collective.z3cform.datagridfield 0.4

Requirements

• Plone 4

• z3c.forms

• A browser with javascript support

• jquery 1.4.3 or later

Installation

Add collective.z3cforms.datagridfield to your buildout eggs.:

eggs=\
    ...
    collective.z3cforms.datagridfield

Example usage

This piece of code demonstrates a schema which has a table within it. The layout of the table is defined by a second schema.:

from zope import schema
from zope import interface
from plone.directives import form

form collective.z3cforms.datagridfield import DataGridFieldFactory

class ITableRowSchema(interface.Interface):
    one = schema.TextLine(title=u"One")
    two = schema.TextLine(title=u"Two")
    three = schema.TextLine(title=u"Three")

class IFormSchema(interface.Interface):
    four = schema.TextLine(title=u"Four")
    table = schema.List(title=u"Table"
        value_type=schema.Object(title=u"tablerow", schema=ITableRowSchema))

class EditForm(form.EditForm):
    extends(form.EditForm)

    grok.context(IFormSchema)
    grok.require('zope2.View')
    fields = field.Fields(IFormSchema)
    label=u"Demo Usage of DataGridField"

    fields['table'].widgetFactory = DataGridFieldFactory

Storage

The data can be stored as either a list of dicts or a list of objects. In both cases the schema content type is ‘schema.Object’.

If you are providing an Object content type (as opposed to dicts) you must provide your own conversion class. The default conversion class returns a list of dicts, not of your object class. See the demos.

Configuration

The widget can be customised via the updateWidgets method.

def updateWidgets(self):

super(EditForm, self).updateWidgets() self.widgets[‘table’].allow_insert = False # Enable/Disable the insert button on the right self.widgets[‘table’].allow_delete = False # Enable/Disable the delete button on the right self.widgets[‘table’].auto_append = False # Enable/Disable the auto-append feature

The widget contains an attribute ‘columns’ which is manipulated to hide column titles.

Manipulating the Sub-form

The DataGridField makes use of a subform to build each line. The main DataGridField contains a DataGridFieldObject for each line in the table. The DataGridFieldObject in turn creates the DataGridFieldObjectSubForm to store the fields.

There are two callbacks to your main form:

datagridInitialise(subform, widget)

• This is called when the subform fields have been initialised, but before the widgets have been created. Field based configuration could occur here.

datagridUpdateWidgets(subform, widgets, widget)

• This is called when the subform widgets have been created. At this point, you can configure the widgets, e.g. specify the size of a widget.

Notes

I have attempted to keep the markup close to Products.DataGridField, so that the styling approach is the same.

TODO

• Testing

• The mechanism to hide/omit columns is cumbersome. This needs to be revisited.

• Better styling of table in read-only mode. Merge the display and input versions of the layout.

• Move rows up / down

• kss validation - highlight the incorrect row, rather than the entire widget. Possibly show the error message in the field cell.

• hidden mode

Demo

Examples of configurations are in the demo folder. Once you install this package, the demo views will be visible on your site.

* http://localhost:8080/Plone/@@demo-collective.z3cform.datagrid

References

• http://pypi.python.org/pypi/Products.DataGridField

0.4 (2011-02-06)

• Renamed the demo pages. The starting point is now @@demo-collective.z3cform.datagrid .

• The widget can now be configured via the updateWidgets method. It is no longer necessary to create a custom factory.

• The columns can now be omitted.

• Provide a set of demo views for Object access.

0.3 (2011-02-04)

• The auto-append functionality did not bind correctly for popup forms. I switched to using jQuery.live() instead of binding at document load time.

• Added a menu to the demo pages

• Added a display only form option.

• Fixed the restructured text of the main README.txt so that it will show more friendly in PyPI.