Literal Escaped Octets

[Chason Hayes]

I am trying to convert raw binary data to data with escaped octets in
order to store it in a bytea field on postgresql server. I could do this
easily in c/c++ but I need to do it in python. I am not sure how to read
and evaluate the binary value of a byte in a long string when it is a non
printable ascii value in python. I read some ways to use unpack from the
struct module, but i really couldn't understand where that would help. I
looked at the MIMIEncode module but I don't know how to convert the object
to a string. Is there a module that will convert the data? It seems to me
that this question must have been answered a million times before but I
can't find anything.



See http://www.postgresql.org/docs/8.1/i...pe-binary.html
for a description of the problem domain.

[Alex Martelli]

Re: Literal Escaped Octets

Chason Hayes <chasonh@hotmai l.com> wrote:
...[color=blue]
> easily in c/c++ but I need to do it in python. I am not sure how to read
> and evaluate the binary value of a byte in a long string when it is a non
> printable ascii value in python.[/color]

If you have a bytestring (AKA plain string) s, the binary value of its
k-th byte is ord(s[k]).


Alex

[Steve Holden]

Re: Literal Escaped Octets

Chason Hayes wrote:[color=blue]
> I am trying to convert raw binary data to data with escaped octets in
> order to store it in a bytea field on postgresql server. I could do this
> easily in c/c++ but I need to do it in python. I am not sure how to read
> and evaluate the binary value of a byte in a long string when it is a non
> printable ascii value in python. I read some ways to use unpack from the
> struct module, but i really couldn't understand where that would help. I
> looked at the MIMIEncode module but I don't know how to convert the object
> to a string. Is there a module that will convert the data? It seems to me
> that this question must have been answered a million times before but I
> can't find anything.
>
>
>
> See http://www.postgresql.org/docs/8.1/i...pe-binary.html
> for a description of the problem domain.
>
>[/color]
The URL you reference is discussing how you represent arbitrary values
in string literals. If you already have the data in a Python string the
best advise is to use a parameterized query - that way your Python DB
API module will do the escaping for you!

regards
Steve
--
Steve Holden +44 150 684 7255 +1 800 494 3119
Holden Web LLC www.holdenweb.com
PyCon TX 2006 www.python.org/pycon/

[Chason Hayes]

Re: Literal Escaped Octets

On Mon, 06 Feb 2006 13:39:17 +0000, Steve Holden wrote:
[color=blue]
> Chason Hayes wrote:[color=green]
>> I am trying to convert raw binary data to data with escaped octets in
>> order to store it in a bytea field on postgresql server. I could do this
>> easily in c/c++ but I need to do it in python. I am not sure how to read
>> and evaluate the binary value of a byte in a long string when it is a non
>> printable ascii value in python. I read some ways to use unpack from the
>> struct module, but i really couldn't understand where that would help. I
>> looked at the MIMIEncode module but I don't know how to convert the object
>> to a string. Is there a module that will convert the data? It seems to me
>> that this question must have been answered a million times before but I
>> can't find anything.
>>
>>
>>
>> See http://www.postgresql.org/docs/8.1/i...pe-binary.html
>> for a description of the problem domain.
>>
>>[/color]
> The URL you reference is discussing how you represent arbitrary values
> in string literals. If you already have the data in a Python string the
> best advise is to use a parameterized query - that way your Python DB
> API module will do the escaping for you!
>
> regards
> Steve[/color]

Thanks for the input. I tried that with a format string and a
dictionary, but I still received a database error indicating illegal
string values. This error went away completely when I used a test file
consisting only of text, but reproduced everytime with a true binary file.
If you can let me know where I am wrong or show me a code snippet with a
sql insert that contains a variable with raw binary data that works,
I would greatly appreciate it.

Chason

[Chason Hayes]

Re: Literal Escaped Octets

On Sun, 05 Feb 2006 21:07:23 -0800, Alex Martelli wrote:
[color=blue]
> Chason Hayes <chasonh@hotmai l.com> wrote:
> ...[color=green]
>> easily in c/c++ but I need to do it in python. I am not sure how to read
>> and evaluate the binary value of a byte in a long string when it is a non
>> printable ascii value in python.[/color]
>
> If you have a bytestring (AKA plain string) s, the binary value of its
> k-th byte is ord(s[k]).
>
>
> Alex[/color]

Thank you very much, That is the function that I was looking for to write
a filter.

