dev-contrib

[vale ibex]

EG like this, maybe namespaced https://github.com/python-discord/sir-lancebot/tree/main/.github/workflows#L52-L56

that way you can do this https://github.com/python-discord/sir-lancebot/tree/main/.github/workflows#L105

[gritty wind]

You’re missing the actual file

In the URL

[vale ibex]

https://github.com/python-discord/sir-lancebot/blob/main/.github/workflows/lint.yaml#L52-L56

[stable mountain]

.github/workflows/lint.yaml lines 52 to 56

- name: Setup python
  id: python
  uses: actions/setup-python@v2
  with:
    python-version: '3.9'```

[vale ibex]

https://github.com/python-discord/sir-lancebot/blob/main/.github/workflows/lint.yaml#L105

[stable mountain]

.github/workflows/lint.yaml line 105

${{ steps.python.outputs.python-version }}-\```

[gritty wind]

Oh lol

[vale ibex]

it's just so that you can grab the exact version of python lower down in the workflow

[gritty wind]

Sure that sounds fine (will GitHub propagate it?)

[vale ibex]

hmmm

good point

[gritty wind]

I can make it a general output

[vale ibex]

yea might be useful

[gritty wind]

I’m out and about atm, but if you PR it I should be able to review it

[vale ibex]

Yea I'll have a play

[gritty wind]

Awesome

Sick scale tips: fork it so you can test your changes

Has to run in actions though, haven’t figured out how to do it locally

[vale ibex]

nah dude, I expect write to main privileges

thanks

[gritty wind]

Devops has write perms to all my projects by default

[vale ibex]

oh lol

[gritty wind]

Very good people, very efficient

[vale ibex]

time to forcepush author tags

ty

[gritty wind]

I stole all this from python 🧌

If you go back far enough, there is probably an actual author credit to you

I wonder if poetry will ever fix that bug I submitted a fix for

[vale ibex]

hmmm this is very weird

[gritty wind]

cough any poetry maintainers here

[vale ibex]

I am guessing the pre-commit-cache is also caching the python version

and not respecting the pythonLocation env var

[gritty wind]

Is this the tests on actions?

[vale ibex]

nah, this is a seperate project at work

I'm just rubber ducking lol

[gritty wind]

Oh lol

[vale ibex]

I have pytohn verison pinned to 3.10

so the cache has 3.10.5

but now it's trying to use 3.10.6

[gritty wind]

Ohhhhhh

[vale ibex]

setup python uses 3.10.6, but the cache has 3.10.5

[gritty wind]

Shouldn’t the cache miss then

Or did I forget to encode python version

I remember there’s a key I forgot to ever add

[vale ibex]

https://github.com/HassanAbouelela/actions/blob/main/setup-python/action.yaml#L84

[stable mountain]

setup-python/action.yaml line 84

${{ inputs.python_version }}-${{ inputs.dev }}-${{ inputs.working_dir }}\```

[vale ibex]

it's there

[gritty wind]

Hm

[vale ibex]

ahhhhhh

it's using the input

not the actual python version

IE input = 3.10, actual =3.10.6

[gritty wind]

Yeah we don’t have access to that till after

Ok I see

[vale ibex]

you have accedss to it further up

[gritty wind]

Actually we can export it

[vale ibex]

https://github.com/HassanAbouelela/actions/blob/main/setup-python/action.yaml#L61

[stable mountain]

setup-python/action.yaml line 61

- name: Setup python```

[gritty wind]

Yeah

[vale ibex]

export it from this

[gritty wind]

Alright that sounds like a good place to export it for the cache and the output

[vale ibex]

time to read up how to export props from a workflow lol

[gritty wind]

I think I do it further down

It’s a CLI command

[vale ibex]

oh, it's the echo thing

[gritty wind]

Export a bar or something

Yup

It’s documented thoroughly

In a million page document

[vale ibex]

https://github.com/HassanAbouelela/actions/blob/main/setup-python/action.yaml#L29-L32

[stable mountain]

setup-python/action.yaml lines 29 to 32

outputs:
  cache-hit:
    description: "Was there a cache hit for the python cache."
    value: ${{ steps.python_cache.outputs.cache-hit }}```

[vale ibex]

ez

well, I can get some OSS credits at work now

[gritty wind]

Oh yeah that should work too

Lmfao

[vale ibex]

your repo has been graced by Biosite Systems OSS team

[gritty wind]

This is OSS in the sense you and pydis use it

Thank you I’ve made it now

[vale ibex]

wait wut

poetry lock --no-update to bump the version removes all the metadata.files

[gritty wind]

What an amazing command

It’s fine to update lol

[vale ibex]

HassanAbouelela/actions#6 should do it

[dusky shore]

GitHub

[actions] #6 Output the Python version used when setting up Python

[vale ibex]

huh wait, all the metadata.files got wiped out anyways

weird

[gritty wind]

I don’t think you pushed any code lol

[vale ibex]

yea I was inside the tests dir so git add . didn't add the code lol

forcepushed

[gritty wind]

Awesome just some doc and version bump stuff

Version bump in root Readme and caches, and doc updates in the setup python Readme

Oooh I should do GitHub docs

[vale ibex]

caches?

[gritty wind]

I encode the action version in the cache keys to prevent cache mishaps

It probably doesn’t matter here though

[vale ibex]

ahhh, yea

pushed

[gritty wind]

Oh you fixed the bug too

Sick

[vale ibex]

lol which bug?

[gritty wind]

The one in your last commit

[vale ibex]

ahhh yea

[gritty wind]

Fun fact

You can’t create releases on GitHub mobile

[vale ibex]

lol

[vale ibex]

give me admin, i can create it

[gritty wind]

I see no issues

https://github.com/HassanAbouelela/actions/releases/tag/setup-python_v1.2.0 @vale ibex

If I haven’t fucked anything up

[vale ibex]

very cool

CI runs fine using that hook 🎉

But it's not a real test since it was a cache-miss due to the version bump

[gritty wind]

The tests do check that caching doesn’t error

[vale ibex]

ahh cool cool

[gritty wind]

I’ll add a regression test for that but later

I think I’ll just use your described setup lol

[vale ibex]

hah ye

[static canyon]

Any chance you're able to look over it? Still have absolutely no clue what's going on

[vale ibex]

running the tests locally I get a nicer traceback

https://paste.pythondiscord.com/famohoyepa

so I'm guessing resolved isn't getting set in your setup?

msg.reference.resolved also needs to be a MockMessage, with a MockMember as the author

fwiw, if you do poetry run task test and get the 3 failures, you can do poetry run task retest just to run those 3 again

[static canyon]

I mean idk

When I run the tests as poetry run task path_to_file.py everything passes, which is what I don't understand

[vale ibex]

poetry run task path_to_file.py doesn't sound valid

That's not a task

The tests are definitely not passing though. I'd suggest using the test and retest tasks.

[gritty wind]

Or run it in pycharm, it’s a very gud runner

I’ll be home soon, so I’ll take a look too

[vale ibex]

this is the relevant test traceback

[vale ibex]

this is what needs changing

[static canyon]

Lemme try fixing it

[vale ibex]

alternativly, you can set the message object such that this if statement runs```
if msg.type != MessageType.reply and not msg.reference:
continue

[static canyon]

If I wanted to do this, I'd need a MockReferencedMessage

Since I'd need the .reference

[static canyon]

So I guess the best thing to do is pass reference=None so this runs

[static canyon]

Yeah, that works

Just a question of whether we want tests to ensure that reply mentions aren't counted

If we don't care about adding that test, then all I need to do is reference=None

If we do care, then we need a new MockReferencedMessage class

[gritty wind]

hmm

!docs discord.Message.reference

[stable mountain]

discord.Message.reference

The message that this message references. This is only applicable to messages of type MessageType.pins_add, crossposted messages created by a followed channel integration, or message replies.

New in version 1.5.

[gritty wind]

It's a relatively simple model, and you can probably auto mock most of it's fields

If you want to set it up, go for it

If you don't, don't

[static canyon]

I'm not that familiar with mocks, but don't mind learning

[gritty wind]

We don't have a particularly strong policy on testing, at least for bot, so it's entirely up to you

[static canyon]

I'll look into it when I have more time

It'll be good to learn

[gritty wind]

Sure, take your time

[vale ibex]

!d r

[stable mountain]

Inventories refreshed

[gritty wind]

Django have finally added support for async views and ORM

Now if DRF can add it 😔

[gritty wind]

Anyone know why django is complaining about a model I created? The field I believe it's complaining about is:

    last_updated = models.DateTimeField(
        help_text="The date and time this data was last fetched.",
        auto_now=True,
    )

With the traceback:

django.db.utils.ProgrammingError: syntax error at or near "None"
LINE 1: ...timestamp with time zone NOT NULL, "name" varchar(None) NOT ...
                                                             ^

But I took that field straight from another model. Makemigrations was fine, the exception happened during migrations

Full migration: https://paste.pythondiscord.com/facilafowe
Startup log and traceback: https://paste.pythondiscord.com/oxizakosiy

Actually it's complaining about this one

    name = models.CharField(
        help_text="The tag's name.",
        primary_key=True,
    )

Alright it's because I'm an idiot, max_length is a required argument

[gritty wind]

Am I safe to assume that all tags in the bot repository are safe and won't be used for XSS or other exploits?

