UnicodeDecodeError when accented chars are present in test result tables

[LeoRochael]

If a zuite contains accented chars, and those get transplanted into test results, the ZuiteResults object cannot render to html properly resulting in an UnicodeDecodeError:

Traceback (most recent call last):
  [...]
  File "/home/leo/opt/zelenium/eggs/Zope2-2.12.5-py2.6-linux-x86_64.egg/Shared/DC/Scripts/Bindings.py", line 324, in __call__
    return self._bindAndExec(args, kw, None)
  File "/home/leo/opt/zelenium/eggs/Zope2-2.12.5-py2.6-linux-x86_64.egg/Shared/DC/Scripts/Bindings.py", line 361, in _bindAndExec
    return self._exec(bound_data, args, kw)
  File "/home/leo/opt/zelenium/eggs/Zope2-2.12.5-py2.6-linux-x86_64.egg/Products/PageTemplates/PageTemplateFile.py", line 130, in _exec
    return self.pt_render(extra_context=bound_names)
  File "/home/leo/opt/zelenium/eggs/Zope2-2.12.5-py2.6-linux-x86_64.egg/Products/PageTemplates/PageTemplate.py", line 80, in pt_render
    showtal=showtal)
  File "/home/leo/opt/zelenium/eggs/zope.pagetemplate-3.5.0-py2.6.egg/zope/pagetemplate/pagetemplate.py", line 115, in pt_render
    strictinsert=0, sourceAnnotations=sourceAnnotations)()
  File "/home/leo/opt/zelenium/eggs/zope.tal-3.5.2-py2.6.egg/zope/tal/talinterpreter.py", line 271, in __call__
    self.interpret(self.program)
  [...]
  File "/home/leo/opt/zelenium/eggs/zope.tal-3.5.2-py2.6.egg/zope/tal/talinterpreter.py", line 742, in do_insertStructure_tal
    structure = self.engine.evaluateStructure(expr)
  File "/home/leo/opt/zelenium/eggs/Zope2-2.12.5-py2.6-linux-x86_64.egg/Products/PageTemplates/Expressions.py", line 221, in evaluateStructure
    return self._handleText(text, expr)
  File "/home/leo/opt/zelenium/eggs/Zope2-2.12.5-py2.6-linux-x86_64.egg/Products/PageTemplates/Expressions.py", line 266, in _handleText
    return unicode(text)
UnicodeDecodeError: 'ascii' codec can't decode byte 0xc3 in position 271: ordinal not in range(128)

This is caused by the TALES engine trying to directly convert the OFS.Image.File object that represents a test table into unicode. Since File objects can never be unicode by definition, this is guaranteed to bork if there are accented chars in it.

The attached patch contains a test and a one-line work-around. Took me a minute to write the work-around, and the whole day for the test...

The work-around converts the OFS.Image.File to string first, giving the TALES unicode-conflict-resolver a chance to do its magic.

[Tres Seaver (tseaver)]

Fix released with version 1.0.2:

 http://pypi.python.org/pypi/Products.Zelenium/1.0.2