hunspell-i18n 0.2.0

data/.rspec

@@ -0,0 +1,2 @@
+--colour
+--format progress

data/README.md

@@ -0,0 +1,16 @@
+# Ruby bindings for Hunspell 1.2.x
+
+This version expects to receive the word in UTF-8 encoding, and it will
+then use iconv behind the scenes in order to convert to and from
+the dictionary encoding.
+
+## Usage
+
+    prefix = '/usr/share/myspell/'  # Fedora
+    speller = Hunspell.new(prefix + 'en_US.aff', prefix + 'en_US.dic')
+    speller.valid?('dog')           # -> true
+    speller.valid?('dpg')           # -> false
+    speller.suggest('dpg')          # -> ["dog", "dg", "pg", "deg", "dig", ...]
+
+
+Copyright 2011 Roman Shterenzon, released under the MIT license

data/Rakefile

@@ -0,0 +1,29 @@
+require 'rake'
+require 'rake/rdoctask'
+
+desc 'Default: create gemspec'
+task :default => :gemspec
+
+desc 'Generate documentation'
+Rake::RDocTask.new(:rdoc) do |rdoc|
+  rdoc.rdoc_dir = 'rdoc'
+  rdoc.title    = 'Hunspell'
+  rdoc.options << '--line-numbers' << '--inline-source'
+  rdoc.rdoc_files.include('README')
+  rdoc.rdoc_files.include('lib/**/*.rb')
+end
+
+begin
+  require 'jeweler'
+  Jeweler::Tasks.new do |gem|
+    gem.name = "hunspell-i18n"
+    gem.summary = %Q{Ruby bindings for libhunspell-1.2 with i18n support}
+    gem.description = gem.summary
+    gem.email = "romanbsd@yahoo.com"
+    gem.homepage = "http://github.com/romanbsd/hunspell"
+    gem.authors = ["Roman Shterenzon"]
+    gem.extensions = ['ext/extconf.rb']
+  end
+  Jeweler::GemcutterTasks.new
+rescue LoadError
+end

data/VERSION

@@ -0,0 +1 @@
+0.2.0

data/ext/extconf.rb

@@ -0,0 +1,7 @@
+require 'mkmf'
+
+unless have_header('hunspell/hunspell.h') and have_library('hunspell-1.2')
+  raise 'Hunspell not found.'
+end
+
+create_makefile('hunspell')

data/ext/hunspell_wrap.c

@@ -0,0 +1,129 @@
+#include <ruby.h>
+#include <hunspell/hunspell.h>
+
+static VALUE rb_cHunspell;
+static VALUE rb_cIconv;
+static VALUE enc_utf8;
+static VALUE enc_iso8859_1;
+
+static void dict_destroy(Hunhandle *handle) {
+    Hunspell_destroy(handle);
+}
+
+/*
+ *  call-seq:
+ *     Hunspell.new(affix_path, dic_path)
+ *
+ *  Instantiate Hunspell with paths pointing to affix and dictionary files.
+ *
+ *  Example:
+ *    speller = Hunspell.new("/usr/share/myspell/en_US.aff", "/usr/share/myspell/en_US.dic")
+ *
+ */
+static VALUE dict_init(VALUE self, VALUE affix_path, VALUE dic_path) {
+    VALUE affpath = StringValue(affix_path), dpath = StringValue(dic_path);
+    Hunhandle *handle = Hunspell_create(RSTRING_PTR(affpath), RSTRING_PTR(dpath));
+    if (!handle) {
+      rb_raise(rb_eRuntimeError, "Failed to initialize Hunspell handle.");
+    }
+    VALUE dict = Data_Wrap_Struct(rb_cHunspell, NULL, dict_destroy, handle);
+    rb_iv_set(self, "dict", dict);
+    return self;
+}
+
+static inline Hunhandle* get_handle(VALUE self) {
+    Hunhandle *handle = NULL;
+    Data_Get_Struct(rb_iv_get(self, "dict"), Hunhandle, handle);
+    if (!handle) {
+      rb_raise(rb_eRuntimeError, "Hunspell handle not found.");
+    }
+    return handle;
+}
+
+/*
+ *  call-seq:
+ *     hunspell.encoding -> string
+ *
+ *  Returns the encoding of the dictionary.
+ */
+static VALUE wrap_encoding(VALUE self) {
+    Hunhandle *handle = get_handle(self);
+    char *enc = Hunspell_get_dic_encoding(handle);
+    return rb_str_new2(enc);
+}
+
+static VALUE recode_if_needed(VALUE self, VALUE str, int dir) {
+    VALUE enc = wrap_encoding(self);
+    VALUE from, to;
+
+    if (dir == 0) {
+        from = enc_utf8;
+        to = enc;
+    } else {
+        from = enc;
+        to = enc_utf8;
+    }
+    if ( rb_str_equal(enc, enc_iso8859_1) == Qfalse && rb_str_equal(enc, enc_utf8) == Qfalse) {
+        rb_funcall(str, rb_intern("force_encoding"), 1, enc_utf8);
+        return rb_funcall(rb_cIconv, rb_intern("conv"), 3, to, from, str);
+    } else {
+        return str;
+    }
+}
+
+/*
+ *  call-seq:
+ *     hunspell.suggest(misspeledword) -> ary
+ *
+ *  Returns a list of suggestions.
+ */
+static VALUE wrap_suggest(VALUE self, VALUE word) {
+    VALUE str = recode_if_needed(self, StringValue(word), 0);
+    VALUE res;
+    char** slst = NULL;
+    int i, count = 0;
+    Hunhandle *handle = get_handle(self);
+
+    count = Hunspell_suggest(handle, &slst, RSTRING_PTR(str));
+
+    res = rb_ary_new2(count);
+    for (i=0; i<count; ++i) {
+        rb_ary_push(res, recode_if_needed(self, rb_str_new2(slst[i]), 1));
+        free(slst[i]);
+    }
+
+    if (slst) {
+        free(slst);
+    }
+
+    return res;
+}
+
+/*
+ *  call-seq:
+ *     hunspell.valid?(word) -> bool
+ *
+ *  Checks if the word is in the dictionary.
+ */
+static VALUE wrap_check(VALUE self, VALUE word) {
+    VALUE str = recode_if_needed(self, StringValue(word), 0);
+    Hunhandle *handle = get_handle(self);
+    int rc = Hunspell_spell(handle, RSTRING_PTR(str));
+    return rc == 0 ? Qfalse : Qtrue;
+}
+
+
+void Init_hunspell(void) {
+    rb_cHunspell = rb_define_class("Hunspell", rb_cObject);
+    rb_define_method(rb_cHunspell, "initialize", dict_init, 2);
+    //rb_define_method(rb_cHunspell, "check?", wrap_check, 1);
+    rb_define_method(rb_cHunspell, "valid?", wrap_check, 1);
+    rb_define_method(rb_cHunspell, "suggest", wrap_suggest, 1);
+    rb_define_method(rb_cHunspell, "encoding", wrap_encoding, 0);
+    rb_require("iconv");
+    rb_cIconv = rb_const_get(rb_cObject, rb_intern("Iconv"));
+    enc_iso8859_1 = rb_str_new2("ISO8859-1");
+    enc_utf8 = rb_str_new2("UTF-8");
+    rb_global_variable(&enc_iso8859_1);
+    rb_global_variable(&enc_utf8);
+}

