fastly · silentbicycle · Oct 30, 2023 · Sep 18, 2023 · Oct 30, 2023
diff --git a/share/ucd/CaseFolding.txt b/share/ucd/CaseFolding.txt
@@ -1,6 +1,6 @@
-# CaseFolding-15.0.0.txt
-# Date: 2022-02-02, 23:35:35 GMT
-# © 2022 Unicode®, Inc.
+# CaseFolding-15.1.0.txt
+# Date: 2023-05-12, 21:53:10 GMT
+# © 2023 Unicode®, Inc.
 # Unicode and the Unicode Logo are registered trademarks of Unicode, Inc. in the U.S. and other countries.
 # For terms of use, see https://www.unicode.org/terms_of_use.html
 #
@@ -929,6 +929,7 @@
 1FCC; S; 1FC3; # GREEK CAPITAL LETTER ETA WITH PROSGEGRAMMENI
 1FD2; F; 03B9 0308 0300; # GREEK SMALL LETTER IOTA WITH DIALYTIKA AND VARIA
 1FD3; F; 03B9 0308 0301; # GREEK SMALL LETTER IOTA WITH DIALYTIKA AND OXIA
+1FD3; S; 0390; # GREEK SMALL LETTER IOTA WITH DIALYTIKA AND OXIA
 1FD6; F; 03B9 0342; # GREEK SMALL LETTER IOTA WITH PERISPOMENI
 1FD7; F; 03B9 0308 0342; # GREEK SMALL LETTER IOTA WITH DIALYTIKA AND PERISPOMENI
 1FD8; C; 1FD0; # GREEK CAPITAL LETTER IOTA WITH VRACHY
@@ -937,6 +938,7 @@
 1FDB; C; 1F77; # GREEK CAPITAL LETTER IOTA WITH OXIA
 1FE2; F; 03C5 0308 0300; # GREEK SMALL LETTER UPSILON WITH DIALYTIKA AND VARIA
 1FE3; F; 03C5 0308 0301; # GREEK SMALL LETTER UPSILON WITH DIALYTIKA AND OXIA
+1FE3; S; 03B0; # GREEK SMALL LETTER UPSILON WITH DIALYTIKA AND OXIA
 1FE4; F; 03C1 0313; # GREEK SMALL LETTER RHO WITH PSILI
 1FE6; F; 03C5 0342; # GREEK SMALL LETTER UPSILON WITH PERISPOMENI
 1FE7; F; 03C5 0308 0342; # GREEK SMALL LETTER UPSILON WITH DIALYTIKA AND PERISPOMENI
@@ -1328,6 +1330,7 @@ FB02; F; 0066 006C; # LATIN SMALL LIGATURE FL
 FB03; F; 0066 0066 0069; # LATIN SMALL LIGATURE FFI
 FB04; F; 0066 0066 006C; # LATIN SMALL LIGATURE FFL
 FB05; F; 0073 0074; # LATIN SMALL LIGATURE LONG S T
+FB05; S; FB06; # LATIN SMALL LIGATURE LONG S T
 FB06; F; 0073 0074; # LATIN SMALL LIGATURE ST
 FB13; F; 0574 0576; # ARMENIAN SMALL LIGATURE MEN NOW
 FB14; F; 0574 0565; # ARMENIAN SMALL LIGATURE MEN ECH

diff --git a/share/ucd/Makefile b/share/ucd/Makefile
@@ -1,4 +1,4 @@
-UCD_URL ?= https://www.unicode.org/Public/15.0.0/ucd/
+UCD_URL ?= https://www.unicode.org/Public/15.1.0/ucd/
 
