sanscript 0.6.2 → 0.7.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 59fade123e6369dd8e65f40aef416b1f3ceb6246
-  data.tar.gz: 4c010e1ca0999f576a8485d385e825fd3cd98a3e
+  metadata.gz: 78a9a4f83c7ebeabdf0a4ce0b02c0efbe2bb99bd
+  data.tar.gz: 1e16e165ab8587163be05d449301c182d70b8037
 SHA512:
-  metadata.gz: 9d2fc3ef1c990703a8f4c5a2e69f6eb535dca84600b375cd030cbc8e1bb30baa2447c7de4d0634afb3303389dd24b5e3a144bce9ac6f8b883768473d8b8d5cc9
-  data.tar.gz: 98a910e4a950ad13ae0501cb3eb9ca841a4f049a42ce46989e6f12ff7d2ad1105a5083b4ae4496878bd6f001b9f29fa1a7323833c42c7234cc23193a6edc82b5
+  metadata.gz: 0354c6bb4d13b5aca33b73febb968f72dfaa843a7c6231fc30482f66d04bd8d239250a2aa0f0dbc957e4fef35d46173800319cd8583e8433cb9a1127e192d519
+  data.tar.gz: f0131f094b3a4e88cb7bba19955c27f1337b4ea7f2d1ca419e5bb8cd61f2546b515f52c2f60b44df1bdc91494f5cd0cb3e9f7a52b4c39b680039db7a23ba6bfb

data/.gitignore

@@ -10,3 +10,4 @@
 /lib/librusty_sanscript.*
 /target
 mkmf.log
+Cargo.lock

data/.travis.yml

@@ -3,12 +3,12 @@ language: ruby
 rvm:
   - 2.3.1
   - 2.2
-  - 2.4.0-preview1
+  - 2.4.0-preview2
 matrix:
   allow_failures:
-  - rvm: 2.4.0-preview1
+  - rvm: 2.4.0-preview2
 
-before_install: gem install bundler -v 1.12.5
+before_install: gem install bundler -v 1.13.1
 script: bundle exec rake spec SPEC_OPTS="--format p"
 addons:
   code_climate:

data/Cargo.toml

@@ -1,16 +1,13 @@
 [package]
 name = "rusty_sanscript"
-version = "0.2.0"
+version = "0.3.0"
 authors = ["Tim Bellefleur <nomoon@phoebus.ca>"]
 publish = false
 
 [dependencies]
 lazy_static = "^0.2.1"
-ruru = "^0.7.8"
-unicode-normalization = "^0.1.2"
-regex = "^0.1.73"
-
-[package.metadata.thermite]
+ruby-sys = "^0.2.13"
+regex = "^0.1.77"
 
 [lib]
-crate-type = ["dylib"]
+crate-type = ["cdylib"]

data/Rakefile

@@ -1,25 +1,9 @@
 # frozen_string_literal: true
 require "bundler/gem_tasks"
 
-# Add enhanced optional_build task into Thermite
 require "thermite/tasks"
-module Thermite
-  class BetterTasks < Tasks
-    def initialize(options = {})
-      super
-      desc "Run thermite:build task or download binaries, but skip without fail if unavailable."
-      task "thermite:optional_build" do
-        if cargo
-          Rake::Task["thermite:build"].invoke
-        elsif !download_binary
-          puts "Rust and downloadable binaries are not available, skipping."
-        end
-      end
-    end
-  end
-end
-Thermite::BetterTasks.new
-task default: :"thermite:optional_build"
+Thermite::Tasks.new(optional_rust_extension: true)
+task default: :"thermite:build"
 
 # Ensure missing RSpec development dependency doesn't kill gem install.
 begin

data/lib/sanscript.rb

@@ -2,6 +2,7 @@
 require "ragabash"
 
 require "sanscript/version"
+require "sanscript/rust"
 require "sanscript/exceptions"
 require "sanscript/detect"
 require "sanscript/transliterate"
@@ -9,6 +10,14 @@ require "sanscript/benchmark"
 
 # Sanscript.rb detection/transliteration module for Sanskrit.
 module Sanscript
+  # :nocov:
+  if RUST_AVAILABLE && ENV["SANSCRIPT_NO_RUST"].nil?
+    rust_enable!
+  else
+    rust_disable!
+  end
+  # :nocov:
+
   module_function
 
