unescapeURI

#!/usr/bin/env python3
#
# unescapeURI: Handle % and + escape in URLs, even for Unicode.
# 2024-04-14: Written by Steven J. DeRose.
#
from urllib.parse import unquote

__metadata__ = {
    "title"        : "unescapeURI",
    "description"  : "Handle % and + escape in URLs, even for Unicode.",
    "rightsHolder" : "Steven J. DeRose",
    "creator"      : "http://viaf.org/viaf/50334488",
    "type"         : "http://purl.org/dc/dcmitype/Software",
    "language"     : "Python 3.9",
    "created"      : "2024-04-14",
    "modified"     : "2024-04-14",
    "publisher"    : "http://github.com/sderose",
    "license"      : "https://creativecommons.org/licenses/by-sa/3.0/"
}
__version__ = __metadata__["modified"]

descr = """
=Name=
unescapeURI: Handle % and + escape in URLs, even for Unicode.


=Description=

==Usage==

    unescapeURI.py [options] [s]

Resolve and %xx escapes in s, Interpret the result as utf-8, and display.
If --plus is set, also turn any unescaped "+" into spaces.

=See also=

My `escapeURI`.


=History=

* 2024-04-14: Written by Steven J. DeRose to replace old Perl version.


=Rights=

By Steven J. DeRose. This work is dedicated to the public domain.

For the most recent version, see [http://www.derose.net/steve/utilities]
or [https://github.com/sderose].


=Options=
"""


###############################################################################
# Main
#
if __name__ == "__main__":
    import argparse

    def processOptions() -> argparse.Namespace:
        try:
            from BlockFormatter import BlockFormatter
            parser = argparse.ArgumentParser(
                description=descr, formatter_class=BlockFormatter)
        except ImportError:
            parser = argparse.ArgumentParser(description=descr)

        parser.add_argument(
            "--plus", action="store_true",
            help="Also turn (unescaped) '+' into space.")
        parser.add_argument(
            "--quiet", "-q", action="store_true",
            help="Suppress most messages.")
        parser.add_argument(
            "--version", action="version", version=__version__,
            help="Display version information, then exit.")

        parser.add_argument(
            "uri", type=str, nargs=argparse.REMAINDER,
            help="A percent-escaped string to fix.")

        args0 = parser.parse_args()

        return(args0)


    ###########################################################################
    #
    args = processOptions()

    if (not args.uri):
        args.uri = "https://example.edu/foo+bar%e2%80%a2+baz"
        print("No string supplied. For example:\n%s  ==>" % (args.uri))

    s = args.uri
    if (args.plus): s = s.replace("+", " ")
    s = unquote(s)
    s = s.encode('utf-8').decode('utf-8')
    print(s)