Add experimental transliterating hint matcher that I apparently wrote for Conkeror. (resume matches résume, mae mæstro, and so on (but résume and mæstro don't...))

2025-12-20 18:17:58 +01:00 · 2009-11-11 08:46:18 -05:00
parent ee91adc5ad
commit 2f8369ddf7
2 changed files with 148 additions and 5 deletions
--- a/common/content/hints.js
+++ b/common/content/hints.js
@@ -570,7 +570,7 @@ const Hints = Module("hints", {
            let tokens = tokenize(/\s+/, hintString);
            return function (linkText) {
                linkText = linkText.toLowerCase();
-                return tokens.every(function (token) linkText.indexOf(token) >= 0);
+                return tokens.every(function (token) indexOf(linkText, token) >= 0);
            };
        } //}}}

@@ -667,7 +667,7 @@ const Hints = Module("hints", {
                        continue;

                    let str = strings[strIdx];
-                    if (str.length == 0 || word.indexOf(str) == 0)
+                    if (str.length == 0 || indexOf(word, str) == 0)
                        strIdx++;
                    else if (!allowWordOverleaping)
                        return false;
@@ -695,7 +695,11 @@ const Hints = Module("hints", {
            };
        } //}}}

-        switch (options["hintmatching"]) {
+        let indexOf = String.indexOf;
+        if (options.get("hintmatching").has("transliterated"))
+            indexOf = Hints.indexOf;
+
+        switch (options.get("hintmatching").values[0]) {
        case "contains"      : return containsMatcher(hintString);
        case "wordstartswith": return wordStartsWithMatcher(hintString, /*allowWordOverleaping=*/ true);
        case "firstletters"  : return wordStartsWithMatcher(hintString, /*allowWordOverleaping=*/ false);
@@ -882,6 +886,142 @@ const Hints = Module("hints", {

