Merge pull request #25 from JustaFrank/master

Added text difference finder
ykdojo · Nov 14, 2018 · ce1d4e6 · ce1d4e6
2 parents 25b73f9 + 93f1680
commit ce1d4e6
Show file tree

Hide file tree

Showing 4 changed files with 157 additions and 0 deletions.
diff --git a/editdojo_project/settings.py b/editdojo_project/settings.py
@@ -132,3 +132,8 @@
 # https://docs.djangoproject.com/en/2.1/howto/static-files/
 
 STATIC_URL = '/static/'
+
+STATICFILES_DIRS = [
+    os.path.join(BASE_DIR, "static"),
+    '/var/www/static/',
+]
diff --git a/static/css/text_difference.css b/static/css/text_difference.css
@@ -0,0 +1,6 @@
+.deleted {
+  background-color: rgb(255, 200, 200);
+}
+.added {
+  background-color: rgb(200, 255, 200);
+}
diff --git a/static/demo.html b/static/demo.html
@@ -0,0 +1,45 @@
+<!DOCTYPE html>
+<html lang="en">
+
+<head>
+  <meta charset="UTF-8">
+  <meta name="viewport" content="width=device-width, initial-scale=1.0">
+  <meta http-equiv="X-UA-Compatible" content="ie=edge">
+  <title>Document</title>
+  <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/materialize/1.0.0/css/materialize.min.css">
+  <link type="text/css" rel="stylesheet" href="css/text_difference.css">
+</head>
+
+<body>
+  <div class="container">
+    <h1>text difference finder</h1>
+    <form id="textForm" method="submit">
+      <label for="original">Original Text</label>
+      <input name="original" type="text">
+      <label for="changed">Changed Text</label>
+      <input name="changed" type="text">
+      <button class="btn waves-effect waves-light" type="submit" name="action">Submit</button>
+    </form>
+    <p id="difference"></p>
+    <p id="deletions"></p>
+    <p id="additions"></p>
+  </div>
+
+  <script src="js/text_difference.js"></script>
+  <script>
+    let form = document.getElementById("textForm");
+    form.addEventListener("submit", event => {
+      event.preventDefault();
+      let formData = new FormData(form);
+      let originalText = formData.get("original");
+      let changedText = formData.get("changed");
+      let [dif, del, add] = findTextDifference(originalText, changedText);
+      document.getElementById("difference").innerHTML = "<b>Total Difference</b><br>" + dif;
+      document.getElementById("deletions").innerHTML = "<b>Deletions</b><br>" + del;
+      document.getElementById("additions").innerHTML = "<b>Additions</b><br>" + add;
+    });
+  </script>
+  <script src="https://cdnjs.cloudflare.com/ajax/libs/materialize/1.0.0/js/materialize.min.js"></script>
+</body>
+
+</html>
diff --git a/static/js/text_difference.js b/static/js/text_difference.js
@@ -0,0 +1,101 @@
+/**
+ * Text difference module
+ * Relevant functions:
+ * @method longestCommonSubsequence - used to find common phrases between 2 pieces of text.
+ * @method getDifferenceStrings - used to construct a string to disply in HTML highlighting differences.
+ */
+
+// Returns difference strings from 2 pieces of text.
+function findTextDifference(originalText, changedText) {
+  let originalWords = originalText.split(" ");
+  let changedWords = changedText.split(" ");
+
+  let commonPhrases = longestCommonSubsequence(originalWords, changedWords);
+
+  return getDifferenceStrings(commonPhrases, originalWords, changedWords);
+}
+
+// Dynamic programming longest common subsequence algorithm.
+let arr = undefined;
+
+function longestCommonSubsequence(a, b) {
+  let m = a.length;
+  let n = b.length;
+  arr = [...Array(m + 1)].map(e => Array(n + 1));
+  return lcs(a, b, m, n);
+}
+
+function lcs(a, b, m, n) {
+  if (arr[m][n] != undefined) {
+    return arr[m][n];
+  }
+
+  let result = undefined;
+  if (m === 0 || n === 0) {
+    result = [];
+  } else if (a[m - 1] === b[n - 1]) {
+    result = lcs(a, b, m - 1, n - 1).concat(a[m - 1]);
+  } else {
+    let lcs1 = lcs(a, b, m - 1, n);
+    let lcs2 = lcs(a, b, m, n - 1);
+    result = lcs1 > lcs2 ? lcs1 : lcs2;
+  }
+
+  arr[m][n] = result;
+  return result;
+}
+
+// Constructs 3 strings that highlight the differences (deletions & additions) between 2 texts.
+function getDifferenceStrings(commonPhrases, originalWords, changedWords) {
+  let differenceStr = "";
+  let deletionsStr = "";
+  let additionsStr = "";
+
+  for (let i = 0; i <= commonPhrases.length; i++) {
+    let phrase = "";
+    let phraseIdxO = originalWords.length;
+    let phraseIdxC = changedWords.length;
+
+    if (i !== commonPhrases.length) {
+      phrase = commonPhrases[i];
+      phraseIdxO = originalWords.indexOf(phrase);
+      phraseIdxC = changedWords.indexOf(phrase);
+    }
+
+    let deletedStr = substringFromWordlist(originalWords, 0, phraseIdxO);
+    let addedStr = substringFromWordlist(changedWords, 0, phraseIdxC);
+
+    differenceStr += `
+      <span class="deleted">${deletedStr}</span>
+      <span class="added">${addedStr}</span>
+      <span class="unchanged">${phrase}</span>
+    `;
+    deletionsStr += `
+      <span class="deleted">${deletedStr}</span>
+      <span class="unchanged">${phrase}</span>
+    `;
+    additionsStr += `
+      <span class="added">${addedStr}</span>
+      <span class="unchanged">${phrase}</span>
+    `;
+
+    originalWords.splice(0, phraseIdxO + 1);
+    changedWords.splice(0, phraseIdxC + 1);
+  }
+
+  return [differenceStr, deletionsStr, additionsStr];
+}
+
+function substringFromWordlist(wordlist, start, end) {
+  if (end < start) return "";
+  let words = wordlist.slice(start, end);
+  return arrToString(words);
+}
+
+function arrToString(arr) {
+  str = "";
+  for (let i = 0; i < arr.length; i++) {
+    str += arr[i] + " ";
+  }
+  return str;
+}