# --------------------------------------------------------------------------
# Rules how to make each one of the binaries
-EXELIST=ascii fire hello mousedemo nachtm plasma sieve tgidemo
+EXELIST=ascii fire gunzip65 hello mousedemo nachtm plasma sieve tgidemo
.PHONY: all
all: $(EXELIST)
-ascii: $(CRT0) ascii.o $(CLIB)
- @$(LD) -t $(SYS) -m ascii.map -Ln ascii.lbl -o $@ $^
+ascii: $(CRT0) ascii.o $(CLIB)
+ @$(LD) -t $(SYS) -m $(basename $@).map -o $@ $^
-fire: $(CRT0) fire.o $(CLIB)
- @$(LD) -t $(SYS) -m fire.map -Ln fire.lbl -o $@ $^
+fire: $(CRT0) fire.o $(CLIB)
+ @$(LD) -t $(SYS) -m $(basename $@).map -o $@ $^
-hello: $(CRT0) hello.o $(CLIB)
- @$(LD) -t $(SYS) -m hello.map -Ln hello.lbl -o $@ $^
+gunzip65: $(CRT0) gunzip65.o $(CLIB)
+ @$(LD) -t $(SYS) -m $(basename $@).map -o $@ $^
+
+hello: $(CRT0) hello.o $(CLIB)
+ @$(LD) -t $(SYS) -m $(basename $@).map -o $@ $^
mousedemo: $(CRT0) mousedemo.o $(CLIB)
- @$(LD) -t $(SYS) -m mousedemo.map -Ln mousedemo.lbl -o $@ $^
+ @$(LD) -t $(SYS) -m $(basename $@).map -o $@ $^
-nachtm: $(CRT0) nachtm.o $(CLIB)
- @$(LD) -t $(SYS) -vm -m nachtm.map -Ln nachtm.lbl -o $@ $^
+nachtm: $(CRT0) nachtm.o $(CLIB)
+ @$(LD) -t $(SYS) -m $(basename $@).map -o $@ $^
-plasma: $(CRT0) plasma.o $(CLIB)
- @$(LD) -t $(SYS) -m plasma.map -Ln nachtm.lbl -o $@ $^
+plasma: $(CRT0) plasma.o $(CLIB)
+ @$(LD) -t $(SYS) -m $(basename $@).map -o $@ $^
-sieve: $(CRT0) sieve.o $(CLIB)
- @$(LD) -t $(SYS) -m sieve.map -Ln sieve.lbl -o $@ $^
+sieve: $(CRT0) sieve.o $(CLIB)
+ @$(LD) -t $(SYS) -m $(basename $@).map -o $@ $^
-tgidemo: $(CRT0) tgidemo.o $(CLIB)
- @$(LD) -t $(SYS) -m tgidemo.map -Ln tgidemo.lbl -o $@ $^
+tgidemo: $(CRT0) tgidemo.o $(CLIB)
+ @$(LD) -t $(SYS) -m $(basename $@).map -o $@ $^
# --------------------------------------------------------------------------
--- /dev/null
+/*
+ * gunzip65 - a gunzip utility for 6502-based machines.
+ *
+ * Piotr Fusik <fox@scene.pl>
+ *
+ * This should be considered as a test of my zlib-compatible library
+ * rather than a real application.
+ * It's not user-friendly, fault-tolerant, whatever.
+ * However, it really works for real GZIP files, provided they are small
+ * enough to fit in buffer[] (after decompression!).
+ */
+
+#include <stdio.h>
+#include <string.h>
+#include <zlib.h>
+
+#ifndef __CC65__
+/*
+ * Emulate inflatemem() if using original zlib.
+ * As you can see, this program is quite portable.
+ */
+unsigned inflatemem(char* dest, const char* source)
+{
+ z_stream stream;
+
+ stream.next_in = (Bytef*) source;
+ stream.avail_in = 65535;
+
+ stream.next_out = dest;
+ stream.avail_out = 65535;
+
+ stream.zalloc = (alloc_func) 0;
+ stream.zfree = (free_func) 0;
+
+ inflateInit2(&stream, -MAX_WBITS);
+ inflate(&stream, Z_FINISH);
+ inflateEnd(&stream);
+
+ return stream.total_out;
+}
+#endif /* __CC65__ */
+
+/*
+ * Structure of a GZIP file:
+ *
+ * 1. GZIP header:
+ * Offset 0: Signature (2 bytes: 0x1f, 0x8b)
+ * Offset 2: Compression method (1 byte: 8 == "deflate")
+ * Offset 3: Flags (1 byte: see below)
+ * Offset 4: File date and time (4 bytes)
+ * Offset 8: Extra flags (1 byte)
+ * Offset 9: Target OS (1 byte: DOS, Amiga, Unix, etc.)
+ * if (flags & FEXTRA) { 2 bytes of length, then length bytes }
+ * if (flags & FNAME) { ASCIIZ filename }
+ * if (flags & FCOMMENT) { ASCIIZ comment }
+ * if (flags & FHCRC) { 2 bytes of CRC }
+ *
+ * 2. Deflate compressed data.
