Fossil SCM
Some optimizations. No change in functionality. - In clean_cmd, use new function vfile_scan2. - When checking for multiple bytes/characters always do that in order of likelihood: '\n' is more likely than '\0', which is more likely than '\r'. - Off-by-one error in looks_like_utf16(). - Allow starts_with_utf16_bom() to be used in detecting UTF-16 without BOM as well: Always determine pbReverse, even if no BOM is present. - Remove unused variable.
Commit
fa506944056737fddefcee01652f6e12e75b9a53
Parent
f4bcdb62fb7fcb6…
3 files changed
+2
-3
+28
-19
-3
+2
-3
| --- src/checkin.c | ||
| +++ src/checkin.c | ||
| @@ -447,12 +447,13 @@ | ||
| 447 | 447 | n = strlen(g.zLocalRoot); |
| 448 | 448 | blob_init(&path, g.zLocalRoot, n-1); |
| 449 | 449 | pIgnore = glob_create(zIgnoreFlag); |
| 450 | 450 | pKeep = glob_create(zKeepFlag); |
| 451 | 451 | pClean = glob_create(zCleanFlag); |
| 452 | - vfile_scan(&path, blob_size(&path), scanFlags, pKeep); | |
| 452 | + vfile_scan2(&path, blob_size(&path), scanFlags, pIgnore, pKeep); | |
| 453 | 453 | glob_free(pKeep); |
| 454 | + glob_free(pIgnore); | |
| 454 | 455 | db_prepare(&q, |
| 455 | 456 | "SELECT %Q || x FROM sfile" |
| 456 | 457 | " WHERE x NOT IN (%s)" |
| 457 | 458 | " ORDER BY 1", |
| 458 | 459 | g.zLocalRoot, fossil_all_reserved_names(0) |
| @@ -461,11 +462,10 @@ | ||
| 461 | 462 | db_multi_exec("DELETE FROM sfile WHERE x=%B", &repo); |
| 462 | 463 | } |
| 463 | 464 | db_multi_exec("DELETE FROM sfile WHERE x IN (SELECT pathname FROM vfile)"); |
| 464 | 465 | while( db_step(&q)==SQLITE_ROW ){ |
| 465 | 466 | const char *zName = db_column_text(&q, 0); |
| 466 | - if( glob_match(pIgnore, zName+n) ) continue; | |
| 467 | 467 | if( !allFlag && !glob_match(pClean, zName+n) ){ |
| 468 | 468 | Blob ans; |
| 469 | 469 | char cReply; |
| 470 | 470 | char *prompt = mprintf("remove unmanaged file \"%s\" (a=all/y/N)? ", |
| 471 | 471 | zName+n); |
| @@ -482,11 +482,10 @@ | ||
| 482 | 482 | if( !dryRunFlag ){ |
| 483 | 483 | file_delete(zName); |
| 484 | 484 | } |
| 485 | 485 | } |
| 486 | 486 | glob_free(pClean); |
| 487 | - glob_free(pIgnore); | |
| 488 | 487 | db_finalize(&q); |
| 489 | 488 | } |
| 490 | 489 | |
| 491 | 490 | /* |
| 492 | 491 | ** Prompt the user for a check-in or stash comment (given in pPrompt), |
| 493 | 492 |
| --- src/checkin.c | |
| +++ src/checkin.c | |
| @@ -447,12 +447,13 @@ | |
| 447 | n = strlen(g.zLocalRoot); |
| 448 | blob_init(&path, g.zLocalRoot, n-1); |
| 449 | pIgnore = glob_create(zIgnoreFlag); |
| 450 | pKeep = glob_create(zKeepFlag); |
| 451 | pClean = glob_create(zCleanFlag); |
| 452 | vfile_scan(&path, blob_size(&path), scanFlags, pKeep); |
| 453 | glob_free(pKeep); |
| 454 | db_prepare(&q, |
| 455 | "SELECT %Q || x FROM sfile" |
| 456 | " WHERE x NOT IN (%s)" |
| 457 | " ORDER BY 1", |
| 458 | g.zLocalRoot, fossil_all_reserved_names(0) |
| @@ -461,11 +462,10 @@ | |
