From: W. Eliot Kimber (eliot@isogen.com)
Date: Mon Nov 04 2002 - 16:17:31 PST
Attached is the usual FO instance and pair of PDFs generated therefrom
by XEP and XSL Formatter.
This set of samples is an attempt to explore most of the useful or
interesting variations in table width and column width markup. I think
that XEP is producing non-conforming results in a few cases and is
failing to report errors or warnings in a few cases.
In addition to table geometry problems, I noticed these problems:
- space-before on table-and-caption isn't honored between tables (but
appears to be when table follows fo:block.
- Keep-with-next between table-caption and table is not being honored
(see tables 13 and 15).
Also, this sample uses explicit inline progression dimension on
fo:inline, which XSL Formatter honors and XEP does not. It's used to get
the hanging indent for the table caption.
Here are the possible problems I've identified, based on my reading of
the XSL and CSS rules for table width calculation. In general, XSL
Formatter gives what I would consider to be the correct result when
there is variance with XEP's results (I've identified those places where
XSL Formatter appears to be in error).
Tables 1 and 2: Tables should be only as wide as required to contain
text of narrowest column without wrapping, with other columns the same
width. Thus the table should not be as wide as the body area.
Table 6: Table should only be as wide as required to contain the cell
content, not body-wide.
Table 7: XEP's rendering is correct, XSL Formatter's is wrong (too much
width in first column).
Table 9: XEP's is correct.
Tables 10, 11, 12: Their markup is erroneous by the FO spec but neither
XEP nor XSL Formatter issue warnings or errors. Both provide identical
fallback behavior. I actually don't know why the FO spec says it's an
error to use proportional column widths with explicit IPD but
table-layout="auto" as there's clearly no ambiguity about the table
width. But the spec says it's an error.
Table 13: Caption isn't kept with table body.
Table 14: XEP's response to limited horizontal width in this case is not
what I would prefer. I think XSL Formatter's response in this case is
more appropriate, especially given the presence of punctuation in the
text (e.g., "-").
Table 15: Another keep failure (but the colors are rendered very nicely).
Cheers,
Eliot
-- W. Eliot Kimber, eliot@isogen.com Consultant, ISOGEN International 1016 La Posada Dr., Suite 240 Austin, TX 78752 Phone: 512.656.4139 [David Tolpin: the file is too big for the list; the message bounced. I've approved the message but placed the file to http://www.renderx.com/~dvd/eliot/table-samples.zip ] ------------------- By using the Service, you expressly agree to these Terms of Service http://www.renderx.com/tos.html
This archive was generated by hypermail 2.1.5 : Wed Dec 18 2002 - 08:41:28 PST