Merge pull request #152 from umccr/dev

Changelog.md

@@ -4,6 +4,18 @@ Changes in this log refer only to changes that make it to the 'main' branch. and
 
 For changes in deployment, please see the [deployment changelog](deploy/cttso-ica-to-pieriandx-cdk/Changelog.md) 
 
+
+## 2023-10-18
+
+> Author: Alexis Lucattini
+> Email: [[email protected]](mailto:[email protected])
+
+
+### Bugfixes
+
+* Update sample type trimming of _sample suffix (https://github.com/umccr/cttso-ica-to-pieriandx/pull/150)
+  * Resolves https://github.com/umccr/cttso-ica-to-pieriandx/issues/149
+
 ## 2023-10-18
 
 > Author: Alexis Lucattini

deploy/cttso-ica-to-pieriandx-cdk/Changelog.md

@@ -8,6 +8,17 @@ are nested under deploy/cttso-ica-to-pieriandx-cdk.
 > Author: Alexis Lucattini
 > Email: [[email protected]](mailto:[email protected])
 
+
+### Bugfixes
+
+* Dont resubmit deleted samples - also remove deleted samples from the lims sheet (https://github.com/umccr/cttso-ica-to-pieriandx/pull/151)
+  * Resolves https://github.com/umccr/cttso-ica-to-pieriandx/issues/146 
+
+## 2023-10-18
+
+> Author: Alexis Lucattini
+> Email: [[email protected]](mailto:[email protected])
+
 ### Enhancements
 
 * Move to project owner / project name mapping logic (https://github.com/umccr/cttso-ica-to-pieriandx/pull/141)

deploy/cttso-ica-to-pieriandx-cdk/lambdas/launch_available_payloads_and_update_cttso_lims_sheet/lambda_code.py

@@ -249,6 +249,24 @@ def get_libraries_for_processing(merged_df) -> pd.DataFrame:
             columns=processing_columns
         )
 
+    # Check none of the processing df libraries are in the list of deleted lists
+    deleted_lims_df, deleted_lims_excel_row_mapping_number = get_deleted_lims_df()
+    process_row: pd.Series
+    already_deleted_list_index = []
+    for index, process_row in to_process_df.iterrows():
+        if not deleted_lims_df.query(
+            f"subject_id == '{process_row['subject_id']}' and "
+            f"library_id == '{process_row['library_id']}' and "
+            f"portal_wfr_id == '{process_row['portal_wfr_id']}'"
+        ).shape[0] == 0:
+            already_deleted_list_index.append(index)
+            logger.warning(f"Already run and deleted this combination {process_row['subject_id']} / {process_row['library_id']} / {process_row['portal_wfr_id']}, not reprocessing")
+
+    # Delete via index
+    to_process_df = to_process_df.iloc[list(
+        set(to_process_df.index.tolist()) - set(already_deleted_list_index)
+    )]
+
     # Update columns to strip glims_ attributes
     new_column_names = [
       "panel",
@@ -1861,7 +1879,7 @@ def bind_pieriandx_case_submission_time_to_merged_df(merged_df: pd.DataFrame, ct
     return merged_lims_df
 
 
-def drop_to_be_deleted_cases(merged_df: pd.DataFrame, cttso_lims_df: pd.DataFrame, excel_row_mapping_number_df: pd.DataFrame) -> (pd.DataFrame, pd.DataFrame, pd.DataFrame):
+def drop_to_be_deleted_cases(merged_df: pd.DataFrame, cttso_lims_df: pd.DataFrame, excel_row_mapping_number_df: pd.DataFrame, existing_pieriandx_cases: List) -> (pd.DataFrame, pd.DataFrame, pd.DataFrame):
     """
     Cases that have been assigned to ToBeDeleted need to be dropped from row list
     and instead attached to a new sheet
@@ -1928,17 +1946,44 @@ def drop_to_be_deleted_cases(merged_df: pd.DataFrame, cttso_lims_df: pd.DataFram
       A pandas DataFrame with the following columns:
         * cttso_lims_index
         * excel_row_number
+    :param existing_pieriandx_cases:
+      A list of pieriandx cases we have
     :return:
     """
     # Split cttso lims df by query
-    to_be_deleted_cases_lims = cttso_lims_df.query("pieriandx_assignee == 'ToBeDeleted'")
+    to_be_deleted_cases_lims = cttso_lims_df.query(
+        "pieriandx_assignee == 'ToBeDeleted' or "
+        "( "
+        "  pieriandx_case_id not in @existing_pieriandx_cases and "
+        "  not pieriandx_case_id.isnull()"
+        ")",
+        engine="python"
+    )
     to_be_deleted_cases_merged_df = merged_df.query("pieriandx_assignee == 'ToBeDeleted'")
 
