full import scripts

04_importArtistSourceReferenceAssignment.py

@@ -0,0 +1,87 @@
+import uuid # For UUID creation
+from initDb import initDb # For database initialization
+from wisski.api import Api, Pathbuilder, Entity # For WissKI API
+import os # For environment variable loading
+from dotenv import load_dotenv # For environment variable loading
+import pandas as pd # For dataframe handling
+
+# Initialize the database
+print('Initializing the database...')
+engine, metadata, Session = initDb(True, './schemas/')
+if engine == False:
+    print('Database initialization failed.')
+    exit()
+
+# Load the environment variables
+load_dotenv()
+
+# Initialize the WissKI API
+print('Initializing the WissKI API...')
+api_url = os.getenv('API_URL')
+auth = (os.getenv('API_USERNAME'), os.getenv('API_PASSWORD'))
+headers = {"Cache-Control": "no-cache"}
+api = Api(api_url, auth, headers)
+api.pathbuilder = api.get_pathbuilder('default')
+
+
+tableName = "c__81kr_que_kt_kue"
+bundleId = 'bf71940d0b18c20511e2141159afb9de' # Artist source reference assignment
+
+try:
+    processedRows = pd.read_csv(f'./logs/processed-{tableName}.csv')
+except FileNotFoundError:
+    processedRows = pd.DataFrame(columns=[ 'docId', 'uuid', 'uri'])
+
+# Load sources table
+sqlTable = pd.read_sql_table(tableName, con=engine)
+
+entityValues = {}
+
+# Create entities
+for index, row in sqlTable.iterrows():
+    # For every row in table...
+    if index < len(processedRows) and sqlTable.iloc[index, 0] == processedRows.iloc[index, 0]:
+        # skip if already processed
+        print(f'Skipping already processed entity {sqlTable.iloc[index, 0]}')
+        continue
+    # Create Entity property dicts
+    entityValues = {}
+    for key, value in row.items():
+        # For every column in row...
+        if (value is None) or (value == ''):
+            #  skip if cell has no value
+            continue
+        # Properties of an entity have to be an array, so...
+        if '&' in str(value):
+            # ...Explode "&"-separated values to array items
+            value = [x.strip() for x in str(value).split('&')]
+        else:
+            # ...Or parse to array
+            value = [value]
+        # Map columns to fields. We use assignments for reification.
+        docId = ''
+        match key:
+            case 'id':
+                continue
+            case 'f__uuid':
+                entityValues['fe3139ac03bd854ac9196fc240e7c68b'] = value # UUID
+                fUuid = value[0]
+            case 'f__8134_stelle':
+                entityValues['f58c13c5502baef24ede2a8a977ae6c6'] = value # Source reference
+            case 'f__81kr_que_kt_kue':
+                entityValues['f14d2d19f879d7398a384bdc132921a3'] = value # Source short title
+
+            case _:
+                print(f'{key} is not a valid field, skipping.')
+
+    # Create Material
+    entity = Entity(api=api, fields=entityValues, bundle_id=bundleId)
+    api.save(entity)
+
+    print(f'Created entity {index}: {entity.uri} of {len(tableName)}')
+
+    # Write log
+    processedRows = processedRows._append({'docId': docId, 'uuid': fUuid, 'uri': entity.uri}, ignore_index=True)
+    processedRows.to_csv(f'./logs/processed-{tableName}.csv', index=False)
+
+print('finish')