(I'm rendering them on site)

[gritty wind]

@vale ibex remember the test failure on site from a while back which I couldn't explain?

Failure
Traceback (most recent call last):
  File "C:\Github\PyDis\site\pydis_site\apps\api\tests\test_infractions.py", line 103, in test_list_all
    self.assertEqual(infractions[0]['id'], self.voiceban_expires_later.id)
AssertionError: 24 != 28

Still no repro though

I hate the coverage requirement on site

I added one feature, and suddenly the entire coverage dropped 10% from a 100

How little code is there on site

[gritty wind]

Coverage failure: total of 100 is less than fail-under=100

Thank you coverage, very cool

The only "miss" is a stupid partial :P

What the fuck, apparently a skipped unittest counts negatively towards your coverage. Why do the unittests themselves count towards the coverage anyway

Hahaha the tests are continuing to troll me

- [<Tag: Tag object (first_tag)>, <Tag: Tag object (second_tag)>]
+ [<Tag: Tag object (second_tag)>, <Tag: Tag object (first_tag)>]

[cold island]

Hey @fossil veldt, I'm looking at your replies in bot#2234 and I'm still confused. If the issue is fixed by rounding the microseconds, then why do we need the new field? and if the new field is calculated in one place, and the inserted_at field in another, then doesn't that mean that one of them is incorrect?

[dusky shore]

GitHub

[bot] #2234 Fixes infraction display duration off-by-one discrepancy

[fossil veldt]

Well so the rounding can potentially fix some of it, depending on network latency, the more robust option would be the new field that we opted for. Though the rounding doesn't hurt for future applications of that humanize datetime function in general.

I can remove the rounding and modifications to humanize datetime if you feel it's out of scope though

To make it clear, it is not required for the infraction duration fix specifically due to our new field

[cold island]

So the real fix is calculating expiry - last_applied?

[fossil veldt]

yeah, and also sending the expiry as a calculation based on last_applied

which required moving the calculation from the individual commands to the post function, and also needed to refactor the d.py typehints since originally the type autoconverted to the final datetime, it is now a delta

[cold island]

ok

What about last_applied and inserted_at being different?

[fossil veldt]

inserted_at is essentially not used for the infraction displays anymore, though it could in the future or just for database reference

inserted_at would be the time the entry was inserted into the database side, last_applied is calculated by the bot before the post

[cold island]

Did you modify the infraction search embed? I don't think I saw it in the PR

[fossil veldt]

hm.. where is that?

[cold island]

There's a command !infraction search

It displays infractions, and one of the things it shows is the time it was added

https://github.com/python-discord/bot/blob/main/bot/exts/moderation/infraction/management.py#L242

[stable mountain]

bot/exts/moderation/infraction/management.py line 242

@infraction_group.group(name="search", aliases=('s',), invoke_without_command=True)```

[fossil veldt]

hm... no I did not

I will add that now 😅

[timid sentinel]

Would it make more sense to do the rounding on the relativedelta rather than the start and end separately?

[cold island]

yeah I think it makes more sense now that you say it

[cold island]

no worries. But that leads to a follow up question of how exactly we should add it

[fossil veldt]

well I can just change the current inserted_at to last_applied in management.infraction_to_string

that should make it use the new last_applied field for the delta calculations

[cold island]

I think that the creation time is more interesting actually, maybe with an indicator that the duration was edited

Which is an issue because atm you can't tell if the duration was edited because last_applied and inserted_at don't match

In the embed, it probably makes sense to calculate the duration as expiry - inserted_at, because if we prolong a one day mute by two days, it will be three days total

[fossil veldt]

well that would be different

wait would it?

hm...

[cold island]

Thanks for your patience btw, you picked a pretty involved PR as your first one 🙃

[cold island]

It can be different in some cases yeah

[fossil veldt]

what paths actually lead to an infraction being re-applied?

[cold island]

A mistake by a mod or an appeal are usually the reasons

Or a very poor reaction by the person being infracted

[fossil veldt]

like what function paths

[cold island]

oooh lol

!src i e

[stable mountain]

Command: infraction edit

Edit the duration and/or the reason of an infraction.

Source Code

Go to GitHub

bot/exts/moderation/infraction/management.py:126

[cold island]

!src i a

[stable mountain]

Command: infraction append

Append text and/or edit the duration of an infraction.

Source Code

Go to GitHub

bot/exts/moderation/infraction/management.py:87

[fossil veldt]

I tried to just mute someone again in the test server but it didn't let me

[cold island]

You have the option to DM the member the details of their infraction with

!src i resend

[stable mountain]

Command: infraction resend

Resend a DM to a user about a given infraction of theirs.

Source Code

Go to GitHub

bot/exts/moderation/infraction/management.py:63

[cold island]

e.g if they didn't get it the first time or if we want to update them on any changes we made

[fossil veldt]

hm..

well currently I didn't update last_applied in infraction_edit, so it's only modifying expires_at, if requested

If I added the updating of last_applied together with the new expires_at, the log would only show the latest time

i.e. a 7 day infraction being edited on day 5 to 1 day would show 1 day when queried again after that

is that acceptable behavior or do we need the original infraction delta as well?

I think maybe it's fine to just only keep track of the last_applied in conjunction with the expires_at of the latest change?

[cold island]

right, which is why I'm saying ^

[fossil veldt]

do you only want to display the original duration, updated, or both?

[cold island]

just the updated, I think that's how it is now

but last_applied should be updated when editing the infraction yeah

[fossil veldt]

hm... the issue is using inserted_at with the new expires_at would result in potential mismatching again

[cold island]

About this, if inserted_at and last_applied were identical in the beginning, we could add an indicator if the infraction has been edited by checking if they're different. e.g by adding *(edited)* to the duration/expiry fields

[fossil veldt]

yeah I think that's possible, I can set both inserted_at and last_applied in the original post then

since currently inserted_at is pretty much a useless field

[cold island]

That might require a change on the site repo

[fossil veldt]

ah, is it not settable?

[cold island]

I'm not sure, don't think it is, since it's currently being set on the site

[cold island]

I don't actually see anything preventing it from being set, so you should try just sending it in post

[fossil veldt]

though uh

we still don't have the original infraction time for displaying both original and modified

we can tell if the infraction is edited, the new delta, and the original application time

[cold island]

Then what do you mean by original infraction time?

[fossil veldt]

the original infraction application time at inserted_at

and the new delta from last_applied to the new expires_at

so we no longer know how long the original infraction delta is for

[cold island]

yeah I think that's fine, that's how it is atm so we're not losing anything. There's an issue to expand the infraction schema to save editing history, but atm I think that's good enough

[fossil veldt]

I might make the current unpacking of DurationOrExpiry types into origin and expiry times into a function within time.py though

If we're doing it in management.py as well it's starting to be pretty repetitive

[cold island]

yeah that's fine

[timid sentinel]

Say we have datetimes:

start = datetime(2022, 8, 13, second=5, microsecond=500_001)
end = datetime(2022, 8, 13, second=6, microsecond=499_999)

The current behaviour is get the difference and display:

diff = relativedelta(end, start)  # relativedelta(microseconds=+999998)
diff.seconds  # 0

Your behaviour to round them first:

start_round  # Rounds up: datetime(2022, 8, 13, second=6)
end_round  # Rounds down: datetime(2022, 8, 13, second=6)
diff = relativedelta(end, start)  # relativedelta(microseconds=0)
diff.seconds  # 0

Correct behaviour of rounding after getting the diff:

diff = relativedelta(end, start)  # relativedelta(microseconds=+999998)
rounded_diff  # relativedelta(seconds=+1)
rounded_diff.seconds  # 1

And looking at that manually it makes sense, the difference is 999998 microseconds which is about a second.

[fossil veldt]

ah okay, so I might do this instead?

def round_delta(delta: relativedelta) -> relativedelta:
    """
    Round a relativedelta to the nearest second.

    Resulting relativedelta objects will have microsecond values of 0.
    """
    if delta.microseconds >= 500_000:
        delta += datetime.timedelta(seconds=1)
    delta.microseconds = 0
    return delta

Is the microseconds setting there a potential mutation of the argument?

since relativedelta doesn't seem to have a .replace()

[fossil veldt]

is this for any edit or only duration edits?

if editing reason only, should we still update both the last_applied and expires_at with a new calculation?

