On 17 Nov 2008, at 14:17, pedro wrote:
Hi
my target is to improve the speed of my print process.
I have a report tha is created by an AreaTreeProcessor and then is
displayed by Acrobat and is too slow.
I debug the application and i see that the creation process have
appreciable lifetime,
You mean the creation of the area tree? So that the issue is about the
speed of the area-tree-to-pdf rendering?
In that case, there are known performance drawbacks in parsing the
area tree (mainly due to a large number of Map lookups, some
suboptimal parsing of the values into Trait objects, and the
relatively large number of container-objects, IIC).
This is precisely one of the key reasons why Jeremias is currently
working on an alternative approach for a FOP-native Intermediate
Format. If I've been following correctly, then the branch in question
is nearing completion, and will be merged back into the trunk in the
very near future.
Read some more at:
http://wiki.apache.org/xmlgraphics-fop/AreaTreeIntermediateXml/NewDesign
In short: I'm not sure whether there is an immediate solution for your
problem right away.
On top of that, IIRC, for standard PDF files, the reader would still
need the whole file before it is able to say anything about the first
page. In basic PDF, the document catalog dictionary, which contains
the info about the document structure, is the final object in the
stream.
For local files the delay for loading the file in-memory (or reading
ahead) is at most a few seconds, but for remote access, if you have
large PDF streams and a bad connection, this can easily reach the
magnitude of minutes.
Cheers
Andreas
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