| 461 | db_multi_exec("DELETE FROM sfile WHERE x=%B", &repo); |
| 462 | } |
| 463 | db_multi_exec("DELETE FROM sfile WHERE x IN (SELECT pathname FROM vfile)"); |
| 464 | while( db_step(&q)==SQLITE_ROW ){ |
| 465 | const char *zName = db_column_text(&q, 0); |
| 466 | if( glob_match(pIgnore, zName+n) ) continue; |
| 467 | if( !allFlag && !glob_match(pClean, zName+n) ){ |
| 468 | Blob ans; |
| 469 | char cReply; |
| 470 | char *prompt = mprintf("remove unmanaged file \"%s\" (a=all/y/N)? ", |
| 471 | zName+n); |
| @@ -482,11 +482,10 @@ | |
| 482 | if( !dryRunFlag ){ |
| 483 | file_delete(zName); |
| 484 | } |
| 485 | } |
| 486 | glob_free(pClean); |
| 487 | glob_free(pIgnore); |
| 488 | db_finalize(&q); |
| 489 | } |
| 490 | |
| 491 | /* |
| 492 | ** Prompt the user for a check-in or stash comment (given in pPrompt), |
| 493 |
| --- src/checkin.c | |
| +++ src/checkin.c | |
| @@ -447,12 +447,13 @@ | |
| 447 | n = strlen(g.zLocalRoot); |
| 448 | blob_init(&path, g.zLocalRoot, n-1); |
| 449 | pIgnore = glob_create(zIgnoreFlag); |
| 450 | pKeep = glob_create(zKeepFlag); |
| 451 | pClean = glob_create(zCleanFlag); |
| 452 | vfile_scan2(&path, blob_size(&path), scanFlags, pIgnore, pKeep); |
| 453 | glob_free(pKeep); |
| 454 | glob_free(pIgnore); |
| 455 | db_prepare(&q, |
| 456 | "SELECT %Q || x FROM sfile" |
| 457 | " WHERE x NOT IN (%s)" |
| 458 | " ORDER BY 1", |
| 459 | g.zLocalRoot, fossil_all_reserved_names(0) |
| @@ -461,11 +462,10 @@ | |
| 462 | db_multi_exec("DELETE FROM sfile WHERE x=%B", &repo); |
| 463 | } |
| 464 | db_multi_exec("DELETE FROM sfile WHERE x IN (SELECT pathname FROM vfile)"); |
| 465 | while( db_step(&q)==SQLITE_ROW ){ |
| 466 | const char *zName = db_column_text(&q, 0); |
| 467 | if( !allFlag && !glob_match(pClean, zName+n) ){ |
| 468 | Blob ans; |
| 469 | char cReply; |
| 470 | char *prompt = mprintf("remove unmanaged file \"%s\" (a=all/y/N)? ", |
| 471 | zName+n); |
| @@ -482,11 +482,10 @@ | |
| 482 | if( !dryRunFlag ){ |
| 483 | file_delete(zName); |
| 484 | } |
| 485 | } |
| 486 | glob_free(pClean); |
| 487 | db_finalize(&q); |
| 488 | } |
| 489 | |
| 490 | /* |
| 491 | ** Prompt the user for a check-in or stash comment (given in pPrompt), |
| 492 |
+28
-19
| --- src/diff.c | ||
| +++ src/diff.c | ||
| @@ -243,26 +243,25 @@ | ||
| 243 | 243 | unsigned int n = blob_size(pContent); |
| 244 | 244 | int j, c, flags = LOOK_NONE; /* Assume UTF-8 text, prove otherwise */ |
| 245 | 245 | |
| 246 | 246 | if( n==0 ) return flags; /* Empty file -> text */ |
| 247 | 247 | c = *z; |
| 248 | - if( c==0 ){ | |
| 248 | + j = (c!='\n'); | |
| 249 | + if( !j ){ | |
| 250 | + flags |= (LOOK_LF | LOOK_LONE_LF); /* Found LF as first char */ | |
| 251 | + }else if( c==0 ){ | |
| 249 | 252 | flags |= LOOK_NUL; /* NUL character in a file -> binary */ |
| 250 | 253 | }else if( c=='\r' ){ |
| 251 | 254 | flags |= LOOK_CR; |
| 252 | 255 | if( n<=1 || z[1]!='\n' ){ |
