PyExcerlerator details

[A_H]

Hi, I'm using PyExcelerator, and it's great, but I can't figure out a
few things:


(1) I set the cell style to '0.00%' but the style does not work.

(2) I want to place a border around the cells( x1, y1, x2, y2 ) but I
can't find any example of doing that.
Well I do see ONE example, but it erases the cells if I do it after,
or when I write to the cells that erases the outline.
Surely I don't have to tediously set each cells border?

Help?

 

[John Machin]

Hi, I'm using PyExcelerator, and it's great,

If you are using the latest released version, it's not, IMO. Reading the
fixed-later bug reports on Sourceforge may prompt you to get the latest
version from svn. Reading the unfixed bug reports on Sourceforge may
prompt you to switch to xlwt (i.e. "Excel write"), a fork of
pyExcelerator (fixed known and unknown bugs, improved performance, and
runs under Python 2.3) -- available from
https://secure.simplistix.co.uk/svn/xlwt/trunk

See also http://groups.google.com/group/python-excel [to which I've CCed
this reply].

but I can't figure out a
few things:


(1) I set the cell style to '0.00%' but the style does not work.

That isn't a style, it's a "number format". See num_formats.py in the
examples directory.

(2) I want to place a border around the cells( x1, y1, x2, y2 ) but I
can't find any example of doing that.
Well I do see ONE example, but it erases the cells if I do it after,
or when I write to the cells that erases the outline.

At the moment, neither pyExcelerator nor xlwt offer a style-setting
facility that's independent of the Worksheet.write method. It could well
be a useful enhancement.

Surely I don't have to tediously set each cells border?

Perhaps you could suggest what you think the API for setting a border on
a rectangle should be.

Each cell's border has to be set somehow. With a style_setting approach,
there are 8 different cases (4 sides and 4 corners). With an independent
border-setting approach, there would be only 4 different cases (4 sides)
.... this needs looking at to see how easily the existing data model
would support it.

Regards,
John