[gs-devel] beginpage and the vector device

[Ralph Giles]

Ken,

Sorry this is a bit long, but I wanted to check with you before I 
started patching things. I've done a little hacking on an SVG output 
device for Ghostscript, and ran into something silly. My understanding 
so far:

We have gx_device_vector which is supposed to track changes in graphics 
state and call a set of notification methods on itself (or rather a 
subclass) when drawing happens. By its own admission it's incomplete, 
but another thing it provides is an output stream for the device, to 
simplify additional manipulation later. I've been using that for 
convenience and conformance.

I ran into trouble with my beginpage method not being called. I was 
relying on this to write the file header if it hadn't been written 
already, because there's no 'begindocument' and it can't be done from 
device_open because the client needs to call put_params to set the 
filename, etc. before the stream can be opened.

As it turns out, the way gx_device_vector decides when to call
beginpage is by testing vdev->in_page INSIDE AN ACCESSOR for
the output stream. So, the subclass is expected to call 
gdev_vector_stream() whenever it wants to write EXCEPT from 
inside the beginpage method, where it must access the stream
directly. The subclass is also expected to set in_page back to 
false after finishing a page if it wants its beginpage method 
called again. Essentially the same hack I was doing, but half
a level up. And gdevpx.c does all this in implementing the 
pxl output device. Whacky.

I think someone was just trying to save some code by blurring a 
boundary, which is unfortunately common in Ghostscript, and I'd
like to fix it. Do you have any thoughts before I try and rework
this? Is there a general page_begin event that I've missed? If
not I guess we just have to make all the gdevvec.c routines check
in_page themselves. More code, but less surprise.

 -r