Non-deterministic creation of perseus exercises

[ivanistheone]

• ricecooker version: 0.6.40
• Python version: 3.5.3
• Operating System: Linux

Description

Running Khan Academy zh-CN and it with nearly identical source trees resulted in all the exercises being marked as different and having to re-import them.

What I Did

• Run KA chef, DEPLOY, PUBLISH
• Import in Kolibri
• Re-run KA chef, DEPLOY, PUBLISH
• Update channel in Kolibri

Expected

Only one or two nodes to have changed (the topics nodes modified).

Actual

Every single exercise had changed and required re-importing.

Possible causes

I suspect two possible cause:

1. this line could be the cause: https://github.com/learningequality/ricecooker/blob/master/ricecooker/classes/questions.py#L275 and since running on Python 3.5 the dict orders are not guaranteed to be consistent between different runs.

2. The other possible cause is the .perseus file generation on Studio could be non-deterministic, see https://github.com/learningequality/studio/blob/develop/contentcuration/contentcuration/utils/publish.py#L342-L359 which uses create predictable zip code code slightly different from the predictable zip used in riececooker.

Real life consequences

Khan Academy channel users will need to redownload many files (small but still many) every time a new channel is published, even though exercises are substantially the same.

[ivanistheone]

An alternative way to store the "cache" of generated persues files --- let's just add a new table:

class GeneratedExerciseFileCache(models.Model):
    """
    A lookup table to avoid re-generating exercise export formats (perseus specifically).
    If the current `md5(exercise_data)` of an exercise node matches the
    `exercise_data_hash` of some row in this table, reuse the perseus .zip file `file`.
    """
    # id = autoincrementing int
    exercise_data_hash = models.CharField(max_length=400, blank=True, db_index=True)
    file = models.ForeignKey('File', null=True, blank=True, related_name='_')
    created = models.DateTimeField(auto_now_add=True, verbose_name=_("created"))

[ivanistheone]

Update post cpUs luncheon (June 17)

Much simpler solution to add a exercise_data_changed (bool) field on ContentNode (only relevant for kind=exercise). Credit @kollivier

on create_perseus_exercise

First PUBLISH:

• lookup for .perseus file fails (exercise_data_changed not considered)
• call create_perseus_zip which
  • generates perseus zip files
  • saves it as a File associated to the exercise ContentNode
  • set exercise_data_changed=False
  • adds perseus zip to exported kolibri db

Subsequent PUBLISH after noop:

• lookup for .perseus finds a candidate
• we see exercise_data_changed is False so OK to reuse cached file
• add existing perseus zip file to exported kolibri db

Subsequent PUBLISH after exercises edited:

• lookup for .perseus finds a candidate
• we see exercise_data_changed is True so NOT OK to reuse cached file
• call create_perseus_zip which
  • generates perseus zip files
  • saves it as a File associated to the exercise ContentNode
  • set exercise_data_changed=False
  • adds perseus zip to exported kolibri db

Frontend requirements

• set exercise_data_changed=True every time an exercise or exercise question is edited

[ivanistheone]

See more recent discussion on Studio here: learningequality/studio#1982