Added get_callee() which determines the right routine of the return values of get_callees()

[schreiberx]

It does what's in the title.

[arporter]

Thanks @schreiberx, unfortunately, linting is still failing.

In case it helps, I have the following pre-push hook setup:

$ cat PSyclone/.git/hooks/pre-push
#!/usr/bin/sh

# An example hook script to verify what is about to be pushed.  Called by "git
# push" after it has checked the remote status, but before anything has been
# pushed.  If this script exits with a non-zero status nothing will be pushed.
#
# This hook is called with the following parameters:
#
# $1 -- Name of the remote to which the push is being done
# $2 -- URL to which the push is being done
#
# If pushing without using a named remote those arguments will be equal.
#
# Information about the commits which are being pushed is supplied as lines to
# the standard input in the form:
#
#   <local ref> <local oid> <remote ref> <remote oid>
#
# This script ensures that the whole of PSyclone is linted successfully
# before the push is executed.

remote="$1"
url="$2"

if ! command -v flake8; then
    echo "WARNING: source not linted because flake8 unavailable"
    exit 0
fi

flake8=$(flake8 src/psyclone)
if (( flake8 == 1 )); then
    echo "Linting failed"
else
    echo "Linting succeeded"
fi
exit $flake8

[schreiberx]

Hi Andy, Thanks for this script.
I think that all these helper scripts should be part of psyclone to quickly run them over the source code. I added this script in the newest pushed version in bin_git/run_flake8.sh.
Cheers,
Martin

[codecov]

Codecov Report

Attention: Patch coverage is 95.65217% with 3 lines in your changes missing coverage. Please review.

Project coverage is 99.87%. Comparing base (a33a99b) to head (e4ac880).

Files with missing lines	Patch %	Lines
src/psyclone/psyir/nodes/call.py	95.38%	3 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #2775      +/-   ##
==========================================
- Coverage   99.88%   99.87%   -0.01%     
==========================================
  Files         356      356              
  Lines       49479    49540      +61     
==========================================
+ Hits        49421    49479      +58     
- Misses         58       61       +3     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

---

🚨 Try these New Features:

• Flaky Tests Detection - Detect and resolve failed and flaky tests

[arporter]

@schreiberx please could you hold off making further changes until I pass it back to you? It's hard to review if code changes underneath me while I'm doing it :-)

[schreiberx]

These changes are only addressing increasing the coverage to 100% since the codecov bot complained with "Please review."
I'll keep my hands off for now :-)

[arporter]

Thanks very much for this Martin. I like the new routine making use of the existing functionality, that seems sensible. I did struggle to get my head around it so you'll see I've requested that some comments be tweaked or added.
As I say in the review, I think a little work on the frontend for those situations where we hit OPTIONAL could really help here.
The only thing I'm really not sure of is the passing around of the list of integers. That seems a bit odd so I'd like to understand why you do that.
I've not looked at the test suite yet (but thanks for addressing the missed lines).

[arporter]

Back to @schreiberx :-)

[schreiberx]

Back to @arporter :-)

[schreiberx]

Back to @arporter :-)

[arporter]

Thanks for the updates Martin. I think the main thing to agree upon now is how best to return the information on the argument mapping. Apart from that it's a matter of tidying a few things. You also still have a couple of lines missing from coverage (but best to leave that until we've hammered-out the implementation).
I have the one-line fix to improve our OPTIONAL support as #2785 which is just waiting on review. Since it's so simple, you could merge that branch into this one.

[arporter]

Please remove all these ifs and print statements (but do add comments describing what specific case each part of the test is for).

[schreiberx]

These if statements are very handy for debugging potential future problems. Setting some of them to '0' allows you to skip this subtest.

Also, the print statements provide helpful output again for debugging.
I don't understand why there should be no output. Is this a convention from PSyclone?

[schreiberx]

PS: The indentation with the "if 1" statements also structures such subtests nicely, hence, improving readability :-)

[schreiberx]

Back to @arporter