coelacanth 0.1.6 → 0.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: '0172679388ce0d8718122988de86589b623e0ab432f9e08b589dd22e2009f74a'
-  data.tar.gz: baa0edff8a94564e33cc8596e44cfe49a040dd920798a10bbca824c2ce4d1f89
+  metadata.gz: 320be95cd3ffe6b905abacda5aa1a7fad00324bd79eafbe9969c3f80ed2695b5
+  data.tar.gz: f53305c6f1a07080ce16412ad4d1eb272691f27f50ccb468ee6f8bcb9bf69382
 SHA512:
-  metadata.gz: 16564c8664acf81dddb9188f828ab7e59da5969c610302e64325e6233f928a908e404332b244c48f3b9e0a7e984d88d3713ca885600cd1cd9a6e3057e273630a
-  data.tar.gz: 543fff73d944bf4b328e0bd3cebf975b9aaa093e2553d36c61ef09cefd1cc76ac997387cebba43580f059d497899006107a6bb0473afc2524f10423fa3cfad54
+  metadata.gz: 3201b30bcde133d5edc963d2522627082783df3565bff9644c8a9c33e2ade577cc8f3ca80b97d6a47ccfa9fc9734a768b94842386883604dd60b69176cc636c0
+  data.tar.gz: 9faaed9ade6c3e6258b7480fe6b2007a715f6b338413b09079b8e523823c4392ee82ec8fe5570dd71236d126c75b7e59fa0161380c5ea15c5df8a62f3d697c74

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    coelacanth (0.1.6)
+    coelacanth (0.2.0)
 
 GEM
   remote: https://rubygems.org/
@@ -31,7 +31,8 @@ GEM
     rainbow (3.1.1)
     rake (13.1.0)
     regexp_parser (2.9.0)
-    rexml (3.2.6)
+    rexml (3.3.6)
+      strscan
     rspec (3.10.0)
       rspec-core (~> 3.10.0)
       rspec-expectations (~> 3.10.0)
@@ -62,6 +63,7 @@ GEM
       ansi
       ast
     ruby-progressbar (1.13.0)
+    strscan (3.1.0)
     unicode-display_width (2.5.0)
     webrick (1.8.1)
     websocket-driver (0.7.6)

data/README.md

@@ -50,11 +50,12 @@ Then, you can easily parse and extract information from a web page like this:
 
 ```ruby
 url = "https://example.com"
-stats = Coelacanth.analyze(url) # Not yet implement
+stats = Coelacanth.analyze(url)
 ```
 
 ## Features
-- More features coming soon!
+- Get dom by oga
+- Get screenshot
 
 ## Commit Message Guidelines
 

data/compose.yml

@@ -18,6 +18,5 @@ services:
       - app-tier
   chrome:
     image: browserless/chrome:latest
-    container_name: chrome
     networks:
       - app-tier

data/config/coelacanth.yml

@@ -1,5 +1,4 @@
 development: &development
-  use_remote_client: false
   remote_client:
     ws_url: "ws://chrome:3000/chrome"
     timeout: 10 # seconds

data/lib/coelacanth/client.rb

@@ -1,76 +1,31 @@
 # frozen_string_literal: true
 
 require "ferrum"
-require "oga"
 
 module Coelacanth
   # Coelacanth::Client
   class Client
-    def initialize(url = nil)
+    def initialize(url)
+      @validator = Validator.new
+      raise URI::InvalidURIError unless @validator.valid_url?(url)
       @config = Coelacanth.config
-      @url = url if url && valid_url?(url)
-    end
-
-    def valid_url?(url = nil)
-      @url = url if url
-      uri = URI.parse(@url)
-      uri.is_a?(URI::HTTP) || uri.is_a?(URI::HTTPS)
-    rescue URI::InvalidURIError
-      false
-    end
-
-    def resolve_redirect(url = nil, limit = 10)
-      @url = url if url && valid_url?(url)
-      raise Coelacanth::DeepRedirectError, "Too many redirect" if limit.zero?
-      raise Coelacanth::RedirectError, "Url or location is nil" if @url.nil?
-
-      get_response(@url)
-      handle_response(@origin_response, limit)
-    end
-
-    def oga(url = nil)
-      @url = url if url && valid_url?(url)
-      Oga.parse_xml(get_response(@url))
+      remote_client.goto(url)
     end
 
     def get_response(url = nil)
-      @url = url if url && valid_url?(url)
-      if @config.read("use_remote_client")
-        response_by_remote_client
-      else
-        response_by_net_http
-      end
-    end
-
-    private
-
-    def handle_response(response, limit)
-      codes = Net::HTTPResponse::CODE_CLASS_TO_OBJ.invert
-      case @status_code.to_s
-      when /^#{codes[Net::HTTPSuccess]}\d\d$/
-        @url
-      when /^#{codes[Net::HTTPRedirection]}\d\d$/
-        @url = response["location"]
-        resolve_redirect(response["location"], limit - 1)
-      else
-        raise Coelacanth::RedirectError
-      end
-    end
-
-    def response_by_remote_client
-      remote_client.goto(@url)
       @status_code = remote_client.network.status
       @origin_response = remote_client
       remote_client.body
     end
 
