Fossil SCM

Further fine-tuning of the check for valid UTF8 characters in filenames.

drh 2013-01-23 13:15 trunk merge
Commit 4d456c9fd17bd289fd38b9b09d4e19a2c1af3183
Parent 3104348ec53e670…
2 files changed +52 -28 +52 -28
~ src/file.c ~ src/file.c
M src/file.c
+52 -28
--- src/file.c
+++ src/file.c
@@ -490,44 +490,68 @@
490490
 ** * Does not contain two or more "/" characters in a row.
491491
 ** * Contains at least one character
492492
 **
493493
 ** Invalid UTF8 characters result in a false return if bStrictUtf8 is
494494
 ** true. If bStrictUtf8 is false, invalid UTF8 characters are silently
495
-** ignored.
495
+** ignored. See http://en.wikipedia.org/wiki/UTF-8#Invalid_byte_sequences
496
+** and http://en.wikipedia.org/wiki/Unicode (for the noncharacters)
497
+**
498
+** The bStrictUtf8 flag is true for new inputs, but is false when parsing
499
+** legacy manifests, for backwards compatibility.
496500
 */
497501
 int file_is_simple_pathname(const char *z, int bStrictUtf8){
498502
 int i;
499
- char c = z[0];
503
+ unsigned char c = (unsigned char) z[0];
500504
 char maskNonAscii = bStrictUtf8 ? 0x80 : 0x00;
501505
 if( c=='/' || c==0 ) return 0;
502506
 if( c=='.' ){
503507
 if( z[1]=='/' || z[1]==0 ) return 0;
504508
 if( z[1]=='.' && (z[2]=='/' || z[2]==0) ) return 0;
505509
 }
506
- for(i=0; (c=z[i])!=0; i++){
507
- if( c & maskNonAscii ){
508
- if( (c & 0xf0) == 0xf0 ) {
509
- /* Unicode characters > U+FFFF are not supported.
510
- * Windows XP and earlier cannot handle them.
511
- */
512
- return 0;
513
- }
514
- if( (c & 0xf0) == 0xe0 ) {
515
- /* This is a 3-byte UTF-8 character */
516
- if ( (c & 0xfe) == 0xee ){
517
- /* Range U+E000 - U+FFFF (Starting with 0xee or 0xef in UTF-8 ) */
518
- if ( !(c & 1) || ((z[i+1] & 0xff) < 0xa4) ){
519
- /* Unicode character in the range U+E000 - U+F8FF are for
520
- * private use, they shouldn't occur in filenames. */
521
- return 0;
522
- }
523
- }else if( ((c & 0xff) == 0xed) && ((z[i+1] & 0xe0) == 0xa0) ){
524
- /* Unicode character in the range U+D800 - U+DFFF are for
525
- * surrogate pairs, they shouldn't occur in filenames. */
526
- return 0;
527
- }
528
- }
510
+ for(i=0; (c=(unsigned char)z[i])!=0; i++){
511
+ if( c & maskNonAscii ){
512
+ if( c<0xc2 ){
513
+ /* Invalid 1-byte UTF-8 sequence, or 2-byte overlong form. */
514
+ return 0;
515
+ }else if( (c&0xe0)==0xe0 ){
516
+ /* 3-byte or more */
517
+ int unicode;
518
+ if( c&0x10 ){
519
+ /* Unicode characters > U+FFFF are not supported.
520
+ * Windows XP and earlier cannot handle them.
521
+ */
522
+ return 0;
523
+ }
524
+ /* This is a 3-byte UTF-8 character */
525
+ unicode = ((c&0x0f)<<12) + ((z[i+1]&0x3f)<<6) + (z[i+2]&0x3f);
526
+ if( unicode <= 0x07ff ){
527
+ /* overlong form */
528
+ return 0;
529
+ }else if( unicode>=0xe000 ){
530
+ /* U+E000..U+FFFF */
531
+ if( (unicode<=0xf8ff) || (unicode>=0xfffe) ){
532
+ /* U+E000..U+F8FF are for private use.
