Format table/view names in lowercase about plsql-formatter-settings HOT 6 CLOSED

There's a setting called "Identifiers case", but if I change that to lower it will also format the parameters and function/procedure/cursor names.

That's correct. Function names, procedure names, cursor names, variable names, synonym names, table names, view names, column names etc. are all identifiers. So this setting works as expected.

There is no setting to change the case for table/view names only.

from plsql-formatter-settings.

Hi @PhilippSalvisberg, thanks for your response.

Do you know if there's any other way of achieving it?
Using an external tool or something else?

from plsql-formatter-settings.

Do you know if there's any other way of achieving it?
Using an external tool or something else?

IMO everything is possible. But it's not that easy. In various cases, you cannot distinguish between a view, table, synonym or function. Here's an example:

select * from x;

In this case x can be a table, a view, a synonym or even a function.

If you want to change the case of views and tables only, you need a tooling that provides semantic information such as PL/Scope. So, it's possible. You could even implement that using Arbori by querying the PL/Scope data (if available) from a callback function. However, this would make the formatter unbearable slow.

from plsql-formatter-settings.

@mmunoz-mic I understand you are using camelCase for function, parameter, and cursor names and would like to preserve them. You would like to change TABLE_1, TABLE_2 and TABLE_3 to lowercase.

What about changing all identifiers to lowercase if they are in uppercase? This could be a single Arbori rule at the end of your formatter. This should produce the result you want. Changing all identifiers to lowercase while preserving existing identifiers in camelCase.

from plsql-formatter-settings.

Hi @PhilippSalvisberg,

Sorry for the late response, I've been out of office.
I understand the implications this would have in performance in the formatter.

I think I can work something out with regex to do the identifier case fix after the formatter.

I'd love to have everything in lowercase but this is used in a corporate environment and camelCase is used for function, parameter, cursor, package and types identifiers.

from plsql-formatter-settings.

As mentioned before, you can achieve what you want with Arbori. Look at i11_enforce_nonquoted_identifiers that is also updating the token stream. It should be more reliable than a RegEx-based solution because you can easily find relevant identifiers.

We will not include such a feature in the formatter provided in this repo. Therefore I'm closing this issue.

from plsql-formatter-settings.