How to Use Regex Extraction

What is it?

Regex extraction runs your regular expression patterns against the full raw source of the page - including all HTML markup, script tags, and embedded data. It's the most flexible text-matching method and works even when there's no clean DOM structure to target.

Use it when the value you need is buried in page source, injected by JavaScript as a string literal, or follows a predictable text pattern like a version number or price format.

When to use it

• Extracting a version number embedded in a script tag (e.g., window.__APP_VERSION__ = "2.4.1")
• Counting occurrences of a pattern (e.g., how many <loc> tags are in a sitemap)
• Pulling a formatted value like a date or price from mixed HTML/text content
• The page has no clean DOM structure to target with CSS or XPath

---

How to configure it

Pick Regex as your extraction method, then map field names to regex patterns:

{
  "method": "regex",
  "fields": {
    "version": "/(\\d+\\.\\d+\\.\\d+)/",
    "url_count": "<loc>"
  }
}

Two modes

1. Pattern with capture groups → extracts the first capture group's value

Wrap a pattern in /…/ with optional flags. The value of the first (…) capture group is returned.

{ "version": "/(\\d+\\.\\d+\\.\\d+)/" }

Returns: "1.2.3"

2. Plain string pattern → counts occurrences

A plain string (no /…/ wrapping) is counted across the full page source.

{ "url_count": "<loc>" }

Returns: 42 (number of times <loc> appears)

---

Example 1 - Extract a version number from a script tag

Goal: Track the app version injected into the page as a JavaScript variable.

Page source (fragment):

<script>
  window.__CONFIG__ = {
    version: "3.14.2",
    env: "production"
  };
</script>

Configuration:

{
  "method": "regex",
  "fields": {
    "version": "/version: \"(\\d+\\.\\d+\\.\\d+)\"/"
  }
}

Output:

{
  "version": "3.14.2"
}

The capture group (\\d+\\.\\d+\\.\\d+) isolates just the version number from the surrounding text.

---

Example 2 - Count sitemap URLs

Goal: Know when a sitemap grows or shrinks (new pages added or removed).

URL: https://example.com/sitemap.xml

File contents (fragment):

<urlset>
  <url><loc>https://example.com/</loc></url>
  <url><loc>https://example.com/about</loc></url>
  <url><loc>https://example.com/pricing</loc></url>
</urlset>

Configuration:

{
  "method": "regex",
  "fields": {
    "url_count": "<loc>"
  }
}

Output:

{
  "url_count": 3
}

No capture group - the plain string counts occurrences.

---

Example 3 - Extract a price from mixed content

Goal: Pull a price from a page where it's written inline in a paragraph with no dedicated element.

Page HTML:

<p>
  The annual plan is currently priced at <strong>$149.00</strong> per seat,
  billed yearly.
</p>

Configuration:

{
  "method": "regex",
  "fields": {
    "annual_price": "/\\$([\\d,]+\\.\\d{2})/"
  }
}

Output:

{
  "annual_price": "149.00"
}

The $ sign is matched but not captured - only the digits inside (…) are returned.

---

Example 4 - Extract a date from a rendered string

Goal: Track the "Last updated" date on a policy or terms page.

Page HTML:

<footer class="doc-meta">
  Last updated: March 15, 2026
</footer>

Configuration:

{
  "method": "regex",
  "fields": {
    "last_updated": "/Last updated: ([A-Za-z]+ \\d{1,2}, \\d{4})/"
  }
}

Output:

{
  "last_updated": "March 15, 2026"
}

---

Example 5 - Count keywords or links

Goal: Track how many external links appear on a page.

Configuration:

{
  "method": "regex",
  "fields": {
    "external_links": "href=\"https://"
  }
}

Output:

{
  "external_links": 17
}

Plain string, no capture group - returns a count.

---

Regex pattern reference

Pattern	What it matches
\\d	Any digit (0–9)
\\d+	One or more digits
\\d+\\.\\d+\\.\\d+	Semantic version like 1.2.3
[A-Za-z]+	One or more letters
\\$[\\d,]+\\.\\d{2}	A price like $1,999.00
(…)	Capture group - what gets returned
/pattern/i	Case-insensitive match
/pattern/g	Global match (used for counting without capture groups)

Tips

• Double-escape backslashes in JSON. Regex \d becomes \\d inside a JSON string. Write \\d+\\.\\d+ not \d+\.\d+.
• Use capture groups to isolate the value. Without a capture group the field returns a count; with one, it returns the matched text.
• Test your regex before saving. Paste a fragment of the page source into regex101.com and verify the match.
• Regex runs on raw source. It sees the full HTML including tags, attributes, and script contents - which makes it great for embedded data, but your pattern needs to be specific enough to avoid accidental matches.

Common issues

Problem	Fix
Returns a count instead of text	Add a capture group (…) around the part you want
Returns 0	The pattern doesn't match - check the page source and adjust
Getting the wrong match	Make the surrounding context more specific so only one match exists
Backslash errors	Double-escape in JSON: \\d not \d