Extract Text Between Two Strings

How to extract text between two strings

Paste your text on the left, or press Load example to try a short application log. Type a start marker and an end marker, or press a preset such as Double quotes or Square brackets to fill both. The result lists every match, one per line, and updates as you type. Copy it or download it as a .txt file. Everything runs in your browser, so nothing you paste is sent anywhere.

Markers are plain text, not patterns. A marker can be one character, such as (, or a longer string, such as user=". Leave the start marker empty to match from the beginning of each line, or the end marker empty to match to the end of the line. Type \t for a tab, \n for a line break and \\ for a backslash.

What the options do

• Include the markers in the result returns [auth] instead of auth.
• Ignore case in markers lets start match START or Start, accented and non-Latin letters included. The results keep their original case.
• Only the first match per line gives at most one result per line, handy when the first bracket in each log line holds the level or module.
• Trim results, Skip empty results and Remove duplicates tidy the list. A result made only of spaces counts as empty.
• Output separator goes between results: a new line by default, , for a comma list or \t for tabs.

Examples

In the table, ␣ stands for a space. Markers made of whole words, like the last three rows, are the quickest way to pull one value out of every line of a log: an order number, a user name or a duration.

Extract or remove

Extract matches returns only the matches. Remove matches returns your text with every match deleted and everything else left as it was. With Include the markers off, removing keeps the markers, so call(x) becomes call(). With it on, they go too, so Hello (world) becomes Hello . When removing, Trim results and Skip empty results only touch the lines you removed text from, so you can strip notes like [1] or inline comments and tidy the leftovers without changing the other lines.

Line by line or whole text

Line by line, the default, a match never crosses a line break. A start marker with no end marker after it on the same line gives no match, and the status below the result says on how many lines that happened. Choose Whole text when a match can span lines, such as a paragraph split over several lines or a /* … */ comment. When extracting, set Line breaks inside a match to Replace with a space to get each of those matches back on a single line.

Why nested brackets are not supported

The tool uses the shortest match: each start marker pairs with the next end marker after it, and the search carries on from there. For (a (b) c) the first opening bracket pairs with the first closing one, so the result is a (b. Matching nested pairs means counting brackets as they open and close, which is a parser's job, and a JavaScript regular expression can't do it either. Instead:

• Use longer markers that the inner pairs can't match, such as total( and ) +.
• If the outer pair closes at the end of the line, leave the end marker empty and delete the last character of each result.
• For JSON, source code or HTML with deep nesting, use a parser for that format.

The regular expression equivalent

Line by line, the tool finds the same matches as a non-greedy regular expression such as "(.*?)", \[(.*?)\] or href="(.*?)" with the g flag. The ? after .* is what makes it stop at the first end marker instead of the last one. Whole text mode is the same pattern with the s flag, which lets the dot match line breaks, and Ignore case is the i flag. The difference is that you don't have to escape characters such as [, ( or ., because markers here are always literal.

Markers work well on HTML you control, but a real parser is safer on HTML from the wild. Attributes can use single quotes or none, carry extra spaces around the equals sign, or appear inside comments and scripts. To list the links on a page, paste its HTML into Extract Links from HTML, which parses it with the browser's HTML parser and lists each link with its anchor text and rel values.

Common uses

Developers use it to pull IDs, user names or durations out of logs, collect quoted strings from code, and list the placeholders in a template. SEOs use it to grab titles, hrefs or image paths from a snippet of HTML, or to strip citation markers from copied text. To narrow the input first, keep only the relevant lines with Filter Lines. To break each result into columns afterwards, use Split Text by Delimiter.