POSIX [[:character class:]] support for standard, locale,

and utf8.  If both utf8 and locale are on, utf8 wins.
I don't fully understand why so many tables changed in
lib/unicode because of "make" -- maybe it was just overdue.

p4raw-id: //depot/cfgperl@3624

16 files changed, 821 insertions, 10 deletions

diff --git a/lib/unicode/Is/ASCII.pl b/lib/unicode/Is/ASCII.pl
new file mode 100644
index 0000000000..b7843e932f
--- /dev/null
+++ b/lib/unicode/Is/ASCII.pl
@@ -0,0 +1,3 @@
+return <<'END';
+0000	007f
+END
diff --git a/lib/unicode/Is/Alnum.pl b/lib/unicode/Is/Alnum.pl
index aa82e4ff89..ac48257e9a 100644
--- a/lib/unicode/Is/Alnum.pl
+++ b/lib/unicode/Is/Alnum.pl
@@ -1,7 +1,6 @@
 return <<'END';
 0030	0039
 0041	005a
-005f	
 0061	007a
 00aa	
 00b5	
@@ -156,8 +155,32 @@ return <<'END';
 1100	1159
 115f	11a2
 11a8	11f9
-1200	135a
-1369	137c
+1200	1206
+1208	1246
+1248	
+124a	124d
+1250	1256
+1258	
+125a	125d
+1260	1286
+1288	
+128a	128d
+1290	12ae
+12b0	
+12b2	12b5
+12b8	12be
+12c0	
+12c2	12c5
+12c8	12ce
+12d0	12d6
+12d8	12ee
+12f0	130e
+1310	
+1312	1315
+1318	131e
+1320	1346
+1348	135a
+1369	1371
 1e00	1e9b
 1ea0	1ef9
 1f00	1f15
diff --git a/lib/unicode/Is/Alpha.pl b/lib/unicode/Is/Alpha.pl
index ea6fa7f8b6..9de0521776 100644
--- a/lib/unicode/Is/Alpha.pl
+++ b/lib/unicode/Is/Alpha.pl
@@ -141,7 +141,31 @@ return <<'END';
 1100	1159
 115f	11a2
 11a8	11f9
-1200	135a
+1200	1206
+1208	1246
+1248	
+124a	124d
+1250	1256
+1258	
+125a	125d
+1260	1286
+1288	
+128a	128d
+1290	12ae
+12b0	
+12b2	12b5
+12b8	12be
+12c0	
+12c2	12c5
+12c8	12ce
+12d0	12d6
+12d8	12ee
+12f0	130e
+1310	
+1312	1315
+1318	131e
+1320	1346
+1348	135a
 1e00	1e9b
 1ea0	1ef9
 1f00	1f15
diff --git a/lib/unicode/Is/BidiL.pl b/lib/unicode/Is/BidiL.pl
index 85de325625..c17ef10a5b 100644
--- a/lib/unicode/Is/BidiL.pl
+++ b/lib/unicode/Is/BidiL.pl
@@ -186,6 +186,32 @@ return <<'END';
 1100	1159
 115f	11a2
 11a8	11f9
+1200	1206
+1208	1246
+1248	
+124a	124d
+1250	1256
+1258	
+125a	125d
+1260	1286
+1288	
+128a	128d
+1290	12ae
+12b0	
+12b2	12b5
+12b8	12be
+12c0	
+12c2	12c5
