docs: mention 'rb' needed to get readable ZipFile file object

[paolopiace]

It works OK with:
wdocx = Document("wfile.docx")

It fails with:
fobj = open("wfile.docx",'r')
wdocx = Document(fobj)

Error Log:
File "/usr/local/lib/python3.5/site-packages/python_docx-0.8.5-py3.5.egg/docx/api.py", line 25, in Document
document_part = Package.open(docx).main_document_part
File "/usr/local/lib/python3.5/site-packages/python_docx-0.8.5-py3.5.egg/docx/opc/package.py", line 116, in open
pkg_reader = PackageReader.from_file(pkg_file)
File "/usr/local/lib/python3.5/site-packages/python_docx-0.8.5-py3.5.egg/docx/opc/pkgreader.py", line 32, in from_file
phys_reader = PhysPkgReader(pkg_file)
File "/usr/local/lib/python3.5/site-packages/python_docx-0.8.5-py3.5.egg/docx/opc/phys_pkg.py", line 101, in init
self._zipf = ZipFile(pkg_file, 'r')
File "/usr/local/lib/python3.5/zipfile.py", line 1026, in init
self._RealGetContents()
File "/usr/local/lib/python3.5/zipfile.py", line 1093, in _RealGetContents
raise BadZipFile("File is not a zip file")
zipfile.BadZipFile: File is not a zip file

[scanny]

Hmm, interesting. A few questions to help with the diagnosis:

• What version of Python are you running?
• What do you get when you print fobj (something like <file object at 0x0...>)? In other words, what kind of object is it?
• What does fobj return for isinstance(fobj, str)?

[scanny]

Also, what OS are you running on. In Windows, if I recall correctly, you need 'rb' for the file mode, not just a plain 'r'. In any case, it's worth trying adding the 'b'. A Zip file is binary, not text.

[paolopiace]

I am using Python 3.5.1 on Ubuntu 12.04

'rb' was the trick. Thanks a lot!

May I suggest to say that in section 2.3.3 of python-docx Documentation

[scanny]

Glad to hear it @paolopiace :)

Is this the section you meant in the documentation?
http://python-docx.readthedocs.org/en/latest/user/documents.html#opening-a-file-like-document

By the way, can you describe your use case? We thought loading from an open file like this would be useful, but haven't learned of too many specific cases in which it is. If you could describe yours, that would help us understand better how it's being used :)

[paolopiace]

Yes, that's the section. It's 2.3.3 in the PDF.
I do not need to open and read the file object explicitly. Providing the file name directly is fine with me.
I reported what I thought to be an issue, which is not.

I'm parsing many table with technical content. All parsing is based on regex.

If I can ask one thing, it would be a method to iterate across the documents, like the function iter_block_items() described in issue 114.

[scanny]

There's a fresh workaround function for the iter_block_items() here if that helps:
#276 (comment)

[scanny]

This is fixed and will appear in the upcoming release.