    //}}}
 }, {
+    indexOf: (function () {
+        const table = [
+            [0x00c0, 0x00c6, ["A"]],
+            [0x00c7, 0x00c7, ["C"]],
+            [0x00c8, 0x00cb, ["E"]],
+            [0x00cc, 0x00cf, ["I"]],
+            [0x00d1, 0x00d1, ["N"]],
+            [0x00d2, 0x00d6, ["O"]],
+            [0x00d8, 0x00d8, ["O"]],
+            [0x00d9, 0x00dc, ["U"]],
+            [0x00dd, 0x00dd, ["Y"]],
+            [0x00e0, 0x00e6, ["a"]],
+            [0x00e7, 0x00e7, ["c"]],
+            [0x00e8, 0x00eb, ["e"]],
+            [0x00ec, 0x00ef, ["i"]],
+            [0x00f1, 0x00f1, ["n"]],
+            [0x00f2, 0x00f6, ["o"]],
+            [0x00f8, 0x00f8, ["o"]],
+            [0x00f9, 0x00fc, ["u"]],
+            [0x00fd, 0x00fd, ["y"]],
+            [0x00ff, 0x00ff, ["y"]],
+            [0x0100, 0x0105, ["A", "a"]],
+            [0x0106, 0x010d, ["C", "c"]],
+            [0x010e, 0x0111, ["D", "d"]],
+            [0x0112, 0x011b, ["E", "e"]],
+            [0x011c, 0x0123, ["G", "g"]],
+            [0x0124, 0x0127, ["H", "h"]],
+            [0x0128, 0x0130, ["I", "i"]],
+            [0x0132, 0x0133, ["IJ", "ij"]],
+            [0x0134, 0x0135, ["J", "j"]],
+            [0x0136, 0x0136, ["K", "k"]],
+            [0x0139, 0x0142, ["L", "l"]],
+            [0x0143, 0x0148, ["N", "n"]],
+            [0x0149, 0x0149, ["n"]],
+            [0x014c, 0x0151, ["O", "o"]],
+            [0x0152, 0x0153, ["OE", "oe"]],
+            [0x0154, 0x0159, ["R", "r"]],
+            [0x015a, 0x0161, ["S", "s"]],
+            [0x0162, 0x0167, ["T", "t"]],
+            [0x0168, 0x0173, ["U", "u"]],
+            [0x0174, 0x0175, ["W", "w"]],
+            [0x0176, 0x0178, ["Y", "y", "Y"]],
+            [0x0179, 0x017e, ["Z", "z"]],
+            [0x0180, 0x0183, ["b", "B", "B", "b"]],
+            [0x0187, 0x0188, ["C", "c"]],
+            [0x0189, 0x0189, ["D"]],
+            [0x018a, 0x0192, ["D", "D", "d", "F", "f"]],
+            [0x0193, 0x0194, ["G"]],
+            [0x0197, 0x019b, ["I", "K", "k", "l", "l"]],
+            [0x019d, 0x01a1, ["N", "n", "O", "O", "o"]],
+            [0x01a4, 0x01a5, ["P", "p"]],
+            [0x01ab, 0x01ab, ["t"]],
+            [0x01ac, 0x01b0, ["T", "t", "T", "U", "u"]],
+            [0x01b2, 0x01d2, ["V", "Y", "y", "Z", "z", "D", "L", "N", "A", "a", "I", "i", "O", "o"]],
+            [0x01d3, 0x01dc, ["U", "u"]],
+            [0x01de, 0x01e1, ["A", "a"]],
+            [0x01e2, 0x01e3, ["AE", "ae"]],
+            [0x01e4, 0x01ed, ["G", "g", "G", "g", "K", "k", "O", "o", "O", "o"]],
+            [0x01f0, 0x01f5, ["j", "D", "G", "g"]],
+            [0x01fa, 0x01fb, ["A", "a"]],
+            [0x01fc, 0x01fd, ["AE", "ae"]],
+            [0x01fe, 0x0217, ["O", "o", "A", "a", "A", "a", "E", "e", "E", "e", "I", "i", "I", "i", "O", "o", "O", "o", "R", "r", "R", "r", "U", "u", "U", "u"]],
+            [0x0253, 0x0257, ["b", "c", "d", "d"]],
+            [0x0260, 0x0269, ["g", "h", "h", "i", "i"]],
+            [0x026b, 0x0273, ["l", "l", "l", "l", "m", "n", "n"]],
+            [0x027c, 0x028b, ["r", "r", "r", "r", "s", "t", "u", "u", "v"]],
+            [0x0290, 0x0291, ["z"]],
+            [0x029d, 0x02a0, ["j", "q"]],
+            [0x1e00, 0x1e09, ["A", "a", "B", "b", "B", "b", "B", "b", "C", "c"]],
+            [0x1e0a, 0x1e13, ["D", "d"]],
+            [0x1e14, 0x1e1d, ["E", "e"]],
+            [0x1e1e, 0x1e21, ["F", "f", "G", "g"]],
+            [0x1e22, 0x1e2b, ["H", "h"]],
+            [0x1e2c, 0x1e8f, ["I", "i", "I", "i", "K", "k", "K", "k", "K", "k", "L", "l", "L", "l", "L", "l", "L", "l", "M", "m", "M", "m", "M", "m", "N", "n", "N", "n", "N", "n", "N", "n", "O", "o", "O", "o", "O", "o", "O", "o", "P", "p", "P", "p", "R", "r", "R", "r", "R", "r", "R", "r", "S", "s", "S", "s", "S", "s", "S", "s", "S", "s", "T", "t", "T", "t", "T", "t", "T", "t", "U", "u", "U", "u", "U", "u", "U", "u", "U", "u", "V", "v", "V", "v", "W", "w", "W", "w", "W", "w", "W", "w", "W", "w", "X", "x", "X", "x", "Y", "y"]],
+            [0x1e90, 0x1e9a, ["Z", "z", "Z", "z", "Z", "z", "h", "t", "w", "y", "a"]],
+            [0x1ea0, 0x1eb7, ["A", "a"]],
+            [0x1eb8, 0x1ec7, ["E", "e"]],
+            [0x1ec8, 0x1ecb, ["I", "i"]],
+            [0x1ecc, 0x1ee3, ["O", "o"]],
+            [0x1ee4, 0x1ef1, ["U", "u"]],
+            [0x1ef2, 0x1ef9, ["Y", "y"]],
+            [0x2071, 0x2071, ["i"]],
+            [0x207f, 0x207f, ["n"]],
+            [0x249c, 0x24b5, "a"],
+            [0x24b6, 0x24cf, "A"],
+            [0x24d0, 0x24e9, "a"],
+            [0xfb00, 0xfb06, ["ff", "fi", "fl", "ffi", "ffl", "st", "st"]],
+            [0xff21, 0xff3a, "A"],
+            [0xff41, 0xff5a, "a"],
+        ].map(function (a) {
+            if (typeof a[2] == "string")
+                a[3] = function (chr) String.fromCharCode(this[2].charCodeAt(0) + chr - this[0])
+            else
+                a[3] = function (chr) this[2][(chr - this[0]) % this[3].length];
+            return a;
+        });
+
+        function translate(chr) {
+            var m, c = chr.charCodeAt(0);
+            var n = table.length;
+            var i = 0;
+            while(n) {
+                m = Math.floor(n / 2);
+                var t = table[i + m];
+                if(c >= t[0] && c <= t[1])
+                    return t[3](c)
+                if(c < t[0] || m == 0)
+                    n = m;
+                else {
+                    i += m;
+                    n = n - m;
+                }
+            }
+            return chr;
+        }
+
+        return function indexOf(dest, src) {
+            var end = dest.length - src.length;
+            if (src.length == 0)
+                return 0;
+        outer:
+            for (var i=0; i < end; i++) {
+                    var j = i;
+                    for (var k=0; k < src.length;) {
+                        var s = translate(dest[j++]);
+                        for (var l=0; l < s.length; l++, k++) {
+                            if (s[l] != src[k])
+                                continue outer;
+                            if (k == src.length - 1)
+                                return i;
+                        }
+                    }
+                }
+            return -1;
+        }
+    })(),
    Mode: new Struct("prompt", "action", "tags"),
 }, {
    mappings: function () {
@@ -966,13 +1106,14 @@ const Hints = Module("hints", {

        options.add(["hintmatching", "hm"],
            "How links are matched",
-            "string", "contains",
+            "stringlist", "contains",
            {
                completer: function (context) [
                    ["contains",       "The typed characters are split on whitespace. The resulting groups must all appear in the hint."],
                    ["wordstartswith", "The typed characters are split on whitespace. The resulting groups must all match the beginings of words, in order."],
                    ["firstletters",   "Behaves like wordstartswith, but all groups much match a sequence of words."],
-                    ["custom",         "Delegate to a custom function: liberator.plugins.customHintMatcher(this._hintString)"]
+                    ["custom",         "Delegate to a custom function: liberator.plugins.customHintMatcher(hintString)"],
+                    ["transliterated", "When true, special latin characters are translated to their ascii equivalent (e.g., \u00e9 -> e)"],
                ],
                validator: Option.validateCompleter
            });
--- a/common/content/liberator.js
+++ b/common/content/liberator.js
@@ -882,6 +882,8 @@ const Liberator = Module("liberator", {
        catch (e) {
            if (e instanceof FailedAssertion)
                liberator.echoerr(e.message);
+            else
+                liberator.reportError(e);
        }
    },