[reportlab-users] VALIGN=TOP, LongTables containing new Para with superscript

Chad Miller Chad.Miller at veritas.com
Fri Jun 24 13:43:04 EDT 2005

> -----Original Message-----
> From: Robin Becker [mailto:robin at reportlab.com]
> Sent: Thursday, 23 June, 2005 13:18
> Chad Miller wrote:
> > I narrowed-down a simplest-case example for a bug I've seen 
> > in using Debian's version 1.19debian-0.2.  With VALIGN set to 
> > TOP or MIDDLE, text before a "super" tag will be pushed 
> > outside the table frame iff there is exactly one line in the 
> > paragraph in the table cell.
> >  
> > Attached is the code (which writes a "./test.pdf" file) and 
> > its output.
> >  
> > Please look over the my code for obvious mistakes, before I 
> > file a bug report.
> >  
> > - chad
> Robin Becker wrote: 
> ... I'm not sure this is a bug. We don't vary the height of 
> lines with content ie lines are assumed to fit in the allotted 
> space. If a single character is exceptional then we will have 
> a problem. If using super the character goes too far up it may 
> collide with something. The table packing algorithm is particu-
> larly simple so it just assumes all lines are what they say 
> they are.

The PDF output clearly shows a bug.  For a cell containing a 
paragraph, if the content has some characteristic, then the cell
contents pop several inches outside of the table.

Look again at this PDF.  The text "strange" at the /bottom/ is 
from table #2.  That's about five inches away!  Surely no one 
has any reason to expect that behavior.

- chad
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test.pdf
Type: application/octet-stream
Size: 2600 bytes
Desc: test.pdf
Url : http://two.pairlist.net/pipermail/reportlab-users/attachments/20050624/b4531086/test.obj
-------------- next part --------------
A non-text attachment was scrubbed...
Name: testcase.py
Type: application/octet-stream
Size: 2837 bytes
Desc: testcase.py
Url : http://two.pairlist.net/pipermail/reportlab-users/attachments/20050624/b4531086/testcase.obj

More information about the reportlab-users mailing list