EFFICIENT STORAGE OF INFORMATION FROM MARKUP LANGUAGE DOCUMENTS

An in-memory document model may be created from a markup language document while parsing the markup language document. The model includes small fixed-size memory structures allocated from a single larger memory pool. The model stores the data contained in the markup language document and the hierarchical relationship between the data items in the markup language document. Thus, random access to the markup language document is achieved without further access to the document and without the overhead of language-specific object construction. When an object-oriented computer program instance references the document model, a language-specific object may be constructed from the model including a pointer to an element of the model. The document model may be created when parsing extensible markup language (XML) documents.