1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703
|
/*++
Module Name:
ReadWriter.cpp
Abstract:
General file writer.
Environment:
User mode service.
Not thread safe.
--*/
#include "stdafx.h"
#include "BigAlloc.h"
#include "Compat.h"
#include "Read.h"
#include "SAM.h"
#include "Tables.h"
#include "RangeSplitter.h"
#include "ParallelTask.h"
#include "Util.h"
#include "ReadSupplierQueue.h"
#include "FileFormat.h"
#include "exit.h"
#include "Error.h"
#include "Genome.h"
class SimpleReadWriter : public ReadWriter
{
public:
SimpleReadWriter(const FileFormat* i_format, DataWriter* i_writer, const Genome* i_genome, bool i_killIfTooSlow, bool i_emitInternalScore, char *i_internalScoreTag, bool i_ignoreAlignmentAdjustmentsForOm)
: format(i_format), writer(i_writer), genome(i_genome), killIfTooSlow(i_killIfTooSlow), lastTooSlowCheck(0), emitInternalScore(i_emitInternalScore), ignoreAlignmentAdjustmentsForOm(i_ignoreAlignmentAdjustmentsForOm)
{
if (emitInternalScore) {
if (strlen(i_internalScoreTag) != 2) {
WriteErrorMessage("SimpleReadWriter: bogus internal score tag\n");
soft_exit(1);
}
strcpy(internalScoreTag, i_internalScoreTag);
} else {
internalScoreTag[0] = '\0';
}
}
virtual ~SimpleReadWriter()
{
delete writer;
}
virtual bool writeHeader(const ReaderContext& context, bool sorted, int argc, const char **argv, const char *version, const char *rgLine, bool omitSQLines);
virtual bool writeReads(const ReaderContext& context, Read *read, SingleAlignmentResult *results, _int64 nResults, bool firstIsPrimary, bool useAffineGap = false);
virtual bool writePairs(const ReaderContext& context, Read **reads /* array of size 2 */, PairedAlignmentResult *result, _int64 nResults,
SingleAlignmentResult **singleResults /* array of size 2*/, _int64 *nSingleResults /* array of size 2*/, bool firstIsPrimary, bool useAffineGap = false);
virtual void close();
private:
const FileFormat* format;
DataWriter* writer;
const Genome* genome;
LandauVishkinWithCigar lvc;
AffineGapVectorizedWithCigar agc;
void checkIfTooSlow();
bool killIfTooSlow;
_int64 lastTooSlowCheck;
_int64 writesSinceLastTooSlowCheck;
bool emitInternalScore;
char internalScoreTag[3];
bool ignoreAlignmentAdjustmentsForOm;
};
bool
SimpleReadWriter::writeHeader(
const ReaderContext& context,
bool sorted,
int argc,
const char **argv,
const char *version,
const char *rgLine,
bool omitSQLines)
{
char* buffer;
size_t size;
size_t used;
char *localBuffer = NULL;
writer->inHeader(true);
if (! writer->getBuffer(&buffer, &size)) {
return false;
}
char *writerBuffer = buffer;
size_t writerBufferSize = size;
while (!format->writeHeader(context, buffer, size, &used, sorted, argc, argv, version, rgLine, omitSQLines)) {
delete[] localBuffer;
size = 2 * size;
localBuffer = new char[size];
buffer = localBuffer;
}
if (NULL == localBuffer) {
_ASSERT(writerBuffer == buffer);
writer->advance((unsigned)used, 0);
writer->nextBatch();
} else {
size_t bytesRemainingToWrite = used;
size_t bytesWritten = 0;
while (bytesRemainingToWrite > 0) {
size_t bytesToWrite = __min(bytesRemainingToWrite, writerBufferSize);
memcpy(writerBuffer, localBuffer + bytesWritten, bytesToWrite);
writer->advance(bytesToWrite);
writer->nextBatch();
if (!writer->getBuffer(&writerBuffer, &writerBufferSize)) {
return false;
}
bytesWritten += bytesToWrite;
bytesRemainingToWrite -= bytesToWrite;
}
delete[] localBuffer;
}
writer->inHeader(false);
return true;
}
void
SimpleReadWriter::checkIfTooSlow()
{
const _int64 tooSlowCheckPeriod = 5 * 60 * 1000; // 5 min in ms
const _int64 tooSlowCheckMinReadsPerCheckPeriod = 5 * 60 * 1000; // One read/ms (or 1000 reads/s, but just on this thread).