| 253 | 256 | flags |= LOOK_LONE_CR; /* More chars, next char is not LF */ |
| 254 | 257 | } |
| 255 | 258 | } |
| 256 | - j = (c!='\n'); | |
| 257 | - if( !j ) flags |= (LOOK_LF | LOOK_LONE_LF); /* Found LF as first char */ | |
| 258 | 259 | while( !(flags&stopFlags) && --n>0 ){ |
| 259 | 260 | int c2 = c; |
| 260 | 261 | c = *++z; ++j; |
| 261 | - if( c==0 ){ | |
| 262 | - flags |= LOOK_NUL; /* NUL character in a file -> binary */ | |
| 263 | - }else if( c=='\n' ){ | |
| 262 | + if( c=='\n' ){ | |
| 264 | 263 | flags |= LOOK_LF; |
| 265 | 264 | if( c2=='\r' ){ |
| 266 | 265 | flags |= (LOOK_CR | LOOK_CRLF); /* Found LF preceded by CR */ |
| 267 | 266 | }else{ |
| 268 | 267 | flags |= LOOK_LONE_LF; |
| @@ -269,10 +268,12 @@ | ||
| 269 | 268 | } |
| 270 | 269 | if( j>LENGTH_MASK ){ |
| 271 | 270 | flags |= LOOK_LONG; /* Very long line -> binary */ |
| 272 | 271 | } |
| 273 | 272 | j = 0; |
| 273 | + }else if( c==0 ){ | |
| 274 | + flags |= LOOK_NUL; /* NUL character in a file -> binary */ | |
| 274 | 275 | }else if( c=='\r' ){ |
| 275 | 276 | flags |= LOOK_CR; |
| 276 | 277 | if( n<=1 || z[1]!='\n' ){ |
| 277 | 278 | flags |= LOOK_LONE_CR; /* More chars, next char is not LF */ |
| 278 | 279 | } |
| @@ -357,32 +358,31 @@ | ||
| 357 | 358 | } |
| 358 | 359 | c = *z; |
| 359 | 360 | if( bReverse ){ |
| 360 | 361 | c = UTF16_SWAP(c); |
| 361 | 362 | } |
| 362 | - if( c==0 ){ | |
| 363 | + j = (c!='\n'); | |
| 364 | + if( !j ){ | |
| 365 | + flags |= (LOOK_LF | LOOK_LONE_LF); /* Found LF as first char */ | |
| 366 | + }else if( c==0 ){ | |
| 363 | 367 | flags |= LOOK_NUL; /* NUL character in a file -> binary */ |
| 364 | 368 | }else if( c=='\r' ){ |
| 365 | 369 | flags |= LOOK_CR; |
| 366 | 370 | if( n<=sizeof(WCHAR_T) || UTF16_SWAP_IF(bReverse, z[1])!='\n' ){ |
| 367 | 371 | flags |= LOOK_LONE_CR; /* More chars, next char is not LF */ |
| 368 | 372 | } |
| 369 | 373 | } |
| 370 | - j = (c!='\n'); | |
| 371 | - if( !j ) flags |= (LOOK_LF | LOOK_LONE_LF); /* Found LF as first char */ | |
| 372 | 374 | while( 1 ){ |
| 373 | 375 | int c2 = c; |
| 374 | 376 | n -= sizeof(WCHAR_T); |
| 375 | 377 | if( (flags&stopFlags) || n<sizeof(WCHAR_T) ) break; |
| 376 | 378 | c = *++z; |
| 377 | 379 | if( bReverse ){ |
| 378 | 380 | c = UTF16_SWAP(c); |
| 379 | 381 | } |
| 380 | 382 | ++j; |
| 381 | - if( c==0 ){ | |
| 382 | - flags |= LOOK_NUL; /* NUL character in a file -> binary */ | |
| 383 | - }else if( c=='\n' ){ | |
| 383 | + if( c=='\n' ){ | |
| 384 | 384 | flags |= LOOK_LF; |
| 385 | 385 | if( c2=='\r' ){ |
| 386 | 386 | flags |= (LOOK_CR | LOOK_CRLF); /* Found LF preceded by CR */ |
| 387 | 387 | }else{ |
| 388 | 388 | flags |= LOOK_LONE_LF; |
| @@ -389,13 +389,15 @@ | ||
| 389 | 389 | } |
| 390 | 390 | if( j>UTF16_LENGTH_MASK ){ |
| 391 | 391 | flags |= LOOK_LONG; /* Very long line -> binary */ |
| 392 | 392 | } |
| 393 | 393 | j = 0; |
| 394 | + }else if( c==0 ){ | |
