Merge pull request #443 from biojppm/fix/442

numbers: fix serialization with `+`, fix emit style heuristics

changelog/current.md

@@ -1,3 +1,7 @@
 ## Fixes
 
+- Fix [#442](https://github.com/biojppm/rapidyaml/issues/442) ([PR#443](https://github.com/biojppm/rapidyaml/pull/443)):
+  - Ensure leading `+` is accepted when deserializing numbers.
+  - Ensure numbers are not quoted by fixing the heuristics in `scalar_style_query_plain()` and `scalar_style_choose()`.
+  - Add quickstart sample for overflow detection (only of integral types).
 - Parse engine: cleanup unused macros

samples/quickstart.cpp

@@ -2488,6 +2488,34 @@ void sample_fundamental_types()
     tree["nan" ] >> f; CHECK(std::isnan(f));
     tree["nan" ] >> d; CHECK(std::isnan(d));
     C4_SUPPRESS_WARNING_GCC_CLANG_POP
+
+    // value overflow detection:
+    // (for integral types only)
+    {
+        // we will be detecting errors below, so we use this sample helper
+        ScopedErrorHandlerExample err = {};
+        ryml::Tree t(err.callbacks()); // instantiate with the error-detecting callbacks
+        // create a simple tree with an int value
+        ryml::parse_in_arena(R"({val: 258})", &t);
+        // by default, overflow is not detected:
+        uint8_t valu8 = 0;
+        int8_t vali8 = 0;
+        t["val"] >> valu8; CHECK(valu8 == 2); // not 257; it wrapped around
+        t["val"] >> vali8; CHECK(vali8 == 2); // not 257; it wrapped around
+        // ...but there are facilities to detect overflow
+        CHECK(ryml::overflows<uint8_t>(t["val"].val()));
+        CHECK(ryml::overflows<int8_t>(t["val"].val()));
+        CHECK( ! ryml::overflows<int16_t>(t["val"].val()));
+        // and there is a format helper
+        CHECK(err.check_error_occurs([&]{
+            auto checku8 = ryml::fmt::overflow_checked(valu8); // need to declare the wrapper type before using it with >>
+            t["val"] >> checku8; // this will cause an error
+        }));
+        CHECK(err.check_error_occurs([&]{
+            auto checki8 = ryml::fmt::overflow_checked(vali8); // need to declare the wrapper type before using it with >>
+            t["val"] >> checki8; // this will cause an error
+        }));
+    }
 }
 
 

src/c4/yml/node.hpp

@@ -1604,31 +1604,49 @@ inline void write(NodeRef *n, T const& v)
     n->set_val_serialized(v);
 }
 
+namespace detail {
+// SFINAE overloads for skipping leading + which cannot be read by the charconv functions
+template<class T>
+C4_ALWAYS_INLINE auto read_skip_plus(csubstr val, T *v)
+    -> typename std::enable_if<std::is_arithmetic<T>::value, bool>::type
+{
+    if(val.begins_with('+'))
+        val = val.sub(1);
+    return from_chars(val, v);
+}
+template<class T>
+C4_ALWAYS_INLINE auto read_skip_plus(csubstr val, T *v)
+    -> typename std::enable_if< ! std::is_arithmetic<T>::value, bool>::type
+{
+    return from_chars(val, v);
+}
+} // namespace detail
+
 /** convert the val of a scalar node to a particular type, by
  * forwarding its val to @ref from_chars<T>(). The full string is
  * used.
  * @return false if the conversion failed */
 template<class T>
-typename std::enable_if< ! std::is_floating_point<T>::value, bool>::type
-inline read(NodeRef const& n, T *v)
+inline auto read(NodeRef const& n, T *v)
+    -> typename std::enable_if< ! std::is_floating_point<T>::value, bool>::type
 {
     csubstr val = n.val();
     if(val.empty())
         return false;
-    return from_chars(val, v);
+    return detail::read_skip_plus(val, v);
 }
 /** convert the val of a scalar node to a particular type, by
  * forwarding its val to @ref from_chars<T>(). The full string is
  * used.
  * @return false if the conversion failed */
 template<class T>
-typename std::enable_if< ! std::is_floating_point<T>::value, bool>::type
-inline read(ConstNodeRef const& n, T *v)
+inline auto read(ConstNodeRef const& n, T *v)
+    -> typename std::enable_if< ! std::is_floating_point<T>::value, bool>::type
 {
     csubstr val = n.val();
     if(val.empty())
         return false;
-    return from_chars(val, v);
+    return detail::read_skip_plus(val, v);
 }
 
