# ARGREP
```json metadata
{
"title": "ARGREP",
"description": "Searches array elements in a range using textual predicates.",
"categories": ["docs","develop","stack","oss","rs","rc","oss","kubernetes","clients"],
"arguments": [{"key_spec_index":0,"name":"key","type":"key"},{"name":"start","type":"string"},{"name":"end","type":"string"},{"arguments":[{"arguments":[{"name":"exact","token":"EXACT","type":"pure-token"},{"name":"string","type":"string"}],"name":"exact","type":"block"},{"arguments":[{"name":"match","token":"MATCH","type":"pure-token"},{"name":"string","type":"string"}],"name":"match","type":"block"},{"arguments":[{"name":"glob","token":"GLOB","type":"pure-token"},{"name":"pattern","type":"string"}],"name":"glob","type":"block"},{"arguments":[{"name":"re","token":"RE","type":"pure-token"},{"name":"pattern","type":"string"}],"name":"re","type":"block"}],"multiple":true,"name":"predicate","type":"oneof"},{"arguments":[{"name":"and","token":"AND","type":"pure-token"},{"name":"or","token":"OR","type":"pure-token"},{"name":"limit","token":"LIMIT","type":"integer"},{"name":"withvalues","token":"WITHVALUES","type":"pure-token"},{"name":"nocase","token":"NOCASE","type":"pure-token"}],"multiple":true,"name":"options","optional":true,"type":"oneof"}],
"syntax_fmt": "ARGREP key start end\n \u003cEXACT string | MATCH string | GLOB pattern | RE pattern [...]\u003e\n [AND | OR | LIMIT limit | WITHVALUES | NOCASE [...]]",
"complexity": "O(P * C) where P is the number of visited positions in touched slices and C is the cost of evaluating the predicates on one existing element.",
"group": "array",
"command_flags": ["readonly"],
"acl_categories": ["@array"],
"since": "8.8.0",
"arity": -6,
"key_specs": [{"begin_search":{"index":{"pos":1}},"find_keys":{"range":{"lastkey":0,"limit":0,"step":1}},"flags":["ro","access"]}],
"tableOfContents": {"sections":[{"id":"required-arguments","title":"Required arguments"},{"id":"optional-arguments","title":"Optional arguments"},{"id":"examples","title":"Examples"},{"id":"return-information","title":"Return information"}]}
,
"codeExamples": []
}
```Searches array elements in a range using textual predicates and returns the indices of the matching elements. Empty slots in the range are skipped.
## Required arguments
key
The name of the key that holds the array.
start
The zero-based integer index at which to begin searching. The special value `-` denotes the first index of the array. If `start` is greater than `end`, matches are returned in reverse index order.
end
The zero-based integer index at which to stop searching (inclusive). The special value `+` denotes the last index of the array.
predicate
One or more textual predicates to evaluate against each non-empty element in the range. Each predicate is one of:
- **`EXACT string`** — Matches elements whose value is exactly equal to `string`.
- **`MATCH string`** — Matches elements whose value contains `string` as a substring.
- **`GLOB pattern`** — Matches elements whose value matches the glob-style `pattern` (with `*`, `?`, and `[...]` wildcards), the same syntax used by [`KEYS`](https://redis.io/docs/latest/commands/keys) and [`SCAN`](https://redis.io/docs/latest/commands/scan) `MATCH`.
- **`RE pattern`** — Matches elements whose value matches the regular expression `pattern`.
When more than one predicate is supplied, use the `AND` or `OR` option to control how they are combined. The default is `OR`.
options
Zero or more of the following modifiers:
- **`AND`** — Combine multiple predicates with logical AND. An element matches only if every predicate matches.
- **`OR`** — Combine multiple predicates with logical OR. An element matches if any predicate matches. This is the default.
- **`LIMIT limit`** — Stop after `limit` matches have been collected. `limit` must be a positive integer. When omitted, all matches in the range are returned.
- **`WITHVALUES`** — In addition to each matching index, return the matching value. The reply becomes a flat list of alternating index-value pairs.
- **`NOCASE`** — Perform case-insensitive comparisons for `EXACT`, `MATCH`, `GLOB`, and `RE`.