[WIP] Make checking optional for dependent encoding profiles #162
Conversation
|
Implements #160. |
| SET ROLE TO postgres; | ||
|
|
||
| CREATE OR REPLACE FUNCTION ticket_state_next(param_project_id bigint, param_ticket_type enum_ticket_type, param_ticket_state enum_ticket_state) | ||
| CREATE OR REPLACE FUNCTION ticket_state_next(param_project_id bigint, param_ticket_type enum_ticket_type, param_ticket_state enum_ticket_state, param_ticket_id bigint default NULL) |
There was a problem hiding this comment.
Mh, not really convinced having that extra parameter is a good idea.
Before this change, the function was generic, independent of a specific ticket, hence all the parameters to answer the question at hand. But now it's for a single ticket, so you could go ahead and just have ticket_id as only parameter, since all the other information is stored in the database as well.
There was a problem hiding this comment.
Mh, not really convinced having that extra parameter is a good idea.
What do you suggest? Moving to ticket_id as only parameter or switching to encoding_profile_[version_]_id?
There was a problem hiding this comment.
Both would work. I don't have a strong opinion on this.
Are there any callers, where no ticket_id is available or hard to determine?
I'm all for simplifying, if possible.
There was a problem hiding this comment.
This should behave the same way as the old function when calling, without ticket_id, but we should check that again.
| ts2.ticket_state = param_ticket_state AND | ||
| (pts.skip_on_dependent = false OR | ||
| ( /* is master encoding ticket */ | ||
| SELECT ep.depends_on |
There was a problem hiding this comment.
Like with other properties before, for performance improvements a column "is_master" in tbl_ticket filled on INSERT e.g. using a trigger might be useful.
|
Superseded by #171. |
Migrated pull request from private repo, cleanup and thoroughly review necessary.