File representation as commit and issue messages

[daomcgill]

Placeholder

[daomcgill]

@carlosparadis part III. Let me know if this should be moved to the code_embedding repo.

[carlosparadis]

@daomcgill I will keep this as a placeholder issue for the time being, this should go on the other repo.

[carlosparadis]

@daomcgill Let's use this as an extension of the #313. Namely, whereas there you should extract subset of words from srcML to represent a file, here you represent a file by a) the commit messages associated to it, and b) the issue messages they are tied to.

This may sound a bit confusing, so let me know if you have questions. This will also serve as input to the original issue here.