Actions

Difference between revisions of "Question attributes redesign"

From LimeSurvey Manual

(Scenarios)
(Scenarios)
Line 32: Line 32:
  
 
[[User:Olle|Olle]] ([[User talk:Olle|talk]]) 17:31, 3 February 2020 (CET)
 
[[User:Olle|Olle]] ([[User talk:Olle|talk]]) 17:31, 3 February 2020 (CET)
 +
 +
For twig : is better to have a single language … {{ Question.i18attribute }} and not {{ Question.i18attribute[????] }}
 +
 +
[[User:DenisChenu|DenisChenu]] ([[User talk:DenisChenu|talk]]) 17:38, 3 February 2020 (CET)
  
 
== Notes ==
 
== Notes ==

Revision as of 18:38, 3 February 2020

Question attribute code saw a rise in complexity between LS3 and LS4. This page will take a step back and analyze if it can be simplified.

Domain diagram

Domain diagram

Class diagram

TODO

Database design

TODO.

Will not change.

Scenarios

  • Get potential question attribute from core
  • Import question attribute from XML/question theme
  • Add (replace ?) question attribute from plugins
  • Save question attribute value (for all language or not)
  • Load question attributes for question type/question id
  • Load question attributes for question type/question id for a single language (for qanda/QuestionTheme and SurveyTheme)

---

> Load question attributes for question type/question id for a single language (for qanda/QuestionTheme and SurveyTheme)

If you already can load for all languages, you don't have to load for one language, IMO. Unless the number of languages are > 20, which I don't think happens very often...? Also, this is done via lazy loading through the ORM. I think. (Unless explicitly loading it eagerly.)

Olle (talk) 17:31, 3 February 2020 (CET)

For twig : is better to have a single language … Template:Question.i18attribute and not {{ Question.i18attribute[????] }}

DenisChenu (talk) 17:38, 3 February 2020 (CET)

Notes

TODO

Future evolution and remark