Regex Cheat Sheet: Regular Expressions For Cleaning Up HTML

This article presents a collection of regular expressions I frequently use to clean up HTML generated from some tools’ export routines, e.g., Typora. A PowerShell script automates the clean-up task.

TL;DR

This PowerShell script implements the techniques explained in the article. It automates the process of cleaning an entire HTML file.

Links: Open in New Window

What This Does

Open links in a new window and replace single with double quotes.

Regex

• Search for:
  ``
• Replace with:
  ``

Example

• Before:
  <a href='https://helgeklein.com/blog/category/home-automation/'>
• After:
  <a href="https://helgeklein.com/blog/category/home-automation/" target="_blank" rel="noopener">

Headings: Remove ID Attributes

What This Does

Remove ID attributes that some export routines add to every heading.

Regex

• Search for:
  ``
• Replace with:
  ``

Example

• Before:
  <h2 id='what-are-grafana-loki--promtail'>
• After:
  <h2>

Code: Remove Duplicate Syntax Highlighting Language Attributes

What This Does

Some export routines add duplicate language attributes to code blocks. This replaces the two attributes with a single new attribute that I prefer.

Regex

• Search for:
  class='language-([^']+)'\s+lang='\1'>
• Replace with:
  class="lang-\1">

Example

• Before:
  <code class='language-yaml' lang='yaml'>
• After:
  <code class="lang-yaml">

Replace Paragraph Tags with Newlines

What This Does

WordPress doesn’t require authors to enclose paragraphs in ``

``

tags. Instead, it adds those automatically to the HTML when it identifies a paragraph.

To optimize readability in the backend, I remove HTML paragraph tags that are added by many tools’ export routines.

Regex

• Search for:
  ?p>
• Replace with:
  \n

Example

• Before:
  <p>Text in a paragraph.</p>
• After:
  \nText in a paragraph.\n

Replace HTML-Encoded Special Characters With Plain Text

What This Does

Many tools’ export routines replace special characters with their corresponding HTML entities, aka they HTML-encode such special characters. With WordPress, this type of encoding is not necessary, as WordPress does it automatically. To optimize readability in the backend, I convert HTML entities back to plain text.

The regex presented below uses a syntax that is available in Notepad++ (regex docs) and all other applications that use the boost regex library. The regex makes the following replacements:

• < → <
• > → >
• ' → '
• " → "
• & → &

Regex

• Search for:
  (<)|(>)|(')|(")|(&)
• Replace with:
  (?1<)(?2>)(?3')(?4")(?5&)

Example

• Before:
  <a href='([^']+)'>
• After:
  <a href='([^']+)'>