A library for retrieving free proxies (HTTP, HTTPS, SOCKS4, SOCKS5). Supports Python 2.7+ and 3.4+.
NOTE: This library isn't designed for production use. It's advised to use your own proxies or purchase a service which provides an API. These are merely free ones that are retrieved from sites and should only be used for development or testing purposes.
import proxyscrape
collector = proxyscrape.create_collector('default', 'http') # Create a collector for http resources
proxy = collector.get_proxy({'country': 'united states'}) # Retrieve a united states proxyThe latest version of proxyscrape is available via pip:
$ pip install proxyscrapeAlternatively, you can download and install from source:
$ python setup.py installCurrent proxies provided are scraped from various sites which offer free HTTP, HTTPS, SOCKS4, and SOCKS5 proxies; and don't require headless browsers or selenium to retrieve. The list of sites proxies retrieved are shown below.
| resource | resource type | url |
|---|---|---|
| anonymous-proxy | http, https | https://free-proxy-list.net/anonymous-proxy.html |
| free-proxy-list | http, https | https://free-proxy-list.net |
| proxy-daily-http proxy-daily-socks4 proxy-daily-socks5 | http socks4 socks5 | http://www.proxy-daily.com |
| socks-proxy | socks4, socks5 | https://www.socks-proxy.net |
| ssl-proxy | https | https://www.sslproxies.org |
| uk-proxy | http, https | https://free-proxy-list.net/uk-proxy.html |
| us-proxy | http, https | https://www.us-proxy.org |
See Integration section for additional proxies.
Proxy Scrape is a library aimed at providing an efficient an easy means of retrieving proxies for web-scraping purposes. The proxies retrieved are available from sites providing free proxies. The proxies provided, as shown in the above table, can be of one of the following types (referred to as a resource type): http, https, socks4, and socks5.
Collectors serve as the interface to retrieving proxies. They are instantiating at module-level and can be retrieved and re-used in different parts of the application (similar to the Python logging library). Collectors can be created and retrieved via the create_collector(...) and get_collector(...) functions.
from proxyscrape import create_collector, get_collector
collector = create_collector('my-collector', ['socks4', 'socks5'])
# Some other section of code
collector = get_collector('my-collector')Each collector should have a unique name and be initialized only once. Typically, only a single collector of a given resource type should be utilized. Filters can then be applied to the proxies if specific criteria is desired.
When given one or more resources, the collector will use those to retrieve proxies. If one or more resource types are given, the resources for each of the types will be used to retrieve proxies.
Once created, proxies can be retrieved via the get_proxy(...) or the get_proxies(...) function. This optionally takes a filter_opts parameter which can filter by the following:
code(us, ca, ...)country(united states, canada, ...)anonymous(True, False)type(http, https, socks4, socks5, ...)
from proxyscrape import create_collector
collector = create_collector('my-collector', 'http')
# Retrieve any http proxy
proxy = collector.get_proxy()
# Retrieve only 'us' proxies
proxy = collector.get_proxy({'code': 'us'})
# Retrieve only anonymous 'uk' or 'us' proxies
proxy = collector.get_proxy({'code': ('us', 'uk'), 'anonymous': True})
# Retrieve all 'ca' proxies
proxies = collector.get_proxies({'code': 'ca'})Filters can be applied to every proxy retrieval from the collector via apply_filter(...). This is useful when the same filter is expected for any proxy retrieved.
from proxyscrape import create_collector
collector = create_collector('my-collector', 'http')
# Only retrieve 'uk' and 'us' proxies
collector.apply_filter({'code': 'us'})
# Filtered proxies
proxy = collector.get_proxy()
# Clear filter
collector.clear_filter()Note that some filters may instead use specific resources to achieve the same results (i.e. 'us-proxy' or 'uk-proxy' for 'us' and 'uk' proxies).
Blacklists can be applied to a collector to prevent specific proxies from being retrieved. They accept either one or more Proxy objects, or a host + port number combination and won't allow retrieval of matching proxies. Proxies can be individually removed from blacklists or the entire blacklist can be cleared.
from proxyscrape import create_collector
collector = create_collector('my-collector', 'http')
# Add proxy to blacklist
collector.blacklist_proxy(Proxy('192.168.1.1', '80', None, None, None, 'http', 'my-resource'))
collector.blacklist_proxy(host='192.168.1.2', port='8080')
# Blacklisted proxies won't be included
proxy = get_proxy()
# Remove individual proxies
collector.remove_blacklist(host='192.168.1.1', port='80')
# Clear blacklist
collector.clear_blacklist()Instead of permanently blacklisting a particular proxies, a proxy can instead be removed from internal memory. This allows it to be re-added to the pool upon a subsequent refresh.
from proxyscrape import create_collector
collector = create_collector('my-collector', 'http')
# Remove proxy from internal pool
collector.remove_proxy(Proxy('192.168.1.1', '80', None, None, 'http', 'my-resource'))Apart from automatic refreshes when retrieving proxies, they can also be forcefully refreshed via the refresh_proxies(...) function.
from proxyscrape import create_collector
collector = create_collector('my-collector', 'http')
# Forcefully refresh
collector.refresh_proxies(force=True)
# Refresh only if proxies not refreshed within `refresh_interval`
collector.refresh_proxies(force=False)Resources refer to a specific function that retrieves a set of proxies; the currently implemented proxies are all retrieves from scraping a particular web site.
Additional user-defined resources can be added to the pool of proxy retrieval functions via the add_resource(...) function. Resources can belong to multiple resource types.
from proxyscrape import add_resource
def func():
return {Proxy('192.168.1.1', '80', 'us', 'united states', False, 'http', 'my-resource'), }
add_resource('my-resource', func, 'http')As shown above, a resource doesn't necessarily have to scrape proxies from a web site. It can be return a hard-coded list of proxies, make a call to an api, read from a file, etc.
The set of library- and user-defined resources can be retrieved via the get_resources(...) function.
from proxyscrape import get_resources
resources = get_resources()Resource types are groupings of resources that can be specified when defining a collector (opposed to giving a collection of resources.
Additional user-defined resource types can be added via the add_resource_type(...) function. Resources can optionally be added to a resource type when defining it.
from proxyscrape import add_resource_type
add_resource_type('my-resource-type')
add_resource_type('my-other-resource-type', 'my-resource') # Define resources for resource typeThe set of library- and user-defined resource types can be retrieved via the get_resource_types(...) function.
from proxyscrape import get_resource_types
resources = get_resource_types()Integrations are proxy implementations that are specific to a particular website or API and have a distinctively separate use case.
The ProxyScrape.com API provides a means of accessing thousands of proxies of various types (HTTP, SOCKS4, SOCKS5) in an efficient manner. These are vetted and validated with a minimal response time.
The get_proxyscrape_resource(...) function is used to dynamically create a new resource for using the proxyscrape API. The resource name can then be added to a resource type and used like any other library- or user-defined resource. The following parameters are used for the API:
proxytype(http, socks4, socks5, all)timeout(milliseconds > 0)ssl(yes, no, all)anonymity(elite, anonymous, transparent, all)country(any Alpha 2 ISO country code, all)
from proxyscrape import get_proxyscrape_resource
resource_name = get_proxyscrape_resource(proxytype='http', timeout=5000, ssl='yes', anonymity='all', country='us')Contributions or suggestions are welcome! Feel free to open an issue if a bug is found or an enhancement is desired, or even a pull request.
All changes and versioning information can be found in the CHANGELOG.
Copyright (c) 2018 Jared Gillespie. See LICENSE for details.