What should the structure and contents of the zip file for MassStoreRun be?

[jstevens176]

We're trying to use the CodeChecker server/database combination such that we can utilise the viewer UI but using a different source for the Sarif files we want to send to CodeChecker. As such, we are using the thrift API to send CodeChecker a massStoreRun message containing a b64 encoded zipfile with our analysis results stored within.

We're having some trouble trying to figure out how to format the contents of the zipfile such that CodeChecker can read the results.

Currently, when Codechecker unzips the message it results in the following files and directories stored at:

/**various_directories**/**and_subdirectories**/.codechecker/tmpABCDEF/ ->
-> reports/**hash_to_sarif_file**/SarifFileName.sarif
-> root/**directory structure and files of the original repo copied here**
-> blame/**directory structure and files of the original repo copied here, but the files contain blame data instead of code**
-> content_hashes.json

Furthermore, currently the sarif file contains no originalUriBaseId, the artifact uri's are relative to the top level directory of the analysed code and each artifact contains a uriBaseId of %SRCROOT%. content hashes have files with the same format of the uri i.e. top-level/sub-levels/other-sublevels/etc/code_file.cpp.

As it currently stands, when running CodeChecker store we encounter errors when trying to __store_source_files in that the source_file_path has repeated elements when trying to "path_for_fake_root()", i.e. /**various_directories**/**and_subdirectories**/.codechecker/tmpABCDEF/home/username/top-level/sub-levels/other-sublevels/etc/code_file.cpp when the file is located at /**various_directories**/**and_subdirectories**/.codechecker/tmpABCDEF/top-level/sub-levels/other-sublevels/etc/code_file.cpp

Is there any documentation regarding what format things being sent to CodeChecker need to be in? Or other instruction as to how to structure things that can sent to CodeChecker?

[jstevens176]

Hi @Discookie,

You made some changes to sanitise the input paths in mass_store_run as part of this commit: 7a98601

I was wondering if you had any insight into the format of the paths you're expecting as input that could help me answer my above question.

Thanks :)