Why wont this loop? - python

Newby working my way through python and sql with mariadb.
Why wont this loop? It updates the first record only. Im pretty hack at this...
cursor1.execute("select recordid, mvavgvol, quote_usd_volume_change_24h from pumped")
records = cursor1.fetchall()
for x in records:
rid = (x[0])
m = (x[1])
o = (x[2])
if (m >= o):
result = 0
else:
result = 1
upd_data=(result,rid)
sql1 = ("UPDATE pumped SET mvavgvolcheck = %s WHERE recordid = %s")
cursor2.execute(sql1,upd_data)
conn.commit()

Since you are fetching multiple rows you have to store the fetched values in an array and use cursor's executemany() method instead.
✂
data= []
for x in records:
rid = (x[0])
result= int(x[1] > x[2])
data+= [(result, rid)]
cursor.executemany(UPDATE pumped SET mvavgvolcheck = %s WHERE recordid = %s", data);
✂
When using mariadb python module (MariaDB Connector/Python) this is much more effective since it reduces network traffic: instead of sending n update commands in a loop (where n is the number of rows in table pumped) only one command will be send to the server.

conn = msql.connect(host=Host,port=Port, user=User, password=Password, database=database)
cursor1 = conn.cursor()
cursor2 = conn.cursor()
cursor1.execute("select recordid, mvavgvol, quote_usd_volume_change_24h from pumped")
records = cursor1.fetchall()
for x in records:
rid = (x[0])
m = (x[1])
o = (x[2])
if (m >= o):
result = 0
cursor2.execute("UPDATE pumped SET mvavgvolcheck = %s WHERE recordid = %s",(result, rid))
conn.commit()
else:
result = 1
cursor2.execute("UPDATE pumped SET mvavgvolcheck = %s WHERE recordid = %s",(result, rid))
conn.commit()

