Javascript Tests, UMD & Node Packaging, and pretty docs

.gitignore

@@ -0,0 +1,3 @@
+.DS_Store
+node_modules
+jsopos.min.js

README.md

@@ -0,0 +1,135 @@
+# jsPOS
+
+jsPOS is a Javascript port of [Mark Watson's FastTag Part of Speech Tagger](http://www.markwatson.com/opensource/) which was itself based on Eric Brill's trained rule set and English lexicon. jsPOS also includes a basic lexer that can be used to extract words and other tokens from text strings.
+
+
+## FILES
+
+* `lexicon.js` - Javascript version of Eric Brill's English lexicon
+* `lexer.js` - Lexer to break a sentence into taggable tokens (e.g. words)
+* `POSTagger.js` - the Part of Speech tagger.
+
+
+## USAGE
+
+See demo in **demos/simple/**.
+
+```javascript
+var words = new Lexer().lex("This is some sample text. This text can contain multiple sentences.");
+var taggedWords = new POSTagger().tag(words);
+for (i in taggedWords) {
+    var taggedWord = taggedWords[i];
+    var word = taggedWord[0];
+    var tag = taggedWord[1];
+}
+```
+
+### node.js
+
+See demo using [vows](http://vowsjs.org/) testing in **test/**
+
+You can install jspos in your project (once jsPOS is published to npm) with this command: `npm install jspos`. You can use it in your code, like this:
+
+```javascript
+var jspos = require('jspos'),
+  lexer =  new jspos.Lexer(),
+  tagger = new jspos.POSTagger();
+
+var tags = tagger.tag(lexer.lex("This is some text. It can be whatever length. Cool!"));
+console.log(tags);
+```
+
+### require.js
+
+See demo in **demos/simple_with_require/**.
+
+```javascript
+define(['jspos/lib/post_tagger', 'jspos/lib/lexer'], function(POSTagger, Lexer){
+  var lexer =  new Lexer(),
+      tagger = new POSTagger();
+
+  var tags = tagger.tag(lexer.lex("This is some text. It can be whatever length. Cool!"));
+  console.log(tags)
+});
+```
+
+### developer tools
+
+There are a few tools linked to npm, via **package.json**:
+
+* `npm install` - install all dependencies
+* `npm test`  - test the library
+* `npm start` - start the demo webserver
+* `node build.js` - creates a minified version of the library (for the web) in jspos.min.js
+
+
+## TAGS
+
+    CC    Coord Conjuncn           and,but,or
+    CD    Cardinal number          one,two
+    DT    Determiner               the,some
+    EX    Existential there        there
+    FW    Foreign Word             mon dieu
+    IN    Preposition              of,in,by
+    JJ    Adjective                big
+    JJR   Adj., comparative        bigger
+    JJS   Adj., superlative        biggest
+    LS    List item marker         1,One
+    MD    Modal                    can,should
+    NN    Noun, sing. or mass      dog
+    NNP   Proper noun, sing.       Edinburgh
+    NNPS  Proper noun, plural      Smiths
+    NNS   Noun, plural             dogs
+    POS   Possessive ending        's
+    PDT   Predeterminer            all,both
+    PP$   Possessive pronoun       my,one's
+    PRP   Personal pronoun         I,you,she
+    RB    Adverb                   quickly
+    RBR   Adverb, comparative      faster
+    RBS   Adverb, superlative      fastest
+    RP    Particle                 up,off
+    SYM   Symbol                   +,%,&
+    TO    ?to?                     to
+    UH    Interjection             oh,oops
+    VB    verb, base form          eat
+    VBD   verb, past tense         ate
+    VBG   verb, gerund             eating
+    VBN   verb, past part          eaten
+    VBP   Verb, present            eat
+    VBZ   Verb, present            eats
+    WDT   Wh-determiner            which,that
+    WP    Wh pronoun               who,what
+    WP$   Possessive-Wh            whose
+    WRB   Wh-adverb                how,where
+    ,     Comma                    ,
+    .     Sent-final punct         .,!,?
+    :     Mid-sent punct.          :,;
+    $     Dollar sign              $
+    #     Pound sign               #
+    "     quote                    "
+    (     Left paren                (
+    )     Right paren               )
+
+
+
+## LICENSE
+
+jsPOS is licensed under the GNU LGPLv3
+
+
+## ACKNOWLEDGEMENTS
+
+Thanks to Mark Watson for writing FastTag, which served as the basis for jsPOS.
+
+
+## AUTHOR
+
+[Percy Wegmann](http://www.percywegmann.com/)
+
+The orignal of this code is avallable [on google code](http://code.google.com/p/jspos/)
+
+Kieren Diment <zarquon@cpan.org> added the demo.html and main.js files.
+
+David Konsumer <konsumer@jetboystudio.com> updated demos, documentation & added npm/node/require.js support
+
+The next step is to add noun phrase extraction routines and other utility functions (see the Perl Module [Lingua::EN::Tagger](http://search.cpan.org/perldoc?Lingua::EN::Tagger)  ).

