Hi Timothy,
I think I’m right in thinking you don’t have enough disc space to copy the file. So to get rid of the “dude” you’ll need to edit in place.
I would write a C program to do this. I would test it on a small file first. Here’s a very rough sketch (no error checking, rued direct on my phone)
#define BlockSize 1024
#define IndexOfDude 123 // needs to be the right value!!
int
main(int argc, char *argv[])
{
int f = open(argv[1], O_RDWR);
int n;
char *buf = malloc(BlockSize);
lseek(f, IndexOfDude + 4, SEEK_SET);
while ((n = read(f, buf, BlockSize)) > 0) {
lseek(f, 0 - (BlockSize + 4), SEEK_CUR);
write(f, buf, BlockSize);
lseek(f, 4, SEEK_CUR);
}
ftruncate(f, lseek(f, -4, SEEK_END);
close(f);
return 0;
}
If your file is > 2Gb you may need to use lseek64. I hope the idea is clear. You overwrite the bytes in the file ffom the start of “dude” with the bytes following “dude”, in chunks. You could probably do this in Smalltalk also. But the key is you only get one chance to get this right on the big file so test thoroughly on a smaller file and then ask what could possibly go wrong with integer sizes if your file is greater than 2^31-1 bytes big.
_,,,^..^,,,_ (phone)
On Jul 15, 2023, at 4:35 AM, gettimothy via Squeak-dev <squeak-dev@lists.squeakfoundation.org> wrote:
oops.
the bzcat method works for appending one bz2 to another.
Unfortunately, during the initial test runs, I concatted a bz2 file with the xml declaration to a bzip file with the word 'dude' in it.
Then, I used that file for the concatenation on the 'true' file.
<?xml version="1.0" encoding="UTF-8"?>
dude
<........rest of properlly formatted XML multi-gigabyte file here....
sigh