What is the matter of my plugin?

Question

i want to decrypt a pdf file when opened that encrypted using the aes.
now i found two method:
1.replace the filesystem
now i just open a normal pdf file,not encrypted.
i read the sample of RplcFileSystem,and modify the MyASFileSysReadProc like this:
static ACCB1 ASSize_t ACCB2 MyASFileSysReadProc (void *ptr, ASSize_t size, ASSize_t count, MDFile f, ASInt32 *pError)
ASSize_t bytecount = defFileSys->read (ptr, size, count, f, pError);
fwrite((char *)ptr,size,count,fp);
return bytecount;
fp is a handle of another file that open before.that's to say i save all the data that read to another file.
but this file is different from the source pdf file,what is the matter,the total bytecount is not equal to the file size.but the content more like the source file.
2.AVConversionConvertStreamToPDFProc
i modified the sample uncompresspdf and register the callback of the AVConversionConvertFromPDFProc,the callback function is like this:
ACCBPROTO1 AVConversionStatus ACCBPROTO2 MYAVConversionConvertStreamToPDFProc(
ASCab inSettings,
AVConversionFlags flags,
ASStm stream,
ASCab metaData,
PDDoc *outPDDoc,
AVStatusMonitorProcs statusMonitor,
AVConversionClientData clientData)
char Buffer[MAX_PATH]={0};
do
count = ASStmRead(Buffer,1,MAX_PATH,stream);
if (count > 0)
fwrite(Buffer,1,count,fp);
ZeroMemory(Buffer,MAX_PATH);
}while(count > 0)
i know this function has some error,but i do not know how to do.
who can tell me how to implement my goal?

Answer 1

> first i use the sdk of version 7
That's OK. 7 supports everything from 40 - 128 bit RC4 and I believe 128-bit AES as well, but I'd have to verify that. RC4 for sure.
> so,at first i should find a way to solve the decrypt the pdf file,and then i can do some other work at that base
Acrobat's internal encryption engine does not expose raw methods for handling various encryption forms - it only has APIs for using it's own forms of encryption in the way Acrobat was designed to use them. You will need to choose an open-source encryption library that supports the algorithms you want to provide your users with, or write your own from scratch. Personally I recommend the crypto++ libraries - I have used it with much success in the past, but there are many other great options as well.
For the workflow you are describing, most of your development won't have anything to do with Acrobat. You'll need to do the file recognition and decryption entirely using non-SDK methods and libraries, and how you do this will depend entirely on which crypto library you decide on.