Merge "Add comment explaining why we need custom logic to parse int"

utils.cc

@@ -7,6 +7,11 @@
 namespace bsdiff {
 
 int64_t ParseInt64(const uint8_t* buf) {
+  // BSPatch uses a non-standard encoding of integers.
+  // Highest bit of that integer is used as a sign bit, 1 = negative
+  // and 0 = positive.
+  // Therefore, if the highest bit is set, flip it, then do 2's complement
+  // to get the integer in standard form
   int64_t result = buf[7] & 0x7F;
   for (int i = 6; i >= 0; i--) {
     result <<= 8;