD)
- A DATES_STARTED vector which allows to know if the day has been started (certainly a useful variable for the front end ...)
- 96 variables : relative to the 96 possible slots per 24 hours (24 * 4 quarters of hours), with a boolean allowing to determine if the slot has been worked or not (Variable names should be defined on 5 positions "00h00","00H15","00H30",......."23H30",23H45")

That would give this JSON extract :

```
"DATES":{
"COLLECTED": [
"2023-05-01",
"2023-05-02",
"2023-05-03",
"2023-05-04",
"2023-05-05",
"2023-05-06",
"2023-05-07"
],
"EDITED": null,
"FORCED": null,
"INPUTED": null,
"PREVIOUS": null
},
"DATES_STARTED":{
"COLLECTED": [
"true",
"false",
"false",
"false",
"false",
"false",
"false"
],
"EDITED": null,
"FORCED": null,
"INPUTED": null,
"PREVIOUS": null
}
,
"00H00": {
"COLLECTED": [
"true", <= means that the slot 00H00-00H15 was worked on 2023-05-01
"false",<= means that the slot 00H00-00H15 was not worked on 2023-05-02
"false",
"false",
"false",
"false",
"false"
],
"EDITED": null,
"FORCED": null,
"INPUTED": null,
"PREVIOUS": null
},
"00H15": {
"COLLECTED": [
"false",
"false",
"false",
"false",
"true", <= means that the slot 00H15-00H30 was worked on 2023-05-05
"false",
"false"
],
"EDITED": null,
"FORCED": null,
"INPUTED": null,
"PREVIOUS": null
},....

```