bug fix: etl engine, sqlite inserts
This commit is contained in:
parent
2de6e51bdb
commit
ccc05acc01
|
@ -16,6 +16,16 @@ THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLI
|
|||
Usage :
|
||||
transport --config <path-to-file.json> --procs <number-procs>
|
||||
@TODO: Create tables if they don't exist for relational databases
|
||||
example of configuration :
|
||||
|
||||
1. Move data from a folder to a data-store
|
||||
transport [--folder <path> ] --config <config.json> #-- assuming the configuration doesn't have folder
|
||||
transport --folder <path> --provider <postgresql|mongo|sqlite> --<database|db> <name> --table|doc <document_name>
|
||||
In this case the configuration should look like :
|
||||
{folder:..., target:{}}
|
||||
2. Move data from one source to another
|
||||
transport --config <file.json>
|
||||
{source:{..},target:{..}} or [{source:{..},target:{..}},{source:{..},target:{..}}]
|
||||
|
||||
|
||||
"""
|
||||
|
@ -46,11 +56,23 @@ if len(sys.argv) > 1:
|
|||
class Post(Process):
|
||||
def __init__(self,**args):
|
||||
super().__init__()
|
||||
|
||||
if 'provider' not in args['target'] :
|
||||
self.PROVIDER = args['target']['type']
|
||||
self.writer = transport.factory.instance(**args['target'])
|
||||
else:
|
||||
self.PROVIDER = args['target']['provider']
|
||||
args['target']['context'] = 'write'
|
||||
|
||||
self.writer = transport.instance(**args['target'])
|
||||
#
|
||||
# If the table doesn't exists maybe create it ?
|
||||
#
|
||||
self.rows = args['rows']
|
||||
|
||||
def run(self):
|
||||
_info = {"values":self.rows} if 'couch' in self.PROVIDER else self.rows
|
||||
|
||||
self.writer.write(_info)
|
||||
self.writer.close()
|
||||
|
||||
|
@ -59,7 +81,19 @@ class ETL (Process):
|
|||
def __init__(self,**_args):
|
||||
super().__init__()
|
||||
self.name = _args['id']
|
||||
if 'provider' not in _args['source'] :
|
||||
#@deprecate
|
||||
self.reader = transport.factory.instance(**_args['source'])
|
||||
else:
|
||||
#
|
||||
# This is the new interface
|
||||
_args['source']['context'] = 'read'
|
||||
|
||||
self.reader = transport.instance(**_args['source'])
|
||||
#
|
||||
# do we have an sql query provided or not ....
|
||||
# self.sql = _args['source']['sql'] if 'sql' in _args['source'] else None
|
||||
self.cmd = _args['source']['cmd'] if 'cmd' in _args['source'] else None
|
||||
self._oargs = _args['target'] #transport.factory.instance(**_args['target'])
|
||||
self.JOB_COUNT = _args['jobs']
|
||||
self.jobs = []
|
||||
|
@ -68,6 +102,9 @@ class ETL (Process):
|
|||
_args['name'] = self.name
|
||||
print (_args)
|
||||
def run(self):
|
||||
if self.cmd :
|
||||
idf = self.reader.read(**self.cmd)
|
||||
else:
|
||||
idf = self.reader.read()
|
||||
idf = pd.DataFrame(idf)
|
||||
idf.columns = [str(name).replace("b'",'').replace("'","").strip() for name in idf.columns.tolist()]
|
||||
|
@ -79,7 +116,8 @@ class ETL (Process):
|
|||
try:
|
||||
self.log(module='write',action='partitioning')
|
||||
rows = np.array_split(np.arange(idf.shape[0]),self.JOB_COUNT)
|
||||
|
||||
#
|
||||
# @TODO: locks
|
||||
for i in rows :
|
||||
_id = 'segment #'.join([str(rows.index(i)),self.name])
|
||||
segment = idf.loc[i,:] #.to_dict(orient='records')
|
||||
|
|
2
setup.py
2
setup.py
|
@ -8,7 +8,7 @@ def read(fname):
|
|||
return open(os.path.join(os.path.dirname(__file__), fname)).read()
|
||||
args = {
|
||||
"name":"data-transport",
|
||||
"version":"1.4.0",
|
||||
"version":"1.4.1",
|
||||
"author":"The Phi Technology LLC","author_email":"info@the-phi.com",
|
||||
"license":"MIT",
|
||||
"packages":["transport"]}
|
||||
|
|
|
@ -71,16 +71,7 @@ from google.cloud import bigquery as bq
|
|||
import nzpy as nz #--- netezza drivers
|
||||
import os
|
||||
|
||||
RDBMS = {
|
||||
|
||||
"postgresql":{"port":"5432","driver":pg},
