Foreign Key processing updates #55
Conversation
| ) | ||
| if ddl: | ||
| tables = DDLParser(ddl).run(group_by_type=True) | ||
| tables = DDLParser(ddl, normalize_names=True).run(group_by_type=True) |
There was a problem hiding this comment.
this feels like a good change to me - i think i would want my output code to have normalized names. I was having problems with my foreign keys until i put this setting in, but i'll figure out where that went wrong if you don't like this change.
| [tool.poetry.dependencies] | ||
| python = ">=3.7,<4.0" | ||
| simple-ddl-parser = "^0.28" | ||
| simple-ddl-parser = "^1.0.0" |
There was a problem hiding this comment.
i hope this is OK. All the changes i submitted to simple-ddl-parser were to make this work for my DDL. ;)
| __tablename__ = 'users_WorkSchedule' | ||
| id = db.Column(ARRAY((1,1)), primary_key=True) | ||
| id = db.Column(db.Integer(), primary_key=True) |
There was a problem hiding this comment.
looking at the code above, I think this was actually a bug in the test/parsing from before. with the updated simple-ddl-parser version the parsing bug has been fixed and so the test updated. please let me know if i interpreted that incorrectly.
|
@cfhowes thanks for opening the PR, I will check little bit later |
|
@cfhowes released in version 0.15.1, thank you one more time! |
2 participants
I was having some problems with how FOREIGN KEYs were being handled for both keys listed after the COLUMN statements (rather then inline in the statement) and for those with compound keys.
If thinks look reasonable, I'll update the changelog as well.