 WGET ?= wget
 

diff --git a/share/ucd/Scripts.txt b/share/ucd/Scripts.txt
@@ -1,6 +1,6 @@
-# Scripts-15.0.0.txt
-# Date: 2022-04-26, 23:15:02 GMT
-# © 2022 Unicode®, Inc.
+# Scripts-15.1.0.txt
+# Date: 2023-07-28, 16:01:07 GMT
+# © 2023 Unicode®, Inc.
 # Unicode and the Unicode Logo are registered trademarks of Unicode, Inc. in the U.S. and other countries.
 # For terms of use, see https://www.unicode.org/terms_of_use.html
 #
@@ -357,7 +357,7 @@
 2E5B          ; Common # Ps       BOTTOM HALF LEFT PARENTHESIS
 2E5C          ; Common # Pe       BOTTOM HALF RIGHT PARENTHESIS
 2E5D          ; Common # Pd       OBLIQUE HYPHEN
-2FF0..2FFB    ; Common # So  [12] IDEOGRAPHIC DESCRIPTION CHARACTER LEFT TO RIGHT..IDEOGRAPHIC DESCRIPTION CHARACTER OVERLAID
+2FF0..2FFF    ; Common # So  [16] IDEOGRAPHIC DESCRIPTION CHARACTER LEFT TO RIGHT..IDEOGRAPHIC DESCRIPTION CHARACTER ROTATION
 3000          ; Common # Zs       IDEOGRAPHIC SPACE
 3001..3003    ; Common # Po   [3] IDEOGRAPHIC COMMA..DITTO MARK
 3004          ; Common # So       JAPANESE INDUSTRIAL STANDARD SYMBOL
@@ -399,6 +399,7 @@
 3192..3195    ; Common # No   [4] IDEOGRAPHIC ANNOTATION ONE MARK..IDEOGRAPHIC ANNOTATION FOUR MARK
 3196..319F    ; Common # So  [10] IDEOGRAPHIC ANNOTATION TOP MARK..IDEOGRAPHIC ANNOTATION MAN MARK
 31C0..31E3    ; Common # So  [36] CJK STROKE T..CJK STROKE Q
+31EF          ; Common # So       IDEOGRAPHIC DESCRIPTION CHARACTER SUBTRACTION
 3220..3229    ; Common # No  [10] PARENTHESIZED IDEOGRAPH ONE..PARENTHESIZED IDEOGRAPH TEN
 322A..3247    ; Common # So  [30] PARENTHESIZED IDEOGRAPH MOON..CIRCLED IDEOGRAPH KOTO
 3248..324F    ; Common # No   [8] CIRCLED NUMBER TEN ON BLACK SQUARE..CIRCLED NUMBER EIGHTY ON BLACK SQUARE
@@ -629,7 +630,7 @@ FFFC..FFFD    ; Common # So   [2] OBJECT REPLACEMENT CHARACTER..REPLACEMENT CHAR
 E0001         ; Common # Cf       LANGUAGE TAG
 E0020..E007F  ; Common # Cf  [96] TAG SPACE..CANCEL TAG
 
-# Total code points: 8301
+# Total code points: 8306
 
 # ================================================
 
@@ -1593,11 +1594,12 @@ FA70..FAD9    ; Han # Lo [106] CJK COMPATIBILITY IDEOGRAPH-FA70..CJK COMPATIBILI
 2B740..2B81D  ; Han # Lo [222] CJK UNIFIED IDEOGRAPH-2B740..CJK UNIFIED IDEOGRAPH-2B81D
 2B820..2CEA1  ; Han # Lo [5762] CJK UNIFIED IDEOGRAPH-2B820..CJK UNIFIED IDEOGRAPH-2CEA1
 2CEB0..2EBE0  ; Han # Lo [7473] CJK UNIFIED IDEOGRAPH-2CEB0..CJK UNIFIED IDEOGRAPH-2EBE0
+2EBF0..2EE5D  ; Han # Lo [622] CJK UNIFIED IDEOGRAPH-2EBF0..CJK UNIFIED IDEOGRAPH-2EE5D
 2F800..2FA1D  ; Han # Lo [542] CJK COMPATIBILITY IDEOGRAPH-2F800..CJK COMPATIBILITY IDEOGRAPH-2FA1D
 30000..3134A  ; Han # Lo [4939] CJK UNIFIED IDEOGRAPH-30000..CJK UNIFIED IDEOGRAPH-3134A
 31350..323AF  ; Han # Lo [4192] CJK UNIFIED IDEOGRAPH-31350..CJK UNIFIED IDEOGRAPH-323AF
 