[cold island]

nah no need

[fossil veldt]

how come it says I can't run the command?

is this a site thing?

[vale ibex]

you can only run it in mod channels

[fossil veldt]

oh.

_>

[vale ibex]

It's to avoid leaking infraction data in public

[fossil veldt]

okay I added an edited indicator and timestamp as such, does that seem right?

[timid sentinel]

Looking at the source I think the delta += datetime.timedelta(seconds=1) will create a new delta anyway so shouldn't be any mutation issues.

What you posted looks good though

[fossil veldt]

yeah I was just wondering about the delta.microseconds = 0 part

apparently it doesn't actually mutate in a function like that? Even if the iadd doesn't run

I'm not sure why though

[timid sentinel]

It doesn't implement __iadd__ so it falls back to essentially delta = delta.__add__(datetime.timedelta(seconds=1)), making delta a different relativedelta to the one passed, so when you later do delta.microseconds = 0 it's only modifying the new delta, which is fine.

Side note: just noticed this is a thing and wtf why

In [87]: relativedelta(second=1) + relativedelta(seconds=1)
Out[87]: relativedelta(seconds=+1, second=1)

[fossil veldt]

yeah uh...

there's also microseconds and microsecond

all the attributes have plural versions?

I have no idea what it is

[fossil veldt]

but like if the microseconds is not over 500k, the if condition doesn't run and the iadd never happens

therefore the delta.microseconds = 0 actually changes the parameter object

is that acceptable or should I do a deepcopy?

[timid sentinel]

ah true

I don't really mind

could do a copy just to be safe

[fossil veldt]

I can alternatively always iadd something to not have to do deepcopy stuff

def round_delta(delta: relativedelta) -> relativedelta:
    if delta.microseconds >= 500000:
        delta += datetime.timedelta(seconds=1)
    delta.microseconds = 0
    return delta

def round_delta(delta: relativedelta) -> relativedelta:
    round_up = delta.microseconds >= 500000
    delta += relativedelta(seconds=round_up)
    delta.microseconds = 0
    return delta

I dunno if it's really more readable than a copy though

[timid sentinel]

I think a copy would be more explicit

copy.copy should be enough rather than copy.deepcopy though right?

I guess it doesn't matter

Either is fine

[fossil veldt]

have it as this rn

def round_delta(delta: relativedelta) -> relativedelta:
    """
    Round a relativedelta to the nearest second.

    Returns a new relativedelta with microsecond values of 0.
    """
    delta = copy(delta)
    if delta.microseconds >= 500000:
        delta += relativedelta(seconds=1)
    delta.microseconds = 0
    return delta

[gritty wind]

@vale ibex is ```py
content = COMMAND_REGEX.search(linkify_tag_ref, content):

meant to be .sub?

[vale ibex]

yea lol

I updated the comment when I noticed

I copied the line from your code and didn't change the func call

[gritty wind]

Noice

[vale ibex]

I only recently discovered re.sub works with func repls too

very cool

!cban 905177072522125363 14d please don't use our server as an ad board

[stable mountain]

:incoming_envelope: :ok_hand: applied ban to @nova pumice until <t:1661628178:f> (13 days and 23 hours).

[gritty wind]

Speaking of, do you know any way to get the links to show up properly

Because they are in a codeblock, they just look like a codeblock

[vale ibex]

yea, I think in our css code block has higher prio that hyperlink

atleast in text colour

[fossil veldt]

uh...

can I do a from __future__ import annotations in time.py?

[gritty wind]

What future do you need

[vale ibex]

yea, the code css is higher than a

[fossil veldt]

I want to type hint a DurationOrExpiry which is imported from bot.converters and it's circular

[gritty wind]

I could mark it as important in those sections, a bit hacky though

[vale ibex]

yeaa

[gritty wind]

Yup that's fine

[vale ibex]

the only thing in a is this

so we could just bump the prio of a entirely

[gritty wind]

Actually would it be that bad if a elements had the highest priority in general

yeah

[vale ibex]

yea that's what I'm thinking too

[gritty wind]

I'll just do it for tag pages for now, we can do more later if we decide to

[vale ibex]

[gritty wind]

Hm it doesn't seem to have worked

Still red after being marked

[vale ibex]

what does dev tools say?

it should show which class it's getting the color from

[gritty wind]

Yeah it's still using code, it's still red

[vale ibex]

also, might be worth trying to clear cache and reloading

[gritty wind]

I did manage to break the github logo again though, which is nice

[vale ibex]

very cool

did a cache reset work btw?

[gritty wind]

The styles had already loaded (I could see the !important in the styles page), but it didn't seem to do anything

[vale ibex]

huh

[gritty wind]

I'm trying to create a fiddle to test it

[vale ibex]

cool cool

css is still magic to me

I know enough to be dangerous

[gritty wind]

MVP https://jsfiddle.net/kchz8pm3/

Doesn't change here either

lmfao what the fuck chris

Oh it's because we need to specify explicitly if applies to the children

[vale ibex]

Yea ```
code {
color: red;
}

a, code {
color: blue;
}

[gritty wind]

The comma is an OR selector, just a space to select descendants

but ye ye

[vale ibex]

ahh cool

[gritty wind]

is this better

[vale ibex]

nice nice

[cold island]

https://tenor.com/view/ken-jeong-community-too-small-to-read-read-reading-gif-5494204

[gritty wind]

New cropping algorithim

Seems to mostly kind of work 🤡

We still end up with some weird cropping like this here and there, thoughts @vale ibex?

Most things go to full sentences though

[vale ibex]

looks good

[gritty wind]

I opted not to include the ellipsis after all because basically everything is trimmed

[vale ibex]

the oddity you describes could be solved by putting something like Show more on a new line underneath, which links to the full tag

[gritty wind]

Hm, interesting

[vale ibex]

only if there is more to show

The show more could link to the tag

or if we wanted to be smart, could unfurl and rerender the whole tag

[cold island]

Can the codeblock background be a tiny bit darker?

[gritty wind]

It would be difficult to do that in a good way because the location where the cropping happens and the links exist are in completely different languages haha

[gritty wind]

Possibly, but this is what's provided by bulma and used throughout the codebase

[vale ibex]

ah, then the latter could be possible then right? if the js already has the full text

[gritty wind]

What does the second suggestion mean

Ohhh gotcha

[vale ibex]

if a tag is cropped, we show something like show more at the end, clicking it removes the cropping

I imagine we'd need to re-render the tag for full md support

[gritty wind]

Nah it's already rendered will MD support

It would make the code a bit more involved, since we'd have to store the removed elements and keep track of the cropped changes

But it should be possible if we wanted

Let me see what that looks like

[vale ibex]

yea if it's too much work, this is already pretty nice

[gritty wind]

I'll leave it for now, we can get back to it later with the other improvements

[vale ibex]

sounds good

[gritty wind]

Right now I need to figure out the tag groups thing, it sounds like it would be really complicated for this system

[vale ibex]

is it not just updating the regex to capture spaces too?

being lazy on closing the `

[gritty wind]

That would read them out, but the code to display the listings is already a hack on the existing system

It would require quite a bit more logic to get it to display properly

[vale ibex]

ahh

[gritty wind]

Off the top of my head, the database might have to store the category, then the get_page_listings would need to return groups too, and the breadcrumbs system will actually need to work properly so that requires special casing the get_category function too

Since that relies on the tag actually existing on disk

@cold island

[cold island]

noice

[gritty wind]

Why are we using an external highlighter, when the markdown library we're using has support

Oh well whatever

This is an interesting page https://deploy-preview-763--pydis-static.netlify.app/pages/tags/ytdl/

Not sure I can do anything about it though

[vale ibex]

yea, ideally we'd want to wrap that text

but obviously wrapping this code block would be good, but we don't want to for other tag codeblocks 😅

it's probably fine

[hoary haven]

oooh tags page!

!ytdl

[stable mountain]

Per Python Discord's Rule 5, we are unable to assist with questions related to youtube-dl, pytube, or other YouTube video downloaders, as their usage violates YouTube's Terms of Service.

For reference, this usage is covered by the following clauses in YouTube's TOS, as of 2021-03-17:

The following restrictions apply to your use of the Service. You are not allowed to:

1. access, reproduce, download, distribute, transmit, broadcast, display, sell, license, alter, modify or otherwise use any part of the Service or any Content except: (a) as specifically permitted by the Service;  (b) with prior written permission from YouTube and, if applicable, the respective rights holders; or (c) as permitted by applicable law;

3. access the Service using any automated means (such as robots, botnets or scrapers) except: (a) in the case of public search engines, in accordance with YouTube’s robots.txt file; (b) with YouTube’s prior written permission; or (c) as permitted by applicable law;

9. use the Service to view or listen to Content other than for personal, non-commercial use (for example, you may not publicly screen videos or stream music from the Service)

[hoary haven]

could just blockquote it?

