librariesio-url-parser 1.0.1 → 1.0.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: df16d2958656fa866177551cb6c6275e6b8303c40d41d3baeb639aff567aad07
-  data.tar.gz: ec4443d5e407716d312746c7157d698a21b9c10eaeea208f42101b42688f4b47
+  metadata.gz: 39533f7a7c92f1b4bcccd5a9a46b3c2ae5d14f04da87b0d643339be188184dfc
+  data.tar.gz: bcf4fd5128de3343a11f46b25fadf2cf0f0daf479a9f37824f2d0ac93534753b
 SHA512:
-  metadata.gz: 8ca1db0935515b6fe9d0f2dc5c27e8af36bb8cd4b652dc70d4f5194a27edd1cd4257144bd83b4b9c7fb4ae6dc333a37789021e852a1bc34cdf4000a61a27b86f
-  data.tar.gz: f253083f04c93d8eff970002d9bcc02ce0ffb362167149371a6a5576dd4ed1ee9dbe2c760841695873487749cab40d121447fdf48f74d429b46be930322db683
+  metadata.gz: 27401dad229d48fd32ec999b91ddf171a708b847fdbfbbb21b283cf6b72e1f8280731ccd2f51a012bdfcf7d73f9087a7c00ab8b8db3109b55913a434be70dba1
+  data.tar.gz: ed74167bc7f4b712209ecf9e0fddbafe7c5c9f77dcda9fcfb883b658d8e99ee8283d89a70f4ddecd2fd3a62912c5cb04c72358804dbedb44932e0adafae75594

data/Gemfile

@@ -3,5 +3,3 @@ ruby "2.6.5"
 
 # Specify your gem's dependencies in librariesio-url-parser.gemspec
 gemspec
-
-gem "pry", "~> 0.14.1", :group => :development

data/Gemfile.lock

@@ -1,17 +1,12 @@
 PATH
   remote: .
   specs:
-    librariesio-url-parser (1.0.1)
+    librariesio-url-parser (1.0.2)
 
 GEM
   remote: https://rubygems.org/
   specs:
-    coderay (1.1.3)
     diff-lcs (1.5.0)
-    method_source (1.0.0)
-    pry (0.14.1)
-      coderay (~> 1.1)
-      method_source (~> 1.0)
     rake (12.3.3)
     rspec (3.11.0)
       rspec-core (~> 3.11.0)
@@ -34,7 +29,6 @@ PLATFORMS
 
 DEPENDENCIES
   librariesio-url-parser!
-  pry (~> 0.14.1)
   rake (~> 12.0)
   rspec (~> 3.0)
   rspec_junit_formatter (~> 0.5)

data/README.md

@@ -45,6 +45,7 @@ URLParser.try_all("git@bitbucket.org:tildeslash/monit.git") #=> "https://bitbuck
 - GitHub
 - GitLab
 - Bitbucket
+- Apache SVN
 
 ## Development
 

data/lib/apache_svn_url_parser.rb

@@ -0,0 +1,66 @@
+# frozen_string_literal: true
+class ApacheSvnUrlParser < URLParser
+  SUBDIR_NAMES = %w[trunk tags branches].freeze
+  private
+
+  def full_domain
+    'https://svn.apache.org/viewvc'
+  end
+
+  def tlds
+    %w(org)
+  end
+
+  def domain
+    'svn.apache'
+  end
+
+  def domain_regex
+    # match only the viewvc endpoint at the domain
+    "#{domain.split("/").first}\.(#{tlds.join('|')})\/viewvc"
+  end
+
+  def remove_domain
+    # find the matches for any github domain characters in the url string
+    # and replace only the first match incase we find a repo with something like github.com as the name
+    url.sub!(/(apache\.org\/(viewvc|repos))+?(:|\/)?/i, '')
+  end
+
+  def extractable_early?
+    false
+  end
+
+  def remove_extra_segments
+    # split the url by / and remove any empty sections
+    self.url = url.split('/').reject{ |s| s.strip.empty? }
+
+    # check to see if any repository subdirectories are included in the segments
+    # this parser is parsing SVN projects, so any common folders used for branching should trip this
+    # truncate the array of segments to stop once we hit a top level sub directory typically seen in SVN repos
+    # and return everything up to that point
+    #
+    # for example apache.org/viewvnc/myproject/subproject/tags/my-1.0.0-release should stop at myproject/subproject
+    # since the tags are just part of that repository
+    subdir_index = url.index{ |s| SUBDIR_NAMES.include?(s) }
+
+    # it looks like the maven/pom directory on the Apache SVN server has a bunch of repositories stored under tags
+    # in this special case, grab the directory name under the subdirectory
+    # it looks like this is most likely to be the first directory under tags/
+    in_maven_pom_dir = url[0..1].join("/").downcase == "maven/pom"
+
+    if in_maven_pom_dir
+      self.url = url[0..subdir_index+1] if subdir_index
+    else
+      self.url = url[0..subdir_index-1] if subdir_index
+    end
+  end
+
+  def format_url
+    # if this is an Array then the url has gone through all the clean up steps
+    #
+    # if this is just a string then the url was not cleaned up and I have no idea how to format it
+    return nil unless url.is_a?(Array) && url.length.positive?
+
+    url.join("/")
+  end
+end

data/lib/librariesio-url-parser.rb

@@ -4,7 +4,8 @@ require_relative "url_parser"
 require_relative "bitbucket_url_parser"
 require_relative "github_url_parser"
 require_relative "gitlab_url_parser"
+require_relative "apache_svn_url_parser"
 
 module LibrariesioURLParser
-  VERSION = "1.0.1"
+  VERSION = "1.0.2"
 end

data/lib/url_parser.rb

@@ -30,12 +30,13 @@ class URLParser
   def self.try_all(url)
     GithubURLParser.parse_to_full_url(url) ||
     GitlabURLParser.parse_to_full_url(url) ||
-    BitbucketURLParser.parse_to_full_url(url)
+    BitbucketURLParser.parse_to_full_url(url) ||
+    ApacheSvnUrlParser.parse_to_full_url(url)
   end
 
   def parse_to_full_url
     path = parse
-    return nil unless path.present?
+    return nil if path.nil? || path.empty?
     [full_domain, path].join('/')
   end
 
@@ -151,11 +152,11 @@ class URLParser
   end
 
   def remove_scheme
-    url.gsub!(/(((git\+https|git|ssh|hg|svn|scm|http|https)+?:)+?)/i, '')
+    url.gsub!(/(((git\+https|git|ssh|hg|svn|scm|http|https)+?:)(\/\/)?)/i, '')
   end
 
   def remove_subdomain
-    url.gsub!(/(www|ssh|raw|git|wiki)+?\./i, '')
+    url.gsub!(/(www|ssh|raw|git|wiki|svn)+?\./i, '')
   end
 
   def remove_whitespace

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: librariesio-url-parser
 version: !ruby/object:Gem::Version
-  version: 1.0.1
+  version: 1.0.2
 platform: ruby
 authors:
 - Matt Pace
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2022-07-26 00:00:00.000000000 Z
+date: 2022-07-27 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rake
@@ -65,6 +65,7 @@ files:
 - Gemfile.lock
 - README.md
 - Rakefile
+- lib/apache_svn_url_parser.rb
 - lib/bitbucket_url_parser.rb
 - lib/github_url_parser.rb
 - lib/gitlab_url_parser.rb