import d6tstack

# fast CSV to SQL import - see SQL examples notebook
d6tstack.utils.pd_to_psql(df, 'postgresql+psycopg2:https://usr:pwd@localhost/db', 'tablename')
d6tstack.utils.pd_to_mysql(df, 'mysql+mysqlconnector:https://usr:pwd@localhost/db', 'tablename')
d6tstack.utils.pd_to_mssql(df, 'mssql+pymssql:https://usr:pwd@localhost/db', 'tablename') # experimental

# ingest mutiple CSVs which may have data schema changes - see CSV examples notebook

import glob
>>> c = d6tstack.combine_csv.CombinerCSV(glob.glob('data/*.csv'))

# show columns of each file
>>> c.columns()

# quick check if all files have consistent columns
>>> c.is_all_equal()
False

# show which files have missing columns
>>> c.is_column_present()
   filename  cost  date profit profit2 sales
0  feb.csv  True  True   True   False  True
2  mar.csv  True  True   True    True  True

>>> c.combine_preview() # keep all columns
   filename  cost        date profit profit2 sales
0   jan.csv  -80  2011-01-01     20     NaN   100
0   mar.csv  -100  2011-03-01    200     400   300

>>> d6tstack.combine_csv.CombinerCSV(glob.glob('*.csv'), columns_select_common=True).combine_preview() # keep common columns
   filename  cost        date profit sales
0   jan.csv  -80  2011-01-01     20   100
0   mar.csv  -100  2011-03-01    200   300

>>> d6tstack.combine_csv.CombinerCSV(glob.glob('*.csv'), columns_rename={'sales':'revenue'}).combine_preview()
   filename  cost        date profit profit2 revenue
0   jan.csv  -80  2011-01-01     20     NaN   100
0   mar.csv  -100  2011-03-01    200     400   300

# to come: check if columns match database
>>> c.is_columns_match_db('postgresql+psycopg2:https://usr:pwd@localhost/db', 'tablename')

# create csv with first nrows_preview rows of each file
>>> c.to_csv_head()

# export to csv, parquet, sql. Out of core with optimized fast imports for postgres and mysql
>>> c.to_pandas()
>>> c.to_csv_align(output_dir='process/')
>>> c.to_parquet_align(output_dir='process/')
>>> c.to_sql_combine('postgresql+psycopg2:https://usr:pwd@localhost/db', 'tablename')
>>> c.to_psql_combine('postgresql+psycopg2:https://usr:pwd@localhost/db', 'tablename') # fast, using COPY FROM
>>> c.to_mysql_combine('mysql+mysqlconnector:https://usr:pwd@localhost/db', 'tablename') # fast, using LOAD DATA LOCAL INFILE

# read Excel files - see Excel examples notebook for more details
import d6tstack.convert_xls

d6tstack.convert_xls.read_excel_advanced('test.xls',
    sheet_name='Sheet1', header_xls_range="B2:E2")

d6tstack.convert_xls.XLStoCSVMultiSheet('test.xls').convert_all(header_xls_range="B2:E2")

d6tstack.convert_xls.XLStoCSVMultiFile(glob.glob('*.xls'), 
    cfg_xls_sheets_sel_mode='name_global',cfg_xls_sheets_sel='Sheet1')
    .convert_all(header_xls_range="B2:E2")