python-copr
1.63
  • Installation
  • Client version 3
    • Example usage
    • Quick overview
      • Client initialization
      • Data structures
      • Error handling
      • Pagination
        • Pagination parameters
      • Working with proxies directly
    • Resources info
    • Parameters
python-copr
  • Docs »
  • Client version 3 »
  • Pagination
  • View page source

Pagination¶

Pagination is not enforced, so users don’t need to care about it if they don’t want to. In this code sample, all packages from given project are returned, regardless their number.

from copr.v3 import Client
client = Client(config)

packages = client.package_proxy.get_list("@copr", "copr")
print(packages)
[Munch({'id': 1, 'ownername': '@copr', 'projectname': 'copr', 'state': 'pending', ...}),
 Munch({'id': 2, 'ownername': '@copr', 'projectname': 'copr', 'state': 'pending', ...}),
 Munch({'id': 3, 'ownername': '@copr', 'projectname': 'copr', 'state': 'importing', ...}),
 Munch({'id': 4, 'ownername': '@copr', 'projectname': 'copr', 'state': 'importing', ...}),
 Munch({'id': 5, 'ownername': '@copr', 'projectname': 'copr', 'state': 'canceled', ...})]

However, in some cases, it may be useful to obtain just a limited number of objects. Querying all builds from a project with hundreds of thousands of them may be painfully slow or even timeout. And if all of them are not even needed, it is a huge waste of resources. It all depends on the specific use-case. When it is useful, you can query limited and/or ordered number of objects.

pagination = {"limit": 3, "order": "name"}
packages = client.package_proxy.get_list("@copr", "copr", pagination=pagination)
print(packages)
print(packages.meta)
[Munch({'id': 1, 'ownername': '@copr', 'projectname': 'copr', 'state': 'pending', ...}),
 Munch({'id': 2, 'ownername': '@copr', 'projectname': 'copr', 'state': 'pending', ...}),
 Munch({'id': 3, 'ownername': '@copr', 'projectname': 'copr', 'state': 'importing', ...})]
Munch({u'offset': 0, u'limit': 3, u'order_type': u'ASC', u'order': u'id'})

And finally, in some cases, it may be useful to iterate through all objects, but not obtaining them all at once (e.g. when projects are so large, that requesting everything timeouts).

from copr.v3.pagination import next_page

package_page = client.package_proxy.get_list("@copr", "copr", pagination={"limit": 3})
while package_page:
    for package in package_page:
        print(package)
    print("---")
    package_page = next_page(package_page)
Munch({'id': 1, 'ownername': '@copr', 'projectname': 'copr', 'state': 'pending', ...})
Munch({'id': 2, 'ownername': '@copr', 'projectname': 'copr', 'state': 'pending', ...})
Munch({'id': 3, 'ownername': '@copr', 'projectname': 'copr', 'state': 'importing', ...})
---
Munch({'id': 4, 'ownername': '@copr', 'projectname': 'copr', 'state': 'importing', ...})
Munch({'id': 5, 'ownername': '@copr', 'projectname': 'copr', 'state': 'canceled', ...})

Warning

Don’t remove projects, packages, builds, etc while paginating over them. It may result in skipping some results. Instead, query all objects (using pagination) first, and then delete them.

Pagination parameters¶

Field Type Description
limit int number of objects to obtain
offset int number of objects from beginning to skip
order str sort objects by this property
order_type str “ASC” or “DESC”
Next Previous

© Copyright 2018, Red Hat, Inc..

Built with Sphinx using a theme provided by Read the Docs.