Glue 0.7.1 cannot handle FITS with missing END card on Windows 7

[astrofrog]

I have Glue 0.7.1 on Windows 7 (Astropy version there is 1.1.2). When I try to open test image, L1448_13CO.fits, provided by @astrofrog, I get the following traceback:

Error message: Header missing END card.

Traceback (most recent call last):
  File "...\glue\dialogs\data_wizard\qt\data_wizard_dialog.py", line 36, in data_wizard
    result = gdd.load_data()
  File "...\glue\utils\qt\decorators.py", line 23, in result
    return func(*args, **kwargs)
  File "...\glue\dialogs\data_wizard\qt\data_wizard_dialog.py", line 105, in load_data
    d = load_data(path, factory=fac.function)
  File "...\glue\core\data_factories\helpers.py", line 237, in load_data
    d = as_list(factory(path, **kwargs))
  File "...\glue\core\data_factories\fits.py", line 56, in fits_reader
    hdulist = fits.open(source)
  File "...\astropy\io\fits\hdu\hdulist.py", line 138, in fitsopen
    return HDUList.fromfile(name, mode, memmap, save_backup, cache, **kwargs)
  File "...\astropy\io\fits\hdu\hdulist.py", line 280, in fromfile
    save_backup=save_backup, cache=cache, **kwargs)
  File "...\astropy\io\fits\hdu\hdulist.py", line 838, in _readfrom
    hdu = _BaseHDU.readfrom(ffo, **kwargs)
  File "...\astropy\io\fits\hdu\base.py", line 423, in readfrom
    **kwargs)
  File "...\astropy\io\fits\hdu\base.py", line 483, in _readfrom_internal
    header = Header.fromfile(data, endcard=not ignore_missing_end)
  File "...\astropy\io\fits\header.py", line 460, in fromfile
    padding)[1]
  File "...\astropy\io\fits\header.py", line 529, in _from_blocks
    raise IOError('Header missing END card.')
IOError: Header missing END card.

I did not see this error on Linux. And others who used the same file on Mac also did not mention seeing this error. On Windows, explicitly choosing "FITS file" from file type also did not help.

I also failed to load w5.fits, w5_psc.vot, and chandra_events.fits on Windows.

[astrofrog]

This used to have a fix that I accidentally undid in #825 - attached code to fix back

[astrofrog]

@pllim - could you confirm that this changes fixes the issue on your Windows machine?

[pllim]

Just an observation -- If ignore_missing_end is also provided via kwargs, there is going to be conflict.

[pllim]

I don't have access to the Windows laptop today; I will test this on Monday. Thanks!

[pllim]

@astrofrog , so it turns out I originally failed to open my files because they were corrupted. After I re-downloaded them, they can open fine on Windows even without this fix.

However, with this fix, I got a more meaningful IOError: Empty or corrupt FITS file message instead of an obscure missing END card.

So, it is up to you whether you still want to apply this or not. I apologize for the confusion!

[astrofrog]

The travis failure is unrelated