| 395 | + flags |= LOOK_NUL; /* NUL character in a file -> binary */ | |
| 394 | 396 | }else if( c=='\r' ){ |
| 395 | 397 | flags |= LOOK_CR; |
| 396 | - if( n<=sizeof(WCHAR_T) || UTF16_SWAP_IF(bReverse, z[1])!='\n' ){ | |
| 398 | + if( n<2*sizeof(WCHAR_T) || UTF16_SWAP_IF(bReverse, z[1])!='\n' ){ | |
| 397 | 399 | flags |= LOOK_LONE_CR; /* More chars, next char is not LF */ |
| 398 | 400 | } |
| 399 | 401 | } |
| 400 | 402 | } |
| 401 | 403 | if( n ){ |
| @@ -436,28 +438,35 @@ | ||
| 436 | 438 | /* |
| 437 | 439 | ** This function returns non-zero if the blob starts with a UTF-16 |
| 438 | 440 | ** byte-order-mark (BOM), either in the endianness of the machine |
| 439 | 441 | ** or in reversed byte order. The UTF-32 BOM is ruled out by checking |
| 440 | 442 | ** if the UTF-16 BOM is not immediately followed by (utf16) 0. |
| 441 | -** pnByte and pbReverse are only set when the function returns 1. | |
| 443 | +** pnByte is only set when the function returns 1. | |
| 444 | +** | |
| 445 | +** pbReverse is always set, even when no BOM is found. Without BOM, | |
| 446 | +** it is set to 1 on little-endian and 0 on big-endian platforms. See | |
| 447 | +** clause D98 of conformance (section 3.10) of the Unicode standard. | |
| 442 | 448 | */ |
| 443 | 449 | int starts_with_utf16_bom( |
| 444 | 450 | const Blob *pContent, /* IN: Blob content to perform BOM detection on. */ |
| 445 | 451 | int *pnByte, /* OUT: The number of bytes used for the BOM. */ |
| 446 | 452 | int *pbReverse /* OUT: Non-zero for BOM in reverse byte-order. */ |
| 447 | 453 | ){ |
| 448 | 454 | const unsigned short *z = (unsigned short *)blob_buffer(pContent); |
| 449 | 455 | int bomSize = sizeof(unsigned short); |
| 450 | 456 | int size = blob_size(pContent); |
| 457 | + static const int one = 1; | |
| 451 | 458 | |
| 452 | - if( size<bomSize ) return 0; /* No: cannot read BOM. */ | |
| 453 | - if( size>=(2*bomSize) && z[1]==0 ) return 0; /* No: possible UTF-32. */ | |
| 454 | - if( z[0]==0xfffe ){ | |
| 459 | + if( size<bomSize ) goto noBom; /* No: cannot read BOM. */ | |
| 460 | + if( size>=(2*bomSize) && z[1]==0 ) goto noBom; /* No: possible UTF-32. */ | |
| 461 | + if( z[0]==0xfeff ){ | |
| 462 | + if( pbReverse ) *pbReverse = 0; | |
| 463 | + }else if( z[0]==0xfffe ){ | |
| 455 | 464 | if( pbReverse ) *pbReverse = 1; |
| 456 | - }else if( z[0]==0xfeff ){ | |
| 457 | - if( pbReverse ) *pbReverse = 0; | |
| 458 | 465 | }else{ |
| 466 | + noBom: | |
| 467 | + if( pbReverse ) *pbReverse = *(char *) &one; | |
| 459 | 468 | return 0; /* No: UTF-16 byte-order-mark not found. */ |
| 460 | 469 | } |
| 461 | 470 | if( pnByte ) *pnByte = bomSize; |
| 462 | 471 | return 1; /* Yes. */ |
| 463 | 472 | } |
| 464 | 473 |
| --- src/diff.c | |
| +++ src/diff.c | |
| @@ -243,26 +243,25 @@ | |
| 243 | unsigned int n = blob_size(pContent); |
| 244 | int j, c, flags = LOOK_NONE; /* Assume UTF-8 text, prove otherwise */ |
