From 64b507b411c73861c5bedd15388c5388bcc47935 Mon Sep 17 00:00:00 2001
From: Marcel Levstek <62072754+marcellevstek@users.noreply.github.com>
Date: Tue, 15 Apr 2025 10:24:05 +0200
Subject: [PATCH] Fix parsing of one-to-one metadata

for samples with no data objects of certain type
---
 src/resdk/tables/base.py | 36 ++++++++++++++++++++++++++----------
 test_meta.py             | 10 ++++++++++
 2 files changed, 36 insertions(+), 10 deletions(-)
 create mode 100644 test_meta.py

diff --git a/src/resdk/tables/base.py b/src/resdk/tables/base.py
index 2eb4c251..d880d181 100644
--- a/src/resdk/tables/base.py
+++ b/src/resdk/tables/base.py
@@ -351,7 +351,7 @@ def _get_relations(self) -> pd.DataFrame:
             for partition in relation.partitions:
                 value = ""
                 if partition["label"] and partition["position"]:
-                    value = f'{partition["label"]} / {partition["position"]}'
+                    value = f"{partition['label']} / {partition['position']}"
                 elif partition["label"]:
                     value = partition["label"]
                 elif partition["position"]:
@@ -371,6 +371,19 @@ def _get_orange_object(self) -> Data:
         )
 
     def _get_orange_data(self) -> pd.DataFrame:
+        def map_and_filter_samples(
+            df: pd.DataFrame, column_name: str, mapping: dict
+        ) -> pd.DataFrame:
+            """Map a predefined column name to a column named ``sample_id``, drop the original
+            and return the constructed DataFrame.
+            Omit samples for which mapping is not defined. These samples do not
+            have any data objects with the defined process type.
+            """
+            df = df[df[column_name].isin(mapping.keys())]
+            df["sample_id"] = df[column_name].map(mapping)
+            df = df.drop(columns=[column_name])
+            return df
+
         try:
             orange_meta = self._get_orange_object()
         except LookupError:
@@ -402,29 +415,32 @@ def _get_orange_data(self) -> pd.DataFrame:
             df = df.rename(columns={"mS#Sample ID": "sample_id"})
         elif "Sample slug" in df.columns:
             mapping = {s.slug: s.id for s in self._samples}
-            df["sample_id"] = [mapping[value] for value in df["Sample slug"]]
-            df = df.drop(columns=["Sample slug"])
+            df = map_and_filter_samples(
+                df=df, column_name="Sample slug", mapping=mapping
+            )
         elif "mS#Sample slug" in df.columns:
             mapping = {s.slug: s.id for s in self._samples}
-            df["sample_id"] = [mapping[value] for value in df["mS#Sample slug"]]
-            df = df.drop(columns=["mS#Sample slug"])
+            df = map_and_filter_samples(
+                df=df, column_name="mS#Sample slug", mapping=mapping
+            )
         elif "Sample name" in df.columns or "Sample name" in df.columns:
             mapping = {s.name: s.id for s in self._samples}
             if len(mapping) != len(self._samples):
                 raise ValueError(
                     "Duplicate sample names. Cannot map orange table data to other metadata"
                 )
-            df["sample_id"] = [mapping[value] for value in df["Sample name"]]
-            df = df.drop(columns=["Sample name"])
+            df = map_and_filter_samples(
+                df=df, column_name="Sample name", mapping=mapping
+            )
         elif "mS#Sample name" in df.columns:
             mapping = {s.name: s.id for s in self._samples}
             if len(mapping) != len(self._samples):
                 raise ValueError(
                     "Duplicate sample names. Cannot map orange table data to other metadata"
                 )
-            df["sample_id"] = [mapping[value] for value in df["mS#Sample name"]]
-            df = df.drop(columns=["mS#Sample name"])
-
+            df = map_and_filter_samples(
+                df=df, column_name="mS#Sample name", mapping=mapping
+            )
         return df.set_index("sample_id")
 
     def _download_metadata(self) -> pd.DataFrame:
diff --git a/test_meta.py b/test_meta.py
new file mode 100644
index 00000000..bb8f371f
--- /dev/null
+++ b/test_meta.py
@@ -0,0 +1,10 @@
+import resdk
+from resdk.tables import QCTables
+
+res = resdk.Resolwe(url="https://qa.genialis.io")
+res.login()
+
+collection = res.collection.get("validation-run-208")
+qt = QCTables(collection=collection)
+
+qt.meta