Problem

At Meta, you need to analyze a large log file from a backend service such as Messenger infrastructure and count how many times specific error codes appear. Implement an efficient Python function that processes the log line by line without loading the entire file into memory.

Task

Given a list of log lines and a list of target error codes, return a dictionary mapping each target code to its total number of occurrences. A log line may contain zero or more error codes. Count only exact matches of codes that appear in the form ERROR_<ALPHANUMERIC_OR_UNDERSCORE>.

Input / Output

• Input:
  • log_lines: a list of strings
  • target_codes: a list of strings
• Output:
  • a dictionary where each key is a target error code and each value is the number of times it appears across all lines

Examples

Example 1

• Input: log_lines = ["INFO start", "WARN retry ERROR_TIMEOUT", "ERROR_TIMEOUT ERROR_DB_FAIL"], target_codes = ["ERROR_TIMEOUT", "ERROR_DB_FAIL"]
• Output: {"ERROR_TIMEOUT": 2, "ERROR_DB_FAIL": 1}
• Explanation: ERROR_TIMEOUT appears once in the second line and once in the third line; ERROR_DB_FAIL appears once.

Example 2

• Input: log_lines = ["ERROR_500 ERROR_500 ok", "noop", "ERROR_AUTH"], target_codes = ["ERROR_500", "ERROR_AUTH", "ERROR_MISSING"]
• Output: {"ERROR_500": 2, "ERROR_AUTH": 1, "ERROR_MISSING": 0}
• Explanation: Missing target codes must still appear in the result with count 0.

Constraints

• 1 <= len(log_lines) <= 10^5
• Total characters across all lines <= 10^7
• 1 <= len(target_codes) <= 10^4
• Each target code is unique