kimurai_dynamic 1.4.1

data/lib/kimurai/runner.rb

@@ -0,0 +1,60 @@
+require 'pmap'
+
+module Kimurai
+  class Runner
+    attr_reader :jobs, :spiders, :session_info
+
+    def initialize(spiders, parallel_jobs)
+      @jobs = parallel_jobs
+      @spiders = spiders
+      @start_time = Time.now
+
+      @session_info = {
+        id: @start_time.to_i,
+        status: :processing,
+        start_time: @start_time,
+        stop_time: nil,
+        environment: Kimurai.env,
+        concurrent_jobs: @jobs,
+        spiders: @spiders
+      }
+
+      if time_zone = Kimurai.configuration.time_zone
+        Kimurai.time_zone = time_zone
+      end
+
+      ENV.store("SESSION_ID", @start_time.to_i.to_s)
+      ENV.store("RBCAT_COLORIZER", "false")
+    end
+
+    def run!(exception_on_fail: true)
+      puts ">>> Runner: started: #{session_info}"
+      if at_start_callback = Kimurai.configuration.runner_at_start_callback
+        at_start_callback.call(session_info)
+      end
+
+      running = true
+      spiders.peach_with_index(jobs) do |spider, i|
+        next unless running
+
+        puts "> Runner: started spider: #{spider}, index: #{i}"
+        pid = spawn("bundle", "exec", "kimurai", "crawl", spider, [:out, :err] => "log/#{spider}.log")
+        Process.wait pid
+
+        puts "< Runner: stopped spider: #{spider}, index: #{i}"
+      end
+    rescue StandardError, SignalException, SystemExit => e
+      running = false
+
+      session_info.merge!(status: :failed, error: e.inspect, stop_time: Time.now)
+      exception_on_fail ? raise(e) : [session_info, e]
+    else
+      session_info.merge!(status: :completed, stop_time: Time.now)
+    ensure
+      if at_stop_callback = Kimurai.configuration.runner_at_stop_callback
+        at_stop_callback.call(session_info)
+      end
+      puts "<<< Runner: stopped: #{session_info}"
+    end
+  end
+end

data/lib/kimurai/template/.gitignore

@@ -0,0 +1,18 @@
+/.bundle
+/cache
+/node_modules
+
+/log/*
+!/log/.keep
+
+/tmp/*
+!/tmp/.keep
+
+/db/*
+!/db/.keep
+
+.byebug_history
+*.swp
+.env
+
+capybara-*.png

data/lib/kimurai/template/Gemfile

@@ -0,0 +1,28 @@
+source 'https://rubygems.org'
+git_source(:github) { |repo| "https://github.com/#{repo}.git" }
+
+ruby '>= 2.5'
+
+# Framework
+gem 'kimurai', '~> 1.4'
+
+# Require files in directory and child directories recursively
+gem 'require_all'
+
+# Dotenv
+gem 'dotenv'
+
+# To debug spiders:
+group :development do
+  gem 'byebug', platforms: :mri
+  gem 'pry'
+end
+
+# If you want to save items to the database, require one of these gems:
+# gem 'sqlite3'
+# gem 'pg'
+# gem 'mysql2'
+
+# And use your preferred ORM/database connector:
+# gem 'activerecord', require: 'active_record'
+# gem 'sequel'

