Fixed printing of nan floats/doubles in Python.

The second assert in _upb_EncodeRoundTripFloat is raised if val is a nan. This fix just returns the output of first spnprintf.

I am not sure how changes to this repo are made so feel free to ignore this CL.

To test this, you could
1. Define a proto with a float field
message Test {
  float val = 1;
}
2. In a python script, import the library and then set the val to nan and try to print it.
proto = Test(val=float('nan'))
print(proto)

This will cause a coredump due to assertion error:

assert.h assertion failed at third_party/upb/upb/lex/round_trip.c:46 in void _upb_EncodeRoundTripFloat(float, char *, size_t): strtof(buf, NULL) == val

Added the corresponding change to double too

PiperOrigin-RevId: 637127851

python/google/protobuf/internal/message_test.py

@@ -382,13 +382,23 @@ def testFloatPrinting(self, message_module):
     message.optional_float = 2.0
     self.assertEqual(str(message), 'optional_float: 2.0\n')
 
+  def testFloatNanPrinting(self, message_module):
+    message = message_module.TestAllTypes()
+    message.optional_float = float('nan')
+    self.assertEqual(str(message), 'optional_float: nan\n')
+
   def testHighPrecisionFloatPrinting(self, message_module):
     msg = message_module.TestAllTypes()
     msg.optional_float = 0.12345678912345678
     old_float = msg.optional_float
     msg.ParseFromString(msg.SerializeToString())
     self.assertEqual(old_float, msg.optional_float)
 
+  def testDoubleNanPrinting(self, message_module):
+    message = message_module.TestAllTypes()
+    message.optional_double = float('nan')
+    self.assertEqual(str(message), 'optional_double: nan\n')
+
   def testHighPrecisionDoublePrinting(self, message_module):
     msg = message_module.TestAllTypes()
     msg.optional_double = 0.12345678912345678

upb/lex/BUILD

@@ -41,6 +41,16 @@ cc_test(
     ],
 )
 
+cc_test(
+    name = "round_trip_test",
+    srcs = ["round_trip_test.cc"],
+    deps = [
+        ":lex",
+        "@com_google_googletest//:gtest",
+        "@com_google_googletest//:gtest_main",
+    ],
+)
+
 # begin:github_only
 filegroup(
     name = "source_files",

upb/lex/round_trip.c

@@ -8,6 +8,8 @@
 #include "upb/lex/round_trip.h"
 
 #include <float.h>
+#include <math.h>
+#include <stdio.h>
 #include <stdlib.h>
 
 // Must be last.
@@ -28,6 +30,10 @@ static void upb_FixLocale(char* p) {
 
 void _upb_EncodeRoundTripDouble(double val, char* buf, size_t size) {
   assert(size >= kUpb_RoundTripBufferSize);
+  if (isnan(val)) {
+    snprintf(buf, size, "%s", "nan");
+    return;
+  }
   snprintf(buf, size, "%.*g", DBL_DIG, val);
   if (strtod(buf, NULL) != val) {
     snprintf(buf, size, "%.*g", DBL_DIG + 2, val);
@@ -38,6 +44,10 @@ void _upb_EncodeRoundTripDouble(double val, char* buf, size_t size) {
 
 void _upb_EncodeRoundTripFloat(float val, char* buf, size_t size) {
   assert(size >= kUpb_RoundTripBufferSize);
+  if (isnan(val)) {
+    snprintf(buf, size, "%s", "nan");
+    return;
+  }
   snprintf(buf, size, "%.*g", FLT_DIG, val);
   if (strtof(buf, NULL) != val) {
     snprintf(buf, size, "%.*g", FLT_DIG + 3, val);

upb/lex/round_trip_test.cc

@@ -0,0 +1,35 @@
+#include "upb/lex/round_trip.h"
+
+#include <math.h>
+
+#include <gtest/gtest.h>
+
+namespace {
+
+TEST(RoundTripTest, Double) {
+  char buf[32];
+
+  _upb_EncodeRoundTripDouble(0.123456789, buf, sizeof(buf));
+  EXPECT_STREQ(buf, "0.123456789");
+
+  _upb_EncodeRoundTripDouble(0.0, buf, sizeof(buf));
+  EXPECT_STREQ(buf, "0");
+
+  _upb_EncodeRoundTripDouble(nan(""), buf, sizeof(buf));
+  EXPECT_STREQ(buf, "nan");
+}
+
+TEST(RoundTripTest, Float) {
+  char buf[32];
+
+  _upb_EncodeRoundTripFloat(0.123456, buf, sizeof(buf));
+  EXPECT_STREQ(buf, "0.123456");
+
+  _upb_EncodeRoundTripFloat(0.0, buf, sizeof(buf));
+  EXPECT_STREQ(buf, "0");
+
+  _upb_EncodeRoundTripFloat(nan(""), buf, sizeof(buf));
+  EXPECT_STREQ(buf, "nan");
+}
+
+}  // namespace