[gritty wind]

It would require the change to be made to the tag, the site only renders the content from the bot repo directly

Do @propertrys work in django models

I don't see any issues yet, so I think they do

[brazen charm]

the fields return normal objects from instances, don't think there should be any issues

[gritty wind]

Awesome, I'll use them to implement your URL suggestion then

[hoary haven]

yeah i mean the tag itself

[gritty wind]

Ah gotcha

[gritty wind]

Is there a way to make a view return and tell django to check the next URLs instead of this view?

Basically, we have:
/pages/tags/...
if .. is a tag group, I want to return from this view, and go to the one managed by
/pages/...

Actually, maybe not

It would be more effort to get that other view to work for this properly

[gritty wind]

Why do django templates not have set, so annoying 😔

Or better yet, why does a with need to be closed

[gritty wind]

Tag groups are live now https://deploy-preview-763--pydis-static.netlify.app/pages/tags/very-cool-group/

Just need to do tests again :P

@vale ibex this is an uhhh interesting tidbit, the static builds are still failing occasionally because... they time out while talking to the site

We might want to extend the timeout on that step as well

[gritty wind]

I'm still getting that occasional infraction test failure, and it's always due to the ID being off by a bit. It seems that the difference is not really consistent.

So far I've had at least three different assertions fail: line 103, 104, 105. I'm consistently getting failures now (8 times in a row so far with no fix), and I have no idea why.

Environment:

DATABASE_URL=postgres://pysite:pysite@localhost:7777/pysite
METRICITY_DB_URL=postgres://pysite:pysite@localhost:7777/metricity
SECRET_KEY=suitable-for-development-only
STATIC_ROOT=./staticfiles
DEBUG=1

Running on host with command: poetry run task test

I managed to get a debugger into it

It seems the issue is... site#741

[dusky shore]

GitHub

[site] #741 Remove IP Specification From Compose Ports

[gritty wind]

If your database access is quick enough, the insertion times will all be the same, which will create just a completely random order

I'll try to explicitly specify the time so this isn't so fragile

That seems to have fixed it, I'll push it in a separate PR

site#764

[dusky shore]

GitHub

[site] #764 Explicitly Specify Infraction Time In Tests

[fossil veldt]

infractions?

my site pr didn't break anything did it

[gritty wind]

Haha no dw this issue is almost as old as the infraction system

Just became a lot more prevalent recently because we sped up database access

(Btw if you haven't recreated your database since site#741 do so, you'll get so much sweet speed)

[dusky shore]

GitHub

[site] #741 Remove IP Specification From Compose Ports

[gritty wind]

Like 5-10x speed improvement 😛

[fossil veldt]

[surreal veldt]

Can anyone help me resolve this error when I run py -m pip install disnake? ``` error: Microsoft Visual C++ 14.0 or greater is required. Get it with "Microsoft C++ Build Tools": https://visualstudio.microsoft.com/visual-cpp-build-tools/
[end of output]

note: This error originates from a subprocess, and is likely not a problem with pip.
ERROR: Failed building wheel for yarl
Failed to build aiohttp frozenlist multidict yarl
ERROR: Could not build wheels for aiohttp, frozenlist, multidict, yarl, which is required to install pyproject.toml-based projects

[outer oasis]

This channel is for questions relating to this servers open source projects, not for general help
That being said, that error is one of the simpler ones, you just need the build tools, the error message already explains what you need do to [do to get them]

[surreal veldt]

oh okay sorry

[outer oasis]

no worries

[static canyon]

How do I only run one specific test? I know about retest but that doesn't work once the test passes (I'm editing the test so need to keep rerunning)

The test I'm wanting to run is tests.bot.rules.test_mentions.TestMentions.test_ignore_reply_mentions, where test_mentions.py is the file, TestMentions is the class, and test_ignore_reply_mentions is the test function

[static canyon]

For now I'm doing poetry run task test tests/bot/rules/test_mentions.py which runs all the tests in that file (which is much better than all tests)

[gritty wind]

Bot tests are pretty quick in general, but to run a specific test:
path/to/file.py -k <name> which checks if name is in the test name

Or press the run button beside the test in pycharm, and experience the best test UI known to man

[static canyon]

Ah yes, beautiful:```
============================= test session starts =============================
collecting ... ERROR: not found: C:\Users\tizzy\bot\tests\bot\rules\test_mentions.py::TestMentions::test_ignore_reply_mentions

[gritty wind]

Haha is this from the command or pycharm

[static canyon]

PyCharm lol

Even running on the class doesn't work

[gritty wind]

Configure the default test template to use the project root as a workdir

Then the run button should work

[static canyon]

Right, yeah

That works

Thanks

[gritty wind]

Awesome

[fossil veldt]

it is possible to only run some tests in bot?

whenever I try it just fails on the mocking thing or something

[gritty wind]

How are you running

[fossil veldt]

uh I just clicked the pycharm run next to the function

[static canyon]

Read ~5 messages up (I just had this issue too)

Configure the workdir to the root

[gritty wind]

Do it on the run template so it’s configured for all future runs

[fossil veldt]

so this template?

[gritty wind]

Pycharm sometimes picks up unittest as well, so I do it on those two and autodetect

[fossil veldt]

oh right we use unittest

oops

I'm too used to everything being pytest

[fossil veldt]

wait by root you mean the bot module or actually the root?

[gritty wind]

The folder where the pyproject file is

[static canyon]

Actual root

[fossil veldt]

hm...

[static canyon]

I.e. in my case it's C:\Users\tizzy\bot

[fossil veldt]

Ran 0 tests in 0.000s

OK

[static canyon]

What's the function you're testing?

[fossil veldt]

ext/moderation/infraction/test_utils.py

one of the classes there

[static canyon]

Works fine for me 🤷

Testing started at 18:26 ...
Launching pytest with arguments tests/bot/exts/moderation/infraction/test_utils.py::ModerationUtilsTests --no-header --no-summary -q in C:\Users\tizzy\bot

============================= test session starts =============================
collecting ... collected 5 items

tests/bot/exts/moderation/infraction/test_utils.py::ModerationUtilsTests::test_get_active_infraction 
tests/bot/exts/moderation/infraction/test_utils.py::ModerationUtilsTests::test_notify_pardon 
tests/bot/exts/moderation/infraction/test_utils.py::ModerationUtilsTests::test_post_user 
tests/bot/exts/moderation/infraction/test_utils.py::ModerationUtilsTests::test_send_infraction_embed 
tests/bot/exts/moderation/infraction/test_utils.py::ModerationUtilsTests::test_send_private_embed 

================== 4 passed, 1 skipped, 5 warnings in 3.20s ===================

Process finished with exit code 0
PASSED [ 20%]PASSED [ 40%]PASSED [ 60%]SKIPPED (Current time needs to be patched so infraction duration is correct.) [ 80%]
Skipped: Current time needs to be patched so infraction duration is correct.
PASSED [100%]```

[fossil veldt]

[static canyon]

What's the full log?

Oh and do the test functions start with test_? Otherwise unittest doesn't detect that it's a test function

[fossil veldt]

hm

they do though

If I right click the entire file it works

[static canyon]

^

[fossil veldt]

/Users/ionite/Library/Caches/pypoetry/virtualenvs/bot-xxxVP0Vd-py3.9/bin/python /Applications/PyCharm.app/Contents/plugins/python/helpers/pycharm/_jb_pytest_runner.py --path /Users/ionite/repos/Py-Dis/bot/bot/exts/moderation/infraction 
Testing started at 1:28 PM ...
Launching pytest with arguments /Users/ionite/repos/Py-Dis/bot/bot/exts/moderation/infraction --no-header --no-summary -q in /Users/ionite/repos/Py-Dis/bot

============================= test session starts ==============================
collecting ... collected 0 items

============================ no tests ran in 0.00s =============================

Process finished with exit code 5

Empty suite

[static canyon]

Launching pytest with arguments /Users/ionite/repos/Py-Dis/bot/bot/exts/moderation/infraction`

That second bot should be tests

I think you're in the wrong dir?

tests should be somewhere in that fp

Probably should be /Py-Dis/bot/tests/bot/exts/moderation/infraction

[fossil veldt]

hm..

when I run the file it does

Testing started at 1:32 PM ...
Launching pytest with arguments /Users/ionite/repos/Py-Dis/bot/tests/bot/exts/moderation/infraction/test_utils.py --no-header --no-summary -q in /Users/ionite/repos/Py-Dis/bot

============================= test session starts ==============================
collecting ... collected 9 items

which yeah I guess that one has the tests path

[static canyon]

Yeah

I guess double check your config or something

[gritty wind]

Are you clicking the run button in the tests dir

Send a screenshot of your run configuration

(Not the template one)

[fossil veldt]

file (runs correctly)

class (no collected)

[static canyon]

That module name seems to be wrong

[gritty wind]

The target in the second screenshot doesn’t look right

Is it autogenerated when you press the run button

[fossil veldt]

yeah

[static canyon]

I mean I guess for now you can manually fix it by correcting the fp

No clue why it's not generating correctly in the first place though

[fossil veldt]

yeah weird

whatever though I'm happy with 1 file working

I was just waiting for all unit tests to run trying to test 1 thing

so this is considerably less painful 🥲

[gritty wind]

Try deleting it and creating again maybe

Not really sure

[cold island]

I need to rebase filters again? 🥲

I'll just merge main lol

[vale ibex]

lol

[static canyon]

Rip my 5 open PRs lol

How do you update the env to use 3.10?

