Initial support for JQL

[doc-sheet]

Hello.

There is some basic support for JQL, could you suggest some impovements?

For now it produces

project IN ("PROJ1","PROJ2") AND issuetype="My issue" AND (labels is EMPTY OR labels NOT IN ("stale","bug")) AND Привет is EMPTY AND repos is not EMPTY AND repos NOT IN ("main","dev")

for a query like this

J = JiraTable()
j = (
    JiraQueryBuilder()
    .where(J.project.isin(["PROJ1", "PROJ2"]))
    .where(J.issuetype == "My issue")
    .where(J.labels.isempty() | J.labels.notin(["stale", "bug"]))
    .where(getattr(J, "Привет").isempty())
    .where(J.repos.notempty() & J.repos.notin(["main", "dev"]))
)
print(j.get_sql())

[AzisK]

By JQL do you mean the selection in Jira or does it have more meanings?

[AzisK]

Could you add tests?

[doc-sheet]

By JQL do you mean the selection in Jira or does it have more meanings?
Yeah, jql as jira query language

I din't aim for full support, just the tools I use myself.
But I can add more if needed. I guess.

[AzisK]

I believe there is no need to start with full support. We can start little and then later extend it. Could you also add this possibility in the documentation?

[doc-sheet]

Sure! Done.

[AzisK]

JiraTable seems not to be imported and these seem unnecessary

[doc-sheet]

oh
sorry, copypaste issue

[doc-sheet]

ugh
There was a dependency conflict in my venv: https://stackoverflow.com/a/71674345

[wd60622]

Should the example use JiraQuery instead of JiraQueryBuilder?

[wd60622]

Think that that would provide consistency with the rest of the API

[doc-sheet]

To be honest I just don't know how to make in consistent.
There are no such thing as tables (so JiraTable is just a dummy placeholder).

Best thing I can do is probably create a bunch of overrides for from_, into and other methods which would return cls._builder(**kwargs)

Is it ok?

>>> table = JiraTable()
>>> JiraQuery._builder().where(table.project.isin(["PROJ1", "PROJ2"]))
project IN ("PROJ1","PROJ2")

[doc-sheet]

Hm. I guess JiraQuery.from_(table).where(table.project.isin(["PROJ1", "PROJ2"])) looks kind consistent.

[doc-sheet]

done

[wd60622]

Maybe the JiraQuery could have a where class method as well that initializes table.

Personally, I tend to used Fields

Maybe it would read like this

query = (
    JiraQuery
    .where(
        (Field("project").isin(["PROJ1", "PROJ2"]))
        & (Field("issuetype") == "my_issue_type")
    )
)

[doc-sheet]

Done.
Working in both ways:

>>> JiraQuery.where(Field("project").isin(["PROJ1", "PROJ2"]) & (Field("issuetype") == "my_issue_type"))
project IN ("PROJ1","PROJ2") AND issuetype="my_issue_type"
>>> table = JiraTable()
>>> JiraQuery.where(table.project.isin(["PROJ1", "PROJ2"]) | table.project == 2)
project IN ("PROJ1","PROJ2") OR project=2

[wd60622]

I'd recommend against using the _ methods

[wd60622]

Looks good to me! Just one more comment on the example

[wd60622]

Some of the other examples show the result explicitly

[doc-sheet]

Done.
Also I forgot to import JiraQuery into main class. So I also had to add Table method for consistency