btscraper 0.1.1 → 0.1.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: af8c80fe5552fae4d6b3ed6d4c8dffe4122f0ac962151ee1339fe2adc36794a8
-  data.tar.gz: 91f557ff0af7a7b256b92ff69f98b74c3be8de2e8363c42e0cad87d090562f6b
+  metadata.gz: 213c402de8c6be9cb33d77c557fbdef58e5737ef08e0f06e6b4fca2676c79442
+  data.tar.gz: '0810754bca603271dcf2d57c016272ce8e13d6a0c136844c63449b04e3b0d64f'
 SHA512:
-  metadata.gz: a724c9161f53173a751c0abd9268c42e9defa63a52bf4c5b91f29c09141c1a29ee3457377b9e981fe87c0d914d36e4013c19adf6ad832ec0805df7f0edf48f5a
-  data.tar.gz: 6ecbd666a84829abc9c2f9d6e4647f3e10e45aa13a0411359fc6181ba00a144b2458a1933d0fe4045531b0f8d87aa9738d9c828db3bfc2783ea3c86e03fc6b5e
+  metadata.gz: 37d768fc93959230ada23bf5d445ab573de61479e3a03661193063306842f14c6597c0551a10266d4275d342332ab823e0c0ebdce2ad69b93aab70d21d38b6e5
+  data.tar.gz: 94d4f3064ea64239ac7154da5533dd54cadc0b04f3ea3429f0d19aef6f7cdfd9476576e7621fb208ba3d108a2719d64112236083073fbb07f81f066f9b2c1656

data/lib/btscraper/checkinfohash.rb

@@ -0,0 +1,10 @@
+module BTScraper
+  Sha1_regex = /^[0-9a-f]{40}$/ # Regex to check SHA1 infohashes
+
+  # This method checks if the infohashes are valid
+  # @raise [BTScraperError] If the infohash is invalid
+  def self.check_info_hash(info_hash)
+    info_hash.each{|x| raise BTScraperError, 'Invalid infohash provided' unless x.match?(Sha1_regex)}
+  end
+  
+end

data/lib/btscraper/httpscrape.rb

@@ -0,0 +1,60 @@
+require 'bencode'
+require 'httparty'
+require 'binascii'
+
+module BTScraper
+  
+  # @author sherkix 
+  # @see https://bittorrent.org/beps/bep_0048.html BEP 48
+  # This class permits you to scrape an HTTP torrent tracker according to the BEP 48
+  class HTTPScrape
+
+    attr_reader :tracker, :info_hash
+    # @!attribute [r] tracker
+    #  @return [String] returns tracker full url
+    #  
+    # @!attribute [r] info_hash
+    #  @return [Array<String>] returns array of infohashes
+    # 
+    # Create a new HTTPScrape object 
+    # 
+    # @param tracker [String] Bittorrent HTTP tracker server
+    # @param info_hash [Array<String>, String] Array of infohashes or single infohash
+    # 
+    # @raise [TypeError] if wrong type of argument is provided
+    # 
+    # @example Default usage
+    #  scrape_object = BTScraper::HTTPScrape.new('https://example.com:443/scrape', ['c22b5f9178342609428d6f51b2c5af4c0bde6a42'], ['aaf4c61ddcc5e8a2dabede0f3b482cd9aea9434d'])
+    #  scrape_object.scrape
+    def initialize(tracker, info_hash)
+      unless tracker.instance_of? String
+        raise TypeError, "String excpected, got #{tracker.class}"
+      end
+      unless info_hash.instance_of? String or info_hash.instance_of? Array
+        raise TypeError, "String or Array excpected, got #{info_hash.class}"
+      end
+
+      if info_hash.instance_of? String
+        info_hash.downcase!
+        BTScraper.check_info_hash Array(info_hash)
+      else
+        info_hash.map(&:downcase!)
+        BTScraper.check_info_hash info_hash
+      end
+      @tracker = tracker
+      @info_hash = Array(info_hash)
+    end
+    # @example Response example
+    #  {"files" => {"xxxxxxxxxxxxxxxxxxxxxxxxxx" => {"complete" => 8, "downloaded" => 9, "incomplete" => 4}, "yyyyyyyyyyyyyyyyyyyyyyyyyy" => {"complete" => 81, "downloaded" => 204, "incomplete" => 23}, "zzzzzzzzzzzzzzzzzzzzzzzzzz" => {"complete" => 3, "downloaded" => 26, "incomplete" => 1}}}
+    # @return [Hash] The method returns a hash with the scraped data
+    def scrape
+      unhex_info_hash = @info_hash.map{|x| Binascii.a2b_hex(x)}
+      params = unhex_info_hash.map{|h| "info_hash=#{CGI.escape(h.to_s)}"}.join('&')
+      begin
+        HTTParty.get(@tracker, :query => params, :headers => {'User-Agent' => "btscraper #{VERSION}"}, :timeout => 10).body.bdecode
+      rescue HTTParty::Error => e
+        raise BTScraperError, e
+      end
+    end
+  end
+end

data/lib/btscraper/udpscrape.rb

@@ -10,11 +10,10 @@ module BTScraper
   Actionerr = 3 # Scrape Error
   Defaulttimeout = 15 # Default timeout is 15s
   Retries = 8 # Maximum number of retransmission
-  Sha1_regex = /^[0-9a-f]{40}$/ # Regex to check SHA1 infohashes
   