if (killIfTooSlow) {
_int64 now = timeInMillis();
if (lastTooSlowCheck + tooSlowCheckPeriod <= now) {
if (lastTooSlowCheck != 0 && writesSinceLastTooSlowCheck < tooSlowCheckMinReadsPerCheckPeriod) {
WriteErrorMessage("Only wrote %lld writes during a %lld minute check period; we're probably out of memory and are giving up because of -kts\n", writesSinceLastTooSlowCheck, tooSlowCheckPeriod / (60 * 1000));
soft_exit(1);
}
lastTooSlowCheck = now;
writesSinceLastTooSlowCheck = 0;
}
writesSinceLastTooSlowCheck++;
} // if (killIfTooSlow)
}
bool
SimpleReadWriter::writeReads(
const ReaderContext& context,
Read *read,
SingleAlignmentResult *results,
_int64 nResults,
bool firstIsPrimary,
bool useAffineGap)
{
char* buffer;
size_t size;
size_t used;
bool result = false;
checkIfTooSlow();
for (_int64 i = 0; i < nResults; i++) {
if (results[i].status == NotFound) {
results[i].location = InvalidGenomeLocation;
}
}
//
// We need to keep track of the offsets of all of the alignments in the output buffer so we can commit them. However,
// we want to avoid dynamic memory allocation as much as possible. So, we have a static buffer on the stack that's big enough
// for the great majority of cases, and then allocate dynamically if that's too small. Makes for annoying, but efficient
// code.
//
const int staticUsedBufferSize = 2000;
size_t staticUsedBuffer[staticUsedBufferSize];
GenomeLocation staticFinalLocationsBuffer[staticUsedBufferSize];
size_t *usedBuffer;
GenomeLocation *finalLocations;
if (nResults <= staticUsedBufferSize) {
usedBuffer = staticUsedBuffer;
finalLocations = staticFinalLocationsBuffer;
} else {
usedBuffer = new size_t[nResults];
finalLocations = new GenomeLocation[nResults];
}
for (int pass = 0; pass < 2; pass++) { // Make two passes, one with whatever buffer space is left and one with a clean buffer.
bool blewBuffer = false;
if (!writer->getBuffer(&buffer, &size)) {
goto done;
}
used = 0;
for (_int64 whichResult = 0; whichResult < nResults; whichResult++) {
int addFrontClipping = 0;
read->setAdditionalFrontClipping(results[whichResult].clippingForReadAdjustment);
int cumulativeAddFrontClipping = 0, cumulativeAddBackClipping = 0;
finalLocations[whichResult] = results[whichResult].location;
unsigned nAdjustments = 0;
if (useAffineGap && (results[whichResult].usedAffineGapScoring || results[whichResult].score > 0)) {
while (!format->writeRead(context, &agc, buffer + used, size - used, &usedBuffer[whichResult], read->getIdLength(), read, results[whichResult].status,
results[whichResult].mapq, finalLocations[whichResult], results[whichResult].direction, (whichResult > 0) || !firstIsPrimary, results[whichResult].supplementary, &addFrontClipping,
results[whichResult].scorePriorToClipping, emitInternalScore, internalScoreTag, results[whichResult].basesClippedBefore,
results[whichResult].basesClippedAfter)) {
_ASSERT(0 == addFrontClipping || ignoreAlignmentAdjustmentsForOm); // Because of the alignment adjuster.