Chason

[Steve Holden]

Re: Literal Escaped Octets

Chason Hayes wrote:[color=blue]
> On Mon, 06 Feb 2006 13:39:17 +0000, Steve Holden wrote:[/color]
[...][color=blue][color=green]
>>
>>The URL you reference is discussing how you represent arbitrary values
>>in string literals. If you already have the data in a Python string the
>>best advise is to use a parameterized query - that way your Python DB
>>API module will do the escaping for you!
>>
>>regards
>> Steve[/color]
>
>
> Thanks for the input. I tried that with a format string and a
> dictionary, but I still received a database error indicating illegal
> string values. This error went away completely when I used a test file
> consisting only of text, but reproduced everytime with a true binary file.
> If you can let me know where I am wrong or show me a code snippet with a
> sql insert that contains a variable with raw binary data that works,
> I would greatly appreciate it.
>[/color]
I tried and my experience was exactly the same, which made me think less
of PostgreSQL.

They don't seem to implement the SQL BLOB type properly, so it looks as
though that rebarbative syntax with all the backslashes is necessary. Sorry.

regards
Steve
--
Steve Holden +44 150 684 7255 +1 800 494 3119
Holden Web LLC www.holdenweb.com
PyCon TX 2006 www.python.org/pycon/

[Bengt Richter]

Re: Literal Escaped Octets

On Mon, 06 Feb 2006 04:40:31 GMT, Chason Hayes <chasonh@hotmai l.com> wrote:
[color=blue]
>I am trying to convert raw binary data to data with escaped octets in
>order to store it in a bytea field on postgresql server. I could do this
>easily in c/c++ but I need to do it in python. I am not sure how to read
>and evaluate the binary value of a byte in a long string when it is a non
>printable ascii value in python. I read some ways to use unpack from the
>struct module, but i really couldn't understand where that would help. I
>looked at the MIMIEncode module but I don't know how to convert the object
>to a string. Is there a module that will convert the data? It seems to me
>that this question must have been answered a million times before but I
>can't find anything.
>[/color]
Have you considered just encoding the data as text in hex or base64, e.g.,
[color=blue][color=green][color=darkred]
>>> import binascii
>>> s = '\x00\x01\x02\x 03ABCD0123'
>>> binascii.hexlif y(s)[/color][/color][/color]
'00010203414243 4430313233'[color=blue][color=green][color=darkred]
>>> binascii.b2a_ba se64(s)[/color][/color][/color]
'AAECA0FCQ0QwMT Iz\n'

which is also reversible later of course:[color=blue][color=green][color=darkred]
>>> h = binascii.hexlif y(s)
>>> binascii.unhexl ify(h)[/color][/color][/color]
'\x00\x01\x02\x 03ABCD0123'[color=blue][color=green][color=darkred]
>>> b64 = binascii.b2a_ba se64(s)
>>> binascii.a2b_ba se64(b64)[/color][/color][/color]
'\x00\x01\x02\x 03ABCD0123'

Regards,
Bengt Richter

[Chason Hayes]

Re: Literal Escaped Octets

On Tue, 07 Feb 2006 15:06:49 +0000, Bengt Richter wrote:
[color=blue]
> On Mon, 06 Feb 2006 04:40:31 GMT, Chason Hayes <chasonh@hotmai l.com> wrote:
>[color=green]
>>I am trying to convert raw binary data to data with escaped octets in
>>order to store it in a bytea field on postgresql server. I could do this
>>easily in c/c++ but I need to do it in python. I am not sure how to read
>>and evaluate the binary value of a byte in a long string when it is a non
>>printable ascii value in python. I read some ways to use unpack from the
>>struct module, but i really couldn't understand where that would help. I
>>looked at the MIMIEncode module but I don't know how to convert the object
>>to a string. Is there a module that will convert the data? It seems to me
>>that this question must have been answered a million times before but I
>>can't find anything.
>>[/color]
> Have you considered just encoding the data as text in hex or base64, e.g.,
>[color=green][color=darkred]
> >>> import binascii
> >>> s = '\x00\x01\x02\x 03ABCD0123'
> >>> binascii.hexlif y(s)[/color][/color]
> '00010203414243 4430313233'[color=green][color=darkred]
> >>> binascii.b2a_ba se64(s)[/color][/color]
> 'AAECA0FCQ0QwMT Iz\n'
>
> which is also reversible later of course:[color=green][color=darkred]
> >>> h = binascii.hexlif y(s)
> >>> binascii.unhexl ify(h)[/color][/color]
> '\x00\x01\x02\x 03ABCD0123'[color=green][color=darkred]
> >>> b64 = binascii.b2a_ba se64(s)
> >>> binascii.a2b_ba se64(b64)[/color][/color]
> '\x00\x01\x02\x 03ABCD0123'
>
> Regards,
> Bengt Richter[/color]

