Replace libdparse in IfElseSameCheck

[Vladiwostok]

#73 cont

[RazvanN7]

I think that using strcmp is better than going through the hurdles of using to.

[Vladiwostok]

Is strcmp behaviour identical to C? If so, wouldn't that leave the program vulnerable? Wouldn't it be better to use D's strings instead of char * ?

[RazvanN7]

Is this PR ready for review? I'm seeing a lot of commented code.

[Vladiwostok]

Is this PR ready for review? I'm seeing a lot of commented code.

Kind of.
It seems that before DMD, this visitor would only check 2 things: logical expressions and assignments:
if (condition1 && condition1) // [warn]
if (condition1 || condition1) // [warn]
x = x; // [warn]
However, only a unittest was implemented for logical expressions.

I've followed your comments, the Visitor Doc Comment and I've added a few more checks. Now, this visitor should check: logical expressions, assignments, comparisons,
if (condition1 && condition1) // [warn]
if (condition1 || condition1) // [warn]
x = x; // [warn]
if (x == x) // [warn]
if (x != x) // [warn]
if (x > x) // [warn]
if (x >= x) // [warn]
if (x < x) // [warn]
if (x <= x) // [warn]

However, d-scanner with those new checks fails when applied to Phobos. There are unit tests in phobos that explictly check comparisons, so expressions like x==x - those tests trigger a d-scanner warning
Run for FILE in $(find std -name '*.d'); for FILE in $(find std -name '*.d'); do echo "$FILE" ../bin/dscanner -S --config=.dscanner.ini "$FILE" done
shell: /usr/bin/bash --noprofile --norc -e -o pipefail {0}
env:
DC: dmd LD_LIBRARY_PATH: /opt/hostedtoolcache/dc/dmd-2.107.1/x64/dmd2/linux/lib64
std/outbuffer.d
std/system.d
std/getopt.d
std/demangle.d
std/bigint.d
std/bigint.d(753:16)[warn]: Left side of '==' operator is identical to right side.
Error: Process completed with exit code 1.
It seems to me that the issue is D-scanner exits with a non '0' exit code, causing the testing stage to fail:
shell: /usr/bin/bash --noprofile --norc -e -o pipefail {0}
From man bash:
pipefail
If set, the return value of a pipeline is the value of the last (rightmost) command to exit with a non-zero status, or zero if all commands in the pipeline exit successfully. This option is disabled by default.
Am I missing something? What do you think @RazvanN7 ?
Link to failed job

[RazvanN7]

This is failing phobos. IIRC, we discussed that you will preserve the behavior of dscanner and not add any new behavior.

[Vladiwostok]

This is failing phobos. IIRC, we discussed that you will preserve the behavior of dscanner and not add any new behavior.

Yep, will revert to old state