+ *
+ * 3. GZIP trailer:
+ * Offset 0: CRC-32 (4 bytes)
+ * Offset 4: uncompressed file length (4 bytes)
+ */
+
+/* Flags in the GZIP header. */
+#define FTEXT 1 /* Extra text */
+#define FHCRC 2 /* Header CRC */
+#define FEXTRA 4 /* Extra field */
+#define FNAME 8 /* File name */
+#define FCOMMENT 16 /* File comment */
+
+/*
+ * We read whole GZIP file into this buffer.
+ * Then we use this buffer for the decompressed data.
+ */
+static unsigned char buffer[26000];
+
+/*
+ * Get a 16-bit little-endian unsigned number, using unsigned char* p.
+ * On many machines this could be (*(unsigned short*) p),
+ * but I really like portability. :-)
+ */
+#define GET_WORD(p) (*(p) + ((unsigned) (p)[1] << 8))
+
+/* Likewise, for a 32-bit number. */
+#define GET_LONG(p) (GET_WORD(p) + ((unsigned long) GET_WORD(p + 2) << 16))
+
+/*
+ * Uncompress a GZIP file.
+ * On entry, buffer[] should contain the whole GZIP file contents,
+ * and the argument complen should be equal to the length of the GZIP file.
+ * On return, buffer[] contains the uncompressed data, and the returned
+ * value is the length of the uncompressed data.
+ */
+unsigned uncompress_buffer(unsigned complen)
+{
+ unsigned char* ptr;
+ unsigned long crc;
+ unsigned long unclen;
+ void* ptr2;
+ unsigned unclen2;
+
+ /* check GZIP signature */
+ if (buffer[0] != 0x1f || buffer[1] != 0x8b) {
+ puts("Not GZIP format");
+ return 0;
+ }
+
+ /* check compression method (it is always (?) "deflate") */
+ if (buffer[2] != 8) {
+ puts("Unsupported compression method");
+ return 0;
+ }
+
+ /* get CRC from GZIP trailer */
+ crc = GET_LONG(buffer + complen - 8);
+
+ /* get uncompressed length from GZIP trailer */
+ unclen = GET_LONG(buffer + complen - 4);
+ if (unclen > sizeof(buffer)) {
+ puts("Uncompressed size too big");
+ return 0;
+ }
+
+ /* skip extra field, file name, comment and crc */
+ ptr = buffer + 10;
+ if (buffer[3] & FEXTRA)
+ ptr = buffer + 12 + GET_WORD(buffer + 10);
+ if (buffer[3] & FNAME)
+ while (*ptr++ != 0);
+ if (buffer[3] & FCOMMENT)
+ while (*ptr++ != 0);
+ if (buffer[3] & FHCRC)
+ ptr += 2;
+
+ /*
+ * calculate length of raw "deflate" data
+ * (without the GZIP header and 8-byte trailer)
+ */
+ complen -= (ptr - buffer) + 8;
+
+ /*
+ * We will move the compressed data to the end of buffer[].
+ * Thus the compressed data and the decompressed data (written from
+ * the beginning of buffer[]) may overlap, as long as the decompressed
+ * data doesn't go further than unread compressed data.
+ * ptr2 points to the beginning of compressed data at the end
+ * of buffer[].
+ */
+ ptr2 = buffer + sizeof(buffer) - complen;
+ /* move the compressed data to end of buffer[] */
+ memmove(ptr2, ptr, complen);
+
+ /* uncompress */
+ puts("Inflating...");
+ unclen2 = inflatemem(buffer, ptr2);
+
+ /* verify uncompressed length */
+ if (unclen2 != (unsigned) unclen) {
+ puts("Uncompressed size does not match");
+ return 0;
+ }
+
+ /* verify CRC */
+ puts("Calculating CRC...");
+ if (crc32(crc32(0L, Z_NULL, 0), buffer, unclen2) != crc) {
+ puts("CRC mismatch");
+ return 0;
+ }
+
+ /* return number of uncompressed bytes */
+ return unclen2;
+}
+
+/*
+ * Get a filename from standard input.
+ */
+char* get_fname(void)
+{
+ static char filename[100];
+ unsigned len;
+ fgets(filename, sizeof(filename), stdin);
+ len = strlen(filename);
+ if (len >= 1 && filename[len - 1] == '\n')
+ filename[len - 1] = '\0';
+ return filename;
+}
+
+int main(void)
+{
+ FILE* fp;
+ unsigned length;
+
+ /* read GZIP file */
+ puts("GZIP file name:");
+ fp = fopen(get_fname(), "rb");
+ if (!fp) {
+ puts("Can't open GZIP file");
+ return 1;
+ }
+ length = fread(buffer, 1, sizeof(buffer), fp);
+ fclose(fp);
+ if (length == sizeof(buffer)) {
+ puts("File is too long");
+ return 1;
+ }
+
+ /* decompress */
+ length = uncompress_buffer(length);
+ if (length == 0)
+ return 1;
+
+ /* write uncompressed file */
+ puts("Uncompressed file name:");
+ fp = fopen(get_fname(), "wb");
+ if (!fp) {
+ puts("Can't create output file");
+ return 1;
+ }
+ if (fwrite(buffer, 1, length, fp) != length) {
+ puts("Error while writing output file");
+ return 1;
+ }
+ fclose(fp);
+
+ puts("Ok.");
+ return 0;
+}
projects / cc65 / commitdiff