533
+ * U+FFFE..U+FFFF are noncharacters. */
534
+ return 0;
535
+ } else if( (unicode>=0xfdd0) && (unicode<=0xfdef) ){
536
+ /* U+FDD0..U+FDEF are noncharacters. */
537
+ return 0;
538
+ }
539
+ }else if( (unicode>=0xD800) && (unicode<=0xDFFF) ){
540
+ /* U+D800..U+DFFF are for surrogate pairs. */
541
+ return 0;
542
+ }
543
+ }
544
+ do{
545
+ if( (z[i+1]&0xc0)!=0x80 ){
546
+ /* Invalid continuation byte (multi-byte UTF-8) */
547
+ return 0;
548
+ }
549
+ /* The hi-bits of c are used to keep track of the number of expected
550
+ * continuation-bytes, so we don't need a separate counter. */
551
+ c<<=1; ++i;
552
+ }while( c>=0xc0 );
529553
 }else if( c=='\\' ){
530554
 return 0;
531555
 }
532556
 if( c=='/' ){
533557
 if( z[i+1]=='/' ) return 0;
@@ -578,11 +602,11 @@
578602
 if( z[i]=='\\' ) z[i] = '/';
579603
 }
580604
 #endif
581605
582606
 /* Removing trailing "/" characters */
583
- if ( !slash ){
607
+ if( !slash ){
584608
 while( n>1 && z[n-1]=='/' ){ n--; }
585609
 }
586610
587611
 /* Remove duplicate '/' characters. Except, two // at the beginning
588612
 ** of a pathname is allowed since this is important on windows. */
@@ -835,11 +859,11 @@
835859
 if( zPwd[i]==0 ){
836860
 blob_append(pOut, ".", 1);
837861
 }else{
838862
 blob_append(pOut, "..", 2);
839863
 for(j=i+1; zPwd[j]; j++){
840
- if( zPwd[j]=='/' ) {
864
+ if( zPwd[j]=='/' ){
841865
 blob_append(pOut, "/..", 3);
842866
 }
843867
 }
844868
 }
845869
 return;
@@ -852,11 +876,11 @@
852876
 return;
853877
 }
854878
 while( zPath[i-1]!='/' ){ i--; }
855879
 blob_set(&tmp, "../");
856880
 for(j=i; zPwd[j]; j++){
857
- if( zPwd[j]=='/' ) {
881
+ if( zPwd[j]=='/' ){
858882
 blob_append(&tmp, "../", 3);
859883
 }
860884
 }
861885
 blob_append(&tmp, &zPath[i], -1);
862886
 blob_reset(pOut);
863887
--- src/file.c
+++ src/file.c
@@ -490,44 +490,68 @@
490 ** * Does not contain two or more "/" characters in a row.
491 ** * Contains at least one character
492 **
493 ** Invalid UTF8 characters result in a false return if bStrictUtf8 is
494 ** true. If bStrictUtf8 is false, invalid UTF8 characters are silently
495 ** ignored.
 
 
 
 
496 */
497 int file_is_simple_pathname(const char *z, int bStrictUtf8){
498 int i;
499 char c = z[0];
500 char maskNonAscii = bStrictUtf8 ? 0x80 : 0x00;
501 if( c=='/' || c==0 ) return 0;
502 if( c=='.' ){
503 if( z[1]=='/' || z[1]==0 ) return 0;
504 if( z[1]=='.' && (z[2]=='/' || z[2]==0) ) return 0;
505 }
506 for(i=0; (c=z[i])!=0; i++){
507 if( c & maskNonAscii ){
508 if( (c & 0xf0) == 0xf0 ) {
509 /* Unicode characters > U+FFFF are not supported.
510 * Windows XP and earlier cannot handle them.