| 245 | |
| 246 | if( n==0 ) return flags; /* Empty file -> text */ |
| 247 | c = *z; |
| 248 | if( c==0 ){ |
| 249 | flags |= LOOK_NUL; /* NUL character in a file -> binary */ |
| 250 | }else if( c=='\r' ){ |
| 251 | flags |= LOOK_CR; |
| 252 | if( n<=1 || z[1]!='\n' ){ |
| 253 | flags |= LOOK_LONE_CR; /* More chars, next char is not LF */ |
| 254 | } |
| 255 | } |
| 256 | j = (c!='\n'); |
| 257 | if( !j ) flags |= (LOOK_LF | LOOK_LONE_LF); /* Found LF as first char */ |
| 258 | while( !(flags&stopFlags) && --n>0 ){ |
| 259 | int c2 = c; |
| 260 | c = *++z; ++j; |
| 261 | if( c==0 ){ |
| 262 | flags |= LOOK_NUL; /* NUL character in a file -> binary */ |
| 263 | }else if( c=='\n' ){ |
| 264 | flags |= LOOK_LF; |
| 265 | if( c2=='\r' ){ |
| 266 | flags |= (LOOK_CR | LOOK_CRLF); /* Found LF preceded by CR */ |
| 267 | }else{ |
| 268 | flags |= LOOK_LONE_LF; |
| @@ -269,10 +268,12 @@ | |
| 269 | } |
| 270 | if( j>LENGTH_MASK ){ |
| 271 | flags |= LOOK_LONG; /* Very long line -> binary */ |
| 272 | } |
| 273 | j = 0; |
| 274 | }else if( c=='\r' ){ |
| 275 | flags |= LOOK_CR; |
| 276 | if( n<=1 || z[1]!='\n' ){ |
| 277 | flags |= LOOK_LONE_CR; /* More chars, next char is not LF */ |
| 278 | } |
| @@ -357,32 +358,31 @@ | |
| 357 | } |
| 358 | c = *z; |
| 359 | if( bReverse ){ |
| 360 | c = UTF16_SWAP(c); |
| 361 | } |
| 362 | if( c==0 ){ |
| 363 | flags |= LOOK_NUL; /* NUL character in a file -> binary */ |
| 364 | }else if( c=='\r' ){ |
| 365 | flags |= LOOK_CR; |
| 366 | if( n<=sizeof(WCHAR_T) || UTF16_SWAP_IF(bReverse, z[1])!='\n' ){ |
| 367 | flags |= LOOK_LONE_CR; /* More chars, next char is not LF */ |
| 368 | } |
| 369 | } |
| 370 | j = (c!='\n'); |
| 371 | if( !j ) flags |= (LOOK_LF | LOOK_LONE_LF); /* Found LF as first char */ |
| 372 | while( 1 ){ |
| 373 | int c2 = c; |
| 374 | n -= sizeof(WCHAR_T); |
| 375 | if( (flags&stopFlags) || n<sizeof(WCHAR_T) ) break; |
| 376 | c = *++z; |
| 377 | if( bReverse ){ |
| 378 | c = UTF16_SWAP(c); |
| 379 | } |
| 380 | ++j; |
| 381 | if( c==0 ){ |
| 382 | flags |= LOOK_NUL; /* NUL character in a file -> binary */ |
| 383 | }else if( c=='\n' ){ |
| 384 | flags |= LOOK_LF; |
| 385 | if( c2=='\r' ){ |
| 386 | flags |= (LOOK_CR | LOOK_CRLF); /* Found LF preceded by CR */ |
| 387 | }else{ |
| 388 | flags |= LOOK_LONE_LF; |
| @@ -389,13 +389,15 @@ | |
| 389 | } |
| 390 | if( j>UTF16_LENGTH_MASK ){ |
| 391 | flags |= LOOK_LONG; /* Very long line -> binary */ |
| 392 | } |
| 393 | j = 0; |
| 394 | }else if( c=='\r' ){ |
| 395 | flags |= LOOK_CR; |
| 396 | if( n<=sizeof(WCHAR_T) || UTF16_SWAP_IF(bReverse, z[1])!='\n' ){ |
| 397 | flags |= LOOK_LONE_CR; /* More chars, next char is not LF */ |
| 398 | } |
| 399 | } |
| 400 | } |
| 401 | if( n ){ |
| @@ -436,28 +438,35 @@ | |
| 436 | /* |
| 437 | ** This function returns non-zero if the blob starts with a UTF-16 |
| 438 | ** byte-order-mark (BOM), either in the endianness of the machine |
