@SmokeDetector The user's been in existence for >1 year, but the question's unclear and could be trolling. It's not clear to me that it's intentionally trolling, but I'm leaning that way. If it's a real question, it could be asked much less controversially, clearly, and on a more appropriate site. They have a new question on Physics, which is conspiracy theory-ish, but is them not being able to read a graph, or the graph's caption.
Can we publish the version of ChatExchange we're using to PyPI, under our name? So pip doesn't have to clone and re-install it every time, especially on CI.
[ SmokeDetector | MS ] URL in title, bad NS for domain in body, bad NS for domain in title, bad keyword in body, blacklisted website in body, +5 more: reviverxtry.com/peraglow/ by yorxgort on english.SE
> Pattern-matching product name in body, pattern-matching product name in title, potentially bad keyword in body, potentially bad keyword in title, potentially bad keyword in username ---------- Title - Position 7-16: Pure Skin Title - Pattern-matching product name *Serum Pure Skin* Body - Position 7-16: Pure Skin Post - Pattern-matching product name *Serum Pure Skin* Username - Position 7-16: Pure Skin
!!/test-a works for me: data<a href="https://www.inspiresayan.com/top-5-best-selfie-camera-smartphones-under-10000-rupees/" rel="nofollow noreferrer">.</a>
> Bad ns for domain in answer, linked punctuation in answer ---------- Post - inspiresayan.com NS suspicious dns2.namecheaphosting.com.,dns1.namecheaphosting.com. Body - Position 115-146: rel="nofollow noreferrer">.</a>
weird, it looks like the bad NS "eats" the linked punctuation
if I switch to another domain it works again
!!/test-a works for me: data<a href="https://www.example.com/top-5-best-selfie-camera-smartphones-under-10000-rupees/" rel="nofollow noreferrer">.</a>
@iBug well HTML is the source for the rendering, but you will appreciate the difference between looking at a web page (a rendering) and looking at its source
Makyen originally reported a couple of cases where the HTML version lacks some of the original input, removing some useful signal
but yeah, it would require rework and heavy regression testing to make sure we still catch what we are supposed to catch
... I did some more testing on the linked punctuation post in Charcoal Test, I don't get a response from Smokey when I !!/test-a the full post so I guess it causes an error somewhere
!!/errorlogs reveals a bunch of SSL errors which doesn't make a lot of sense in this context
@iBug regex compilation is supposed to make things faster but in some scenarios doesn't actually ... I would expect a big ole comment to explain if it's done on purpose though
the Python re module had a bug where it was (or still is?) much slower to hand it a compiled object than to just pretend it's a new object and have it find it from its cache
not sure if this affects regex too or what precise versions to look for this in
@Makyen That pattern looks like it's already caught by Bad ns for domain in body and Bad ns for domain in answer; append -force if you really want to do that.
@iBug the regex module has a nifty feature which lets you keep a list of alternatives and refer to it like that; we use it for the cities list and some other places too IIRC
I don't have any timings or actual insights but intuitively, I would guess that the engine would be able to optimize a bit if it knows you are explicitly referring to the same list of alternatives multiple times
also informally (only looking at a fragment of the code) should it perhaps extract the matches into a set and check that the set has at least three members, rather than requiring all the matches to be unique? (if indeed the function name is correct and that's what it does)
... I remember thinking these thoughts before but not being bold enough to actually make these changes, but now that you are changing it anyway, maybe ...?
BTW: With respect to scanning HTML vs. scanning Markdown: We can easily get both at the same time from the SE API call (assuming we're using /questions/{ids}, etc). Doing so would allow us to write some detections which test the Markdown instead, or in addition to, the HTML. Cutting-over to using Markdown would be a significant effort, but just writing some tests which fill in where we identify issues, shouldn't be that difficult.
The difficulty in getting to that point would be the infrastructure that needs to be in place to enable it. This would include having MS store the Markdown, in addition to the HTML. But, once the infrastructure was in place, adding additional detections for the Markdown would be no more difficult than any other detections.
> Pattern-matching product name in body, pattern-matching product name in title ---------- Title - Pattern-matching product name *max boost extreme* Post - Pattern-matching product name *max boost extreme*
> Pattern-matching product name in body, pattern-matching product name in title ---------- Title - Pattern-matching product name *max boost* Post - Pattern-matching product name *max boost*
> Bad keyword in body, bad keyword in title, bad keyword in username, pattern-matching product name in body, pattern-matching product name in title ---------- Title - Position 11-22: max extreme Title - Pattern-matching product name *max boost* Body - Position 11-22: max extreme Post - Pattern-matching product name *max boost* Username - Position 11-22: max extreme
> Bad keyword in answer ---------- Body - Position 11-22: max extreme
A small notice. The repaired "pattern product name" may catch some FPs as it contains a few really common words, and when they appear two in-a-row it'd get caught. Current keywords
[ SmokeDetector | MS ] Bad NS for domain in body, bad keyword in body, bad keyword in title, blacklisted website in body, body starts with title and ends in URL, +2 more: Konect Nutra Keto by RogFoster on askubuntu.com
tpu- by iBug
CI on 5af4766 succeeded. Message contains 'autopull', pulling...
> Bad keyword in body, bad keyword in title, bad keyword in username ---------- Title - Position 1-10: Keto Trim Body - Position 1-10: Keto Trim Username - Position 1-10: Keto Trim
> Bad keyword in body, bad keyword in title, bad keyword in username ---------- Title - Position 1-10: Keto Slim Body - Position 1-10: Keto Slim Username - Position 1-10: Keto Slim
@tripleee I think it's OK for us to refuse update a dependency if it isn't necessary (e.g. contains no security update and only backward-incompatible changes) in the future
@iBug you mean for tld? in the long term no, sticking to an old version kind of sucks ... but I didn't expect to have yesterday's ticket resolved during the same day, I created it just as a reminder for us to upgrade at some point
@iBug That pattern looks like it's already caught by Pattern-matching website in answer, Pattern-matching website in body, Bad ns for domain in answer, and Bad ns for domain in body; append -force if you really want to do that.
@iBug That pattern looks like it's already caught by Pattern-matching website in answer and Pattern-matching website in body; append -force if you really want to do that.
@iBug Added supplements4test\.com to watchlist
CI on 8cccd4b succeeded. Message contains 'autopull', pulling...
@Makyen That pattern looks like it's already caught by Potentially bad keyword in body and Potentially bad keyword in answer; append -force if you really want to do that.
In the requirements.txt for a Python library I am using, one of the requirements is specified like:
mock-django~=0.6.10
What does ~= mean?