   # Attempts to detect the encoding scheme of the provided string.
@@ -56,29 +65,4 @@ module Sanscript
     end
     Transliterate.transliterate(text, from, to, opts)
   end
-
-  # Override
-  # :nocov:
-  begin
-    require "fiddle"
-    require "thermite/config"
-
-    toplevel_dir = File.dirname(File.dirname(__FILE__))
-    config = Thermite::Config.new(cargo_project_path: toplevel_dir, ruby_project_path: toplevel_dir)
-    library = Fiddle.dlopen(config.ruby_extension_path)
-    module ::RustySanscriptDetect; end # rubocop:disable Style/ClassAndModuleChildren
-    func = Fiddle::Function.new(library["init_rusty_sanscript"],
-                                [], Fiddle::TYPE_VOIDP)
-    func.call
-    module Detect
-      extend ::RustySanscriptDetect
-      class << self
-        alias detect_scheme rust_detect_scheme
-      end
-    end
-    RUST_ENABLED = true
-  rescue Fiddle::DLError
-    RUST_ENABLED = false
-  end
-  # :nocov:
 end

data/lib/sanscript/benchmark.rb

@@ -1,6 +1,6 @@
 # frozen_string_literal: true
-#:nocov:
 
+# :nocov:
 begin
   require "benchmark/ips"
 rescue LoadError
@@ -10,6 +10,7 @@ rescue LoadError
     end
   end
 end
+# :nocov:
 
 module Sanscript
   # Benchmark/testing module.
@@ -17,7 +18,7 @@ module Sanscript
     module_function
 
     # Runs benchmark-ips test on detection methods.
-    def detect!
+    def detect!(time = 2, warmup = 1)
       deva_string = "नानाशास्त्रसुभाषितामृतरसैः श्रोत्रोत्सवं कुर्वतां येषां यान्ति दिनानि पण्डितजनव्यायामखिन्नात्मनाम् तेषां जन्म च जीवितं च सुकृतं तैर् एव भूर् भूषिता शेषैह् किं पशुवद् विवेकरहितैर् भूभारभूतैर् नरः"
       malayalam_string = "നാനാശാസ്ത്രസുഭാഷിതാമൃതരസൈഃ ശ്രോത്രോത്സവം കുര്വതാം യേഷാം യാന്തി ദിനാനി പണ്ഡിതജനവ്യായാമഖിന്നാത്മനാമ് തേഷാം ജന്മ ച ജീവിതം ച സുകൃതം തൈര് ഏവ ഭൂര് ഭൂഷിതാ ശേഷൈഹ് കിം പശുവദ് വിവേകരഹിതൈര് ഭൂഭാരഭൂതൈര് നരഃ"
       iast_string = "nānāśāstrasubhāṣitāmṛtarasaiḥ śrotrotsavaṃ kurvatāṃ yeṣāṃ yānti dināni paṇḍitajanavyāyāmakhinnātmanām teṣāṃ janma ca jīvitaṃ ca sukṛtaṃ tair eva bhūr bhūṣitā śeṣaih kiṃ paśuvad vivekarahitair bhūbhārabhūtair naraḥ"
@@ -25,21 +26,21 @@ module Sanscript
       hk_string = "nAnAzAstrasubhASitAmRtarasaiH zrotrotsavaM kurvatAM yeSAM yAnti dinAni paNDitajanavyAyAmakhinnAtmanAm teSAM janma ca jIvitaM ca sukRtaM tair eva bhUr bhUSitA zeSaih kiM pazuvad vivekarahitair bhUbhArabhUtair naraH"
 
       ::Benchmark.ips do |x|
-        x.config(time: 5, warmup: 1)
+        x.config(time: time, warmup: warmup)
         x.report("Detect Devanagari") do
-          raise unless Sanscript::Detect.detect_scheme(deva_string) == :devanagari
+          Sanscript::Detect.detect_scheme(deva_string)
         end
         x.report("Detect Malayalam") do
-          raise unless Sanscript::Detect.detect_scheme(malayalam_string) == :malayalam
+          Sanscript::Detect.detect_scheme(malayalam_string)
         end
         x.report("Detect IAST") do
-          raise unless Sanscript::Detect.detect_scheme(iast_string) == :iast
+          Sanscript::Detect.detect_scheme(iast_string)
         end
         x.report("Detect SLP1") do
-          raise unless Sanscript::Detect.detect_scheme(slp1_string) == :slp1
+          Sanscript::Detect.detect_scheme(slp1_string)
         end
         x.report("Detect HK") do
-          raise unless Sanscript::Detect.detect_scheme(hk_string) == :hk
+          Sanscript::Detect.detect_scheme(hk_string)
         end
         x.compare!
       end
@@ -47,13 +48,13 @@ module Sanscript
     end
 
