> I am not sure if this is the right mailing list to ask these
This list is exactly the right list to ask such questions!
> The XCQL is only a segment of an XML document. Studying the
> XCQL schema,
> I think for CQL statements that start with prefixes, like the
> example in
> that page, it seems not possible to create a valid XML document to
> contain the statement. In another word, the XCQL created only can be a
> part of a bigger XML document.
> Does this mean that XCQL is intended to be embedded in a SOAP
> message or
> other types of XML documents, rather than be contained in a single XML
More or less - the XCQL schema does not define a top level element per
se. What it does define however is an xml schema type (namely
operandType) which can contain the XCQL fragment - i.e. a top level
element should be of this type.
So in SRW schema we have such a top level element xQuery to contain the
<xsd:element name="xQuery" type="xcql:operandType" nillable="false" />
> Another question is about how CQL is expected to be used. Right now
> there are several CQL parsers/compilers but most of them compiles a
> statement into XCQL. This makes sense since it's easier to parse and
> analyze an XCQL document semantically. However, how this
> query should be
> used to retrieve results from the local index or databases is not
> defined. Is this left for the SRW/SRU implementers to decide?
Yes. Our JAFER implementation at Oxford does does go via XCQL as we can
manipulate that via XSLT etc. but this seems to be the exception rather
than the rule and most others just go straight from CQL to the database
translation layer. As a result XCQL may be deprecated (it is no longer
possible for a client to send XCQL to a server directly which was
possible in SRW 1.0; and there have been calls to drop it entirely...)