From b28df24362efd6c70b6d4ebc43f90b18eecbe3dc Mon Sep 17 00:00:00 2001
From: hjp <hjp>
Date: Fri, 20 Mar 1998 20:09:53 +0000
Subject: [PATCH] Initial release: prints arguments as strings, floats, and
 doubles decimal and binary.

---
 ieeefloat/GNUmakefile |  13 ++++
 ieeefloat/ieeefloat.c | 167 ++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 180 insertions(+)
 create mode 100644 ieeefloat/GNUmakefile
 create mode 100644 ieeefloat/ieeefloat.c

diff --git a/ieeefloat/GNUmakefile b/ieeefloat/GNUmakefile
new file mode 100644
index 0000000..2d0ac06
--- /dev/null
+++ b/ieeefloat/GNUmakefile
@@ -0,0 +1,13 @@
+include GNUmakevars
+include GNUmakerules
+
+all: ieeefloat
+install: $(BINDIR)/ieeefloat
+
+clean:
+	rm -f *.bak *.o core ieeefloat
+
+distclean: clean
+	rm -f *.d
+
+-include *.d
diff --git a/ieeefloat/ieeefloat.c b/ieeefloat/ieeefloat.c
new file mode 100644
index 0000000..866c76e
--- /dev/null
+++ b/ieeefloat/ieeefloat.c
@@ -0,0 +1,167 @@
+char ieeefloat_c_rcs_id[] = 
+    "$Id: ieeefloat.c,v 1.1 1998-03-20 20:09:53 hjp Exp $";
+/* ieeefloat: print binary representations of IEEE 754 FP numbers.
+ *
+ * $Log: ieeefloat.c,v $
+ * Revision 1.1  1998-03-20 20:09:53  hjp
+ * Initial release:
+ * prints arguments as strings, floats, and doubles decimal and binary.
+ *
+ */
+#include <assert.h>
+#include <limits.h>
+#include <stdio.h>
+#include <stdlib.h>
+
+
+#define BITS_FLT 32
+#define EXC_FLT 127
+#define MANT_FLT 23
+#define SIGN_FLT ((floatint)1<<(BITS_FLT-1))
+
+#define BITS_DBL 64
+#define EXC_DBL 1023
+#define MANT_DBL 52
+#define SIGN_DBL ((doubleint)1<<(BITS_DBL-1))
+
+#if ULONG_MAX/2+1 == (1UL << (BITS_FLT-1)) 
+typedef unsigned long int floatint;
+#elif UINT_MAX/2+1 == (1U << (BITS_FLT-1))
+typedef unsigned int floatint;
+#else
+#error "no integral type with BITS_FLT bits"
+#endif
+
+#if ULONG_MAX/2+1 == (1UL << (BITS_DBL-1))
+typedef unsigned long int doubleint;
+#else
+/* We cannot portably test for the existance of long long.
+ * If it doesn't exist, either the compiler or sanitychecks
+ * will complain
+ */
+typedef unsigned long long int doubleint;
+#endif
+
+
+
+
+char *cmnd;
+
+
+static void usage(void) {
+    fprintf(stderr, "Usage: %s fp-number ...\n", cmnd);
+    exit(1);
+}
+
+
+static void printbinary (doubleint b, int d) {
+    while (--d >= 0) {
+	putchar('0' + ((b >> d) & 1));
+    }
+}
+
+
+static void printfloat(float f) {
+    union {
+	floatint i;
+	float f;
+    } u;
+    floatint e;
+    floatint m;
+
+    u.f = f;
+
+    printf("%24.7g: ", f);
+    printf("%c ", (u.i & SIGN_FLT) ? '-' : '+');
+    u.i &= ~SIGN_FLT;
+    e = u.i >> MANT_FLT;
+    printf("   ");
+    printbinary(e, BITS_FLT-1-MANT_FLT);
+
+    m = (u.i & (((floatint)1 << MANT_FLT) - 1));
+    if (e == 0) {	/* denormalized */
+	printf(" [0.]");
+    } else {
+	printf(" [1.]");
+    }
+    printbinary(m, MANT_FLT);
+    printf("\n");
+}
+
+
+static void printdouble(double f) {
+    union {
+	doubleint i;
+	double f;
+    } u;
+    doubleint e;
+    doubleint m;
+
+    u.f = f;
+
+    printf("%24.17g: ", f);
+    printf("%c ", (u.i & SIGN_DBL) ? '-' : '+');
+    u.i &= ~SIGN_DBL;
+    e = u.i >> MANT_DBL;
+    printbinary(e, BITS_DBL-1-MANT_DBL);
+
+    m = (u.i & (((doubleint)1 << MANT_DBL) - 1));
+    if (e == 0) {
+	printf(" [0.]");
+    } else {
+	printf(" [1.]");
+    }
+    printbinary(m, MANT_DBL);
+    printf("\n");
+}
+
+
+static void sanitychecks(void) {
+    union {
+	floatint i;
+	float f;
+    } fu;
+    union {
+	doubleint i;
+	double f;
+    } du;
+
+    /* size */
+    assert (sizeof(floatint) == sizeof(float));
+    assert (sizeof(doubleint) == sizeof(double));
+
+    /* some numbers */
+    fu.f = 0;
+    assert (fu.i == 0);
+    du.f = 0;
+    assert (du.i == 0);
+
+    fu.f = 1.0;
+    assert (fu.i == (floatint)(0 + EXC_FLT) << MANT_FLT);
+    du.f = 1.0;
+    assert (du.i == (doubleint)(0 + EXC_DBL) << MANT_DBL);
+
+    fu.f = 1.5;
+    assert (fu.i ==
+            (((floatint)(0 + EXC_FLT) << MANT_FLT) | (1 << (MANT_FLT-1))));
+    du.f = 1.5;
+    assert (du.i == 
+            (((doubleint)(0 + EXC_DBL) << MANT_DBL) | ((doubleint)1 << (MANT_DBL-1))));
+}
+
+
+int main(int argc, char**argv) {
+    int i;
+
+    cmnd = argv[0];
+    sanitychecks();
+
+    for (i = 1; i < argc; i++) {
+	double d = strtod(argv[i], NULL);
+	printf("%s\n", argv[i]);
+	printdouble(d);
+	printfloat(d);
+    }
+
+    return 0;
+}