jekyll-url-metadata 1.0.1 → 1.0.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: a3048e805f8d9c5aebcec7168ade03463791061b48eca4819656d9954b3a08e5
-  data.tar.gz: a62831160686205005734c3c4ded035627fda472f316778e9065f414c28f15c0
+  metadata.gz: 07f8f6a1a9019b147240b9c24ccdf325351a438430401937ee09586786318688
+  data.tar.gz: 8629259c9beed38570135fa3eebbadcd5b708e76bb33154fbae89fc54f95ddef
 SHA512:
-  metadata.gz: 8591ced4c0ba81dd04a28f84a45b2f1342b5117f020b3591d88ed98b471261d4ba77e1df7e31d9d97b8d3cb4aa4b24a397ec12eceebfa2ffacdf0673dcdd505a
-  data.tar.gz: e4102af544b8d0ad4a83b58ab641563e83b658639d9fbe85bc2a185449d3b5409f09bca434ab95ec88d37a77c5f325510e6d81e8e85ca00d9979edce6900b126
+  metadata.gz: 7b335e1dadb10c5cc7bb228e692b6f3aba28005b720e47cbf5c34033d5713194b8e8c603c607059285bae213bfc70d8736197f8dca2f30fdc5480347858d01cb
+  data.tar.gz: 4dc6c36426d180d2e83ed8eefe75bfca97687c342651bdc2a09235a33a8c5d986b768d82a2be0d7d88c67db16bf544f63cee0074744fa8648a9f56864fcf6f0d

data/jekyll-url-metadata.gemspec

@@ -24,5 +24,6 @@ Gem::Specification.new do |spec|
   # Dependencies
   spec.add_runtime_dependency "jekyll", ">= 3.0.0"
   spec.add_runtime_dependency "nokogiri", ">= 1.10.0"
-
+  spec.add_development_dependency "rake", "~> 13.0"
+  spec.add_development_dependency "minitest", "~> 5.14"
 end

data/lib/jekyll-url-metadata/main.rb

@@ -5,8 +5,12 @@ module Jekyll
   module URLMetadata
 
     def url_config
-      @@url_config ||= @context.registers[:site].config['url_metadata'].nil? ? {} : 
-        @context.registers[:site].config['url_metadata']
+      begin
+        @@url_config ||= @context.registers[:site].config['url_metadata'].nil? ? {} : 
+          @context.registers[:site].config['url_metadata']
+      rescue
+        {}
+      end
     end
 
     def cache
@@ -14,27 +18,7 @@ module Jekyll
     end
 
     def metadata(input)
-      if !input.is_a?(String)
-        log("Expected input type \"String\". Got \"#{input.class}\".")
-        return
-      end
-
-      if input.nil? || input == ""
-        log("Empty input string.")
-        return
-      end
-
-      if !url_config["open_timeout"].nil? && !url_config["open_timeout"].is_a?(Integer)
-        log("Expected an \"Integer\" value for config \"open_timeout\". Got #{url_config["open_timeout"]}.")
-
-        return
-      end
-
-      if !url_config["read_timeout"].nil? && !url_config["read_timeout"].is_a?(Integer)
-        log("Expected an \"Integer\" value for config \"read_timeout\". Got #{url_config["read_timeout"]}.")
-
-        return
-      end
+      return if !is_input_valid(input) || !is_config_valid()
 
       # The getset() API implements a cache first strategy
       cache.getset(input) do
@@ -42,23 +26,23 @@ module Jekyll
       end
     end
 
-    def generate_hashmap(input)
+    def generate_hashmap(input, testCase = false)
       # parse HTML from URL
-      doc = Nokogiri::HTML(URI.open(input, { 
-        :open_timeout => url_config["open_timeout"].nil? ? 1 : url_config["open_timeout"],
-        :read_timeout => url_config["read_timeout"].nil? ? 1 : url_config["read_timeout"]
-      }))
-
-      if !doc
-        log("Failed to parse HTML from #{input}. Please double check for URL validity.")
+      begin
+        doc = Nokogiri::HTML(URI.open(input, { 
+          :open_timeout => url_config["open_timeout"].nil? ? 1 : url_config["open_timeout"],
+          :read_timeout => url_config["read_timeout"].nil? ? 1 : url_config["read_timeout"]
+        }))
+      rescue
+        log("Failed to parse HTML from '#{input}'. Please double check for URL validity.") if !testCase
         return
       end
 
       hash = Hash.new
 
-      # add <title> tag's value to the hash
-      doc.search("title").each do | title |
-        next if title.content == ""
+      # add first <title> tag's value to the hash
+      doc.search("head title").each do | title |
+        break if exists(hash["title"])
         hash["title"] = title.content
       end
 
@@ -86,8 +70,46 @@ module Jekyll
       hash
     end
 
-    def log(msg)
-      Jekyll.logger.error "URL Metadata:", msg
+    def is_input_valid(input, testCase = false)
+      if !input.is_a?(String)
+        log("Expected input type 'String'. Got '#{input.class}'.") if !testCase
+
+        return false
+      end
+
+      if input.nil? || input == ""
+        log("Empty input string.") if !testCase
+
+        return false
+      end
+
+      begin
+        if URI.parse(input).kind_of?(URI::HTTP)
+          return true
+        else
+          log("'#{input}' does not seem to be a valid URL.") if !testCase
+        end
+      rescue
+        log("'#{input}' does not seem to be a valid URL.") if !testCase
+      end
+
+      false
+    end
+
+    def is_config_valid()
+      if !url_config["open_timeout"].nil? && !url_config["open_timeout"].is_a?(Integer)
+        log("Expected an 'Integer' value for config 'open_timeout'. Got '#{url_config["open_timeout"].class}'.")
+
+        return false
+      end
+
+      if !url_config["read_timeout"].nil? && !url_config["read_timeout"].is_a?(Integer)
+        log("Expected an 'Integer' value for config 'read_timeout'. Got '#{url_config["read_timeout"].class}'.")
+
+        return false
+      end
+
+      true
     end
 
     def exists(obj)
@@ -98,6 +120,10 @@ module Jekyll
       obj.get_attribute(attr)
     end
 
+    def log(msg)
+      Jekyll.logger.error "URL Metadata:", msg
+    end
+
   end # module Jekyll
 end # module URLMetadata
 

data/lib/jekyll-url-metadata/version.rb

@@ -1,5 +1,5 @@
 module Jekyll
   module URLMetadata
-    VERSION = "1.0.1"
+    VERSION = "1.0.3"
   end # module Jekyll
 end # module URLMetadata

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: jekyll-url-metadata
 version: !ruby/object:Gem::Version
-  version: 1.0.1
+  version: 1.0.3
 platform: ruby
 authors:
 - Gourav Khunger
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2022-07-21 00:00:00.000000000 Z
+date: 2022-10-07 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: jekyll
@@ -38,6 +38,34 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: 1.10.0
+- !ruby/object:Gem::Dependency
+  name: rake
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '13.0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '13.0'
+- !ruby/object:Gem::Dependency
+  name: minitest
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '5.14'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '5.14'
 description: A plugin to expose meta data information from the head tag of a webapge
   to liquid variables just from its url string.
 email: