MBS-11854: Recognize unicode hyphen in guess case (#2199)

Words with a unicode hyphen were not being recognized as split words: the hyphen was just being considered one more character. This changes that, and then ensures that guess case recognizes the hyphen as such and does the same as with a hyphen-minus. Sadly, part of my fix for MBS-10156 (for re‐mode with unicode hyphen) will no longer work now since it is no longer considered one single word, but I think this makes a lot more sense for now. The current code already didn't do anything with re-mode with a hyphen-minus, so this actually adds consistency.
metabrainz · Aug 12, 2021 · 61ee8af · 61ee8af
1 parent 6e3e634
commit 61ee8af
Show file tree

Hide file tree

Showing 4 changed files with 14 additions and 7 deletions.
diff --git a/root/static/scripts/guess-case/MB/GuessCase/Handler/Base.js b/root/static/scripts/guess-case/MB/GuessCase/Handler/Base.js
@@ -117,7 +117,7 @@ MB.GuessCase.Handler.Base = function (gc) {
        */
       var handled = false;
       if (!gc.regexes.SPECIALCASES) {
-        gc.regexes.SPECIALCASES = /(&|¿|¡|\?|\!|;|:|'|‘|’|‹|›|"|“|”|„|“|«|»|\-|\+|,|\*|\.|#|%|\/|\(|\)|\{|\}|\[|\])/;
+        gc.regexes.SPECIALCASES = /(&|¿|¡|\?|\!|;|:|'|‘|’|‹|›|"|“|”|„|“|«|»|\-|‐|\+|,|\*|\.|#|%|\/|\(|\)|\{|\}|\[|\])/;
       }
       if (input.matchCurrentWord(gc.regexes.SPECIALCASES)) {
         handled = !!(
@@ -332,7 +332,7 @@ MB.GuessCase.Handler.Base = function (gc) {
    */
   self.doHyphen = function () {
     if (!gc.regexes.HYPHEN) {
-      gc.regexes.HYPHEN = '-';
+      gc.regexes.HYPHEN = /^[\-‐]$/;
     }
     if (input.matchCurrentWord(gc.regexes.HYPHEN)) {
       output.appendWordPreserveWhiteSpace(true);

diff --git a/root/static/scripts/guess-case/MB/GuessCase/Input.js b/root/static/scripts/guess-case/MB/GuessCase/Input.js
@@ -182,7 +182,7 @@ class GuessCaseInput {
     const splitwords = [];
     let word = [];
     if (!gc.regexes.SPLITWORDSANDPUNCTUATION) {
-      gc.regexes.SPLITWORDSANDPUNCTUATION = /[^!¿¡\"%&'´`‘’‹›“”„“«»()\[\]\{\}\*\+,-\.\/:;<=>\?\s#]/;
+      gc.regexes.SPLITWORDSANDPUNCTUATION = /[^!¿¡\"%&'´`‘’‹›“”„“«»()\[\]\{\}\*\+‐\-,\.\/:;<=>\?\s#]/;
     }
     for (let i = 0; i < chars.length; i++) {
       if (chars[i].match(gc.regexes.SPLITWORDSANDPUNCTUATION)) {

diff --git a/root/static/scripts/guess-case/utils.js b/root/static/scripts/guess-case/utils.js
@@ -61,7 +61,6 @@ const preBracketSingleWordsList = [
   'rehearsal',
   'remixed',
   'remode',
-  're‐mode',
   'rework',
   'reworked',
   'session',

diff --git a/root/static/scripts/tests/GuessCase.js b/root/static/scripts/tests/GuessCase.js
@@ -175,7 +175,7 @@ test('Recording', function (t) {
 });
 
 test('Work', function (t) {
-  t.plan(23);
+  t.plan(24);
 
   const tests = [
     {
@@ -344,6 +344,14 @@ test('Work', function (t) {
       roman: false,
       keepuppercase: false,
     },
+    {
+      input: 'hyphen-minus? hyphen‐maximus!',
+      expected: 'Hyphen-Minus? Hyphen‐Maximus!',
+      bug: 'MBS-11854',
+      mode: 'English',
+      roman: false,
+      keepuppercase: false,
+    },
   ];
 
   for (const test of tests) {
@@ -457,8 +465,8 @@ test('BugFixes', function (t) {
       mode: 'French',
     },
     {
-      input: 'We Love Techno (Re‐Mode)',
-      expected: 'We Love Techno (re‐mode)',
+      input: 'We Love Techno (Remode)',
+      expected: 'We Love Techno (remode)',
       bug: 'MBS-10156',
       mode: 'English',
     },