Each time I perform an analysis on my project (on an unchanged code base), the number of outputted errors/warnings/information randomly changes. I would think that the output should always be identical since nothing has changed. Further down are the statisics from two back-to-back scans. Version: Cppcheck v1.87
Note: Performing a "Clear results" between scans does not fix it.
Workaround 1: Close Cppcheck, reopen and press "Reanalyze all files"
Workaround 2: Delete the myproject-cppcheck-build-dir folder then "Reanalyze all files".
The results between the two work around is also different.
Attached files:
My .cppcheck file
lastResults - 1st.xml (results from "good" scan)
lastResults - 2nd.xml (results after back-to-back scan)
Notes About lastResults files:
In the 2nd file, there are results missing, as would be expected from what the Statistics output reported. What's strange is that for the lines that are the same, the escape character for a newline is a different format. For the first, the escape sequence is \012 and for the 2nd it is 

Another strange output between the files is that there is the following output about a line of code. In the 2nd lastResults.xml file, the order in which the lines are listed are reversed. Should that not be consistent between scans?
<errorid="redundantAssignment"severity="style"msg="Variable ... (I deleted code here for clarity) <location file="Sources\FlexCMBApp.c"line="996"info=""/><locationfile="Sources\FlexCMBApp.c"line="1009"info=""/></error>
Based on both of these observations, is it possible the outputs being piped through a different "outputter" in different conditions?
@@@@ Good scan @@@@@@@@@@@@@@@@@@@@@@@
Project Settings
Project:
Paths: Sources;test
Include paths:
Defines:
Undefines:
Previous Scan
Path selected: C:/Share/Projects/CMBGen2/FW/CMBGen2
Number of files scanned: 99
Scan duration: 0.2 seconds
Statistics
Errors: 13
Warnings: 0
Style warnings: 384
Portability warnings: 0
Performance warnings: 0
Information messages: 2
Each time I perform an analysis on my project (on an unchanged code base), the number of outputted errors/warnings/information randomly changes. I would think that the output should always be identical since nothing has changed. Further down are the statisics from two back-to-back scans. Version: Cppcheck v1.87
Note: Performing a "Clear results" between scans does not fix it.
Workaround 1: Close Cppcheck, reopen and press "Reanalyze all files"
Workaround 2: Delete the myproject-cppcheck-build-dir folder then "Reanalyze all files".
The results between the two work around is also different.
Attached files:
Notes About lastResults files:
In the 2nd file, there are results missing, as would be expected from what the Statistics output reported. What's strange is that for the lines that are the same, the escape character for a newline is a different format. For the first, the escape sequence is \012 and for the 2nd it is 

Another strange output between the files is that there is the following output about a line of code. In the 2nd lastResults.xml file, the order in which the lines are listed are reversed. Should that not be consistent between scans?
Based on both of these observations, is it possible the outputs being piped through a different "outputter" in different conditions?
@@@@ Good scan @@@@@@@@@@@@@@@@@@@@@@@
Project Settings
Project:
Paths: Sources;test
Include paths:
Defines:
Undefines:
Previous Scan
Path selected: C:/Share/Projects/CMBGen2/FW/CMBGen2
Number of files scanned: 99
Scan duration: 0.2 seconds
Statistics
Errors: 13
Warnings: 0
Style warnings: 384
Portability warnings: 0
Performance warnings: 0
Information messages: 2
@@@@ "Bad" scan @@@@@@@@@@@@@@@@@@@@@@@
Project Settings
Project:
Paths: Sources;test
Include paths:
Defines:
Undefines:
Previous Scan
Path selected: C:/Share/Projects/CMBGen2/FW/CMBGen2
Number of files scanned: 99
Scan duration: 0.1 seconds
Statistics
Errors: 10
Warnings: 0
Style warnings: 168
Portability warnings: 0
Performance warnings: 0
Information messages: 2
I have the same problem with version 1.87. The behaviour is reproducible.
Edit:
It could be that this is just a counter problem, I tried to analyze a file with this code:
And the counter was always 0, even after restarting the program. But the error was always in the list.
Last edit: Adrian Haider 2019-04-10