511 */
512 return 0;
513 }
514 if( (c & 0xf0) == 0xe0 ) {
515 /* This is a 3-byte UTF-8 character */
516 if ( (c & 0xfe) == 0xee ){
517 /* Range U+E000 - U+FFFF (Starting with 0xee or 0xef in UTF-8 ) */
518 if ( !(c & 1) || ((z[i+1] & 0xff) < 0xa4) ){
519 /* Unicode character in the range U+E000 - U+F8FF are for
520 * private use, they shouldn't occur in filenames. */
521 return 0;
522 }
523 }else if( ((c & 0xff) == 0xed) && ((z[i+1] & 0xe0) == 0xa0) ){
524 /* Unicode character in the range U+D800 - U+DFFF are for
525 * surrogate pairs, they shouldn't occur in filenames. */
526 return 0;
527 }
528 }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
529 }else if( c=='\\' ){
530 return 0;
531 }
532 if( c=='/' ){
533 if( z[i+1]=='/' ) return 0;
@@ -578,11 +602,11 @@
578 if( z[i]=='\\' ) z[i] = '/';
579 }
580 #endif
581
582 /* Removing trailing "/" characters */
583 if ( !slash ){
584 while( n>1 && z[n-1]=='/' ){ n--; }
585 }
586
587 /* Remove duplicate '/' characters. Except, two // at the beginning
588 ** of a pathname is allowed since this is important on windows. */
@@ -835,11 +859,11 @@
835 if( zPwd[i]==0 ){
836 blob_append(pOut, ".", 1);
837 }else{
838 blob_append(pOut, "..", 2);
839 for(j=i+1; zPwd[j]; j++){
840 if( zPwd[j]=='/' ) {
841 blob_append(pOut, "/..", 3);
842 }
843 }
844 }
845 return;
@@ -852,11 +876,11 @@
852 return;
853 }
854 while( zPath[i-1]!='/' ){ i--; }
855 blob_set(&tmp, "../");
856 for(j=i; zPwd[j]; j++){
857 if( zPwd[j]=='/' ) {
858 blob_append(&tmp, "../", 3);
859 }
860 }
861 blob_append(&tmp, &zPath[i], -1);
862 blob_reset(pOut);
863
--- src/file.c
+++ src/file.c
@@ -490,44 +490,68 @@
490 ** * Does not contain two or more "/" characters in a row.
491 ** * Contains at least one character
492 **
493 ** Invalid UTF8 characters result in a false return if bStrictUtf8 is
494 ** true. If bStrictUtf8 is false, invalid UTF8 characters are silently
495 ** ignored. See http://en.wikipedia.org/wiki/UTF-8#Invalid_byte_sequences
496 ** and http://en.wikipedia.org/wiki/Unicode (for the noncharacters)
497 **
498 ** The bStrictUtf8 flag is true for new inputs, but is false when parsing
499 ** legacy manifests, for backwards compatibility.
500 */
501 int file_is_simple_pathname(const char *z, int bStrictUtf8){
502 int i;
503 unsigned char c = (unsigned char) z[0];
504 char maskNonAscii = bStrictUtf8 ? 0x80 : 0x00;
505 if( c=='/' || c==0 ) return 0;
506 if( c=='.' ){
507 if( z[1]=='/' || z[1]==0 ) return 0;
508 if( z[1]=='.' && (z[2]=='/' || z[2]==0) ) return 0;
509 }
510 for(i=0; (c=(unsigned char)z[i])!=0; i++){
511 if( c & maskNonAscii ){
512 if( c<0xc2 ){
513 /* Invalid 1-byte UTF-8 sequence, or 2-byte overlong form. */
514 return 0;
515 }else if( (c&0xe0)==0xe0 ){
516 /* 3-byte or more */
517 int unicode;
518 if( c&0x10 ){
519 /* Unicode characters > U+FFFF are not supported.
520 * Windows XP and earlier cannot handle them.
521 */
522 return 0;
523 }
524 /* This is a 3-byte UTF-8 character */
525 unicode = ((c&0x0f)<<12) + ((z[i+1]&0x3f)<<6) + (z[i+2]&0x3f);
526 if( unicode <= 0x07ff ){
527 /* overlong form */
528 return 0;
529 }else if( unicode>=0xe000 ){
530 /* U+E000..U+FFFF */
531 if( (unicode<=0xf8ff) || (unicode>=0xfffe) ){
532 /* U+E000..U+F8FF are for private use.
