[Bug]: Latest monty release doesn't find regrep matches on last line

[kavanase]

Email (Optional)

No response

Version

v2025.1.9

Which OS(es) are you using?

• MacOS
• Windows
• Linux

What happened?

In the latest release of monty, regrep fails to match text which is on the last line of the file:

Works fine with previous release:

Noticed by automated tests with ShakeNBreak

Code snippet

from monty.re import regrep

with open("test_file", "w+") as f:
    f.write("some_text\n"*70 + "final line")


regrep(filename="test_file",
       patterns={"match": "final"},
       reverse=True,
       terminate_on_match=True,
       )["match"]

Log output

Code of Conduct

• I agree to follow this project's Code of Conduct

[kavanase]

Works fine when reverse=False, so it's something related to changes in reverse_readfile:

I think due to changes in #730?

[DanielYang59]

Hi @kavanase happy new year and thanks for reporting this.

However I believe this is expected behaviour (introduced in #712) to treat the last matched line ending character (\n in this case) as the marker of last line.

Reason being: according to the POSIX standard definition of "line":

3.185 Line
A sequence of zero or more non- characters plus a terminating character.

In other words, a "correctly formatted" (agreeing with POSIX standard) text file should have every line ended with a line ending char including the last line (i.e. it should end with last line\n instead of last line).

[kavanase]

Hi @DanielYang59 , happy new year!

Ah ok, thanks for linking the PR. I guess it could be an issue with incorrectly-formatted outputs, but I see that was discussed in #712 already.
Thanks for the quick reply!

[DanielYang59]

No worries at all.

I guess it could be an issue with incorrectly-formatted outputs, but I see that was discussed in #712 already.

Perhaps you could take my answer with a pinch of salt as I'm not 100% sure if "all text files should follow the POSIX standard" or it's just the "recommended way". Meanwhile we noticed some output file of LOBSTER may also have some similar behaviour

Free feel to comment if you have any findings :)