Compute letter frequencies from encoded string

Problem

You are given an encoded string s representing a string of lowercase English letters (a–z). The encoding follows these rules:

1. Letters a to i are encoded as digits 1 to 9.
   • a -> "1", b -> "2", …, i -> "9"
2. Letters j to z are encoded as two digits followed by #.
   • j -> "10#", k -> "11#", …, z -> "26#"
3. If a letter repeats consecutively k >= 2 times, the repetition is encoded by appending "(k)" immediately after that letter’s code.
   • Example: "aa" -> "1(2)", "ccc" -> "3(3)", "jj" -> "10#(2)"

Task

Return an integer array counts of length 26, where counts[0] is the number of 'a' characters in the decoded string, counts[1] is the number of 'b', …, and counts[25] is the number of 'z'.

Examples

• s = "1226#24#" decodes to "abzx".
• s = "1(2)23(3)" decodes to "aabccc".
• s = "2110#(2)" decodes to "bajj".
• s = "23#(2)24#25#26#23#(3)" decodes to "wwxyzwww".

Input/Output

• Input: a string s consisting of digits, #, (, and ) that follows the encoding rules above.
• Output: an array of 26 integers with letter frequencies.

Notes / Constraints (reasonable interview assumptions)

• s is validly encoded.
• Counts in parentheses can be multiple digits (e.g., (12)).
• Aim for O(|s|) time and O(1) extra space (excluding the output array).

Compute letter frequencies from encoded string

Hints

1. Scan left to right. The only way to tell a two-digit letter (10#..26#) from a single digit is to peek two characters ahead for a '#'.
2. After consuming a letter's code, check whether the next character is '('. If so, read the integer up to ')' — it can be more than one digit — and use it as the multiplier; otherwise the letter occurs once.
3. Map the numeric value v (1..26) to array index v-1 and add the repetition count there.