Financial Web Crawler

Web Crawler for gathering financial news for scientific research purposes.

Stack

The stack bellow was used mostly due to its ease of installation, configuration, and also efficiency and portability.

• Language: Python (3.7.4)
• Core crawler engine: Scrapy (1.8.0)

Pre-installation

This system was developed in Ubuntu 16.04 but will work properly on any other Operational System(OS X, Windows, etc.).

However, this guide will only include instructions for plugins and packages that are not already installed on this OS. For this reason, we assume that technologies like a python interpreter and pipenv are ready for use, and should not be on the scope of this document.

• Now install pipenv dependency manager:

$ pip install --user pipenv

Project configuration

Now we'll start setting up the project.

• Clone the repo from GitHub and change to project root directory. After that, install project dependencies and go to python virtual env, by running:

$ pipenv install
$ pipenv shell

Running the project

Now we will run the project and get query answers.

• The project will do the following actions:
  1. Download a raw data like a JSON or an HTML in a controlled not abusive manner, respecting instructions from robots.txt files
  2. Save data to a specific folder according to webpage source and stock or asset selected
  3. Parse the HTML or JSON and extract the data to a CSV file selecting only preselected fields

$ python scraper

• To run just the step iii above, pass the argument --step=gen_csv:

$ python scraper --step=gen_csv

• Only stocks present in the companies_list.csv are allowed to run if --website=wsj_news. So if necessary, include new entries in this file following previous entries as examples.

• Required params:

  --crawl_type allowed values: api, website

  --stock a single stock ticker present in the file companies_list.csv or the value list to run for all stocks in companies_list.csv

• When no params are passed the default mode of this scraper runs a crawler with the following characteristics:

  --website=wsj_news for wall street journal API

  --mode=default default operation mode

  --step=all run all steps mentioned before

  --max_requests=2 maximum number of consecutive requests

  --max_elements=100 gets a maximum number of elements

  --start_time is the current time

  --end_time is unused as the systems defaults to the number maximum of consecutive requests

• Example to run with all possible fields:

$ python scraper --crawl_type=api --website=wsj_news --stock=btcusd --mode=default --step=all --max_requests=2 --end_time=2019-10-18T18:59:08 --start_time=2019-11-12T22:30:00

• Example to run greedy mode:

  Beware the greedy mode will run trough exhaustion and may lead to blockage if not used properly. This mode ignores end_time and max_requests, but it can consider start_time.

$ python scraper --website=wsj_news --stock=btcusd --mode=greedy

Further Improvements

Tests

Unitary and integration tests will guarantee consistency and good maintainability

Job scheduler

Adding a feature for scheduling each job separately with improve performance

Work for any site

Right now the code only works for a specific news website, it will be adjusted for some other websites candidates:

• Yahoo finance

• Business Insider

• News BTC

• Be In Crypto

Design improvements

• It seem scrapy expects all loops in the same website are self-contained

Known issues

• See open issues in github

Final considerations

• Go play around! =)