odoo 关于self的迭代器（0）_for record in self

万恶之源：

    @api.depends("girls", "boys")
    def _compute_boys_per(self):
        for record in self:
            record.per = record.boys/(record.boys + record.girls)

这是一段典型的odoo的代码，作为初学者，我看了只有有点懵，在for循环内部对self和record做了print操作，发现居然是一样的，瞬间傻了。接下来的内容中，简单的说下。

class BaseModel:
    def __iter__(self):
        """ Return an iterator over ``self``. """
        for id in self._ids:
            yield self._browse((id,), self.env, self._prefetch)
 
 
    #
    # Instance creation
    # 创建实例
    # 
    # An instance represents an ordered collection of records in a given
    # execution environment. The instance object refers to the environment, and
    # the records themselves are represented by their cache dictionary. The 'id'
    # of each record is found in its corresponding cache dictionary.
    # 创建一个给定执行环境中记录的有序集合的实例。实例对象提供了环境，并且记录
    # 本身由缓存字典表示。每一个记录的id都在相应的缓存字典中找到
    #
    # This design has the following advantages:
    #  - cache access is direct and thus fast;
    #  - one can consider records without an 'id' (see new records);
    #  - the global cache is only an index to "resolve" a record 'id'.
    #
 
 
    @classmethod
    def _browse(cls, ids, env, prefetch=None):
        """ Create a recordset instance.
        :param ids: a tuple of record ids
        :param env: an environment
        :param prefetch: an optional prefetch object
        """
        records = object.__new__(cls)
        records.env = env
        records._ids = ids
        if prefetch is None:
            prefetch = defaultdict(set)         # {model_name: set(ids)}
        records._prefetch = prefetch
        prefetch[cls._name].update(ids)
 
 
 
 
    def browse(self, arg=None, prefetch=None):
        """ browse([ids]) -> records
        Returns a recordset for the ids provided as parameter in the current
        environment.
        Can take no ids, a single id or a sequence of ids.
        这个就是格式化ids的结构
        """
        ids = _normalize_ids(arg) 
        #assert all(isinstance(id, IdType) for id in ids), "Browsing invalid ids: %s" % ids
        return self._browse(ids, self.env, prefetch)
 
 
    @api.returns('self')
    def exists(self):
        """  exists() -> records
        Returns the subset of records in ``self`` that exist, and marks deleted
        records as such in cache. It can be used as a test on records::
		返回self中存在的记录的子集，并且标记缓存中删除的记录。他可以用来在记录上做测试。
            if record.exists():
                ...
        By convention, new records are returned as existing.
        """
        ids, new_ids = [], []
        for i in self._ids:
            (ids if isinstance(i, pycompat.integer_types) else new_ids).append(i)
        if not ids:
            return self
        query = """SELECT id FROM "%s" WHERE id IN %%s""" % self._table
        self._cr.execute(query, [tuple(ids)])
        ids = [r[0] for r in self._cr.fetchall()]
        existing = self.browse(ids + new_ids)
        if len(existing) < len(self):
            # mark missing records in cache with a failed value
            exc = MissingError(_("Record does not exist or has been deleted."))
            self.env.cache.set_failed(self - existing, self._fields.values(), exc)
        return existing

以上来自odoo/models.py

我们从下往上看

从数据库中拿到记录相关的ids

_browse创建了一个实例

__iter__用的是yield格式，每次从ids中拿出一个id，通过_browse生成这条记录的一个实例。

如果对__iter__迭代器这块有不理解请看：

odoo 关于self的迭代器（1）

odoo 关于self的迭代器（2）