Reconfigure locate_data to bypass execute method to create a post request instead of get.

Author: d-giles

astroquery/heasarc/core.py

@@ -518,34 +518,17 @@ def locate_data(self, query_result=None, catalog_name=None):
 
         # datalink url
         dlink_url = f'{self.VO_URL}/datalink/{catalog_name}'
-        def query_func(query_chunk):
-                return pyvo.dal.adhoc.DatalinkQuery(
-                    baseurl=dlink_url,
-                    id=query_chunk['__row'],
-                    session=self._session
-                )
+        query = pyvo.dal.adhoc.DatalinkQuery(
+            baseurl=dlink_url,
+            id=query_result['__row'],
+            session=self._session
+            )
 
-        # Standard URL limit of ~2000 characters for GET
-        # With query formula https://[urlbase]?ID=###, could get 100 16 digit IDs 
-        URI_MAX = 100
-        if len(query_result) > URI_MAX:
-            # execute a set amount at a time
-            chunks = len(query_result) % URI_MAX
-            for chunk in range(chunks):
-                query_chunk = query_result[URI_MAX*chunk: URI_MAX*chunk+URI_MAX]
-                query = query_func(query_chunk)
-                dl_result_chunk = query.execute().to_table()
-                if chunk == 0:
-                    dl_result = dl_result_chunk
-                else:
-                    dl_result = vstack([dl_result, dl_result_chunk])
-            query_chunk = query_result[URI_MAX*chunk+URI_MAX:]
-            query = query_func(query_chunk)
-            dl_result_chunk = query.execute().to_table()
-            dl_result = vstack([dl_result, dl_result_chunk])
-        else:
-            query = query_func(query_result)
-            dl_result = query.execute().to_table()
+        dl_result = pyvo.dal.DALResults(
+            query.execute_votable(post=True),
+            url=query.queryurl,
+            session=query._session
+            ).to_table()
 
         # include rows that have directory links (i.e. data) and those
         # that report errors (usually means there are no data products)