Most versioned components have their own method of identifying their versions. This method is often termed “versioning” or called a “version scheme”. The method is simply a set of rules that dictate what is a valid version identifier and is usually separated (often by delimiters, such as a dot) in a number of segments to signify hierarchy among the versions. The hierarchy determines ordering along with the values of the segments within the hierarchy.
In the verschemes library, a version scheme is a set of rules pertaining to the identification of versions for a given scope or purpose. Each version contains a number of segments often separated by delimiters. Each segment is allowed to have multiple fields to identify portions of the segment, though many segments have just one field.
Many version schemes are just integers separated by dots. The base Version class works fine for generic version numbers that fit this scheme, but the real power of this library is in defining version schemes (Version subclasses) with segments that specifically describe the scheme and automatically implement validation and normalized rendering of a version identifier or a sequence of version segment values. The SEGMENT_DEFINITIONS attribute of a Version subclass can be used to define the specific parameters of the version scheme that is represented by that class.
The library also contains some implementations of specific version schemes including Python, PEP 440, PostgreSQL, and X.org versioning. More are sure to be added, and submissions of version shemes for popular, public projects/systems are welcome. These implementations also serve as examples for those wishing to subclass Version for their own (or another’s) version scheme.