New data.stripHtml() to strip HTML from text.

igoramadas · Oct 30, 2019 · 52ee71b · 52ee71b
1 parent 5e141bb
commit 52ee71b
Show file tree

Hide file tree

Showing 7 changed files with 304 additions and 29 deletions.
diff --git a/History.md b/History.md
@@ -1,5 +1,10 @@
 # Changelog for JAUL
 
+1.2.0
+=====
+* New data.stripHtml() to remove HTML tags from text.
+* Updated dependencies.
+
 1.1.2
 =====
 * Updated dependencies.

diff --git a/lib/data.d.ts b/lib/data.d.ts
@@ -24,6 +24,13 @@ declare class DataUtils {
      * @returns The minified JSON as object or string, depending on asString.
      */
     static minifyJson(source: string, asString?: boolean): any;
+    /**
+     * Strips all the HTML tags from the specified value and returns only the text.
+     * Tags will be replaced with an empty space by default.
+     * @param value The HTML string to be converted to only text.
+     * @param tagReplace Replace tags with that value
+     */
+    static stripHtml(html: string, tagReplace?: string): string;
     /**
      * Generates a RFC4122-compliant unique ID using random numbers.
      * @returns A unique ID.

diff --git a/lib/data.js b/lib/data.js
@@ -155,6 +155,114 @@ class DataUtils {
             return JSON.parse(result);
         }
     }
+    /**
+     * Strips all the HTML tags from the specified value and returns only the text.
+     * Tags will be replaced with an empty space by default.
+     * @param value The HTML string to be converted to only text.
+     * @param tagReplace Replace tags with that value
+     */
+    static stripHtml(html, tagReplace) {
+        if (!html || html == "") {
+            return "";
+        }
+        // Default tagReplace is an empty space.
+        if (!tagReplace && tagReplace !== "") {
+            tagReplace = " ";
+        }
+        // Output will be set here.
+        let output = "";
+        // Default symbols.
+        const PLAINTEXT = Symbol("plaintext");
+        const HTML = Symbol("html");
+        const COMMENT = Symbol("comment");
+        // Initial state variables.
+        let tagBuffer = "";
+        let qChar = "";
+        let depth = 0;
+        let state = PLAINTEXT;
+        // Iterate HTML content to replace tags.
+        for (let i = 0, length = html.length; i < length; i++) {
+            let char = html[i];
+            if (state === PLAINTEXT) {
+                switch (char) {
+                    case "<":
+                        state = HTML;
+                        tagBuffer += char;
+                        break;
+                    default:
+                        output += char;
+                        break;
+                }
+            }
+            else if (state === HTML) {
+                switch (char) {
+                    case "<":
+                        if (qChar) {
+                            break;
+                        }
+                        depth++;
+                        break;
+                    case ">":
+                        if (qChar) {
+                            break;
+                        }
+                        if (depth) {
+                            depth--;
+                            break;
+                        }
+                        qChar = "";
+                        state = PLAINTEXT;
+                        tagBuffer += ">";
+                        output += tagReplace;
+                        tagBuffer = "";
+                        break;
+                    case '"':
+                    case "'":
+                        if (char === qChar) {
+                            qChar = "";
+                        }
+                        else {
+                            qChar = qChar || char;
+                        }
+                        tagBuffer += char;
+                        break;
+                    case "-":
+                        if (tagBuffer === "<!-") {
+                            state = COMMENT;
+                        }
+                        tagBuffer += char;
+                        break;
+                    case " ":
+                    case "\n":
+                        if (tagBuffer === "<") {
+                            state = PLAINTEXT;
+                            output += "< ";
+                            tagBuffer = "";
+                            break;
+                        }
+                        tagBuffer += char;
+                        break;
+                    default:
+                        tagBuffer += char;
+                        break;
+                }
+            }
+            else if (state === COMMENT) {
+                switch (char) {
+                    case ">":
+                        if (tagBuffer.slice(-2) == "--") {
+                            state = PLAINTEXT;
+                        }
+                        tagBuffer = "";
+                        break;
+                    default:
+                        tagBuffer += char;
+                        break;
+                }
+            }
+        }
+        return output;
+    }
     /**
      * Generates a RFC4122-compliant unique ID using random numbers.
      * @returns A unique ID.

diff --git a/package-lock.json b/package-lock.json
diff --git a/package.json b/package.json
@@ -1,6 +1,6 @@
 {
     "name": "jaul",
-    "version": "1.1.2",
+    "version": "1.2.0",
     "description": "Just another utilities library. But a good one.",
     "keywords": [
         "jaul",
@@ -32,13 +32,13 @@
         "moment": "^2.24.0"
     },
     "devDependencies": {
-        "@types/node": "^12.7.12",
+        "@types/node": "^12.12.1",
         "chai": "^4.2.0",
         "coveralls": "^3.0.7",
         "express": "^4.17.1",
         "get-port": "^5.0.0",
         "nyc": "^14.1.1",
-        "mocha": "^6.2.1",
+        "mocha": "^6.2.2",
         "mocha-lcov-reporter": "^1.3.0",
         "socket.io-client": "^2.3.0",
         "supertest": "^4.0.2",