 /** convert the val of a scalar node to a floating point type, by

src/c4/yml/node_type.cpp

@@ -153,6 +153,10 @@ bool scalar_style_query_plain(csubstr s) noexcept
         else if(s.sub(2).is_number())
             return true;
     }
+    else if(s.begins_with_any("0123456789.-+") && s.is_number())
+    {
+        return true;
+    }
     return s != ':'
         && ( ! s.begins_with_any("-:?*&,'\"{}[]|>%#@`\r")) // @ and ` are reserved characters
         && ( ! s.ends_with_any(":#"))

src/c4/yml/tree.hpp

@@ -64,6 +64,10 @@ bool from_chars_float(csubstr buf, T *C4_RESTRICT val)
     {
         return true;
     }
+    else if(C4_UNLIKELY(buf.begins_with('+')))
+    {
+        return from_chars(buf.sub(1), val);
+    }
     else if(C4_UNLIKELY(buf == ".nan" || buf == ".NaN" || buf == ".NAN"))
     {
         *val = std::numeric_limits<T>::quiet_NaN();

test/test_node_type.cpp

@@ -129,29 +129,132 @@ TEST(NodeType, type_str)
 #undef teststr
 }
 
-TEST(NodeType, scalar_style_choose)
-{
-    EXPECT_EQ(scalar_style_choose(" \n\t"), SCALAR_DQUO);
-    EXPECT_EQ(scalar_style_choose("01"), SCALAR_PLAIN);
-}
-
 TEST(NodeType, scalar_style_choose_json)
 {
     EXPECT_EQ(scalar_style_json_choose("true"), SCALAR_PLAIN);
     EXPECT_EQ(scalar_style_json_choose("false"), SCALAR_PLAIN);
     EXPECT_EQ(scalar_style_json_choose("null"), SCALAR_PLAIN);
     EXPECT_EQ(scalar_style_json_choose("0.1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_json_choose("-0.1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_json_choose("+0.1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_json_choose(".1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_json_choose("+.1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_json_choose("-.1"), SCALAR_PLAIN);
     EXPECT_EQ(scalar_style_json_choose("01"), SCALAR_DQUO);
     EXPECT_EQ(scalar_style_json_choose("foo"), SCALAR_DQUO);
     EXPECT_EQ(scalar_style_json_choose("bar"), SCALAR_DQUO);
 }
 
