问题
-- I'm loading test data into my SQL Server Database using python and was able to successfully take images and break them down into bytes and store them in the database, but when trying to fetch back the bytes and decode it to save it as a new file type, all i get is a blank image file. Not sure what i am doing wrong here...
-- I've tried several iterations using base64 from other tutorials and similar questions, but cant seem to find one that will solve my problem.
SQLCommand = ("SELECT Photo FROM Validation")
cursor.execute(SQLCommand)
data = cursor.fetchone()[0]
image_64_decode = base64.decodebytes(data)
image_result = open('booking.png', 'wb')
image_result.write(image_64_decode)
image_result.close()
connection.close()
The expected result is that I should be able to fetch the bytes from the database which the database column is varbinary(max) the equivalent of bytes in python. once the bytes are fetched using the script in python it should save a file as booking.png which should replicate the image i stored in the database.
When i run the script i don't get an error, and in fact it saves a file, but the file is empty containing 1kb and does not reproduce the image. Not sure where i am going wrong, but it seems like it's not properly fetching the bytes.
回答1:
I was able to get my code to work and can successfully convert an image to bytes, store it in the sql server datebase and retrieve it by fetching the bytes and reproducing the image.
There is only one problem- this only works if i use the nvarchar(max) data type for the column where I am storing the image bytes. I get errors when using varbinary(max) or when i solve the error, it just doesn't actually fetch the bites and properly convert it-- any guidance on what I might be doing wrong as I have a feeling it's something small. The updated code below is what I am doing using nvarchar(max) that is working.
import pypyodbc
import base64
from base64 import *
connection = pypyodbc.connect('Driver=SQL Server;'
'Server=DESKTOP-MSSQLSERVER;'
'Database=Test;'
'Trusted_Connection=yes;'
)
cursor = connection.cursor()
a = 'bob@bob.com'
b = 'mack jones'
filename = 'bookingsuccessful.PNG'
image = open(filename, 'rb')
image_read = image.read()
image_64_encode = base64.encodebytes(image_read)
image.close()
SQLCommand = ("INSERT INTO Validation(email, myname, photo) VALUES(?,?,?)")
Values = [a,b,image_64_encode]
cursor.execute(SQLCommand, Values)
connection.commit()
SQLCommand = ("SELECT Photo FROM validation")
cursor.execute(SQLCommand)
data = cursor.fetchone()[0]
data = bytes(data.strip("\n"), 'utf-8')
image_64_decode = base64.decodebytes(data)
image_result = open('testfile.gif', 'wb')
image_result.write(image_64_decode)
image_result.close()
connection.close()
回答2:
There really is no need for base64 encoding. If you use pyodbc instead of pypyodbc then it is as simple as
# test data
photo_path = r'C:\Users\Public\Pictures' + '\\'
email = 'bob@example.com'
# test environment
cursor.execute("""\
CREATE TABLE #validation (
email nvarchar(255) PRIMARY KEY,
photo varbinary(max))
""")
# save binary file
with open(photo_path + 'generic_man.jpg', 'rb') as photo_file:
photo_bytes = photo_file.read()
cursor.execute("INSERT INTO #validation (email, photo) VALUES (?, ?)", email, photo_bytes)
print(f'{len(photo_bytes)}-byte file written for {email}')
# 5632-byte file written for bob@example.com
# retrieve binary data and save as new file
retrieved_bytes = cursor.execute("SELECT photo FROM #validation WHERE email = ?", email).fetchval()
with open(photo_path + 'new.jpg', 'wb') as new_jpg:
new_jpg.write(retrieved_bytes)
print(f'{len(retrieved_bytes)} bytes retrieved and written to new file')
# 5632 bytes retrieved and written to new file
来源：https://stackoverflow.com/questions/56927045/how-to-fetch-bytes-from-sql-server-database-and-convert-to-image-in-python