Related

MySQL update or insert based on fetchall results in Python

I need to set some user meta in my wordpress through local python script. Hence I can't use the WP update_user_meta for it - it has to be done manually.
import mysql.connector as mysql
cnx = mysql.connect(host=HOST, database=DATABASE, user=USER, password=PASSWORD)
cursor = cnx.cursor()
get_meta = ("SELECT * FROM `ff_usermeta` WHERE `user_id`= 1 AND (`meta_key`='nickname' OR `meta_key`='info' OR `meta_key`='bg' OR `meta_key`='avatar' OR `meta_key`='profile_updated')")
cursor.execute(get_meta)
meta = cursor.fetchall()
#some processing of the result
cursor.execute(q, (...))
cnx.commit()
cursor.close()
cnx.close()
Now I need to check if the result has meta with each of the keys.
If the key already exists for this user, it needs to run UPDATE for this meta.
If this user still has no meta of this key, it has to INSERT new row.
if(there's no 'nickname' in meta_key on either of 5 or less rows):
q = ("INSERT INTO `ff_usermeta` ...")
else:
q = ("UPDATE `ff_usermeta` ...")
...and 4 more times like that?.. Seems like a good place for a cycle, but I don't really like the idea to make it 5x queues, especially since there might be more fields in the future.
I was thinking along the lines of searching the fetchall result for matches in meta_key, and if found, adding required data to one array, if not - to another. And then just running one update and one insert at the end, assuming both are not empty. If I were to write it in semi-php style, it would look roughly like this:
if(in_array("nickname", meta))
for_update .= "`nickname`='"+data[0]+"', "
else:
fields .= "`nickname`, "
vals .= "'"+data[0]+"', "
if(in_array("bg", meta)):
for_update .= "`bg`='"+data[1]+"', "
else:
fields .= "`bg`, "
vals .= "'"+data[1]+"', "
if(for_update):
update = ("UPDATE `ff_usermeta` SET "+for_update+" WHERE 1")
if(fields):
insert = ("INSERT INTO `ff_usermeta`("+fields+") VALUES ("+vals+")")
But absolutely no clue how to translate it correctly to python. I had to google it up to things like "why dot not working to add one string to another". Any advice? Or perhaps there is a better way? Thanks!
It is not complete, you can not update your rows in that way.
But with this you can start to make your query
The frist select gets exactly 1 row, if the user_id exists.
The user_id doesn't seem the right choice for this, but to get what you can do it is enough.
If the query doesn't have an entry, the it will insert some data you get from anywhere
The update as the insert are in that form wrong as you have to insert 5 new orws or update max 5 rows, but that is more for you to programm
import mysql.connector as mysql
HOST = "localhost"
DATABASE = ""
USER = "root"
PASSWORD = "mypassword"
cnx = mysql.connect(host=HOST, database=DATABASE, user=USER, password=PASSWORD)
cnx = mysql.connect(host=HOST, database=DATABASE, user=USER, password=PASSWORD)
cursor = cnx.cursor()
user_id = 1
get_meta = ("""SELECT umeta_id, user_id , MAX(IF( `meta_key`='nickname', meta_value,'')) AS 'nickname' , MAX(IF( `meta_key`='info', meta_value,'')) AS 'info' , MAX(IF( `meta_key`='bg', meta_value,'')) AS 'bg' , MAX(IF( `meta_key`='avatar', meta_value,''NULL'')) AS 'avatar' , MAX(IF (`meta_key`='profile_updated', meta_value,'')) AS 'profile_updated' FROM `ff_usermeta` WHERE `user_id`= %s GROUP BY umeta_id, user_id:""")
result = cursor.execute(get_meta,(user_id,))
if result > 0:
data = cursor.fetchone()
for_update = "";
#some processing of the result
if not data["nickname"]:
for_update += "`nickname`='"+data["nickname"]+"', "
if not data["bg"]:
for_update += "`bg`='"+data["bg"]+"', "
query = ("UPDATE `ff_usermeta` SET "+for_update+" WHERE user_id = " + user_id)
else:
#here are no data to be gathered as there is no user_id present add new user
nickname = ""
bg= ""
info = ""
avatar = ""
profile_updated = ""
fields= ""
vals = ""
fields += "`nickname`,`info`, `bg`,`avatar`,`profile_updated`"
vals += "'"+nickname+"', "+"'"+info+"', "+"'"+bg+"', "+"'"+avatar+"', "+"'"+profile_updatedfo+"'"
query = ("INSERT INTO `ff_usermeta`("+fields+") VALUES ("+vals+")")
cursor.execute(query)
cnx.commit()
cursor.close()
cnx.close()
I tried my best to adapt the suggestion above, but couldn't figure out how to make it work. Eventually I went another way, and it seems to work somehow, so I'll post the full code in case anyone would find it useful.
What it does: checks the queue in table with validation request, then parses a page (separate function) and updates user profile accodringly.
import mysql.connector as mysql
import time
from datetime import datetime
cnx = mysql.connect(host=HOST, database=DATABASE, user=USER, password=PASSWORD)
while True: #endless loop as a temporary scheduler
cursor = cnx.cursor()
#getting first request in the queue - 0: id, 1: url, 2: parse, 3: status, 4: user, 5: user_page, 6: req_date, 7: action
cursor.execute("SELECT * FROM `ff_qq` WHERE status = 0 LIMIT 1")
row = cursor.fetchone()
if row:
status = 1 #processed
if row[7] == "verify":
get_user = ("SELECT * FROM `ff_users` WHERE ID = %s LIMIT 1")
cursor.execute(get_user, (row[4],))
user = cursor.fetchone() #0 - ID, 5 - user_url, 8 - user_status, 9 - display_name
#separate function that returns data to insert into mysql
udata = verify(row) #0 - nickname, 1 - fb_av, 2 - fb_bg, 3 - fb_info, 4 - owner
ustat = row[1].split("/authors/")
if udata['owned'] or user[8] == ustat[1]:
update_user = ("UPDATE `ff_users` SET user_status = %s, display_name = %s, user_url = %s WHERE ID = %s LIMIT 1")
cursor.execute(update_user, (ustat[1], udata['nickname'], row[1], user[0]))
status = 2 #success
get = ("SELECT `meta_value` FROM `ff_usermeta` WHERE `user_id`= %s AND `meta_key`='ff_capabilities' LIMIT 1")
cursor.execute(get, (row[4],))
rights = cursor.fetchone()
if rights == 'a:1:{s:10:"subscriber";b:1;}':
promote = ("UPDATE `ff_usermeta` SET `meta_value` = 'a:1:{s:6:\"author\";b:1;}' "
"WHERE `user_id` = %s AND `meta_key`='ff_capabilities' LIMIT 1")
cursor.execute(promote, (row[0],))
#list of meta_key values in same order as returned data
ff = ['nickname', 'fb_av', 'fb_bg', 'fb_info']
for x in range(0,3): #goes through each one of the above list
if udata[ff[x]]: #yes this actually works, who would've thought?..
#current meta_key added directly into the string
get = ("SELECT `meta_value` FROM `ff_usermeta` WHERE `user_id`= %s AND `meta_key`='" + ff[x] + "' LIMIT 1")
cursor.execute(get, (row[4],))
meta = cursor.fetchone()
if(meta): #update if it exists, otherwise insert new row
qq = ("UPDATE `ff_usermeta` SET `meta_value` = %s "
"WHERE `user_id` = %s AND `meta_key`='" + ff[x] + "' LIMIT 1")
else:
qq = ("INSERT INTO `ff_usermeta`(`meta_value`, `meta_key`, `user_id`) "
"VALUES ('%s','" + ff[x] + "','%s'")
cursor.execute(qq, (udata[ff[x]], row[0])) #same execute works for both
else:
status = 3 #verification failed
#update queue to reflect its status
update = ("UPDATE `ff_qq` SET status = %s WHERE id = %s LIMIT 1")
cursor.execute(update, (status, row[0]))
cnx.commit()
cursor.close()
now = datetime.now()
print(now.strftime("%d.%m.%Y %H:%M:%S"))
time.sleep(180) #sleep until it's time to re-check the queue
cnx.close()