[vale ibex]

I deleted and remade mine

[static canyon]

Fair enough

[vale ibex]

Not sure if there's a way to update

[austere hornet]

Will Sir Lancebot run on Python 3.10 eventually as well?

[vale ibex]

yup

[austere hornet]

Cool

[fossil veldt]

So uh, bot embeds seem to be broken?

[lyric temple]

!d os.wait

[stable mountain]

os.wait

b'py\n\nos.wait()\nWait for completion of a child process, and return a tuple containing its pid and exit status indication: a 16-bit number, whose low byte is the signal number that killed the process, and whose high byte is the exit status (if the signal number is zero); the high bit of the low byte is set if a core file was produced.\n\nwaitstatus_to_exitcode() can be used to convert the exit status into an exit code.\n\nAvailability: Unix.\n\nSee also\n\nwaitpid() can be used to wait for the completion of a specific child process and has more options.'

[brazen charm]

!d str

[stable mountain]

str

b'py\n\nclass str(object=\'\')``````py\n\nclass str(object=b\'\', encoding=\'utf-8\', errors=\'strict\')\nReturn a string version of object. If object is not provided, returns the empty string. Otherwise, the behavior of str() depends on whether encoding or errors is given, as follows.\n\nIf neither encoding nor errors is given, str(object) returns type(object).__str__(object), which is the \xe2\x80\x9cinformal\xe2\x80\x9d or nicely printable string representation of object. For string objects, this is the string itself. If object does not have a __str__() method, then str() falls back to returning repr(object).'

[lyric temple]

3.10 side-effect? I didn't think anything changed with strings/bytestrings

[brazen charm]

I think it may just be the cache

[vale ibex]

!d r

[stable mountain]

Inventories refreshed

[brazen charm]

ah the results from redis are broken

[vale ibex]

Yea, looks like redis is sending raw bytes and the bot isn't encoding

This wasn't happening in dev

so might be a migration thing

!d str

[stable mountain]

str

b'py\n\nclass str(object=\'\')``````py\n\nclass str(object=b\'\', encoding=\'utf-8\', errors=\'strict\')\nReturn a string version of object. If object is not provided, returns the empty string. Otherwise, the behavior of str() depends on whether encoding or errors is given, as follows.\n\nIf neither encoding nor errors is given, str(object) returns type(object).__str__(object), which is the \xe2\x80\x9cinformal\xe2\x80\x9d or nicely printable string representation of object. For string objects, this is the string itself. If object does not have a __str__() method, then str() falls back to returning repr(object).'

[vale ibex]

kk gonna drop the doc cache

[brazen charm]

the things in the cache should be fine, it's just returning the bytes for some reason

[fossil veldt]

we removed encoding='utf8' in the 3.10 merge?

[vale ibex]

Yea, it's a global setting now

but might actually cause an issue if they're bytes in the cache on upgrade

[fossil veldt]

oh so it caches strings now?

[brazen charm]

I don't think redis differentiates that

[vale ibex]

runs alright on main in dev

[brazen charm]

second invocations too?

[vale ibex]

Ah, nope

[slate gorge]

!d threading.Thread.join 🔨 blame on redis?

[stable mountain]

threading.Thread.join

b'py\n\njoin(timeout=None)\nWait until the thread terminates. This blocks the calling thread until the thread whose join() method is called terminates \xe2\x80\x93 either normally or through an unhandled exception \xe2\x80\x93 or until the optional timeout occurs.\n\nWhen the timeout argument is present and not None, it should be a floating point number specifying a timeout for the operation in seconds (or fractions thereof). As join() always returns None, you must call is_alive() after join() to decide whether a timeout happened \xe2\x80\x93 if the thread is still alive, the join() call timed out.\n\nWhen the timeout argument is not present or None, the operation will block until the thread terminates.\n\nA thread can be join()ed many times.'

[vale ibex]

bot#2249

[dusky shore]

GitHub

[bot] #2249 Globally decode binary responses from redis to strings

[gritty wind]

Hey I want to document this in the async-rediscache guide and aioredis one, could you summarize where the change is exactly

Looking at the diff it seems to be something we removed?

I don’t entirely understand

[vale ibex]

aiordis <1.3.X did this by default

[gritty wind]

Gotcha, thanks

[vale ibex]

with the move to redis-py, it returns bytes by default now instead

well, iirc string in used to be string out

but now it always returns bytes, regardless of how you put it in there

[gritty wind]

I’ll investigate tomorrow too and try to provide the info

[fossil veldt]

main runs on 3.10 only now right?

not compat with 3.9?

[gritty wind]

Correct

[vale ibex]

you could probably run it on 3.9 if you really wanted to

but you'd need to have build tools installed

so probably better to just use 3.10 :P

[gritty wind]

bot-core will refuse to install so you’d have to bump down its version

[fossil veldt]

what should I do if my PR was forked when it was 3.9?

[gritty wind]

Rebase

Or merge main

[vale ibex]

rebase/merge main, and then remake your venv

[cold island]

No more Union and Optional

[fossil veldt]

hm?

those weren't deprecated though were they?

[cold island]

No need for them I mean 😄

[fossil veldt]

oh also um

do you think I can add a fix for bot#2247 in my infraction duration PR?

[dusky shore]

GitHub

[bot] #2247 Modify bot's automute messages to indicate that moderators role has already been pinged

[cold island]

wdym

[fossil veldt]

since if I make another fork there will a large amount of conflicts depending on which one gets merged first due to the extensive argument name changes in the infraction duration PR

[cold island]

Sorry, I don't follow

How is the issue related?

It's a string change in the antispam cog

[fossil veldt]

well, they're not really I guess

nvm >_>

[fossil veldt]

well it still ends up as an infraction

[cold island]

yeah but it doesn't conflict with your PR

[fossil veldt]

ah, were you thinking it would just be a change in the reason for the auto mute?

[cold island]

yep

[fossil veldt]

I guess that would work, I was thinking of something more complex in recognizing auto-mutes on the infraction side

if it's just adding the moderators have been alerted to the actual reason string then it shouldn't have any conflicts

[gritty wind]

Was optional replaced? What’s the new syntax

[fossil veldt]

just union none I guess

-> str | None

[gritty wind]

Hmm I think I’ll stick with optional lol

[fossil veldt]

yeah sometimes it's actually longer than optional

[gritty wind]

Union and optional were marked as changed rather than deprecated

The 3.9 changes were deprecations tho

[fossil veldt]

I guess cuz | ultimately is just sugar for Union?

and Optional was sugar for Union[x, None] anyways

[gritty wind]

It is, but by that same argument generics wouldn’t have deprecated their alternatives

[fossil veldt]

wouldn't be a python release if it didn't break backward compatibility with something

[gritty wind]

It didn’t break backward compat haha

We can still use the same types we’ve been using since 3.7

Just there are better ways now

[fossil veldt]

I probably had like 7 ML related libraries broken by the collections.abc switch in 3.8

[gritty wind]

Haha

[grave plank]

Anyone want to take challenge 15 hrs of study ?

🤔

[outer oasis]

!rule 9

[stable mountain]

Rules

9. Do not offer or ask for paid work of any kind.

[grave plank]

I am not offering

ok

[static canyon]

Installing dependencies from lock file

Package operations: 91 installs, 0 updates, 0 removals

  • Installing pyparsing (3.0.9)
  • Installing wrapt (1.14.1)

  JSONDecodeError

  Expecting value: line 1 column 1 (char 0)

  at ~\AppData\Local\Programs\Python\Python39\lib\json\decoder.py:355 in raw_decode

[static canyon]

I keep getting this error when trying to setup the 3.10 poetry env

It also says \...\Python39\... so not sure why it seems to be using 3.9 even though it's set to use 3.10 as the base

[fossil veldt]

Wait your poetry is a bat? Huh

Mine is an exe?

I think your pycharm env is not synced with your shell env?

