[reportlab-users] errors parsing & > in paragraphs

[Andy Robinson]

> 
> it appears ther emight be some kind of inconsistency between the windows 
> and source release? can anyone verify this by testing it.

> my freebsd installation simply chops out the & (also not really 
> desirable) but windows actually errors
> 

If they are different the most likely case is that you have
a difference in which extensions are installed.  Can you do
'import pyRXP' and 'import _rl_accel' in both?  In an ideal
world the pure python implementation should do the same as the C ones,
but it probably didn't at that time, and I suspect you have
an extension in one which is not in the other.

IMHO a paragraph with naked >,  and & SHOULD cause an error, because
paragraph source has always been treated as XML and there are
lots of genuine tags you can embed.  But we were not strict about it 
so usage has varied.  

The daily builds, which are a long way past 1.20 now, have had
all of this reworked extensively and should fix the problem.
We're long overdue for a 1.21 release, although project pressures
mean it will be nearer the end of the year.

Best Regards,

Andy