data/lib/kimurai/template/README.md

@@ -0,0 +1,3 @@
+# README
+
+New Kimurai project readme

data/lib/kimurai/template/config/application.rb

@@ -0,0 +1,37 @@
+Kimurai.configure do |config|
+  # Default logger has colored mode in development.
+  # If you would like to disable it, set `colorize_logger` to false.
+  # config.colorize_logger = false
+
+  # Logger level for default logger:
+  # config.log_level = :info
+
+  # Custom logger:
+  # config.logger = Logger.new(STDOUT)
+
+  # Custom time zone (for logs):
+  # config.time_zone = "UTC"
+  # config.time_zone = "Europe/Moscow"
+
+  # At start callback for a runner. Accepts argument with info as hash with
+  # keys: id, status, start_time, environment, concurrent_jobs, spiders list.
+  # For example, you can use this callback to send notification when runner was started:
+  # config.runner_at_start_callback = lambda do |info|
+  #   json = JSON.pretty_generate(info)
+  #   Sender.send_notification("Started session: #{json}")
+  # end
+
+  # At stop callback for a runner. Accepts argument with info as hash with
+  # all `runner_at_start_callback` keys plus additional `stop_time` key. Also `status` contains
+  # stop status of a runner (completed or failed).
+  # You can use this callback to send notification when runner has been stopped:
+  # config.runner_at_stop_callback = lambda do |info|
+  #   json = JSON.pretty_generate(info)
+  #   Sender.send_notification("Stopped session: #{json}")
+  # end
+
+  # Provide custom chrome binary path (default is any available chrome/chromium in the PATH):
+  # config.selenium_chrome_path = "/usr/bin/chromium-browser"
+  # Provide custom selenium chromedriver path (default is "/usr/local/bin/chromedriver"):
+  # config.chromedriver_path = "/usr/local/bin/chromedriver"
+end

data/lib/kimurai/template/config/automation.yml

@@ -0,0 +1,13 @@
+# software versions to install for `setup` command
+setup:
+  ruby: 2.5.1
+  # check latest here http://phantomjs.org/download.html
+  phantomjs: 2.1.1
+  # check latest here https://github.com/mozilla/geckodriver/releases/
+  geckodriver: 0.21.0
+  # check latest here https://sites.google.com/a/chromium.org/chromedriver/downloads
+  chromedriver: 2.39
+# settings for deploy command, you can use cli options as well (--repo-url, --git-key-path)
+deploy:
+  # repo_url: git@bitbucket.org:username/repo_name.git
+  # repo_key_path: ~/.ssh/id_rsa

data/lib/kimurai/template/config/boot.rb

@@ -0,0 +1,22 @@
+# require project gems
+require 'bundler/setup'
+Bundler.require(:default, Kimurai.env)
+
+# require custom ENV variables located in .env file
+require 'dotenv/load'
+
+# require initializers
+Dir.glob(File.join("./config/initializers", "*.rb"), &method(:require))
+
+# require helpers
+Dir.glob(File.join("./helpers", "*.rb"), &method(:require))
+
+# require pipelines
+Dir.glob(File.join("./pipelines", "*.rb"), &method(:require))
+
+# require spiders recursively in the `spiders/` folder
+require_relative '../spiders/application_spider'
+require_all "spiders"
+
+# require Kimurai configuration
+require_relative 'application'

