fix for #801

README.md

@@ -826,7 +826,7 @@ or if you don't care about POS-tagging, you can use the tokenize-only build: (90
       <summary>✨ Partial builds?</summary>
       <p></p>
       <ul>
-        we do offer a [compromise-tokenize](./builds/compromise-tokenize.js) build, which has the POS-tagger pulled-out.
+        we do offer a <a href="./builds/compromise-tokenize.js">compromise-tokenize</a> build, which has the POS-tagger pulled-out.
         <br/> 
         but otherwise, compromise isn't easily tree-shaken.
         <br/> 

scratch.js

@@ -1,5 +1,6 @@
 const nlp = require('./src/index')
 nlp.extend(require('./plugins/sentences/src'))
+nlp.extend(require('./plugins/penn-tags/src'))
 // nlp.verbose(true)
 
 //
@@ -15,7 +16,7 @@ nlp.extend(require('./plugins/sentences/src'))
 // doc.match(reg).debug()
 
 // #801
-// nlp('79-years-old').debug()
+nlp('79-years-old').debug()
 // nlp('foo-bar').match('@hasDash').debug()
 
 // #802

src/01-tokenizer/02-words.js

@@ -20,7 +20,7 @@ const hasHyphen = function (str) {
     return true
   }
   //number-letter '20-aug'
-  let reg2 = /^([0-9]{1,4})(-|–|—)([a-z\u00C0-\u00FF`"'/]+$)/i
+  let reg2 = /^([0-9]{1,4})(-|–|—)([a-z\u00C0-\u00FF`"'/-]+$)/i
   if (reg2.test(str) === true) {
     return true
   }

tests/_pennSample.js

@@ -1151,14 +1151,6 @@ module.exports = [
     text: 'It should be 4 - easy.',
     tags: 'PRP, MD, VB, CD, JJ',
   },
-  // {
-  //   text: '1579-EBS Network Co. Division of 17H.',
-  //   tags: 'NNP, NNP, NNP, NN, IN, NN',
-  // },
-  {
-    text: '1691-EPI-EBS Europe',
-    tags: 'NNP, NNP',
-  },
   {
     text: 'Set up last month to centralize merchant asset activities.',
     tags: 'VBN, RP, JJ, NN, TO, VB, NN, NN, NNS',
@@ -1167,14 +1159,6 @@ module.exports = [
     text: 'Broke out the activities of 1179.',
     tags: 'VBD, RP, DT, NNS, IN, CD',
   },
-  {
-    text: '1307-EBIC-Apache, LLC',
-    tags: 'NNP, NNP',
-  },
-  {
-    text: '1689-EPI-EBS Ventures, LLC',
-    tags: 'NNP, NNPS, NNP',
-  },
   {
     text: 'Set up last month to centralize merchant asset activities.',
     tags: 'VBN, RP, JJ, NN, TO, VB, NN, NN, NNS',

tests/tokenize.test.js

@@ -27,6 +27,10 @@ test('em-dash, en-dash', function (t) {
   t.equal(doc.terms().length, 3, 'en-dash-num')
   doc = nlp('20—20')
   t.equal(doc.terms().length, 3, 'em-dash-num')
+
+  doc = nlp('79-years-old')
+  t.equal(doc.terms().length, 3, 'x-years-old')
+
   t.end()
 })