+12c8	12ce
+12d0	12d6
+12d8	12ee
+12f0	130e
+1310	
+1312	1315
+1318	131e
+1320	1346
+1348	135a
+1361	137c
 1e00	1e9b
 1ea0	1ef9
 1f00	1f15
diff --git a/lib/unicode/Is/Cntrl.pl b/lib/unicode/Is/Cntrl.pl
new file mode 100644
index 0000000000..274239f9d2
--- /dev/null
+++ b/lib/unicode/Is/Cntrl.pl
@@ -0,0 +1,12 @@
+return <<'END';
+0000	001f
+007f	009f
+200c	200f
+202a	202e
+206a	206f
+d800	db7f
+db80	dbff
+dc00	dfff
+e000	f8ff
+feff	
+END
diff --git a/lib/unicode/Is/Digit.pl b/lib/unicode/Is/Digit.pl
index 2181f150ff..a25e28f9e9 100644
--- a/lib/unicode/Is/Digit.pl
+++ b/lib/unicode/Is/Digit.pl
@@ -14,6 +14,6 @@ return <<'END';
 0e50	0e59
 0ed0	0ed9
 0f20	0f29
-1369	137c
+1369	1371
 ff10	ff19
 END
diff --git a/lib/unicode/Is/Graph.pl b/lib/unicode/Is/Graph.pl
new file mode 100644
index 0000000000..7a8c225176
--- /dev/null
+++ b/lib/unicode/Is/Graph.pl
@@ -0,0 +1,327 @@
+return <<'END';
+0021	007e
+00a0	01f5
+01fa	0217
+0250	02a8
+02b0	02de
+02e0	02e9
+0300	0345
+0360	0361
+0374	0375
+037a	
+037e	
+0384	038a
+038c	
+038e	03a1
+03a3	03ce
+03d0	03d6
+03da	
+03dc	
+03de	
+03e0	
+03e2	03f3
+0401	040c
+040e	044f
+0451	045c
+045e	0486
+0490	04c4
+04c7	04c8
+04cb	04cc
+04d0	04eb
+04ee	04f5
+04f8	04f9
+0531	0556
+0559	055f
+0561	0587
+0589	
+0591	05a1
+05a3	05b9
+05bb	05c4
+05d0	05ea
+05f0	05f4
+060c	
+061b	
+061f	
+0621	063a
+0640	0652
+0660	066d
+0670	06b7
+06ba	06be
+06c0	06ce
+06d0	06ed
+06f0	06f9
+0901	0903
+0905	0939
+093c	094d
+0950	0954
+0958	0970
+0981	0983
+0985	098c
+098f	0990
+0993	09a8
+09aa	09b0
+09b2	
+09b6	09b9
+09bc	
+09be	09c4
+09c7	09c8
+09cb	09cd
+09d7	
+09dc	09dd
+09df	09e3
+09e6	09fa
+0a02	
+0a05	0a0a
+0a0f	0a10
+0a13	0a28
+0a2a	0a30
+0a32	0a33
+0a35	0a36
+0a38	0a39
+0a3c	
+0a3e	0a42
+0a47	0a48
+0a4b	0a4d
+0a59	0a5c
+0a5e	
+0a66	0a74
+0a81	0a83
+0a85	0a8b
+0a8d	
+0a8f	0a91
+0a93	0aa8
+0aaa	0ab0
+0ab2	0ab3
+0ab5	0ab9
+0abc	0ac5
+0ac7	0ac9
+0acb	0acd
+0ad0	
+0ae0	
+0ae6	0aef
+0b01	0b03
+0b05	0b0c
+0b0f	0b10
+0b13	0b28
+0b2a	0b30
+0b32	0b33
+0b36	0b39
+0b3c	0b43
+0b47	0b48
+0b4b	0b4d
+0b56	0b57
+0b5c	0b5d
+0b5f	0b61
