String manipulation related problem

The title is too vague to support one factual algorithm, so model the behavior an interviewer expects: turn ambiguity into a short, testable contract. Ask what counts as a character, whether matching is case-sensitive, whether whitespace matters, and whether the result is a boolean, index, or transformed string. Then select a known pattern: a scan for validation, a map for counts, a window for contiguous constraints, a stack for nesting, or a builder for output. Keep state incremental and describe the invariant in one sentence. For example, a filter can maintain that the output buffer contains exactly the accepted prefix processed so far. This produces O(n) time with O(n) output space in the common rebuilding case. If the full question later includes regular expressions, escaped quotes, or Unicode normalization, those are requirements that change the implementation and test set. Calling that out is more accurate than pretending the label determines an answer.