pycraigslist

A fast and expressive Craigslist API wrapper.

Disclaimer

I do not work or have an affiliation with Craigslist.
This library is intended for educational purposes. It is not advised to crawl and download data from Craigslist.

Installation

pip install pycraigslist

Jump Start

Let's find cars & trucks for sale with keyword "Mazda Miata" in the East Bay Area, California:

import pycraigslist

miatas = pycraigslist.forsale.cta(site="sfbay", area="eby", query="Mazda Miata")
for miata in miatas.search():
    print(miata)

>>> {'country': 'US',
    'region': 'CA',
    'site': 'sfbay',
    'area': 'eby',
    'category': 'cto',
    'id': '7291715564',
    'repost_of': '',
    'last_updated': '2021-03-15 09:06',
    'title': '1990 Mazda Miata',
    'neighborhood': 'oakland lake merritt / grand',
    'price': '$5,000',
    'url': 'https://sfbay.craigslist.org/eby/cto/d/oakland-1990-mazda-miata/7291715564.html'}
    # ...

Background

This library is intended to be expressive and easy to use.

pycraigslist classes

pycraigslist.community (craigslist.org > community)
pycraigslist.events (craigslist.org > event calendar)
pycraigslist.forsale (craigslist.org > for sale)
pycraigslist.gigs (craigslist.org > gigs)
pycraigslist.housing (craigslist.org > housing)
pycraigslist.jobs (craigslist.org > jobs)
pycraigslist.resumes (craigslist.org > resumes)
pycraigslist.services (craigslist.org > services)

We can search for posts in parent classes. Let's find paid gigs in Portland, Oregon:

import pycraigslist

paid_gigs = pycraigslist.gigs(site="portland", is_paid=True)
for gig in paid_gigs.search():
    print(gig)

>>> {'country': 'US',
    'region': 'OR',
    'site': 'portland',
    'area': 'mlt',
    'category': 'lbg',
    'id': '7295392821',
    'repost_of': '7292985211',
    'last_updated': '2021-03-22 13:00',
    'title': 'Packing and moving',
    'neighborhood': 'SE Portland',
    'price': '',
    'url': 'https://portland.craigslist.org/mlt/lbg/d/portland-packing-and-moving/7295392821.html'}
    # ...

pycraigslist subclasses

Most pycraigslist classes have subclasses to allow for categorical searches. For example:

pycraigslist.forsale.bia (craigslist.org > for sale > bikes)
pycraigslist.forsale.cta (craigslist.org > for sale > cars & trucks)
pycraigslist.housing.apa (craigslist.org > housing > apartments / housing for rent)
pycraigslist.housing.roo (craigslist.org > housing > apartments / rooms & shares)

Finding pycraigslist subclasses

To search for subclasses, use the class method .get_categories(). The resulting keys are the subclass names.

import pycraigslist

print(pycraigslist.housing.get_categories())

>>> {'apa': 'apartments / housing for rent',
    'swp': 'housing swap',
    'off': 'office & commercial',
    'prk': 'parking & storage',
    'rea': 'real estate',
    'reb': 'real estate - by dealer',
    'reo': 'real estate - by owner',
    'roo': 'rooms & shares',
    'sub': 'sublets & temporary',
    'vac': 'vacation rentals',
    'hou': 'wanted: apts',
    'rew': 'wanted: real estate',
    'sha': 'wanted: room/share',
    'sbw': 'wanted: sublet/temp'}

If we're interested in searching for vacation rentals, we'd use the subclass pycraigslist.housing.vac.

Finding and using filters

As demonstrated in the jump-start example, we can apply filters to our Craigslist search. To find valid filters for a class or subclass instance, use .get_filters().

import pycraigslist

tokyo_autos = pycraigslist.forsale.cta(site="tokyo")
print(tokyo_autos.get_filters())

>>> {'query': '...', 'search_titles': 'True/False', 'has_image': 'True/False',
    'posted_today': 'True/False', 'bundle_duplicates': 'True/False',
    'search_distance': '...', 'zip_code': '...', 'min_price': '...', 'max_price': '...',
    'make_model': '...', 'min_year': '...', 'max_year': '...', 'min_miles': '...',
    'max_miles': '...', 'min_engine_displacement': '...', 'max_engine_displacement': '...',
    'condition': ['新品', 'ほぼ新品', '美品', '良品', '使用に問題なし', 'サルベージ'],
    'auto_cylinders': ['3気筒', '4気筒', '5気筒', '6気筒', '8気筒', '10気筒', '12気筒', 'その他'],
    'auto_drivetrain': ['前輪', '後輪', '4WD'],
    'auto_fuel_type': ['ガソリン', 'ディーゼル', 'ハイブリッド', '電気', 'その他'],
    'auto_paint': ['ブラック', 'ブルー', 'ブラウン', 'グリーン', 'グレー', 'オレンジ', 'パープル',
                   'レッド', 'シルバー', 'ホワイト', 'イエロー', 'カスタム'],
    'auto_size': ['コンパクト', 'フルサイズ', '中型', 'サブコンパクト'],
    'auto_title_status': ['クリーン', 'サルベージ', '再生', '部品のみ', '先取特権', '不明'],
    'auto_transmission': ['MT', 'AT', 'その他'],
    'auto_bodytype': ['バス', 'コンバーチブル', 'クーペ', 'ハッチバック', 'ミニバン', 'オフロード',
                      'ピックアップ', 'セダン', 'トラック', 'SUV', 'ワゴン', 'バン', 'その他'],
    'language': ['afrikaans', 'català', 'dansk', 'deutsch', 'english', 'español', 'suomi',
                 'français', 'italiano', 'nederlands', 'norsk', 'português', 'svenska',
                 'filipino', 'türkçe', '中文', 'العربية', '日本語', '한국말', 'русский',
                 'tiếng việt']}

