I simplified the test cases. The test is now making sure that the language is detected correctly for each file and instead of checking the language ratio we're now checking the confidence score

Author: gignupg

src/components/processContent.js

@@ -20,10 +20,6 @@ module.exports = (data, fileInfo) => {
 
   const calculations = calculateConfidenceScore(data, fileInfo);
 
-  if (data.testFile) {
-    return calculations;
-  }
-
   if (fileInfo.confidence.encoding) {
     fileInfo.confidence.language = calculations;
   } else {

src/components/processing-content/calculateConfidenceScore.js

@@ -6,7 +6,6 @@ module.exports = (data, fileInfo) => {
   const totalCharacters = data.content.replace(charRegex, "").length;
   const langArr = data.languageArr;
   const pos = data.pos;
-  const testFile = data.testFile;
 
   const secondLanguage = langArr.reduce((acc, val) => {
     if (acc.name === fileInfo.language) return val;
@@ -57,24 +56,5 @@ module.exports = (data, fileInfo) => {
     );
   }
 
-  // If the test script is running
-  if (testFile) {
-    return {
-      name: testFile.substr(testFile.lastIndexOf("/") + 1),
-      path: testFile,
-      encoding: fileInfo.encoding,
-      language: fileInfo.language,
-      languageConfidence: confidenceScore,
-      ratio: Number(languageRatio.toFixed(2)),
-      count: langArr[pos].count,
-      totalCharacters: totalCharacters,
-      characterWordRatio: characterWordRatio.toFixed(6),
-      secondLanguage: {
-        name: secondLanguage.name,
-        count: secondLanguage.count,
-      },
-    };
-  }
-
   return confidenceScore;
 };

src/index-node.js

@@ -3,7 +3,7 @@ const checkUTF = require("./components/checkUTF.js");
 const processContent = require("./components/processContent.js");
 const checkByteOrderMark = require("./components/checkByteOrderMark.js");
 
-module.exports = (file, test) => {
+module.exports = (file) => {
   return new Promise((resolve, reject) => {
     const fileInfo = {
       encoding: null,
@@ -14,7 +14,6 @@ module.exports = (file, test) => {
       },
     };
     const data = {};
-    data.testFile = test ? file : null;
 
     // Reading the first four bytes and checking if they coincide with one of the predefined byte order marks.
     const readStream = fs.createReadStream(file, { start: 0, end: 3 });

testing/language-encoding.test.js

@@ -12,11 +12,44 @@ const folderPath = "/home/gignu/Documents/Subtitle Database/Language Folders/";
 const testFiles = getFiles(folderPath);
 
 testFiles.forEach((file) => {
-  languageEncoding(file, true)
+  languageEncoding(file)
     .then((fileInfo) => {
-      if (fileInfo.ratio <= 0.85) {
+      // language = language
+      const testFileArray = file.split("/");
+      const expectedLanguage = testFileArray[testFileArray.length - 2]
+        .toLowerCase()
+        .replace(" ", "-");
+
+      if (
+        fileInfo.language !== expectedLanguage &&
+        expectedLanguage !== "japanese"
+      )
+        testFailed("language");
+
+      // confidence >= 0.95
+      if (
+        fileInfo.confidence.encoding < 0.95 &&
+        expectedLanguage !== "japanese"
+      )
+        testFailed("confidence");
+
+      function testFailed(issue) {
         console.log("Test case failed:");
+
+        switch (issue) {
+          case "language":
+            console.log("Expected language:", expectedLanguage);
+            console.log("Detected language:", fileInfo.language);
+          case "confidence":
+            console.log("Confidence score too low!");
+            console.log(
+              "fileInfo.confidence.encoding:",
+              fileInfo.confidence.encoding
+            );
+        }
+
         console.log(fileInfo);
+        console.log("file:", file);
         process.exit(1);
       }
     })