data/lib/kimurai/template/config/schedule.rb

@@ -0,0 +1,57 @@
+### Settings ###
+require 'tzinfo'
+
+# Export current PATH to the cron
+env :PATH, ENV["PATH"]
+
+# Use 24 hour format when using `at:` option
+set :chronic_options, hours24: true
+
+# Use local_to_utc helper to setup execution time using your local timezone instead
+# of server's timezone (which is probably and should be UTC, to check run `$ timedatectl`).
+# Also maybe you'll want to set same timezone in kimurai as well (use `Kimurai.configuration.time_zone =` for that),
+# to have spiders logs in a specific time zone format.
+# Example usage of helper:
+# every 1.day, at: local_to_utc("7:00", zone: "Europe/Moscow") do
+#   crawl "google_spider.com", output: "log/google_spider.com.log"
+# end
+def local_to_utc(time_string, zone:)
+  TZInfo::Timezone.get(zone).local_to_utc(Time.parse(time_string))
+end
+
+# Note: by default Whenever exports cron commands with :environment == "production".
+# Note: Whenever can only append log data to a log file (>>). If you want
+# to overwrite (>) log file before each run, pass lambda:
+# crawl "google_spider.com", output: -> { "> log/google_spider.com.log 2>&1" }
+
+# Project job types
+job_type :crawl,  "cd :path && KIMURAI_ENV=:environment bundle exec kimurai crawl :task :output"
+job_type :runner, "cd :path && KIMURAI_ENV=:environment bundle exec kimurai runner --jobs :task :output"
+
+# Single file job type
+job_type :single, "cd :path && KIMURAI_ENV=:environment ruby :task :output"
+# Single with bundle exec
+job_type :single_bundle, "cd :path && KIMURAI_ENV=:environment bundle exec ruby :task :output"
+
+### Schedule ###
+# Usage (check examples here https://github.com/javan/whenever#example-schedulerb-file):
+# every 1.day do
+  # Example to schedule a single spider in the project:
+  # crawl "google_spider.com", output: "log/google_spider.com.log"
+
+  # Example to schedule all spiders in the project using runner. Each spider will write
+  # it's own output to the `log/spider_name.log` file (handled by a runner itself).
+  # Runner output will be written to log/runner.log file.
+  # Argument number it's a count of concurrent jobs:
+  # runner 3, output:"log/runner.log"
+
+  # Example to schedule single spider (without project):
+  # single "single_spider.rb", output: "single_spider.log"
+# end
+
+### How to set a cron schedule ###
+# Run: `$ whenever --update-crontab --load-file config/schedule.rb`.
+# If you don't have whenever command, install the gem: `$ gem install whenever`.
+
+### How to cancel a schedule ###
+# Run: `$ whenever --clear-crontab --load-file config/schedule.rb`.

data/lib/kimurai/template/helpers/application_helper.rb

@@ -0,0 +1,3 @@
+module ApplicationHelper
+  # Put here custom methods which are will be available for any spider
+end

data/lib/kimurai/template/pipelines/saver.rb

@@ -0,0 +1,11 @@
+class Saver < Kimurai::Pipeline
+  def process_item(item, options: {})
+    # Here you can save item to the database, send it to a remote API or
+    # simply save item to a file format using `save_to` helper:
+
+    # To get the name of a current spider: `spider.class.name`
+    # save_to "db/#{spider.class.name}.json", item, format: :pretty_json
+
+    item
+  end
+end

data/lib/kimurai/template/pipelines/validator.rb

@@ -0,0 +1,24 @@
+class Validator < Kimurai::Pipeline
+  def process_item(item, options: {})
+    # Here you can validate item and raise `DropItemError`
+    # if one of the validations failed. Examples:
+
+    # Check item sku for uniqueness using buit-in `unique?` helper:
+    # unless unique?(:sku, item[:sku])
+    #   raise DropItemError, "Item sku is not unique"
+    # end
+
+    # Drop item if title length shorter than 5 symbols:
+    # if item[:title].size < 5
+    #   raise DropItemError, "Item title is short"
+    # end
+
+    # Drop item if it doesn't contains any images:
+    # unless item[:images].present?
+    #   raise DropItemError, "Item images are not present"
+    # end
+
+    # Pass item to the next pipeline (if it wasn't dropped)
+    item
+  end
+end