Using this information, let's find cars & trucks with clean (クリーン) titles in Tokyo, Japan:

import pycraigslist

tokyo_autos = pycraigslist.forsale.cta(site="tokyo", auto_title_status="クリーン")
for auto in tokyo_autos.search():
    print(auto)

>>> {'country': 'JP',
    'region': '',
    'site': 'tokyo',
    'area': '',
    'category': 'cto',
    'id': '7301105503',
    'repost_of': '',
    'last_updated': '2021-04-03 14:04',
    'title': 'Suzuki Jimny 660 XG 4WD Keyless Entry Aluminum Wheel Non-Smoking Car',
    'neighborhood': 'Chiba Ken, Noda shi, Funakata 1630-1',
    'price': '¥650,000',
    'url': 'https://tokyo.craigslist.org/cto/d/suzuki-jimny-660-xg-4wd-keyless-entry/7301105503.html'}
    # ...

If we want to apply a bunch of filters, pass a dictionary of filters into the filters keyword parameter. Note: keyword argument filters will override filters if there are conflicting keys. For example:

import pycraigslist

bike_filters = {
    "bicycle_frame_material": "steel",
    # array of filter values are accepted
    "bicycle_wheel_size": ["650C", "700C"],
    "bicycle_type": "road",
}
# we'd still get titanium road bikes with size 650C or 700C wheels
titanium_bikes = pycraigslist.forsale.bia(
    site="sfbay", area="sfc", bicycle_frame_material="titanium", filters=bike_filters
)

Searching for posts

General search

To search for Craigslist posts, use .search(). .search() will return a dictionary of post attributes (type str) and will search for every post by default. Use the limit keyword parameter to add a stop limit to a query. For example, use limit=50 if we want 50 posts. There is a maximum of 3000 posts per query.

Let's find the first 20 posts for farming and gardening services in Denver, Colorado:

import pycraigslist

gardening_services = pycraigslist.services.fgs(site="denver")
for service in gardening_services.search(limit=20):
    print(service)

>>> {'country': 'US',
    'region': 'CO',
    'site': 'denver',
    'area': '',
    'category': 'fgs',
    'id': '7301324564',
    'repost_of': '6974119634',
    'last_updated': '2021-04-03 11:47',
    'title': '🌲 Tree Removal/Trimming, Stump Grind: LICENSED/INSURED! 720-605-1584',
    'neighborhood': 'All Areas',
    'price': '',
    'url': 'https://denver.craigslist.org/fgs/d/littleton-tree-removal-trimming-stump/7301324564.html'}
    # ...

Detailed search

Use .search_detail() to get detailed Craigslist posts. The limit keyword parameter in .search also applies to .search_detail. Set include_body=True to include the post's body in the output. By default, include_body=False. Disclaimer: .search_detail is more time consuming than .search.

Let's get detailed posts with the post body for all cars & trucks for sale in Abilene, Texas:

import pycraigslist

all_autos = pycraigslist.forsale.cta(site="abilene")
for auto in all_autos.search_detail(include_body=True):
    print(auto)

>>> {'country': 'US',
    'region': 'TX',
    'site': 'abilene',
    'area': '',
    'category': 'cto',
    'id': '7309894792',
    'repost_of': '',
    'last_updated': '2021-04-20 12:17',
    'title': '2009 Mercedes GL-320',
    'neighborhood': 'Brownwood',
    'price': '$12,000',
    'url': 'https://abilene.craigslist.org/cto/d/brownwood-2009-mercedes-gl-320/7309894792.html',
    'lat': '31.729000',
    'lon': '-99.019000',
    'address': '',
    'misc': ['2009 mercedes-benz gl-class'],
    'condition': 'excellent',
    'drive': 'fwd',
    'fuel': 'diesel',
    'odometer': '100700',
    'paint_color': 'black',
    'title_status': 'clean',
    'transmission': 'automatic',
    'body': 'BEAUTIFUL car inside and out!! Diesel with only 100k, mechanic says its in great condition.'}
    # ...

Additional attributes

__doc__: Gets category name.
url: Gets full URL.
count: Gets number of posts.

import pycraigslist

east_bay_apa = pycraigslist.housing.apa(site="sfbay", area="eby", max_price=800)

# 1
print(east_bay_apa.__doc__)
>>> 'apartments / housing for rent'

# 2
print(east_bay_apa.url)
>>> 'https://sfbay.craigslist.org/search/eby/apa?searchNearby=1&s=0&max_price=800'

# 3
print(east_bay_apa.count)
>>> 56

Exceptions

pycraigslist has the following exceptions:

MaximumRequestsError : exceeds maximum retries for a query

To use pycraigslist exceptions, import / import from pycraigslist.exceptions. For example:

import pycraigslist
from pycraigslist.exceptions import MaximumRequestsError

try:
    sf_bikes = pycraigslist.forsale.bia(site="sfbay", area="sfc", min_price=50)
    for bike in sf_bikes.search():
        print(bike)
except MaximumRequestsError:
    print("Yikes! Something's up with the network.")

Contribute

Support

If you are having issues or would like to propose a new feature, please use the issues tracker.

License

The project is licensed under the MIT license.

jgrugru / pycraigslist