Merge pull request #1 from 3droj7/master (Fix incorrect handling of byte responses (e.g., images) by decoding base64 data)

This PR fixes an issue where binary responses (such as images) were not handled correctly in `python-tls-client-async`. Previously, the response body was treated as plain text, leading to corrupted data when downloading binary content like PNGs or PDFs.

The root cause was that the underlying Go library returns binary responses encoded in base64 if `isByteResponse=true`, but the Python code did not decode this properly.

This patch:
- Modifies `response.py` to split and decode base64-encoded bodies.
- Ensures correct handling of binary responses (e.g., images).
- Sets `"isByteResponse": True` in the session payload for consistency.

Related issues:
- FlorianREGAZ#83
- FlorianREGAZ#119
- FlorianREGAZ#117
- FlorianREGAZ#122

Thanks to @3droj7 for PR.

Author: diprog

async_tls_client/response.py

@@ -3,6 +3,7 @@
 
 from typing import Union
 import json
+import base64
 
 
 class Response:
@@ -72,7 +73,7 @@ def build_response(res: Union[dict, list], res_cookies: RequestsCookieJar) -> Re
     # Add cookies
     response.cookies = res_cookies
     # Add response body
-    response.text = res["body"]
+    response.text = base64.b64decode(res["body"].split(',')[1]).decode(errors='ignore')
     # Add response content (bytes)
-    response._content = res["body"].encode()
+    response._content = base64.b64decode(res["body"].split(',')[1])
     return response

async_tls_client/sessions.py

@@ -427,6 +427,7 @@ def build_payload():
                 "headerOrder": self.header_order,
                 "insecureSkipVerify": insecure_skip_verify,
                 "isByteRequest": is_byte_request,
+                "isByteResponse": True,
                 "additionalDecode": self.additional_decode,
                 "proxyUrl": final_proxy,
                 "requestUrl": final_url,