Filter Lines

How to filter lines

Paste your list into Your lines, one item per line. Choose Keep matching lines or Remove matching lines, then pick what a match is: words, a regular expression, a line length, or text at the start or end of the line. The result updates as you type, with counts of input, kept and removed lines. Tick Show removed lines to see what the filter threw away, which is the quickest way to check that a rule does what you meant. Copy the result or download it as a .txt file.

Everything runs in your browser. Nothing you paste is uploaded, so it is safe to use with server logs, customer lists and unpublished keyword research.

Options

• Ignore case (on by default) treats Error and error as the same in words, regex and starts-or-ends-with matching.
• Whole words only makes cat match "the cat." but not "category". Letters in any script, digits and underscores count as part of a word, similar to grep -w, so caf does not match "café". Leave it off for languages written without spaces, such as Chinese or Japanese.
• Trim lines removes spaces at both ends before matching, and the output is trimmed too.
• Remove empty lines drops blank and whitespace-only lines before filtering.
• Remove duplicates from the result keeps the first copy of each kept line. The comparison is exact and case-sensitive; turn on Trim lines if copies differ only in surrounding spaces.

Examples for each mode

Each row filters these five lines with Keep matching lines, unless it says otherwise:

1. https://example.com/pricing
2. https://blog.example.com/serp-api-guide
3. http://example.org/old-page
4. https://example.com/files/report.pdf
5. https://www.example.net/a

To keep lines that contain none of your words, use Remove matching lines with Any of these. To keep lines that lack at least one of them, use Remove matching lines with All of these.

Keep only URLs from one domain

A words filter for example.com is quick, but it also keeps blog.example.com, notexample.com and example.com.evil.net, because it matches the text anywhere in the line. For an exact host, switch to Regular expression and anchor the pattern: ^https?://(www\.)?example\.com([/?#]|$) keeps example.com and www.example.com over HTTP or HTTPS and nothing else. To include every subdomain, use ^https?://([^/]+\.)?example\.com([/?#]|$). This is handy for trimming a crawl export or a sitemap down to one site before you compare it with another list in Compare Lists.

Remove lines with certain words from keyword lists

Keyword exports are full of terms you will never target. Choose Remove matching lines, Words and Any of these, and list the negatives one per line: free, jobs, salary, competitor brand names. Turn on Whole words only so that free removes "free seo tools" but keeps "freelance seo writer". Add Trim lines and Remove duplicates to tidy the result in the same pass, or run it through Remove Duplicate Lines for more control over what counts as a copy.

Regex cheat sheet

Patterns use JavaScript syntax, and each line is tested on its own. Unicode mode is on by default: it adds \p{L} and makes . match a whole emoji, but it rejects needless escapes such as \- outside brackets. If you see that error, remove the backslash or turn Unicode mode off.

A group that repeats and contains a repeat, such as (\w+\s?)+$, can take minutes on a single 35-character line, because the engine tries every way of splitting it. The tool doesn't run such patterns as you type; it warns you and waits for Filter lines. It also checks the clock every thousand lines and stops after about 2 seconds with a message, though a single match already in progress can't be interrupted. For inputs over about 1 MB, regex filtering runs only when you press Filter lines, and above 2 MB every mode works that way.

Find lines by length

Length mode keeps lines that are at least, at most, exactly or between a number of characters. Use it to find page titles longer than about 60 characters, meta descriptions over 160, or short junk lines in a scraped list. Google shortens long titles by pixel width rather than a fixed count, so treat those numbers as rough guides. Length counts Unicode code points: 😀 is 1, but a flag or an emoji with a skin tone is 2, and an accent typed as a separate combining mark adds 1. With Trim lines on, the trimmed line is measured.

The same filters with grep and spreadsheets

On the command line, grep keeps matching lines, grep -v removes them and grep -E enables extended regular expressions. In a spreadsheet, FILTER is available in Google Sheets and in Excel 365 and 2021 or later; SEARCH ignores case, FIND does not, and REGEXMATCH is a Sheets function.

Two differences to watch for. In a grep pattern file, an empty line matches every line, while this tool ignores blank lines in the word list. And grep has no "all of these"; you chain it instead, as in grep -i 'red' list.txt | grep -i 'apple'. If you need the text between two markers rather than whole lines, use Extract Text Between.