-# Total code points: 98408
+# Total code points: 99030
 
 # ================================================
 

diff --git a/share/ucd/UnicodeData.txt b/share/ucd/UnicodeData.txt
@@ -11231,6 +11231,10 @@
 2FF9;IDEOGRAPHIC DESCRIPTION CHARACTER SURROUND FROM UPPER RIGHT;So;0;ON;;;;;N;;;;;
 2FFA;IDEOGRAPHIC DESCRIPTION CHARACTER SURROUND FROM LOWER LEFT;So;0;ON;;;;;N;;;;;
 2FFB;IDEOGRAPHIC DESCRIPTION CHARACTER OVERLAID;So;0;ON;;;;;N;;;;;
+2FFC;IDEOGRAPHIC DESCRIPTION CHARACTER SURROUND FROM RIGHT;So;0;ON;;;;;N;;;;;
+2FFD;IDEOGRAPHIC DESCRIPTION CHARACTER SURROUND FROM LOWER RIGHT;So;0;ON;;;;;N;;;;;
+2FFE;IDEOGRAPHIC DESCRIPTION CHARACTER HORIZONTAL REFLECTION;So;0;ON;;;;;N;;;;;
+2FFF;IDEOGRAPHIC DESCRIPTION CHARACTER ROTATION;So;0;ON;;;;;N;;;;;
 3000;IDEOGRAPHIC SPACE;Zs;0;WS;<wide> 0020;;;;N;;;;;
 3001;IDEOGRAPHIC COMMA;Po;0;ON;;;;;N;;;;;
 3002;IDEOGRAPHIC FULL STOP;Po;0;ON;;;;;N;IDEOGRAPHIC PERIOD;;;;
@@ -11705,6 +11709,7 @@
 31E1;CJK STROKE HZZZG;So;0;ON;;;;;N;;;;;
 31E2;CJK STROKE PG;So;0;ON;;;;;N;;;;;
 31E3;CJK STROKE Q;So;0;ON;;;;;N;;;;;
+31EF;IDEOGRAPHIC DESCRIPTION CHARACTER SUBTRACTION;So;0;ON;;;;;N;;;;;
 31F0;KATAKANA LETTER SMALL KU;Lo;0;L;;;;;N;;;;;
 31F1;KATAKANA LETTER SMALL SI;Lo;0;L;;;;;N;;;;;
 31F2;KATAKANA LETTER SMALL SU;Lo;0;L;;;;;N;;;;;
@@ -34035,6 +34040,8 @@ FFFD;REPLACEMENT CHARACTER;So;0;ON;;;;;N;;;;;
 2CEA1;<CJK Ideograph Extension E, Last>;Lo;0;L;;;;;N;;;;;
 2CEB0;<CJK Ideograph Extension F, First>;Lo;0;L;;;;;N;;;;;
 2EBE0;<CJK Ideograph Extension F, Last>;Lo;0;L;;;;;N;;;;;
+2EBF0;<CJK Ideograph Extension I, First>;Lo;0;L;;;;;N;;;;;
+2EE5D;<CJK Ideograph Extension I, Last>;Lo;0;L;;;;;N;;;;;
 2F800;CJK COMPATIBILITY IDEOGRAPH-2F800;Lo;0;L;4E3D;;;;N;;;;;
 2F801;CJK COMPATIBILITY IDEOGRAPH-2F801;Lo;0;L;4E38;;;;N;;;;;
 2F802;CJK COMPATIBILITY IDEOGRAPH-2F802;Lo;0;L;4E41;;;;N;;;;;

diff --git a/src/libre/ast.c b/src/libre/ast.c
@@ -997,13 +997,31 @@ ast_expr_is_literal(const struct ast_expr *e,
 		*anchor_end = 1;
 	}
 
+done:
+
 	for (i = 0; i < count; i++) {
 		if (nodes[i] == NULL || nodes[i]->type != AST_EXPR_LITERAL) {
 			return 0;
 		}
-	}
 
-done:
+		/* mask out AST flags that we don't care about; these may be present or not */
+		int ast_flags = nodes[i]->flags
+			& ~(AST_FLAG_FIRST | AST_FLAG_LAST | AST_FLAG_ALWAYS_CONSUMES | AST_FLAG_CAN_CONSUME);
+
+		/* we reject anything else */
+		if (ast_flags != 0) {
+			return 0;
+		}
+
+		/* mask out re flags that we don't handle here yet (the caller is responsible for these) */
+		int re_flags = nodes[i]->re_flags
+			& ~(RE_END_NL);
+
+		/* and we don't permit any other re_flags */
+		if (re_flags != 0) {
+			return 0;
+		}
+	}
 
