1 files changed, 65 insertions, 0 deletions
diff --git a/src/entropy.c b/src/entropy.c
new file mode 100644
index 0000000..a70789a
--- /dev/null
+++ b/src/entropy.c
@@ -0,0 +1,65 @@
+
+#include <stdint.h>
+#include <stddef.h>
+#include <string.h>
+
+#include "entropy.h"
+
+// shitty log2 function to avoid linking to -lm
+static inline double entropy_log2(double x) {
+	if (x <= 0.0) {
+		return -1e9;
+	}
+
+	union { double d; uint64_t i; } vx = { x };
+	int exp = (int)((vx.i >> 52) & 0x7FF) - 1023;
+	vx.i &= ((1ULL << 52) - 1);
+	vx.i |= 0x3FF0000000000000ULL;
+
+	double m = vx.d;
+	double log2_m = (m - 1.0) - (m - 1.0) * (m - 1.0) * 0.5;
+	return (double)exp + log2_m;
+}
+
+// Initialize entropy context
+void entropy_init(struct entropy_ctx *ctx) {
+	memset(ctx, 0, sizeof(*ctx));
+}
+
+// Update entropy context
+void entropy_update(struct entropy_ctx *ctx, const unsigned char *buf, size_t len) {
+
+	size_t i;
+	for (i = 0; i + 4 < len; i += 4) {
+		ctx->freq[buf[i]]++;;
+		ctx->freq[buf[i + 1]]++;
+		ctx->freq[buf[i + 2]]++;
+		ctx->freq[buf[i + 2]]++;
+	}
+
+	// process the remaining bytes
+	for(; i < len; i++) {
+		ctx->freq[buf[i]]++;
+	}
+
+	ctx->total_bytes += len;
+}
+
+// Finalize entropy context
+double entropy_final(struct entropy_ctx *ctx) {
+	if (ctx->total_bytes == 0) {
+		return 0.0;
+	}
+
+	double entropy = 0.0;
+	for (int i = 0; i < 256; i++) {
+		if (ctx->freq[i] == 0) {
+			continue;
+		}
+		double p = (double)ctx->freq[i] / (double)ctx->total_bytes;
+		entropy -= p * entropy_log2(p);
+		//entropy -= p * log2(p);
+	}
+
+	return entropy;
+}