After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 317427 - xml parse problem "Sheet!$A:$A"
xml parse problem "Sheet!$A:$A"
Status: RESOLVED FIXED
Product: Gnumeric
Classification: Applications
Component: General
unspecified
Other All
: Normal normal
: ---
Assigned To: Jody Goldberg
Jody Goldberg
Depends on:
Blocks:
 
 
Reported: 2005-09-28 12:13 UTC by Jean Bréfort
Modified: 2005-09-30 09:50 UTC
See Also:
GNOME target: ---
GNOME version: ---



Description Jean Bréfort 2005-09-28 12:13:34 UTC
insert a name such as my_column with Sheet1!$A:$A as value

=sum(my_column) gives the correct result.

Save the file, close and reopen. The name has disappeared.
The file contains :

  <gnm:Names>
    <gnm:Name>
      <gnm:name>my_column</gnm:name>
      <gnm:value>Sheet1!$A:$A</gnm:value>
      <gnm:position>C7</gnm:position>
    </gnm:Name>
  </gnm:Names>
Comment 1 Andreas J. Guelzow 2005-09-28 13:00:00 UTC
I can replicate this. Note that the problems happens when the file is opened and
seems to be restricted to full columns.
Comment 2 Jody Goldberg 2005-09-28 13:08:57 UTC
losing data is a blocker.
I'll have a look on the flight back.
Comment 3 Morten Welinder 2005-09-28 21:13:11 UTC
Several problems:

1. "exp" (the function pointer) is not null.  Fixed.
2. io context needed to be set before names were read.  Fixed.
3. Discrepancy between xml expression conventions and gui ditto.  [Worked around]

---> downgrading.

We use gnm_1_0_rangeref_parse in xml_io_conventions.
gnm_expr_conventions_default (as used inside cell_set_text and now
xml_read_names) uses rangeref_parse.

I really don't know why we use gnm_1_0_rangeref_parse, but it does not
understand Sheet!$A:$A.  Inside regular expression we survive this since
we fall back to using cell_set_text.
Comment 4 Morten Welinder 2005-09-29 01:08:03 UTC
To elaborate:

either we really need to fix gnm_1_0_rangeref_parse to handle this kind of
reference, or else we need to switch to using rangeref_parse.
Comment 5 Jody Goldberg 2005-09-30 09:50:40 UTC
The current arrangement seems backwards.  We should export 1.0 and import
anything.  I'll do it on the flight back