| 439 | ** or in reversed byte order. The UTF-32 BOM is ruled out by checking |
| 440 | ** if the UTF-16 BOM is not immediately followed by (utf16) 0. |
| 441 | ** pnByte and pbReverse are only set when the function returns 1. |
| 442 | */ |
| 443 | int starts_with_utf16_bom( |
| 444 | const Blob *pContent, /* IN: Blob content to perform BOM detection on. */ |
| 445 | int *pnByte, /* OUT: The number of bytes used for the BOM. */ |
| 446 | int *pbReverse /* OUT: Non-zero for BOM in reverse byte-order. */ |
| 447 | ){ |
| 448 | const unsigned short *z = (unsigned short *)blob_buffer(pContent); |
| 449 | int bomSize = sizeof(unsigned short); |
| 450 | int size = blob_size(pContent); |
| 451 | |
| 452 | if( size<bomSize ) return 0; /* No: cannot read BOM. */ |
| 453 | if( size>=(2*bomSize) && z[1]==0 ) return 0; /* No: possible UTF-32. */ |
| 454 | if( z[0]==0xfffe ){ |
| 455 | if( pbReverse ) *pbReverse = 1; |
| 456 | }else if( z[0]==0xfeff ){ |
| 457 | if( pbReverse ) *pbReverse = 0; |
| 458 | }else{ |
| 459 | return 0; /* No: UTF-16 byte-order-mark not found. */ |
| 460 | } |
| 461 | if( pnByte ) *pnByte = bomSize; |
| 462 | return 1; /* Yes. */ |
| 463 | } |
| 464 |
| --- src/diff.c | |
| +++ src/diff.c | |
| @@ -243,26 +243,25 @@ | |
| 243 | unsigned int n = blob_size(pContent); |
| 244 | int j, c, flags = LOOK_NONE; /* Assume UTF-8 text, prove otherwise */ |
| 245 | |
| 246 | if( n==0 ) return flags; /* Empty file -> text */ |
| 247 | c = *z; |
| 248 | j = (c!='\n'); |
| 249 | if( !j ){ |
| 250 | flags |= (LOOK_LF | LOOK_LONE_LF); /* Found LF as first char */ |
| 251 | }else if( c==0 ){ |
| 252 | flags |= LOOK_NUL; /* NUL character in a file -> binary */ |
| 253 | }else if( c=='\r' ){ |
| 254 | flags |= LOOK_CR; |
| 255 | if( n<=1 || z[1]!='\n' ){ |
| 256 | flags |= LOOK_LONE_CR; /* More chars, next char is not LF */ |
| 257 | } |
| 258 | } |
| 259 | while( !(flags&stopFlags) && --n>0 ){ |
| 260 | int c2 = c; |
| 261 | c = *++z; ++j; |
| 262 | if( c=='\n' ){ |
| 263 | flags |= LOOK_LF; |
| 264 | if( c2=='\r' ){ |
| 265 | flags |= (LOOK_CR | LOOK_CRLF); /* Found LF preceded by CR */ |
| 266 | }else{ |
| 267 | flags |= LOOK_LONE_LF; |
| @@ -269,10 +268,12 @@ | |
| 268 | } |
| 269 | if( j>LENGTH_MASK ){ |
| 270 | flags |= LOOK_LONG; /* Very long line -> binary */ |
| 271 | } |
| 272 | j = 0; |
| 273 | }else if( c==0 ){ |
| 274 | flags |= LOOK_NUL; /* NUL character in a file -> binary */ |
| 275 | }else if( c=='\r' ){ |
| 276 | flags |= LOOK_CR; |
| 277 | if( n<=1 || z[1]!='\n' ){ |
| 278 | flags |= LOOK_LONE_CR; /* More chars, next char is not LF */ |
| 279 | } |
| @@ -357,32 +358,31 @@ | |
| 358 | } |
| 359 | c = *z; |
| 360 | if( bReverse ){ |
| 361 | c = UTF16_SWAP(c); |
| 362 | } |
| 363 | j = (c!='\n'); |
| 364 | if( !j ){ |
| 365 | flags |= (LOOK_LF | LOOK_LONE_LF); /* Found LF as first char */ |
| 366 | }else if( c==0 ){ |
