usr: xxd - hex dump a file or stdin

Eight bytes per line rather than the usual sixteen: offset + 8 hex pairs +
the ASCII column is 38 of the terminal's 40 columns, where 16 would wrap
every line and make the dump unreadable. -c overrides it.

No -s (skip): there is no seek syscall, so it streams from the start - the
same reason tail(1) has to keep a ring.

    /# xxd h
    0000: 48 65 6c 6c 6f 20 67 62 Hello gb
    0008: 6f 73 0d 0a             os..

blk(1) could already peek at raw disk blocks; nothing could look inside a
file. Verified against known bytes, from stdin, with -c, through a pipe, and
on a missing file.

5 files changed, 92 insertions, 2 deletions

diff --git a/Makefile b/Makefile
index de7175b..b3f1b0e 100644
--- a/Makefile
+++ b/Makefile
@@ -29,7 +29,7 @@ PROGS  := chello echo true false uname pid cat wc head args \
           ls save rm sh ps kill spin blk mkdir count ptest necho \
           wget chat netd ping nslookup dhcp irc uptime ansi \
           grep more cp nc httpd ircd \
-          df mv sleep tee tail help poweroff
+          df mv sleep tee tail help poweroff xxd
 CBLOBS := $(patsubst %,$(BUILD)/usr/%.bin,$(PROGS))
 $(BUILD)/programs.o: $(CBLOBS)
 
diff --git a/README.md b/README.md
index 8596fe2..0dffb18 100644
--- a/README.md
+++ b/README.md
@@ -49,7 +49,7 @@ into the on-screen console. Regenerate it anytime with `tools/gbdemo` (below).*
   way to stop the machine (the battery save is written on a clean exit).
 - **A C toolchain**: userland is plain C compiled with SDCC (`sm83`); ~45
   tools ship in `usr/` (`cat`, `wc`, `head`, `tail`, `grep`, `cp`, `mv`,
-  `tee`, `more`, `ls`, `df`, `sleep`, `uptime`, …) — `help` lists them all on
+  `tee`, `more`, `ls`, `df`, `xxd`, `sleep`, `uptime`, …) — `help` lists them all on
   the Game Boy itself. Adding a program is one `.c` file + a table entry.
 
 ## Quick start
diff --git a/include/gbos.inc b/include/gbos.inc
index 76981da..1af3fc8 100644
--- a/include/gbos.inc
+++ b/include/gbos.inc
@@ -330,5 +330,6 @@ DEF PROG_TEE    EQU 42
 DEF PROG_TAIL   EQU 43
 DEF PROG_HELP   EQU 44
 DEF PROG_POWEROFF EQU 45
+DEF PROG_XXD    EQU 46
 
     ENDC
diff --git a/src/programs.asm b/src/programs.asm
index 11da2b1..a602350 100644
--- a/src/programs.asm
+++ b/src/programs.asm
@@ -178,6 +178,9 @@ ProgHelp:
 SECTION "prog_poweroff", ROMX[$4000], BANK[47]
 ProgPoweroff:
     INCBIN "build/usr/poweroff.bin"
+SECTION "prog_xxd",    ROMX[$4000], BANK[48]
+ProgXxd:
+    INCBIN "build/usr/xxd.bin"
 
 ; -----------------------------------------------------------------------------
 ; PROG_SH (bank 3) - the shell, now written in C (usr/sh.c): parses >/</| and
@@ -312,6 +315,8 @@ ProgramTable::
     dw ProgHelp
     db LOW(BANK(ProgPoweroff)), HIGH(BANK(ProgPoweroff))
     dw ProgPoweroff
+    db LOW(BANK(ProgXxd)), HIGH(BANK(ProgXxd))
+    dw ProgXxd
 ProgramTableEnd::
 
 ; -----------------------------------------------------------------------------
@@ -411,4 +416,6 @@ NameTable::
     db PROG_HELP
     db "poweroff", 0
     db PROG_POWEROFF
+    db "xxd", 0
+    db PROG_XXD
     db 0
diff --git a/usr/xxd.c b/usr/xxd.c
new file mode 100644
index 0000000..8e255ae
--- /dev/null
+++ b/usr/xxd.c
@@ -0,0 +1,82 @@
+#include "gbos.h"
+/* xxd [-c N] [file]: hex dump a file, or stdin.
+ *
+ * Eight bytes per line, not the usual sixteen: offset + 8 hex pairs + the
+ * ASCII column is 38 of the terminal's 40 columns, where 16 would wrap every
+ * line and make the dump unreadable. -c overrides it anyway.
+ *
+ * There is no seek syscall, so this streams from the start - which is also
+ * why there is no -s (skip). */
+
+#define MAXC 16
+
+static char hexd(unsigned char v) { return (char)(v < 10 ? '0' + v : 'a' + (v - 10)); }
+static void hex2(unsigned char v) {
+    putc(hexd((unsigned char)(v >> 4)));
+    putc(hexd((unsigned char)(v & 15)));
+}
+static void hex4(unsigned int v) {
+    hex2((unsigned char)(v >> 8));
+    hex2((unsigned char)v);
+}
+
+void main(void) {
+    char *argv[8];
+    unsigned char buf[MAXC];
+    unsigned char argc = argv_parse(argv, 8);
+    char *cv = optval(argv, argc, 'c');
+    unsigned char cols = 8, fd = NOFD, n = 0, i;
+    unsigned int off = 0;
+    char *fname = 0;
+    int ch;
+
+    if (cv) {
+        cols = atou(cv);
+        if (cols == 0 || cols > MAXC) cols = 8;
+    }
+    for (i = 0; i < argc; i++) {
+        if (argv[i][0] == '-') continue;
+        if (cv == argv[i]) continue;
+        fname = argv[i];
+        break;
+    }
+    if (fname) {
+        fd = open(fname, O_READ);
+        if (fd == EISDIR) {
+            puts("xxd: is a directory");
+            nl();
+            sexit(2);
+        }
+        if (fd == NOFD) {
+            puts("xxd: no such file");
+            nl();
+            sexit(2);
+        }
+    }
+
+    for (;;) {
+        if (fd == NOFD) {
+            char c = readc();
+            ch = (c == EOF) ? -1 : c;
+        } else ch = fgetc(fd);
+        if (ch >= 0) buf[n++] = (unsigned char)ch;
+        if (n == cols || (ch < 0 && n)) { /* full line, or the last one */
+            hex4(off);
+            puts(": ");
+            for (i = 0; i < cols; i++) {
+                if (i < n) hex2(buf[i]);
+                else {
+                    putc(' ');
+                    putc(' ');
+                } /* pad so ASCII stays aligned */
+                putc(' ');
+            }
+            for (i = 0; i < n; i++) putc((buf[i] >= 32 && buf[i] < 127) ? (char)buf[i] : '.');
+            nl();
+            off += n;
+            n = 0;
+        }
+        if (ch < 0) break;
+    }
+    if (fd != NOFD) close(fd);
+}