I had just about come to that conclusion last night while I was working on
it. I was going to use
import base64
base64.stringen code(binarydata )
and
base64.stringde code(stringdata )

I then wasn't sure if I should still use the bytea field or just use a
text field.

Do you have a suggestion?

[Chason Hayes]

Re: Literal Escaped Octets

On Tue, 07 Feb 2006 01:58:00 +0000, Steve Holden wrote:
[color=blue]
> Chason Hayes wrote:[color=green]
>> On Mon, 06 Feb 2006 13:39:17 +0000, Steve Holden wrote:[/color]
> [...][color=green][color=darkred]
>>>
>>>The URL you reference is discussing how you represent arbitrary values
>>>in string literals. If you already have the data in a Python string the
>>>best advise is to use a parameterized query - that way your Python DB
>>>API module will do the escaping for you!
>>>
>>>regards
>>> Steve[/color]
>>
>>
>> Thanks for the input. I tried that with a format string and a
>> dictionary, but I still received a database error indicating illegal
>> string values. This error went away completely when I used a test file
>> consisting only of text, but reproduced everytime with a true binary file.
>> If you can let me know where I am wrong or show me a code snippet with a
>> sql insert that contains a variable with raw binary data that works,
>> I would greatly appreciate it.
>>[/color]
> I tried and my experience was exactly the same, which made me think less
> of PostgreSQL.
>
> They don't seem to implement the SQL BLOB type properly, so it looks as
> though that rebarbative syntax with all the backslashes is necessary. Sorry.
>
> regards
> Steve[/color]

with regards to escaping data parameters I have found that I have to
specifically add quotes to my strings for them to be understood by
pstgresql. For example

ifs=open("binar ydatafile","r")
binarydata=ifs. read()
stringdata=base 64.encodestring (binarydata)

#does not work
cursor.execute( "insert into binarytable values(%s)" % stringdata)

#need to do this first
newstringdata = "'" + stringdata + "'"

then the select statment works.
Is this expected behavior? Is there a better way of doing this?

thanks for any insight
Chason

[Steve Holden]

Re: Literal Escaped Octets

Chason Hayes wrote:[color=blue]
> On Tue, 07 Feb 2006 01:58:00 +0000, Steve Holden wrote:
>
>[color=green]
>>Chason Hayes wrote:
>>[color=darkred]
>>>On Mon, 06 Feb 2006 13:39:17 +0000, Steve Holden wrote:[/color]
>>
>>[...]
>>[color=darkred]
>>>>The URL you reference is discussing how you represent arbitrary values
>>>>in string literals. If you already have the data in a Python string the
>>>>best advise is to use a parameterized query - that way your Python DB
>>>>API module will do the escaping for you!
>>>>
>>>>regards
>>>> Steve
>>>
>>>
>>>Thanks for the input. I tried that with a format string and a
>>>dictionary , but I still received a database error indicating illegal
>>>string values. This error went away completely when I used a test file
>>>consisting only of text, but reproduced everytime with a true binary file.
>>>If you can let me know where I am wrong or show me a code snippet with a
>>>sql insert that contains a variable with raw binary data that works,
>>>I would greatly appreciate it.
>>>[/color]
>>
>>I tried and my experience was exactly the same, which made me think less
>>of PostgreSQL.
>>
>>They don't seem to implement the SQL BLOB type properly, so it looks as
>>though that rebarbative syntax with all the backslashes is necessary. Sorry.
>>
>>regards
>> Steve[/color]
>
>
> with regards to escaping data parameters I have found that I have to
> specifically add quotes to my strings for them to be understood by
> pstgresql. For example
>
> ifs=open("binar ydatafile","r")
> binarydata=ifs. read()
> stringdata=base 64.encodestring (binarydata)
>
> #does not work
> cursor.execute( "insert into binarytable values(%s)" % stringdata)
>
> #need to do this first
> newstringdata = "'" + stringdata + "'"
>
> then the select statment works.
> Is this expected behavior? Is there a better way of doing this?
>
> thanks for any insight[/color]

Yes, parameterize your queries. I assume you are using psycopg or
something similar to create the database connection (i.e. I something
that expects the "%s" parameter style - there are other options, but we
needn't discuss them here).

