PQL - Path Query Language

top | about (long) | download | documentation | source

The latest release of PQLBRL is 0.4.1.


PQL, or Path Query Language or "pickle", is a database query language designed for working with graph-structured data. Specifically, it is for handling ordinary scalar data contained in a database whose structure is a graph of objects and relationships instead of a system of tabular relations. (Other tools and languages exist for other graph-related database problems, such as storing and retrieving graphs as data or computing transformations on graphs.) The syntax of PQL is similar to SQL, with additional bits drawn from familiar sources such as Unix regular expressions. This makes it familiar and reasonably legible. more

PQLBRL, the PQL Base Reference Library ("picklebarrel") is the implementation of the PQL engine. PQLBRL is not a DBMS or even a database engine; it is just a query engine. This allows it to be plugged into arbitrary storage backends and arbitrary query submission frontends. Currently PQLBRL is used as part of the custom provenance querying architecture in the PASS project.

Project Members


(in progress)

Also see the language documentation.

Source Code

The current release of the PQLBRL implementation is 0.4.1. You can download or browse the source.

The latest development version can be checked out via Mercurial. To check out a copy:

   hg clone http://www.eecs.harvard.edu/~dholland/hg/pqlbrl/
You can also browse the source or follow the RSS commit log.


Source of Support: NSF
Proposal Number: 0849392
Title: SGER: PQL: A Path Query Language