533 * U+FFFE..U+FFFF are noncharacters. */
534 return 0;
535 } else if( (unicode>=0xfdd0) && (unicode<=0xfdef) ){
536 /* U+FDD0..U+FDEF are noncharacters. */
537 return 0;
538 }
539 }else if( (unicode>=0xD800) && (unicode<=0xDFFF) ){
540 /* U+D800..U+DFFF are for surrogate pairs. */
541 return 0;
542 }
543 }
544 do{
545 if( (z[i+1]&0xc0)!=0x80 ){
546 /* Invalid continuation byte (multi-byte UTF-8) */
547 return 0;
548 }
549 /* The hi-bits of c are used to keep track of the number of expected
550 * continuation-bytes, so we don't need a separate counter. */
551 c<<=1; ++i;
552 }while( c>=0xc0 );
553 }else if( c=='\\' ){
554 return 0;
555 }
556 if( c=='/' ){
557 if( z[i+1]=='/' ) return 0;
@@ -578,11 +602,11 @@
602 if( z[i]=='\\' ) z[i] = '/';
603 }
604 #endif
605
606 /* Removing trailing "/" characters */
607 if( !slash ){
608 while( n>1 && z[n-1]=='/' ){ n--; }
609 }
610
611 /* Remove duplicate '/' characters. Except, two // at the beginning
612 ** of a pathname is allowed since this is important on windows. */
@@ -835,11 +859,11 @@
859 if( zPwd[i]==0 ){
860 blob_append(pOut, ".", 1);
861 }else{
862 blob_append(pOut, "..", 2);
863 for(j=i+1; zPwd[j]; j++){
864 if( zPwd[j]=='/' ){
865 blob_append(pOut, "/..", 3);
866 }
867 }
868 }
869 return;
@@ -852,11 +876,11 @@
876 return;
877 }
878 while( zPath[i-1]!='/' ){ i--; }
879 blob_set(&tmp, "../");
880 for(j=i; zPwd[j]; j++){
881 if( zPwd[j]=='/' ){
882 blob_append(&tmp, "../", 3);
883 }
884 }
885 blob_append(&tmp, &zPath[i], -1);
886 blob_reset(pOut);
887
M src/file.c
+52 -28
--- src/file.c
+++ src/file.c
@@ -490,44 +490,68 @@
490490
 ** * Does not contain two or more "/" characters in a row.
491491
 ** * Contains at least one character
492492
 **
493493
 ** Invalid UTF8 characters result in a false return if bStrictUtf8 is
494494
 ** true. If bStrictUtf8 is false, invalid UTF8 characters are silently
495
-** ignored.
495
+** ignored. See http://en.wikipedia.org/wiki/UTF-8#Invalid_byte_sequences
496
+** and http://en.wikipedia.org/wiki/Unicode (for the noncharacters)
497
+**
498
+** The bStrictUtf8 flag is true for new inputs, but is false when parsing
499
+** legacy manifests, for backwards compatibility.
496500
 */
497501
 int file_is_simple_pathname(const char *z, int bStrictUtf8){
498502
 int i;
499
- char c = z[0];
503
+ unsigned char c = (unsigned char) z[0];
500504
 char maskNonAscii = bStrictUtf8 ? 0x80 : 0x00;
501505
 if( c=='/' || c==0 ) return 0;
502506
 if( c=='.' ){
503507
 if( z[1]=='/' || z[1]==0 ) return 0;
504508
 if( z[1]=='.' && (z[2]=='/' || z[2]==0) ) return 0;
505509
 }
506
- for(i=0; (c=z[i])!=0; i++){
507
- if( c & maskNonAscii ){
508
- if( (c & 0xf0) == 0xf0 ) {
509
- /* Unicode characters > U+FFFF are not supported.
510
- * Windows XP and earlier cannot handle them.
