Re: HTML encoded spreadsheets?

On Wed, 2005-09-07 at 18:40 -0600, Andreas J. Guelzow wrote:
On Wed, 2005-07-09 at 18:06 -0500, David Stanaway wrote:
Also, when I open with the html format module, lots of errors scroll
about namespace errors. It is an html file with excel xml markup eg:


  <td class=xl29 width=71

x:num flags an error

I really think that x:num is not acceptable in a td tag in any version
of html. 

I agree it is not html, but it is I assume covered in the
xmlns:x="urn:schemas-microsoft-com:office:excel" namespace. (This I do
not expect to be supported by the html module, except by specific design
of course)

 and the css formatting is ignored, eg:

        mso-number-format:"Short Date";
        border:.5pt solid black;
        mso-pattern:auto none;

Do you really expect gnumeric to parse some random css and extract
formatting from it?

Sorry, Maybe my question came out wrong. I don't expect the html format
to support the nearly arbritary complexity that is possible in html.
This is however a predictable (?) subset of html/css and some extra xml
covered in the urn:schemas-microsoft-com:office:excel namespace etc. I
was just suggesting that maybe it would be good to support this html
hybrid format that ms seems to be generating in office 11 (Which I only
have limited access to).

FYI, I tried to open the file with the latest version of the free Excel
viewer from ms, and it bailed out saying that it did not support this
format. Maybe there is someone here that does have access to office 11
that can do a couple of quick tests to see what the format options for
save as are. Maybe my vendor is going out of their way to make a dodgy
file? EG, save as web page, then renaming the file to .xls and ignoring
the warning.

If in fact, it turns out that this format is automagically chosen by
excel in certain conditions (Very broken on ms's part), then there could
be a subset of .xls files that will be incompatible with gnumeric. This
would be pretty well fixed if the html format had a probe method, the
only loss would be formatting, which although important, is easier to
work around than html loaded with the text file module, as is currently
what seems to be the case.

David Stanaway <david stanaway net>

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]