extracting_datetime() returns incorrect values if a week day is included with the utterance

[krisgesling]

Describe the bug
extracting_datetime() returns incorrect values if a week day is included with the utterance.

To Reproduce

extract_datetime("monday april 5th")

returns

[datetime.datetime(2022, 4, 11, 0, 0, tzinfo=tzfile('/usr/share/zoneinfo/Australia/Darwin')), '']

Removing "monday" from the utterance returns the correct value.

[ChanceNCounter]

May be addressed by WIP #96 (I haven't checked)

[emphasize]

not to say that 04/11/2022 isn't a monday.
edit: shoot, looked at may (yet the code wouldn't care if day mismatches datum)

this one year offset also occurs at _de, yet not the day offset (in your example +6) - but i guess you parsed the day before today

these are the codelines where the +6 should originate from:

elif word in days and not fromFlag:
            d = days.index(word)
            dayOffset = (d + 1) - int(today) #where today = anchorDate.strftime("%w")
            used = 1
            if dayOffset < 0:
                dayOffset += 7

[emphasize]

Ok, got it. (one year offset) Since

temp = datetime.strptime(datestr, "%B %d")
extractedDate = extractedDate.replace(hour=0, minute=0, second=0)

eg (if called on the same day)

>>>2021-05-01 00:00:00+02:00 (temp)
>>>2021-05-01 00:00:00+02:00 (extractedDate)

if it has no year the case would be if extractedDate <= temp: (instead of if extractedDate < temp:)
otherwise extractedDate = extractedDate.replace(year=int(currentYear) + 1,...

skimmed through the WIP refactor and it seems it is not resolved with #96

[emphasize]

dayOffset:
The problem is that weekdays in on itself are treated as relative words (what is reflected in dayOffset from now) -or in other words durations. And those durations collide with the specific date given
If you are parsing "monday may 3rd" you're asking "2 days from now after may 3rd"

Furthermore things like "Friday after May 3rd" are also not correctly parsed since the relative date (or anchorDate) should be may 3rd in that case, not datenow

[ChanceNCounter]

I have now identified that the refactor heavily addresses weekdays. Still haven't tested as I don't know the state of it; maintainers have obviously been stretched thin lately.

Just in case, @JarbasAl ⬆️ and keep 'em coming! Edge cases galore on the way to rewrite glory, after all.

[emphasize]

@ChanceNCounter Hell yeah! There's a lot to identify. And tackles (from what i can tell) those pitfalls i measly outlined. Don't sleep on that one.

And you've said there's nothing to localize :P