nAdjustments++;
if (0 == addFrontClipping) {
blewBuffer = true;
break;
}
// redo if read modified (e.g. to add soft clipping, or move alignment for a leading I.
const Genome::Contig *originalContig = results[whichResult].status == NotFound ? NULL
: genome->getContigAtLocation(results[whichResult].location);
const Genome::Contig *newContig = results[whichResult].status == NotFound ? NULL
: genome->getContigAtLocation(results[whichResult].location + addFrontClipping);
if (newContig == NULL || newContig != originalContig || finalLocations[whichResult] + addFrontClipping > originalContig->beginningLocation + originalContig->length - genome->getChromosomePadding() ||
nAdjustments > read->getDataLength()) {
//
// Altering this would push us over a contig boundary, or we're stuck in a loop. Just give up on the read.
//
results[whichResult].status = NotFound;
results[whichResult].location = InvalidGenomeLocation;
finalLocations[whichResult] = InvalidGenomeLocation;
} else {
if (addFrontClipping < 0) { // Insertion (soft-clip)
cumulativeAddFrontClipping += addFrontClipping;
if (results[whichResult].direction == FORWARD) {
read->setAdditionalFrontClipping(-cumulativeAddFrontClipping);
}
else {
read->setAdditionalBackClipping(-cumulativeAddFrontClipping);
}
}
else { // Deletion
finalLocations[whichResult] = results[whichResult].location + addFrontClipping;
}
}
} // while formatting doesn't work
} else {
while (!format->writeRead(context, &lvc, buffer + used, size - used, &usedBuffer[whichResult], read->getIdLength(), read, results[whichResult].status,
results[whichResult].mapq, finalLocations[whichResult], results[whichResult].direction, (whichResult > 0) || !firstIsPrimary, results[whichResult].supplementary, &addFrontClipping,
results[whichResult].scorePriorToClipping, emitInternalScore, internalScoreTag)) {
_ASSERT(0 == addFrontClipping || ignoreAlignmentAdjustmentsForOm); // Because of the alignment adjuster.
nAdjustments++;
if (0 == addFrontClipping) {
blewBuffer = true;
break;
}
// redo if read modified (e.g. to add soft clipping, or move alignment for a leading I.
const Genome::Contig *originalContig = results[whichResult].status == NotFound ? NULL
: genome->getContigAtLocation(results[whichResult].location);
const Genome::Contig *newContig = results[whichResult].status == NotFound ? NULL
: genome->getContigAtLocation(results[whichResult].location + addFrontClipping);
if (newContig == NULL || newContig != originalContig || finalLocations[whichResult] + addFrontClipping > originalContig->beginningLocation + originalContig->length - genome->getChromosomePadding() ||
nAdjustments > read->getDataLength()) {
//
// Altering this would push us over a contig boundary, or we're stuck in a loop. Just give up on the read.
//
results[whichResult].status = NotFound;
results[whichResult].location = InvalidGenomeLocation;
finalLocations[whichResult] = InvalidGenomeLocation;
}
else {
if (addFrontClipping > 0) {
cumulativeAddFrontClipping += addFrontClipping;
read->setAdditionalFrontClipping(cumulativeAddFrontClipping);
}
finalLocations[whichResult] += addFrontClipping;
}
} // while formatting doesn't work
}
if (blewBuffer) {
break;
}
used += usedBuffer[whichResult];
_ASSERT(used <= size);
if (used > 0xffffffff) {
WriteErrorMessage("SimpleReadWriter:writeReads: used too big\n");
soft_exit(1);
}
} // for each result.
if (!blewBuffer) {
//
// Everything worked OK.
