Fails to load Pivot without a dxfId

Created originally on Bitbucket by uluttner (Ulrich Luttner)

I came across xlsx files with PivotTables that couldn't be opened for writing.

load_workbook fails:

#!python


---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
C:\anaconda3\lib\site-packages\openpyxl\descriptors\base.py in _convert(expected_type, value)
     56         try:
---> 57             value = expected_type(value)
     58         except:

TypeError: int() argument must be a string, a bytes-like object or a number, not 'NoneType'

During handling of the above exception, another exception occurred:

TypeError                                 Traceback (most recent call last)
C:\anaconda3\lib\site-packages\openpyxl\reader\excel.py in load_workbook(filename, read_only, keep_vba, data_only, guess_types, keep_links)
    284                     src = archive.read(pivot_path)
    285                     tree = fromstring(src)
--> 286                     pivot = TableDefinition.from_tree(tree)
    287                     pivot.cache = pivot_caches[pivot.cacheId]
    288                     ws.add_pivot(pivot)

C:\anaconda3\lib\site-packages\openpyxl\descriptors\serialisable.py in from_tree(cls, node)
     78             if hasattr(desc, 'from_tree'):
     79                 #descriptor manages conversion
---> 80                 obj = desc.from_tree(el)
     81             else:
     82                 if hasattr(desc.expected_type, "from_tree"):

C:\anaconda3\lib\site-packages\openpyxl\descriptors\sequence.py in from_tree(self, node)
     84
     85     def from_tree(self, node):
---> 86         return [self.expected_type.from_tree(el) for el in node]
     87
     88

C:\anaconda3\lib\site-packages\openpyxl\descriptors\sequence.py in <listcomp>(.0)
     84
     85     def from_tree(self, node):
---> 86         return [self.expected_type.from_tree(el) for el in node]
     87
     88

C:\anaconda3\lib\site-packages\openpyxl\descriptors\serialisable.py in from_tree(cls, node)
     98                 attrib[tag] = obj
     99
--> 100         return cls(**attrib)
    101
    102

C:\anaconda3\lib\site-packages\openpyxl\pivot\table.py in __init__(self, action, dxfId, pivotArea, extLst)
    501                 ):
    502         self.action = action
--> 503         self.dxfId = dxfId
    504         self.pivotArea = pivotArea
    505         self.extLst = extLst

C:\anaconda3\lib\site-packages\openpyxl\descriptors\base.py in __set__(self, instance, value)
     67         if ((self.allow_none and value is not None)
     68             or not self.allow_none):
---> 69             value = _convert(self.expected_type, value)
     70         super(Convertible, self).__set__(instance, value)
     71

C:\anaconda3\lib\site-packages\openpyxl\descriptors\base.py in _convert(expected_type, value)
     57             value = expected_type(value)
     58         except:
---> 59             raise TypeError('expected ' + str(expected_type))
     60     return value
     61

TypeError: expected <class 'int'>

It's pretty much the same problem as described for tableStyleElement in issue #902 (closed):

I changed in pivot/table.py:

dxfId = Integer()

to

dxfId = Integer(allow_none=True)

and gave it a try.

Seemed to work, I could write my files with the Pivots just fine.

I don't know if this may have other implications, though.

Would be good to have this fixed. Thanks ever so much, also for your cool library.

Attachments: test.xlsx