ots 0.4.4 → 0.5.0

data/README.md

@@ -1,18 +1,18 @@
 # OTS
 
-ots is an interface to libots - The open text summarizer
+ots is an interface to libots - The [open text summarizer](http://libots.sourceforge.net/).
 
 ## Dependencies
 
   * ruby 1.9.1 or later
   * libxml2
-  * glib2.0 
+  * glib2.0
   * homebrew (on MacOSX)
 
 ## Installation
 
 ### Debian flavors of Linux
- 
+
 ```
 
   # ruby & ruby development libraries (not needed if you use rvm)
@@ -34,6 +34,9 @@ ots is an interface to libots - The open text summarizer
  # update homebrew to latest & greatest version
  GIT_SSL_NO_VERIFY=1 brew update
 
+ # optional: macosx normally has libxml2 installed if not try
+ brew install libxml2
+
  # install glib
  brew install glib
 
@@ -47,13 +50,13 @@ ots is an interface to libots - The open text summarizer
 ```
   OTS
     .parse        #=> OTS::Article
-    .dictionaries #=> Array
+    .languages    #=> Array
 
   OTS::Article
     .new
-    #summarize    #=> Array
+    #topics       #=> Array
     #keywords     #=> Array
-    #title        #=> String
+    #summarize    #=> Array
 
 ```
 
@@ -62,17 +65,15 @@ ots is an interface to libots - The open text summarizer
 ```ruby
   require 'ots'
   article = OTS.parse("I think I need some ice cream to cool me off. It is too hot down under")
+  article = OTS.parse("j'ai besoin de la crème glacée. il fait trop chaud en australie.", language: "fr")
+  article = OTS.parse("j'ai besoin de la crème glacée. il fait trop chaud en australie.", dictionary: "custom.xml")
 
+  article.topics
   article.keywords
-  article.summarize(lines: 1)
-  article.summarize(percent: 50)
-
-  article = OTS.parse("j'ai besoin de la crème glacée. il fait trop chaud en australie.", "fr")
-  article.keywords
-  article.summarize(lines: 1)
   article.summarize(percent: 50)
+  article.summarize(sentences: 1)
 