//
for (int whichResult = 0; whichResult < nResults; whichResult++) {
writer->advance((unsigned)usedBuffer[whichResult], finalLocations[whichResult]);
}
result = true;
goto done;
}
if (pass == 1) {
WriteErrorMessage("Failed to write into fresh buffer; trying providing the -wbs switch with a larger value\n");
soft_exit(1);
}
if (!writer->nextBatch()) {
goto done;
}
} // for each pass (i.e., not empty, empty buffer)
done:
if (usedBuffer != staticUsedBuffer) {
delete[] usedBuffer;
usedBuffer = NULL;
delete[] finalLocations;
finalLocations = NULL;
}
read->setAdditionalFrontClipping(0);
return result;
}
bool
SimpleReadWriter::writePairs(
const ReaderContext& context,
Read **reads /* array of size NUM_READS_PER_PAIR */,
PairedAlignmentResult *result,
_int64 nResults,
SingleAlignmentResult **singleResults /* array of size NUM_READS_PER_PAIR*/,
_int64 *nSingleResults /* array of size NUM_READS_PER_PAIR*/,
bool firstIsPrimary,
bool useAffineGap)
{
bool retVal = false;
//
// We need to write all alignments for the pair into the same buffer, so that a write from
// some other thread doesn't separate them. We make two passes, trying to write into the
// existing buffer, and then into a clean one. If that doesn't work, abort the alignment
// run and ask for a bigger write buffer.
//
const int staticUsedBufferSize = 2000;
size_t staticUsedBuffer[NUM_READS_PER_PAIR][staticUsedBufferSize];
GenomeLocation staticLocationBuffer[NUM_READS_PER_PAIR][staticUsedBufferSize];
_int64 NullNSingleResults[NUM_READS_PER_PAIR] = { 0,0 };
if (NULL == nSingleResults) {
nSingleResults = NullNSingleResults;
}
checkIfTooSlow();
GenomeLocation *finalLocations[NUM_READS_PER_PAIR];
size_t *usedBuffer[NUM_READS_PER_PAIR];
if (nResults + nSingleResults[0] <= staticUsedBufferSize && nResults + nSingleResults[1] <= staticUsedBufferSize) {
usedBuffer[0] = staticUsedBuffer[0];
usedBuffer[1] = staticUsedBuffer[1];
finalLocations[0] = staticLocationBuffer[0];
finalLocations[1] = staticLocationBuffer[1];
} else {
usedBuffer[0] = new size_t[nResults * NUM_READS_PER_PAIR + nSingleResults[0] + nSingleResults[1]];
usedBuffer[1] = usedBuffer[0] + nResults + nSingleResults[0];
finalLocations[0] = new GenomeLocation[nResults * NUM_READS_PER_PAIR + nSingleResults[0] + nSingleResults[1]];
finalLocations[1] = finalLocations[0] + nResults + nSingleResults[0];
}
//
// For paired reads, we need to have the same QNAME for both of them, and it needs to be unique among all other
// reads in the dataset. For now, all we do is see if the read names end in /1 and /2, and if so truncate them.
//
size_t idLengths[NUM_READS_PER_PAIR];
idLengths[0] = reads[0]->getIdLength();
idLengths[1] = reads[1]->getIdLength();
if (idLengths[0] == idLengths[1] && idLengths[0] > 2 && reads[0]->getId()[idLengths[0]-2] == '/' && reads[1]->getId()[idLengths[0]-2] == '/') {
char lastChar0, lastChar1;
lastChar0 = reads[0]->getId()[idLengths[0] - 1];
lastChar1 = reads[1]->getId()[idLengths[1] - 1];
if ((lastChar0 == '1' || lastChar0 == '2') && (lastChar1 == '1' || lastChar1 == '2') &&
lastChar0 != lastChar1) {
idLengths[0] -= 2;
idLengths[1] -= 2;
}
}
for (int pass = 0; pass < 2; pass++) {
char* buffer;
size_t size;
size_t used = 0;
bool fitInBuffer = true;
if (!writer->getBuffer(&buffer, &size)) {
goto done;
}
//
// Write all of the pair alignments into the buffer.
