1. [Home](https://www.toolcomb.com/) 2. [Text Tools](https://www.toolcomb.com/text-tools) 3. Regex Named Capture Tester # Regex Named Capture Group Tester - Test (?...) Online Free Free online regex named capture group tester. Enter a pattern with (?...) syntax, flags, and a test string. Highlight matches, inspect numbered and named groups in a table, copy matches as JSON - browser-only. ### Pattern & flags Regular expression _(?\...)_ g global i ignore case m multiline s dotAll u unicode y sticky ### Test string _Release 2024-03-15 shipped. Backup dated 2023-11-01 remains._ Copy matches JSON Clear ### Highlights (2 matches) ``` Release 2024-03-15 shipped. Backup dated 900/60 px-0.5 rounded-sm">2023-11-01 remains. ``` ### Capture groups | # | Match | Index | Numbered | Named | | --- | --- | --- | --- | --- | | 1 | 2024-03-15 | 8 | [1]=2024 [2]=03 [3]=15 | year=2024, month=03, day=15 | | 2 | 2023-11-01 | 41 | [1]=2023 [2]=11 [3]=01 | year=2023, month=11, day=01 | ## What is a regex named capture group tester? A regex named capture group tester is a browser tool that runs a regular expression against sample text and shows what each (?...) segment captured. Instead of mentally mapping group 1 to “year” and group 2 to “month,” you see labeled keys beside every match. This page is built for that workflow: paste a pattern, paste a log line or URL, and read a table of named and numbered captures without writing a throwaway script. Named captures shine when patterns parse semi-structured strings. ISO dates, query strings, user agents, invoice numbers, and Kubernetes pod names all benefit from readable group labels. Numbered groups still exist - parentheses increment indices in source order - but namedGroups.year is easier to discuss in code review than “capture three.” The tester highlights every match in the test string and lists each occurrence as its own row so global searches do not collapse into one ambiguous summary. This tool is intentionally separate from ToolComb’s general regex tester at /regex-testing. That page optimizes for broad pattern exploration, samples, and layout flexibility. Here the copy actions and UI center on (?...) syntax and JSON export of captures. If you only need yes/no matching, either page works; if you are documenting a parser or writing test fixtures from real logs, start here. Errors are first-class. Invalid group names, dangling backslashes, and unbalanced parentheses surface a clear message immediately. That saves a trip through stack traces when a teammate pastes a pattern with unsupported features. Privacy matches other ToolComb text tools: nothing is posted to a server. Security teams can paste redacted production samples locally. You should still avoid pasting live secrets; the guarantee is local execution, not magic scrubbing. Copy matches JSON serializes match, index, groups, and namedGroups for each hit. Drop the array into Jest snapshots, OpenAPI examples, or internal runbooks. The format is plain JSON - no proprietary wrapper - so jq and Python json.loads consume it directly. Flags control whether the tool finds all matches, ignores case, treats line starts separately, lets dots cross newlines, supports Unicode-aware patterns, or anchors matching at the current position. Toggle them to match the behavior you expect in your application. Understanding numbered versus named groups prevents off-by-one bugs. Group 0 is always the full match in API terms, while match[1] is the first parenthesized capture. Named groups occupy the same numbering sequence as if you had only parentheses - (?x)(?y) still yields match[1] and match[2] plus groups.a and groups.b. When you refactor a pattern to insert a new named group in the middle, downstream code that relied on numeric indices breaks unless you update references; names stabilize refactors because destructuring follows labels. Lookahead and lookbehind assertions can wrap naming decisions. A pattern like (?https?)://(?[\w.-]+) keeps protocol and host labeled even when a non-capturing cluster (?:...) sits between optional www. and the host. Non-capturing groups do not appear in the table - only capturing parentheses and named captures - which keeps the UI focused on data you will actually extract in application code. For security reviewers, local testing reduces the temptation to paste production payloads into cloud regex sites with unknown retention policies. You still must treat pasted secrets carefully, but the attack surface is your own browser tab rather than a third-party SaaS analytics pipeline. Educators use named capture tables to bridge syntax cheat sheets and runnable examples. Showing match.groups beside highlighted text connects abstract grammar rules to concrete substrings students can click-copy into assignments. Microservice teams parsing trace IDs and correlation headers often encode structure in a single string. Named groups document which segment is tenant, shard, and sequence without maintaining a separate parsing spec that drifts from the regex in repo. ## How to test named capture groups online You can validate a named-group parser in under two minutes. Follow these steps when porting patterns from docs, logs, or code review comments. 1. **Write the pattern without slashes** Enter only the pattern body - not /pattern/flags. Wrap each capture you want labeled in (?