rtidd@ccels3.mitre.org (Randy Tidd) (08/22/90)
For the application i'm writing, I call a routine written by somebody else that loads an image from a file. This routine takes a file pointer (FILE *) as an argument, and reads the image from there. However, my application uses a database rather than the filesystem, and I don't have files or filenames or file pointers that this routine could use. What I *can* do is get a block of memory that holds the image i'm trying to read from the database. This block of memory is exactly the same size and holds exactly the same thing as the file, except that it resides in memory rather than a file. What I want to do is somehow point a file pointer (FILE *) to this block of memory, and pass the file pointer into the routine to read the image. I'm a fairly experienced Unix systems programmer, but I can't seem to find the right way to do this. All this is going on in a single process so using popen() isn't really the right way. I thought of dumping the image to a file then reading it in, but the images can be very large (upwards of a meg) and for performance reasons this is undesirable. Please respond through e-mail to rtidd@mwunix.mitre.org or the address(es) in the header. Randy Tidd GOOD rtidd%ccels3@mwunix.mitre.org FAST #define DISCLAIM TRUE CHEAP -pick any two