Add conversion from float to fixed-point i16

To encode float values between -1 and 1.

PR #3369 <https://github.com/Genymobile/scrcpy/pull/3369>

app/src/util/binary.h

@@ -58,4 +58,19 @@ sc_float_to_u16fp(float f) {
     return (uint16_t) u;
 }
 
+/**
+ * Convert a float between -1 and 1 to a signed 16-bit fixed-point value
+ */
+static inline int16_t
+sc_float_to_i16fp(float f) {
+    assert(f >= -1.0f && f <= 1.0f);
+    int32_t i = f * 0x1p15f; // 2^15
+    assert(i >= -0x8000);
+    if (i >= 0x7fff) {
+        assert(i == 0x8000); // for f == 1.0f
+        i = 0x7fff;
+    }
+    return (int16_t) i;
+}
+
 #endif

app/tests/test_binary.c

@@ -78,6 +78,25 @@ static void test_float_to_u16fp(void) {
     assert(sc_float_to_u16fp(1.0f) == 0xffff);
 }
 
+static void test_float_to_i16fp(void) {
+    assert(sc_float_to_i16fp(0.0f) == 0);
+    assert(sc_float_to_i16fp(0.03125f) == 0x400);
+    assert(sc_float_to_i16fp(0.0625f) == 0x800);
+    assert(sc_float_to_i16fp(0.125f) == 0x1000);
+    assert(sc_float_to_i16fp(0.25f) == 0x2000);
+    assert(sc_float_to_i16fp(0.5f) == 0x4000);
+    assert(sc_float_to_i16fp(0.75f) == 0x6000);
+    assert(sc_float_to_i16fp(1.0f) == 0x7fff);
+
+    assert(sc_float_to_i16fp(-0.03125f) == -0x400);
+    assert(sc_float_to_i16fp(-0.0625f) == -0x800);
+    assert(sc_float_to_i16fp(-0.125f) == -0x1000);
+    assert(sc_float_to_i16fp(-0.25f) == -0x2000);
+    assert(sc_float_to_i16fp(-0.5f) == -0x4000);
+    assert(sc_float_to_i16fp(-0.75f) == -0x6000);
+    assert(sc_float_to_i16fp(-1.0f) == -0x8000);
+}
+
 int main(int argc, char *argv[]) {
     (void) argc;
     (void) argv;
@@ -90,5 +109,6 @@ int main(int argc, char *argv[]) {
     test_read64be();
 
     test_float_to_u16fp();
+    test_float_to_i16fp();
     return 0;
 }