Keep content of textual, unparsed files to the dataset

[nguyenhoan]

Useful for tasks that processing contents as textual documents.
Users might also write simple parser in Boa to parse and extract information.

[nguyenhoan]

53daba4

[psybers]

Should we store all textual files, including ones we parsed? That gives users a consistent capability of being able to analyze any text file (including source).

[psybers]

The data should be stored into its own data file, not in the AST sequence file.

[psybers]

There needs to be a Boa function to read the file and return the contents for a given ChangedFile.

[nguyenhoan]

I don't see benefit of storing parsed files because they are already in the asts.
Storing them would increase the space significantly.

[hridesh]

Agreed. Perhaps limit textual contents to unparsed ASCII files?

[psybers]

Parsed files do not retain all information (such as whitespace and comments) however. Hence my thoughts to including them.