//
for (int whichAlignmentPair = 0; whichAlignmentPair < nResults; whichAlignmentPair++) {
reads[0]->setAdditionalFrontClipping(result[whichAlignmentPair].clippingForReadAdjustment[0]);
reads[1]->setAdditionalFrontClipping(result[whichAlignmentPair].clippingForReadAdjustment[1]);
GenomeLocation locations[2];
locations[0] = result[whichAlignmentPair].status[0] != NotFound ? result[whichAlignmentPair].location[0] : InvalidGenomeLocation;
locations[1] = result[whichAlignmentPair].status[1] != NotFound ? result[whichAlignmentPair].location[1] : InvalidGenomeLocation;
bool secondReadLocationChanged;
int cumulativePositiveAddFrontClipping[NUM_READS_PER_PAIR] = { 0, 0 };
bool writeOrderChanged;
do {
size_t tentativeUsed = 0;
secondReadLocationChanged = false;
writeOrderChanged = false;
int writeOrder[2]; // The order in which we write the reads, which is just numerical by genome location. SO writeOrder[0] gets written first, and writeOrder[1] second.
if (locations[0] <= locations[1]) {
writeOrder[0] = 0;
writeOrder[1] = 1;
} else {
writeOrder[0] = 1;
writeOrder[1] = 0;
}
size_t spaceUsed[2] = { usedBuffer[0][whichAlignmentPair], usedBuffer[1][whichAlignmentPair] };
bool outOfSpace = false;
if (!format->writePairs(context, &lvc, &agc, useAffineGap, buffer + used + tentativeUsed, size - used - tentativeUsed,
spaceUsed, idLengths, reads, locations, &result[whichAlignmentPair], whichAlignmentPair != 0 || !firstIsPrimary,
emitInternalScore, internalScoreTag, writeOrder, cumulativePositiveAddFrontClipping, &secondReadLocationChanged, &outOfSpace)) {
if (outOfSpace) {
//
// We failed because we ran out of buffer.
//
goto blownBuffer;
}
}
// Check if the write order is correct, if not redo
int newWriteOrder[2];
if (locations[0] <= locations[1]) {
newWriteOrder[0] = 0;
newWriteOrder[1] = 1;
} else {
newWriteOrder[0] = 1;
newWriteOrder[1] = 0;
}
if (writeOrder[0] != newWriteOrder[0] || writeOrder[1] != newWriteOrder[1]) {
writeOrder[0] = newWriteOrder[0];
writeOrder[1] = newWriteOrder[1];
writeOrderChanged = true;
}
usedBuffer[0][whichAlignmentPair] = spaceUsed[0];
usedBuffer[1][whichAlignmentPair] = spaceUsed[1];
} while (secondReadLocationChanged || writeOrderChanged);
used += usedBuffer[0][whichAlignmentPair] + usedBuffer[1][whichAlignmentPair];
//
// Both reads are written into the buffer. Save the final locations we used for when we commit.
//
for (int whichRead = 0; whichRead < NUM_READS_PER_PAIR; whichRead++) {
finalLocations[whichRead][whichAlignmentPair] = locations[whichRead];
}
} // for each pair.
//
// Now write the single alignments.
