Nomad List Web Scrapper

The Nomad List Web Scrapper, as the name states, is a program that navigates to every city on the Nomad List homepage and scraps all the information that Nomad List has on each and every one of these cities. The goal of this program is to make the vast amounts of information on Nomad List more accessible to the user, so that they can more easily find the best place in the world to live, work and travel as a remote worker.

The information on each city is taken from the tabs that are displayed when a city tile on the Nomad List home page is selected. The tabs that are scraped are:

• Scores
• Digital Nomad Guide
• Cost of Living
• Pro and Cons
• Reviews
• Weather
• Photos
• Near
• Next
• Similar

Table of Contents

• Nomad List Web Scrapper
  • Technologies Used
  • Installation
    • Environment Variables
    • Database Creation
  • Usage
    • Command Line Interface (CLI)
  • Storage
  • Project Status
  • Roadmap
  • Authors
  • Support

Technologies Used

• Coding Language: Python 3 🐍
• Main Packages: requests, bs4, selenium, argparse, pymysql.

Installation

Before running the program it is recommended that the following installation steps are carried out:

1. Clone the repo by running the following code in Git Bash

   gh repo clone kevinszuchet/data-mining

2. Install the dependencies (eg: Beautiful Soup, Requests, etc) required to run the program by executing the following code in the Terminal (macOS/Linux) or Command Line (Windows):

   pip3 install -r requirements.txt

3. If the code is not loaded on and running from a server, the user must ensure that some form of MySQL host is installed and running on their computer before running the program. The following MySQL installation guide should be followed in order to configure the storage properly.

Environment variables

Selenium

An env var can be configured in order to make Selenium work. Selenium, in order to run, needs the path to the chrome driver in your computer.

For Linux users:

export NOMAD_LIST_CHROME_DRIVER_PATH = '/path/to/chrome/driver'

or

NOMAD_LIST_CHROME_DRIVER_PATH='/path/to/chrome/driver' python3 main.py

For MacOs users:

brew install chromedriver

#Grant access:
xattr -d com.apple.quarantine /usr/local/bin/chromedriver

MySQL

In order to be able to save the nomad_list database the user needs to input the appropriate login information for the specific host server that they are using. The user is free to choose which MySQL host they want to use to save their nomad_list database. The login details of the host server can be saved by running the following code:

For Linux users:

export NOMAD_LIST_MYSQL_HOST = 'your_host'
export NOMAD_LIST_MYSQL_USER = 'your_user'
export NOMAD_LIST_MYSQL_PASSWORD = 'your_password'
export NOMAD_LIST_MYSQL_DATABASE = 'nomad_list'

or

NOMAD_LIST_MYSQL_HOST='your_host' NOMAD_LIST_MYSQL_USER='your_user' NOMAD_LIST_MYSQL_PASSWORD='your_password' NOMAD_LIST_MYSQL_DATABASE='nomad_list' python3 main.py

Environment file

A better option than running the above blocks of code is to create a .env in the root of the project. This will help to avoid having to pass each variable manually every time the Nomad List scrapper is run. The user needs to only set the value of the variables once in the .env like so:

NOMAD_LIST_CHROME_DRIVER_PATH='/path/to/chrome/driver'
NOMAD_LIST_MYSQL_HOST='your_host'
NOMAD_LIST_MYSQL_USER='your_user'
NOMAD_LIST_MYSQL_PASSWORD='your_password'
NOMAD_LIST_MYSQL_DATABASE='nomad_list'

Database Creation

The database can be created using one of two methods. The first method to create the database is to execute the following code in the MySQL Command Line:

mysql -u your_user -p your_password < create_schemas.sql

which runs the create_schema.sql file which contains the MySQL code for creating the nomad_list database.

The second method, is to execute the setup-db function provided in the command line interface CLI. This can be done by executing the following code in the terminal/command line:

python3 main.py setup-db

Usage

The Nomad List scrapper can be executed by navigating to the directory where the data-mining program has been saved and then running the main.py file. To navigating to the correct directory, run the following commands in the terminal (Linux/masOS) or command line (Windows):

cd <directory path>

In the same Terminal/ Command Line as before, after navigating to the directory where the Nomad List scrapper main.py file is saved, the Nomad List scrapper can be run by executing the following code:

python3 main.py

If the main.py file has been given execution permissions this command can be executed:

./main.py --help

which will display the manual of the CLI. Alternatively, if permission has not been granted, the equivalent code can be run:

python3 main.py --help

As mentioned above, the web scrapper can be run by loading the main.py file in a Python IDE like Pycharm or IDLE and running the file there with a special configuration. It will need the corresponding arguments, scrape [OPTIONS], (as discussed below) in order to fetch the desired cities from Nomad List, and store them in the nomad_list MySQL database.