When I switched to 3.10 I removed the old env, selected no interpreter in pycharm, did poetry env use 3.10 poetry install, and then selected the new poetry env by adding existing in pycharm

[static canyon]

PS C:\Users\tizzy\bot> poetry env use 3.10
Traceback (most recent call last):
  File "C:\Users\tizzy\.poetry\bin\poetry", line 17, in <module>
    from poetry.console import main                                                   
  File "C:\Users\tizzy\.poetry\lib\poetry\console\__init__.py", line 1, in <module>   
    from .application import Application                                              
  File "C:\Users\tizzy\.poetry\lib\poetry\console\application.py", line 3, in <module>
    from cleo import Application as BaseApplication                                   
ModuleNotFoundError: No module named 'cleo'    ```

My poetry got broken at some point in the process

All poetry commands give that

[fossil veldt]

Uh...

Reinstall poetry maybe? Did you use the new script

https://python-poetry.org/docs/master/#installing-with-the-official-installer

get-poetry.py is deprecated if you used that

[static canyon]

That works for the install but still same error

[fossil veldt]

For any poetry command? Even outside pycharm?

[static canyon]

Nvm, needed to restart the terminal

[static canyon]

PS C:\Users\tizzy\bot> poetry env use 3.10
'python3.10' is not recognized as an internal or external command,
operable program or batch file.

  EnvCommandError

  Command python3.10 -c "import sys; print('.'.join([str(s) for s in sys.version_info[:3]]))" errored with the following return code 1, and output:

And I still have this```
PS C:\Users\tizzy\bot> poetry install
Installing dependencies from lock file

Package operations: 91 installs, 0 updates, 0 removals

• Installing pyparsing (3.0.9)
• Installing wrapt (1.14.1)

JSONDecodeError

Expecting value: line 1 column 1 (char 0)

at ~\AppData\Local\Programs\Python\Python39\lib\json\decoder.py:355 in raw_decode
351│ """
352│ try:
353│ obj, end = self.scan_once(s, idx)
354│ except StopIteration as err:
→ 355│ raise JSONDecodeError("Expecting value", s, err.value) from None
356│ return obj, end
357│

[fossil veldt]

Uh, okay try setting it to the full path of your python executable

The documentation says you can just use minor version if python is in $PATH

Which is what I've always done, weird it doesn't work for yours

[static canyon]

Yeah, it is in path

Maybe it's because I have both 3.9 and 3.10 in path

There's also this ig PS C:\Users\tizzy> python -V Python 3.9.5 PS C:\Users\tizzy> py -V Python 3.10.0

[fossil veldt]

Is that a windows store alias?

How did you end up with python

[static canyon]

From the path ig

[static canyon]

PS C:\Users\tizzy> gcm python

CommandType     Name                                               Version    Source
-----------     ----                                               -------    ------
Application     python.exe                                         3.9.515... C:\Users\tizzy\AppData\Local\Programs\...


PS C:\Users\tizzy> gcm py

CommandType     Name                                               Version    Source
-----------     ----                                               -------    ------
Application     py.exe                                             3.10.15... C:\Windows\py.exe```

That shows where they're installed ig

[fossil veldt]

Was that from windows store?

[static canyon]

Idk tbh

I doubt it

[outer oasis]

(Get-Command python).Path

Please

[static canyon]

I apparently don't have these installed so it wouldn't be from windows store

[outer oasis]

Also what's the one to see if py recognizes your versions?
py -0?

[static canyon]

PS C:\Users\tizzy> (gcm python).Path
C:\Users\tizzy\AppData\Local\Programs\Python\Python39\python.exe```

[static canyon]

Yeah, and that shows 3.9+3.10

[fossil veldt]

py --list also shows both?

[static canyon]

Yeah

With 3.10 being the selected one

[outer oasis]

AppData
So you unchecked "Install Python for all users"?
Yeah, might just need to give the specific path to Poetry

[fossil veldt]

You should be able to just poetry env use 3.10 then

Or maybe "3.10"?

[static canyon]

Yeah I most probably did

[static canyon]

Same error as above

[static canyon]

Specifying the entire fp seems to have worked?```
PS C:\Users\tizzy\bot> poetry env use C:\Users\tizzy\AppData\Local\Programs\Python\Python310\python.exe
Using virtualenv: C:\Users\tizzy\AppData\Local\pypoetry\Cache\virtualenvs\bot-6mLjCNeK-py3.10

[outer oasis]

Looks like it

[static canyon]

poetry install still fails though, and is using 3.9

[fossil veldt]

Try

poetry config experimental.new-installer false

[static canyon]

No difference

[outer oasis]

They redid their package installer?

[fossil veldt]

From python-poetry/poetry#4210

[dusky shore]

GitHub

[poetry] #4210 JSONDecodeError on Python 3.10

[static canyon]

Ah wait no that works for poetry install

Thanks

Things are installing

[fossil veldt]

But in 3.9?

[static canyon]

No clue 🤷

It doesn't say what py version

Just "Installing dependencies from lock file"

[fossil veldt]

Are you using pyenv btw?

[static canyon]

🤷

No clue

[fossil veldt]

I think if you do poetry --version it should show the python version?

[static canyon]

Ah nice, back to this```
PS C:\Users\tizzy\bot> poetry --version
Traceback (most recent call last):
File "C:\Users\tizzy.poetry\bin\poetry", line 17, in <module>
from poetry.console import main
File "C:\Users\tizzy.poetry\lib\poetry\console_init_.py", line 1, in <module>
from .application import Application
File "C:\Users\tizzy.poetry\lib\poetry\console\application.py", line 3, in <module>
from cleo import Application as BaseApplication
ModuleNotFoundError: No module named 'cleo'

[fossil veldt]

Eh

Wtf

[fossil veldt]

Uh

So apparently...

Latest version of poetry is just missing dependencies for python 3.10.0 to 3.10.1 versions

pip install cleo tomlkit poetry.core requests cachecontrol cachy html5lib pkginfo virtualenv lockfile

[outer oasis]

I'm sorry what

[fossil veldt]

Idk

[static canyon]

It's no longer giving that error```
PS C:\Users\tizzy\bot> poetry --version
Poetry version 1.1.7

[fossil veldt]

Huh, did you do the pip install?

[static canyon]

Yeah

[fossil veldt]

Interesting

I guess it's literally missing dependencies then

Wonderful

[static canyon]

But it doesn't show python version so still no clue if it's using 3.10

[fossil veldt]

Just do py --version?

[static canyon]

However this does```
PS C:\Users\tizzy\bot> pip -V
pip 21.1.2 from C:\Users\tizzy\bot\venv\lib\site-packages\pip (python 3.10)

So I'm assuming it is 3.10

[fossil veldt]

I like how this is casually a thing buried in a closed issue though 😂

[static canyon]

Ah, wonderful ModuleNotFoundError: No module named 'botcore'

Guess I need to do poetry install again

[fossil veldt]

Mhm yeah

Should be fine this time... Maybe?

[static canyon]

Ah, beautiful```
PS C:\Users\tizzy\bot> poetry install
Installing dependencies from lock file

Package operations: 0 installs, 0 updates, 0 removals

Installing the current project: bot (1.0.0)

[outer oasis]

Yeah, I'm still not even seeing what actually caused it

[static canyon]

I think PyCharm is using a different venv to my PyCharm terminal

[fossil veldt]

Yeah that happens

Remove your current one and delete (-) it on the interpreter list

And then click add new interpreter

Go to poetry, and do add existing interpreter

And the folder should show up...?

[static canyon]

So I did it a roundabout way but I think it's good

Ugh ModuleNotFoundError: No module named 'async_rediscache'

But if I poetry install then it says that everything's already installed

[fossil veldt]

poetry update -vv ?

[static canyon]

So I've just done this, but now PyCharm seems to be stuck on the scan

The folder did show up this time but it's just seemingly infinitely scanning

And for some reason it has two tasks doing that

[fossil veldt]

Uh

Set no interpreter, quit pycharm, then start pycharm and put it in the current interpreter...?

[static canyon]

So now there's only one, but still seems stuck

[fossil veldt]

Uh..

But the shell is working? You're in the correct poetry version?

The shell poetry env name matches the pycharm one?

[static canyon]

Yep

No clue if this has anything to do with it

00:03    Plugin error: Plugin 'Poetry' (version '1.1.4') is not compatible with the current version of the IDE, because it requires build 211.* or older but the current build is PC-213.6777.50

[fossil veldt]

Tbh I would just close the project, then delete the entire .idea folder in it, then restart and let it reinitialize

[gritty wind]

Uninstall the poetry plugin

It’s no longer maintained since it’s built in

[static canyon]

Right, done

Just restarted PyCharm and still the same issue 🤷

[static canyon]

Gonna try this

Didn't work & now I can't even use the project lol

I have to wait for it to finish scanning, which it never does

[gritty wind]

It might be the case that your poetry env is corrupted. I’d delete the environment, update python to 10.6, and recreate the pycharm project as well

[static canyon]

Wdym by recreate the pycharm project?