data/hunspell-i18n.gemspec

@@ -0,0 +1,44 @@
+# Generated by jeweler
+# DO NOT EDIT THIS FILE DIRECTLY
+# Instead, edit Jeweler::Tasks in Rakefile, and run 'rake gemspec'
+# -*- encoding: utf-8 -*-
+
+Gem::Specification.new do |s|
+  s.name = "hunspell-i18n"
+  s.version = "0.2.0"
+
+  s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
+  s.authors = ["Roman Shterenzon"]
+  s.date = "2012-01-21"
+  s.description = "Ruby bindings for libhunspell-1.2 with i18n support"
+  s.email = "romanbsd@yahoo.com"
+  s.extensions = ["ext/extconf.rb"]
+  s.extra_rdoc_files = [
+    "README.md"
+  ]
+  s.files = [
+    ".rspec",
+    "README.md",
+    "Rakefile",
+    "VERSION",
+    "ext/extconf.rb",
+    "ext/hunspell_wrap.c",
+    "hunspell-i18n.gemspec",
+    "spec/hunspell_spec.rb",
+    "spec/spec_helper.rb"
+  ]
+  s.homepage = "http://github.com/romanbsd/hunspell"
+  s.require_paths = ["lib"]
+  s.rubygems_version = "1.8.15"
+  s.summary = "Ruby bindings for libhunspell-1.2 with i18n support"
+
+  if s.respond_to? :specification_version then
+    s.specification_version = 3
+
+    if Gem::Version.new(Gem::VERSION) >= Gem::Version.new('1.2.0') then
+    else
+    end
+  else
+  end
+end
+

data/spec/hunspell_spec.rb

@@ -0,0 +1,33 @@
+require 'spec_helper'
+
+describe Hunspell do
+
+  before :all do
+    @speller = Hunspell.new('en_US.aff', 'en_US.dic')
+  end
+
+  it "should check if a word is valid" do
+    @speller.should be_valid('dog')
+    @speller.should_not be_valid('dxg')
+  end
+
+  it "survives GC" do
+    10.times do
+      @speller.suggest('dxg')
+    end
+    GC.start
+    10.times do
+      @speller.suggest('dxg')
+    end
+  end
+
+  it "should suggest alternate spellings for words" do
+    @speller.suggest('arbitrage').should == %w[
+      arbitrage
+      arbitrages
+      arbitrager
+      arbitraged
+      arbitrate
+    ]
+  end
+end

data/spec/spec_helper.rb

@@ -0,0 +1 @@
+require 'hunspell'

metadata

@@ -0,0 +1,62 @@
+--- !ruby/object:Gem::Specification 
+name: hunspell-i18n
+version: !ruby/object:Gem::Version 
+  prerelease: 
+  version: 0.2.0
+platform: ruby
+authors: 
+- Roman Shterenzon
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2012-01-21 00:00:00 Z
+dependencies: []
+
+description: Ruby bindings for libhunspell-1.2 with i18n support
+email: romanbsd@yahoo.com
+executables: []
+
+extensions: 
+- ext/extconf.rb
+extra_rdoc_files: 
+- README.md
+files: 
+- .rspec
+- README.md
+- Rakefile
+- VERSION
+- ext/extconf.rb
+- ext/hunspell_wrap.c
+- hunspell-i18n.gemspec
+- spec/hunspell_spec.rb
+- spec/spec_helper.rb
+homepage: http://github.com/romanbsd/hunspell
+licenses: []
+
+post_install_message: 
+rdoc_options: []
+
+require_paths: 
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+requirements: []
+
+rubyforge_project: 
+rubygems_version: 1.8.15
+signing_key: 
+specification_version: 3
+summary: Ruby bindings for libhunspell-1.2 with i18n support
+test_files: []
+