Remove master json when diaggregate json tests #1627
Conversation
| @@ -521,6 +530,8 @@ def write_ruletest_json(section, rule, ruleset_doc): | |||
|
|
|||
| # Write out final rule dictionary | |||
| write_ruletest_json(prev_section, prev_rule, ruleset_doc) | |||
| # Remove the master json file: | |||
| os.remove(master_json_path) | |||
There was a problem hiding this comment.
I am thinking now that having this master json exist in the repo will be helpful for the rule renumbering process. Because the spreadsheets are not part of the repo, I want to have the renumbering script be able to renumber all test jsons independent of the spreadsheets. I am imagining that the spreadsheets and master json will only use the unique id so that they have no work necessary when renumbering. The disaggregate function will handle the sectionNruleM mapping and renumber the disaggregated rule tests with the appropriate values for N and M.
There was a problem hiding this comment.
@weilixu @jugonzal07 does this sound like a good workflow to you? The only drawback that I see is that it makes it difficult for a developer to look at the spreadsheet and know which rules the tests correspond to.
I think there is also an automated solution to this drawback where the rule id map can be written into a dedicated spot of each of the spreadsheets so that each spreadsheet can perform the same mapping and display the Section N Rule M (for developer convenience only - the row would be ignored by the Excel to master JSON script)
There was a problem hiding this comment.
Would it be possible to keep the master jsons in the repo but have them excluded from the build?
There was a problem hiding this comment.
The workflow works good. I can make the code only run in build process (when releasing to PYPI). I think that was one of my intentions.
I will remove this line in the json_generation_utilities file.
… into remove_master_json_when_diaggregate_json_tests
No description provided.