From eab76c5312504aacba2ff4c82cf44726171a9e32 Mon Sep 17 00:00:00 2001
From: nemtrif <ntrifunovic@hotmail.com>
Date: Sun, 6 Aug 2023 07:32:16 -0400
Subject: [PATCH] next16

---
 source/utf8/checked.h   | 14 ++++++++++++++
 source/utf8/core.h      | 36 ++++++++++++++++++++++++++++++++++++
 source/utf8/unchecked.h |  9 +++++++++
 3 files changed, 59 insertions(+)

diff --git a/source/utf8/checked.h b/source/utf8/checked.h
index 2d159f5..c10f9b3 100644
--- a/source/utf8/checked.h
+++ b/source/utf8/checked.h
@@ -167,6 +167,20 @@ namespace utf8
         return cp;
     }
 
+    template <typename word_iterator>
+    utfchar32_t next16(word_iterator& it, word_iterator end)
+    {
+        utfchar32_t cp = 0;
+        internal::utf_error err_code = utf8::internal::validate_next16(it, end, cp);
+        switch (err_code) {
+            case internal::UTF8_OK :
+                break;
+            case internal::NOT_ENOUGH_ROOM :
+                throw not_enough_room();
+        }
+        return cp;
+    }
+
     template <typename octet_iterator>
     utfchar32_t peek_next(octet_iterator it, octet_iterator end)
     {
diff --git a/source/utf8/core.h b/source/utf8/core.h
index 096e529..4494c53 100644
--- a/source/utf8/core.h
+++ b/source/utf8/core.h
@@ -305,6 +305,42 @@ namespace internal
         return utf8::internal::validate_next(it, end, ignored);
     }
 
+    template <typename word_iterator>
+    utf_error validate_next16(word_iterator& it, word_iterator end, utfchar32_t& code_point)
+    {
+        if (it == end)
+            return NOT_ENOUGH_ROOM;
+        // Save the original value of it so we can go back in case of failure
+        // Of course, it does not make much sense with i.e. stream iterators
+        word_iterator original_it = it;
+
+        utf_error err = UTF8_OK;
+
+        const utfchar16_t first_word = *it++;
+        if (!is_surrogate(first_word)) {
+            code_point = first_word;
+            return UTF8_OK;
+        }
+        else {
+            if (it == end)
+                err = NOT_ENOUGH_ROOM;
+            else if (is_lead_surrogate(first_word)) {
+                const utfchar16_t second_word = *it++;
+                if (is_trail_surrogate(second_word)) {
+                    code_point = (first_word << 10) + second_word + SURROGATE_OFFSET;
+                    return UTF8_OK;
+                } else 
+                    err = INCOMPLETE_SEQUENCE; 
+                
+            } else {
+                err = INVALID_LEAD;               
+            }
+        }
+        // error branch
+        it = original_it;
+        return err;
+    }
+
     // Internal implementation of both checked and unchecked append() function
     // This function will be invoked by the overloads below, as they will know
     // the octet_type.
diff --git a/source/utf8/unchecked.h b/source/utf8/unchecked.h
index 835c429..94f10a5 100644
--- a/source/utf8/unchecked.h
+++ b/source/utf8/unchecked.h
@@ -136,6 +136,15 @@ namespace utf8
             return utf8::unchecked::next(it);
         }
 
+        template <typename word_iterator>
+        utfchar32_t next16(word_iterator& it)
+        {
+            utfchar32_t cp = utf8::internal::mask16(*it++);
+            if (utf8::internal::is_lead_surrogate(cp))
+                return (cp << 10) + *it++ + utf8::internal::SURROGATE_OFFSET;
+            return cp;
+        }
+
         template <typename octet_iterator>
         utfchar32_t prior(octet_iterator& it)
         {