Re: [Exist-open] Large DOM processing

[Wolfgang M. <wol...@ex...>]

> Does the embedded version of Exist still scale with very large XML documents being parsed as DOM? In other words, can one work  with the DOM API but have
> eXist still manage the serialization/deserialization for very large XML documents?

eXist keeps the persistent DOM on disk. DOM operations will only load
the node they are processing, never the document as a whole. This
makes it possible to work with large documents.

However, eXist is mainly designed to evaluate queries based on indexes
and not by traversing the DOM. From this perspective, it tries to
avoid accessing the DOM wherever possible. Walking the DOM tree will
usually be much slower than doing an XQuery to select the nodes you
want. My recommendation would thus always be to do as much as possible
in XQuery and only use DOM operations to process query results.

Wolfgang