data/lib/kimurai/template/spiders/application_spider.rb

@@ -0,0 +1,143 @@
+# ApplicationSpider is a default base spider class. You can set here
+# default settings for all spiders inherited from ApplicationSpider.
+# To generate a new spider, run: `$ kimurai generate spider spider_name`
+
+class ApplicationSpider < Kimurai::Base
+  include ApplicationHelper
+
+  # Default engine for spiders (available engines: :mechanize, :poltergeist_phantomjs,
+  # :selenium_firefox, :selenium_chrome)
+  @engine = :poltergeist_phantomjs
+
+  # Pipelines list, by order.
+  # To process item through pipelines pass item to the `send_item` method
+  @pipelines = [:validator, :saver]
+
+  # Default config. Set here options which are default for all spiders inherited
+  # from ApplicationSpider. Child's class config will be deep merged with this one
+  @config = {
+    # Custom headers, format: hash. Example: { "some header" => "some value", "another header" => "another value" }
+    # Works only for :mechanize and :poltergeist_phantomjs engines (Selenium doesn't allow to set/get headers)
+    # headers: {},
+
+    # Custom User Agent, format: string or lambda.
+    # Use lambda if you want to rotate user agents before each run:
+    # user_agent: -> { ARRAY_OF_USER_AGENTS.sample }
+    # Works for all engines
+    # user_agent: "Mozilla/5.0 Firefox/61.0",
+
+    # Custom cookies, format: array of hashes.
+    # Format for a single cookie: { name: "cookie name", value: "cookie value", domain: ".example.com" }
+    # Works for all engines
+    # cookies: [],
+
+    # Proxy, format: string or lambda. Format of a proxy string: "ip:port:protocol:user:password"
+    # `protocol` can be http or socks5. User and password are optional.
+    # Use lambda if you want to rotate proxies before each run:
+    # proxy: -> { ARRAY_OF_PROXIES.sample }
+    # Works for all engines, but keep in mind that Selenium drivers doesn't support proxies
+    # with authorization. Also, Mechanize doesn't support socks5 proxy format (only http)
+    # proxy: "3.4.5.6:3128:http:user:pass",
+
+    # If enabled, browser will ignore any https errors. It's handy while using a proxy
+    # with self-signed SSL cert (for example Crawlera or Mitmproxy)
+    # Also, it will allow to visit webpages with expires SSL certificate.
+    # Works for all engines
+    ignore_ssl_errors: true,
+
+    # Custom window size, works for all engines
+    # window_size: [1366, 768],
+
+    # Skip images downloading if true, works for all engines
+    disable_images: true,
+
+    # Selenium engines only: headless mode, `:native` or `:virtual_display` (default is :native)
+    # Although native mode has a better performance, virtual display mode
+    # sometimes can be useful. For example, some websites can detect (and block)
+    # headless chrome, so you can use virtual_display mode instead
+    # headless_mode: :native,
+
+    # This option tells the browser not to use a proxy for the provided list of domains or IP addresses.
+    # Format: array of strings. Works only for :selenium_firefox and selenium_chrome
+    # proxy_bypass_list: [],
+
+    # Option to provide custom SSL certificate. Works only for :poltergeist_phantomjs and :mechanize
+    # ssl_cert_path: "path/to/ssl_cert",
+
+    # Inject some JavaScript code to the browser.
+    # Format: array of strings, where each string is a path to JS file.
+    # Works only for poltergeist_phantomjs engine (Selenium doesn't support JS code injection)
+    # extensions: ["lib/code_to_inject.js"],
+
+    # Automatically skip duplicated (already visited) urls when using `request_to` method.
+    # Possible values: `true` or `hash` with options.
+    # In case of `true`, all visited urls will be added to the storage's scope `:requests_urls`
+    # and if url already contains in this scope, request will be skipped.
+    # You can configure this setting by providing additional options as hash:
+    # `skip_duplicate_requests: { scope: :custom_scope, check_only: true }`, where:
+    # `scope:` - use custom scope than `:requests_urls`
+    # `check_only:` - if true, then scope will be only checked for url, url will not
+    # be added to the scope if scope doesn't contains it.
+    # works for all drivers
+    # skip_duplicate_requests: true,
+
+    # Automatically skip provided errors while requesting a page.
+    # If raised error matches one of the errors in the list, then this error will be caught,
+    # and request will be skipped.
+    # It is a good idea to skip errors like NotFound(404), etc.
+    # Format: array where elements are error classes or/and hashes. You can use hash format
+    # for more flexibility: `{ error: "RuntimeError", message: "404 => Net::HTTPNotFound" }`.
+    # Provided `message:` will be compared with a full error message using `String#include?`. Also
+    # you can use regex instead: `{ error: "RuntimeError", message: /404|403/ }`.
+    # skip_request_errors: [{ error: RuntimeError, message: "404 => Net::HTTPNotFound" }],
+
+    # Automatically retry provided errors with a few attempts while requesting a page.
+    # If raised error matches one of the errors in the list, then this error will be caught
+    # and the request will be processed again within a delay. There are 3 attempts:
+    # first: delay 15 sec, second: delay 30 sec, third: delay 45 sec.
+    # If after 3 attempts there is still an exception, then the exception will be raised.
+    # It is a good idea to try to retry errros like `ReadTimeout`, `HTTPBadGateway`, etc.
+    # Format: same like for `skip_request_errors` option.
+    # retry_request_errors: [Net::ReadTimeout],
+
+    # Handle page encoding while parsing html response using Nokogiri. There are two modes:
+    # Auto (`:auto`) (try to fetch correct encoding from <meta http-equiv="Content-Type"> or <meta charset> tags)
+    # Set required encoding manually, example: `encoding: "GB2312"` (Set required encoding manually)
+    # Default this option is unset.
+    # encoding: nil,
+
+    # Restart browser if one of the options is true:
+    restart_if: {
+      # Restart browser if provided memory limit (in kilobytes) is exceeded (works for all engines)
+      # memory_limit: 350_000,
+
+      # Restart browser if provided requests limit is exceeded (works for all engines)
+      # requests_limit: 100
+    },
+
+    # Perform several actions before each request:
+    before_request: {
+      # Change proxy before each request. The `proxy:` option above should be presented
+      # and has lambda format. Works only for poltergeist and mechanize engines
+      # (Selenium doesn't support proxy rotation).
+      # change_proxy: true,
+
+      # Change user agent before each request. The `user_agent:` option above should be presented
+      # and has lambda format. Works only for poltergeist and mechanize engines
+      # (selenium doesn't support to get/set headers).
+      # change_user_agent: true,
+
+      # Clear all cookies before each request, works for all engines
+      # clear_cookies: true,
+
+      # If you want to clear all cookies + set custom cookies (`cookies:` option above should be presented)
+      # use this option instead (works for all engines)
+      # clear_and_set_cookies: true,
+
+      # Global option to set delay between requests.
+      # Delay can be `Integer`, `Float` or `Range` (`2..5`). In case of a range,
+      # delay number will be chosen randomly for each request: `rand (2..5) # => 3`
+      # delay: 1..3
+    }
+  }
+end

