ash/Anemone3DS
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Automatically set BGM flag to enabled when doing a BGM install

Browse Source
This commit is contained in:
Alex Taber
2020-06-16 18:45:59 -04:00
parent 6bafa6a22e
commit 36e8104dbb
5 changed files with 67 additions and 70 deletions
Download Patch File Download Diff File Expand all files Collapse all files

53
source/fs.c
View File

@@ -28,7 +28,6 @@
#include "fs.h"
#include "unicode.h"
#include "search.h"
#include <archive.h>
#include <archive_entry.h>
@@ -197,11 +196,11 @@ Result buf_to_file(u32 size, FS_Path path, FS_Archive archive, char *buf)
return 0;
}
u32 decompress_lz_file(FS_Path file_name, char **buf)
u32 decompress_lz_file(FS_Path file_name, FS_Archive archive, char **buf)
{
Handle handle;
Result res = 0;
if (R_FAILED(res = FSUSER_OpenFile(&handle, ArchiveSD, file_name, FS_OPEN_READ, 0))) {
if (R_FAILED(res = FSUSER_OpenFile(&handle, archive, file_name, FS_OPEN_READ, 0))) {
DEBUG("%lu\n", res);
return 0;
}
@@ -218,6 +217,7 @@ u32 decompress_lz_file(FS_Path file_name, char **buf)
FSFILE_Close(handle);
if (temp_buf[0] != 0x11) {
free(temp_buf);
return 0;
}
@@ -290,9 +290,52 @@ u32 decompress_lz_file(FS_Path file_name, char **buf)
return cur_written;
}
u32 compress_lz_file(FS_Path path, char *buf, u32 len)
// This is an awful algorithm to "compress" LZ11 data.
// We do this instead of actual LZ11 compression because of time -
// LZ11 requires a lot of mem searching which is painfully slow on 3DS.
// This process is nearly instant but means the resulting file is actually larger
// than the input file. I don't think this is a problem as the only file we compress like this
// is the theme data installed to extdata in some rare cases, which means only 1 file at most
// is ever compressed like this. I don't think 400 KB is that big a sacrifice for probably
// half a minute or so of time save - I may change my mind on this in the future, especially
// if i figure out a dynamic programming algorithm which ends up being significantly
// faster. Otherwise, I think this is probably a fine implementation.
u32 compress_lz_file_fast(FS_Path path, FS_Archive archive, char *in_buf, u32 size)
{
char *output_buf = calloc(1, size * 2);
u32 output_size = 0;
u32 mask_pos = 0;
u32 bytes_processed = 0;
u8 counter = 0;
if (output_buf == NULL) return 0;
// Set header data for the LZ11 file - 0x11 is version (LZ11), next 3 bytes are size
output_buf[0] = 0x11;
output_buf[3] = (size & 0xFF0000) >> 16;
output_buf[2] = (size & 0xFF00) >> 8;
output_buf[1] = (size & 0xFF);
output_size += 4;
while (bytes_processed < size)
{
if (counter == 0)
{
mask_pos = output_size++;
output_buf[mask_pos] = 0;
}
output_buf[output_size++] = in_buf[bytes_processed++];
if (++counter == 8) counter = 0;
}
buf_to_file(output_size, path, archive, output_buf);
free(output_buf);
return output_size;
}
void remake_file(FS_Path path, FS_Archive archive, u32 size)