| 367 | flags |= LOOK_NUL; /* NUL character in a file -> binary */ |
| 368 | }else if( c=='\r' ){ |
| 369 | flags |= LOOK_CR; |
| 370 | if( n<=sizeof(WCHAR_T) || UTF16_SWAP_IF(bReverse, z[1])!='\n' ){ |
| 371 | flags |= LOOK_LONE_CR; /* More chars, next char is not LF */ |
| 372 | } |
| 373 | } |
| 374 | while( 1 ){ |
| 375 | int c2 = c; |
| 376 | n -= sizeof(WCHAR_T); |
| 377 | if( (flags&stopFlags) || n<sizeof(WCHAR_T) ) break; |
| 378 | c = *++z; |
| 379 | if( bReverse ){ |
| 380 | c = UTF16_SWAP(c); |
| 381 | } |
| 382 | ++j; |
| 383 | if( c=='\n' ){ |
| 384 | flags |= LOOK_LF; |
| 385 | if( c2=='\r' ){ |
| 386 | flags |= (LOOK_CR | LOOK_CRLF); /* Found LF preceded by CR */ |
| 387 | }else{ |
| 388 | flags |= LOOK_LONE_LF; |
| @@ -389,13 +389,15 @@ | |
| 389 | } |
| 390 | if( j>UTF16_LENGTH_MASK ){ |
| 391 | flags |= LOOK_LONG; /* Very long line -> binary */ |
| 392 | } |
| 393 | j = 0; |
| 394 | }else if( c==0 ){ |
| 395 | flags |= LOOK_NUL; /* NUL character in a file -> binary */ |
| 396 | }else if( c=='\r' ){ |
| 397 | flags |= LOOK_CR; |
| 398 | if( n<2*sizeof(WCHAR_T) || UTF16_SWAP_IF(bReverse, z[1])!='\n' ){ |
| 399 | flags |= LOOK_LONE_CR; /* More chars, next char is not LF */ |
| 400 | } |
| 401 | } |
| 402 | } |
| 403 | if( n ){ |
| @@ -436,28 +438,35 @@ | |
| 438 | /* |
| 439 | ** This function returns non-zero if the blob starts with a UTF-16 |
| 440 | ** byte-order-mark (BOM), either in the endianness of the machine |
| 441 | ** or in reversed byte order. The UTF-32 BOM is ruled out by checking |
| 442 | ** if the UTF-16 BOM is not immediately followed by (utf16) 0. |
| 443 | ** pnByte is only set when the function returns 1. |
| 444 | ** |
| 445 | ** pbReverse is always set, even when no BOM is found. Without BOM, |
| 446 | ** it is set to 1 on little-endian and 0 on big-endian platforms. See |
| 447 | ** clause D98 of conformance (section 3.10) of the Unicode standard. |
| 448 | */ |
| 449 | int starts_with_utf16_bom( |
| 450 | const Blob *pContent, /* IN: Blob content to perform BOM detection on. */ |
| 451 | int *pnByte, /* OUT: The number of bytes used for the BOM. */ |
| 452 | int *pbReverse /* OUT: Non-zero for BOM in reverse byte-order. */ |
| 453 | ){ |
| 454 | const unsigned short *z = (unsigned short *)blob_buffer(pContent); |
| 455 | int bomSize = sizeof(unsigned short); |
| 456 | int size = blob_size(pContent); |
| 457 | static const int one = 1; |
| 458 | |
| 459 | if( size<bomSize ) goto noBom; /* No: cannot read BOM. */ |
| 460 | if( size>=(2*bomSize) && z[1]==0 ) goto noBom; /* No: possible UTF-32. */ |
| 461 | if( z[0]==0xfeff ){ |
| 462 | if( pbReverse ) *pbReverse = 0; |
| 463 | }else if( z[0]==0xfffe ){ |
| 464 | if( pbReverse ) *pbReverse = 1; |
| 465 | }else{ |
| 466 | noBom: |
| 467 | if( pbReverse ) *pbReverse = *(char *) &one; |
| 468 | return 0; /* No: UTF-16 byte-order-mark not found. */ |
| 469 | } |
