DropDatabase

##MicroDB Graph Database The MicroDB Graph Database is a graph database written in python. A graph database embodies the properties of a graph, consisting entirely of nodes and edges (relationships). Each node and relationship in the database is defined by its attributes. A user can query the database to find useful information about the graph, such as connections between nodes, shortest path to a particular node, and other interesting statistics.

##Usage
The database can be started by typing in

python start.py   

This will load any graph database files from disk to restore the graph state, if one existed.
Commands can be entered at the prompt ">>>".
Commands are terminated by a space followed by a semicolon. For example:

>>> Create n: a asdf:12;

Here are some examples of usage (Look at testfile.txt for more examples):

>>> create n: a b:c;
Unique node id: 1
Nodes
[(1, {'b': 'c'})]
Edges
[]
Identifier
{'a': (1, {'b': 'c'})}
>>> return n: a;
a = (1, {'b': 'c'})
Unique node id: 1
Nodes
[(1, {'b': 'c'})]
Edges
[]
Identifier
{'a': (1, {'b': 'c'})}
>>>

##Query Language
The query language takes the form:

    CREATE          ID ATTR ...
    CREATEEDGE      ID ATTR REL ATTR ID ATTR
    MATCH           ID ATTR (PRED) REL ATTR ID ATTR (PRED) REL ATTR ID ATTR ...
    MODIFYNODE      ID ATTR ID ATTR BOOL
    MODIFYEDGE      REL ATTR REL ATTR BOOL
    DELETENODE      ID ATTR
    DELETEEDGE      REL ATTR
    RETURN          ID ID ...

    HASPATH         ID ATTR ID ATTR
    SHORTESTPATH    ID ATTR ID ATTR
    NEIGHBOR        ID ATTR
    HASEDGE         ID ATTR ID ATTR

    SHOW
    CLEAR
    VISUALIZE
      
    REL ATTR = e: a b:c
    ID ATTR = n: a a:b
    BOOL = b: a val:0/1 - 1 means add the attribute, and 0 means delete the attribute
    ID = n: a
    PRED = a(pred)b

##Framework
The database is started by initializing the StartDatabase class. This loads the graph files from disk and starts the prompt to take input from the user. The parser is then called (from parser.py) and parses commands from the user, passing in the objects parsed into the linker (from linker.py). The linker then calls the appropriate evaluation methods in query_evaluator.py.

##Libraries Libraries needed for this database are:

NetworkX

matplotlib

pprint

##Testing Testing is currently only available for the query_evaluator.py methods. The python unittest library is used for testing. Tests for the query_evaluator methods can be run by

python test_query_evaluator.py

##Documentation
The documentation files can be generated by

epydoc --config=doc_config.txt   

The files are then generated into the pythondoc folder (open index.html to view the documentation)