Source code for octavia.api.common.pagination
# Copyright 2016 Intel Corporation
#
# Licensed under the Apache License, Version 2.0 (the "License"); you may
# not use this file except in compliance with the License. You may obtain
# a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
# License for the specific language governing permissions and limitations
# under the License.
import copy
import itertools
from oslo_log import log as logging
from pecan import request
import sqlalchemy
from sqlalchemy.orm import aliased
import sqlalchemy.sql as sa_sql
from octavia.api.common import types
from octavia.common.config import cfg
from octavia.common import constants
from octavia.common import exceptions
from octavia.db import base_models
from octavia.db import models
CONF = cfg.CONF
LOG = logging.getLogger(__name__)
[docs]
class PaginationHelper:
"""Class helping to interact with pagination functionality
Pass this class to `db.repositories` to apply it on query
"""
_auxiliary_arguments = ('limit', 'marker',
'sort', 'sort_key', 'sort_dir',
'fields', 'page_reverse',
)
def __init__(self, params, sort_dir=constants.DEFAULT_SORT_DIR):
"""Pagination Helper takes params and a default sort direction
:param params: Contains the following:
limit: maximum number of items to return
marker: the last item of the previous page; we return
the next results after this value.
sort: array of attr by which results should be sorted
:param sort_dir: default direction to sort (asc, desc)
"""
self.marker = params.get('marker')
self.sort_dir = self._validate_sort_dir(sort_dir)
self.limit = self._parse_limit(params)
self.sort_keys = self._parse_sort_keys(params)
self.params = params
self.filters = None
self.page_reverse = params.get('page_reverse', 'False')
@staticmethod
def _parse_limit(params):
if CONF.api_settings.pagination_max_limit == 'infinite':
page_max_limit = None
else:
page_max_limit = int(CONF.api_settings.pagination_max_limit)
limit = params.get('limit', page_max_limit)
try:
# Deal with limit being a string or int meaning 'Unlimited'
if limit == 'infinite' or int(limit) < 1:
limit = None
# If we don't have a max, just use whatever limit is specified
elif page_max_limit is None:
limit = int(limit)
# Otherwise, we need to compare against the max
else:
limit = min(int(limit), page_max_limit)
except ValueError as e:
raise exceptions.InvalidLimit(key=limit) from e
return limit
def _parse_sort_keys(self, params):
sort_keys_dirs = []
sort = params.get('sort')
sort_keys = params.get('sort_key')
if sort:
for sort_dir_key in sort.split(","):
comps = sort_dir_key.split(":")
if len(comps) == 1: # Use default sort order
sort_keys_dirs.append((comps[0], self.sort_dir))
elif len(comps) == 2:
sort_keys_dirs.append(
(comps[0], self._validate_sort_dir(comps[1])))
else:
raise exceptions.InvalidSortKey(key=comps)
elif sort_keys:
sort_keys = sort_keys.split(',')
sort_dirs = params.get('sort_dir')
if not sort_dirs:
sort_dirs = [self.sort_dir] * len(sort_keys)
else:
sort_dirs = sort_dirs.split(',')
if len(sort_dirs) < len(sort_keys):
sort_dirs += [self.sort_dir] * (len(sort_keys) -
len(sort_dirs))
for sk, sd in zip(sort_keys, sort_dirs):
sort_keys_dirs.append((sk, self._validate_sort_dir(sd)))
return sort_keys_dirs
def _parse_marker(self, session, model):
return session.query(model).filter_by(id=self.marker).one_or_none()
@staticmethod
def _get_default_column_value(column_type):
"""Return the default value of the columns from DB table
In postgreDB case, if no right default values are being set, an
psycopg2.DataError will be thrown.