     # Runs benchmark-ips test on roman-source transliteration methods.
-    def transliterate_roman!
+    def transliterate_roman!(time = 2, warmup = 1)
       iast_string = "nānāśāstrasubhāṣitāmṛtarasaiḥ śrotrotsavaṃ kurvatāṃ yeṣāṃ yānti dināni paṇḍitajanavyāyāmakhinnātmanām teṣāṃ janma ca jīvitaṃ ca sukṛtaṃ tair eva bhūr bhūṣitā śeṣaih kiṃ paśuvad vivekarahitair bhūbhārabhūtair naraḥ"
       slp1_string = "nAnASAstrasuBAzitAmftarasEH SrotrotsavaM kurvatAM yezAM yAnti dinAni paRqitajanavyAyAmaKinnAtmanAm tezAM janma ca jIvitaM ca sukftaM tEr eva BUr BUzitA SezEh kiM paSuvad vivekarahitEr BUBAraBUtEr naraH"
       hk_string = "nAnAzAstrasubhASitAmRtarasaiH zrotrotsavaM kurvatAM yeSAM yAnti dinAni paNDitajanavyAyAmakhinnAtmanAm teSAM janma ca jIvitaM ca sukRtaM tair eva bhUr bhUSitA zeSaih kiM pazuvad vivekarahitair bhUbhArabhUtair naraH"
 
       ::Benchmark.ips do |x|
-        x.config(time: 3, warmup: 2)
+        x.config(time: time, warmup: warmup)
 
         x.report("IAST==>Devanagari") do
           Sanscript.transliterate(iast_string, :iast, :devanagari)
@@ -88,11 +89,11 @@ module Sanscript
     end
 
     # Runs benchmark-ips test on brahmic-source transliteration methods.
-    def transliterate_brahmic!
+    def transliterate_brahmic!(time = 2, warmup = 1)
       deva_string = "नानाशास्त्रसुभाषितामृतरसैः श्रोत्रोत्सवं कुर्वतां येषां यान्ति दिनानि पण्डितजनव्यायामखिन्नात्मनाम् तेषां जन्म च जीवितं च सुकृतं तैर् एव भूर् भूषिता शेषैह् किं पशुवद् विवेकरहितैर् भूभारभूतैर् नरः"
 
       ::Benchmark.ips do |x|
-        x.config(time: 5, warmup: 2)
+        x.config(time: time, warmup: warmup)
         x.report("Devanagari==>IAST") do
           Sanscript.transliterate(deva_string, :devanagari, :iast)
         end

data/lib/sanscript/detect.rb

@@ -53,8 +53,9 @@ module Sanscript
     #   Attempts to detect the encoding scheme of the provided string.
     #
     #   Uses the most efficient implementation for your ruby version
-    #   (either {Ruby2x#ruby_detect_scheme} or {Ruby24#ruby_detect_scheme}) or
-    #   the Rust native extension if available.
+    #   (either {Ruby2x#ruby_detect_scheme} or {Ruby24#ruby_detect_scheme})
+    #   at first, which may be then overriden by the Rust native extension
+    #   (see {Sanscript#rust_enable!} and {Sanscript#rust_disable!}})
     #
     #   @param text [String] a string of Sanskrit text
     #   @return [Symbol, nil] the Symbol of the scheme, or nil if no match
@@ -68,9 +69,13 @@ module Sanscript
       require "sanscript/detect/ruby2x"
       extend Ruby2x
     end
+    # :nocov:
     class << self
       alias detect_scheme ruby_detect_scheme
     end
-    # :nocov:
+
+    if defined?(Rust) && Rust.instance_methods.include?(:rust_detect_scheme)
+      extend Rust
+    end
   end
 end

data/lib/sanscript/rust.rb

@@ -0,0 +1,47 @@
+# frozen_string_literal: true
+module Sanscript
+  begin
+    require "thermite/fiddle"
+
+    Thermite::Fiddle.load_module("init_rusty_sanscript",
+                                 cargo_project_path: GEM_ROOT,
+                                 ruby_project_path: GEM_ROOT)
+    #:nocov:#
+    RUST_AVAILABLE = true
+  rescue Fiddle::DLError
+    RUST_AVAILABLE = false
+    #:nocov:#
+  end
+
+  module_function
+
+  # @return [bool] the enabled status of the Rust extension
+  def rust_enabled?
+    @rust_enabled
+  end
+
+  # Turns on Rust extension, if available.
+  # @return [bool] the enabled status of the Rust extension
+  def rust_enable!
+    if RUST_AVAILABLE
+      Detect.module_eval do
+        class << self
+          alias_method :detect_scheme, :rust_detect_scheme
+        end
+      end
+      @rust_enabled = true
+    end
+    @rust_enabled
+  end
+
+  # Turns off Rust native extension.
+  # @return [bool] the enabled status of the Rust extension
+  def rust_disable!
+    Detect.module_eval do
+      class << self
+        alias_method :detect_scheme, :ruby_detect_scheme
+      end
+    end
+    @rust_enabled = false
+  end
+end

data/lib/sanscript/version.rb

@@ -1,7 +1,7 @@
 # frozen_string_literal: true
 module Sanscript
   # The version number
-  VERSION = "0.6.2"
+  VERSION = "0.7.0"
 
