[poppler] Loading documents from GInputStream?

[Leonard Rosenthol]

On May 12, 2008, at 12:08 PM, Tommi Komulainen wrote:
> Well, not entirely. It would still allow getting started with
> processing the document earlier,

	Not unless the file is linearized.  If it is not, then you have to  
download the entire document anyway - since you need the xref (and  
it's at the end of the file).


> and if the document happens to be
> linear the first pages could be displayed faster than waiting for the
> whole document to finish downloading first before starting up the PDF
> reader.
>
	True - provided you modified poppler/Xpdf to support reading the  
linearization tables AND THEN modified the browser plugin that wraps  
poppler/Xpdf to handle byte serving requests along with modifications  
to your GStream to integrate the fetches (since it won't be a linear  
stream any longer - it will be a dynamic random-access set of fetches).

Leonard