|
||||
"redshift":{"port":"5432","driver":pg},
|
||||
"netezza":{"port":"5480","driver":nz},
|
||||
"mysql":{"port":"3306","driver":my},
|
||||
"mariadb":{"port":"3306","driver":my},
|
||||
"mongodb":{"port":"27017","class":{"read"}},
|
||||
"couchdb":{"port":"5984"}
|
||||
}
|
||||
class factory :
|
||||
TYPE = {"sql":{"providers":["postgresql","mysql","neteeza","bigquery","mariadb","redshift"]}}
|
||||
PROVIDERS = {
|
||||
|
@ -94,6 +85,11 @@ class factory :
|
|||
"mongo":{"port":27017,"host":"localhost","class":{"read":mongo.MongoReader,"write":mongo.MongoWriter}},
|
||||
"couch":{"port":5984,"host":"localhost","class":{"read":couch.CouchReader,"write":couch.CouchWriter}},
|
||||
"netezza":{"port":5480,"driver":nz,"default":{"type":"VARCHAR(256)"}}}
|
||||
#
|
||||
# creating synonyms
|
||||
PROVIDERS['mongodb'] = PROVIDERS['mongo']
|
||||
PROVIDERS['couchdb'] = PROVIDERS['couch']
|
||||
PROVIDERS['sqlite3'] = PROVIDERS['sqlite']
|
||||
|
||||
@staticmethod
|
||||
def instance(**args):
|
||||
|
@ -126,14 +122,17 @@ class factory :
|
|||
return anObject
|
||||
|
||||
import time
|
||||
def instance(provider,context,**_args):
|
||||
def instance(**_args):
|
||||
"""
|
||||
|
||||
@param provider {file,sqlite,postgresql,redshift,bigquery,netezza,mongo,couch ...}
|
||||
@param context read|write|rw
|
||||
@param _args argument to got with the datastore (username,password,host,port ...)
|
||||
"""
|
||||
_id = context if context in ['read','write'] else None
|
||||
|
||||
provider = _args['provider']
|
||||
context = _args['context']
|
||||
_id = context if context in ['read','write'] else 'read'
|
||||
if _id :
|
||||
args = {'provider':_id}
|
||||
for key in factory.PROVIDERS[provider] :
|
||||
|
|
|
@ -114,6 +114,7 @@ class DiskWriter(Writer):
|
|||
class SQLiteReader (DiskReader):
|
||||
def __init__(self,**args):
|
||||
DiskReader.__init__(self,**args)
|
||||
self.path = args['database'] if 'database' in args else args['path']
|
||||
self.conn = sqlite3.connect(self.path,isolation_level=None)
|
||||
self.conn.row_factory = sqlite3.Row
|
||||
self.table = args['table']
|
||||
|
@ -145,7 +146,7 @@ class SQLiteWriter(DiskWriter) :
|
|||
DiskWriter.__init__(self,**args)
|
||||
self.table = args['table']
|
||||
|
||||
self.conn = sqlite3.connect(self.path,isolation_level=None)
|
||||
self.conn = sqlite3.connect(self.path,isolation_level="IMMEDIATE")
|
||||
self.conn.row_factory = sqlite3.Row
|
||||
self.fields = args['fields'] if 'fields' in args else []
|
||||
|
||||
|
@ -184,20 +185,27 @@ class SQLiteWriter(DiskWriter) :
|
|||
if not self.fields :
|
||||
self.init(list(info.keys()))
|
||||
|
||||
if type(info) != list :
|
||||
if type(info) == object :
|
||||
info = [info]
|
||||
elif type(info) == pd.DataFrame :
|
||||
info = info.to_dict(orient='records')
|
||||
|
||||
SQLiteWriter.LOCK.acquire()
|
||||
try:
|
||||
cursor = self.conn.cursor()
|
||||
sql = " " .join(["INSERT INTO ",self.table,"(", ",".join(self.fields) ,")", "values(':values')"])
|
||||
for row in info :
|
||||
stream = json.dumps(row)
|
||||
stream = stream.replace("'","''")
|
||||
cursor.execute(sql.replace(":values",stream) )
|
||||
|
||||
# self.conn.commit()
|
||||
cursor = self.conn.cursor()
|
||||
sql = " " .join(["INSERT INTO ",self.table,"(", ",".join(self.fields) ,")", "values(:values)"])
|
||||
for row in info :
|
||||
stream =["".join(["",value,""]) if type(value) == str else value for value in row.values()]
|
||||
stream = json.dumps(stream).replace("[","").replace("]","")
|
||||
|
||||
|
||||
self.conn.execute(sql.replace(":values",stream) )
|
||||
# cursor.commit()
|
||||
|
||||
self.conn.commit()
|
||||
# print (sql)
|
||||
except Exception as e :
|
||||
print (e)
|
||||
pass
|
||||
SQLiteWriter.LOCK.release()
|
Loading…
Reference in New Issue