`namingng.py` somewhat supported specifying a dict instead of a list for
regular expressions, until the feature was broken by a patch of mine
recently. This PR contains a patch rewriting the feature and expanding
relevant unit tests.
To improve maintainability, a second patch is added that refactors the
code for better readability and structure.
For naming issues reported, column was always set to `0`, which is now
fixed.
Global variable naming errors were reported as "Public member" issues,
which is also fixed.
The unit test now covers namespaces, class names, public and private
member variables.
Include guard naming can be validated against various patterns:
- prefixes/suffixes (`_FILE_H`, `PROJECT_FILE_H`, `FILE_H_`)
- basename/full path (`FILE_H`, `SUB_DIR_INC_FILE_H`)
- upper- or lowercase (`FILE_H`, `file_h`)
- any combination of the above (`project_sub_dir_inc_file_h_`)
A regexp can be specified to match header filenames. The example matches
any filename not starting with / and ending with `.h`, intended to match
C header files while exluding system files.
The test is not limited to naming only; validity and presence of include
guards can also be tested by setting `"required":true` in the config
file.
Enabling this feature requires adding the key `"include_guard"` to the
namingng config file used.
The namingng unit test is extended to test various features of the
include guard test.
Also, config handling is improved, adding (superficial) validation and a
unit test.
namingng.py was only usable in standalone mode, but now supports CLI
mode, i.e. with cppcheck --addon=namingng. It uses the generic reporting
provided by cppcheckdata.reportError(). All output other than reported
errors is suppressed.
A local function reportNamingError() is implemented to call through to
cppcheckdata.reportError(), filling in common defaults.
The collection of errors and the --verify feature are removed, including
related workflow and a test file. These are replaced by a unit test.
In case a user accidentally uses a wrong JSON file (e.g. naming.json,
which is the config file for namingng.py), the code could give a
confusing exception. This happens when the key 'script' is not defined
as a string.
This is solved by testing the key for existence and type. In case
'script' is not a key or refers to a type other than a string, a clear
error is given, stating for example: 'Loading naming.json failed. script
must be set to a string value.'
The message is kept in line with other messages. Maybe it can be
clarified further, e.g. 'Loading naming.json failed. A key "script" must
be set with a string value referring to a Python script.' - in which
case the errors relating to other keys may also be clarified.
Currently the `AddonInfo` is generated and discarded on each addon
invocation. This leads to an unnecessary process invocation for each
addon on each file.
Also if an addon is completely broken we will still perform the whole
analysis only for it to be failed at the end so we should bail out early
if we know it doesn't work at all.
This is a step onto leveraging the `ThreadExecutor` implementation for
`ProcessExecutor` which is a follow-up to #4870. We need to have the
proper test coverage and the existing implementations working as
expected before we move to the shared code.
Fixes:
- added `--showtime=` tests for all executor implementations
- only print `--showtime=summary` once at the end
- prevents `--showtime=` by multiple threads to be written at the same
time - essentially breaking the output
- reset the timer results before each test
- deprecated `top5` in favor of `top5_file`
- fixed printing for all executors except `ProcessExecutor`
Encountered while investigating https://trac.cppcheck.net/ticket/11708.
This has been like this since the introduction of `internalError` in
b6bcdf2936 (almost ten years ago to the
day). Logging internal errors which bail out(!) of the analysis simply
to `std::cout` for them possibly never to be seen (and also not affected
the exitcode) is pretty bad IMO. They should always be visible.
I also removed the filename from the message as it is already available
(and thus redundant) and its existence should be defined by the
template.
This was introduced in #5279. We were transferring the terminating `\0`
via the pipe and also added another one in the parsing. As we are now
directly writing into a `std::string` these extra characters will now
show up in it. So just get rid of them.
* Preprocessor: cleaned up `missingInclude()`
* Preprocessor: relaxed dependency on `Suppressions` / adjusted `TestPreProcessor::inline_suppression_for_missing_include()` which was not testing production behavior
* test/cli/test-other.py: added test for `missingInclude` and `missingIncludeSystem` inline suppressions
* fixed `constParameterReference` selfcheck warning
* cleaned up triage.pro and pass all triage files to selfcheck
* CI-unixish.yml: factored out redundant selfcheck flags
* CI-unixish.yml: cleaned up selfcheck suppressions and fixed warnings
* added explicit `missingInclude` checks to selfcheck to work around current issues/limitations / fixed selfcheck includes / added some unit tests for `missingInclude`
* applied the selfcheck changes to the sanitizer jobs
* only fail TSAN CI build in case of TSAN error / updated sanitizer CI build TODOs
Maarten van der Schrieck
thingsconnected
Daniel Marjamäki
Oliver Stöneberg
chrchr-github