[gritty wind]

What you did here

[static canyon]

Right

I guess I'll try that

Apparently uninstalling python takes a really long time lol

Right, all done

Now we try again

It's.... still stuck

I've created an entirely new poetry env and still the exact same issue

It just forever stays on this

[fossil veldt]

you reinstalled poetry already right?

Maybe delete the install poetry folder in Local/and then reinstall, not sure

[static canyon]

Yep

[static canyon]

Already done this numerous times

Left it running overnight in case it was just talking an absurd amount of time but it's still not finished

[gritty wind]

At this point we’ve reinstalled everything, so I guess reinstall windows? No that’s a joke

Have you tried the builtin “fix IDE” option

[static canyon]

I guess I can try that

The "restart option" section of that seems to have actually fixed that lmao

Even though I've restarted the project myself numerous times

But the libraries pycharm shows in venv still don't match pip list

[static canyon]

So this fixed the infinite "Scanning installed packages" but stuff still doesn't match

[gritty wind]

Which environment is selected in settings

[static canyon]

Poetry (bot) [3.10.6]

[gritty wind]

Hm, try importing a package in the python interpreter?

Or remove and add it again (the environment)

[static canyon]

>>> import botcore
botcore
<module 'botcore' from 'C:\\Users\\tizzy\\AppData\\Local\\pypoetry\\Cache\\virtualenvs\\bot-6mLjCNeK-py3.10\\lib\\site-packages\\botcore\\__init__.py'>```

venv folder still hasn't updated though

[static canyon]

I've done that already numerous times

Currently when I run tests I getpy from async_rediscache import RedisCache E ModuleNotFoundError: No module named 'async_rediscache'

Which oddly looks like it might be correct since there's not async_rediscache folder in site-packages

Huh

After manually installing async_rediscache the test works 🤷

So for some reason poetry install didn't install that

arg = <MagicMock name='mock.created_at' id='2237042899456'>

E               TypeError: Cannot parse single argument of type <class 'unittest.mock.MagicMock'>.

..\AppData\Local\pypoetry\Cache\virtualenvs\bot-6mLjCNeK-py3.10\lib\site-packages\arrow\factory.py:267: TypeError```for some reason `arrow.get(mockmessage.created_at)` is giving this error -- the actual code being `arrow.get(incident.created_at)`

[fossil veldt]

this is main?

[static canyon]

The test is from main, yeah

I'm on a different branch though

incident-archive-msg-improvements

[fossil veldt]

did you pull the stuff from the 3.10 PR yet?

or is that a new branch

[static canyon]

Yeah, that's what broke the tests

And I've spent ~3hrs trying to fix them with the biggest issue being the poetry thing from yesterday that still isn't fixed

[fossil veldt]

well it appears to fail on github as well, so I don't think it's related to your poetry?

[static canyon]

Yeah, poetry is something else

[fossil veldt]

yeah your env is like the most cursed I've seen it

[static canyon]

I still have this running infinitely, and PyCharm's venv doesn't match the linked /pypoetry/Cache/virtualenv/bot-.../

[fossil veldt]

I usually have to do a bit of fiddling every now and then but nothing like yours 😅

[fossil veldt]

if you just use pycharm to make a brand new unrelated poetry project for 3.10.6

does it work normally?

[static canyon]

Testing now

Yeah, it does

[fossil veldt]

[static canyon]

🤷

[fossil veldt]

the env you're using for bot now is created via shell right?

or through pycharm

[static canyon]

Via poetry install iirc

[fossil veldt]

hm

[static canyon]

It was created through pycharm

[fossil veldt]

oh?

[static canyon]

From the "Add interpreter"

[fossil veldt]

oh but the folder was already there?

[static canyon]

🤷

[fossil veldt]

like was it from existing here or new

[static canyon]

I think I originally did a new one, then set interpreter to nothing and did existing

[fossil veldt]

you could try deleting the current one and letting pycharm make a new one?

[static canyon]

So the top option?

[fossil veldt]

yeah

and after you delete the venv the existing should be empty

(restart pycharm after deleting)

I mean if your new projects are working fine then it's just something bork with this specific venv?

[static canyon]

This yeah?

[fossil veldt]

yeah, if you click existing its empty right?

[static canyon]

Yes

[fossil veldt]

okay yeah

[static canyon]

It's installing everything atm

[fossil veldt]

hopefully it works and you can escape the pycharm poetry env mismatch purgatory

[static canyon]

I mean I've done this a few times now so I doubt it

[static canyon]

Yeah, still not matching

[fossil veldt]

wtf

[static canyon]

From the looks of it PyCharm is using its own env instead of the poetry one still

[fossil veldt]

wait so

in the terminal the poetry env name is not the pycharm one?

[static canyon]

I.e when I right-click venv and select "open in explorer" it opens C:\Users\tizzy\bot\venv\Lib\site-packages and not ...\pypoetry\Cache\virtualenv\bot-....\Lib\site-packages

[fossil veldt]

wait do you have an in-project venv?

[static canyon]

I've just deleted that for now

Cause it does show the poetry stuff and external

[fossil veldt]

but what created that, pycharm or manual

the in-project one

[static canyon]

Who knows 🤷

PyCharm ig because it wasn't me

[fossil veldt]

wait you reinstalled poetry earlier right

[static canyon]

Yeah

[fossil veldt]

can you run poetry config --list

[static canyon]

