marcoconti83 / read-ods-with-odfpy

Read an ODS file with Python and Odfpy

read-ods-with-odfpy

As seen on: http://www.marco83.com/work/173/read-an-ods-file-with-python-and-odfpy

Odfpy is a python library to read and write OpenDocument documents (such as the .odt or.ods created with LibreOffice or OpenOffice). However, the documentation and examples shipped with Odfpy are more oriented to writing new documents rather than reading existing ones.

Failing to find any simple spreadsheet reading code snippet on the internet, I wrote a simple ODS reader in python that reads an entire .ods file in a dictionary of sheets, where each sheets is stored as an array of arrays (rows, columns). It still requires Odfpy to run.

It has been tested with odfpy 0.9.3 and 1.3.0, python 2.7 and 3.4, using ods files created with OpenOffice.org.

Usage example:

from ODSReader import ODSReader

doc = ODSReader(u'films.ods', clonespannedcolumns=True)
table = doc.getSheet(u'Sheet1')
for i in range(len(table)):
    for j in range(len(table[i])):
        print (table[i][j])

Requirements

• odfpy 0.9.3 or 1.3.0
• python 2.7 or 3.4

read-ods-with-odfpy's Issues

Wrong object used in line 72

Lines 71/72 currently read:

if (c.nodeType == 3):
     textContent = u'{}{}'.format(textContent, n.data)

However, I think they should read:

if (c.nodeType == 3):
    textContent = u'{}{}'.format(textContent, c.data)

Otherwise, ODSReader crashes immediately while trying to parse an ODS file with the Exception:

File "xxx/ODSReader.py", line 72, in readSheet
    textContent = u'{}{}'.format(textContent, n.data)

The parser works with the fix suggested above.

What about reading numbers?

Your example works great for reading text, but there seems to be no example for reading numbers. It would be nice if there was!

Rename odf-to-array.py to ODSReader.py

In examples, code

from ODSReader import ODSReader

not work before renaming odf-to-array.py to ODSReader.py.
I think it is need to rename file in git repository or change examples.

return a pandas data frame

Can you change the code in a way to return a pandas dataframe instead?

spreadsheet is a text object?

I've been struggling with odfpy and every library that depends on it for a few hours.

line 35:
for sheet in self.doc.spreadsheet.getElementsByType(Table):
gives an error:
AttributeError: 'Text' object has no attribute 'getElementsByType'

remove 'spreadsheet' seems to fix the problem.
new line 35:
for sheet in self.doc.getElementsByType(Table):

I don't know if this is just fixing the symptom but has solved all my current problems.