+0b66	0b70
+0b82	0b83
+0b85	0b8a
+0b8e	0b90
+0b92	0b95
+0b99	0b9a
+0b9c	
+0b9e	0b9f
+0ba3	0ba4
+0ba8	0baa
+0bae	0bb5
+0bb7	0bb9
+0bbe	0bc2
+0bc6	0bc8
+0bca	0bcd
+0bd7	
+0be7	0bf2
+0c01	0c03
+0c05	0c0c
+0c0e	0c10
+0c12	0c28
+0c2a	0c33
+0c35	0c39
+0c3e	0c44
+0c46	0c48
+0c4a	0c4d
+0c55	0c56
+0c60	0c61
+0c66	0c6f
+0c82	0c83
+0c85	0c8c
+0c8e	0c90
+0c92	0ca8
+0caa	0cb3
+0cb5	0cb9
+0cbe	0cc4
+0cc6	0cc8
+0cca	0ccd
+0cd5	0cd6
+0cde	
+0ce0	0ce1
+0ce6	0cef
+0d02	0d03
+0d05	0d0c
+0d0e	0d10
+0d12	0d28
+0d2a	0d39
+0d3e	0d43
+0d46	0d48
+0d4a	0d4d
+0d57	
+0d60	0d61
+0d66	0d6f
+0e01	0e3a
+0e3f	0e5b
+0e81	0e82
+0e84	
+0e87	0e88
+0e8a	
+0e8d	
+0e94	0e97
+0e99	0e9f
+0ea1	0ea3
+0ea5	
+0ea7	
+0eaa	0eab
+0ead	0eb9
+0ebb	0ebd
+0ec0	0ec4
+0ec6	
+0ec8	0ecd
+0ed0	0ed9
+0edc	0edd
+0f00	0f47
+0f49	0f69
+0f71	0f8b
+0f90	0f95
+0f97	
+0f99	0fad
+0fb1	0fb7
+0fb9	
+10a0	10c5
+10d0	10f6
+10fb	
+1100	1159
+115f	11a2
+11a8	11f9
+1200	1206
+1208	1246
+1248	
+124a	124d
+1250	1256
+1258	
+125a	125d
+1260	1286
+1288	
+128a	128d
+1290	12ae
+12b0	
+12b2	12b5
+12b8	12be
+12c0	
+12c2	12c5
+12c8	12ce
+12d0	12d6
+12d8	12ee
+12f0	130e
+1310	
+1312	1315
+1318	131e
+1320	1346
+1348	135a
+1361	137c
+1e00	1e9b
+1ea0	1ef9
+1f00	1f15
+1f18	1f1d
+1f20	1f45
+1f48	1f4d
+1f50	1f57
+1f59	
+1f5b	
+1f5d	
+1f5f	1f7d
+1f80	1fb4
+1fb6	1fc4
+1fc6	1fd3
+1fd6	1fdb
+1fdd	1fef
+1ff2	1ff4
+1ff6	1ffe
+2000	200b
+2010	2029
+2030	2046
+2070	
+2074	208e
+20a0	20ac
+20d0	20e1
+2100	2138
+2153	2182
+2190	21ea
+2200	22f1
+2300	
+2302	237a
+2400	2424
+2440	244a
+2460	24ea
+2500	2595
+25a0	25ef
+2600	2613
+261a	266f
+2701	2704
+2706	2709
+270c	2727
+2729	274b
+274d	
+274f	2752
+2756	
+2758	275e
+2761	2767
+2776	2794
+2798	27af
+27b1	27be
+3000	3037
+303f	
+3041	3094
+3099	309e
+30a1	30fe
+3105	312c
+3131	318e
+3190	319f
+3200	321c
+3220	3243
+3260	327b
+327f	32b0
+32c0	32cb
+32d0	32fe
+3300	3376
+337b	33dd
+33e0	33fe
+4e00	9fa5
+ac00	d7a3
+f900	fa2d
+fb00	fb06
+fb13	fb17
+fb1e	fb36
+fb38	fb3c
+fb3e	
+fb40	fb41
+fb43	fb44
+fb46	fbb1
+fbd3	fd3f
+fd50	fd8f
+fd92	fdc7
+fdf0	fdfb
+fe20	fe23
+fe30	fe44
+fe49	fe52