511
- */
512
- return 0;
513
- }
514
- if( (c & 0xf0) == 0xe0 ) {
515
- /* This is a 3-byte UTF-8 character */
516
- if ( (c & 0xfe) == 0xee ){
517
- /* Range U+E000 - U+FFFF (Starting with 0xee or 0xef in UTF-8 ) */
518
- if ( !(c & 1) || ((z[i+1] & 0xff) < 0xa4) ){
519
- /* Unicode character in the range U+E000 - U+F8FF are for
520
- * private use, they shouldn't occur in filenames. */
521
- return 0;
522
- }
523
- }else if( ((c & 0xff) == 0xed) && ((z[i+1] & 0xe0) == 0xa0) ){
524
- /* Unicode character in the range U+D800 - U+DFFF are for
525
- * surrogate pairs, they shouldn't occur in filenames. */
526
- return 0;
527
- }
528
- }
510
+ for(i=0; (c=(unsigned char)z[i])!=0; i++){
511
+ if( c & maskNonAscii ){
512
+ if( c<0xc2 ){
513
+ /* Invalid 1-byte UTF-8 sequence, or 2-byte overlong form. */
514
+ return 0;
515
+ }else if( (c&0xe0)==0xe0 ){
516
+ /* 3-byte or more */
517
+ int unicode;
518
+ if( c&0x10 ){
519
+ /* Unicode characters > U+FFFF are not supported.
520
+ * Windows XP and earlier cannot handle them.
521
+ */
522
+ return 0;
523
+ }
524
+ /* This is a 3-byte UTF-8 character */
525
+ unicode = ((c&0x0f)<<12) + ((z[i+1]&0x3f)<<6) + (z[i+2]&0x3f);
526
+ if( unicode <= 0x07ff ){
527
+ /* overlong form */
528
+ return 0;
529
+ }else if( unicode>=0xe000 ){
530
+ /* U+E000..U+FFFF */
531
+ if( (unicode<=0xf8ff) || (unicode>=0xfffe) ){
532
+ /* U+E000..U+F8FF are for private use.
533
+ * U+FFFE..U+FFFF are noncharacters. */
534
+ return 0;
535
+ } else if( (unicode>=0xfdd0) && (unicode<=0xfdef) ){
536
+ /* U+FDD0..U+FDEF are noncharacters. */
537
+ return 0;
538
+ }
539
+ }else if( (unicode>=0xD800) && (unicode<=0xDFFF) ){
540
+ /* U+D800..U+DFFF are for surrogate pairs. */
541
+ return 0;
542
+ }
543
+ }
544
+ do{
545
+ if( (z[i+1]&0xc0)!=0x80 ){
546
+ /* Invalid continuation byte (multi-byte UTF-8) */
547
+ return 0;
548
+ }
549
+ /* The hi-bits of c are used to keep track of the number of expected
550
+ * continuation-bytes, so we don't need a separate counter. */
551
+ c<<=1; ++i;
552
+ }while( c>=0xc0 );
529553
 }else if( c=='\\' ){
530554
 return 0;
531555
 }
532556
 if( c=='/' ){
533557
 if( z[i+1]=='/' ) return 0;
@@ -578,11 +602,11 @@
578602
 if( z[i]=='\\' ) z[i] = '/';
579603
 }
580604
 #endif
581605
582606
 /* Removing trailing "/" characters */
583
- if ( !slash ){
607
+ if( !slash ){
584608
 while( n>1 && z[n-1]=='/' ){ n--; }
585609
 }
586610
587611
 /* Remove duplicate '/' characters. Except, two // at the beginning
588612
 ** of a pathname is allowed since this is important on windows. */
@@ -835,11 +859,11 @@
835859
 if( zPwd[i]==0 ){
836860
 blob_append(pOut, ".", 1);
837861
 }else{
838862
 blob_append(pOut, "..", 2);
839863
 for(j=i+1; zPwd[j]; j++){
840
- if( zPwd[j]=='/' ) {
864
+ if( zPwd[j]=='/' ){
841865
 blob_append(pOut, "/..", 3);
842866
 }
843867
 }
844868
 }
845869
 return;
@@ -852,11 +876,11 @@
852876
 return;
853877
 }
854878
 while( zPath[i-1]!='/' ){ i--; }
855879
 blob_set(&tmp, "../");
856880
 for(j=i; zPwd[j]; j++){
857
- if( zPwd[j]=='/' ) {
881
+ if( zPwd[j]=='/' ){
858882
 blob_append(&tmp, "../", 3);
859883
 }
860884
 }
861885
 blob_append(&tmp, &zPath[i], -1);
862886
 blob_reset(pOut);
863887
--- src/file.c
+++ src/file.c
@@ -490,44 +490,68 @@
490 ** * Does not contain two or more "/" characters in a row.
