# treewalker

[![Build Status](https://github.com/AntoineGagne/treewalker/actions/workflows/erlang.yml/badge.svg)](https://github.com/AntoineGagne/treewalker/actions)
[![Hex Pm](http://img.shields.io/hexpm/v/treewalker.svg?style=flat)](https://hex.pm/packages/treewalker)
[![Coverage](https://coveralls.io/repos/github/AntoineGagne/treewalker/badge.svg?branch=master)](https://coveralls.io/github/AntoineGagne/treewalker?branch=master)

A web crawler in Erlang that respects `robots.txt`.

## Installation

This library is available on [hex.pm](https://hex.pm/packages/treewalker).

Keep in mind that the library is not yet stable and its API may be
subject to changes.

## Usage

```erlang
%% This will add the specified crawler to the supervision tree
{ok, _} = treewalker:add_crawler(example, #{scraper => example_scraper,
                                            fetcher => example_fetcher,
                                            max_depth => 3,
                                            link_filter => example_link_filter,
                                            store => example_store}),
%% Starts crawling
ok = treewalker:start_crawler(example),
%% ...
%% Stops the crawler
%% The pending requests will be completed and dropped
ok = treewalker:stop_crawler(example),
```

## Options

The following settings are available via the `sys.config` configuration:

```erlang
{treewalker, [
              %% The minimum delay to wait before retrying a failed request
              {min_retry_delay, pos_integer()},
              %% The maximum delay to wait before retrying a failed request
              {max_retry_delay, pos_integer()},
              %% The maximum amount of retries of a failed request
              {max_retries, pos_integer()},
              %% The maximum amount of delay before starting a request (in seconds)
              {max_worker_delay, pos_integer()},
              %% The maximum amount of concurrent workers making HTTP requests
              {max_concurrent_worker, pos_integer()},
              %% The user agent making the HTTP requests
              {user_agent, binary()}]},
```

## Development

### Running all the tests and linters

You can run all the tests and linters with the `rebar3` alias:

```sh
rebar3 check
```