initial commit 2

3 files changed, 198 insertions, 0 deletions

diff --git a/addons/base_sparse_field/models/__init__.py b/addons/base_sparse_field/models/__init__.py
new file mode 100644
index 00000000..37c0f934
--- /dev/null
+++ b/addons/base_sparse_field/models/__init__.py
@@ -0,0 +1,4 @@
+# -*- coding: utf-8 -*-
+
+from . import fields
+from . import models
diff --git a/addons/base_sparse_field/models/fields.py b/addons/base_sparse_field/models/fields.py
new file mode 100644
index 00000000..94242297
--- /dev/null
+++ b/addons/base_sparse_field/models/fields.py
@@ -0,0 +1,95 @@
+# -*- coding: utf-8 -*-
+
+import json
+
+from odoo import fields
+
+
+def monkey_patch(cls):
+    """ Return a method decorator to monkey-patch the given class. """
+    def decorate(func):
+        name = func.__name__
+        func.super = getattr(cls, name, None)
+        setattr(cls, name, func)
+        return func
+    return decorate
+
+
+#
+# Implement sparse fields by monkey-patching fields.Field
+#
+
+fields.Field.__doc__ += """
+
+        .. _field-sparse:
+
+        .. rubric:: Sparse fields
+
+        Sparse fields have a very small probability of being not null. Therefore
+        many such fields can be serialized compactly into a common location, the
+        latter being a so-called "serialized" field.
+
+        :param sparse: the name of the field where the value of this field must
+            be stored.
+"""
+fields.Field.sparse = None
+
+@monkey_patch(fields.Field)
+def _get_attrs(self, model, name):
+    attrs = _get_attrs.super(self, model, name)
+    if attrs.get('sparse'):
+        # by default, sparse fields are not stored and not copied
+        attrs['store'] = False
+        attrs['copy'] = attrs.get('copy', False)
+        attrs['compute'] = self._compute_sparse
+        if not attrs.get('readonly'):
+            attrs['inverse'] = self._inverse_sparse
+    return attrs
+
+@monkey_patch(fields.Field)
+def _compute_sparse(self, records):
+    for record in records:
+        values = record[self.sparse]
+        record[self.name] = values.get(self.name)
+    if self.relational:
+        for record in records:
+            record[self.name] = record[self.name].exists()
+
+@monkey_patch(fields.Field)
+def _inverse_sparse(self, records):
+    for record in records:
+        values = record[self.sparse]
+        value = self.convert_to_read(record[self.name], record, use_name_get=False)
+        if value:
+            if values.get(self.name) != value:
+                values[self.name] = value
+                record[self.sparse] = values
+        else:
+            if self.name in values:
+                values.pop(self.name)
+                record[self.sparse] = values
+
+
+#
+# Definition and implementation of serialized fields
+#
+
+class Serialized(fields.Field):
+    """ Serialized fields provide the storage for sparse fields. """
+    type = 'serialized'
+    column_type = ('text', 'text')
+
+    prefetch = False                    # not prefetched by default
+
+    def convert_to_column(self, value, record, values=None, validate=True):
+        return self.convert_to_cache(value, record, validate=validate)
+
+    def convert_to_cache(self, value, record, validate=True):
+        # cache format: json.dumps(value) or None
+        return json.dumps(value) if isinstance(value, dict) else (value or None)
+
+    def convert_to_record(self, value, record):
+        return json.loads(value or "{}")
+
+
+fields.Serialized = Serialized
diff --git a/addons/base_sparse_field/models/models.py b/addons/base_sparse_field/models/models.py
new file mode 100644
index 00000000..179f6fbd
--- /dev/null
+++ b/addons/base_sparse_field/models/models.py
@@ -0,0 +1,99 @@
+# -*- coding: utf-8 -*-
+
+from collections import defaultdict
+
+from odoo import models, fields, api, _
+from odoo.exceptions import UserError
+
+
+class Base(models.AbstractModel):
+    _inherit = 'base'
+
+    def _valid_field_parameter(self, field, name):
+        return name == 'sparse' or super()._valid_field_parameter(field, name)
+
+
+class IrModelFields(models.Model):
+    _inherit = 'ir.model.fields'
+
+    ttype = fields.Selection(selection_add=[
+        ('serialized', 'serialized'),
+    ], ondelete={'serialized': 'cascade'})
+    serialization_field_id = fields.Many2one('ir.model.fields', string='Serialization Field',
+        ondelete='cascade', domain="[('ttype','=','serialized'), ('model_id', '=', model_id)]",
+        help="If set, this field will be stored in the sparse structure of the "
+             "serialization field, instead of having its own database column. "
+             "This cannot be changed after creation.",
+    )
+
+    def write(self, vals):
+        # Limitation: renaming a sparse field or changing the storing system is
+        # currently not allowed
+        if 'serialization_field_id' in vals or 'name' in vals:
+            for field in self:
+                if 'serialization_field_id' in vals and field.serialization_field_id.id != vals['serialization_field_id']:
+                    raise UserError(_('Changing the storing system for field "%s" is not allowed.', field.name))
+                if field.serialization_field_id and (field.name != vals['name']):
+                    raise UserError(_('Renaming sparse field "%s" is not allowed', field.name))
+
+        return super(IrModelFields, self).write(vals)
+
+    def _reflect_fields(self, model_names):
+        super()._reflect_fields(model_names)
+
+        # set 'serialization_field_id' on sparse fields; it is done here to
+        # ensure that the serialized field is reflected already
+        cr = self._cr
+
+        # retrieve existing values
+        query = """
+            SELECT model, name, id, serialization_field_id
+            FROM ir_model_fields
+            WHERE model IN %s
+        """
+        cr.execute(query, [tuple(model_names)])
+        existing = {row[:2]: row[2:] for row in cr.fetchall()}
+
+        # determine updates, grouped by value
+        updates = defaultdict(list)
+        for model_name in model_names:
+            for field_name, field in self.env[model_name]._fields.items():
+                field_id, current_value = existing[(model_name, field_name)]
+                try:
+                    value = existing[(model_name, field.sparse)][0] if field.sparse else None
+                except KeyError:
+                    msg = _("Serialization field %r not found for sparse field %s!")
+                    raise UserError(msg % (field.sparse, field))
+                if current_value != value:
+                    updates[value].append(field_id)
+
+        if not updates:
+            return
+
+        # update fields
+        query = "UPDATE ir_model_fields SET serialization_field_id=%s WHERE id IN %s"
+        for value, ids in updates.items():
+            cr.execute(query, [value, tuple(ids)])
+
+        records = self.browse(id_ for ids in updates.values() for id_ in ids)
+        self.pool.post_init(records.modified, ['serialization_field_id'])
+
+    def _instanciate_attrs(self, field_data):
+        attrs = super(IrModelFields, self)._instanciate_attrs(field_data)
+        if attrs and field_data.get('serialization_field_id'):
+            serialization_record = self.browse(field_data['serialization_field_id'])
+            attrs['sparse'] = serialization_record.name
+        return attrs
+
+
+class TestSparse(models.TransientModel):
+    _name = 'sparse_fields.test'
+    _description = 'Sparse fields Test'
+
+    data = fields.Serialized()
+    boolean = fields.Boolean(sparse='data')
+    integer = fields.Integer(sparse='data')
+    float = fields.Float(sparse='data')
+    char = fields.Char(sparse='data')
+    selection = fields.Selection([('one', 'One'), ('two', 'Two')], sparse='data')
+    partner = fields.Many2one('res.partner', sparse='data')