The code in main.py will call the NomadListScrapper class in the nomad_list_scrapper.py file, which is responsible for handling the scrapper in the Nomad List website. The class NomadListScrapper creates an object of the CityScrapper class, which is the class that actually scrapes the information from the website, and it does so by accessing the various scrapping functions in the TabScrapper class in tab_scrapper.py file.

The web scrapper, once it has completed scrapping all the information from Nomad List, saves all the data to a MySQL database nomad_list.

Command Line Interface (CLI)

The command line interface grants the user much greater flexibility and functionality when it comes to using the Nomad List scrapper as well as scrapping Nomad List. There are 3 main methods for executing the Nomad List scrapper in the command line:

1. Database setup: Creates the necessary schemas to store the scrape data into a MySQL database.

   python3 main.py setup-db [-h] [--verbose]
   
   Options:
     -h, --help
     -v, --verbose: Enable verbosity.

2. Scrape cities from Nomad List: Scrap specific cities from the Nomad List website.

   python3 main.py scrape [-h] [--num-of-cities NUM_OF_CITIES] [--scrolls SCROLLS] [--verbose]
   
   optional arguments:
     -h, --help
     -n --num-of-cities:   Number of required cities.
     -s --scrolls:         Number of scrolls to make in the site to fetch the cities.
     -v --verbose:         Enable verbosity.

3. Show the scrapped cities: Fetch cities stored in the nomad_list database that match the user specified filters.

   python3 main.py show [-h] [--num-of-cities NUM_OF_CITIES] [--country COUNTRY] [--continent CONTINENT]
                          [--rank-from RANK_FROM] [--rank-to RANK_TO]
                          [--sorted-by SORTING_CRITERIA] [--order SORTING_ORDER] [--output {table,json,csv}] [--verbose]
   
   Options:
     -h, --help
     -n, --num-of-cities:  Number of required cities.
     --country:            Name of the country.
     --continent:          Name of the continent.
     --rank-from:          From rank <rank-from>.
     --rank-to:            To rank <rank-to>.
     --sorted-by:          Sorting criteria. Default: rank.
                           {rank,name,country,continent,overall score,cost,internet,fun,safety}
     --order:              Order of sorting. Default: ASC.
                           {ASC,DESC}
     -o, --output:         Output format. Default: table.
     -v, --verbose:        Enable verbosity.

scrape

The scrape Command Line Function (CLF) can be executed by running the following code in the CLI:

python main.py scrape

The scrape function gives the user the capability to specify how many cities they want the scrapper to scrape. The user can do this in one of two ways: either by specifying how many cities they want the scrapper to scrape or by specifying how far they want the scrapper to scroll-down the Nomad List homepage.

For example, if the user would like to specify to the Nomad List Scrapper that the number of cities it must scrape is 10, the following code will do that:

python main.py scrape -n 10

or

python main.py scrape --num-of-cities 10

If on the other hand the user wants to scroll 5 times and take all the visible cities, then the following code would be run:

python main.py scrape -s 5 

or

python main.py scrape --scrolls 5

show_by

The show_by Command Line Function (CLF) can be executed by running the following code in the CLI:

python main.py show

The show function allows the user to filter through the scrapped data stored in the nomad_list database in order to search for cities that meet specific requirements or criteria. Such requirements can be which country or continent the city is situated, what the rank of the city is, or what is the Cost of living for nomad.

For example, if the user wants to find which are the top 10 cities in Europe for a nomad to go to based on the overall rank of the city, the code to implement this query is:

python main.py show -n 10 --continent 'Europe' --sorted-by 'rank' --order 'DESC' 

Autocompletion

To take advantage of the autocomplete, the argcomplete module was installed.

To configure it, with all the requirements installed, it is necessary to run the following commands in the terminal:

Linux:

activate-global-python-argcomplete --user

sudo mv ~/.bash_completion.d/python-argcomplete /etc/.bash_completion.d

It's not mandatory to move the autocompletion file to the /etc/.bash_completion.d directory but, this dependent on it being globally accessible.

After resetting the terminal, the autocomplete will be ready to be used.

Storage

ERD

Project Status

Project is: in progress

Roadmap

This project is at the second check point of the roadmap and further developments will be released in the near future. Some the intended functionality of this program are:

• Scrape data source (Checkpoint #1)
• Design db using MySQL and store the data (Checkpoint #2)
• Add a CLI for interacting with the scrapper (Checkpoint #2)
• Enrich the data using external APIs (Checkpoint #3)
• Deploy the code to AWS (Checkpoint #3)
• Build a dashboard using ReDash (Checkpoint #4)

Authors

• David Demby (@david613)
• Kevin Szuchet (@kevinszuchet)

Support

If you have any queries regarding this program, please feel free to contact one of the authors.