data/lib/kimurai/version.rb

@@ -0,0 +1,3 @@
+module Kimurai
+  VERSION = "1.4.1"
+end

data/lib/kimurai.rb

@@ -0,0 +1,54 @@
+require 'ostruct'
+require 'logger'
+require 'json'
+require 'active_support'
+require 'active_support/core_ext'
+require 'rbcat'
+
+require_relative 'kimurai/version'
+
+require_relative 'kimurai/core_ext/numeric'
+require_relative 'kimurai/core_ext/string'
+require_relative 'kimurai/core_ext/array'
+require_relative 'kimurai/core_ext/hash'
+
+require_relative 'kimurai/browser_builder'
+require_relative 'kimurai/base_helper'
+require_relative 'kimurai/pipeline'
+require_relative 'kimurai/base'
+
+module Kimurai
+  class << self
+    def configuration
+      @configuration ||= OpenStruct.new
+    end
+
+    def configure
+      yield(configuration)
+    end
+
+    def env
+      ENV.fetch("KIMURAI_ENV") { "development" }
+    end
+
+    def time_zone
+      ENV["TZ"]
+    end
+
+    def time_zone=(value)
+      ENV.store("TZ", value)
+    end
+
+    def list
+      Base.descendants.map do |klass|
+        next unless klass.name
+        [klass.name, klass]
+      end.compact.to_h
+    end
+
+    def find_by_name(name)
+      return unless name
+      Base.descendants.find { |klass| klass.name == name }
+    end
+  end
+end