   GEM_ROOT = Pathname.new(File.realpath(File.join(__dir__, "..", "..")))
   private_constant :GEM_ROOT

data/sanscript.gemspec

@@ -22,16 +22,16 @@ Gem::Specification.new do |spec|
 
   spec.required_ruby_version = "~> 2.2"
 
-  spec.add_development_dependency "bundler", "~> 1.12"
-  spec.add_development_dependency "rake", "~> 11.2"
+  spec.add_development_dependency "bundler", "~> 1.13"
+  spec.add_development_dependency "rake", "~> 11.3"
   spec.add_development_dependency "rspec", "~> 3.5"
   spec.add_development_dependency "codeclimate-test-reporter", "~> 0.6"
-  spec.add_development_dependency "rubocop", "~> 0.41"
-  spec.add_development_dependency "rubocop-rspec", "~> 1.5"
+  spec.add_development_dependency "rubocop", "~> 0.43"
+  spec.add_development_dependency "rubocop-rspec", "~> 1.7"
   spec.add_development_dependency "pry", "~> 0.10"
-  spec.add_development_dependency "benchmark-ips", "~> 2.6"
+  spec.add_development_dependency "benchmark-ips", "~> 2.7"
   spec.add_development_dependency "yard", "~> 0.9"
 
   spec.add_runtime_dependency "ragabash", "~> 0.2"
-  spec.add_runtime_dependency "thermite", "~> 0.5"
+  spec.add_runtime_dependency "thermite", "~> 0.7"
 end

data/src/detect/mod.rs

@@ -0,0 +1,96 @@
+pub mod ruby;
+
+use regex::Regex;
+
+//
+// Initialize all of the generic static variables
+//
+lazy_static! {
+  // Match escaped control characters
+  static ref RE_ESCAPED_CONTROL_CHAR: Regex = Regex::new(r"\\(?:\{#|##|#\})").unwrap();
+
+  // Match ##...## or {#...#} control blocks.
+  static ref RE_CONTROL_BLOCK: Regex = Regex::new(r"##.*?##|\{#.*?#\}").unwrap();
+
+  // Match any character in the block of Brahmic scripts
+  // between Devanagari and Malayalam.
+  static ref RE_BRAHMIC_RANGE: Regex = Regex::new(r"[\x{0900}-\x{0d7f}]").unwrap();
+
+  // Match on special Roman characters
+  static ref RE_IAST_OR_KOLKATA_ONLY: Regex = Regex::new(r"(?i)[āīūṛṝḷḹēōṃḥṅñṭḍṇśṣḻ]").unwrap();
+
+  // Match on Kolkata-specific Roman characters
+  static ref RE_KOLKATA_ONLY: Regex = Regex::new(r"(?i)[ēō]").unwrap();
+
+  // Match on ITRANS-only
+  static ref RE_ITRANS_ONLY: Regex = Regex::new(r"ee|oo|\^[iI]|RR[iI]|L[iI]|~N|N\^|Ch|chh|JN|sh|Sh|\.a").unwrap();
+
+  // Match on SLP1-only characters and bigrams
+  static ref RE_SLP1_ONLY: Regex = Regex::new(r"[fFxXEOCYwWqQPB]|kz|N[kg]|tT|dD|S[cn]|[aAiIuUeo]R|G[yr]").unwrap();
+
+  // Match on Velthuis-only characters
+  static ref RE_VELTHUIS_ONLY: Regex = Regex::new(r"\.[mhnrlntds]|\x22n|~s").unwrap();
+
+  // Match on chars shared by ITRANS and Velthuis
+  static ref RE_ITRANS_OR_VELTHUIS_ONLY: Regex = Regex::new(r"aa|ii|uu|~n").unwrap();
+
+  // Match on characters available in Harvard-Kyoto
+  static ref RE_HARVARD_KYOTO: Regex = Regex::new(r"[aAiIuUeoRMHkgGcjJTDNtdnpbmyrlvzSsh]").unwrap();
+}
+
+//
+// The function itself!
+//
+#[no_mangle]
+pub extern fn detect_scheme(s: &str) -> usize {
+  // Clean-up string of control characters.
+  let r_str = &RE_CONTROL_BLOCK.replace_all(
+                &RE_ESCAPED_CONTROL_CHAR.replace_all(s, ""), "");
+
+  // Brahmic schemes are all within a specific range of code points.
+  let brahmic_match = RE_BRAHMIC_RANGE.find(r_str);
+  if brahmic_match != None {
+    let brahmic_match = brahmic_match.unwrap();
+    let brahmic_codepoint = r_str.chars().nth(brahmic_match.0).unwrap() as usize;
+
+    if brahmic_codepoint < 0x0980 {
+      return 1; // Devanagari
+    } else if brahmic_codepoint < 0x0A00 {
+      return 2; // Bengali
+    } else if brahmic_codepoint < 0x0A80 {
+      return 3; // Gurmukhi
+    } else if brahmic_codepoint < 0x0B00 {
+      return 4; // Gujarati
+    } else if brahmic_codepoint < 0x0B80 {
+      return 5; // Oriya
+    } else if brahmic_codepoint < 0x0C00 {
+      return 6; // Tamil
+    } else if brahmic_codepoint < 0x0C80 {
+      return 7; // Telugu
+    } else if brahmic_codepoint < 0x0D00 {
+      return 8; // Kannada
+    } else {
+      return 9; // Malayalam
+    }
+  }
+
+  // Romanizations
+  if RE_IAST_OR_KOLKATA_ONLY.is_match(r_str) {
+    if RE_KOLKATA_ONLY.is_match(r_str) {
+      return 11; // Kolkata
+    } else {
+      return 10; // IAST
+    }
+  } else if RE_ITRANS_ONLY.is_match(r_str) {
+    return 12; // ITRANS
+  } else if RE_SLP1_ONLY.is_match(r_str) {
+    return 13; // SLP1
+  } else if RE_VELTHUIS_ONLY.is_match(r_str) {
+    return 14; // Velthuis
+  } else if RE_ITRANS_OR_VELTHUIS_ONLY.is_match(r_str) {
+    return 12; // ITRANS
+  } else if RE_HARVARD_KYOTO.is_match(r_str) {
+    return 15; // HK
+  }
+  return 0; // Unknown
+}