build.js

@@ -0,0 +1,25 @@
+/**
+ * Simple javascript minifier
+ * run with node build.js or npm run-script build
+ */
+
+var uglify = require('uglify-js'),
+  path = require('path');
+  fs = require('fs');
+
+var fname = 'jspos.min.js';
+
+var orig_code = "";
+orig_code += fs.readFileSync(path.join(__dirname, 'lib', 'lexicon.js'), 'utf8');
+orig_code += fs.readFileSync(path.join(__dirname, 'lib', 'lexer.js'), 'utf8');
+orig_code += fs.readFileSync(path.join(__dirname, 'lib', 'POSTagger.js'), 'utf8');
+
+var uglify_options = {
+  strict_semicolons: true,
+  mangle_options: {except: ['$super']},
+  gen_options: {ascii_only: true},
+};
+var minified_code = uglify(orig_code, uglify_options);
+
+fs.writeFileSync(path.join(__dirname, fname), minified_code, 'utf8');
+console.log("Wrote minified file to " + fname);

demo/eliza/index.html

@@ -0,0 +1,32 @@
+<html>
+  <head>
+    <title>Eliza Demo</title>
+    <style type="text/css">
+      body, input, textarea {
+        font-size: 20px;
+      }
+
+      textarea, input {
+        width: 100%;
+      }
+
+      textarea {
+        height: 500px;
+        display:block;
+      }
+    </style>
+  </head>
+
+  <body>
+    <form id="eliza">
+      <textarea id="output" disabled="disabled">This is an Eliza simulation using jspos. Press [Enter] to say something.</textarea>
+      <input id="input">
+    </form>
+  </body>
+
+  <script src="https://ajax.googleapis.com/ajax/libs/jquery/1.8.1/jquery.min.js" type="text/javascript"></script>
+  <script src="../../lib/lexer.js" type="text/javascript"></script>
+  <script src="../../lib/lexicon.js" type="text/javascript"></script>
+  <script src="../../lib/POSTagger.js" type="text/javascript"></script>
+  <script src="main.js" type="text/javascript"></script>
+</html>

demo/eliza/main.js

@@ -0,0 +1,101 @@
+/**
+   *  from prototype
+   *
+   *  Array#filter(iterator[, context]) -> Array
+   *  - iterator (Function): An iterator function to use to test the
+   *    elements.
+   *  - context (Object): An optional object to use as `this` within
+   *    calls to the iterator.
+   *
+   *  Returns a new array containing all the items in this array for which
+   *  `iterator` returned a truthy value.
+   *
+   *  `Array#filter` acts as an ECMAScript 5 [polyfill](http://remysharp.com/2010/10/08/what-is-a-polyfill/).
+   *  It is only defined if not already present in the user's browser, and it
+   *  is meant to behave like the native version as much as possible. Consult
+   *  the [ES5 specification](http://es5.github.com/#x15.4.4.20) for more
+   *  information.
+  **/
+if (!Array.prototype.filter){
+  Array.prototype.filter = function(iterator) {
+    var object = Object(this);
+    var results = [], context = arguments[1], value;
+
+    for (var i = 0, length = object.length >>> 0; i < length; i++) {
+      if (i in object) {
+        value = object[i];
+        if (iterator.call(context, value, i, object)) {
+          results.push(value);
+        }
+      }
+    }
+    return results;
+  }
+}
+
+$(function(){
+  var lexer = new Lexer();
+  var tagger =  new POSTagger();
+
+  // talk data
+  var data = {
+    "init": [
+      "How do you do.  Please tell me your problem.",
+      "Please tell me what's been bothering you.",
+      "Is something troubling you?"
+    ],
+
+    "synons" : {
+      "be": ["am", "is", "are", "was"],
+      "belief": ["feel", "think", "believe", "wish"],
+      "cannot": ["can't"],
+      "desire": ["want", "need"],
+      "everyone": ["everybody", "nobody", "noone"],
+      "family": ["mother", "mom", "father", "dad", "sister", "brother", "wife", "children", "child"],
+      "happy": ["elated", "glad", "better"],
+      "sad": ["unhappy", "depressed", "sick"]
+    },
+
+    "sorry":[
+       "Please don't apologise.",
+       "Apologies are not necessary.",
+       "I've told you that apologies are not required.",
+       "It did not bother me.  Please continue."
+     ]
+  };
+
+  function randq(name){
+    var q = data[name];
+    return q[Math.floor(Math.random()*q.length)];
+  }
+
+  $('#eliza').submit(function(){
+    var v = $('#input').val(),
+      out="";
+
+    $('#output').append("YOU: " + v + "\n\n");
+
+    var good_nouns = [];
+
+    if (v.match(/\w+/) !== null){
+      var words = tagger.tag(lexer.lex(v));
+      console.log("words", words);
+      // find the most relevant noun/pronouns
+      var subjects = words.filter(function(s, i, a){ s.push(i); return (s[1] == 'PP$' || s[1] == 'PRP$' || s[1][0] == 'N'); });
+      console.log("subjects", subjects);
+
+    }
+
+    if (out.length){
+      $('#output').append("ELIZA: " + out +"\n\n");
+    }
+
+    return false;
+  });
+
+  $('#input').focus();
+  $('#output').append("\n\nELIZA: " + randq('init') + "\n\n");
+
+
+
+});