Another idea would be to save the doc in the db with a different file extension
so ML would automatically treat it as a binary. Or just change the mimetype for
HTML in ML to be binary.
Sent from my iPhone
On Jun 28, 2012, at 10:10 AM, "Michael Blakeley" <m...@blakeley.com> wrote:
> You could keep the documents as XML but exclude their root element from
> word-query indexing. There would still be XPath indexing but I estimate this
> would be 25-50% of the ordinary indexing done.
>
> Or yes, you could store them as binary. From XCC or xdmp:document-load that
> is as simple as setting the format. From XQuery with constructed XML it might
> be a little harder. But try this:
>
> binary {
> xs:hexBinary(
> xs:base64-binary(
> xdmp:base64-encode(
> xdmp:quote($xml))) }
>
> That looks ugly, but seems to be reasonably efficient. I think the UTF8 bytes
> will end up in the database, so there should be no need for output
> conversion. Remember to set the right content-type, though.
>
> -- Mike
>
> On Jun 28, 2012, at 8:32, Tim Finney <t...@tfinney.net> wrote:
>
>> Hi All,
>>
>> I'd like to know how to minimize the overhead associated with reading
>> and writing an HTML document to a MarkLogic database. I want to create a
>> cache of previously generated HTML pages to improve response times. Is
>> there a way to, say, make the server think an HTML page is a binary and
>> therefore not do any indexing? Or is there another strategy that would
>> give me good read and write performance and avoid loading the server
>> with indexing that will never be used?
>>
>> Best,
>>
>> Tim Finney
>>
>>
>> _______________________________________________
>> General mailing list
>> General@developer.marklogic.com
>> http://community.marklogic.com/mailman/listinfo/general
>>
> _______________________________________________
> General mailing list
> General@developer.marklogic.com
> http://community.marklogic.com/mailman/listinfo/general
_______________________________________________
General mailing list
General@developer.marklogic.com
http://community.marklogic.com/mailman/listinfo/general