data/src/detect/ruby.rs

@@ -0,0 +1,42 @@
+use detect;
+
+use rb;
+use rb::{CallbackPtr, Value};
+
+// Initialize all of the Ruby-specific static variables.
+lazy_static! {
+  // Lookup table for Ruby Results
+  static ref RUBY_RESULTS: [Value; 16] = [
+    rb::RB_NIL,                    // 0
+    rb::str_to_sym("devanagari"),  // 1
+    rb::str_to_sym("bengali") ,    // 2
+    rb::str_to_sym("gurmukhi"),    // 3
+    rb::str_to_sym("gujarati"),    // 4
+    rb::str_to_sym("oriya"),       // 5
+    rb::str_to_sym("tamil"),       // 6
+    rb::str_to_sym("telugu"),      // 7
+    rb::str_to_sym("kannada"),     // 8
+    rb::str_to_sym("malayalam"),   // 9
+    rb::str_to_sym("iast"),        // 10
+    rb::str_to_sym("kolkata"),     // 11
+    rb::str_to_sym("itrans"),      // 12
+    rb::str_to_sym("slp1"),        // 13
+    rb::str_to_sym("velthuis"),    // 14
+    rb::str_to_sym("hk")           // 15
+  ];
+}
+
+fn rbstr_detect_scheme(_rself: Value, s: Value) -> Value {
+  let r_str = rb::rbstr_to_str(&s);
+  let result = detect::detect_scheme(r_str);
+  return RUBY_RESULTS[result];
+}
+
+#[no_mangle]
+pub extern fn init_rusty_sanscript() {
+  let m_sanscript = rb::define_module("Sanscript");
+  let m_detect = rb::define_module_under(&m_sanscript, "Detect");
+  let m_rust = rb::define_module_under(&m_detect, "Rust");
+  rb::define_method(&m_rust, "rust_detect_scheme",
+    rbstr_detect_scheme as CallbackPtr, 1);
+}

data/src/lib.rs

@@ -1,122 +1,7 @@
 #[macro_use] extern crate lazy_static;
-#[macro_use] extern crate ruru;
-extern crate unicode_normalization;
+extern crate ruby_sys;
 extern crate regex;
 
