2. How to Extend

vcsSHARK has two different extension possibilities. First, you can add new parsers for new repository types (e.g. SVN, CVS). Currently, only GIT repositories can be parsed. Second, there is the possibility to add new datastores (e.g. MySQL). Currently, only MongoDB is supported as a datastore backend. This chapter will show what needs to be done to extend vcsSHARK.

2.1. Parsers

All parsers are stored in pyvcsshark/parser folder. There are conditions, which must be fulfilled by a parser so that it is accepted by vcsSHARK:

  1. The *.py file for this parser must be stored in the pyvcsshark/parser folder.

  2. It must inherit from BaseParser and implement the methods defined there.

The process of chosing the parser is the following:

  • Every parser gets instantiated

  • The detect method is executed

  • If a parser returns true for the detect method, this parser is used for the repository

There are several important things to note:

  1. If you want to use a logger for your implementation, get it via

    logger = logging.getLogger("parser")

2. You must call the add_commit() function of the datastore, that is given to the parser as argument of the parse method, to add commits.

3. You must use the models as they are defined in in the module pyvcsshark.dbmodels.models, as otherwise the interfaces do not match

  1. The execution logic is in the application class and explained here Application.

2.2. Datastores

All datastores are stored in the pyvcsshark/datastores folder. There are conditions, which must be fulfilled by a datastore so that it is accepted by vcsSHARK:

  1. The *.py file for this datastore must be stored in the pyvcsshark/datastore folder.

  2. It must inherit from BaseStore and implement the methods defined there.

    Note

    The store_identifier property must just return a string, which represents your datastore and is not used by another datastore

The process of choosing the datastore is the following:

  • Every datastore gets instantiated

  • The store_identifier property is compared to the db-driver that was chosen by the user

  • If they are equal, the correct datastore was found

There are several important things to note:

  1. If you want to use a logger for your implementation, get it via

    logger = logging.getLogger("datastore")

2. The add_commit() method gets a commit model of class CommitModel as parameter. Think about this first!

  1. The execution logic is in the application class and explained here Application.

Note

Dont mind if your datastore do not need all the information of the initialize() function.