sanscript 0.6.0 → 0.6.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: dd5f98881a344f630109b020f7614bf3c860a919
-  data.tar.gz: e3a1353bf994c3aabee9b779cc35e70c98f70bb3
+  metadata.gz: 7faf23075373caded757c32e4e708d85e6e39c10
+  data.tar.gz: 049ea58c3690d4dc6627d6a8a205f19914e69b33
 SHA512:
-  metadata.gz: 925e8af64d8eeed22da8b35ffa3b569b80da03956b419bc5055c7bfafc6e52f80dffbdf810f7c538ef50ec103df9431c334295f84811f861d5a78d88a30dc2c5
-  data.tar.gz: 82d5cba98f7a7f38899df4bdc20a72597882743ed7fc2a484da9c2d96097dea075cbcfa433b3db20a0ef15075a962d0022851c250a80f337e218d4f2411dc31d
+  metadata.gz: a8bf83a24936460b5331580b094bac0c3aff7e6db4b7839e4fe46b282161ff4d927aa5af8e5fa76afdea3eeb72ecd5353e7f32f15db5bc49bd02153d317ae4e1
+  data.tar.gz: bc16c1526a2d3c89c047c7dbe4b27e5698a7aeb59822505006f2da6e6b7237fb15b7ecd3870ba6b024bb244bcb62912e5ac83cf915e0166f64d9e48d1f4ae77f

data/.gitignore

@@ -7,6 +7,6 @@
 /pkg/
 /spec/reports/
 /tmp/
-/rust/target
-/rust/Makefile
-/rust/libsanscript.*
+/lib/librusty_sanscript.*
+/target
+mkmf.log

data/{rust/Cargo.lock → Cargo.lock}

@@ -1,10 +1,11 @@
 [root]
-name = "sanscript"
-version = "0.1.0"
+name = "rusty_sanscript"
+version = "0.2.0"
 dependencies = [
  "lazy_static 0.2.1 (registry+https://github.com/rust-lang/crates.io-index)",
- "libc 0.2.15 (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]]
@@ -59,6 +60,22 @@ 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"
@@ -76,6 +93,11 @@ 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"

data/Cargo.toml

@@ -0,0 +1,16 @@
+[package]
+name = "rusty_sanscript"
+version = "0.2.0"
+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]
+
+[lib]
+crate-type = ["dylib"]

data/Rakefile

@@ -1,13 +1,31 @@
 # frozen_string_literal: true
 require "bundler/gem_tasks"
-require "rspec/core/rake_task"
 
-RSpec::Core::RakeTask.new(:spec)
-
-task default: :spec
+# 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"
 
-task :compile do
-  Dir.chdir("#{File.dirname(__FILE__)}/rust")
-  require "./extconf.rb"
-  sh "make clean && make && make install" if File.exist?("Makefile")
+# Ensure missing RSpec development dependency doesn't kill gem install.
+begin
+  require "rspec/core/rake_task"
+  RSpec::Core::RakeTask.new(:spec)
+  task default: :spec
+rescue LoadError
+  nil
 end

data/lib/sanscript.rb

@@ -56,4 +56,29 @@ 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/detect.rb

@@ -53,7 +53,8 @@ module Sanscript
     #   Attempts to detect the encoding scheme of the provided string.
     #
     #   Uses the most efficient implementation for your ruby version
-    #   (either {Ruby2x#detect_scheme} or {Ruby24#detect_scheme})
+    #   (either {Ruby2x#ruby_detect_scheme} or {Ruby24#ruby_detect_scheme}) or
+    #   the Rust native extension if available.
     #
     #   @param text [String] a string of Sanskrit text
     #   @return [Symbol, nil] the Symbol of the scheme, or nil if no match
@@ -67,24 +68,9 @@ module Sanscript
       require "sanscript/detect/ruby2x"
       extend Ruby2x
     end
-    # :nocov:
-
-    # Rust FFI
     class << self
-      begin
-        require "ffi"
-        extend FFI::Library
-        ffi_lib Dir.glob(File.join(GEM_ROOT, "rust/libsanscript.*")).first
-        attach_function :_rust_detect, :detect, [:string], :int
-        RUST_SCHEMES = %i[devanagari bengali gurmukhi gujarati oriya tamil telugu kannada malayalam iast kolkata itrans slp1 velthuis hk].unshift(nil).freeze
-        private_constant :RUST_SCHEMES
-        def rust_detect_scheme(text)
-          RUST_SCHEMES[_rust_detect(text)]
-        end
-        alias detect_scheme rust_detect_scheme
-      rescue LoadError
-        alias detect_scheme ruby_detect_scheme
-      end
+      alias detect_scheme ruby_detect_scheme
     end
+    # :nocov:
   end
 end

data/lib/sanscript/version.rb

@@ -1,7 +1,7 @@
 # frozen_string_literal: true
 module Sanscript
   # The version number
-  VERSION = "0.6.0"
+  VERSION = "0.6.1"
 
