In-mem compression / decompression and json serialization

compress_json.py

import base64
import io
import json
import urllib.request
import zipfile

from pathlib import Path
from urllib.parse import urlparse


def get_file_from_uri(uri):
    """Get a file from a URI and return it as a string."""
    return urllib.request.urlopen(uri).read().decode("utf-8")


# Create a buffer 
zip_buffer = io.BytesIO()

with zipfile.ZipFile(zip_buffer, "a", zipfile.ZIP_DEFLATED, allowZip64=False, compresslevel=9) as zip_file:

    for uri in urls:
        # Extract filename from URI
        filename = Path(urlparse(uri).path).name

        # GET file as str
        s = get_file_from_uri(uri)
        
        # Convert to bytes
        data = io.BytesIO(bytes(s, encoding="utf-8"))

        # Write bytes to zipfile as filename
        zip_file.writestr(filename, data=data.getvalue())

    # Validate zip archive
    zip_file.testzip()

# Write the zip file directly
with open("pre_json.zip", "wb") as f:
    f.write(zip_buffer.getvalue())
    

######### Convert to json #########

# Base64-encode the zipfile and convert it to string
zipped_str = base64.b64encode(zip_buffer.getvalue()).decode("ascii")

# Create json 
jsonstr = json.dumps({"result": zipped_str})

# Convert back from json, extract zipfile, and save zipfile to disk
with open("post_json.zip", "wb") as f:
    f.write(base64.b64decode(json.loads(jsonstr)["result"]))