491 ** * Contains at least one character
492 **
493 ** Invalid UTF8 characters result in a false return if bStrictUtf8 is
494 ** true. If bStrictUtf8 is false, invalid UTF8 characters are silently
495 ** ignored.
 
 
 
 
496 */
497 int file_is_simple_pathname(const char *z, int bStrictUtf8){
498 int i;
499 char c = z[0];
500 char maskNonAscii = bStrictUtf8 ? 0x80 : 0x00;
501 if( c=='/' || c==0 ) return 0;
502 if( c=='.' ){
503 if( z[1]=='/' || z[1]==0 ) return 0;
504 if( z[1]=='.' && (z[2]=='/' || z[2]==0) ) return 0;
505 }
506 for(i=0; (c=z[i])!=0; i++){
507 if( c & maskNonAscii ){
508 if( (c & 0xf0) == 0xf0 ) {
509 /* Unicode characters > U+FFFF are not supported.
510 * Windows XP and earlier cannot handle them.
511 */
512 return 0;
513 }
514 if( (c & 0xf0) == 0xe0 ) {
515 /* This is a 3-byte UTF-8 character */
516 if ( (c & 0xfe) == 0xee ){
517 /* Range U+E000 - U+FFFF (Starting with 0xee or 0xef in UTF-8 ) */
518 if ( !(c & 1) || ((z[i+1] & 0xff) < 0xa4) ){
519 /* Unicode character in the range U+E000 - U+F8FF are for
520 * private use, they shouldn't occur in filenames. */
521 return 0;
522 }
523 }else if( ((c & 0xff) == 0xed) && ((z[i+1] & 0xe0) == 0xa0) ){
524 /* Unicode character in the range U+D800 - U+DFFF are for
525 * surrogate pairs, they shouldn't occur in filenames. */
526 return 0;
527 }
528 }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
529 }else if( c=='\\' ){
530 return 0;
531 }
532 if( c=='/' ){
533 if( z[i+1]=='/' ) return 0;
@@ -578,11 +602,11 @@
578 if( z[i]=='\\' ) z[i] = '/';
579 }
580 #endif
581
582 /* Removing trailing "/" characters */
583 if ( !slash ){
584 while( n>1 && z[n-1]=='/' ){ n--; }
585 }
586
587 /* Remove duplicate '/' characters. Except, two // at the beginning
588 ** of a pathname is allowed since this is important on windows. */
@@ -835,11 +859,11 @@
835 if( zPwd[i]==0 ){
836 blob_append(pOut, ".", 1);
837 }else{
838 blob_append(pOut, "..", 2);
839 for(j=i+1; zPwd[j]; j++){
840 if( zPwd[j]=='/' ) {
841 blob_append(pOut, "/..", 3);
842 }
843 }
844 }
845 return;
@@ -852,11 +876,11 @@
852 return;
853 }
854 while( zPath[i-1]!='/' ){ i--; }
855 blob_set(&tmp, "../");
856 for(j=i; zPwd[j]; j++){
857 if( zPwd[j]=='/' ) {
858 blob_append(&tmp, "../", 3);
859 }
860 }
861 blob_append(&tmp, &zPath[i], -1);
862 blob_reset(pOut);
863
--- src/file.c
+++ src/file.c
@@ -490,44 +490,68 @@
490 ** * Does not contain two or more "/" characters in a row.
491 ** * Contains at least one character
492 **
493 ** Invalid UTF8 characters result in a false return if bStrictUtf8 is
494 ** true. If bStrictUtf8 is false, invalid UTF8 characters are silently
495 ** ignored. See http://en.wikipedia.org/wiki/UTF-8#Invalid_byte_sequences
496 ** and http://en.wikipedia.org/wiki/Unicode (for the noncharacters)
497 **
498 ** The bStrictUtf8 flag is true for new inputs, but is false when parsing
499 ** legacy manifests, for backwards compatibility.
