Introduce plist.err and parse --status

[barnabasdomozi]

Currently, it is a problem to determine if the analysis was successful or failed during execution.

With this patch, we will create plist.err files to indicate a failed analysis action.
The parse --status command was also introduced to
get the analysis summary of an existing report directory.

[bruntib]

Aren't .err files generated only when analysis fails?
Why clangsa and clang-tidy are hard-coded when printing stats?

[barnabasdomozi]

@bruntib

Aren't .err files generated only when analysis fails?

If the return_code is 0 we return from error_file.py's create() method and no .plist.err will be generated.

Why clangsa and clang-tidy are hard-coded when printing stats?

Is there a list of supported analyzers defined somewhere else?
I think in the regex pattern we should try to match for existing analyzers and skip the rest of the files.

[barnabasdomozi]

@bruntib

I added error file generation for the other analyzers as well.