patch 9.1.1258: regexp: max \U and \%U value is limited by INT_MAX

Problem:  regexp: max \U and \%U value is limited by INT_MAX but gives a
          confusing error message (related: v8.1.0985).
Solution: give a better error message when the value reaches INT_MAX

When searching Vim allows to get up to 8 hex characters using the /\V
and /\%V regex atoms.  However, when using "/\UFFFFFFFF" the code point is
already above what an integer variable can hold, which is 2,147,483,647.

Since patch v8.1.0985, Vim already limited the max codepoint to INT_MAX
(otherwise it caused a crash in the nfa regex engine), but instead of
error'ing out it silently fell back to parse the number as a backslash
value and not as a codepoint value and as such this "/[\UFFFFFFFF]" will
happily find a "\" or an literal "F".  And this "/[\d127-\UFFFFFFFF]"
will error out as "reverse range in character class).

Interestingly, the max Unicode codepoint value is U+10FFFF which still
fits into an ordinary integer value,  which means, that we don't even
need to parse 8 hex characters, but 6 should have been enough.

However, let's not limit Vim to search for only max 6 hex characters
(which would be a backward incompatible change), but instead allow all 8
characters and only if the codepoint reaches INT_MAX, give a more
precise error message (about what the max unicode codepoint value is).
This allows to search for "[\U7FFFFFFE]" (will likely return "E486
Pattern not found") and "[/\U7FFFFFF]" now errors "E1517: Value too
large, max Unicode codepoint is U+10FFFF".

While this change is straight forward on architectures where long is 8
bytes, this is not so simple on Windows or 32bit architectures where long
is 4 bytes (and therefore the test fails there).  To account for that,
let's make use of the vimlong_T number type and make a few corresponding
changes in the regex engine code and cast the value to the expected data
type. This however may not work correctly on systems that doesn't have
the long long datatype (e.g. OpenVMS) and probably the test will fail
there.

fixes: #16949
closes: #16994

Signed-off-by: Christian Brabandt <cb@256bit.org>

runtime/doc/pattern.txt

@@ -1,4 +1,4 @@
-*pattern.txt*   For Vim version 9.1.  Last change: 2025 Mar 21
+*pattern.txt*   For Vim version 9.1.  Last change: 2025 Mar 28
 
 
 		  VIM REFERENCE MANUAL    by Bram Moolenaar
@@ -1222,7 +1222,8 @@ x	A single character, with no special meaning, matches itself
 		\o40	octal number of character up to 0o377
 		\x20	hexadecimal number of character up to 0xff
 		\u20AC	hex. number of multibyte character up to 0xffff
-		\U1234	hex. number of multibyte character up to 0xffffffff
+		\U1234	hex. number of multibyte character up to 8 characters
+			0xffffffff |E1541|
 	  NOTE: The other backslash codes mentioned above do not work inside
 	  []!
 	- Matching with a collection can be slow, because each character in
@@ -1263,7 +1264,8 @@ x	A single character, with no special meaning, matches itself
 \%u20AC	Matches the character specified with up to four hexadecimal
 	characters.
 \%U1234abcd	Matches the character specified with up to eight hexadecimal
-	characters, up to 0x7fffffff
+	characters, up to 0x7fffffff (the maximum allowed value is INT_MAX
+	|E1541|, but the maximum valid Unicode codepoint is U+10FFFF).
 
 ==============================================================================
 7. Ignoring case in a pattern					*/ignorecase*

runtime/doc/tags

@@ -4621,6 +4621,7 @@ E1538	eval.txt	/*E1538*
 E1539	vim9.txt	/*E1539*
 E154	helphelp.txt	/*E154*
 E1540	eval.txt	/*E1540*
+E1541	vi_diff.txt	/*E1541*
 E155	sign.txt	/*E155*
 E156	sign.txt	/*E156*
 E157	sign.txt	/*E157*

runtime/doc/vi_diff.txt

@@ -1,4 +1,4 @@
-*vi_diff.txt*   For Vim version 9.1.  Last change: 2024 Nov 10
+*vi_diff.txt*   For Vim version 9.1.  Last change: 2025 Mar 28
 