-use ruru::{AnyObject, Class, NilClass, Symbol, RString};
-use ruru::traits::Object;
-#[allow(unused_imports)] use unicode_normalization::UnicodeNormalization;
-use regex::Regex;
-
-methods! {
-  Class,
-  _itself,
-  fn detect(s: RString) -> AnyObject {
-    lazy_static! {
-      // # Match escaped control characters
-      static ref RE_ESCAPED_CONTROL_CHAR: Regex = Regex::new(r"\\(?:\{#|##|#\})").unwrap();
-
-      // # Match ##...## or {#...#} control blocks.
-      static ref RE_CONTROL_BLOCK: Regex = Regex::new(r"##.*?##|\{#.*?#\}").unwrap();
-
-      // Match any character in the block of Brahmic scripts
-      // between Devanagari and Malayalam.
-      static ref RE_BRAHMIC_RANGE: Regex = Regex::new(r"[\x{0900}-\x{0d7f}]").unwrap();
-
-      // Match on special Roman characters
-      static ref RE_IAST_OR_KOLKATA_ONLY: Regex = Regex::new(r"(?i)[āīūṛṝḷḹēōṃḥṅñṭḍṇśṣḻ]").unwrap();
-
-      // Match on Kolkata-specific Roman characters
-      static ref RE_KOLKATA_ONLY: Regex = Regex::new(r"(?i)[ēō]").unwrap();
-
-      // Match on ITRANS-only
-      static ref RE_ITRANS_ONLY: Regex = Regex::new(r"ee|oo|\^[iI]|RR[iI]|L[iI]|~N|N\^|Ch|chh|JN|sh|Sh|\.a").unwrap();
-
-      // Match on SLP1-only characters and bigrams
-      static ref RE_SLP1_ONLY: Regex = Regex::new(r"[fFxXEOCYwWqQPB]|kz|N[kg]|tT|dD|S[cn]|[aAiIuUeo]R|G[yr]").unwrap();
-
-      // Match on Velthuis-only characters
-      static ref RE_VELTHUIS_ONLY: Regex = Regex::new(r"\.[mhnrlntds]|\x22n|~s").unwrap();
-
-      // Match on chars shared by ITRANS and Velthuis
-      static ref RE_ITRANS_OR_VELTHUIS_ONLY: Regex = Regex::new(r"aa|ii|uu|~n").unwrap();
-
-      // Match on characters available in Harvard-Kyoto
-      static ref RE_HARVARD_KYOTO: Regex = Regex::new(r"[aAiIuUeoRMHkgGcjJTDNtdnpbmyrlvzSsh]").unwrap();
-
-      static ref _DEVANAGARI: Symbol = Symbol::new("devanagari");
-      static ref _BENGALI: Symbol = Symbol::new("bengali");
-      static ref _GURMUKHI: Symbol = Symbol::new("gurmukhi");
-      static ref _GUJARATI: Symbol = Symbol::new("gujarati");
-      static ref _ORIYA: Symbol = Symbol::new("oriya");
-      static ref _TAMIL: Symbol = Symbol::new("tamil");
-      static ref _TELUGU: Symbol = Symbol::new("telugu");
-      static ref _KANNADA: Symbol = Symbol::new("kannada");
-      static ref _MALAYALAM: Symbol = Symbol::new("malayalam");
-      static ref _IAST: Symbol = Symbol::new("iast");
-      static ref _KOLKATA: Symbol = Symbol::new("kolkata");
-      static ref _ITRANS: Symbol = Symbol::new("itrans");
-      static ref _SLP1: Symbol = Symbol::new("slp1");
-      static ref _VELTHUIS: Symbol = Symbol::new("velthuis");
-      static ref _HK: Symbol = Symbol::new("hk");
-      static ref _NIL: NilClass = NilClass::new();
-    }
-
-    let r_replaced_str = &RE_ESCAPED_CONTROL_CHAR.replace_all(&s.to_string(), "");
-    let r_str = &RE_CONTROL_BLOCK.replace_all(r_replaced_str, "");
-
-    // Brahmic schemes are all within a specific range of code points.
-    let brahmic_match = RE_BRAHMIC_RANGE.find(r_str);
-    if brahmic_match != None {
-      let brahmic_match = brahmic_match.unwrap();
-      let brahmic_codepoint = r_str.chars().nth(brahmic_match.0).unwrap() as u32;
-
-      if brahmic_codepoint < 0x0980 {
-        return _DEVANAGARI.to_any_object();
-      } else if brahmic_codepoint < 0x0A00 {
-        return _BENGALI.to_any_object();
-      } else if brahmic_codepoint < 0x0A80 {
-        return _GURMUKHI.to_any_object();
-      } else if brahmic_codepoint < 0x0B00 {
-        return _GUJARATI.to_any_object();
-      } else if brahmic_codepoint < 0x0B80 {
-        return _ORIYA.to_any_object();
-      } else if brahmic_codepoint < 0x0C00 {
-        return _TAMIL.to_any_object();
-      } else if brahmic_codepoint < 0x0C80 {
-        return _TELUGU.to_any_object();
-      } else if brahmic_codepoint < 0x0D00 {
-        return _KANNADA.to_any_object();
-      } else {
-        return _MALAYALAM.to_any_object();
-      }
-    }
-
-    // Romanizations
-    if RE_IAST_OR_KOLKATA_ONLY.is_match(r_str) {
-      if RE_KOLKATA_ONLY.is_match(r_str) {
-        return _KOLKATA.to_any_object();
-      } else {
-        return _IAST.to_any_object();
-      }
-    } else if RE_ITRANS_ONLY.is_match(r_str) {
-      return _ITRANS.to_any_object();
-    } else if RE_SLP1_ONLY.is_match(r_str) {
-      return _SLP1.to_any_object();
-    } else if RE_VELTHUIS_ONLY.is_match(r_str) {
-      return _VELTHUIS.to_any_object();
-    } else if RE_ITRANS_OR_VELTHUIS_ONLY.is_match(r_str) {
-      return _ITRANS.to_any_object();
-    } else if RE_HARVARD_KYOTO.is_match(r_str) {
-      return _HK.to_any_object();
-    }
-    return _NIL.to_any_object();
-  }
-}
-
-#[no_mangle]
-pub extern fn init_rusty_sanscript() {
-  Class::from_existing("RustySanscriptDetect").define(|itself| {
-    itself.def("rust_detect_scheme", detect);
-  });
-}
+mod rb;
+// Exports a Sanscript::Detect::Rust module
+pub mod detect;

