UnicodeDecodeError: 'utf-8' codec can't decode byte 0x89 in position 0: invalid start byte

[Hans Mulder]

On 6/12/12 11:07:51, iMath wrote:
> the following code originally from http://zetcode.com/databases/mysqlpythontutorial/
> within the "Writing images" part .
> 
> 
> import MySQLdb as mdb
> import sys
> 
> try:
>     fin = open("Chrome_Logo.svg.png",'rb')
>     img = fin.read()
>     fin.close()
> 
> except IOError as e:
> 
>     print ("Error %d: %s" % (e.args[0],e.args[1]))
>     sys.exit(1)
> 
> 
> try:
>     conn = mdb.connect(host='localhost',user='testuser',
>        passwd='test623', db='testdb')
>     cursor = conn.cursor()
>     cursor.execute("INSERT INTO Images SET Data='%s'" % \
>         mdb.escape_string(img))

You shouldn't call mdb.escape_string directly.  Instead, you
should put placeholders in your SQL statement and let MySQLdb
figure out how to properly escape whatever needs escaping.

Somewhat confusingly, placeholders are written as %s in MySQLdb.
They differ from strings in not being enclosed in quotes.
The other difference is that you'd provide two arguments to
cursor.execute; the second of these is a tuple; in this case
a tuple with only one element:

    cursor.execute("INSERT INTO Images SET Data=%s", (img,))

>     conn.commit()
> 
>     cursor.close()
>     conn.close()
> 
> except mdb.Error as e:
> 
>     print ("Error %d: %s" % (e.args[0],e.args[1]))
>     sys.exit(1)
> 
> 
> I port it to python 3 ,and also change 
> fin = open("chrome.png") 
> to 
> fin = open("Chrome_Logo.png",'rb')
> but when I run it ,it gives the following error :
> 
> Traceback (most recent call last): 
>   File "E:\Python\py32\itest4.py", line 20, in <module>
>     mdb.escape_string(img))
> UnicodeDecodeError: 'utf-8' codec can't decode byte 0x89 in position 0: invalid start byte
> 
> so how to fix it ?

Python 3 distinguishes between binary data and Unicode text.
Trying to apply string functions to images or other binary
data won't work.

Maybe correcting this bytes/strings confusion and porting
to Python 3 in one go is too large a transformation.  In
that case, your best bet would be to go back to Python 2
and fix all the bytes/string confusion there.  When you've
got it working again, you may be ready to port to Python 3.


Hope this helps,

-- HansM