Add a new test pattern for asserting command outputs and exit status
through a run() wrapper function and corresponding assertion functions.

Addresses issue: bash-unit#134
--------------------------------------
User story suggested pipe syntax:
  command_under_test | should_display "${expected_pattern}"
  command_under_test | err_should_contain "${expected_pattern}"

Implemented wrapper pattern:
  run command_under_test && assert_run_output_matches "${expected_pattern}"
  run command_under_test && assert_run_error_matches "${expected_pattern}"

Works also as separate lines:
  run command_under_test
  assert_run_output_matches "${expected_pattern}"
  assert_run_error_matches "${expected_pattern}"

The wrapper pattern achieves the same readability while providing more
flexibility and reliability.

Problem with pipes:
- pipes carry only stdout
- harder to resolve exit status if needed
- can't verify multiple conditions independently

Usage Examples
--------------
Simple one-liners:
  run my_script && assert_run_success
  run my_command && assert_run_output_equals "expected"
  run failing_cmd && assert_run_status_code 42

Multiple assertions:
  run my_script.sh input.txt
  assert_run_success
  assert_run_output_matches "^Processed.*successfully"
  assert_run_error_equals ""