 
 		  VIM REFERENCE MANUAL    by Bram Moolenaar
@@ -91,8 +91,11 @@ Maximum display width	   Unix and Win32: 1024 characters, otherwise 255
 Maximum lhs of a mapping   50 characters.
 Number of different highlighting types: over 30000
 Range of a Number variable:  -2147483648 to 2147483647 (might be more on 64
-			   bit systems)
+			   bit systems)  See also: |v:numbermax|,
+			   |v:numbermin| and |v:numbersize|
 Maximum length of a line in a tags file: 512 bytes.
+							*E1541*
+Maximum value for |/\U| and |/\%U|: 2147483647 (for 32bit integer).
 
 Information for undo and text in registers is kept in memory, thus when making
 (big) changes the amount of (virtual) memory available limits the number of

src/errors.h

@@ -3716,3 +3716,5 @@ EXTERN char e_variadic_tuple_must_end_with_list_type_str[]
 EXTERN char e_cannot_use_variadic_tuple_in_concatenation[]
 	INIT(= N_("E1540: Cannot use a variadic tuple in concatenation"));
 #endif
+EXTERN char e_unicode_val_too_large[]
+	INIT(= N_("E1541: Value too large, max Unicode codepoint is U+10FFFF"));

src/regexp.c

@@ -427,9 +427,9 @@ static void	skipchr_keepstart(void);
 static int	peekchr(void);
 static void	skipchr(void);
 static void	ungetchr(void);
-static long	gethexchrs(int maxinputlen);
+static vimlong_T	gethexchrs(int maxinputlen);
 static long	getoctchrs(void);
-static long	getdecchrs(void);
+static vimlong_T	getdecchrs(void);
 static int	coll_get_char(void);
 static int	prog_magic_wrong(void);
 static int	cstrncmp(char_u *s1, char_u *s2, int *n);
@@ -979,7 +979,7 @@ ungetchr(void)
  * The parameter controls the maximum number of input characters. This will be
  * 2 when reading a \%x20 sequence and 4 when reading a \%u20AC sequence.
  */
-    static long
+    static vimlong_T
 gethexchrs(int maxinputlen)
 {
     long_u	nr = 0;
@@ -998,14 +998,14 @@ gethexchrs(int maxinputlen)
 
     if (i == 0)
 	return -1;
-    return (long)nr;
+    return nr;
 }
 
 /*
  * Get and return the value of the decimal string immediately after the
  * current position. Return -1 for invalid.  Consumes all digits.
  */
-    static long
+    static vimlong_T
 getdecchrs(void)
 {
     long_u	nr = 0;
@@ -1025,7 +1025,7 @@ getdecchrs(void)
 
     if (i == 0)
 	return -1;
-    return (long)nr;
+    return nr;
 }
 
 /*

src/regexp_bt.c

@@ -1589,7 +1589,7 @@ regatom(int *flagp)
 		case 'u':   // %uabcd hex 4
 		case 'U':   // %U1234abcd hex 8
 			  {
-			      long i;
+			      vimlong_T i;
 
 			      switch (c)
 			      {
@@ -1612,7 +1612,7 @@ regatom(int *flagp)
 			      if (i == 0)
 				  regc(0x0a);
 			      else
-				  regmbc(i);
+				  regmbc((int)i);
 			      regc(NUL);
 			      *flagp |= HASWIDTH;
 			      break;
@@ -1831,6 +1831,10 @@ collection:
 				|| *regparse == 'U')
 			{
 			    startc = coll_get_char();
+			    // max UTF-8 Codepoint is U+10FFFF,
+			    // but allow values until INT_MAX
+			    if (startc == INT_MAX)
+				EMSG_RET_NULL(_(e_unicode_val_too_large));
 			    if (startc == 0)
 				regc(0x0a);
 			    else
@@ -2131,7 +2135,7 @@ regpiece(int *flagp)
 		int	lop = END;
 		long	nr;
 