+fe54	fe66
+fe68	fe6b
+fe70	fe72
+fe74	
+fe76	fefc
+ff01	ff5e
+ff61	ffbe
+ffc2	ffc7
+ffca	ffcf
+ffd2	ffd7
+ffda	ffdc
+ffe0	ffe6
+ffe8	ffee
+fffc	fffd
+END
diff --git a/lib/unicode/Is/L.pl b/lib/unicode/Is/L.pl
index 9c8e3cf21f..06796fd1bb 100644
--- a/lib/unicode/Is/L.pl
+++ b/lib/unicode/Is/L.pl
@@ -145,7 +145,31 @@ return <<'END';
 1100	1159
 115f	11a2
 11a8	11f9
-1200	135a
+1200	1206
+1208	1246
+1248	
+124a	124d
+1250	1256
+1258	
+125a	125d
+1260	1286
+1288	
+128a	128d
+1290	12ae
+12b0	
+12b2	12b5
+12b8	12be
+12c0	
+12c2	12c5
+12c8	12ce
+12d0	12d6
+12d8	12ee
+12f0	130e
+1310	
+1312	1315
+1318	131e
+1320	1346
+1348	135a
 1e00	1e9b
 1ea0	1ef9
 1f00	1f15
diff --git a/lib/unicode/Is/Lo.pl b/lib/unicode/Is/Lo.pl
index 30f776da52..07da29ed1b 100644
--- a/lib/unicode/Is/Lo.pl
+++ b/lib/unicode/Is/Lo.pl
@@ -107,7 +107,31 @@ return <<'END';
 1100	1159
 115f	11a2
 11a8	11f9
-1200	135a
+1200	1206
+1208	1246
+1248	
+124a	124d
+1250	1256
+1258	
+125a	125d
+1260	1286
+1288	
+128a	128d
+1290	12ae
+12b0	
+12b2	12b5
+12b8	12be
+12c0	
+12c2	12c5
+12c8	12ce
+12d0	12d6
+12d8	12ee
+12f0	130e
+1310	
+1312	1315
+1318	131e
+1320	1346
+1348	135a
 2135	2138
 3041	3094
 30a1	30fa
diff --git a/lib/unicode/Is/Print.pl b/lib/unicode/Is/Print.pl
index eef2d314c0..c2e607006b 100644
--- a/lib/unicode/Is/Print.pl
+++ b/lib/unicode/Is/Print.pl
@@ -199,6 +199,32 @@ return <<'END';
 1100	1159
 115f	11a2
 11a8	11f9
+1200	1206
+1208	1246
+1248	
+124a	124d
+1250	1256
+1258	
+125a	125d
+1260	1286
+1288	
+128a	128d
+1290	12ae
+12b0	
+12b2	12b5
+12b8	12be
+12c0	
+12c2	12c5
+12c8	12ce
+12d0	12d6
+12d8	12ee
+12f0	130e
+1310	
+1312	1315
+1318	131e
+1320	1346
+1348	135a
+1361	137c
 1e00	1e9b
 1ea0	1ef9
 1f00	1f15
diff --git a/lib/unicode/Is/Punct.pl b/lib/unicode/Is/Punct.pl
new file mode 100644
index 0000000000..0d52205b0a
--- /dev/null
+++ b/lib/unicode/Is/Punct.pl
@@ -0,0 +1,70 @@
+return <<'END';
+0021	0023
+0025	002a
+002c	002f
+003a	003b
+003f	0040
+005b	005d
+005f	
+007b	
+007d	
+00a1	
+00ab	
+00ad	
+00b7	
+00bb	
+00bf	
+0374	0375
+037e	
+0387	
+055a	055f