-    def response_by_net_http
-      response = Net::HTTP.get_response(URI.parse(@url))
-      @status_code = response.code
-      @origin_response = response
-      response.body
+    def get_screenshot
+      tempfile = Tempfile.new
+      remote_client.screenshot(path: tempfile.path, format: "png")
+      File.read(tempfile.path)
     end
 
+    private
+
     def remote_client
       if @remote_client.nil?
         headers = @config.read("remote_client.headers")

data/lib/coelacanth/dom.rb

@@ -0,0 +1,12 @@
+# frozen_string_literal: true
+
+require "oga"
+
+module Coelacanth
+  # Coelacanth::Dom
+  class Dom
+    def oga(url)
+      Oga.parse_xml(Client.new(url).get_response)
+    end
+  end
+end

data/lib/coelacanth/redirect.rb

@@ -0,0 +1,36 @@
+# frozen_string_literal: true
+
+require "ferrum"
+require "oga"
+
+module Coelacanth
+  # Coelacanth::Redirect
+  class Redirect
+    def resolve_redirect(url, limit = 10)
+      @url = url if url && Validator.new.valid_url?(url)
+      raise Coelacanth::DeepRedirectError, "Too many redirect" if limit.zero?
+      raise Coelacanth::RedirectError, "Url or location is nil" if @url.nil?
+
+      response = Net::HTTP.get_response(URI.parse(@url))
+      @status_code = response.code
+      @origin_response = response
+
+      handle_response(@origin_response, limit)
+    end
+
+    private
+
+    def handle_response(response, limit)
+      codes = Net::HTTPResponse::CODE_CLASS_TO_OBJ.invert
+      case @status_code.to_s
+      when /^#{codes[Net::HTTPSuccess]}\d\d$/
+        @url
+      when /^#{codes[Net::HTTPRedirection]}\d\d$/
+        @url = response["location"]
+        resolve_redirect(response["location"], limit - 1)
+      else
+        raise Coelacanth::RedirectError
+      end
+    end
+  end
+end

data/lib/coelacanth/validator.rb

@@ -0,0 +1,15 @@
+# frozen_string_literal: true
+
+require "ferrum"
+
+module Coelacanth
+  # Coelacanth::Validator
+  class Validator
+    def valid_url?(url)
+      uri = URI.parse(url)
+      uri.is_a?(URI::HTTP) || uri.is_a?(URI::HTTPS)
+    rescue URI::InvalidURIError
+      false
+    end
+  end
+end

data/lib/coelacanth/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Coelacanth
-  VERSION = "0.1.6"
+  VERSION = "0.2.0"
 end

data/lib/coelacanth.rb

@@ -1,9 +1,12 @@
 # frozen_string_literal: true
 
 require "net/http"
-require_relative "coelacanth/version"
 require_relative "coelacanth/configure"
 require_relative "coelacanth/client"
+require_relative "coelacanth/dom"
+require_relative "coelacanth/redirect"
+require_relative "coelacanth/validator"
+require_relative "coelacanth/version"
 
 # Coelacanth
 module Coelacanth
@@ -13,10 +16,10 @@ module Coelacanth
 
   def self.analyze(url)
     @client = Client.new(url)
-    @client.resolve_redirect
+    regular_url = Redirect.new.resolve_redirect(url)
     {
-      remote_client: @config.read("use_remote_client"),
-      oga: @client.oga
+      dom: Dom.new.oga(regular_url),
+      screenshot: @client.get_screenshot,
     }
   end
 

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: coelacanth
 version: !ruby/object:Gem::Version
-  version: 0.1.6
+  version: 0.2.0
 platform: ruby
 authors:
 - Yusuke
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2024-08-29 00:00:00.000000000 Z
+date: 2024-09-12 00:00:00.000000000 Z
 dependencies: []
 description: |
   coelacanth is a gem that allows you to easily parse and analyze web pages,
@@ -21,7 +21,6 @@ extra_rdoc_files: []
 files:
 - ".rspec"
 - ".rubocop.yml"
-- CHANGELOG.md
 - CODE_OF_CONDUCT.md
 - Dockerfile
 - Gemfile
@@ -34,6 +33,9 @@ files:
 - lib/coelacanth.rb
 - lib/coelacanth/client.rb
 - lib/coelacanth/configure.rb
+- lib/coelacanth/dom.rb
+- lib/coelacanth/redirect.rb
+- lib/coelacanth/validator.rb
 - lib/coelacanth/version.rb
 homepage: https://github.com/slidict/coelacanth
 licenses:

data/CHANGELOG.md

@@ -1,5 +0,0 @@
-## [Unreleased]
-
-## [0.1.0] - 2024-08-24
-
-- Initial release