500 */
501 int file_is_simple_pathname(const char *z, int bStrictUtf8){
502 int i;
503 unsigned char c = (unsigned char) z[0];
504 char maskNonAscii = bStrictUtf8 ? 0x80 : 0x00;
505 if( c=='/' || c==0 ) return 0;
506 if( c=='.' ){
507 if( z[1]=='/' || z[1]==0 ) return 0;
508 if( z[1]=='.' && (z[2]=='/' || z[2]==0) ) return 0;
509 }
510 for(i=0; (c=(unsigned char)z[i])!=0; i++){
511 if( c & maskNonAscii ){
512 if( c<0xc2 ){
513 /* Invalid 1-byte UTF-8 sequence, or 2-byte overlong form. */
514 return 0;
515 }else if( (c&0xe0)==0xe0 ){
516 /* 3-byte or more */
517 int unicode;
518 if( c&0x10 ){
519 /* Unicode characters > U+FFFF are not supported.
520 * Windows XP and earlier cannot handle them.
521 */
522 return 0;
523 }
524 /* This is a 3-byte UTF-8 character */
525 unicode = ((c&0x0f)<<12) + ((z[i+1]&0x3f)<<6) + (z[i+2]&0x3f);
526 if( unicode <= 0x07ff ){
527 /* overlong form */
528 return 0;
529 }else if( unicode>=0xe000 ){
530 /* U+E000..U+FFFF */
531 if( (unicode<=0xf8ff) || (unicode>=0xfffe) ){
532 /* U+E000..U+F8FF are for private use.
533 * U+FFFE..U+FFFF are noncharacters. */
534 return 0;
535 } else if( (unicode>=0xfdd0) && (unicode<=0xfdef) ){
536 /* U+FDD0..U+FDEF are noncharacters. */
537 return 0;
538 }
539 }else if( (unicode>=0xD800) && (unicode<=0xDFFF) ){
540 /* U+D800..U+DFFF are for surrogate pairs. */
541 return 0;
542 }
543 }
544 do{
545 if( (z[i+1]&0xc0)!=0x80 ){
546 /* Invalid continuation byte (multi-byte UTF-8) */
547 return 0;
548 }
549 /* The hi-bits of c are used to keep track of the number of expected
550 * continuation-bytes, so we don't need a separate counter. */
551 c<<=1; ++i;
552 }while( c>=0xc0 );
553 }else if( c=='\\' ){
554 return 0;
555 }
556 if( c=='/' ){
557 if( z[i+1]=='/' ) return 0;
@@ -578,11 +602,11 @@
602 if( z[i]=='\\' ) z[i] = '/';
603 }
604 #endif
605
606 /* Removing trailing "/" characters */
607 if( !slash ){
608 while( n>1 && z[n-1]=='/' ){ n--; }
609 }
610
611 /* Remove duplicate '/' characters. Except, two // at the beginning
612 ** of a pathname is allowed since this is important on windows. */
@@ -835,11 +859,11 @@
859 if( zPwd[i]==0 ){
860 blob_append(pOut, ".", 1);
861 }else{
862 blob_append(pOut, "..", 2);
863 for(j=i+1; zPwd[j]; j++){
864 if( zPwd[j]=='/' ){
865 blob_append(pOut, "/..", 3);
866 }
867 }
868 }
869 return;
@@ -852,11 +876,11 @@
876 return;
877 }
878 while( zPath[i-1]!='/' ){ i--; }
879 blob_set(&tmp, "../");
880 for(j=i; zPwd[j]; j++){
881 if( zPwd[j]=='/' ){
882 blob_append(&tmp, "../", 3);
883 }
884 }
885 blob_append(&tmp, &zPath[i], -1);
886 blob_reset(pOut);
887

Keyboard Shortcuts

Open search /
Next entry (timeline) j
Previous entry (timeline) k
Open focused entry Enter
Show this help ?
Toggle theme Top nav button