refactor(strip_html): re-implementation (#232)

lib/strip_html.js

@@ -1,3 +1,120 @@
 'use strict';
 
-module.exports = require('striptags');
+const STATE_PLAINTEXT = Symbol('plaintext');
+const STATE_HTML = Symbol('html');
+const STATE_COMMENT = Symbol('comment');
+
+function striptags(html = '') {
+  let state = STATE_PLAINTEXT;
+  let tag_buffer = '';
+  let depth = 0;
+  let in_quote_char = '';
+  let output = '';
+
+  const { length } = html;
+
+  for (let idx = 0; idx < length; idx++) {
+    const char = html[idx];
+
+    if (state === STATE_PLAINTEXT) {
+      switch (char) {
+        case '<':
+          state = STATE_HTML;
+          tag_buffer = tag_buffer + char;
+          break;
+
+        default:
+          output += char;
+          break;
+      }
+    } else if (state === STATE_HTML) {
+      switch (char) {
+        case '<':
+          // ignore '<' if inside a quote
+          if (in_quote_char) break;
+
+          // we're seeing a nested '<'
+          depth++;
+          break;
+
+        case '>':
+          // ignore '>' if inside a quote
+          if (in_quote_char) {
+            break;
+          }
+
+          // something like this is happening: '<<>>'
+          if (depth) {
+            depth--;
+
+            break;
+          }
+
+          // this is closing the tag in tag_buffer
+          in_quote_char = '';
+          state = STATE_PLAINTEXT;
+          // tag_buffer += '>';
+
+          tag_buffer = '';
+          break;
+
+        case '"':
+        case '\'':
+          // catch both single and double quotes
+
+          if (char === in_quote_char) {
+            in_quote_char = '';
+          } else {
+            in_quote_char = in_quote_char || char;
+          }
+
+          tag_buffer = tag_buffer + char;
+          break;
+
+        case '-':
+          if (tag_buffer === '<!-') {
+            state = STATE_COMMENT;
+          }
+
+          tag_buffer = tag_buffer + char;
+          break;
+
+        case ' ':
+        case '\n':
+          if (tag_buffer === '<') {
+            state = STATE_PLAINTEXT;
+            output += '< ';
+            tag_buffer = '';
+
+            break;
+          }
+
+          tag_buffer = tag_buffer + char;
+          break;
+
+        default:
+          tag_buffer = tag_buffer + char;
+          break;
+      }
+    } else if (state === STATE_COMMENT) {
+      switch (char) {
+        case '>':
+          if (tag_buffer.slice(-2) === '--') {
+            // close the comment
+            state = STATE_PLAINTEXT;
+          }
+
+          tag_buffer = '';
+          break;
+
+        default:
+          tag_buffer = tag_buffer + char;
+          break;
+      }
+    }
+  }
+
+  return output;
+}
+
+module.exports = striptags;

package.json

@@ -50,8 +50,7 @@
     "highlight.js": "^10.0.0",
     "htmlparser2": "^4.0.0",
     "prismjs": "^1.17.1",
-    "strip-indent": "^3.0.0",
-    "striptags": "^3.1.1"
+    "strip-indent": "^3.0.0"
   },
   "engines": {
     "node": ">=10.13.0"

test/strip_html.spec.js

@@ -0,0 +1,54 @@
+'use strict';
+
+const stripHTML = require('../lib/strip_html');
+
+describe('stripHTML', () => {
+  it('should not strip invalid tags', () => {
+    const text = 'lorem ipsum < a> < div>';
+
+    stripHTML(text).should.eql(text);
+  });
+
+  it('should remove simple HTML tags', () => {
+    const html = '<a href="">lorem <strong>ipsum</strong></a>';
+    const text = 'lorem ipsum';
+
+    stripHTML(html).should.eql(text);
+  });
+
+  it('should remove comments', () => {
+    const html = '<!-- lorem -- ipsum -- --> dolor sit amet';
+    const text = ' dolor sit amet';
+
+    stripHTML(html).should.eql(text);
+  });
+
+  it('should strip tags within comments', () => {
+    const html = '<!-- <strong>lorem ipsum</strong> --> dolor sit';
+    const text = ' dolor sit';
+
+    stripHTML(html).should.eql(text);
+  });
+
+
+  it('should not fail with nested quotes', () => {
+    const html = '<article attr="foo \'bar\'">lorem</article> ipsum';
+    const text = 'lorem ipsum';
+
+    stripHTML(html).should.eql(text);
+  });
+
+  it('should strip extra < within tags', () => {
+    const html = '<div<>>lorem ipsum</div>';
+    const text = 'lorem ipsum';
+
+    stripHTML(html).should.eql(text);
+  });
+
+  it('should strip <> within quotes', () => {
+    const html = '<a href="<script>">lorem ipsum</a>';
+    const text = 'lorem ipsum';
+
+    stripHTML(html).should.eql(text);
+  });
+});