PyQt custom MongoDB model - model-view-controller

Question

I'm developing an app using PyQt with MongoDB as a backend. There is a table (QTableView) in my application that should be populated with data from MongoDB and I would like to use model->view architecture. Since Qt doesn't have a model for MongoDB I need to write a custom one.
This data is currently (can be changed to fit this problem) organized as a list of dictionaries, like this (actual data is more complex):
[{"name":"some name","phone":"283891273218"}, {"name":"some other name","phone":"56958656556"}]
Each dictionary represents a row, and each key of a dictionary is a column. After a few hours of searching I got almost nowhere, and the code is:
class CustomModel(QtCore.QAbstractTableModel):
def __init__(self, parent=None, *args):
super(CustomModel, self).__init__()
self.datatable = None
def update(self, dataIn):
print 'Updating Model'
self.datatable = dataIn
#print 'Datatable : {0}'.format(self.datatable)
def rowCount(self, parent=QtCore.QModelIndex()):
return len(self.datatable)
def columnCount(self, parent=QtCore.QModelIndex()):
return len(self.datatable[0])
def data(self, index, role=QtCore.Qt.DisplayRole):
if role == QtCore.Qt.DisplayRole:
i = index.row()
j = index.column()
#print self.datatable
else:
return QtCore.QVariant()
def flags(self, index):
return QtCore.Qt.ItemIsEnabled
If I understand correctly, method data should fill the rows and columns with data but I don't know how to do it. This code is currently generating the correct number of rows and columns without data.
I would appreciate any help or advice.

Answer 1

A good thing to start with is to ensure a consistent column ordering when pulling data from the source. Since the data sources rows are dictionaries in this case, you are not guaranteed a consistent ordering when iterating (indexing is not available anyhow). To ensure consistent ordering, setup an ordered list of column indices that map to the data sources keys:
class CustomModel(QtCore.QAbstractTableModel):
columns = ['name', 'phone']
If you want to re-use this custom model for wrapping other data sets, you probably don't want to couple the column names so tightly with the class. It could easily be made more generic by using a required instance variable passed to the initializer:
class CustomModel(QtCore.QAbstractTableModel):
def __init__(self, columns, parent=None):
super(CustomModel, self).__init__(parent)
self.columns = columns
self.datatable = []
...
model = CustomModel(['name', 'phone'])
The column names can then be used to determine the columnCount which is also more robust because len(self.datatable[0]) will fail with an empty model:
def columnCount(self, parent=QtCore.QModelIndex()):
return len(self.columns)
Furthermore, you can use the ordered columns class variable for use in a default implementation of headerData():
def headerData(self, section, orientation, role=QtCore.Qt.DisplayRole):
if orientation == QtCore.Qt.Horizontal and role == QtCore.Qt.DisplayRole:
return self.columns[section].title()
Finally, use the columns list as a lookup for converting a model column index into a dictionary key for use in the source data and return the source data value from the data() method:
def data(self, index, role=QtCore.Qt.DisplayRole):
if role == QtCore.Qt.DisplayRole:
row = self.datatable[index.row()]
column_key = self.columns[index.column()]
return row[column_key]
else:
return None
(Note PyQt4 and PySide don't need to use QVariant AFAICT)