   GEM_ROOT = Pathname.new(File.realpath(File.join(__dir__, "..", "..")))
   private_constant :GEM_ROOT

data/sanscript.gemspec

@@ -18,7 +18,7 @@ Gem::Specification.new do |spec|
   spec.bindir        = "exe"
   spec.executables   = spec.files.grep(%r{^exe/}) { |f| File.basename(f) }
   spec.require_paths = ["lib"]
-  spec.extensions    = Dir["rust/extconf.rb"]
+  spec.extensions    = "Rakefile"
 
   spec.required_ruby_version = "~> 2.2"
 
@@ -33,5 +33,5 @@ Gem::Specification.new do |spec|
   spec.add_development_dependency "yard", "~> 0.9"
 
   spec.add_runtime_dependency "ragabash", "~> 0.2"
-  spec.add_runtime_dependency "ffi"
+  spec.add_runtime_dependency "thermite", "~> 0.5"
 end

data/src/lib.rs

@@ -0,0 +1,122 @@
+#[macro_use] extern crate lazy_static;
+#[macro_use] extern crate ruru;
+extern crate unicode_normalization;
+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);
+  });
+}

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: sanscript
 version: !ruby/object:Gem::Version
-  version: 0.6.0
+  version: 0.6.1
 platform: ruby
 authors:
 - Tim Bellefleur
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2016-08-14 00:00:00.000000000 Z
+date: 2016-08-16 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -151,25 +151,25 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '0.2'
 - !ruby/object:Gem::Dependency
-  name: ffi
+  name: thermite
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '0.5'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '0.5'
 description: 
 email:
 - nomoon@phoebus.ca
 executables: []
 extensions:
-- rust/extconf.rb
+- Rakefile
 extra_rdoc_files: []
 files:
 - ".codeclimate.yml"
@@ -178,6 +178,8 @@ files:
 - ".rubocop.yml"
 - ".travis.yml"
 - CODE_OF_CONDUCT.md
+- Cargo.lock
+- Cargo.toml
 - Gemfile
 - LICENSE.txt
 - README.md
@@ -193,11 +195,8 @@ files:
 - lib/sanscript/transliterate.rb
 - lib/sanscript/transliterate/schemes.rb
 - lib/sanscript/version.rb
-- rust/Cargo.lock
-- rust/Cargo.toml
-- rust/extconf.rb
-- rust/src/lib.rs
 - sanscript.gemspec
+- src/lib.rs
 homepage: https://github.com/ubcsanskrit/sanscript.rb
 licenses:
 - MIT

data/rust/Cargo.toml

@@ -1,12 +0,0 @@
-[package]
-name = "sanscript"
-version = "0.1.0"
-authors = ["Tim Bellefleur <nomoon@phoebus.ca>"]
-
-[dependencies]
-libc = "*"
-lazy_static = "*"
-regex = "*"
-
-[lib]
-crate-type = ["dylib"]

data/rust/extconf.rb

@@ -1,6 +0,0 @@
-# frozen_string_literal: true
-require "mkmf"
-unless !system("cargo --version") || !system("rustc --version")
-  create_makefile("sanscript")
-  File.write("Makefile", "all:\n\tcargo build --release\n\nclean:\n\trm -rf target\n\ninstall:\n\tmv target/release/libsanscript.* .\n\trm -rf target\n")
-end

data/rust/src/lib.rs

@@ -1,100 +0,0 @@
-extern crate libc;
-#[macro_use] extern crate lazy_static;
-extern crate regex;
-
-use libc::{c_char, int32_t};
-use std::ffi::CStr;
-use std::str;
-use regex::Regex;
-
-#[no_mangle]
-pub extern fn detect(s: *const c_char) -> int32_t {
-  let c_str = unsafe {
-    assert!(!s.is_null());
-
-    CStr::from_ptr(s)
-  };
-
-  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();
-  }
-
-  let r_replaced_str = &RE_ESCAPED_CONTROL_CHAR.replace_all(c_str.to_str().unwrap(), "");
-  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 1;
-    } else if brahmic_codepoint < 0x0A00 {
-      return 2;
-    } else if brahmic_codepoint < 0x0A80 {
-      return 3;
-    } else if brahmic_codepoint < 0x0B00 {
-      return 4;
-    } else if brahmic_codepoint < 0x0B80 {
-      return 5;
-    } else if brahmic_codepoint < 0x0C00 {
-      return 6;
-    } else if brahmic_codepoint < 0x0C80 {
-      return 7;
-    } else if brahmic_codepoint < 0x0D00 {
-      return 8;
-    } else {
-      return 9;
-    }
-  }
-
-  // Romanizations
-  if RE_IAST_OR_KOLKATA_ONLY.is_match(r_str) {
-    if RE_KOLKATA_ONLY.is_match(r_str) {
-      return 11;
-    } else {
-      return 10;
-    }
-  } else if RE_ITRANS_ONLY.is_match(r_str) {
-    return 12;
-  } else if RE_SLP1_ONLY.is_match(r_str) {
-    return 13;
-  } else if RE_VELTHUIS_ONLY.is_match(r_str) {
-    return 14;
-  } else if RE_ITRANS_OR_VELTHUIS_ONLY.is_match(r_str) {
-    return 12;
-  } else if RE_HARVARD_KYOTO.is_match(r_str) {
-    return 15;
-  }
-  return 0;
-}