Also detect windows-1250 every time windows-1252 is detected

This is in relation to #70. The differences between the two are minimal, so this will be a workaround for now. These two encodings use very different models for detection. The windows-1252 detector is purely based on the occurance probability of each character's class. The windows-1250 uses a Hungarian language model to detect the text is in Hungarian. This is brittle as there are other languages using windows-1250.

src/universaldetector.js

@@ -235,14 +235,35 @@ function UniversalDetector(options) {
             return this.result;
         }
 
-        if( this._mInputState == _state.highbyte ) {
-            for( var i = 0, prober; prober = this._mCharsetProbers[i]; i++ ) {
-                if( !prober || !prober.getCharsetName() || !canDetectEncoding(prober.getCharsetName()) ) continue;
+        if (this._mInputState == _state.highbyte) {
+            let windows_1252_confidence = 0;
+            let windows_1250_detected = false;
+            for (var i = 0, prober; prober = this._mCharsetProbers[i]; i++) {
+                if (!prober) continue;
+                const charsetName = prober.getCharsetName();
+                const confidence = prober.getConfidence();
+                if (prober.getCharsetName() === "windows-1252") {
+                    windows_1252_confidence = confidence;
+                }
+                if (!charsetName || !canDetectEncoding(charsetName)) continue;
                 this.results.push({
                     "encoding": prober.getCharsetName(),
-                    "confidence": prober.getConfidence()
+                    "confidence": confidence
+                });
+                if (prober.getCharsetName() === "windows-1250") {
+                    windows_1250_detected = true;
+                }
+                logger.log(prober.getCharsetName() + " confidence " + confidence);
+            }
+            // HACK: When windows-1252 is detected it's almost sure that it can
+            // also be windows-1250.
+            // https://en.wikipedia.org/wiki/Windows-1250 (Central European)
+            if (windows_1252_confidence && !windows_1250_detected && canDetectEncoding("windows-1250")) {
+                this.results.push({
+                    "encoding": "windows-1250",
+                    // Report the confidence just a bit under windows-1252's.
+                    "confidence": windows_1252_confidence - 5/10**(String(windows_1252_confidence).length - 1),
                 });
-                logger.log(prober.getCharsetName() + " confidence " + prober.getConfidence());
             }
             this.results.sort(function(a, b) {
                 return b.confidence - a.confidence;

tests/detectEncodings.test.js

@@ -50,3 +50,13 @@ test('detectEncodings locks down which encodings to detect (SHIFT_JIS)', async (
     });
     expect(shortSingleEncoding.encoding).toBe("SHIFT_JIS")
 });
+
+test('detectEncodings should also report windows-1250 when it detects windows-1252', async () => {
+    const fixturePath = `${__dirname}/fixtures/windows-1250.txt`;
+    fileContents = await utils.readFileAsBuffer(fixturePath);
+    const possibleEncodings = jschardet.detectAll(fileContents, {
+        detectEncodings: ["windows-1250"],
+    });
+    expect(possibleEncodings.length).toBe(1);
+    expect(possibleEncodings[0].encoding).toBe("windows-1250")
+});

tests/fixtures/windows-1250.txt

@@ -0,0 +1 @@
+P��li� �lu�ou�k� k�� �p�l ��belsk� �dy