+0589	
+05be	
+05c0	
+05c3	
+05f3	05f4
+060c	
+061b	
+061f	
+066a	066d
+06d4	
+0964	0965
+0970	
+0e2f	
+0e5a	0e5b
+0eaf	
+0f04	0f12
+0f3a	0f3d
+0f85	
+10fb	
+1361	1368
+2010	2027
+2030	2043
+2045	2046
+207d	207e
+208d	208e
+2329	232a
+3001	3003
+3006	
+3008	3011
+3014	301f
+3030	
+30fb	
+fd3e	fd3f
+fe30	fe44
+fe49	fe52
+fe54	fe61
+fe63	
+fe68	
+fe6a	fe6b
+ff01	ff03
+ff05	ff0a
+ff0c	ff0f
+ff1a	ff1b
+ff1f	ff20
+ff3b	ff3d
+ff3f	
+ff5b	
+ff5d	
+ff61	ff65
+END
diff --git a/lib/unicode/Is/Space.pl b/lib/unicode/Is/Space.pl
index 903f854416..715afc3ef3 100644
--- a/lib/unicode/Is/Space.pl
+++ b/lib/unicode/Is/Space.pl
@@ -3,7 +3,6 @@ return <<'END';
 000c	000d
 0020	
 00a0	
-1361
 2000	200b
 2028	2029
 3000	
diff --git a/lib/unicode/Is/Word.pl b/lib/unicode/Is/Word.pl
new file mode 100644
index 0000000000..6a30246b20
--- /dev/null
+++ b/lib/unicode/Is/Word.pl
@@ -0,0 +1,250 @@
+return <<'END';
+0030	0039
+0041	005a
+005f	
+0061	007a
+00aa	
+00b5	
+00ba	
+00c0	00d6
+00d8	00f6
+00f8	01c4
+01c6	01c7
+01c9	01ca
+01cc	01f1
+01f3	01f5
+01fa	0217
+0250	02a8
+0386	
+0388	038a
+038c	
+038e	03a1
+03a3	03ce
+03d0	03d6
+03da	
+03dc	
+03de	
+03e0	
+03e2	03f3
+0401	040c
+040e	044f
+0451	045c
+045e	0481
+0490	04c4
+04c7	04c8
+04cb	04cc
+04d0	04eb
+04ee	04f5
+04f8	04f9
+0531	0556
+0561	0587
+05d0	05ea
+05f0	05f2
+0621	063a
+0641	064a
+0660	0669
+0671	06b7
+06ba	06be
+06c0	06ce
+06d0	06d3
+06d5	
+06f0	06f9
+0905	0939
+093d	
+0958	0961
+0966	096f
+0985	098c
+098f	0990
+0993	09a8
+09aa	09b0
+09b2	
+09b6	09b9
+09dc	09dd
+09df	09e1
+09e6	09f1
+0a05	0a0a
+0a0f	0a10
+0a13	0a28
+0a2a	0a30
+0a32	0a33
+0a35	0a36
+0a38	0a39
+0a59	0a5c
+0a5e	
+0a66	0a6f
+0a72	0a74
+0a85	0a8b
+0a8d	
+0a8f	0a91
+0a93	0aa8
+0aaa	0ab0
+0ab2	0ab3
+0ab5	0ab9
+0abd	
+0ae0	
+0ae6	0aef
+0b05	0b0c
+0b0f	0b10
+0b13	0b28
+0b2a	0b30
+0b32	0b33
+0b36	0b39
+0b3d	
+0b5c	0b5d
+0b5f	0b61
+0b66	0b6f
+0b85	0b8a
+0b8e	0b90
+0b92	0b95
+0b99	0b9a
+0b9c	
+0b9e	0b9f
+0ba3	0ba4
+0ba8	0baa
+0bae	0bb5
+0bb7	0bb9
+0be7	0bef
+0c05	0c0c
+0c0e	0c10
+0c12	0c28
+0c2a	0c33
+0c35	0c39
+0c60	0c61
+0c66	0c6f
+0c85	0c8c