//
for (int whichRead = 0; whichRead < NUM_READS_PER_PAIR; whichRead++) {
for (int whichAlignment = 0; whichAlignment < nSingleResults[whichRead]; whichAlignment++) {
int addFrontClipping;
reads[whichRead]->setAdditionalFrontClipping(singleResults[whichRead]->clippingForReadAdjustment);
GenomeLocation location = singleResults[whichRead][whichAlignment].status != NotFound ? singleResults[whichRead][whichAlignment].location : InvalidGenomeLocation;
int cumulativePositiveAddFrontClipping = 0;
if (useAffineGap && (singleResults[whichRead][whichAlignment].usedAffineGapScoring || singleResults[whichRead][whichAlignment].score > 0)) {
while (!format->writeRead(context, &agc, buffer + used, size - used, &usedBuffer[whichRead][nResults + whichAlignment], reads[whichRead]->getIdLength(),
reads[whichRead], singleResults[whichRead][whichAlignment].status, singleResults[whichRead][whichAlignment].mapq, location, singleResults[whichRead][whichAlignment].direction,
true, singleResults[whichRead][whichAlignment].supplementary, &addFrontClipping, singleResults[whichRead][whichAlignment].scorePriorToClipping, emitInternalScore, internalScoreTag,
singleResults[whichRead][whichAlignment].basesClippedBefore, singleResults[whichRead][whichAlignment].basesClippedAfter)) {
if (0 == addFrontClipping) {
goto blownBuffer;
}
const Genome::Contig *originalContig = genome->getContigAtLocation(location);
const Genome::Contig *newContig = genome->getContigAtLocation(location + addFrontClipping);
if (newContig != originalContig || NULL == newContig || location + addFrontClipping > originalContig->beginningLocation + originalContig->length - genome->getChromosomePadding()) {
//
// Altering this would push us over a contig boundary. Just give up on the read.
//
singleResults[whichRead][whichAlignment].status = NotFound;
location = InvalidGenomeLocation;
}
else {
if (addFrontClipping < 0) { // Insertion (soft-clip)
cumulativePositiveAddFrontClipping += addFrontClipping;
if (singleResults[whichRead][whichAlignment].direction == FORWARD) {
reads[whichRead]->setAdditionalFrontClipping(-cumulativePositiveAddFrontClipping);
}
else {
reads[whichRead]->setAdditionalBackClipping(-cumulativePositiveAddFrontClipping);
}
}
else { // Deletion
location += addFrontClipping;
}
}
}
} else {
while (!format->writeRead(context, &lvc, buffer + used, size - used, &usedBuffer[whichRead][nResults + whichAlignment], reads[whichRead]->getIdLength(),
reads[whichRead], singleResults[whichRead][whichAlignment].status, singleResults[whichRead][whichAlignment].mapq, location, singleResults[whichRead][whichAlignment].direction,
true, singleResults[whichRead][whichAlignment].supplementary, &addFrontClipping, singleResults[whichRead][whichAlignment].scorePriorToClipping, emitInternalScore, internalScoreTag)) {
if (0 == addFrontClipping) {
goto blownBuffer;
}
const Genome::Contig *originalContig = genome->getContigAtLocation(location);
const Genome::Contig *newContig = genome->getContigAtLocation(location + addFrontClipping);
if (newContig != originalContig || NULL == newContig || location + addFrontClipping > originalContig->beginningLocation + originalContig->length - genome->getChromosomePadding()) {
//
// Altering this would push us over a contig boundary. Just give up on the read.
//
singleResults[whichRead][whichAlignment].status = NotFound;
location = InvalidGenomeLocation;
}
else {
if (addFrontClipping > 0) {
cumulativePositiveAddFrontClipping += addFrontClipping;
reads[whichRead]->setAdditionalFrontClipping(cumulativePositiveAddFrontClipping);
}
location += addFrontClipping;
}
}
}
finalLocations[whichRead][nResults + whichAlignment] = location;
used += usedBuffer[whichRead][nResults + whichAlignment];
} // For each single alignment of a read
} // For each read
//
// They all fit into the buffer.
//
//
// Commit the updates for the pairs.
