string_utils 1.0.0

data/.gitignore

@@ -0,0 +1,3 @@
+pkg/*
+*.gem
+.bundle

data/Gemfile

@@ -0,0 +1,4 @@
+source "http://rubygems.org"
+
+# Specify your gem's dependencies in string_utils.gemspec
+gemspec

data/Gemfile.lock

@@ -0,0 +1,30 @@
+PATH
+  remote: .
+  specs:
+    string_utils (1.0.0)
+      activesupport (> 2)
+      i18n
+
+GEM
+  remote: http://rubygems.org/
+  specs:
+    activesupport (3.0.3)
+    diff-lcs (1.1.2)
+    i18n (0.5.0)
+    rspec (2.2.0)
+      rspec-core (~> 2.2)
+      rspec-expectations (~> 2.2)
+      rspec-mocks (~> 2.2)
+    rspec-core (2.2.1)
+    rspec-expectations (2.2.0)
+      diff-lcs (~> 1.1.2)
+    rspec-mocks (2.2.0)
+
+PLATFORMS
+  ruby
+
+DEPENDENCIES
+  activesupport (> 2)
+  i18n
+  rspec (>= 2.0.0)
+  string_utils!

data/README

@@ -0,0 +1,8 @@
+StringUtils gem repo
+
+StringUtils is a library that allows various string manipulation
+
+Example usage:
+
+ * StringUtils.truncate("hello world", 10, "...") #=> "hello..."
+ * StringUtils.normalize_name "\302\240  Gran Via/Avda.de Asturias " #=> :Gran Via / Avda. de Asturias"

data/Rakefile

@@ -0,0 +1,9 @@
+require 'bundler'
+require 'rspec/core/rake_task'
+
+RSpec::Core::RakeTask.new do |t|
+  t.rspec_opts = ["-c", "-f progress"]
+  t.pattern = 'test/**/*_spec.rb'
+end
+
+Bundler::GemHelper.install_tasks

data/lib/string_utils.rb

@@ -0,0 +1,134 @@
+# -*- coding: utf-8 -*-
+if RUBY_VERSION < '1.9'
+  $KCODE = "UTF8" if $KCODE == "NONE"
+  require 'active_support/core_ext/string/multibyte'
+end
+
+
+# StringUtils is a library that allows various string manipulation
+# Example usage:
+#   * StringUtils.truncate("hello world", 10, "...") #=> "hello..."
+#   * StringUtils.normalize_name "\302\240  Gran Via/Avda.de Asturias " #=> :Gran Via / Avda. de Asturias"
+module StringUtils
+  extend self
+
+  NBSP               = "\302\240"
+  WHITESPACE_MATCHER = "(?:\s|#{NBSP})"
+  WHITESPACE         = /#{WHITESPACE_MATCHER}/
+  NOT_WHITESPACE     = "[^\s#{NBSP}]"
+  WHITESPACES        = /#{WHITESPACE_MATCHER}+/
+
+
+  # Normalizes whitespace
+  # "a , a" => "a, a"
+  # "a ,a"  => "a, a"
+  # "a,a"   => "a, a"
+  # "a/b"   => "a / b", "a/ b" => "a / b", "a /b" => "a / b"
+  # Removes trailing and leading [.,]
+  # options: {:titleize => true (default false)}
+  def normalize_name(value, options = {})
+    value = mb_charify(value)
+
+    # Normalize whitespace
+    value.gsub!("\n", ' ')
+    value.gsub!(WHITESPACES, ' ')
+    value.strip!
+
+    # Remove trailing and leading .,
+    value.gsub!(/^[.,]/, '')
+    value.gsub!(/[.,]$/, '')
+
+    # "a ,a"  => "a, a"
+    # "a,a"   => "a, a"
+    # "a , a" => "a, a"
+    value.gsub!(/#{WHITESPACE_MATCHER}([,.])/, '\1')
+    value.gsub!(/([,.])(#{NOT_WHITESPACE})/, '\1 \2')
+
+    # "//" => "/"
+    value.gsub!(/\/+/, '/')
+
+    # "a/b" => "a / b", "a/ b" => "a / b", "a /b" => "a / b"
+    value.gsub!(/(#{NOT_WHITESPACE})\//, '\1 /')
+    value.gsub!(/\/(#{NOT_WHITESPACE})/, '/ \1')
+
+    if options[:titleize]
+      value = value.titleize
+      value.gsub!(/#{WHITESPACE_MATCHER}(Of|And|A)#{WHITESPACE_MATCHER}/) { |m| " #{m.downcase} " }
+    end
+    value.to_s
+  end
+
+  # Truncates the string
+  # The result will be +:length+ or shorter, and the words will not be cut in the middle
+  # Arguments:
+  # :length => Integer (default: 30)
+  # :omission => String (default: "...")
+  def truncate(text, *args)
+    options = args.last.is_a?(Hash) ? args.pop : {}
+
+    # support either old or Rails 2.2 calling convention:
+    unless args.empty?
+      options[:length]   = args[0] || 30
+      options[:omission] = args[1] || "…"
+    end
+
+    options          = {:length => 30, :omission => "…"}.merge(options)
+    options[:length] = options[:length].to_i
+
+    return "" if !text
+    chars = mb_charify(text)
+
+
+    # If we can return it straight away or rstrip it and return it, we do it here
+    if chars.length <= options[:length]
+      return text
+    elsif (chars = rstrip_with_nbsp(chars)).length <= options[:length]
+      return chars.to_s
+    end
+
+    omission           = mb_charify(options[:omission])
+
+    # Here we know we have to remove at least 1 word
+    # 1. Get the first l characters
+    # 2. Remove the last word if it's a part
+    # 3. Add omission
+    length_wo_omission = options[:length] - omission.length
+
+    return '' if length_wo_omission < 0
+
+    result = rstrip_with_nbsp(chars[0...length_wo_omission] || "")
+
+    # Remove the last word unless we happened to trim it exactly already
+    unless chars[length_wo_omission] =~ WHITESPACE || result.length < length_wo_omission
+      len    = result.split(WHITESPACES).last
+      len    &&= len.length
+      result = rstrip_with_nbsp(result[0...(result.length - (len || 0))])
+    end
+
+    result += options[:omission]
+    result.to_s
+  end
+
+  # Returns a unicode compatible version of the string
+  #
+  # support any of:
+  #  * ruby 1.9 sane utf8 support
+  #  * rails 2.1 workaround for crappy ruby 1.8 utf8 support
+  #  * rails 2.2 workaround for crappy ruby 1.8 utf8 support
+  # hooray!
+  def mb_charify(text)
+    if RUBY_VERSION >= '1.9'
+      text
+    elsif text.respond_to?(:mb_chars)
+      text.mb_chars
+    else
+      raise "StringUtils: No unicode support for strings"
+    end
+  end
+
+  private
+
+  def rstrip_with_nbsp(s)
+    s.gsub(/#{WHITESPACE_MATCHER}+\z/, '')
+  end
+end

