🐛 fix url for glad dataset (#577)

Author: philippemiron

clouddrift/adapters/glad.py

@@ -25,7 +25,7 @@
 
 def get_dataframe() -> pd.DataFrame:
     """Get the GLAD dataset as a pandas DataFrame."""
-    url = "https://data.gulfresearchinitiative.org/pelagos-symfony/api/file/download/169841"
+    url = "https://data.gulfresearchinitiative.org/api/file/download/169841"
     # GRIIDC server doesn't provide Content-Length header, so we'll hardcode
     # the expected data length here.
     file_size = 155330876

clouddrift/adapters/utils.py

@@ -23,6 +23,9 @@
 from tqdm import tqdm
 
 _DISABLE_SHOW_PROGRESS = False  # purely to de-noise our test suite output, should never be used/configured outside of that.
+_BROWSER_HEADERS = {
+    "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/126.0.0.0 Safari/537.36"
+}
 
 
 def _before_call(rcs: RetryCallState):
@@ -147,7 +150,7 @@ def _download_with_progress(
 
         # Get last modified time of the remote file
         try:
-            res = requests.head(url, timeout=5)
+            res = requests.head(url, headers=_BROWSER_HEADERS, timeout=10)
             remote_last_modified_str = res.headers.get("Last-Modified")
             if remote_last_modified_str:
                 remote_last_modified = datetime.strptime(
@@ -173,7 +176,7 @@ def _download_with_progress(
     buffer: BufferedWriter | BufferedIOBase | None = None
 
     try:
-        resp = requests.get(url, timeout=10, stream=True)
+        resp = requests.get(url, headers=_BROWSER_HEADERS, timeout=60, stream=True)
         temp_output = f"{output}.part" if isinstance(output, str) else None
 
         if isinstance(output, str) and temp_output is not None:

tests/adapters/utils_test.py

@@ -117,7 +117,7 @@ def test_download_new_update_buffer(self):
 
             # Assertions to verify the behavior
             self.requests_mock.get.assert_called_with(
-                "some.url.com", timeout=10, stream=True
+                "some.url.com", headers=utils._BROWSER_HEADERS, timeout=60, stream=True
             )
             buffer.write.assert_has_calls(
                 [call(b"a"), call(b"b"), call(b"c"), call(b"d")]
@@ -153,7 +153,7 @@ def test_download_new_update_file(self):
 
             # Assertions to verify the behavior
             self.requests_mock.get.assert_called_with(
-                "some.url.com", timeout=10, stream=True
+                "some.url.com", headers=utils._BROWSER_HEADERS, timeout=60, stream=True
             )
             self.open_mock.assert_called_with(output_file + ".part", "wb")
             handle = self.open_mock()
@@ -255,7 +255,7 @@ def test_progress_mechanism_enabled_file(self):
             tqdm_mock.assert_called_once()
             self.bar_mock.update.assert_has_calls([call(1), call(1), call(1), call(1)])
             self.requests_mock.get.assert_called_with(
-                "some.url.com", timeout=10, stream=True
+                "some.url.com", headers=utils._BROWSER_HEADERS, timeout=60, stream=True
             )
             self.open_mock.assert_called_with(output_file + ".part", "wb")
 
@@ -285,7 +285,7 @@ def test_progress_mechanism_disabled_file(self):
             tqdm_mock.assert_not_called()
             self.bar_mock.update.assert_not_called()
             self.requests_mock.get.assert_called_with(
-                "some.url.com", timeout=10, stream=True
+                "some.url.com", headers=utils._BROWSER_HEADERS, timeout=60, stream=True
             )
             self.open_mock.assert_called_with(output_file + ".part", "wb")