String Compression (HackerRank OA, LeetCode 443)

First problem on the recent Staff SWE HackerRank OA: classic run-length string compression done in-place on a character array. Requirements Input: a character array chars. Compress consecutive runs of the same character. A run of length 1 stays as the single character; a run of length k ≥ 2 becomes the character followed by the decimal digits of k. Modify chars in place. Return the new length. Use O(1) extra space (multi-digit counts must be written digit-by-digit; cannot allocate a separate buffer of size n). Examples chars = ['a','a','b','b','c','c','c'] → length 6, chars = ['a','2','b','2','c','3', _ ] chars = ['a'] → length 1, chars = ['a'] chars = ['a','b','b','b','b','b','b','b','b','b','b','b','b'] → length 4, chars = ['a','b','1','2', _ , _ , _ , _ , _ , _ , _ , _ , _ ] # 'b' run of 12 Notes Two pointers: read walks the input, write walks the output. While the run continues, advance read and bump a count; once the run breaks, write the character at write, then if count > 1, write each digit of count in order. Writing digits: convert count to its decimal string, then copy character by character. Use a small fixed buffer (up to 6 digits is enough since len(chars) ≤ 2000 per the LC bound, more than enough in practice). Subtle bug: do not write the run-length digits inside the run loop — wait until the run ends so you don't overwrite unread characters that still belong to the same run. Edge cases: empty array (length 0); single character (no count written); run of length 10+ (multi-digit count must be written left-to-right so the decimal reads correctly). Preparation Implement the in-place two-pointer compressor from scratch in <10 minutes. Verify on the three examples above. Practise the multi-digit count branch explicitly — most bugs land there. Be ready for the decompression follow-up — reverse the operation, expanding "a2b3" back to "aabbb". Watch for ambiguity with multi-digit counts.