-  OTS.dictionaries #=> list of supported dictionaries
+  OTS.languages #=> list of supported languages dictionaries baked-in to libots
 ```
 
 ## License

data/ext/dictionary.c

@@ -59,12 +59,13 @@ ots_load_xml_dictionary (OtsArticle * Doc, const char *name)
   char *local_dict_name;
 
   dict_name = g_strdup_printf ("%s%s.xml", DICTIONARY_DIR, name);
-  local_dict_name = g_strdup_printf ("%s.xml", name);
+  local_dict_name = g_strdup_printf ("%s", name);
 
+  if (g_file_test(local_dict_name,G_FILE_TEST_EXISTS))
+    doc = xmlReadFile (local_dict_name, 0, XML_PARSE_NOERROR | XML_PARSE_NOWARNING);
+  if (doc == NULL && g_file_test(dict_name,G_FILE_TEST_EXISTS))
+    doc = xmlReadFile (dict_name, 0, XML_PARSE_NOERROR | XML_PARSE_NOWARNING);
 
-	if (g_file_test(local_dict_name,G_FILE_TEST_EXISTS))
-		  doc = xmlParseFile (local_dict_name); /* it warns to the screen so we cant use it; enable for web services only */
-  if (doc == NULL)   doc = xmlParseFile (dict_name);
   if (doc == NULL) return (FALSE);
 
   head = xmlDocGetRootElement (doc);

data/ext/ots.c

@@ -30,18 +30,28 @@ void article_load_dictionary(OtsArticle *article, char *name) {
 }
 
 VALUE article_initialize(int argc, VALUE *argv, VALUE self) {
-    VALUE text, dictionary;
+    VALUE text, options, language, dictionary = Qnil;
     OtsArticle *article = article_handle(self);
 
-    rb_scan_args(argc, argv, "11", &text, &dictionary);
+    rb_scan_args(argc, argv, "11", &text, &options);
+
+    language = rb_str_new2("en");
 
     if (TYPE(text) != T_STRING)
         rb_raise(rb_eArgError, "invalid +text+");
 
-    if (NIL_P(dictionary))
-        article_load_dictionary(article, "en");
-    else
+    if (!NIL_P(options)) {
+        if (TYPE(options) != T_HASH)
+            rb_raise(rb_eArgError, "invalid +options+ hash");
+
+        dictionary = rb_hash_aref(options, ID2SYM(rb_intern("dictionary")));
+        language   = rb_hash_aref(options, ID2SYM(rb_intern("language")));
+    }
+
+    if (!NIL_P(dictionary))
         article_load_dictionary(article, CSTRING(dictionary));
+    else
+        article_load_dictionary(article, CSTRING(language));
 
     ots_parse_stream(RSTRING_PTR(text), RSTRING_LEN(text), article);
     ots_grade_doc(article);
@@ -87,11 +97,11 @@ VALUE article_summarize(VALUE self, VALUE options) {
     if (TYPE(options) != T_HASH)
         rb_raise(rb_eArgError, "expect an options hash");
 
-    lines   = rb_hash_aref(options, ID2SYM(rb_intern("lines")));
+    lines   = rb_hash_aref(options, ID2SYM(rb_intern("sentences")));
     percent = rb_hash_aref(options, ID2SYM(rb_intern("percent")));
 
     if (NIL_P(lines) && NIL_P(percent))
-        rb_raise(rb_eArgError, "expect +lines+ or +percent+ to be provided");
+        rb_raise(rb_eArgError, "expect +sentences+ or +percent+");
 
     if (lines != Qnil)
         ots_highlight_doc_lines(article, NUM2INT(lines));
@@ -101,9 +111,13 @@ VALUE article_summarize(VALUE self, VALUE options) {
     return article_summary(article, (rb_encoding *)rb_iv_get(self, "@encoding"));
 }
 
-VALUE article_title(VALUE self) {
+VALUE article_topics(VALUE self) {
     OtsArticle *article = article_handle(self);
-    return (article->title ? rb_enc_str_new2(article->title, (rb_encoding*)rb_iv_get(self, "@encoding")) : Qnil);
+
+    return
+        article->title ?
+            rb_str_split(rb_enc_str_new2(article->title, (rb_encoding*)rb_iv_get(self, "@encoding")), ",") :
+            Qnil;
 }
 
 typedef struct {
@@ -136,16 +150,16 @@ VALUE ots_parse(int argc, VALUE *argv, VALUE self) {
     return article;
 }
 
-VALUE ots_dictionaries(VALUE self) {
+VALUE ots_languages(VALUE self) {
     DIR *dir;
     struct dirent *entry;
-    VALUE dictionaries = rb_ary_new();
+    VALUE languages = rb_ary_new();
 
     if ((dir = opendir(DICTIONARY_DIR))) {
         while ((entry = readdir(dir))) {
             // entry->d_type is not portable.
             if (strstr(entry->d_name, ".xml"))
-                rb_ary_push(dictionaries, rb_str_new(entry->d_name, strlen(entry->d_name) - 4));
+                rb_ary_push(languages, rb_str_new(entry->d_name, strlen(entry->d_name) - 4));
         }
     }
     else {
@@ -153,7 +167,7 @@ VALUE ots_dictionaries(VALUE self) {
     }
 
     closedir(dir);
-    return dictionaries;
+    return languages;
 }
 
 /* init */
@@ -164,11 +178,11 @@ void Init_ots(void) {
 
     rb_define_method(cArticle, "initialize", RUBY_METHOD_FUNC(article_initialize), -1);
     rb_define_method(cArticle, "summarize",  RUBY_METHOD_FUNC(article_summarize),   1);