"""
type_schema = {
'datetime': None,
'big_integer': 0,
'integer': 0,
'string': ''
}
if isinstance(column_type, sa_sql.type_api.Variant):
return PaginationHelper._get_default_column_value(column_type.impl)
return type_schema[column_type.__visit_name__]
@staticmethod
def _validate_sort_dir(sort_dir):
sort_dir = sort_dir.lower()
if sort_dir not in constants.ALLOWED_SORT_DIR:
raise exceptions.InvalidSortDirection(key=sort_dir)
return sort_dir
def _make_links(self, model_list):
if CONF.api_settings.api_base_uri:
path_url = "{api_base_url}{path}".format(
api_base_url=CONF.api_settings.api_base_uri.rstrip('/'),
path=request.path)
else:
path_url = request.path_url
links = []
if model_list:
prev_attr = [f"limit={self.limit}"]
if self.params.get('sort'):
prev_attr.append("sort={}".format(self.params.get('sort')))
if self.params.get('sort_key'):
prev_attr.append("sort_key={}".format(
self.params.get('sort_key')))
next_attr = copy.copy(prev_attr)
if self.marker:
prev_attr.append("marker={}".format(model_list[0].get('id')))
prev_attr.append("page_reverse=True")
prev_link = {
"rel": "previous",
"href": "{url}?{params}".format(
url=path_url,
params="&".join(prev_attr))
}
links.append(prev_link)
# TODO(rm_work) Do we need to know when there are more vs exact?
# We safely know if we have a full page, but it might include the
# last element or it might not, it is unclear
if self.limit is None or len(model_list) >= self.limit:
next_attr.append("marker={}".format(model_list[-1].get('id')))
next_link = {
"rel": "next",
"href": "{url}?{params}".format(
url=path_url,
params="&".join(next_attr))
}
links.append(next_link)
links = [types.PageType(**link) for link in links]
return links
def _apply_tags_filtering(self, params, model, query):
if not getattr(model, "_tags", None):
return query
if 'tags' in params:
tags = params.pop('tags')
for tag in tags:
# This requires a multi-join to the tags table,
# so me must use aliases for each one.
tag_alias = aliased(base_models.Tags)
query = query.join(tag_alias, model._tags)
query = query.filter(tag_alias.tag == tag)
if 'tags-any' in params:
tags = params.pop('tags-any')
tag_alias = aliased(base_models.Tags)
query = query.join(tag_alias, model._tags)
query = query.filter(tag_alias.tag.in_(tags))
if 'not-tags' in params:
tags = params.pop('not-tags')
subq = query.session.query(model.id)
for tag in tags:
tag_alias = aliased(base_models.Tags)
subq = subq.join(tag_alias, model._tags)
subq = subq.filter(tag_alias.tag == tag)
query = query.filter(~model.id.in_(subq))
if 'not-tags-any' in params:
tags = params.pop('not-tags-any')
query = query.filter(
~model._tags.any(base_models.Tags.tag.in_(tags)))
return query
@staticmethod
def _prepare_tags_list(param):
"""Split comma separated tags and return a flat list of tags."""
if not isinstance(param, list):
param = [param]
return list(itertools.chain.from_iterable(
tag.split(',') for tag in param))
[docs]
def apply(self, query, model, enforce_valid_params=True):
"""Returns a query with sorting / pagination criteria added.
Pagination works by requiring a unique sort_key specified by sort_keys.
(If sort_keys is not unique, then we risk looping through values.)
We use the last row in the previous page as the pagination 'marker'.
So we must return values that follow the passed marker in the order.
With a single-valued sort_key, this would be easy: sort_key > X.
With a compound-values sort_key, (k1, k2, k3) we must do this to repeat
the lexicographical ordering:
(k1 > X1) or (k1 == X1 && k2 > X2) or (k1 == X1 && k2 == X2 && k3 > X3)
We also have to cope with different sort_directions.
Typically, the id of the last row is used as the client-facing
pagination marker, then the actual marker object must be fetched from
the db and passed in to us as marker.
:param query: the query object to which we should add
paging/sorting/filtering
:param model: the ORM model class
:param enforce_valid_params: check for invalid entries in self.params
:rtype: sqlalchemy.orm.query.Query
:returns: The query with sorting/pagination/filtering added.
"""
# Add filtering
if CONF.api_settings.allow_filtering:
# Exclude (valid) arguments that are not used for data filtering
filter_params = {k: v for k, v in self.params.items()
if k not in self._auxiliary_arguments}
secondary_query_filter = filter_params.pop(
"project_id", None) if (model == models.Amphora) else None
# Translate arguments from API standard to data model's field name
filter_params = (
model.__v2_wsme__.translate_dict_keys_to_data_model(
filter_params)
)
if 'loadbalancer_id' in filter_params:
filter_params['load_balancer_id'] = filter_params.pop(
'loadbalancer_id')
