nlpsum/nlp-node-master/tag/helpers/coffeejs/chunker.js

Powerful text summarization algorithms from research papers and dedicated research.

// Generated by CoffeeScript 1.6.3
var chunker, data;

chunker = (function() {
  var build;
  build = function(before, obj, after) {
    var all, i;
    all = before;
    all.push(obj);
    for (i in after) {
      all.push(after[i]);
    }
    return all;
  };
  chunker = function(data, options) {
    var chunked, i;
    for (i in data) {
      i = parseInt(i);
      if (data[i].pos.tag === "NNO") {
        data = build(data.slice(0, i + 1), {
          word: "",
          pos: parts_of_speech["IN"]
        }, data.slice(i + 1, data.length));
      }
      if (data[i].pos.tag === "VBG" && data[i + 2]) {
        if (data[i + 1].pos.tag === "IN" && data[i + 2].pos.parent === "glue") {
          data[i].word += " " + data[i + 1].word;
          data[i + 1].word = null;
          continue;
        }
      }
      if (data[i].pos.tag === "VBG") {
        if (data[i + 1] && (data[i + 2] == null) && data[i + 1].pos.tag === "IN") {
          data[i].word += " " + data[i + 1].word;
          data[i + 1].word = null;
        }
      }
    }
    if (options.gerund) {
      for (i in data) {
        if (data[i].pos.tag === "VBG") {
          data[i].pos = parts_of_speech["NG"];
        }
      }
    }
    chunked = [data[0]];
    i = 1;
    while (i <= data.length - 1) {
      console.log(i);
      if (chunked[chunked.length - 1].pos.parent === data[i].pos.parent) {
        if (chunked[chunked.length - 1].word && !chunked[chunked.length - 1].word.match(/(,|")/)) {
          chunked[chunked.length - 1].word += " " + data[i].word;
          continue;
        }
      }
      chunked.push(data[i]);
      i++;
    }
    return chunked;
  };
  if (typeof define !== "undefined" && define.amd) {
    define([], function() {
      return chunker;
    });
  } else {
    if (typeof module !== "undefined" && module.exports) {
      module.exports = chunker;
    }
  }
  return chunker;
})();

data = [
  {
    word: "sally",
    pos: {
      description: "verb, base form",
      example: "eat",
      parent: "verb",
      tag: "VB"
    },
    clues: [],
    rule: "lexicon"
  }, {
    word: "walked",
    pos: {
      description: "verb, past tense",
      example: "ate",
      parent: "verb",
      tense: "past",
      tag: "VBD"
    },
    clues: [],
    rule: "regex"
  }, {
    word: "to",
    pos: {
      description: "to",
      example: "to",
      parent: "glue",
      tag: "TO"
    },
    clues: [],
    rule: "lexicon"
  }, {
    word: "the",
    pos: {
      description: "Determiner",
      example: "the,some",
      parent: "title",
      tag: "DT"
    },
    clues: [],
    rule: "lexicon"
  }, {
    word: "store",
    pos: {
      description: "Noun, sing. or mass",
      example: "dog, rain",
      parent: "noun",
      tag: "NN"
    },
    clues: [],
    rule: "unknown"
  }
];

console.log(chunker(data, {}));