[DRAFT] Add CallsiteInfoAdder class which populates the event dictionary with callsite info

[aucampia]

Summary

This PR creates a CallsiteInfoAdder processor which will populate the event dictionary with callsite info:

Attribute name	Description
pathname	Full pathname of the source file where the logging call was issued (if available).
filename	Filename portion of pathname.
module	Module (name portion of filename).
funcName	Name of function containing the logging call.
lineno	Source line number where the logging call was issued (if available).
thread	Thread ID (if available).
threadName	Thread name (if available).
process	Process ID (if available).
processName	Process name (if available).

Fixes #378

Pull Request Check List

This is just a friendly reminder about the most common mistakes. Please make sure that you tick all boxes. But please read our contribution guide at least once, it will save you unnecessary review cycles!

If an item doesn't apply to your pull request, check it anyway to make it apparent that there's nothing left to do.

• Added tests for changed code.
• New APIs are added to typing_examples.py.
• Updated documentation for changed code.
  • New functions/classes have to be added to docs/api.rst by hand.
  • Changed/added classes/methods/functions have appropriate versionadded, versionchanged, or deprecated directives. Find the appropriate next version in our __init__.py file.
• Documentation in .rst files is written using semantic newlines.
• Changes (and possible deprecations) are documented in the changelog.

If you have any questions to any of the points above, just submit and ask! This checklist is here to help you, not to deter you from contributing!

[hynek]

I like the approach and I have only two concerns:

1. The thread info is probably gonna be wrong in async setups. We can say that's fine, but then we'd have to document it.
2. Similar problem: we probably should allow the user to supply more names for additional_ignores?

[aucampia]

I like the approach and I have only two concerns:

1. The thread info is probably gonna be wrong in async setups. We can say that's fine, but then we'd have to document it.
2. Similar problem: we probably should allow the user to supply more names for additional_ignores?

Thanks for the heads up on these, I will do some testing with async to see what is what, and I will allow user control of additional_ignores and I think the default in this case is redundant.

[aucampia]

deleting this PR as I will create a new one from same branch to work around the branch editing issue.

[gzmarstone]

I know its closed, but what about class if available as well... use case would be to find all of the logging calls in a particular class... You can check if the first arg is self, if it is then get the class and the name of that?
I used this example to log the classname via structlog: https://stackoverflow.com/questions/2203424/python-how-to-retrieve-class-information-from-a-frame-object

[aucampia]

I used this example to log the classname via structlog: https://stackoverflow.com/questions/2203424/python-how-to-retrieve-class-information-from-a-frame-object

@gzmarstone Nice to know about this, I will consider adding this - would be nice to have - but will only get time to look at it next year if at all. The main goal was to get parity with capabilities of the stdlib logger regarding callsite info.

Update:

You could create a logger on a class, named for the class and the module, and then use the logger name.

[gzmarstone]

No problem, I just noticed the change with the latest release and it would let me delete the code that I had to implemented to log the stdlib items, but then realized that it was missing the Class which I have, so I couldn't delete all of it yet... I can just leave that part in my code and if it gets added obsolete it, otherwise no big deal...