From 442b0e7ceadffa6ff380d6c7d5a89af3f09bd66a Mon Sep 17 00:00:00 2001
From: Patrick Pelissier <patrick.pelissier@gmail.com>
Date: Sat, 10 Sep 2022 09:59:06 +0200
Subject: [PATCH] Reduce size of struct string_it_t on 64 bits by storing an
 offset to the next character instead of a pointer. An offset can be stored as
 a BYTE instead of a PTR since UT-8 can use at most 4 characters per code
 point.

---
 m-string.h           | 14 ++++++++------
 tests/test-mstring.c |  2 +-
 2 files changed, 9 insertions(+), 7 deletions(-)

diff --git a/m-string.h b/m-string.h
index 083b5f6f..c453e6d4 100644
--- a/m-string.h
+++ b/m-string.h
@@ -114,9 +114,9 @@ typedef uint32_t m_string_unicode_t;
 
 /* Iterator on a string over UTF8 encoded characters */
 typedef struct m_string_it_s {
-  m_string_unicode_t u;
+  m_string_unicode_t u;            // Decoded Unicode character for the iterator
+  int         next_ptr;            // Offset to the next character
   const char *ptr;
-  const char *next_ptr;
 } m_string_it_t[1];
 
 /* PREFIX:
@@ -1695,9 +1695,9 @@ m_string_it(m_string_it_t it, const m_string_t str)
   M_STR1NG_CONTRACT(str);
   M_ASSERT(it != NULL);
   it->ptr      = m_string_get_cstr(str);
-  it->next_ptr = it->ptr;
+  it->next_ptr = 0;
   it->u        = 0;
-}
+} 
 
 /* Set the iterator to the end of string 
    The iterator references therefore nothing.
@@ -1741,7 +1741,9 @@ m_string_end_p (m_string_it_t it)
     str++;
   } while (state != M_STR1NG_UTF8_STARTING && state != M_STR1NG_UTF8_ERROR && *str != 0);
   // Save where the current unicode value ends in the UTF8 steam
-  it->next_ptr = str;
+  // UTF-8 is maximum 4 characters
+  M_ASSERT( (str > it->ptr) && (str - it->ptr) <= 4);
+  it->next_ptr = (int) (str - it->ptr);
   // Save the decoded unicode value
   it->u = M_UNLIKELY (state == M_STR1NG_UTF8_ERROR) ? M_STRING_UNICODE_ERROR : u;
   return false;
@@ -1764,7 +1766,7 @@ m_string_next (m_string_it_t it)
   M_ASSERT (it != NULL);
   // We already decoded the UTF8 stream
   // Read the decoded value
-  it->ptr = it->next_ptr;
+  it->ptr += it->next_ptr;
 }
 
 /* Return the unicode value associated to the iterator */
diff --git a/tests/test-mstring.c b/tests/test-mstring.c
index 144e3756..ef1504e2 100644
--- a/tests/test-mstring.c
+++ b/tests/test-mstring.c
@@ -46,7 +46,7 @@ static void test_utf8_basic(void)
       assert (b == false);
       assert (it->u == i);
       assert (string_get_cref(it) == i);
-      assert (*it->next_ptr == 0);
+      assert (it->ptr[it->next_ptr] == 0);
       assert (m_str1ng_utf8_length(buf) == 1);
       assert (m_str1ng_utf8_valid_str_p(buf) == true);
       /* Higher level access */