traverse.py is a python script that recursively traverses through a directory and counts the number of files matching a given keyword/pattern in each subdirectory.

Script has dependencies on matplotlib.pyplot and regex modules, to ensure you have required dependencies, type

pip install matplotlib regex

To run, execute the script specifying arguments in the format:

./traverse.py <dir_path> <keyword> [-tx]

keyword is specified as a regex pattern

An optional -t specifier may be provided to make use of multithreaded processing when running (e.g. -t4 for 4 threads). This can speed up the routine if the directory structure is large.

The program will output to stdout a dictionary of key-values where key specifies a given subdirectory and the value the count of matching files found in the subdirectory (e.g. {’a/b’: 6, ’a/b/c’: 7, ‘a/b/c/d’:0}).

A bar chart is generated as filecount.png visualising the data output.

Test cases to consider when implementing a test-suite for this routine:

• Basic functional correctness – testing a simple directory structure and asserting the expected output (stdout and graph file generated)
• Testing empty directory - returns proper count (0) even when given an empty directory
• Testing different regex patterns – testing different operators (e.g. ?,*,.,{} etc.) produces correct count with matching and unmatching files
• Testing correct count – count should only count files and not directories that match the specified pattern
• Test invalid inputs correctly responds with error:
  • root_dir:
    • Specified directory does not exist
    • Specified path is a file
  • Regex
    • Check invalid regex patterns (e.g. '[')
• Test different directory path formats work – e.g. Unix systems using path/to/dir vs Windows C:\\path\to\dir
• Scale test – e.g. testing if large numbers of nested directories or large number of subdirs under root can cause program to crash (e.g. stack overflow in deep recursion).
• Test all the above works the same on multiple threads as on single thread