Implement 5 MB file size limit enforcement

Enhanced both FicheServer and LinesServer to reject data transfers exceeding a 5MB limit by introducing a max_size parameter across server classes, request handlers, and CLI arguments. This change improves system stability and security by preventing excessive resource use. Additionally, customized HTTP responses are provided for missing or invalid Content-Length headers and oversized files, creating a clearer client-server communication and ensuring better request validation logic.

src/pyfiche/classes/fiche.py

@@ -21,6 +21,7 @@ class FicheServer:
     slug_size: int = 8
     https: bool = False
     buffer_size: int = 4096
+    max_size: int = 5242880 # 5 MB by default
     _output_dir: pathlib.Path = pathlib.Path('data/')
     _log_file: Optional[pathlib.Path] = None
     _banlist: Optional[pathlib.Path] = None
@@ -92,6 +93,7 @@ class FicheServer:
         fiche.banlist = args.banlist or fiche.banlist
         fiche.allowlist = args.allowlist or fiche.allowlist
         fiche.buffer_size = args.buffer_size or fiche.buffer_size
+        fiche.max_size = args.max_size or fiche.max_size
 
         fiche.logger = logging.getLogger('pyfiche')
         fiche.logger.setLevel(logging.INFO if not args.debug else logging.DEBUG)
@@ -181,6 +183,11 @@ class FicheServer:
 
                     received_data.extend(data)
 
+                    if len(received_data) > self.max_size:
+                        self.logger.error(f"Received data exceeds maximum size ({self.max_size} bytes), terminating connection.")
+                        conn.sendall(b"Data exceeds maximum size.\n")
+                        return
+
             except socket.timeout:
                 pass
 

src/pyfiche/classes/lines.py

@@ -67,7 +67,13 @@ body {{
         # Reject any POST requests that don't have a Content-Length header
 
         if "Content-Length" not in self.headers:
-            return self.not_found()
+            return self.invalid_request()
+
+        if not self.headers["Content-Length"].isdigit():
+            return self.invalid_request()
+
+        if int(self.headers["Content-Length"]) > self.max_size:
+            return self.file_too_large()
 
         # Upload the file
 
@@ -91,7 +97,17 @@ body {{
         self.send_response(303)
         self.send_header("Location", f"/{slug}")
         self.end_headers()
-        
+
+    def invalid_request(self):
+        self.send_response(400)
+        self.end_headers()
+        self.wfile.write(b"Invalid request")
+
+    def file_too_large(self):
+        self.send_response(413)
+        self.end_headers()
+        self.wfile.write(b"File too large")
+
     def not_found(self):
         self.send_response(404)
         self.end_headers()
@@ -211,13 +227,16 @@ body {{
             self.wfile.write(full_html.encode("utf-8"))
 
 
-def make_lines_handler(data_dir, logger, banlist=None, allowlist=None):
+def make_lines_handler(
+    data_dir, logger, banlist=None, allowlist=None, max_size=5242880
+):
     class CustomHandler(LinesHTTPRequestHandler):
         def __init__(self, *args, **kwargs):
             self.data_dir: pathlib.Path = data_dir
             self.logger: logging.Logger = logger
             self.banlist: Optional[pathlib.Path] = banlist
             self.allowlist: Optional[pathlib.Path] = allowlist
+            self.max_size: int = max_size
 
             super().__init__(*args, **kwargs)
 
@@ -227,6 +246,7 @@ def make_lines_handler(data_dir, logger, banlist=None, allowlist=None):
 class LinesServer:
     port: int = 9997
     listen_addr: str = "0.0.0.0"
+    max_size: int = 5242880  # 5 MB by default
     _data_dir: pathlib.Path = pathlib.Path("data/")
     _log_file: Optional[pathlib.Path] = None
     _banlist: Optional[pathlib.Path] = None
@@ -295,6 +315,7 @@ class LinesServer:
 
         lines.port = args.port or lines.port
         lines.listen_addr = args.listen_addr or lines.listen_addr
+        lines.max_size = args.max_size or lines.max_size
         lines.data_dir = args.data_dir or lines.data_dir
         lines.log_file = args.log_file or lines.log_file
         lines.banlist = args.banlist or lines.banlist
@@ -316,7 +337,7 @@ class LinesServer:
 
     def run(self):
         handler_class = make_lines_handler(
-            self.data_dir, self.logger, self.banlist, self.allowlist
+            self.data_dir, self.logger, self.banlist, self.allowlist, self.max_size
         )
 
         with HTTPServer((self.listen_addr, self.port), handler_class) as httpd:

src/pyfiche/fiche_server.py

@@ -17,6 +17,7 @@ def main():
     parser.add_argument('-S', '--https', action='store_true', help='HTTPS (requires reverse proxy)')
     parser.add_argument('-o', '--output_dir', help='Output directory path (default: data/)')
     parser.add_argument('-B', '--buffer_size', type=int, help='Buffer size (default: 4096)')
+    parser.add_argument('-M', '--max_size', type=int, help='Maximum file size (in bytes) (default: 5242880)')
     parser.add_argument('-l', '--log_file', help='Log file path (default: None - log to stdout)')
     parser.add_argument('-b', '--banlist', help='Banlist file path')
     parser.add_argument('-w', '--allowlist', help='Allowlist file path')

src/pyfiche/lines_server.py

@@ -16,6 +16,7 @@ def main():
     parser.add_argument('-l', '--log_file', help='Log file path (default: None - log to stdout)')
     parser.add_argument('-b', '--banlist', help='Banlist file path')
     parser.add_argument('-w', '--allowlist', help='Allowlist file path')
+    parser.add_argument('-M', '--max_size', type=int, help='Maximum file size (in bytes) (default: 5242880)')
     parser.add_argument('-D', '--debug', action='store_true', help='Debug mode')
 
     # Parse the arguments