Extract data from Craigslist.org by python3 and pomp framework

Example how to build scalable cluster of web crawlers with centralized jobs queue on python3.

• redis for queue management with unique jobs
• Apache Kafka for committing gathered data
• django with postgres database backend to store and visualize gathered data
• grafana with kamon dashboards for metric collecting via kamon-io/docker-grafana-graphite
• pomp lightweight web scraping framework
• docker-compose to run all of this zoo in one machine

Crawler instance uses:

• latest python3
• works on asyncio
• aiohttp for async fetch html pages from craigslist.org
• lxml for parsing content

Screencast without audio.

Notice

This software is not associated with Craigslist and for research purposes only. Please check Craigslist terms of use and robots.txt.

Keep in mind

All this stuff is overhead stack of technologies to scrape data from Craigslist.org.

You can get all of you need from Craigslist.org in few steps with urllib/requests/etc + re/lxml/beautifulsoup/etc.

But this example is a good entry point to make own cluster of web crawlers like Scrapy Cluster.

Item's price saved and represented in cents.

Crawler have some issues when trying to parse some pages for example purposes of exception handling.

Hard-coded in crawler sources:

• Craigslist's cities to scrape are: newyork, sfbay, chicago
• max number of list's pages to scrape is 3, first 3 pages from pagination
• max number of concurrent requests for one crawler instance is 3
• concurrent parse content by 2 workers, one crawler instance would be extract data by 2 workers

Installation

Prepare:

$ git clone https://github.com/estin/pomp-craigslist-example.git
$ cd pomp-craigslist-example
$ mkdir logs
$ chmod +x wait-for-it.sh
$ chmod +x wait-pg-and-kafka.sh
$ docker-compose pull

Install crawler and requirements:

$ docker-compose run --rm crawler python3 setup.py develop --user

Create database and admin user for django app:

$ docker-compose up -d postgres
$ docker-compose run --rm dataview manage dataview migrate
$ docker-compose run --rm dataview sh -c "echo \"from django.contrib.auth.models import User; User.objects.create_superuser('admin', '[email protected]', 'admin')\" | manage dataview shell"

Configure grafana/kamon dashboard to view metrics:

• login to dashboard http://localhost:8090/
• go to Data Sources section http://localhost:8090/datasources and click on add new or open link http://localhost:8090/datasources/new
• set field Name to graphite
• set field Url to http://localhost:81/
• click on Add button
• then import dashboard.json from this repo on http://localhost:8090/import/dashboard page

Usage

Run:

$ docker-compose up -d

Check status:

$ docker-compose ps

Check actuality of xpath:

$ docker-compose run --rm crawler manage check-xpath "https://sfbay.craigslist.org/search/bia?is_paid=all&search_distance_type=mi&query=kid+bike"

Start new crawling session:

$ docker-compose run --rm crawler manage session kidbike "search/bia?is_paid=all&search_distance_type=mi&query=kid+bike"

Where kidbike is the crawling session id and other part is the target query from browser url.

This command put next requests to the job's queue:

• https://newyork.craigslist.org/search/bia?is_paid=all&search_distance_type=mi&query=kid+bike
• https://sfbay.craigslist.org/search/bia?is_paid=all&search_distance_type=mi&query=kid+bike
• https://chicago.craigslist.org/search/bia?is_paid=all&search_distance_type=mi&query=kid+bike

And then check (with username: admin and pass: admin):

• http://localhost:8080/admin to view parsed and imported data to the postgres, do not forget periodically refresh the page
• http://localhost:8090/ to view metrics in kamon dashboard
• logs in ./logs directory or docker-compose logs <service name>

Increase crawler instances to speedup:

$ docker-compose scale crawler=2

Start another one crawling session:

$ docker-compose run --rm crawler manage session mountainbike "search/bia?is_paid=all&search_distance_type=mi&query=mountain+bike"

Project structure

• entry point craigslist/manage.py
• craigslist crawler, downloader, pipelines, middlewares and utils
• dataview django app
• dashboard.json kamon dashboard
• tests unit tests for craigslist

$ tree -I "*.pyc|__pycache__"
.
├── base-compose.yml
├── craigslist
│   ├── crawler.py
│   ├── downloader.py
│   ├── __init__.py
│   ├── item.py
│   ├── log.py
│   ├── manage.py
│   ├── middleware.py
│   ├── pipeline.py
│   ├── queue.py
│   └── utils.py
├── dashboard.json
├── dataview
│   ├── __init__.py
│   ├── items
│   │   ├── admin.py
│   │   ├── apps.py
│   │   ├── __init__.py
│   │   ├── management
│   │   │   └── commands
│   │   │       ├── dbimport.py
│   │   │       └── __init__.py
│   │   ├── migrations
│   │   │   ├── 0001_initial.py
│   │   │   └── __init__.py
│   │   ├── models.py
│   │   ├── tests.py
│   │   └── views.py
│   ├── settings.py
│   ├── urls.py
│   └── wsgi.py
├── docker-compose.yml
├── README.md
├── requires.pip
├── setup.py
└── tests
    ├── data
    │   ├── item.html
    │   └── list.html
    ├── test_crawler.py
    ├── test_downloader.py
    ├── test_pipeline.py
    ├── test_queue.py
    └── utils.py

8 directories, 37 files

TODO

• draw architecture diagram
• use asyncio kafka implementation
• fix kafka node not ready error on startup
• use native kafka consumer poll when doing bulk import data to postgres
• gather metrics of queue size only by one crawler instance or separate django management command

License

(The MIT License)

Copyright (c) 2016 Evgeniy Tatarkin

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.