data/lib/string_utils/version.rb

@@ -0,0 +1,3 @@
+module StringUtils
+  VERSION = "1.0.0"
+end

data/string_utils.gemspec

@@ -0,0 +1,34 @@
+# -*- encoding: utf-8 -*-
+$:.push File.expand_path("../lib", __FILE__)
+require "string_utils/version"
+
+Gem::Specification.new do |s|
+  s.name              = "string_utils"
+  s.version           = StringUtils::VERSION
+  s.platform          = Gem::Platform::RUBY
+  s.authors           = ["Gleb Mazovetskiy"]
+  s.email             = ["glex.spb@gmail.com"]
+  s.homepage          = "http://github.com/glebm"
+  s.summary           = %q{Provides useful string utils like "truncate to word"}
+  s.description       = %q{Provides useful string utils like "truncate to word".
+Compatible with ruby >= 1.8. Benefits from active_support if available on ruby < 1.9.
+
+Tested with:
+
+  * ruby 1.8.7 (2010-04-19 patchlevel 253) [i686-linux], MBARI 0x8770, Ruby Enterprise Edition 2010.02
+  * 
+}
+
+  s.rubyforge_project = "string_utils"
+
+  
+  s.add_dependency "activesupport", "> 2"
+  s.add_dependency "i18n"
+
+  s.add_development_dependency "rspec", ">= 2.0.0"
+
+  s.files             = `git ls-files`.split("\n")
+  s.test_files        = `git ls-files -- {test,spec,features}/*`.split("\n")
+  s.executables       = `git ls-files -- bin/*`.split("\n").map { |f| File.basename(f) }
+  s.require_paths     = ["lib"]
+end

data/test/normalize_name_spec.rb

@@ -0,0 +1,19 @@
+# -*- coding: utf-8 -*-
+require "string_utils"
+
+describe "StringUtils" do
+  describe "#normalize_name" do
+
+    it 'normalizes whitespace' do
+      StringUtils.normalize_name("\302\240   Hello \302\240  World!   ").should == "Hello World!"
+    end
+
+    it 'normalizes periods, commas and /' do
+      StringUtils.normalize_name("Av . Valle , Tetuan.Aqui ,a").should == "Av. Valle, Tetuan. Aqui, a"
+    end
+
+    it 'removes trailing and leading [,.]' do
+      StringUtils.normalize_name(".Here,").should == "Here"
+    end
+  end
+end