data/src/rb.rs

@@ -0,0 +1,48 @@
+use std::ffi::{CStr, CString};
+
+use ruby_sys::{class, string, symbol, util};
+use ruby_sys::types::c_char;
+use ruby_sys::value::RubySpecialConsts::Nil;
+
+pub use ruby_sys::types::{CallbackPtr, Value};
+pub const RB_NIL: Value = Value { value: Nil as usize };
+
+//
+// Helper functions for dealing with Ruby and CStrings
+//
+
+#[inline(always)]
+fn str_to_cstrp(s: &str) -> *const c_char {
+  CString::new(s).unwrap().as_ptr()
+}
+
+#[inline]
+pub fn rbstr_to_str<'a>(s: *const Value) -> &'a str {
+  unsafe {
+    let c_strp = string::rb_string_value_cstr(s);
+    CStr::from_ptr(c_strp).to_str().unwrap()
+  }
+}
+
+pub fn str_to_sym(s: &str) -> Value {
+  let c_strp = str_to_cstrp(s);
+  unsafe {
+    let id = util::rb_intern(c_strp);
+    symbol::rb_id2sym(id)
+  }
+}
+
+pub fn define_module(name: &str) -> Value {
+  let c_strp = str_to_cstrp(name);
+  unsafe { class::rb_define_module(c_strp) }
+}
+
+pub fn define_module_under(parent: &Value, name: &str) -> Value {
+  let c_strp = str_to_cstrp(name);
+  unsafe { class::rb_define_module_under(*parent, c_strp) }
+}
+
+pub fn define_method(module: &Value, name: &str, method: CallbackPtr, argc: i32) {
+  let c_strp = str_to_cstrp(name);
+  unsafe { class::rb_define_method(*module, c_strp, method, argc) }
+}

