Upstream spencermountain/compromise state of the touched files before 0866b22381

src/1-one/match/methods/parseMatch/02-parseToken.js

@@ -31,7 +31,7 @@ const stripBoth = function (str) {
   return str
 }
 //
-const parseToken = function (w, opts, aliases) {
+const parseToken = function (w, opts) {
   const obj = {}
   //collect any flags (do it twice)
   for (let i = 0; i < 2; i += 1) {
@@ -73,7 +73,7 @@ const parseToken = function (w, opts, aliases) {
       }
     }
     //back-flags
-    if (end(w) === '+' && !w.endsWith('\\+')) {
+    if (end(w) === '+') {
       obj.greedy = true
       w = stripEnd(w)
     }
@@ -98,7 +98,7 @@ const parseToken = function (w, opts, aliases) {
     }
 
     //regex
-    if (start(w) === '/' && end(w) === '/' && w.length > 2) {
+    if (start(w) === '/' && end(w) === '/') {
       w = stripBoth(w)
       if (opts.caseSensitive) {
         obj.use = 'text'
@@ -156,7 +156,7 @@ const parseToken = function (w, opts, aliases) {
       obj.choices = obj.choices.filter(s => s)
       //recursion alert!
       obj.choices = obj.choices.map(str => {
-        return str.split(/ /g).map(s => parseToken(s, opts, aliases))
+        return str.split(/ /g).map(s => parseToken(s, opts))
       })
       w = ''
     }
@@ -197,12 +197,8 @@ const parseToken = function (w, opts, aliases) {
   }
   //do the actual token content
   if (start(w) === '#') {
-    const name = stripStart(w)
-    obj.tag = aliases?.[name]
-    if (!obj.tag) {
-      obj.tag = titleCase(name)
-      obj.tag = aliases?.[obj.tag] || obj.tag
-    }
+    obj.tag = stripStart(w)
+    obj.tag = titleCase(obj.tag)
     return obj
   }
   //dynamic function on a term object
@@ -225,8 +221,6 @@ const parseToken = function (w, opts, aliases) {
     //somehow handle encoded-chars?
     w = w.replace('\\*', '*')
     w = w.replace('\\.', '.')
-    w = w.replace('\\+', '+')
-    w = w.replace('\\/', '/')
     if (opts.caseSensitive) {
       obj.use = 'text'
     } else {

tests/one/match/encoding.test.js

@@ -0,0 +1,64 @@
+import test from 'tape'
+import nlp from '../_lib.js'
+const here = '[one/encoding] '
+
+// test('encoding-match:', function (t) {
+//   let r = nlp('it is * nice')
+//   let str = r.match('is \\*').out().trim()
+//   t.equal(str, 'is *', 'encode asterix')
+
+//   r = nlp('it is + nice');
+//   str = r.match('is \\+ nice').trim().out();
+//   t.equal(str, 'is + nice', 'encode plus');
+
+//   t.end()
+// })
+
+test('reserved-word-in-src:', function (t) {
+  const r = nlp('buy eggs constructor yeah prototype')
+  t.equal(r.has(`backburner`), false, here + 'single')
+  t.equal(r.has(`#Foo`), false, here + 'tag')
+  t.equal(r.has(`(upcoming|backburner)`), false, here + 'anyOf')
+  t.equal(r.has(`lala+`), false, here + 'manyOf')
+  t.equal(r.has(`nword{2,4}`), false, here + 'someOf')
+  t.end()
+})
+
+test('reserved-word-in-match:', function (t) {
+  const r = nlp('fo foo fulala repeat')
+  t.equal(r.has(`constructor`), false, here + 'single')
+  t.equal(r.has(`#constructor`), false, here + 'tag')
+  t.equal(r.has(`(upcoming|constructor)`), false, here + 'anyOf')
+  t.equal(r.has(`constructor+`), false, here + 'manyOf')
+  t.equal(r.has(`constructor{2,4}`), false, here + 'someOf')
+  t.end()
+})
+
+test('test-infinite-loop', function (t) {
+  const weirdDoc = nlp('^ ? * . + $')
+  weirdDoc.match('is?')
+  weirdDoc.match('.?')
+  weirdDoc.match('*')
+  weirdDoc.match('.+')
+  weirdDoc.match('+')
+  weirdDoc.match('?')
+  weirdDoc.match('.')
+  weirdDoc.match('? * . +')
+  weirdDoc.not('?')
+  weirdDoc.not('*')
+  weirdDoc.not('^')
+  weirdDoc.not('$')
+  weirdDoc.not('+')
+  weirdDoc.not('? * . +')
+  t.ok(true, here + 'didnt regress')
+
+  let str = 'And you are?. Marshal'
+  let have = nlp(str).all().out()
+  t.equal(have, str, here + 'regression #1')
+
+  str = `- where is she.Oh.  you guys don't know?`
+  have = nlp(str).all().out()
+  t.equal(have, str, here + 'regression #2')
+
+  t.end()
+})