Skip to main content
Regular Expressions (RegEx) let you match patterns instead of exact strings. A single RegEx entry can cover an entire category of sites or apps.

​
Choosing the Regex Match Type

The What to block step — click a row to open its TYPE / VALUE / MATCH editor

Every entry in a plan has a target type and a match rule. To use a regular expression:
  1. In the Plan Wizard’s What to block step, add your pattern with the add row (or click an existing entry)
  2. Click the row to open its inline editor — the hint reads Click a row to edit its type and match
  3. Set MATCH to Regex
  4. Check the TYPE is what you want the pattern to run against (see below)
The match options are Contains, Regex, Exact, Starts With and Ends With. Regex is only applied when the row’s match is set to Regex — otherwise your entry is treated as plain text.

​
What the pattern is matched against

Keyword with Regex is the broadest website rule: it catches a term wherever it appears — in the address or in the tab title. URL is the safer choice when you only want to match addresses.
FocusMe’s RegEx matching is always case-insensitive — you don’t need to add (?i) to your patterns.

​
How a website address is matched

Your pattern is never tested against the address exactly as the browser shows it. FocusMe first removes https:// or http://, removes a leading www. and removes any trailing /, then searches for your pattern anywhere in what is left, ignoring case. The path is part of that text. So for https://www.pinterest.com/ideas the pattern is tested against:
Three things follow from that:
  • Do not anchor the end of a domain with $. $ is the end of the whole string, path and all, so \.org$ matches example.org but never example.org/page. To block a whole top-level domain, leave room for the path: ^[^/]*\.(org|it|net)(/|$)
  • Anchor on the dots to catch country variants. (^|\.)pinterest\. matches pinterest.com, pinterest.es and pinterest.jp, but not notpinterest.com
  • There is no https:// or www. left to match. A pattern that begins with either of them never matches anything
  • Do not reach for RegEx when a plain entry will do. If you simply want everything containing a word, a Contains entry says so with no pattern to get wrong
This section is about patterns. A Contains entry that is a bare domain — a dot and no slash, space or colon — is not searched for as text at all: FocusMe matches it as a domain, so reddit.com covers reddit.com and its subdomains but not notreddit.com. Every other Contains entry (one with a path, or a bare word such as game) is the plain substring test described here. See Blocking Websites.

​
Wildcards Without RegEx

For simple cases you may not need RegEx at all. An entry that contains * (and no |) is treated as a wildcard pattern automatically, whatever its match type: reddit.com/r/* matches any subreddit on reddit.com. A wildcard entry is anchored at both ends, so it has to describe the whole cleaned address:
  • A leading *. demands a literal dot before the domain, so *.reddit.com/r/* never matches reddit.com/r/funny — the www. has already been removed by the time the pattern is applied
  • By the same token reddit.com/r/* does not cover old.reddit.com/r/funny. Add *.reddit.com/r/* as a second entry if you want the subdomains as well
Use full RegEx when you need alternation, word boundaries or character classes.

​
Essential Patterns

​
Word Boundaries

Use \b to match whole words only. A \b sits only between a word character (a letter, digit or _) and a non-word character, so a . or a / creates a boundary but another letter does not:
Matches steam.com and store.steam.com — “steam” is followed by a .. Does not match upstream.example.com (a letter precedes “steam”) or store.steampowered.com (a letter follows it). That second case catches people out. To match a name wherever it appears, including where it runs into other letters, drop the \b and use steam on its own — or, simpler still, a plain Contains entry, which is the better tool whenever you only want everything containing a word. To match a domain and all of its country variants, anchor on the dots instead:
Matches steam.com, steam.es and store.steam.com, but not notsteam.com.

​
OR Matching

Use | (pipe) to match any of several options:
Matches any address (or, with the Keyword type, any title) containing “reddit”, “tiktok” or “instagram”.
Do not end a pattern with a stray | — an empty alternative matches everything, which would block every page. FocusMe strips a trailing | when it loads a plan, but it is better not to save one in the first place.

​
Combining Patterns

One entry covers the major social networks. The . in x\.com is escaped as \. because a bare . in RegEx matches any character.

​
Practical Examples

​
Block All Gaming Platforms

No \b here on purpose: \bsteam\b would miss store.steampowered.com, because a letter follows “steam”. Naming the store and community domains covers Steam without a word boundary.

​
Block All News Sites

​
Block YouTube Shorts and Reels

Use the URL type for this one so that a video titled “shorts” on another site is not caught.

​
Block All Subdomains of a Domain

Matches example.com, www.example.com, mail.example.com and so on, but not notexample.com. Anchoring on the dot is what creates the boundary: (.+\.)?example\.com looks like it does the same job, but its optional group can simply not participate, so it still matches notexample.com. For this particular job you do not need RegEx at all. A plain Contains entry for example.com is a bare domain, so FocusMe matches it as a domain — the domain and its subdomains, but not notexample.com. Reach for a pattern when the target is not a bare domain, or when one entry has to cover several domains.

​
Match Several Versions of an App

On the Applications tab, with the Process Name type:

​
RegEx Quick Reference

​
Testing Your Patterns

  • Check the wizard’s warning. If a pattern is not valid RegEx, the What to block step shows an invalid regex pattern warning and the entry is ignored until you fix it.
  • Use the Helper. Open Helper from the sidebar to see the exact URL, Window Title, Process Name and other values of every open window — these are the strings your pattern must match. See Using the Helper.
  • Try it on a sample. Paste a few real addresses or titles into regex101.com alongside your pattern before adding it to a protected plan.
  • Start unprotected. Test a new pattern in a plan with Unlocked protection, confirm it blocks what you expect on the Activity page, then raise the protection.

​
Common Mistakes

  1. Leaving the match on Contains — the pattern is never compiled. It is treated as literal text, or as a bare domain if it happens to look like one, so it matches nothing
  2. Unescaped dots — google.com also matches googleXcom; use google\.com
  3. Too broad a pattern — game blocks game.com but also gamestop.com and any address containing “game”; add word boundaries with \b
  4. Using Keyword when you meant URL — Keyword also searches page titles, so a news article about “steam engines” would match \bsteam\b
  5. Anchoring the end with $ — the path is part of the matched text, so \.com$ misses example.com/feed
  6. Starting a pattern with https:// or www\. — both are removed before matching, so the pattern never matches
  7. A trailing | — makes the pattern match everything