# Pop the 'tags' related parameters off before handling the
# other filters. Then apply the 'tags' filters after the
# other filters have been applied.
tag_params = {}
if 'tags' in filter_params:
tag_params['tags'] = self._prepare_tags_list(
filter_params.pop('tags'))
if 'tags-any' in filter_params:
tag_params['tags-any'] = self._prepare_tags_list(
filter_params.pop('tags-any'))
if 'not-tags' in filter_params:
tag_params['not-tags'] = self._prepare_tags_list(
filter_params.pop('not-tags'))
if 'not-tags-any' in filter_params:
tag_params['not-tags-any'] = self._prepare_tags_list(
filter_params.pop('not-tags-any'))
# Drop invalid arguments
self.filters = {k: v for (k, v) in filter_params.items()
if k in vars(model.__data_model__())}
if enforce_valid_params and (
len(self.filters) < len(filter_params)
):
raise exceptions.InvalidFilterArgument()
query = model.apply_filter(query, model, self.filters)
if secondary_query_filter is not None:
query = query.filter(model.load_balancer.has(
project_id=secondary_query_filter))
# Apply tags filtering for the models which support tags.
query = self._apply_tags_filtering(tag_params, model, query)
# Add sorting
if CONF.api_settings.allow_sorting:
# Add default sort keys (if they are OK for the model)
keys_only = [k[0] for k in self.sort_keys]
for key in constants.DEFAULT_SORT_KEYS:
if key not in keys_only and hasattr(model, key):
self.sort_keys.append((key, self.sort_dir))
for current_sort_key, current_sort_dir in self.sort_keys:
# Translate sort_key from API standard to data model's name
current_sort_key = (
model.__v2_wsme__.translate_key_to_data_model(
current_sort_key))
sort_dir_func = {
constants.ASC: sqlalchemy.asc,
constants.DESC: sqlalchemy.desc,
}[current_sort_dir]
try:
# The translated object may be a nested parameter
# such as vip.ip_address, so handle that case by
# joining with the nested table.
if '.' in current_sort_key:
parent, child = current_sort_key.split('.')
parent_obj = getattr(model, parent)
query = query.join(parent_obj)
sort_key_attr = child
else:
sort_key_attr = getattr(model, current_sort_key)
except AttributeError as e:
raise exceptions.InvalidSortKey(
key=current_sort_key) from e
query = query.order_by(sort_dir_func(sort_key_attr))
# Add pagination
if CONF.api_settings.allow_pagination:
default = '' # Default to an empty string if NULL
if self.marker is not None:
marker_object = self._parse_marker(query.session, model)
if not marker_object:
raise exceptions.InvalidMarker(key=self.marker)
marker_values = []
for sort_key, _ in self.sort_keys:
v = getattr(marker_object, sort_key)
if v is None:
v = default
marker_values.append(v)
# Build up an array of sort criteria as in the docstring
criteria_list = []
for i in range(len(self.sort_keys)):
crit_attrs = []
for j in range(i):
model_attr = getattr(model, self.sort_keys[j][0])
default = PaginationHelper._get_default_column_value(
model_attr.property.columns[0].type)
attr = sa_sql.expression.case(
(model_attr.isnot(None), model_attr),
else_=default)
crit_attrs.append(attr == marker_values[j])
model_attr = getattr(model, self.sort_keys[i][0])
default = PaginationHelper._get_default_column_value(
model_attr.property.columns[0].type)
attr = sa_sql.expression.case(
(model_attr.isnot(None), model_attr),
else_=default)
this_sort_dir = self.sort_keys[i][1]
if this_sort_dir == constants.DESC:
if self.page_reverse == "True":
crit_attrs.append(attr > marker_values[i])
else:
crit_attrs.append(attr < marker_values[i])
elif this_sort_dir == constants.ASC:
if self.page_reverse == "True":
crit_attrs.append(attr < marker_values[i])
else:
crit_attrs.append(attr > marker_values[i])
else:
raise exceptions.InvalidSortDirection(
key=this_sort_dir)
criteria = sa_sql.and_(*crit_attrs)
criteria_list.append(criteria)
f = sa_sql.or_(*criteria_list)
query = query.filter(f)
if self.limit is not None:
query = query.limit(self.limit)
model_list = query.all()
links = None
if CONF.api_settings.allow_pagination:
links = self._make_links(model_list)
return model_list, links
