String Decompression with Nested Counts

NYC onsite coding round 2: decode a custom compressed string format where parenthesized groups are repeated using a `{k}` suffix. Variant of LC 394 with different bracketing and an explicit assumption that input is well-formed. Requirements Input: a string using the format <literals>(<group>){k}<literals>... where: (...) marks a repeatable group; the immediately following {k} means repeat k times. Groups may be nested arbitrarily. Literal characters outside of any group are concatenated as-is. k is an integer in [2, 99]. Output: the fully decompressed string. Assume input is well-formed: every { follows a ), and parentheses are properly balanced. Examples Input: "a(abc){3}" Output: "aabcabcabc" Input: "a(b(c){2}){3}d" Output: "abccbccbccd" Notes Recursion with an index pointer is the cleanest implementation: each call returns the decoded substring of a single (...){k} block plus literals up to its terminator. An iterative stack-based approach also works: push (StringBuilder, repeatCount) on (, pop and append on ){k}. The 2-digit k matters only for parsing — read until non-digit. Watch out for literals between two groups: a(b){2}c(d){3} → abbcddd. Preparation Drill the canonical k[group] decode-string problem to muscle memory; the input grammar there is k[group] instead of (group){k}, but the algorithm is identical. Practice both recursive and iterative versions — the iterative stack version generalizes more cleanly if the interviewer asks about iterative streaming. Test with: deeply nested groups, groups containing only literals, two-digit counts ({12}).