The Galago query language presented here is similar to query languages used
in open source search engines that are based on the inference network retrieval
model. This version focuses on the most useful aspects of those languages for a
variety of search applications, and adds the ability to use arbitrary features. Note
that the Galago search engine is not based on a specific retrieval model, but instead
provides an efficient framework for implementing retrieval models.
Although the query language can easily handle simple unstructured text documents,
many of the more interesting features make use of evidence based on document
structure. We assume that structure is specified using tag pairs, as in HTML
or XML. Consider the following document: