goodByeHida/18_importNumDating.py
2025-09-09 10:16:31 +02:00

71 lines
2.9 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
def importNumDating(api, engine):
print('Importing num dating...')
tableName = "c__5064_num__dat_"
bundleId = 'b9cfb95e627e1710cf8d736d4ca5db64' # Dating Information Assignment
try:
processedRows = pd.read_csv(f'./logs/{tableName}.csv')
except FileNotFoundError:
processedRows = pd.DataFrame(columns=[ 'id', 'uuid', 'uri'])
# Load sources table
sqlTable = pd.read_sql_table(tableName, con=engine)
# 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, 'id']:
# 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...
value = str(value).replace('&###{{new_line}}###'.format(), '&')
value = str(value).replace('###{{new_line}}###', '&')
value = str(value).replace(' & ', '&')
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['f74baaf58e49393cc89d6616ee197901'] = value # UUID
fUuid = value[0]
case 'f__5064_num__dat_':
entityValues['f0da3b36d16e16602bb550aff7d36297'] = value # Date
case 'f__50bm_bem__datierung':
entityValues['fe7870b5a86040d81140bccb01697765'] = value # Note
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 num dating {index}: {entity.uri} of {len(tableName)}')
# Write log
processedRows = processedRows._append({'id': row['id'], 'uuid': fUuid, 'uri': entity.uri}, ignore_index=True)
processedRows.to_csv(f'./logs/{tableName}.csv', index=False)
print('finish')