The magic incantation you seek is:

cursor.execute( "insert into binarytable values(%s)", (stringdata, ))

Note that here there are TWO arguments to the .execute() method. The
first is a parameterized SQL statement, and the second is a tuple of
data items, one for each parameter mark in the SQL.

Using this technique all necessary quoting (and even data conversion
with a good database module) is performed inside the database driver,
meaning (among other things) that your program is no longer vulnerable
to the dreaded SQL injection errors.

This is the technique I was hoping would work with the bytea datatype,
but alas it doesn't. ISTM that PostgreSQL needs a bit of work there,
even though it is otherwise a very polished product.

regards
Steve
--
Steve Holden +44 150 684 7255 +1 800 494 3119
Holden Web LLC www.holdenweb.com
PyCon TX 2006 www.python.org/pycon/

[Chason Hayes]

Re: Literal Escaped Octets

On Wed, 08 Feb 2006 00:57:45 -0500, Steve Holden wrote:
[color=blue]
> Chason Hayes wrote:[color=green]
>> On Tue, 07 Feb 2006 01:58:00 +0000, Steve Holden wrote:
>>
>>[color=darkred]
>>>Chason Hayes wrote:
>>>
>>>>On Mon, 06 Feb 2006 13:39:17 +0000, Steve Holden wrote:
>>>
>>>[...]
>>>
>>>>>The URL you reference is discussing how you represent arbitrary values
>>>>>in string literals. If you already have the data in a Python string the
>>>>>best advise is to use a parameterized query - that way your Python DB
>>>>>API module will do the escaping for you!
>>>>>
>>>>>regards
>>>>> Steve
>>>>
>>>>
>>>>Thanks for the input. I tried that with a format string and a
>>>>dictionar y, but I still received a database error indicating illegal
>>>>string values. This error went away completely when I used a test file
>>>>consistin g only of text, but reproduced everytime with a true binary file.
>>>>If you can let me know where I am wrong or show me a code snippet with a
>>>>sql insert that contains a variable with raw binary data that works,
>>>>I would greatly appreciate it.
>>>>
>>>
>>>I tried and my experience was exactly the same, which made me think less
>>>of PostgreSQL.
>>>
>>>They don't seem to implement the SQL BLOB type properly, so it looks as
>>>though that rebarbative syntax with all the backslashes is necessary. Sorry.
>>>
>>>regards
>>> Steve[/color]
>>
>>
>> with regards to escaping data parameters I have found that I have to
>> specifically add quotes to my strings for them to be understood by
>> pstgresql. For example
>>
>> ifs=open("binar ydatafile","r")
>> binarydata=ifs. read()
>> stringdata=base 64.encodestring (binarydata)
>>
>> #does not work
>> cursor.execute( "insert into binarytable values(%s)" % stringdata)
>>
>> #need to do this first
>> newstringdata = "'" + stringdata + "'"
>>
>> then the select statment works.
>> Is this expected behavior? Is there a better way of doing this?
>>
>> thanks for any insight[/color]
>
> Yes, parameterize your queries. I assume you are using psycopg or
> something similar to create the database connection (i.e. I something
> that expects the "%s" parameter style - there are other options, but we
> needn't discuss them here).
>
> The magic incantation you seek is:
>
> cursor.execute( "insert into binarytable values(%s)", (stringdata, ))
>
> Note that here there are TWO arguments to the .execute() method. The
> first is a parameterized SQL statement, and the second is a tuple of
> data items, one for each parameter mark in the SQL.
>
> Using this technique all necessary quoting (and even data conversion
> with a good database module) is performed inside the database driver,
> meaning (among other things) that your program is no longer vulnerable
> to the dreaded SQL injection errors.
>
> This is the technique I was hoping would work with the bytea datatype,
> but alas it doesn't. ISTM that PostgreSQL needs a bit of work there,
> even though it is otherwise a very polished product.
>
> regards
> Steve[/color]

That was it. Thanks for your great help.

Chason