[Octave-bug-tracker] [bug #36356] textscan: error using CollectOutput with uneven-length data columns

[Philip Nienhuis]

URL:
  <http://savannah.gnu.org/bugs/?36356>

                 Summary: textscan: error using CollectOutput with
uneven-length data columns
                 Project: GNU Octave
            Submitted by: philipnienhuis
            Submitted on: Mon 30 Apr 2012 09:53:43 PM CEST
                Category: None
                Severity: 3 - Normal
                Priority: 5 - Normal
              Item Group: Incorrect Result
                  Status: In Progress
             Assigned to: philipnienhuis
         Originator Name: Philip Nienhuis
        Originator Email: 
             Open/Closed: Open
         Discussion Lock: Any
                 Release: dev
        Operating System: Any

    _______________________________________________________

Details:

While investigating a bug report by Alexander Graf

https://mailman.cae.wisc.edu/pipermail/octave-maintainers/2012-April/028137.html

(under "another small problem"), I found that textscan.m always removes a
trailing "\n" before handing the data text string to strread().
If the last char of the data stream is a "\n", strread.m pads data columns
with 0x0[] or NaN so that all columns have the same size.
If strread.m gets data w/o trailing "\n", it leaves uneven-length data column
sizes as-is.
This of course is potentially the wrong thing for later CollectOutput
processing by textscan.m - combining columns of the same class requires all
columns to be of the same length.

I already have a fix + new test ready; I'm awaiting feedback from Alexander.




    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/bugs/?36356>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.gnu.org/