For Loop with input confirmation after n iterations with pymssql

I am updating a database table with a dictionary with this code:
cursor = conn.cursor()
for key in dRt:
x = dRt[key]
sql = 'UPDATE table SET R = %s WHERE %s = ID'
cursor.execute(sql, (x, key))
conn.commit()
conn.close()
My dictionary got several thousand entries. Is it possible to add a part to the code which asks, after writing 1000 rows into the database, if it should continue?
I tried something like this:
cursor = conn.cursor()
counter = 0
for key in dRt:
x = dRt[key]
sql = 'UPDATE table SET R = %s WHERE %s = ID'
if counter == 1000:
break
eingabe = input("Beenden? Enter drücken!\n")
cursor.execute(sql, (x, key))
conn.commit()
conn.close()
But this does not work correctly.
Conditionally initialising an input based on counter should work. After initialising an input you should also reset the counter to 0.
cursor = conn.cursor()
counter = 0
for key in dRt:
counter += 1
x = dRt[key]
sql = 'UPDATE table SET R = %s WHERE %s = ID'
if counter == 1000:
input("Beenden? Enter drücken!\n")
counter = 0
cursor.execute(sql, (x, key))
conn.commit()
conn.close()
It's okay to have an input statement in the loop, but execution will stop until something is entered. You must then check to see what was entered to determine whether you want to continue or not:
cursor = conn.cursor()
sql = 'UPDATE table SET R = %s WHERE %s = ID' # moved out of the loop
counter = 0
for key in dRt:
counter += 1
x = dRt[key]
cursor.execute(sql, (x, key)) # moved higher in the loop
if counter == 1000: # now we have done the next 1000 updates
reply = input("Beenden? Enter drücken!\n")
if reply == '':
break # anything else entered than just enter, for example "geh", and we continue
counter = 0
conn.commit()
conn.close()

How to execute more than once the same query with different data?