   # @author sherkix 
   # @see https://bittorrent.org/beps/bep_0015.html BEP 15
-  # This class permits you to scrape an UDP torrent tracker according to BEP 15
+  # This class permits you to scrape an UDP torrent tracker according to the BEP 15
   class UDPScrape
     
     attr_reader :tracker, :info_hash, :hostname, :port
@@ -32,7 +31,7 @@ module BTScraper
     #
     # Create a new UDPScrape object 
     # 
-    # @param tracker [String] Bittorrent tracker server
+    # @param tracker [String] Bittorrent UDP tracker server
     # @param info_hash [Array<String>, String] Array of infohashes or single infohash
     #
     # 
@@ -50,17 +49,17 @@ module BTScraper
         raise TypeError, "String or Array excpected, got #{info_hash.class}"
       end
       
-      # Maximum number of infohashes is 74 according to BEP 15
+      # Maximum number of infohashes is 74
       if info_hash.instance_of? Array and info_hash.count > 74
         raise BTScraperError, 'The number of infohashes must be less than 74'
       end
 
       if info_hash.instance_of? String
         info_hash.downcase!
-        check_info_hash Array(info_hash)
+        BTScraper.check_info_hash Array(info_hash)
       else
         info_hash.map(&:downcase!)
-        check_info_hash info_hash
+        BTScraper.check_info_hash info_hash
       end
       @tracker = tracker 
       @hostname = URI(@tracker).hostname
@@ -72,7 +71,7 @@ module BTScraper
     # @return [Hash] The method returns a hash with the scraped data
     # @raise [BTScraperError] If the response is less than 8 bytes
     # @raise [BTScraperError] If the scraping request fails
-    # @raise [BTScraperError] If the tracker response with a different transaction_id provided by the client
+    # @raise [BTScraperError] If the tracker responds with a different transaction_id provided by the client
     # @raise [BTScraperError] After 8 timeouts
     def scrape
       attempt = 0
@@ -97,7 +96,7 @@ module BTScraper
         end
       rescue Timeout::Error
         attempt+=1
-        puts "#{attempt} Request to #{@hostname} timed out, retying after #{Defaulttimeout * 2 ** attempt}s"
+        puts "#{attempt} Request to #{@hostname} timed out, retying after #{Defaulttimeout * 2**attempt}s"
         retry if attempt <= Retries
         raise BTScraperError, 'Max retries exceeded'
       ensure
@@ -109,7 +108,7 @@ module BTScraper
     
     private
 
-    # @return [Array<Integer>] This method makes request to the bittorrent tracker to get a connection_id
+    # @return [Array<Integer>] This method makes a request to the bittorrent tracker to get a connection_id
     def get_connection_id
       attempt = 0
       client = connect_to_tracker
@@ -132,7 +131,7 @@ module BTScraper
         end
       rescue Timeout::Error
         attempt+=1
-        puts "#{attempt} Request to #{@hostname} timed out, retrying after #{Defaulttimeout * 2 ** attempt}s"
+        puts "#{attempt} Request to #{@hostname} timed out, retrying after #{Defaulttimeout * 2**attempt}s"
         retry if attempt <= Retries
         raise BTScraperError, 'Max retries exceeded'
       ensure
@@ -151,10 +150,6 @@ module BTScraper
       rand(0..4294967295)
     end
 
-    def check_info_hash(info_hash)
-      info_hash.each{|x| raise BTScraperError, 'Invalid infohash provided' unless x.match?(Sha1_regex)}
-    end
-
     def create_scrape_hash(info_hash, response, hash)
       i = 2
       info_hash.each do |x|

data/lib/btscraper.rb

@@ -1,9 +1,10 @@
 
 # Scrape library for bittorrent trackers
+# @author sherkix
 
 module BTScraper
   # @!visibility private
-  VERSION = '0.1.1'
+  VERSION = '0.1.2'
 
   # Base class for exceptions
   class BTScraperError < StandardError

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: btscraper
 version: !ruby/object:Gem::Version
-  version: 0.1.1
+  version: 0.1.2
 platform: ruby
 authors:
 - Sherkix
@@ -51,6 +51,34 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: bencode
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
+  name: httparty
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: minitest
   requirement: !ruby/object:Gem::Requirement
@@ -72,6 +100,8 @@ extensions: []
 extra_rdoc_files: []
 files:
 - lib/btscraper.rb
+- lib/btscraper/checkinfohash.rb
+- lib/btscraper/httpscrape.rb
 - lib/btscraper/udpscrape.rb
 homepage: https://github.com/sherkix/btscraper
 licenses:
@@ -91,7 +121,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.7.2
+rubygems_version: 4.0.0
 specification_version: 4
 summary: Scrape library for bittorrent trackers
 test_files: []