data/test/truncate_spec.rb

@@ -0,0 +1,50 @@
+# -*- coding: utf-8 -*-
+require 'string_utils'
+
+describe "StringUtils" do
+  describe "#truncate" do
+
+    it "does not cut words shorter than the truncate length" do
+      tests = ["hello", "hello there this is a test nothing more"]
+      tests.each do |test|
+        StringUtils.truncate(test, test.length).should == test
+      end
+    end
+
+    it "truncates even if the only word is longer than the truncate length" do
+      StringUtils.truncate("hello", 3, "...").should == "..."
+    end
+
+    it "never goes above the length provided" do
+      (0..2).each do |length|
+        StringUtils.truncate("ɦɛĺłø ŵőřļđ", length, "...").should == ""
+      end
+
+      (3..7).each do |length|
+        StringUtils.truncate("ɦɛĺłø ŵőřļđ", length, "...").should == "..."
+      end
+
+      (8..10).each do |length|
+        StringUtils.truncate("ɦɛĺłø ŵőřļđ", length, "...").should == "ɦɛĺłø..."
+      end
+
+      (11..15).each do |length|
+        StringUtils.truncate("ɦɛĺłø ŵőřļđ", length, "...").should == "ɦɛĺłø ŵőřļđ"
+      end
+    end
+
+    it "keeps the whole word" do
+      StringUtils.truncate("ab cd", 2, '.').should == "."
+    end
+
+    it "does not cut character entities" do
+      StringUtils.truncate("ä Ä", 9, '..').should == "ä.."
+    end
+
+    it "handles empty strings" do
+      StringUtils.truncate("", 0, "...").should == ""
+      StringUtils.truncate("", 1, "...").should == ""
+      StringUtils.truncate("", 30, "...").should == ""
+    end
+  end
+end

metadata

@@ -0,0 +1,128 @@
+--- !ruby/object:Gem::Specification 
+name: string_utils
+version: !ruby/object:Gem::Version 
+  hash: 23
+  prerelease: false
+  segments: 
+  - 1
+  - 0
+  - 0
+  version: 1.0.0
+platform: ruby
+authors: 
+- Gleb Mazovetskiy
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2010-12-07 00:00:00 +01:00
+default_executable: 
+dependencies: 
+- !ruby/object:Gem::Dependency 
+  name: activesupport
+  prerelease: false
+  requirement: &id001 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">"
+      - !ruby/object:Gem::Version 
+        hash: 7
+        segments: 
+        - 2
+        version: "2"
+  type: :runtime
+  version_requirements: *id001
+- !ruby/object:Gem::Dependency 
+  name: i18n
+  prerelease: false
+  requirement: &id002 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        hash: 3
+        segments: 
+        - 0
+        version: "0"
+  type: :runtime
+  version_requirements: *id002
+- !ruby/object:Gem::Dependency 
+  name: rspec
+  prerelease: false
+  requirement: &id003 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        hash: 15
+        segments: 
+        - 2
+        - 0
+        - 0
+        version: 2.0.0
+  type: :development
+  version_requirements: *id003
+description: |
+  Provides useful string utils like "truncate to word".
+  Compatible with ruby >= 1.8. Benefits from active_support if available on ruby < 1.9.
+  
+  Tested with:
+  
+    * ruby 1.8.7 (2010-04-19 patchlevel 253) [i686-linux], MBARI 0x8770, Ruby Enterprise Edition 2010.02
+    * 
+
+email: 
+- glex.spb@gmail.com
+executables: []
+
+extensions: []
+
+extra_rdoc_files: []
+
+files: 
+- .gitignore
+- Gemfile
+- Gemfile.lock
+- README
+- Rakefile
+- lib/string_utils.rb
+- lib/string_utils/version.rb
+- string_utils.gemspec
+- test/normalize_name_spec.rb
+- test/truncate_spec.rb
+has_rdoc: true
+homepage: http://github.com/glebm
+licenses: []
+
+post_install_message: 
+rdoc_options: []
+
+require_paths: 
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      hash: 3
+      segments: 
+      - 0
+      version: "0"
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      hash: 3
+      segments: 
+      - 0
+      version: "0"
+requirements: []
+
+rubyforge_project: string_utils
+rubygems_version: 1.3.7
+signing_key: 
+specification_version: 3
+summary: Provides useful string utils like "truncate to word"
+test_files: []
+