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: sanscript
 version: !ruby/object:Gem::Version
-  version: 0.6.2
+  version: 0.7.0
 platform: ruby
 authors:
 - Tim Bellefleur
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2016-08-16 00:00:00.000000000 Z
+date: 2016-09-27 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -16,28 +16,28 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.12'
+        version: '1.13'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.12'
+        version: '1.13'
 - !ruby/object:Gem::Dependency
   name: rake
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '11.2'
+        version: '11.3'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '11.2'
+        version: '11.3'
 - !ruby/object:Gem::Dependency
   name: rspec
   requirement: !ruby/object:Gem::Requirement
@@ -72,28 +72,28 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '0.41'
+        version: '0.43'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '0.41'
+        version: '0.43'
 - !ruby/object:Gem::Dependency
   name: rubocop-rspec
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.5'
+        version: '1.7'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.5'
+        version: '1.7'
 - !ruby/object:Gem::Dependency
   name: pry
   requirement: !ruby/object:Gem::Requirement
@@ -114,14 +114,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '2.6'
+        version: '2.7'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '2.6'
+        version: '2.7'
 - !ruby/object:Gem::Dependency
   name: yard
   requirement: !ruby/object:Gem::Requirement
@@ -156,14 +156,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '0.5'
+        version: '0.7'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '0.5'
+        version: '0.7'
 description: 
 email:
 - nomoon@phoebus.ca
@@ -178,7 +178,6 @@ files:
 - ".rubocop.yml"
 - ".travis.yml"
 - CODE_OF_CONDUCT.md
-- Cargo.lock
 - Cargo.toml
 - Gemfile
 - LICENSE.txt
@@ -192,11 +191,15 @@ files:
 - lib/sanscript/detect/ruby24.rb
 - lib/sanscript/detect/ruby2x.rb
 - lib/sanscript/exceptions.rb
+- lib/sanscript/rust.rb
 - lib/sanscript/transliterate.rb
 - lib/sanscript/transliterate/schemes.rb
 - lib/sanscript/version.rb
 - sanscript.gemspec
+- src/detect/mod.rs
+- src/detect/ruby.rs
 - src/lib.rs
+- src/rb.rs
 homepage: https://github.com/ubcsanskrit/sanscript.rb
 licenses:
 - MIT

data/Cargo.lock

@@ -1,115 +0,0 @@
-[root]
-name = "rusty_sanscript"
-version = "0.2.0"
-dependencies = [
- "lazy_static 0.2.1 (registry+https://github.com/rust-lang/crates.io-index)",
- "regex 0.1.73 (registry+https://github.com/rust-lang/crates.io-index)",
- "ruru 0.7.8 (registry+https://github.com/rust-lang/crates.io-index)",
- "unicode-normalization 0.1.2 (registry+https://github.com/rust-lang/crates.io-index)",
-]
-
-[[package]]
-name = "aho-corasick"
-version = "0.5.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-dependencies = [
- "memchr 0.1.11 (registry+https://github.com/rust-lang/crates.io-index)",
-]
-
-[[package]]
-name = "kernel32-sys"
-version = "0.2.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-dependencies = [
- "winapi 0.2.8 (registry+https://github.com/rust-lang/crates.io-index)",
- "winapi-build 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)",
-]
-
-[[package]]
-name = "lazy_static"
-version = "0.2.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-
-[[package]]
-name = "libc"
-version = "0.2.15"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-
-[[package]]
-name = "memchr"
-version = "0.1.11"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-dependencies = [
- "libc 0.2.15 (registry+https://github.com/rust-lang/crates.io-index)",
-]
-
-[[package]]
-name = "regex"
-version = "0.1.73"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-dependencies = [
- "aho-corasick 0.5.2 (registry+https://github.com/rust-lang/crates.io-index)",
- "memchr 0.1.11 (registry+https://github.com/rust-lang/crates.io-index)",
- "regex-syntax 0.3.4 (registry+https://github.com/rust-lang/crates.io-index)",
- "thread_local 0.2.6 (registry+https://github.com/rust-lang/crates.io-index)",
- "utf8-ranges 0.1.3 (registry+https://github.com/rust-lang/crates.io-index)",
-]
-
-[[package]]
-name = "regex-syntax"
-version = "0.3.4"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-
-[[package]]
-name = "ruby-sys"
-version = "0.2.6"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-dependencies = [
- "libc 0.2.15 (registry+https://github.com/rust-lang/crates.io-index)",
-]
-
-[[package]]
-name = "ruru"
-version = "0.7.8"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-dependencies = [
- "ruby-sys 0.2.6 (registry+https://github.com/rust-lang/crates.io-index)",
-]
-
-[[package]]
-name = "thread-id"
-version = "2.0.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-dependencies = [
- "kernel32-sys 0.2.2 (registry+https://github.com/rust-lang/crates.io-index)",
- "libc 0.2.15 (registry+https://github.com/rust-lang/crates.io-index)",
-]
-
-[[package]]
-name = "thread_local"
-version = "0.2.6"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-dependencies = [
- "thread-id 2.0.0 (registry+https://github.com/rust-lang/crates.io-index)",
-]
-
-[[package]]
-name = "unicode-normalization"
-version = "0.1.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-
-[[package]]
-name = "utf8-ranges"
-version = "0.1.3"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-
-[[package]]
-name = "winapi"
-version = "0.2.8"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-
-[[package]]
-name = "winapi-build"
-version = "0.1.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-