I'm trying to execute the same query but with different data but I always get data the first time. The others times, dispite of there are data for the querys in the data base, mysql returns empty data.
This is the code:
def get_team_colour_map(self, players, id_competition):
tcm = FIBAColourMap()
for p in players:
args = [p["id"], id_competition]
conn = pymysql.Connect(host = DDBB.DDBB_FIBA_HOST,
user = DDBB.DDBB_FIBA_USER,
password = DDBB.DDBB_FIBA_PSWD,
db = DDBB.DDBB_FIBA_NAME,
charset = DDBB.DDBB_FIBA_CHARSET,
cursorclass=pymysql.cursors.DictCursor)
with conn.cursor() as cursor:
print("id player: {}".format(p["id"]))
print("args: {}".format(args))
cursor.execute("select sc.* from tbl030_shots_chart sc, tbl006_player_team pt, tbl007_game g, tbl004_jornada j, tbl012_competition c where pt.id = %s and pt.id_player_feb = sc.id_fiba and sc.id_game = g.id and g.id_jornada = j.id and j.id_competition = c.id and c.id = %s", args)
data = cursor.fetchall()
print("data: {}".format(data))
print("Total rows: {}".format(cursor.rowcount))
if cursor.rowcount > 0:
for s in data:
x = float(FIBASCReport.adjust_x(s["x"]))
y = float(FIBASCReport.adjust_y(s["y"]))
color = tcm.image.getpixel((x,y))
color = ("#%02x%02x%02x" % color).upper()
if tcm.exists_color(color):
if int(s["m"]) == 0:
tcm.set_scored_shots(color, 1)
else:
tcm.set_failed_shots(color, 1)
else:
if int(s["m"]) == 0:
tcm.set_scored_shots("OTROS", 1)
else:
tcm.set_failed_shots("OTROS", 1)
else:
#tcm = None
print("Jugadora con id: {} NO ha realizado ningún tiro en competición: {}".format(p["id"], id_competition))
return tcm
In this code, cursor.fetchall() returns data the first query but the next querys returns empty results.
How can I run several querys? I'm using mySQL 8.0 and Python 3.6
Its because you are using the same cursor each time. create a new instance of the cursor each time you loop through to excecute the query. After the first query is run the cursor is already positioned after all the data. Hence no rows returned after that
You can also try this:
Look at the documentation for MySQLCursor.execute().
It claims that you can pass in a multi parameter that allows you to run multiple queries in one string.
If multi is set to True, execute() is able to execute multiple statements specified in the operation string.
multi is an optional second parameter to the execute() call:
operation = 'SELECT 1; INSERT INTO t1 VALUES (); SELECT 2'
for result in cursor.execute(operation, multi=True):

Python: Print only one value of an SQL query result set

from sqlalchemy import create_engine
engine = create_engine('mssql+pymssql://myusername:mypassword#127.0.0.1:1433/AQOA_Core')
connection = engine.connect()
result = connection.execute("""SELECT DISTINCT Distributor FROM Product""")
for row in result:
print row[0]
connection.close()
The above code returns a result set as:
FRANCETV
GAUMONT
M6SND
PATHE
STUDIOCANAL
TF1
WARNER
What if I want to print just one value without changing the query?
Tried print row[0][1], print row[0:1] This was tried to print just the first value FRANCETV
Basically I want to be able to print each one of the values in the result set seperately without making changes to the query.
You can try to access the data by:
connection = engine.connect()
result = connection.execute("""SELECT DISTINCT Distributor FROM Product""")
result_list = result.fetchall()
result_list[0][0]
connection.close()
If you take a look at the pymssql docs, you'll see the cursor object has a fetchone() method.
with engine.connect() as connection:
cursor = connection.cursor()
cursor.execute("""SELECT DISTINCT Distributor FROM Product""")
first_row = cursor.fetchone()
print first_row
file = open("testfile.txt", "w", encoding='utf-8')
mycursor.execute('SELECT * from COMPANY')
myresult = mycursor.fetchall()
counter = 0
y = 0
for x in myresult:
y = 0
while y < 3:
print(y)
file.write("%s\n" % myresult[counter][y])
y = y + 1
counter = counter + 1
file.close()
This is what I concluded from the answers above

