Fix #14317 Syntax error using macro inside ifdef block by ceJce · Pull Request #8123 · danmar/cppcheck

ceJce · 2026-01-14T09:05:53Z

Changed so that auto-configured values (without specified values) are set to itself instead of 1.

From issue 14317

isless=islesss

instead of
isless=1

danmar · 2026-01-14T09:56:28Z

I would expect that we test this in some way.

chrchr-github · 2026-01-14T12:49:19Z

This needs to be fixed in simplecpp, probably here:
const std::string rhs(eq==std::string::npos ? std::string("1") : macrostr.substr(eq+1));
https://github.com/danmar/simplecpp/blob/6d45cd7bf7f12878ed99a5dcf5ee410ef6492a4a/simplecpp.cpp#L3343

danmar · 2026-01-14T14:51:06Z

This needs to be fixed in simplecpp, probably here:

I am not sure. Then what happens if user executes cppcheck like this:

cppcheck -DX test.c

-DX should work the same as -DX=1 imho.

danmar · 2026-01-14T16:02:05Z

hmm.. the fix seems to work when I test it quickly. but is in wrong repo.

ceJce · 2026-01-19T12:05:35Z

This needs to be fixed in simplecpp, probably here: const std::string rhs(eq==std::string::npos ? std::string("1") : macrostr.substr(eq+1)); https://github.com/danmar/simplecpp/blob/6d45cd7bf7f12878ed99a5dcf5ee410ef6492a4a/simplecpp.cpp#L3343

The solution did not work, so I'm working on a new solution.

chrchr-github · 2026-01-23T14:21:23Z

test/testpreprocessor.cpp

                                "#error abcd\n"
                                "#endif\n";
-        ASSERT_EQUALS("\nA\n", getConfigsStr(filedata));
+        ASSERT_EQUALS("\nA=A\n", getConfigsStr(filedata));


This raises the question, do we want to print A=A when it was just A before? @danmar

Yes that is a question. I think it's good to explicitly show that A=A expansion is used. And the "=1" is not wrong to show but it's redundant.

When configuring the defines in the GUI project file dialog, "A;B" means that A and B will be 1. While "A=A;B" would mean that A macro will get value "A" and B macro would get value "1". As far as I know this is how it worked from the beginning.

"no value" implied 1. A=A is misleading and to make it makes no sense. If you specify -DA on the CLI it will still show just as A.

See also https://trac.cppcheck.net/ticket/14458.

And see also https://trac.cppcheck.net/ticket/14166.

ceJce · 2026-01-26T10:01:40Z

getConfig is updated into returning configs:

"ifdef X" and "if define(X)",return as 'X=X'
For ifndef, if etc nothing is changed, returns 'X'
If there are multiple X:s in the code, for example "ifdef X" and "if X" (or an "ifndef X"), the more general 'X' is returned.

Expected testresults updated accordingly and 2 new issue specific tests added to testpreprocessor.cpp

lib/preprocessor.cpp

danmar

is there any test that contains both #ifdef X and #if X and checks the config output? I guess test/cfg does test this indirectly but testpreprocessor could also have a direct test.

test/testpreprocessor.cpp

ceJce · 2026-01-27T07:56:46Z

is there any test that contains both #ifdef X and #if X and checks the config output? I guess test/cfg does test this indirectly but testpreprocessor could also have a direct test.

Good catch. The existing tests indirectly tested it, but no test that specifically test. Added new test.

lib/preprocessor.cpp

* Updated so that getConfig extracts ifdefs and if defined as itself, example X=X, and for all else just X * Updated hasDefine checker to detect configs even in pattern X=X

* Updated expected results from getConfigs according to defines X=X, else X * Added two TestPreprocessor tests, getConfigsIssue14317 & getCodeIssue1431, verifying issue 14317 fixed.

