90 lines
3.3 KiB
Python
90 lines
3.3 KiB
Python
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__8490_fotograf"
|
|
bundleId = 'b821fb6c518948b7f40d17803b6ce293' # Photographer 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.loc[index, 'id'] == processedRows.loc[index, 'docId']:
|
|
# skip if already processed
|
|
print(f'Skipping already processed entity {sqlTable.loc[index, 'id']}')
|
|
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':
|
|
docId = value[0]
|
|
case 'f__uuid':
|
|
entityValues['f6c3c3e35af2f2073fd517aabf88fa7c'] = value # UUID
|
|
docUuid = value[0]
|
|
case 'f__8490_fotograf':
|
|
entityValues['fe8f8b235f896862b74caa0fa8f3682d'] = value # Photographer
|
|
case 'f__8494_aufn_datum':
|
|
entityValues['f12c7538643314f0f46ba76a5140a87d'] = value # Recording Date
|
|
case 'f__8470_aufnahmenr_':
|
|
entityValues['ff6ec986fb4cc5a2f34deb7144f2f817'] = value # Recording number
|
|
case 'f__849r_repro_datei': # Image Assignment
|
|
entityValues['f24a609593559a904a0a0f2e215db584'] = value # Reproduction Number
|
|
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(sqlTable)}')
|
|
|
|
# Write log
|
|
processedRows = processedRows._append({'docId': docId, 'uuid': docUuid, 'uri': entity.uri}, ignore_index=True)
|
|
processedRows.to_csv(f'./logs/processed-{tableName}.csv', index=False)
|
|
|
|
print('finish')
|