//
for (int whichReadPair = 0; whichReadPair < nResults; whichReadPair++) {
for (int firstOrSecond = 0; firstOrSecond < NUM_READS_PER_PAIR; firstOrSecond++) {
// adjust for write order
int writeFirstOrSecond = (!!firstOrSecond) ^ (finalLocations[0][whichReadPair] > finalLocations[1][whichReadPair]); // goofy looking !! converts int to bool
writer->advance((unsigned)usedBuffer[firstOrSecond][whichReadPair],
finalLocations[writeFirstOrSecond][whichReadPair] == InvalidGenomeLocation ? finalLocations[1 - writeFirstOrSecond][whichReadPair] : finalLocations[writeFirstOrSecond][whichReadPair]);
}
}
//
// Now commit the updates for the single reads.
//
for (int whichRead = 0; whichRead < NUM_READS_PER_PAIR; whichRead++) {
for (int whichAlignment = 0; whichAlignment < nSingleResults[whichRead]; whichAlignment++) {
writer->advance((unsigned)usedBuffer[whichRead][nResults + whichAlignment], finalLocations[whichRead][nResults + whichAlignment]);
}
}
retVal = true;
break;
blownBuffer:
if (pass > 0) {
WriteErrorMessage("Unable to fit all alignments for one read pair into a single write buffer. Increase the size of the write buffer with -wbs, or reduce the number of alignments with -om or -omax\n");
WriteErrorMessage("Read id: '%.*s'\n", reads[0]->getIdLength(), reads[0]->getId());
soft_exit(1);
}
if (!writer->nextBatch()) {
goto done;
}
} // For each buffer full pass
done:
if (usedBuffer[0] != staticUsedBuffer[0]) {
delete[] usedBuffer[0];
usedBuffer[0] = usedBuffer[1] = NULL;
delete[] finalLocations[0];
finalLocations[0] = finalLocations[1] = NULL;
}
reads[0]->setAdditionalFrontClipping(0);
reads[1]->setAdditionalFrontClipping(0);
return retVal;
}
void
SimpleReadWriter::close()
{
writer->close();
}
class SimpleReadWriterSupplier : public ReadWriterSupplier
{
public:
SimpleReadWriterSupplier(const FileFormat* i_format, DataWriterSupplier* i_dataSupplier, const Genome* i_genome, bool i_killIfTooSlow, bool i_emitInternalScore, char *i_internalScoreTag, bool i_ignoreAlignmentAdjustmentsForOm)
:
format(i_format),
dataSupplier(i_dataSupplier),
genome(i_genome),
killIfTooSlow(i_killIfTooSlow),
emitInternalScore(i_emitInternalScore),
ignoreAlignmentAdjustmentsForOm(i_ignoreAlignmentAdjustmentsForOm)
{
if (emitInternalScore) {
if (strlen(i_internalScoreTag) != 2) {
WriteErrorMessage("SimpleReadWriterSupplier: bad internal score tag\n");
soft_exit(1);
}
strcpy(internalScoreTag, i_internalScoreTag);
} else {
internalScoreTag[0] = '\0';
}
}
~SimpleReadWriterSupplier()
{
delete dataSupplier;
}
virtual ReadWriter* getWriter()
{
return new SimpleReadWriter(format, dataSupplier->getWriter(), genome, killIfTooSlow, emitInternalScore, internalScoreTag, ignoreAlignmentAdjustmentsForOm);
}
virtual void close()
{
dataSupplier->close();
}
private:
const FileFormat* format;
DataWriterSupplier* dataSupplier;
const Genome* genome;
bool killIfTooSlow;
bool emitInternalScore;
char internalScoreTag[3];
bool ignoreAlignmentAdjustmentsForOm;
};
ReadWriterSupplier*
ReadWriterSupplier::create(
const FileFormat* format,
DataWriterSupplier* dataSupplier,
const Genome* genome,
bool killIfTooSlow,
bool emitInternalScore,
char *internalScoreTag,
bool ignoreAlignmentAdjustmentsForOm)
{
return new SimpleReadWriterSupplier(format, dataSupplier, genome, killIfTooSlow, emitInternalScore, internalScoreTag, ignoreAlignmentAdjustmentsForOm);
}
|