Gsoap bug note

The bug of using the _ write Method for socket communication check before sending socket data

Version: gsoap_2.7.13

Operate System: WindowsXP

Describe:

I want to send the HTTP request mulitie times by the same socket. And I coded as follow:

For (size_t retrytimes = 0; retrytimes <3; retrytimes ++)
{
// Whenever connected; send the HTTP request;
If (
Soap_end_send (DOM. Soap)
)
{
// If failed, try to create the socket session;
Soap_connect (DOM. Soap, URL, null );
}
// Once send success, break the loop;
Else
That is {
Break;
}
}

 

The problam is: wheath the CONNECT is estabilsh or not, soap_end_send () always success.

I find that is because some code valide the socket connnect as follow:

In soapstd2.cpp: fsend,

# Ifdef Win32
Nwritten = _ write (soap-> sendfd, S, (unsigned INT) N );
# Else
Nwritten = write (soap-> sendfd, S, (unsigned INT) N );
# Endif

The write is just a normaly is function and it can't sure weather the fild handle is a socket handle. So, if I just open a normal file, it is just return success.

I think send () is more appropriate.

 

 

2 \ Fault Tolerance:

 

Such Tag:

 

<Div> <ul> <span> hello! </Span>
</Div>

Soap_element_end_in (struct soap * soap, const char * tag)

If (TAG & (soap-> mode & soap_xml_strict ))
{Soap_pop_namespace (SOAP );
If (soap_match_tag (soap, soap-> tag, tag ))
{Dbglog (test, soap_message (fdebug, "End Element tag name does not match \ n "));
//// If the end string does not match, manually add the end string to match it.
Soap-> bufidx-= strlen (soap-> tag );
Soap-> bufidx = offset;
Soap-> ahead = soap_tt;
Strncpy (soap-> tag, Tag, strlen (TAG ));

 

 

If the first character in a tag is a Chinese character, garbled characters may occur.

The following is a UTF-8 encoded string:

Soap_get

<Ul>
<Li> company: xianyuan property Jialin store </LI>
<Li> address: No. 134, Jialin Road, Pudong New Area, Shanghai </LI>
</Ul>

 

The parsed result is

<Ul>
<Li> L Division: Lixin Housing </LI>
<Li> 0 address: No.-1041, Changde road </LI>
</Ul>

"Public" and "land" are garbled. This is an algorithm defect in gosp parsing XML.

In the soap_wchar soap_get (struct soap * soap) function, the algorithm for reading a single character is:

If gsoap> ahead contains a character, this character is returned. Otherwise, take a character from gsoap> Buf, gsoap> bufidx ++.

The value of gsoap-> ahead comes from soap_peek_element (), and its logic is:

When the tag <li> is obtained, a UTF-8 character is read. If it is not <, it is determined to be the data (rather than the child node) of the label <li> ). In this case, we need to return the read UTF-8.

Here we do not return gsoap-> bufidx --, but store this value in gosp-> ahead. Here, gsoap-> ahead stores the Unicode encoding of characters.

 

If (C! = Soap_lt)
{* Soap-> tag = '\ 0 ';
If (INT) C = EOF)
Return soap-> error = soap_eof;
Soap_unget (soap, C );

3. Abstraction Level

The gosp Dom has two basic objects: node and attribute. Each object has two attributes: Name and value. Simple rules. Easy to use. However, if the following HTML content exists:

<Div>
The sea is the hometown of <span> You </span>
<Div>

 

Gosp considers the sea as a value attribute in the hometown of <span> You </span>. Instead of parsing <span> as a node.

 

Xmlspy provides a higher level of abstraction than gsoap. In XML, node is an abstract DOM tree, and value represents the real value. The value may be a tag, a value, or a xmpspy. The preceding HTML statement is parsed

 

<Div>
The sea is
<Span>
You
</Span>
Hometown
<Div>

That is, a div node has three subnodes.