It's right, at least it should be right... but why are you stripping out
the "\r" characters? They're required according to the HTTP specs.
···
On 11 May 2015 at 21:45, Bar Hofesh <bar.hofesh@safe-t.com> wrote:
Hi,
I'm trying to re-calculate the content length of HTTP packages, but it
seems that I get wrong values (mostly too small)
This is the code --> Rewrite · GitHub
def self.rewrite_content_length(data)
new_size = data[:http_body].bytesize
if data[:http_headers].match(/Content-Length: (.*?)\r?\n/i)
data[:http_headers].gsub!(/Content-Length: (.*?)\r?\n/i,
"Content-Length: #{new_size}\n")
else
data[:http_headers].strip!
data[:http_headers] << "\nContent-Length: #{new_size}\n\n"
end
data
end
Isn't .bytesize is the way to go for this ?
--
*[image: Safe-T.com] Bar Hofesh*
Information Security Architect
Support: (IL)1700700139, 927-9-8666110(ext 231)
Haatzmaut 40 St, first floor.
Beer-Sheva 84150, Israel www.Safe-T.com
My other question is: if there's no Content-Length header, are you checking
for Transfer-Encoding? If you're receiving data, and it doesn't have a
specified length, there's a good chance the server is sending it in chunks.
In that case you could have only part of the final response inside
data[:http_body]
Check this wikipedia article for a starting point:
···
On 11 May 2015 at 21:45, Bar Hofesh <bar.hofesh@safe-t.com> wrote:
Hi,
I'm trying to re-calculate the content length of HTTP packages, but it
seems that I get wrong values (mostly too small)
This is the code --> Rewrite · GitHub
def self.rewrite_content_length(data)
new_size = data[:http_body].bytesize
if data[:http_headers].match(/Content-Length: (.*?)\r?\n/i)
data[:http_headers].gsub!(/Content-Length: (.*?)\r?\n/i,
"Content-Length: #{new_size}\n")
else
data[:http_headers].strip!
data[:http_headers] << "\nContent-Length: #{new_size}\n\n"
end
data
end
Isn't .bytesize is the way to go for this ?
--
*[image: Safe-T.com] Bar Hofesh*
Information Security Architect
Support: (IL)1700700139, 927-9-8666110(ext 231)
Haatzmaut 40 St, first floor.
Beer-Sheva 84150, Israel www.Safe-T.com
I'm trying to re-calculate the content length of HTTP
packages, but it seems that I get wrong values (mostly
too small)
This is the code --> [https://gist.github.com/bararchy/00b50eca203f0999a79e](https://gist.github.com/bararchy/00b50eca203f0999a79e)
def self.rewrite_content_length(data)
new_size = data[:http_body].bytesize
if data[:http_headers].match(/Content-Length:
Yeha, this is a specific part of the code, if
the header is telling me the data will be chunked, I’ll just
buffer it, remove the “chunked” headers and recalculate the
Content-Length of the whole thing
···
On 05/11/2015 02:56 PM, Matthew Kerwin wrote:
My other
question is: if there’s no Content-Length header, are you
checking for Transfer-Encoding? If you’re receiving data, and
it doesn’t have a specified length, there’s a good chance the
server is sending it in chunks. In that case you could have
only part of the final response inside data[:http_body]
I'm trying to re-calculate the content length of HTTP
packages, but it seems that I get wrong values (mostly
too small)
This is the code --> [https://gist.github.com/bararchy/00b50eca203f0999a79e](https://gist.github.com/bararchy/00b50eca203f0999a79e)
def self.rewrite_content_length(data)
new_size = data[:http_body].bytesize
if data[:http_headers].match(/Content-Length:
If you're sticking to one thing, it should be the one in the spec.
···
On 11 May 2015 at 21:55, Bar Hofesh <bar.hofesh@safe-t.com> wrote:
Well, TBH both \r\n and \n works the same, so I just prefer to stick to
one thing along my code
I seems that the bytesize issue arises when UTF8 chars or binary data is
used.
On 05/11/2015 02:53 PM, Matthew Kerwin wrote:
It's right, at least it should be right... but why are you stripping out
the "\r" characters? They're required according to the HTTP specs.
On 11 May 2015 at 21:45, Bar Hofesh <bar.hofesh@safe-t.com> wrote:
Hi,
I'm trying to re-calculate the content length of HTTP packages, but it
seems that I get wrong values (mostly too small)
This is the code --> Rewrite · GitHub
def self.rewrite_content_length(data)
new_size = data[:http_body].bytesize
if data[:http_headers].match(/Content-Length: (.*?)\r?\n/i)
data[:http_headers].gsub!(/Content-Length: (.*?)\r?\n/i,
"Content-Length: #{new_size}\n")
else
data[:http_headers].strip!
data[:http_headers] << "\nContent-Length: #{new_size}\n\n"
end
data
end
Isn't .bytesize is the way to go for this ?
