As a matter of specification, to be valid itself, a schema must import all the namespaces it references. In the <import> element it need not specify where to find the imported schema, and if it does specify it, the specification may be ignored by a validator that can access the grammar another way.
As a matter of implementation of the validator, the validator may find an imported schema in a variety of ways. One of these is to use a grammar for the namespace that it already holds, e.g. in a Xerces grammar pool. Thus, in (pre)parsing a schema with an import, a parser with a grammar pool does not necessarily need to open a URL and read a serialized schema in order to build the importing grammar. If a Xerces parser is configured to use a grammar pool that contains the imported schema's grammar, the imported grammar should be used by default, and not need to be parsed again. Jeff ----- Original Message ----- From: "Guy Sharon" <[EMAIL PROTECTED]> To: <[email protected]> Sent: Thursday, June 23, 2005 4:08 AM Subject: Preparse with existing Grammar Pool | Hi, | | I am interested in extending my grammar pool during the execution of my | program. So in the beginning I want to preparse a schema for grammar and | later receive another schema to have its grammar be added to the pool. | However, the additional schemas all reference elements that were defined in | the first schema. | | Is it possible to preparse the additional schemas without each importing the | first schema but rather validate via the existing GrammarPool? --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
