[py]: guess mimetypes in webserver for content serving (#16449)

Author: navin772

py/test/selenium/webdriver/common/webserver.py

@@ -24,6 +24,8 @@
 import re
 import threading
 
+import filetype
+
 try:
     from urllib import request as urllib_request
 except ImportError:
@@ -70,8 +72,20 @@ def _serve_page(self, page_number):
 
     def _serve_file(self, file_path):
         """Serve a file from the HTML root directory."""
-        with open(file_path, encoding="latin-1") as f:
-            return f.read().encode("utf-8")
+        with open(file_path, "rb") as f:
+            content = f.read()
+
+        kind = filetype.guess(content)
+        if kind is not None:
+            return content, kind.mime
+
+        # fallback for text files that filetype can't detect
+        if file_path.endswith(".txt"):
+            return content, "text/plain"
+        elif file_path.endswith(".json"):
+            return content, "application/json"
+        else:
+            return content, "text/html"
 
     def _send_response(self, content_type="text/html"):
         """Send a response."""
@@ -89,8 +103,7 @@ def do_GET(self):
                 self._send_response("text/html")
                 self.wfile.write(html)
             elif os.path.isfile(file_path):
-                content_type = "application/json" if file_path.endswith(".json") else "text/html"
-                content = self._serve_file(file_path)
+                content, content_type = self._serve_file(file_path)
                 self._send_response(content_type)
                 self.wfile.write(content)
             else:

py/test/selenium/webdriver/remote/remote_downloads_tests.py

@@ -28,11 +28,9 @@
 def test_get_downloadable_files(driver, pages):
     _browser_downloads(driver, pages)
     file_names = driver.get_downloadable_files()
-    # TODO: why is Chrome downloading files as .html???
-    # assert "file_1.txt" in file_names
-    # assert "file_2.jpg" in file_names
-    assert any(f in file_names for f in ("file_1.txt", "file_1.htm", "file_1.html"))
-    assert any(f in file_names for f in ("file_2.jpg", "file_2.htm", "file_2.html"))
+
+    assert "file_1.txt" in file_names
+    assert "file_2.jpg" in file_names
     assert type(file_names) is list
 
 
@@ -42,12 +40,8 @@ def test_download_file(driver, pages):
 
     # Get a list of downloadable files and find the txt file
     downloadable_files = driver.get_downloadable_files()
-    # TODO: why is Chrome downloading files as .html???
-    # text_file_name = next((file for file in downloadable_files if file.endswith(".txt")), None)
-    text_file_name = next(
-        (f for f in downloadable_files if all((f.endswith((".txt", ".htm", ".html")), f.startswith("file_1")))), None
-    )
-    assert text_file_name is not None, "Could not find file in downloadable files"
+    text_file_name = next((file for file in downloadable_files if file.endswith(".txt")), None)
+    assert text_file_name is not None, "Could not find a .txt file in downloadable files"
 
     with tempfile.TemporaryDirectory() as target_directory:
         driver.download_file(text_file_name, target_directory)
@@ -69,8 +63,4 @@ def _browser_downloads(driver, pages):
     pages.load("downloads/download.html")
     driver.find_element(By.ID, "file-1").click()
     driver.find_element(By.ID, "file-2").click()
-    # TODO: why is Chrome downloading files as .html???
-    # WebDriverWait(driver, 5).until(lambda d: "file_2.jpg" in d.get_downloadable_files())
-    WebDriverWait(driver, 5).until(
-        lambda d: any(f in d.get_downloadable_files() for f in ("file_2.jpg", "file_2.htm", "file_2.html"))
-    )
+    WebDriverWait(driver, 3).until(lambda d: "file_2.jpg" in d.get_downloadable_files())