2.3.8: finish_input helper for the strict-EOF block in the parser (#5891)

sax_parse keeps its if (result) guard and return value. Also adds the
clang-tidy NOLINT for the 2.3.6 message table.

Signed-off-by: Niels Lohmann <mail@nlohmann.me>

include/nlohmann/detail/input/lexer.hpp

@@ -688,7 +688,7 @@ class lexer : public lexer_base<BasicJsonType>
                 case 0x1E:
                 case 0x1F:
                 {
-                    static const char* const control_character_messages[32] =
+                    static const char* const control_character_messages[32] = // NOLINT(cppcoreguidelines-avoid-c-arrays,hicpp-avoid-c-arrays,modernize-avoid-c-arrays)
                     {
                         "invalid string: control character U+0000 (NUL) must be escaped to \\u0000",
                         "invalid string: control character U+0001 (SOH) must be escaped to \\u0001",

include/nlohmann/detail/input/parser.hpp

@@ -165,20 +165,7 @@ class parser
 
         if (result)
         {
-            if (strict)
-            {
-                // strict mode: next byte must be EOF
-                if (get_token() != token_type::end_of_input)
-                {
-                    return syntax_error(*sax, exception_message(token_type::end_of_input, "value"));
-                }
-            }
-            else
-            {
-                // the caller keeps using the input: position it right after
-                // the value by leaving the character that terminated it
-                m_lexer.release_lookahead();
-            }
+            return finish_input(*sax, strict);
         }
 
         return result;
@@ -202,6 +189,39 @@ class parser
         return m_lexer.get_string();
     }
 
+    /*!
+    @brief apply the end-of-input handling after a value was parsed
+
+    In strict mode, the next token must be the end of the input; otherwise
+    parse_error.101 is reported through @a sax. In non-strict mode, the
+    lookahead is released so the caller can keep reading the input right after
+    the parsed value.
+
+    @param[in,out] sax  the SAX parser to report an error to
+    @param[in] strict   whether to expect the last token to be EOF
+    @return true, or what @a sax returned for the error
+    */
+    template<typename SAX>
+    bool finish_input(SAX& sax, const bool strict)
+    {
+        if (strict)
+        {
+            // strict mode: next byte must be EOF
+            if (get_token() != token_type::end_of_input)
+            {
+                return syntax_error(sax, exception_message(token_type::end_of_input, "value"));
+            }
+        }
+        else
+        {
+            // the caller keeps using the input: position it right after
+            // the value by leaving the character that terminated it
+            m_lexer.release_lookahead();
+        }
+
+        return true;
+    }
+
     /*!
     @brief run a DOM SAX parser to completion and position the lexer
 
@@ -221,20 +241,7 @@ class parser
     {
         sax_parse_internal(&sdp);
 
-        if (strict)
-        {
-            // in strict mode, input must be completely read
-            if (get_token() != token_type::end_of_input)
-            {
-                syntax_error(sdp, exception_message(token_type::end_of_input, "value"));
-            }
-        }
-        else
-        {
-            // the caller keeps using the input: position it right after
-            // the value by leaving the character that terminated it
-            m_lexer.release_lookahead();
-        }
+        static_cast<void>(finish_input(sdp, strict));
 
         return !sdp.is_errored();
     }