Python function - test if given file is considered binary.

README.md

Python function - is file binary?

Function which determines if a given file is binary.

Test is based on the following algorithm (similar to that implemented within Perl):

• Empty files are considered text.
• If not empty, read up to 512 bytes as a buffer. File will be binary if:
  • Null byte is encountered.
  • More than 30% of the buffer consists of "non text" characters.
• Otherwise, file is text.

Reference

• https://eli.thegreenplace.net/2011/10/19/perls-guess-if-file-is-text-or-binary-implemented-in-python
• https://stackoverflow.com/questions/1446549/how-to-identify-binary-and-text-files-using-python/1446870#1446870

isbinary.py

#!/usr/bin/env python

class IsFileBinary:
    READ_BYTES = 512
    CHAR_THRESHOLD = 0.3
    TEXT_CHARACTERS = "".join(
        [chr(code) for code in range(32, 127)] + list("\b\f\n\r\t")
    )

    def test(self, file_path):
        # read chunk of file
        fh = open(file_path, "r")
        file_data = fh.read(IsFileBinary.READ_BYTES)
        fh.close()

        # store chunk length read
        data_length = len(file_data)
        if not data_length:
            # empty files considered text
            return False

        if "\x00" in file_data:
            # file containing null bytes is binary
            return True

        # remove all text characters from file chunk, get remaining length
        binary_length = len(file_data.translate(None, IsFileBinary.TEXT_CHARACTERS))

        # if percentage of binary characters above threshold, binary file
        return (float(binary_length) / data_length) >= IsFileBinary.CHAR_THRESHOLD


def main():
    is_file_binary = IsFileBinary()

    print("Is binary file: {0}".format(is_file_binary.test("./first")))
    print("Is binary file: {0}".format(is_file_binary.test("./second")))
    print("Is binary file: {0}".format(is_file_binary.test("./third")))


if __name__ == "__main__":
    main()

If all you know is OOP, what you do every day is adjusting your problem to a predetermined solution.

one of the strangest comments I've read for a while. 😄