Support all IANA mime types #82
Conversation
| @@ -0,0 +1,9 @@ | |||
| # build-mime-types | |||
|
|
|||
| Builds a class file for ContentType enum from the IANA mime types list. | |||
There was a problem hiding this comment.
I really like the approach with code-generation based on the standards, but since there is no automation to get the latest version of those csv files - it makes the module useful just for one-time execution. I think there is no need to include it into the repository since the process of extending will be identical (update csv of the actual python enum file)
There was a problem hiding this comment.
Where would the code live if it doesn't live in this repo?
since the process of extending will be identical (update csv of the actual python enum file)
Process:
- Download CSV files
- Run build_mime_types.py
If you delete build_mime_types.py you'll have to add thousands of enum lines by hand or do a manual diff. Maybe I'm missing something?
There was a problem hiding this comment.
I agree with this approach, but for now it's semi-automated. I meant that if we go with code generation, then we can exclude CSV files from the git and make them downloadable directly in the script to have all the latest versions w/o manual actions
There was a problem hiding this comment.
In this case your suggestion with enum map for existing values would work fine as well, so it is up to you: update it now or leave PR as is and wait till the major version update w/o BC (but I have no time estimate on that for now)
There was a problem hiding this comment.
I added the maps. There are some open api specific content-types not listed in IANA that I added / generated and some that this library names specifically so I added logic for those to keep everything backwards compatible.
build_mime_types: Add tool to create ContentType enum from all IANA mime type CSVS pyproject.toml: Run build_mime_types.py during `poetry build` src/tests: Add more tests around mime types and adjust to mime_types.ContentType from enumeration.ContentType
dkoston
force-pushed
the
support_all_mime_types
branch
from
9677b45
to
592423c
Compare
| "multipart": "MULTIPART_ANY", | ||
| "text/*": "TEXT_ANY", | ||
| "video/*": "VIDEO_ANY", | ||
| 'text/json': "JSON_TEXT", # This isn't an OpenAPI specific mime type but it is needed as it already is in use |
There was a problem hiding this comment.
FWIW: This content type is deprecated. We may want to mark as such so it can be removed later?
Overview
We have some content types in our schemas that were not supported. Instead of just adding the few we needed, I decided to grab all the IANA types as CSVs and build a tool to output the ContentType enum. I moved it out to mime_types.py as that's a generated file so, it should be kept separate.
NOTE: this should be merged after the switch to poetry PR as it is branched off that branch
Changes
build_mime_types: Add tool to create ContentType enum from all IANA mime type CSVS
pyproject.toml: Run build_mime_types.py during
poetry buildsrc/tests: Add more tests around mime types and adjust to mime_types.ContentType from enumeration.ContentType