+0c8e	0c90
+0c92	0ca8
+0caa	0cb3
+0cb5	0cb9
+0cde	
+0ce0	0ce1
+0ce6	0cef
+0d05	0d0c
+0d0e	0d10
+0d12	0d28
+0d2a	0d39
+0d60	0d61
+0d66	0d6f
+0e01	0e2e
+0e30	
+0e32	0e33
+0e40	0e45
+0e50	0e59
+0e81	0e82
+0e84	
+0e87	0e88
+0e8a	
+0e8d	
+0e94	0e97
+0e99	0e9f
+0ea1	0ea3
+0ea5	
+0ea7	
+0eaa	0eab
+0ead	0eae
+0eb0	
+0eb2	0eb3
+0ebd	
+0ec0	0ec4
+0ed0	0ed9
+0edc	0edd
+0f20	0f29
+0f40	0f47
+0f49	0f69
+10a0	10c5
+10d0	10f6
+1100	1159
+115f	11a2
+11a8	11f9
+1200	1206
+1208	1246
+1248	
+124a	124d
+1250	1256
+1258	
+125a	125d
+1260	1286
+1288	
+128a	128d
+1290	12ae
+12b0	
+12b2	12b5
+12b8	12be
+12c0	
+12c2	12c5
+12c8	12ce
+12d0	12d6
+12d8	12ee
+12f0	130e
+1310	
+1312	1315
+1318	131e
+1320	1346
+1348	135a
+1369	1371
+1e00	1e9b
+1ea0	1ef9
+1f00	1f15
+1f18	1f1d
+1f20	1f45
+1f48	1f4d
+1f50	1f57
+1f59	
+1f5b	
+1f5d	
+1f5f	1f7d
+1f80	1fb4
+1fb6	1fbc
+1fbe	
+1fc2	1fc4
+1fc6	1fcc
+1fd0	1fd3
+1fd6	1fdb
+1fe0	1fec
+1ff2	1ff4
+1ff6	1ffc
+207f	
+2102	
+2107	
+210a	2113
+2115	
+2118	211d
+2124	
+2126	
+2128	
+212a	2131
+2133	2138
+3041	3094
+30a1	30fa
+3105	312c
+3131	318e
+4e00	9fa5
+ac00	d7a3
+f900	fa2d
+fb00	fb06
+fb13	fb17
+fb1f	fb28
+fb2a	fb36
+fb38	fb3c
+fb3e	
+fb40	fb41
+fb43	fb44
+fb46	fbb1
+fbd3	fd3d
+fd50	fd8f
+fd92	fdc7
+fdf0	fdfb
+fe70	fe72
+fe74	
+fe76	fefc
+ff10	ff19
+ff21	ff3a
+ff41	ff5a
+ff66	ff6f
+ff71	ff9d
+ffa0	ffbe
+ffc2	ffc7
+ffca	ffcf
+ffd2	ffd7
+ffda	ffdc
+END
diff --git a/lib/unicode/Is/XDigit.pl b/lib/unicode/Is/XDigit.pl
new file mode 100644
index 0000000000..f0b7044eb6
--- /dev/null
+++ b/lib/unicode/Is/XDigit.pl
@@ -0,0 +1,5 @@
+return <<'END';
+0030	0039
+0041	0046
+0061	0066
+END
diff --git a/lib/unicode/Is/Z.pl b/lib/unicode/Is/Z.pl
index af595da354..9e83d9427f 100644
--- a/lib/unicode/Is/Z.pl
+++ b/lib/unicode/Is/Z.pl
@@ -1,7 +1,6 @@
 return <<'END';
 0020	
 00a0	
-1361
 2000	200b
 2028	2029
 3000	
diff --git a/lib/unicode/Is/Zs.pl b/lib/unicode/Is/Zs.pl
index 403728c0db..87d4455d02 100644
--- a/lib/unicode/Is/Zs.pl
+++ b/lib/unicode/Is/Zs.pl
@@ -1,7 +1,6 @@
 return <<'END';
 0020	
 00a0	
-1361
 2000	200b
 3000	
 END