--
*[image: Safe-T.com] Bar Hofesh*
Information Security Architect
Support: (IL)1700700139, 927-9-8666110(ext 231)
Haatzmaut 40 St, first floor.
Beer-Sheva 84150, Israel www.Safe-T.com
*[image: Safe-T.com] Bar Hofesh*
Information Security Architect
Support: (IL)1700700139, 927-9-8666110(ext 231)
Haatzmaut 40 St, first floor.
Beer-Sheva 84150, Israel www.Safe-T.com
same, so I just prefer to stick to one thing along my
code
I seems that the bytesize issue arises when UTF8 chars
or binary data is used.
On 05/11/2015 02:53 PM, Matthew Kerwin wrote:
It's
right, at least it should be right… but why
are you stripping out the “\r” characters?
They’re required according to the HTTP specs.
On 11 May 2015 at 21:45, > > > Bar Hofesh <bar.hofesh@safe-t.com> > > > wrote:
Hi,
I'm trying to re-calculate the content
length of HTTP packages, but it seems that
I get wrong values (mostly too small)
This is the code --> [https://gist.github.com/bararchy/00b50eca203f0999a79e](https://gist.github.com/bararchy/00b50eca203f0999a79e)
def self.rewrite_content_length(data)
new_size = data[:http_body].bytesize
if
"Although the line terminator for the start-line and header fields is
the sequence CRLF, a recipient MAY recognize a single LF as a line
terminator and ignore any preceding CR."
The same thing, but with different emphasis.
Anyway, back on topic, I really don't know why String#bytesize would be
miscounting the number of bytes in a binary (or UTF-8) string. Are you sure
it's getting it wrong? Have you tried dumping the bytes and comparing them
with your replaced header? I use something like this:
class String
def hexdump
puts each_byte.to_a.map{|b|'%02X' % b}.each_slice(16).map{|s|s.join
' '}
end
end
... although it also relies on the 'byte' methods of String. You could use
string.unpack('C*') instead of each_byte.to_a if you don't trust it.
···
On 11 May 2015 at 22:00, Bar Hofesh <bar.hofesh@safe-t.com> wrote:
The line terminator for message-header fields is the sequence CRLF.
However, we recommend that applications, when parsing such headers,
recognize a single LF as a line terminator and ignore the leading CR.
But, I guess you are right, I'll change to \r\n
On 05/11/2015 02:56 PM, Matthew Kerwin wrote:
If you're sticking to one thing, it should be the one in the spec.
On 11 May 2015 at 21:55, Bar Hofesh <bar.hofesh@safe-t.com> wrote:
Well, TBH both \r\n and \n works the same, so I just prefer to stick to
one thing along my code
I seems that the bytesize issue arises when UTF8 chars or binary data is
used.
On 05/11/2015 02:53 PM, Matthew Kerwin wrote:
It's right, at least it should be right... but why are you stripping
out the "\r" characters? They're required according to the HTTP specs.
On 11 May 2015 at 21:45, Bar Hofesh <bar.hofesh@safe-t.com> wrote:
Hi,
I'm trying to re-calculate the content length of HTTP packages, but it
seems that I get wrong values (mostly too small)
This is the code --> Rewrite · GitHub
def self.rewrite_content_length(data)
new_size = data[:http_body].bytesize
if data[:http_headers].match(/Content-Length: (.*?)\r?\n/i)
data[:http_headers].gsub!(/Content-Length: (.*?)\r?\n/i,
"Content-Length: #{new_size}\n")
else
data[:http_headers].strip!
data[:http_headers] << "\nContent-Length: #{new_size}\n\n"
end
data
end
Isn't .bytesize is the way to go for this ?
--
*[image: Safe-T.com] Bar Hofesh*
Information Security Architect
Support: (IL)1700700139, 927-9-8666110(ext 231)
Haatzmaut 40 St, first floor.
Beer-Sheva 84150, Israel www.Safe-T.com
*[image: Safe-T.com] Bar Hofesh*
Information Security Architect
Support: (IL)1700700139, 927-9-8666110(ext 231)
Haatzmaut 40 St, first floor.
Beer-Sheva 84150, Israel www.Safe-T.com
*[image: Safe-T.com] Bar Hofesh*
Information Security Architect
Support: (IL)1700700139, 927-9-8666110(ext 231)
Haatzmaut 40 St, first floor.
Beer-Sheva 84150, Israel www.Safe-T.com