-    rb_define_method(cArticle, "title",      RUBY_METHOD_FUNC(article_title),       0);
+    rb_define_method(cArticle, "topics",     RUBY_METHOD_FUNC(article_topics),      0);
     rb_define_method(cArticle, "keywords",   RUBY_METHOD_FUNC(article_keywords),    0);
 
-    rb_define_module_function(mOTS, "parse",        RUBY_METHOD_FUNC(ots_parse),       -1);
-    rb_define_module_function(mOTS, "dictionaries", RUBY_METHOD_FUNC(ots_dictionaries), 0);
+    rb_define_module_function(mOTS, "parse",     RUBY_METHOD_FUNC(ots_parse),      -1);
+    rb_define_module_function(mOTS, "languages", RUBY_METHOD_FUNC(ots_languages),   0);
 
     rb_define_alloc_func(cArticle, article_allocate);
 

data/ext/version.h

@@ -1,2 +1,2 @@
 #pragma once
-#define RUBY_OTS_VERSION            "0.4.4"
+#define RUBY_OTS_VERSION            "0.5.0"

data/test/test_article.rb

@@ -12,8 +12,8 @@ describe 'OTS::Article' do
     @article = OTS::Article.new(@sample)
   end
 
-  it 'should extract title keywords from given document' do
-    assert_equal 'species,turtle,subspecies,pacific,atlantic', @article.title
+  it 'should extract topic keywords from given document' do
+    assert_equal %w(species turtle subspecies pacific atlantic), @article.topics
   end
 
   it 'should extract keywords from given document' do
@@ -27,7 +27,7 @@ describe 'OTS::Article' do
 
 
   it 'should extract keywords from given document' do
-    lines  = @article.summarize(lines: 2).map {|line| [line[:sentence].gsub(/\s+/, ' ').strip, line[:score]]}
+    lines  = @article.summarize(sentences: 2).map {|line| [line[:sentence].gsub(/\s+/, ' ').strip, line[:score]]}
     expect = [
       ["The hawksbill turtle is a critically endangered sea turtle belonging to the family Cheloniidae.", 48],
       ["The species has a worldwide distribution, with Atlantic and Pacific subspecies.", 20],
@@ -39,14 +39,37 @@ describe 'OTS::Article' do
   it 'should utf8 encode strings properly' do
     text    = "The hawksbill turtle\xE2\x80\x93is critically endangered.".force_encoding('utf-8')
     article = OTS.parse(text)
-    summary = article.summarize(lines: 1).first[:sentence]
+    summary = article.summarize(sentences: 1).first[:sentence]
     assert_equal text, summary
   end
 
   describe 'dictionaries' do
+    before do
+      @text = "j'ai besoin de la crème glacée. il fait trop chaud en australie."
+    end
+
     it 'should load the french dictionary' do
-      article = OTS.parse("j'ai besoin de la crème glacée. il fait trop chaud en australie.", "fr")
-      assert_equal "j'ai besoin de la crème glacée.", article.summarize(lines: 1).first[:sentence]
+      article = OTS.parse(@text, language: "fr")
+      assert_equal "j'ai besoin de la crème glacée.", article.summarize(sentences: 1).first[:sentence]
+    end
+
+    it 'should load the french dictionary given path' do
+      article = OTS.parse(@text, dictionary: File.join(File.dirname(__FILE__), '..', 'dictionaries', 'fr.xml'))
+      assert_equal "j'ai besoin de la crème glacée.", article.summarize(sentences: 1).first[:sentence]
+    end
+
+    it 'should raise LoadError on invalid language or dictionaries' do
+      assert_raises(LoadError) do
+        OTS.parse('hello world', language: "xxx")
+      end
+
+      assert_raises(LoadError) do
+        OTS.parse('hello world', dictionary: "xxx")
+      end
+
+      assert_raises(LoadError) do
+        OTS.parse('hello world', dictionary: __FILE__)
+      end
     end
   end
 end

data/test/test_ots.rb

@@ -12,12 +12,12 @@ describe 'OTS' do
   end
 
   it 'should return a list of dictonaries' do
-    dictionaries = OTS.dictionaries
+    languages = OTS.languages
 
     %w(en fr it es de ru).each do |name|
-      assert dictionaries.include?(name), "has #{name} dictionary"
+      assert languages.include?(name), "has #{name} language dictionary"
     end
 
-    assert_empty dictionaries.reject {|name| name.size == 2}, "dictionaries path should not have other junk"
+    assert_empty languages.reject {|name| name.size == 2}, "dictionaries path should not have other junk"
   end
 end

metadata

@@ -4,9 +4,9 @@ version: !ruby/object:Gem::Version
   prerelease: false
   segments: 
   - 0
-  - 4
-  - 4
-  version: 0.4.4
+  - 5
+  - 0
+  version: 0.5.0
 platform: ruby
 authors: 
 - Bharanee Rathna
@@ -14,7 +14,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2012-01-09 00:00:00 +11:00
+date: 2012-01-10 00:00:00 +11:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency