pydevDAG is configurable and flexible. At this point, it has no configuration file or options. It is necessary to take advantage of its configurability by editing its top-level file, _graphs.py. Each class in _graphs.py represents a separate, configurable aspect of the graphing operations.
Graph generation makes use of the _structure module. This module defines a number of classes which handle the addition of different types of nodes or different categories of relationships to the graph. The contents of the graph are determined by the choice of which of these classes participate in building the graph.
Graph decoration is the process of adding additional information to nodes or edges of the graph as desired. At present, all information is obtained either from udev properties or sysfs attributes. The classes that manage the decoration process and also obtaining suitable decorations are defined in the _decorations module.
It is necessary to convert some values in the graph to text in order to write them in various graphical output formats. The _write module handles these tasks as well as the inverse task of reading a file and converting text to an object suitable for the graph.
Display of the graph is currently handled by means of the graphviz library. The graph is first converted to a an internal graphviz graph and is then annotated with various graphviz attributes to enrich the visual display with the most relevant information in the graph. This enrichment is handled by the _display module. Each of the classes in the _display module implements a different enrichment, and the choice of modules determines the appearance of the graph.
Textual representation of the graph, in the form of a table is also useful. The _print module handles the high-level organization of the output, while the _print_helpers module consists of small classes that handle the choice of values for individual columns.
It is helpful to determine programatically if there is a difference between one graph and another. What makes two graphs equivalent is itself a configurable notion. The _compare module allows defining multiple different notions of equivalence. If two graphs are different, then the difference between them can be calculated in various ways; the result is itself a graph.