…Config. * Added checks that makes sure that no duplicated of a configuration (like X & X=X, in that case only X is kept) ends up in getConfig ret.

sonarqubecloud · 2026-01-28T12:49:12Z

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

mptre · 2026-01-30T06:04:29Z

This introduced a regression: https://trac.cppcheck.net/ticket/14455#ticket

ceJce force-pushed the fix14317_syntax_error branch from 05b5536 to eb9075c Compare January 14, 2026 15:08

ceJce marked this pull request as draft January 14, 2026 15:12

ceJce force-pushed the fix14317_syntax_error branch from eb9075c to 8321b48 Compare January 19, 2026 09:18

ceJce force-pushed the fix14317_syntax_error branch 7 times, most recently from 2ec411a to 52f76ef Compare January 23, 2026 13:40

chrchr-github reviewed Jan 23, 2026

View reviewed changes

ceJce force-pushed the fix14317_syntax_error branch from 52f76ef to b87ba98 Compare January 26, 2026 09:36

ceJce marked this pull request as ready for review January 26, 2026 13:59

danmar reviewed Jan 26, 2026

View reviewed changes

lib/preprocessor.cpp Outdated Show resolved Hide resolved

danmar reviewed Jan 26, 2026

View reviewed changes

test/testpreprocessor.cpp Show resolved Hide resolved

ceJce force-pushed the fix14317_syntax_error branch from b87ba98 to caa892a Compare January 27, 2026 07:53

danmar reviewed Jan 27, 2026

View reviewed changes

lib/preprocessor.cpp Show resolved Hide resolved

lib/preprocessor.cpp Show resolved Hide resolved

ceJce added 4 commits January 28, 2026 13:25

Fix #14317 Syntax error using macro inside ifdef block

52046ff

* Updated so that getConfig extracts ifdefs and if defined as itself, example X=X, and for all else just X * Updated hasDefine checker to detect configs even in pattern X=X

Updated expected results due to updated getConfig & added 2 new tests

4f307d2

* Updated expected results from getConfigs according to defines X=X, else X * Added two TestPreprocessor tests, getConfigsIssue14317 & getCodeIssue1431, verifying issue 14317 fixed.

Add look for duplicates and select most general in config list in get…

33a2d21

…Config. * Added checks that makes sure that no duplicated of a configuration (like X & X=X, in that case only X is kept) ends up in getConfig ret.

Updated expected results for SOME_CONFIG config in helloworld_test.py

cc4c880

ceJce force-pushed the fix14317_syntax_error branch from caa892a to cc4c880 Compare January 28, 2026 12:48

danmar approved these changes Jan 29, 2026

View reviewed changes

danmar merged commit 1b82675 into danmar:main Jan 29, 2026
77 of 78 checks passed

ceJce deleted the fix14317_syntax_error branch January 30, 2026 13:59

Conversation

ceJce commented Jan 14, 2026

Uh oh!

danmar commented Jan 14, 2026

Uh oh!

chrchr-github commented Jan 14, 2026

Uh oh!

danmar commented Jan 14, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

danmar commented Jan 14, 2026

Uh oh!

ceJce commented Jan 19, 2026

Uh oh!

chrchr-github Jan 23, 2026

Choose a reason for hiding this comment

Uh oh!

danmar Jan 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

firewave Feb 1, 2026

Choose a reason for hiding this comment

Uh oh!

firewave Feb 1, 2026

Choose a reason for hiding this comment

Uh oh!

ceJce commented Jan 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

danmar left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

ceJce commented Jan 27, 2026

Uh oh!

Uh oh!

Uh oh!

sonarqubecloud bot commented Jan 28, 2026

Quality Gate passed

Uh oh!

Uh oh!

mptre commented Jan 30, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

danmar commented Jan 14, 2026 •

edited

Loading

danmar Jan 23, 2026 •

edited

Loading

ceJce commented Jan 26, 2026 •

edited

Loading

danmar left a comment •

edited

Loading