2023-01-05 04:12:15 +01:00
#!/usr/bin/env python3
import argparse
import html
import sys
2024-07-12 02:30:17 +02:00
from typing import NamedTuple
2023-01-05 04:12:15 +01:00
2024-11-14 01:30:36 +01:00
import orjson
2023-01-05 04:12:15 +01:00
class CLIArgs(NamedTuple):
unescape_html: bool
filename: str
indent_level: int
def parse_args() -> CLIArgs:
parser = argparse.ArgumentParser(
prog="unescape-contents",
description="Unescape Unicode and, optionally, HTML entities in a JSON file. Input file must be a JSON dictionary. Output will always be unescaped UTF-8.",
)
parser.add_argument("filename", type=str)
parser.add_argument(
"--unescape-html",
action="store_true",
dest="unescape_html",
help="If the key of a dictionary field does not contain HTML escapes, unescape any HTML escapes found in the value",
)
parser.add_argument("--indent-level", dest="indent_level", type=int, default=2, required=False)
args = parser.parse_args()
return CLIArgs(args.unescape_html, args.filename, args.indent_level)
if __name__ == "__main__":
args = parse_args()
print(f"unescaping file {args.filename}", file=sys.stderr)
2024-07-12 02:30:17 +02:00
json_data: dict[str, str] = {}
2023-01-05 04:12:15 +01:00
2024-11-14 01:30:36 +01:00
with open(args.filename, "rb") as source:
json_data = orjson.loads(source.read())
2023-01-05 04:12:15 +01:00
if args.unescape_html:
for key, value in json_data.items():
if key == html.unescape(key):
json_data[key] = html.unescape(value)
else:
print(
f'{args.filename}: key "{key}" contains HTML, not escaping HTML in value',
file=sys.stderr,
)
2024-11-14 01:30:36 +01:00
with open(args.filename, mode="wb") as dest:
# orjson ensures our output uses real UTF-8 codepoints for
# human readability, rather than \u0000 style escape
# sequences, providing us a somewhat-implicit JSON unescape.
dest.write(
orjson.dumps(
json_data,
option=orjson.OPT_APPEND_NEWLINE | orjson.OPT_INDENT_2 | orjson.OPT_SORT_KEYS,
)
)