+TEST(NodeType, scalar_style_choose)
+{
+    EXPECT_EQ(scalar_style_choose(" \n\t"), SCALAR_DQUO);
+    EXPECT_EQ(scalar_style_choose("-.inf"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("-.INF"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("-.034"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("-.034x"), SCALAR_SQUO);
+    EXPECT_EQ(scalar_style_choose("2.35e-10"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("-2.35e-10"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("+2.35e-10"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("0.1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("-0.1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("+0.1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("01"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("0x1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("0o1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("0b1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("0x1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("0o1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("0b1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("-01"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("-0x1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("-0o1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("-0b1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("-0x1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("-0o1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("-0b1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("+01"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("+0x1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("+0o1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("+0b1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("+0x1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("+0o1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("+0b1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("01"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("0X1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("0O1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("0B1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("0X1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("0O1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("0B1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("-01"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("-0X1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("-0O1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("-0B1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("-0X1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("-0O1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("-0B1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("+01"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("+0X1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("+0O1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("+0B1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("+0X1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("+0O1"), SCALAR_PLAIN);
+    EXPECT_EQ(scalar_style_choose("+0B1"), SCALAR_PLAIN);
+}
+
 TEST(NodeType, scalar_style_query_plain)
 {
     EXPECT_TRUE(scalar_style_query_plain("-.inf"));
     EXPECT_TRUE(scalar_style_query_plain("-.INF"));
     EXPECT_TRUE(scalar_style_query_plain("-.034"));
     EXPECT_FALSE(scalar_style_query_plain("-.034x"));
+    EXPECT_TRUE(scalar_style_query_plain("2.35e-10"));
+    EXPECT_TRUE(scalar_style_query_plain("-2.35e-10"));
+    EXPECT_TRUE(scalar_style_query_plain("+2.35e-10"));
+    EXPECT_TRUE(scalar_style_query_plain("0.1"));
+    EXPECT_TRUE(scalar_style_query_plain("-0.1"));
+    EXPECT_TRUE(scalar_style_query_plain("+0.1"));
+    EXPECT_TRUE(scalar_style_query_plain("01"));
+    EXPECT_TRUE(scalar_style_query_plain("0x1"));
+    EXPECT_TRUE(scalar_style_query_plain("0o1"));
+    EXPECT_TRUE(scalar_style_query_plain("0b1"));
+    EXPECT_TRUE(scalar_style_query_plain("0x1"));
+    EXPECT_TRUE(scalar_style_query_plain("0o1"));
+    EXPECT_TRUE(scalar_style_query_plain("0b1"));
+    EXPECT_TRUE(scalar_style_query_plain("+01"));
+    EXPECT_TRUE(scalar_style_query_plain("+0x1"));
+    EXPECT_TRUE(scalar_style_query_plain("+0o1"));
+    EXPECT_TRUE(scalar_style_query_plain("+0b1"));
+    EXPECT_TRUE(scalar_style_query_plain("+0x1"));
+    EXPECT_TRUE(scalar_style_query_plain("+0o1"));
+    EXPECT_TRUE(scalar_style_query_plain("+0b1"));
+    EXPECT_TRUE(scalar_style_query_plain("-01"));
+    EXPECT_TRUE(scalar_style_query_plain("-0x1"));
+    EXPECT_TRUE(scalar_style_query_plain("-0o1"));
+    EXPECT_TRUE(scalar_style_query_plain("-0b1"));
+    EXPECT_TRUE(scalar_style_query_plain("-0x1"));
+    EXPECT_TRUE(scalar_style_query_plain("-0o1"));
+    EXPECT_TRUE(scalar_style_query_plain("-0b1"));
+    EXPECT_TRUE(scalar_style_query_plain("0X1"));
+    EXPECT_TRUE(scalar_style_query_plain("0O1"));
+    EXPECT_TRUE(scalar_style_query_plain("0B1"));
+    EXPECT_TRUE(scalar_style_query_plain("0X1"));
+    EXPECT_TRUE(scalar_style_query_plain("0O1"));
+    EXPECT_TRUE(scalar_style_query_plain("0B1"));
+    EXPECT_TRUE(scalar_style_query_plain("+01"));
+    EXPECT_TRUE(scalar_style_query_plain("+0X1"));
+    EXPECT_TRUE(scalar_style_query_plain("+0O1"));
+    EXPECT_TRUE(scalar_style_query_plain("+0B1"));
+    EXPECT_TRUE(scalar_style_query_plain("+0X1"));
+    EXPECT_TRUE(scalar_style_query_plain("+0O1"));
+    EXPECT_TRUE(scalar_style_query_plain("+0B1"));
+    EXPECT_TRUE(scalar_style_query_plain("-01"));
+    EXPECT_TRUE(scalar_style_query_plain("-0X1"));
+    EXPECT_TRUE(scalar_style_query_plain("-0O1"));
+    EXPECT_TRUE(scalar_style_query_plain("-0B1"));
+    EXPECT_TRUE(scalar_style_query_plain("-0X1"));
+    EXPECT_TRUE(scalar_style_query_plain("-0O1"));
+    EXPECT_TRUE(scalar_style_query_plain("-0B1"));
 }