 	*n = count + is_end_nl;
 

diff --git a/src/libre/class/utf8_Common.c b/src/libre/class/utf8_Common.c
diff --git a/src/libre/class/utf8_Han.c b/src/libre/class/utf8_Han.c
diff --git a/src/libre/class/utf8_L.c b/src/libre/class/utf8_L.c
diff --git a/src/libre/class/utf8_Lo.c b/src/libre/class/utf8_Lo.c
diff --git a/src/libre/class/utf8_S.c b/src/libre/class/utf8_S.c
diff --git a/src/libre/class/utf8_So.c b/src/libre/class/utf8_So.c
diff --git a/src/libre/class/utf8_assigned.c b/src/libre/class/utf8_assigned.c
diff --git a/tests/re_literal/in13.re b/tests/re_literal/in13.re
@@ -0,0 +1 @@
+^(?i)a$
diff --git a/tests/re_literal/in14.re b/tests/re_literal/in14.re
@@ -0,0 +1 @@
+(?:x(?i)a)
diff --git a/tests/re_literal/in15.re b/tests/re_literal/in15.re
@@ -0,0 +1 @@
+(?i)
diff --git a/tests/re_literal/in16.re b/tests/re_literal/in16.re
@@ -0,0 +1 @@
+(?i)xy
diff --git a/tests/re_literal/in17.re b/tests/re_literal/in17.re
@@ -0,0 +1 @@
+a(?i)xy
diff --git a/tests/re_literal/in18.re b/tests/re_literal/in18.re
@@ -0,0 +1 @@
+a(?:x(?i)xy)
diff --git a/tests/re_literal/in19.re b/tests/re_literal/in19.re
@@ -0,0 +1 @@
+a(?:x(?-i)xy)
diff --git a/tests/re_literal/in20.re b/tests/re_literal/in20.re
@@ -0,0 +1 @@
+(?i)(?:(?-i)a)(?:x(?-i)xy)
diff --git a/tests/re_literal/in21.re b/tests/re_literal/in21.re
@@ -0,0 +1 @@
+(?i)(?:(?-i)a)(?:(?-i)x(?-i)xy)
diff --git a/tests/re_literal/in22.re b/tests/re_literal/in22.re
@@ -0,0 +1 @@
+(?s)a
diff --git a/tests/re_literal/in23.re b/tests/re_literal/in23.re
@@ -0,0 +1 @@
+(?x)a
diff --git a/tests/re_literal/in24.re b/tests/re_literal/in24.re
@@ -0,0 +1 @@
+(?n)a
diff --git a/tests/re_literal/out13.txt b/tests/re_literal/out13.txt
@@ -0,0 +1 @@
+non-literal
diff --git a/tests/re_literal/out14.txt b/tests/re_literal/out14.txt
@@ -0,0 +1 @@
+non-literal
diff --git a/tests/re_literal/out15.txt b/tests/re_literal/out15.txt
@@ -0,0 +1,2 @@
+anchors: 
+literal: 
diff --git a/tests/re_literal/out16.txt b/tests/re_literal/out16.txt
@@ -0,0 +1 @@
+non-literal
diff --git a/tests/re_literal/out17.txt b/tests/re_literal/out17.txt
@@ -0,0 +1 @@
+non-literal
diff --git a/tests/re_literal/out18.txt b/tests/re_literal/out18.txt
@@ -0,0 +1 @@
+non-literal
diff --git a/tests/re_literal/out19.txt b/tests/re_literal/out19.txt
@@ -0,0 +1,2 @@
+anchors: 
+literal: axxy
diff --git a/tests/re_literal/out20.txt b/tests/re_literal/out20.txt
@@ -0,0 +1 @@
+non-literal
diff --git a/tests/re_literal/out21.txt b/tests/re_literal/out21.txt
@@ -0,0 +1,2 @@
+anchors: 
+literal: axxy
diff --git a/tests/re_literal/out22.txt b/tests/re_literal/out22.txt
@@ -0,0 +1 @@
+non-literal
diff --git a/tests/re_literal/out23.txt b/tests/re_literal/out23.txt
@@ -0,0 +1 @@
+non-literal
diff --git a/tests/re_literal/out24.txt b/tests/re_literal/out24.txt
@@ -0,0 +1,2 @@
+anchors: 
+literal: a