| 470 | if( pnByte ) *pnByte = bomSize; |
| 471 | return 1; /* Yes. */ |
| 472 | } |
| 473 |
-3
| --- src/main.c | ||
| +++ src/main.c | ||
| @@ -513,11 +513,10 @@ | ||
| 513 | 513 | #endif |
| 514 | 514 | { |
| 515 | 515 | const char *zCmdName = "unknown"; |
| 516 | 516 | int idx; |
| 517 | 517 | int rc; |
| 518 | - int timerId; | |
| 519 | 518 | sqlite3_config(SQLITE_CONFIG_LOG, fossil_sqlite_log, 0); |
| 520 | 519 | memset(&g, 0, sizeof(g)); |
| 521 | 520 | g.now = time(0); |
| 522 | 521 | #ifdef FOSSIL_ENABLE_JSON |
| 523 | 522 | #if defined(NDEBUG) |
| @@ -536,13 +535,11 @@ | ||
| 536 | 535 | #ifdef FOSSIL_ENABLE_TCL |
| 537 | 536 | memset(&g.tcl, 0, sizeof(TclContext)); |
| 538 | 537 | g.tcl.argc = g.argc; |
| 539 | 538 | g.tcl.argv = copy_args(g.argc, g.argv); /* save full arguments */ |
| 540 | 539 | #endif |
| 541 | - | |
| 542 | 540 | g.mainTimerId = fossil_timer_start(); |
| 543 | - | |
| 544 | 541 | if( fossil_getenv("GATEWAY_INTERFACE")!=0 && !find_option("nocgi", 0, 0)){ |
| 545 | 542 | zCmdName = "cgi"; |
| 546 | 543 | g.isHTTP = 1; |
| 547 | 544 | }else if( g.argc<2 ){ |
| 548 | 545 | fossil_print( |
| 549 | 546 |
| --- src/main.c | |
| +++ src/main.c | |
| @@ -513,11 +513,10 @@ | |
| 513 | #endif |
| 514 | { |
| 515 | const char *zCmdName = "unknown"; |
| 516 | int idx; |
| 517 | int rc; |
| 518 | int timerId; |
| 519 | sqlite3_config(SQLITE_CONFIG_LOG, fossil_sqlite_log, 0); |
| 520 | memset(&g, 0, sizeof(g)); |
| 521 | g.now = time(0); |
| 522 | #ifdef FOSSIL_ENABLE_JSON |
| 523 | #if defined(NDEBUG) |
| @@ -536,13 +535,11 @@ | |
| 536 | #ifdef FOSSIL_ENABLE_TCL |
| 537 | memset(&g.tcl, 0, sizeof(TclContext)); |
| 538 | g.tcl.argc = g.argc; |
| 539 | g.tcl.argv = copy_args(g.argc, g.argv); /* save full arguments */ |
| 540 | #endif |
| 541 | |
| 542 | g.mainTimerId = fossil_timer_start(); |
| 543 | |
| 544 | if( fossil_getenv("GATEWAY_INTERFACE")!=0 && !find_option("nocgi", 0, 0)){ |
| 545 | zCmdName = "cgi"; |
| 546 | g.isHTTP = 1; |
| 547 | }else if( g.argc<2 ){ |
| 548 | fossil_print( |
| 549 |
| --- src/main.c | |
| +++ src/main.c | |
| @@ -513,11 +513,10 @@ | |
| 513 | #endif |
| 514 | { |
| 515 | const char *zCmdName = "unknown"; |
| 516 | int idx; |
| 517 | int rc; |
| 518 | sqlite3_config(SQLITE_CONFIG_LOG, fossil_sqlite_log, 0); |
| 519 | memset(&g, 0, sizeof(g)); |
| 520 | g.now = time(0); |
| 521 | #ifdef FOSSIL_ENABLE_JSON |
| 522 | #if defined(NDEBUG) |
| @@ -536,13 +535,11 @@ | |
| 535 | #ifdef FOSSIL_ENABLE_TCL |
| 536 | memset(&g.tcl, 0, sizeof(TclContext)); |
| 537 | g.tcl.argc = g.argc; |
| 538 | g.tcl.argv = copy_args(g.argc, g.argv); /* save full arguments */ |
| 539 | #endif |
| 540 | g.mainTimerId = fossil_timer_start(); |
| 541 | if( fossil_getenv("GATEWAY_INTERFACE")!=0 && !find_option("nocgi", 0, 0)){ |
| 542 | zCmdName = "cgi"; |
| 543 | g.isHTTP = 1; |
| 544 | }else if( g.argc<2 ){ |
| 545 | fossil_print( |
| 546 |