Hello Berin,
Yes, patch solves first problem. Thank you. Do you have any idea
what could be solution for second problem?
Best regards,
Milan
> -----Original Message-----
> From: Berin Lautenbach [mailto:[EMAIL PROTECTED]
> Sent: Saturday, October 02, 2004 2:31 AM
> To: [EMAIL PROTECTED]
> Subject: Re: [java & c++] interoperability & encryption
>
>
> Matej/Milan,
>
> There appears to be two separate problems. The first (cpp
> encrypt->java
> decrypt) is the one Matej indicated - the RSA encryption within the
> windows crypto API stores the bytes in reverse order. I have
> pasted a
> diff against WinCapiCryptoKeyRSA.cpp that will fix that problem.
>
> The second problem (java->cpp) is proving harder to find. I
> find that I
> cannot decrypt your Java using OpenSSL as the provider either
> - which is
> strange. So I'm going to have a play with the Java library
> and see if I
> can work out what it is.
>
> Can you verify whether the patch below fixes the cpp->java problem?
>
> Cheers,
> Berin
>
> cvs -z3 diff WinCAPICryptoKeyRSA.cpp (in directory
> C:\prog\SRC\xml-security\c\src\enc\WinCAPI\)
> Index: WinCAPICryptoKeyRSA.cpp
> ===================================================================
> RCS file:
> /home/cvs/xml-security/c/src/enc/WinCAPI/WinCAPICryptoKeyRSA.cpp,v
> retrieving revision 1.7
> diff -r1.7 WinCAPICryptoKeyRSA.cpp
> 507a508
> > // Have to reverse ordering of input :
> 509c510,512
> < memcpy(plainBuf, inBuf, inLength);
> ---
> > // memcpy(plainBuf, inBuf, inLength);
> > for (unsigned int i = 0; i < inLength; ++i)
> > plainBuf[i] = inBuf[inLength - 1 - i];
> 633a637,644
> > // Reverse the output
> > unsigned char *tbuf;
> > XSECnew(tbuf, unsigned char[encryptSize]);
> > ArrayJanitor<unsigned char> j_tbuf(tbuf);
> > memcpy(tbuf, cipherBuf, encryptSize);
> >
> > for (unsigned int i = 0; i < encryptSize; ++i)
> > cipherBuf[i] = tbuf[encryptSize - 1 - i];
>
>
>
