#twisted-dev

Tickets pending review: https://tm.tl/#7495, #9299, #9305, #9333, #9337, #9338, #9340, #8966 (the0id), #9100, #9118 (the0id), #4964 (jameshilliard), #9138, #9176

Should I file a bug against code or docs for this? I just did this to fix a bunch of Twisted logging. https://github.com/ElementalAlchemist/txircd/co... However, the "before" bit appears to be explicitly allowed according to the logger docs: https://twistedmatrix.com/documents/current/cor...

There's even an example of the same exact sort of thing I removed because Twisted was giving me errors for my log string

ElementAlchemist: seems like a bug

ElementAlchemist: but in the docs

ElementAlchemist: new-style formatting doesn't support function invocation

ElementAlchemist: however, f-strings do

ElementAlchemist: that might be the root of the confusion

even more confusing, the logger _does_ allow function invocation, but not method invocation

I get .format(); my issue is specifically with logger

ElementAlchemist: given that this works: logger.info("{call()}", call=someFunc), but this doesn't: logger.info("obj.call()", obj=someObj)

It lets you do {function()} but not {thing.method()}, so I just wanted to figure out whether what the docs say (that {thing.method()} is OK) was intended or if the code is what we're going for so I know what to call a bug :)

ElementAlchemist: i think this is a bug in the logger

That said, I've apparently been too lazy to do something about reporting it for over a year until now

thanks for bringing it up now

OK, I'll write it against the code

Thanks!

ElementAlchemist: thank you!

new logger defect https://tm.tl/#9347 by ElementalAlchemist: twisted.logger doesn't execute methods despite documentation

Issue created

this is going to be an unpleasant ticket

Did I miss something I should put in?

Or did I just unrelatedly make everyone sad? :(

ElementAlchemist: my issue is with the state of string formatting in python :)

Oh

Fair.

runciter: can't you just use str.format instead of converting things to % formatting internally? is the problem that you want to maintain compatibility with... py25, or what?

or am I misreading what's goin gon?

oh no, I see, there'es some random % formatting in the middle of that, but the problem is that str.format doesn't support method calls

that's too bad

njs: that' sit

njs: format doesn't support invocation of any kind

njs: we just have a hack for function calls

well, hack is the wrong word

we support it

Tickets pending review: https://tm.tl/#7495, #9299, #9305, #9333, #9337, #9338, #9340, #8966 (the0id), #9100, #9118 (the0id), #4964 (jameshilliard), #9138, #9176

Tickets pending review: https://tm.tl/#7495, #9299, #9305, #9333, #9337, #9338, #9340, #8966 (the0id), #9100, #9118 (the0id), #4964 (jameshilliard), #9138, #9176

str.format's language is completely awful, and wontfixed on making it not awful

tos9: do you have a reference for that?

tos9: (the wontfix part)

Tickets pending review: https://tm.tl/#8966 (the0id), #9100, #9118 (the0id), #4964 (jameshilliard), #9138, #9176, #7495, #9299, #9305, #9333, #9337, #9338, #9340

runciter: will have to find it

tos9: ty

runciter: it's a comment from raymond saying "we don't want to make the format language support 'everything'" (so instead it supports a random collection of things)

awesome

tos9: and f-strings support a superset of that random collection

Yep.

python's truly a joy

https://bugs.python.org/issue7951 is one issue, but that's not the one I was remembering

Tickets pending review: https://tm.tl/#8966 (the0id), #9100, #9118 (the0id), #4964 (jameshilliard), #9138, #9176, #7495, #9299, #9305, #9333, #9337, #9338, #9340