From 0321a08cf37a79adf1d0a2091c26220c46e4e2df Mon Sep 17 00:00:00 2001
From: Santiago Soler <santisoler@fastmail.com>
Date: Mon, 13 Mar 2023 11:10:43 -0700
Subject: [PATCH 1/4] Support Dataverse files without a persistentID

Add support for downloading Dataverse files that don't have a persistent
ID. Use the file ID instead.
---
 pooch/downloaders.py | 37 ++++++++++++++++++++++++-------------
 1 file changed, 24 insertions(+), 13 deletions(-)

diff --git a/pooch/downloaders.py b/pooch/downloaders.py
index f391fb19..ccc8c48c 100644
--- a/pooch/downloaders.py
+++ b/pooch/downloaders.py
@@ -1034,20 +1034,31 @@ def download_url(self, file_name):
         download_url : str
             The HTTP URL that can be used to download the file.
         """
-
         parsed = parse_url(self.archive_url)
-
-        # Iterate over the given files until we find one of the requested name
-        for filedata in self.api_response.json()["data"]["latestVersion"]["files"]:
-            if file_name == filedata["dataFile"]["filename"]:
-                return (
-                    f"{parsed['protocol']}://{parsed['netloc']}/api/access/datafile/"
-                    f":persistentId?persistentId={filedata['dataFile']['persistentId']}"
-                )
-
-        raise ValueError(
-            f"File '{file_name}' not found in data archive {self.archive_url} (doi:{self.doi})."
-        )
+        response = self.api_response.json()
+        files = {
+            file["dataFile"]["filename"]: file["dataFile"]
+            for file in response["data"]["latestVersion"]["files"]
+        }
+        if file_name not in files:
+            raise ValueError(
+                f"File '{file_name}' not found in data archive "
+                f"{self.archive_url} (doi:{self.doi})."
+            )
+        # Generate download_url using persistentId or file id
+        persistent_id = files[file_name]["persistentId"]
+        if persistent_id:
+            download_url = (
+                f"{parsed['protocol']}://{parsed['netloc']}/api/access/datafile/"
+                f":persistentId?persistentId={persistent_id}"
+            )
+        else:
+            file_id = files[file_name]["id"]
+            download_url = (
+                f"{parsed['protocol']}://{parsed['netloc']}/api/access/datafile/"
+                f"{file_id}"
+            )
+        return download_url
 
     def populate_registry(self, pooch):
         """

From 7ad5073b2f5e99481edbdd5e86e1fa0c8dc27c36 Mon Sep 17 00:00:00 2001
From: Santiago Soler <santisoler@fastmail.com>
Date: Mon, 20 Mar 2023 14:29:47 -0700
Subject: [PATCH 2/4] Don't assume that persistentId is always present

The `persistentId` key might be missing in the API response, while the
`ID` is always there. So, don't assume it exists when deciding which id
should be used to download the files.
---
 pooch/downloaders.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/pooch/downloaders.py b/pooch/downloaders.py
index ccc8c48c..d7a85a08 100644
--- a/pooch/downloaders.py
+++ b/pooch/downloaders.py
@@ -1046,8 +1046,8 @@ def download_url(self, file_name):
                 f"{self.archive_url} (doi:{self.doi})."
             )
         # Generate download_url using persistentId or file id
-        persistent_id = files[file_name]["persistentId"]
-        if persistent_id:
+        persistent_id = files[file_name].get("persistentId")
+        if persistent_id is not None and persistent_id:
             download_url = (
                 f"{parsed['protocol']}://{parsed['netloc']}/api/access/datafile/"
                 f":persistentId?persistentId={persistent_id}"

From 1e670cfcd76bc230358cae3985847a90306bfac3 Mon Sep 17 00:00:00 2001
From: Santiago Soler <santisoler@fastmail.com>
Date: Mon, 20 Mar 2023 17:38:06 -0700
Subject: [PATCH 3/4] Simplify the if statement

Both for a persistent_id as a None or as an empty string, we can
evaluate them with `if persistent_id:`.
---
 pooch/downloaders.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/pooch/downloaders.py b/pooch/downloaders.py
index d7a85a08..fe1af254 100644
--- a/pooch/downloaders.py
+++ b/pooch/downloaders.py
@@ -1047,7 +1047,7 @@ def download_url(self, file_name):
             )
         # Generate download_url using persistentId or file id
         persistent_id = files[file_name].get("persistentId")
-        if persistent_id is not None and persistent_id:
+        if persistent_id:
             download_url = (
                 f"{parsed['protocol']}://{parsed['netloc']}/api/access/datafile/"
                 f":persistentId?persistentId={persistent_id}"

From 15f7536563d02224042426ecf0b69a6cd7bd8944 Mon Sep 17 00:00:00 2001
From: Leonardo Uieda <leo@uieda.com>
Date: Mon, 19 Feb 2024 16:59:46 -0300
Subject: [PATCH 4/4] Only rely on the file ID, not PID

---
 pooch/downloaders.py | 18 +++++-------------
 1 file changed, 5 insertions(+), 13 deletions(-)

diff --git a/pooch/downloaders.py b/pooch/downloaders.py
index 56fbbffb..fa3b52da 100644
--- a/pooch/downloaders.py
+++ b/pooch/downloaders.py
@@ -1134,19 +1134,11 @@ def download_url(self, file_name):
                 f"File '{file_name}' not found in data archive "
                 f"{self.archive_url} (doi:{self.doi})."
             )
-        # Generate download_url using persistentId or file id
-        persistent_id = files[file_name].get("persistentId")
-        if persistent_id:
-            download_url = (
-                f"{parsed['protocol']}://{parsed['netloc']}/api/access/datafile/"
-                f":persistentId?persistentId={persistent_id}"
-            )
-        else:
-            file_id = files[file_name]["id"]
-            download_url = (
-                f"{parsed['protocol']}://{parsed['netloc']}/api/access/datafile/"
-                f"{file_id}"
-            )
+        # Generate download_url using the file id
+        download_url = (
+            f"{parsed['protocol']}://{parsed['netloc']}/api/access/datafile/"
+            f"{files[file_name]['id']}"
+        )
         return download_url
 
     def populate_registry(self, pooch):