[NumPy-Tickets] [NumPy] #1071: loadtxt fails if the last column contains empty value

[NumPy Trac]

#1071: loadtxt fails if the last column contains empty value
---------------------------------+------------------------------------------
 Reporter:  Electrion            |       Owner:  somebody    
     Type:  defect               |      Status:  needs_review
 Priority:  normal               |   Milestone:  1.6.0       
Component:  numpy.lib            |     Version:  devel       
 Keywords:  loadtxt ascii strip  |  
---------------------------------+------------------------------------------

Comment(by derek):

 Replying to [comment:7 bsouthey]:
 > Replying to [comment:5 derek]:
 > If I just use 'line = line.split(comments)[0].strip('\r\n')' in version
 2.0.0.dev-7134a93 then this example now works for me.
 Not converting the string with {{{asbytes('\r\n')}}}  raises an error for
 me with Python3, and without the blank it fails this test:
 {{{
 ERROR: Test using an explicit dtype with an object
 ----------------------------------------------------------------------
 Traceback (most recent call last):
   File "/sw/lib/python2.6/site-packages/numpy/lib/tests/test_io.py", line
 422, in test_dtype_with_object
     converters=converters)
   File "/sw/lib/python2.6/site-packages/numpy/lib/npyio.py", line 784, in
 loadtxt
     items = [conv(val) for (conv, val) in zip(converters, vals)]
   File "/sw/lib/python2.6/site-packages/numpy/lib/npyio.py", line 569, in
 <lambda>
     return lambda x: int(float(x))
 ValueError: empty string for float()
 }}}
 because the test passes two blank lines as input. It is not explicitly
 documented that loadtxt silently skips blank lines, and it could raise
 other issues because empty lines have been proposed to delimit different
 data blocks in #1107.
 But if we want to preserve this functionality, the better fix I now
 realise is
 {{{
 @@ -724,8 +728,8 @@ def loadtxt(fname, dtype=float, comments='#',
 delimiter=None,

      def split_line(line):
          """Chop off comments, strip, and split at delimiter."""
 -        line = asbytes(line).split(comments)[0].strip()
 +        line = asbytes(line).split(comments)[0].strip(asbytes('\r\n'))
 +        if line.strip():
              return line.split(delimiter)
          else:
              return []
 }}}

-- 
Ticket URL: <http://projects.scipy.org/numpy/ticket/1071#comment:8>
NumPy <http://projects.scipy.org/numpy>
My example project