-    if to_be_deleted_cases_lims.shape[0] == 0 and to_be_deleted_cases_merged_df.shape[0] == 0:
+    # Check cases that are not in pieriandx
+    if len(existing_pieriandx_cases) == 0:
+        logger.error("Something seriously wrong has happened! Got an empty list of cases")
+        raise ValueError
+    # Cases that have already been deleted are ones in the lims df that cannot be found in pieriandx
+    cases_already_deleted = list(
+        set(
+            cttso_lims_df["pieriandx_case_id"].dropna().tolist()
+        ) -
+        set(
+            existing_pieriandx_cases
+        )
+    )
+
+    if to_be_deleted_cases_lims.shape[0] == 0 and to_be_deleted_cases_merged_df.shape[0] == 0 and len(cases_already_deleted) == 0:
         logger.info("Nothing to transfer to delete pile")
         return merged_df, cttso_lims_df, excel_row_mapping_number_df
 
-    cttso_lims_df_cleaned = cttso_lims_df.query("pieriandx_assignee != 'ToBeDeleted'")
+    # Remove cases assigned to the ToBeDeleted user or cases that have already been deleted
+    cttso_lims_df_cleaned = cttso_lims_df.query(
+        "pieriandx_assignee != 'ToBeDeleted' and not "
+        "pieriandx_case_id in @cases_already_deleted"
+    )
     clean_cttso_case_ids_list = cttso_lims_df_cleaned["pieriandx_case_id"].tolist()
     deleted_case_ids_list = to_be_deleted_cases_lims["pieriandx_case_id"].tolist()
 
@@ -1963,6 +2008,12 @@ def drop_to_be_deleted_cases(merged_df: pd.DataFrame, cttso_lims_df: pd.DataFram
         engine="python"
     )
 
+    # Remove cases from merged df that have already been deleted
+    merged_df = merged_df.query(
+        "pieriandx_case_id not in @cases_already_deleted",
+        engine="python"
+    )
+
     # Update cttso lims sheet with replacement
     append_df_to_cttso_lims(cttso_lims_df_cleaned, replace=True)
     # Wait for doc population
@@ -2109,7 +2160,7 @@ def lambda_handler(event, context):
 
     # Clean out to-be-deleted cases
     merged_df, cttso_lims_df, excel_row_number_mapping_df = \
-        drop_to_be_deleted_cases(merged_df, cttso_lims_df, excel_row_number_mapping_df)
+        drop_to_be_deleted_cases(merged_df, cttso_lims_df, excel_row_number_mapping_df, pieriandx_df["pieriandx_case_id"].tolist())
 
     merged_df, cttso_lims_df, excel_row_number_mapping_df = \
         cleanup_duplicate_rows(merged_df, cttso_lims_df, excel_row_number_mapping_df)

deploy/cttso-ica-to-pieriandx-cdk/lambdas/layers/lambda_utils/pieriandx_helpers.py

@@ -12,6 +12,8 @@
 import pandas as pd
 import time
 
+from pyriandx.utils import retry_session
+
 from .globals import \
     PIERIANDX_CDK_SSM_LIST, \
     PIERIANDX_CDK_SSM_PATH, \
@@ -352,6 +354,40 @@ def validate_case_accession_number(subject_id: str, library_id: str, case_access
         raise ValueError
 
 
+def check_case_exists(case_id: str) -> bool:
+    """
+    Check a case actually exists and has not been deleted
+    :param case_id:
+    :return:
+    """
+    email, auth_token, institution, base_url = get_pieriandx_env_vars()
+
+    pyriandx_client = get_pieriandx_client(
+        email=email,
+        auth_token=auth_token,
+        institution=institution,
+        base_url=base_url
+    )
+
+    # Go around _get_api to collect error code if it exists
+    url = pyriandx_client.baseURL + f"/case/{case_id}"
+    response = retry_session(pyriandx_client.headers).get(url, params=None)
+
+    if response.status_code == 200:
+        return True
+
+    if response.status_code == 400:
+        logger.info(f"Case {case_id} is not found, it may have been deleted")
+        return False
+
+    if response.status_code == 401:
+        logger.error("Got unauthorized status code 401. Cannot continue with script")
+        raise ChildProcessError
+
+    logger.warning(f"Got status_code {response.status_code}. Assuming case does not exist")
+    return False
+
+
 def get_pieriandx_status_for_missing_sample(case_id: str) -> pd.Series:
     """
     Get pieriandx results for a sample with incomplete results
@@ -382,7 +418,7 @@ def get_pieriandx_status_for_missing_sample(case_id: str) -> pd.Series:
         iter_count += 1
         if iter_count >= MAX_ATTEMPTS_GET_CASES:
             logger.error(f"Tried to get all cases {str(MAX_ATTEMPTS_GET_CASES)} times and failed")
-            raise EnvironmentError
+            raise ChildProcessError
 
         # Attempt to get cases
         response: Dict = pyriandx_client._get_api(endpoint=f"/case/{case_id}")