    from cleo import Application as BaseApplication
ModuleNotFoundError: No module named 'cleo'
```wooo

[fossil veldt]

lmao what

[static canyon]

Time to reinstall the dependencies again lmao

[fossil veldt]

aren't you on 3.10.6 now?

[static canyon]

Because why would a project come installed with its dependencies 🤷

[fossil veldt]

that dependency issue was only for 3.10.0-3.10.1 iirc

[static canyon]

Who tf knows lol

PS C:\Users\tizzy\bot> pip -V
pip 21.1.2 from C:\Users\tizzy\AppData\Local\pypoetry\Cache\virtualenvs\bot-6mLjCNeK-py3.10\lib\site-packages\pip (python 3.10)

I don't even have 3.10 installed 🤷

[fossil veldt]

ah

[static canyon]

So I'm assuming that just doesn't include the minor version

[fossil veldt]

can you cd to C:\Users\tizzy\AppData\Local\pypoetry

is there a poetry executable in there?

[static canyon]

No

[fossil veldt]

hm

[fossil veldt]

wait what do you mean

but you have 3.10.6 installed?

[static canyon]

I have 3.10.6 installed, not 3.10

I did have 3.10 but Scaleios said to update to 3.10.6 and try, so I did

Just gonna manually install the poetry dependencies

[fossil veldt]

that's the default install for the new poetry though...

[static canyon]

PS C:\Users\tizzy\bot> poetry config --list
cache-dir = "C:\\Users\\tizzy\\AppData\\Local\\pypoetry\\Cache"
experimental.new-installer = false
installer.parallel = true
virtualenvs.create = true
virtualenvs.in-project = null
virtualenvs.path = "{cache-dir}\\virtualenvs"  # C:\Users\tizzy\AppData\Local\pypoetry\Cache\virtualenvs

[fossil veldt]

if you do (Get-Command poetry).Path?

my working theory is your pycharm and shell are trying to use completely different versions of poetry

[static canyon]

PS C:\Users\tizzy\bot> (gcm poetry).Path
C:\Users\tizzy\.poetry\bin\poetry.bat

[fossil veldt]

you installed with this earlier right?

[static canyon]

Yeah

[fossil veldt]

...

poetry self update?

[static canyon]

I already had the tizzy\.poetry though I think not sure

[fossil veldt]

I dunno what's going on anymore

something is broken with your poetry though

maybe just delete the whole .poetry folder before installing again, I remember I saw that on a few comments on the poetry issue

[static canyon]

PS C:\Users\tizzy\bot> poetry self update
Updating to 1.1.14

  PermissionError

  [WinError 5] Access is denied: 'C:\\Users\\tizzy\\.poetry\\lib\\poetry\\_vendor\\py3.9\\msgpack\\_cmsgpack.cp39-win_amd64.pyd'

[fossil veldt]

nice

oh its back to 3.9 as well

😂

[static canyon]

Yeah lol

[fossil veldt]

yeah get rid of that whole .poetry directory and install again

completely cursed

also close pycharm when you do that

and reopen pycharm after you reinstall

it only reindexes poetry paths on startup

[static canyon]

So I've deleted .poetry

[static canyon]

Now this?

[fossil veldt]

yeah

[static canyon]

Before installing I'm gonna uninstall 3.9 actually

So it doesn't use that

[fossil veldt]

also for good measure maybe disable the windows store alias as well

[fossil veldt]

it's gonna somehow resurrect a zombie version of 3.9

[static canyon]

I really hope not 😢

3.9 is gone

This installer is deprecated. Poetry versions installed using this script will not be able to use 'self update' command to upgrade to 1.2.0a1 or later. It is recommended to use https://install.python-poetry.org instead. Instructions are available at https://python-poetry.org/docs/#installation```

Ah nice

PS C:\Users\tizzy> (Invoke-WebRequest -Uri https://install.python-poetry.org -UseBasicParsing).Content | py -
Retrieving Poetry metadata

The latest version (1.1.14) is already installed.```wtf lol

I deleted the .poetry but apparently it's still installed

Yeah poetry isn't a valid command in the terminal 🤷

Wtf is going on

Right, I had to delete pypoetry from appdata/local and appdata/roaming

Okay, it's reinstalled PS C:\Users\tizzy> poetry --version Poetry version 1.1.14

Now to setup the env in PyCharm

[fossil veldt]

Okay so if you just do self update does it error

[static canyon]

I assume you mean poetry self update, and that says I'm using the latest version

[fossil veldt]

Okay... That's an improvement 👀

[fossil veldt]

Yeah the path makes sense now

I always had an exe

I don't know how you have a .bat as poetry before

[static canyon]

🤷

I also don't have tizzy/.poetry anymore 🤷

It still seems to be hanging 😩

[fossil veldt]

[static canyon]

Maybe not actually

I think it's just taking its time

Since the text underneath is updating

[fossil veldt]

Okay..

Yeah it might take a while to reindex python completely

[static canyon]

Yep, it's working 😄

[fossil veldt]

🎉

[static canyon]

And it installed everything properly (last time it didn't installed async_rediscache)

[fossil veldt]

Well check if your unit tests work I guess

And if it's using 3.10 and dependencies correctly

[static canyon]

Nope 😢

Exact same error

[fossil veldt]

Was it supposed to work?

[static canyon]

E               TypeError: Cannot parse single argument of type <class 'unittest.mock.MagicMock'>.

..\AppData\Local\pypoetry\Cache\virtualenvs\bot-6mLjCNeK-py3.10\lib\site-packages\arrow\factory.py:267: TypeError```

self = <arrow.factory.ArrowFactory object at 0x0000022F93228B80>
args = (<MagicMock name='mock.created_at' id='2403377807024'>,), kwargs = {}
arg_count = 1, locale = 'en-us', tz = None, normalize_whitespace = False
arg = <MagicMock name='mock.created_at' id='2403377807024'>```

It seems to be passing a magicmock instead of that datetime from .created_at

This is the full log: https://paste.pythondiscord.com/raturebope

[static canyon]

I don't think so, but I equally have no clue why it's not

[fossil veldt]

Where was your pr again?

I saw it earlier and now GitHub won't show it

[static canyon]

bot#2031

[dusky shore]

GitHub

[bot] #2031 Incident archive msg improvements

[static canyon]

The github one is up 2 date, except from stuff I did trying to fix the tests that didn't work

Should probably undo the edits actually since they don't work

[fossil veldt]

Oh you didn't even touch the failing tests part in diff

Weird

[static canyon]

Yeah I know

That's why I'm so confused

[fossil veldt]

Also this thing is broken too? Lol

[static canyon]

From what I can tell this should never have been merged into main (the 3.10 PR)

But obvs the tests have to work first for that to happen so 🤷

[fossil veldt]

The tests are working on the exact main version though I think

[vale ibex]

The test seems like it is working as intended. You have updated make_embed to use incident.created_at

but in the other tests, you don't update incident.created at to be a datetime

incident being from the message

[static canyon]

But I have for the one I'm running

    async def test_make_embed_content(self):
        """Incident content appears as embed description."""
        current_time = datetime.datetime(2022, 1, 1, 0, 0, 0, tzinfo=datetime.timezone.utc)
        incident = MockMessage(content="this is an incident", created_at=current_time)

        reported_timestamp = discord_timestamp(current_time)
        relative_timestamp = discord_timestamp(current_time, TimestampFormats.RELATIVE)

        embed, file = await incidents.make_embed(incident, incidents.Signal.ACTIONED, MockMember())

        self.assertEqual(
            f"{incident.content}\n\n*Reported {reported_timestamp} ({relative_timestamp}).*",
            embed.description
        )```

[vale ibex]

What is the test that's failing?

[static canyon]

Right I'm dumb

That one's passing

[vale ibex]

yup :D

[static canyon]

     async def setUp(self):
        self.current_time = datetime.datetime(2022, 1, 1, tzinfo=datetime.timezone.utc)

    async def test_make_embed_actioned(self):
        """Embed is coloured green and footer contains 'Actioned' when `outcome=Signal.ACTIONED`."""
        embed, file = await incidents.make_embed(MockMessage(created_at=self.current_time), incidents.Signal.ACTIONED, MockMember())

        self.assertEqual(embed.colour.value, Colours.soft_green)
        self.assertIn("Actioned", embed.footer.text)```

What was the setup function?

Nvm, it's not async

[vale ibex]

the setup function gets called before every test

there is an async setup if you need one

[static canyon]

Nah I don't

I just need to add created_at=self.current_time to all of the MockMessage()

[vale ibex]

👍

yoiu could create the whole message in there

if it's the same code in every test

[static canyon]

Nah it's not

It changes

There's 3 different messages

I guess I could assign each one but that seems overkill

[vale ibex]

Yea, setup is just there to remove code duplication from your tests

if there isn't duplication then it doesn't need to be there

[static canyon]

The current_time is needed in more than one test class, so would it be worth making it a constant in the file?

[vale ibex]

There is a module wide setup function that exists, similar to how there are test specific setup functions

[static canyon]

So just def setUp() but outside the classes?

[vale ibex]

I think it's setUpModule

[static canyon]

Is there a preference to whether it goes at the top or bottom of the file?

[vale ibex]

but I'm also not sure how that works with variable scoping

so it might just be better to create a screaming case constant

[static canyon]

Yeah

Okay

Now moderation tests about the rediscache are failing

FAILED tests/bot/exts/moderation/test_silence.py::UnsilenceTests::test_unsilence_role - async_rediscache.session.RedisSessionNotInitialized: the redis session has not been initialized ...

FAILED tests/bot/exts/moderation/test_silence.py::UnsilenceTests::test_skipped_already_unsilenced - async_rediscache.session.RedisSessionNotInitialized: the redis session has not been ...```

How do I fix that? @vale ibex

[vale ibex]

Uhhh, what version of redis-py is installed?

or async-rediscache

This should be initialised by the super class

Could you try doing poetry install --remove-untracked

and then running tests again

[static canyon]

I'll try that when I'm back on my laptop

[static canyon]

1.0.0rc2

[static canyon]

This didn't change anything

Err

Now I've got even more tests failing (for the same reason)

FAILED tests/bot/exts/moderation/test_silence.py::UnsilenceTests::test_skipped_already_unsilenced - async_rediscache.session.RedisSessionNotInitialized: the redis session has not been ...

FAILED tests/bot/exts/moderation/test_silence.py::UnsilenceTests::test_unsilence_role - async_rediscache.session.RedisSessionNotInitialized: the redis session has not been initialized ...

FAILED tests/bot/exts/moderation/test_silence.py::UnsilenceTests::test_preserved_other_overwrites_voice - async_rediscache.session.RedisSessionNotInitialized: the redis session has not...

FAILED tests/bot/exts/moderation/test_silence.py::UnsilenceTests::test_removed_notifier - async_rediscache.session.RedisSessionNotInitialized: the redis session has not been initialize...

FAILED tests/bot/exts/moderation/test_silence.py::UnsilenceTests::test_restored_overwrites_text - async_rediscache.session.RedisSessionNotInitialized: the redis session has not been in...

FAILED tests/bot/exts/moderation/test_silence.py::UnsilenceTests::test_restored_overwrites_voice - async_rediscache.session.RedisSessionNotInitialized: the redis session has not been i...

FAILED tests/bot/exts/moderation/test_silence.py::UnsilenceTests::test_sent_correct_message - async_rediscache.session.RedisSessionNotInitialized: the redis session has not been initialized ...```

[vale ibex]

just to be sure, could you delete any __pycache__ folders you have ?

[static canyon]

I don't seem to have any

[vale ibex]

hmmm, I have no idea how to troubleshoot this then

it works in CI, and it works on my machine, so I'm not sure why it isn't working in your env

[static canyon]

I swear the env just hates me

I've had so many issues with getting it to 3.10

And now it's finally 3.10 but the tests fail for no good reason

I've not even made any changes to the silence command 🤷

No clue what to do 🤷