ualberta-smr / merganser Goto Github PK

validation_repository_name --> validate_repository_name

Please check all code.

https://github.com/ualberta-smr/code-owhadi-msr19/blob/1724d8a2ff6b334c0775a8498359ce892249d320/merge_excavator/validation.py#L5

Document reason for or (mentioning the weird scenario you ran to if you can look it up)

https://github.com/ualberta-smr/code-owhadi-msr19/blob/1724d8a2ff6b334c0775a8498359ce892249d320/merge_excavator/code_quality.py#L28

Maybe change to get_build_status

https://github.com/ualberta-smr/code-owhadi-msr19/blob/1724d8a2ff6b334c0775a8498359ce892249d320/merge_excavator/code_quality.py#L8

There are several warnings when inserting data into the Merge_Related_Commit Table.

Please check this line:

https://github.com/ualberta-smr/code-owhadi-msr19/blob/1724d8a2ff6b334c0775a8498359ce892249d320/merge_excavator/merge_replay.py#L77

When running the tests, a project may be using maven but has no test suites. Such projects should be clearly marked. E.g., -1 as the value of the passing test or some other field marked for the project

What are these:

They are not described anywhere?

It would be good if there's a parameter to indicate the starting date of the mining (i.e., mine commits after date X). In that case, we should check if the repo is already cloned and if it is, we shouldn't clone again.

For the -q flag for the search module, what if you have multiple search terms, should you use quotes? If so, clarify in instructions/example.

https://github.com/ualberta-smr/code-owhadi-msr19/blob/1724d8a2ff6b334c0775a8498359ce892249d320/merge_excavator/merge_scenario_data.py#L55

https://github.com/ualberta-smr/code-owhadi-msr19/blob/1724d8a2ff6b334c0775a8498359ce892249d320/merge_excavator/GitUtil.py#L138

There is encoding error while parsing the output of git to extract the conflicting regions since some commit messages use non-ASCII characters.

https://github.com/ualberta-smr/code-owhadi-msr19/blob/1724d8a2ff6b334c0775a8498359ce892249d320/merge_excavator/GitUtil.py#L136

The wiki page says "The directory path for reading the list of repositories, as *.txt files", which suggests that you may have multiple input list files for the set of repos to analyze. In the ReadMe instructions, it suggests that there is only one list

I would suggest that you encourage people to use Issues for bugs or even questions rather than to email us, since this allows better tracking of things.

I would also suggest adding a Contributors list, where you can have both our names (with yours first as the main author) and contact information.

I think the command may return multiple branches and you only consider the first one. It would be good to verify if it does indeed return multiple branches and if for our purposes, considering the first branch is enough

https://github.com/ualberta-smr/code-owhadi-msr19/blob/1724d8a2ff6b334c0775a8498359ce892249d320/merge_excavator/GitUtil.py#L108

In the parameter page, add the default values for the parameters you indicate don't necessarily need to be changed?

https://github.com/ualberta-smr/code-owhadi-msr19/blob/1724d8a2ff6b334c0775a8498359ce892249d320/merge_excavator/merge_replay.py#L86

https://github.com/ualberta-smr/code-owhadi-msr19/blob/1724d8a2ff6b334c0775a8498359ce892249d320/merge_excavator/validation.py#L18

The checksumdir pytohnlibrary reports is cannot find the directory and returns Directory Not Found eror.

Maybe change to "Temp variables for reading username and repo"

https://github.com/ualberta-smr/code-owhadi-msr19/blob/1724d8a2ff6b334c0775a8498359ce892249d320/merge_excavator/main.py#L71

The current code is using setup.py which helps us to install the package py pip. Another option would be using docker.

Do we need to have a docker file too?

If you could not run the build or test for any reason, or the repo doesn't support mvn, then store -2 (And update documentation in ReadMe or data schema description) and log a failure

https://github.com/ualberta-smr/code-owhadi-msr19/blob/1724d8a2ff6b334c0775a8498359ce892249d320/merge_excavator/code_quality.py#L33

I'm assuming that if all the max parameters (e.g., max number of days, merge scenarios etc.) are not set, that all the merge scenarios in a given repo are analyzed. Is that correct?