Data extraction and transform efficiency

I've got a Python script that connects to a MySQL database and executes a number of nested SELECT queries. It's basically a giant for loop. The database is structured such that Businesses have Menus, Menus have Sections, and Sections have Items. The script queries all the Businesses, and for each Business, it queries all of its Menus, and so on. It builds a big dictionary along the way that it then spits out as JSON.
It looks something like this:
#!/usr/bin/env python
from bottle import route, run
import mysql.connector
import json
import collections
import datetime
def getBusinesses():
conn = mysql.connector.connect(user="APIUser", password="abc123", host="12.34.56.78", port="54321", database="businesses")
cursor = conn.cursor()
objects = {}
businessesQuery = ("SELECT * FROM business")
cursor.execute(businessesQuery)
businessRows = cursor.fetchall()
businessObjects = []
for businessRow in businessRows:
print businessRow[0]
businessDict = collections.OrderedDict()
businessDict['id'] = businessRow[0]
businessDict['business_name'] = businessRow[1]
businessDict['business_address1'] = businessRow[2]
businessDict['business_address2'] = businessRow[3]
businessDict['business_city'] = businessRow[4]
businessDict['business_state'] = businessRow[5]
businessDict['business_zip'] = businessRow[6]
businessObjects.append(businessDict)
menuQuery = ("SELECT * FROM menu WHERE business_id = %s" % businessRow[0])
cursor.execute(menuQuery)
menuRows = cursor.fetchall()
menuObjects = []
for menuRow in menuRows:
menuDict = collections.OrderedDict()
menuDict['id'] = menuRow[0]
menuDict['menu_name'] = menuRow[1]
menuDict['menu_description'] = menuRow[2]
menuDict['menu_note'] = menuRow[3]
menuDict['business_id'] = menuRow[4]
menuObjects.append(menuDict)
businessDict['menus'] = menuObjects
for menuIdx, menuRow in enumerate(menuRows):
sectionQuery = ("SELECT * FROM menu_section WHERE menu_id = %s" % menuRow[0])
cursor.execute(sectionQuery)
sectionRows = cursor.fetchall()
sectionObjects = []
for sectionIdx, sectionRow in enumerate(sectionRows):
sectionDict = collections.OrderedDict()
sectionDict['id'] = sectionRow[0]
sectionDict['section_name'] = sectionRow[1]
sectionDict['section_note'] = sectionRow[2]
sectionDict['section_description'] = sectionRow[3]
sectionDict['menu_id'] = sectionRow[4]
sectionObjects.append(sectionDict)
businessDict['menus'][menuIdx]['sections'] = sectionObjects
itemQuery = ("SELECT * FROM menu_item WHERE section_id = %s" % sectionRow[0])
cursor.execute(itemQuery)
itemRows = cursor.fetchall()
itemObjects = []
for itemIdx, itemRow in enumerate(itemRows):
itemDict = collections.OrderedDict()
itemDict['id'] = itemRow[0]
itemDict['item_name'] = itemRow[1]
itemDict['item_description'] = itemRow[2]
itemDict['item_note'] = itemRow[3]
itemDict['item_price'] = itemRow[4]
itemDict['section_id'] = itemRow[5]
itemObjects.append(itemDict)
businessDict['menus'][menuIdx]['sections'][sectionIdx]['items'] = itemObjects
objects['businesses'] = businessObjects
return objects
#route('/test')
def index():
return json.dumps(getBusinesses())
run(host='192.168.1.70', port=7070)
I want to know if this is an efficient way of doing things. When I deployed my database remotely (WebFaction) and ran the Bottle server locally, it took almost 40 seconds to return a few hundred rows. So it seems like something is amiss. I have a gut feeling that there could be a better way of doing this. Just not sure what that way is!
if I had to venture a guess: notice the rough structure of your code is:
def getBusinesses():
businessesQuery = ("SELECT * FROM business")
businessRows = cursor.fetchall()
businessObjects = []
for businessRow in businessRows:
menuQuery = ("SELECT * FROM menu WHERE business_id = %s" % businessRow[0])
menuRows = cursor.fetchall()
for menuIdx, menuRow in enumerate(menuRows):
sectionQuery = ("SELECT * FROM menu_section WHERE menu_id = %s" % menuRow[0])
cursor.execute(sectionQuery)
sectionRows = cursor.fetchall()
sectionObjects = []
for sectionIdx, sectionRow in enumerate(sectionRows):
itemQuery = ("SELECT * FROM menu_item WHERE section_id = %s" % sectionRow[0])
itemRows = cursor.fetchall()
That is, you execute nearly identical queries in a loop for menu, menu_section and especially menu_item. Also, you're using fetchall() to return the full contents of the result set but examine each element only once, in a loop, where you create another list of objects.
what you might want instead is something more like:
businesses = []
cursor.execute("select * from business")
row = cursor.fetchone()
while row is not None:
business.append(...(row))
row = cursor.fetchone()
cursor.execute("select * from menu")
row = cursor.fetchone()
while row is not None:
business[row['business_id']].menus.append(...(row))
row = cursor.fetchone()
cursor.execute("select menu.business_id, menu_section.*"
" from menu_section"
" join menu on menu.id = menu_section.menu_id")
row = cursor.fetchone()
while row is not None:
business[row['business_id']][row['menu_id']].sections.append(...(row))
row = cursor.fetchone()
cursor.execute("select menu.business_id, menu_section.menu_id, menu_item.*"
" from menu_item"
" join menu_section on menu_section.id = menu_item.section_id"
" join menu on menu.id = menu_section.menu_id")
row = cursor.fetchone()
while row is not None:
business[row['business_id']][row['menu_id']][row['section_id'].items.append(...(row))
row = cursor.fetchone()
so that you're issuing a much smaller number of queries, and only loading the amount of data you can process in one go.

Categories

Resources