-		nr = getdecchrs();
+		nr = (long)getdecchrs();
 		switch (no_Magic(getchr()))
 		{
 		    case '=': lop = MATCH; break;		  // \@=
@@ -2610,7 +2614,7 @@ vim_regcomp_had_eol(void)
     static int
 coll_get_char(void)
 {
-    long	nr = -1;
+    vimlong_T	nr = -1;
 
     switch (*regparse++)
     {
@@ -2620,13 +2624,15 @@ coll_get_char(void)
 	case 'u': nr = gethexchrs(4); break;
 	case 'U': nr = gethexchrs(8); break;
     }
-    if (nr < 0 || nr > INT_MAX)
+    if (nr < 0)
     {
 	// If getting the number fails be backwards compatible: the character
 	// is a backslash.
 	--regparse;
 	nr = '\\';
     }
+    if (nr > INT_MAX)
+	nr = INT_MAX;
     return nr;
 }
 

src/regexp_nfa.c

@@ -1560,7 +1560,7 @@ nfa_regatom(void)
 		case 'u':   // %uabcd hex 4
 		case 'U':   // %U1234abcd hex 8
 		    {
-			long nr;
+			vimlong_T nr;
 
 			switch (c)
 			{
@@ -1577,7 +1577,7 @@ nfa_regatom(void)
 						       reg_magic == MAGIC_ALL);
 			// A NUL is stored in the text as NL
 			// TODO: what if a composing character follows?
-			EMIT(nr == 0 ? 0x0a : nr);
+			EMIT(nr == 0 ? 0x0a : (long)nr);
 		    }
 		    break;
 
@@ -1953,6 +1953,10 @@ collection:
 			    {
 				// TODO(RE) This needs more testing
 				startc = coll_get_char();
+				// max UTF-8 Codepoint is U+10FFFF,
+				// but allow values until INT_MAX
+				if (startc == INT_MAX)
+				    EMSG_RET_FAIL(_(e_unicode_val_too_large));
 				got_coll_char = TRUE;
 				MB_PTR_BACK(old_regparse, regparse);
 			    }
@@ -2218,7 +2222,7 @@ nfa_regpiece(void)
 	    break;
 
 	case Magic('@'):
-	    c2 = getdecchrs();
+	    c2 = (long)getdecchrs();
 	    op = no_Magic(getchr());
 	    i = 0;
 	    switch(op)

src/testdir/test_search.vim

@@ -1541,17 +1541,46 @@ func Test_large_hex_chars2()
   try
     /[\Ufffffc1f]
   catch
-    call assert_match('E486:', v:exception)
+    call assert_match('E1541:', v:exception)
   endtry
   try
     set re=1
     /[\Ufffffc1f]
   catch
-    call assert_match('E486:', v:exception)
+    call assert_match('E1541:', v:exception)
   endtry
   set re&
 endfunc
 
+func Test_large_hex_chars3()
+  " Validate max number of Unicode char
+  try
+    /[\UFFFFFFFF]
+  catch
+    call assert_match('E1541:', v:exception)
+  endtry
+  try
+    /[\UFFFFFFF]
+  catch
+    call assert_match('E486:', v:exception)
+  endtry
+  try
+    /\%#=2[\d32-\UFFFFFFFF]
+  catch
+    call assert_match('E1541:', v:exception)
+  endtry
+  try
+    /\%#=1[\UFFFFFFFF]
+  catch
+    call assert_match('E1541:', v:exception)
+  endtry
+  try
+    /\%#=1[\d32-\UFFFFFFFF]
+  catch
+    call assert_match('E945:', v:exception)
+  endtry
+endfunc
+
 func Test_one_error_msg()
   " This was also giving an internal error
   call assert_fails('call search(" \\((\\v[[=P=]]){185}+             ")', 'E871:')

src/version.c

@@ -704,6 +704,8 @@ static char *(features[]) =
 
 static int included_patches[] =
 {   /* Add new patch number below this line */
+/**/
+    1258,
 /**/
     1257,
 /**/