fixes after deploy 03.25

[antgonza]

No description provided.

[wasade]

Do we know the structure of the filename sufficiently to express this as a regex?

[antgonza]

Not sure about regex and if adding that will actually help. Anyway, for clarity, as the comment says the file name is: run-id, project_qid, other - for example: 20231009_FS10001773_73_BTR67718-2515.U19_ADRC_14736.1.tsv so we could do something like '{self.pipeline.run_id}.{project_qid}.{extra}' where we are looking for the digits after '_', in this case '14736'. Does this help?

[wasade]

I think it would be this, right? Which would allow for matching 5 or 6 digit study IDs to future proof for a while?

>>> x = "20231009_FS10001773_73_BTR67718-2515.U19_ADRC_14736.1.tsv"
>>> re.compile(r"(?P<runid>[a-zA-Z0-9_-]+)\.(?P<qname>[a-zA-Z0-9_]+)(?P<qid>[0-9]{5,6})\..\.tsv").match(x).groups()
('20231009_FS10001773_73_BTR67718-2515', 'U19_ADRC_', '14736')
>>>

If we know the pattern, then it seems reasonable to formally describe the pattern as presumably anything which does not fit is erroneous. If we do not know the pattern, then it suggests we should enforce control upstream to minimize inference and favor explicit behavior

[antgonza]

This sounds reasonable but before making the change, how do you think we should handle errors, like in this example:

>>> x = 'XX'
>>> re.compile(r"(?P<runid>[a-zA-Z0-9_-]+)\.(?P<qname>[a-zA-Z0-9_]+)(?P<qid>[0-9]{5,6})\..\.tsv").match(x).groups()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
AttributeError: 'NoneType' object has no attribute 'groups'

just a try/except-pass?

[wasade]

Why not just store the result of match in a variable and test whether it is None?

[antgonza]

Thank you for the review @wasade !