rc_t SRAFastqFile_Open(const KFile** cself, const SRAListNode* sra, const FileOptions* opt)
{
rc_t rc = 0;
SRAFastqFile* self;
CALLOC( self, 1, sizeof( *self ) );
if( self == NULL )
{
rc = RC( rcExe, rcFile, rcConstructing, rcMemory, rcExhausted );
}
else
{
if ( ( rc = KFileInit( &self->dad, (const KFile_vt*)&SRAFastqFile_vtbl, "SRAFastqFile", "no-name", true, false ) ) == 0 )
{
if ( ( rc = SRAListNode_TableOpen( sra, &self->stbl ) ) == 0 )
{
if ( ( rc = SRATableGetKTableRead( self->stbl, &self->ktbl ) ) == 0 )
{
if ( ( rc = KTableOpenIndexRead( self->ktbl, &self->kidx, opt->index ) ) == 0 )
{
if ( ( rc = KLockMake( &self->lock ) ) == 0 )
{
MALLOC( self->buf, opt->buffer_sz * ( opt->f.fastq.gzip ? 2 : 1 ) );
if ( self->buf == NULL )
{
rc = RC( rcExe, rcFile, rcOpening, rcMemory, rcExhausted );
}
else
{
self->file_sz = opt->file_sz;
self->buffer_sz = opt->buffer_sz;
if ( opt->f.fastq.gzip )
{
self->gzipped = &self->buf[ opt->buffer_sz ];
}
self->from = ~0; /* reset position beyond file end */
rc = FastqReaderMake( &self->reader, self->stbl,
opt->f.fastq.accession, opt->f.fastq.colorSpace,
opt->f.fastq.origFormat, false, opt->f.fastq.printLabel,
opt->f.fastq.printReadId, !opt->f.fastq.clipQuality, false,
opt->f.fastq.minReadLen, opt->f.fastq.qualityOffset,
opt->f.fastq.colorSpaceKey,
opt->f.fastq.minSpotId, opt->f.fastq.maxSpotId );
}
}
}
}
}
if ( rc == 0 )
{
*cself = &self->dad;
}
else
{
KFileRelease( &self->dad );
}
}
}
return rc;
}
static
rc_t FastqGzip_Idx(const SRATable* sratbl, SIndexObj* obj, char* buffer, const size_t buffer_sz)
{
rc_t rc = 0;
const FastqReader* reader = NULL;
uint16_t zlib_ver = ZLIB_VERNUM;
uint8_t colorSpace = false;
char* colorSpaceKey = "\0";
uint8_t origFormat = false;
uint8_t printLabel = true;
uint8_t printReadId = true;
uint8_t clipQuality = true;
uint32_t minReadLen = 0;
uint16_t qualityOffset = 0;
{{
const SRAColumn* c = NULL;
const uint8_t *platform = SRA_PLATFORM_UNDEFINED;
bitsz_t o, z;
if( (rc = SRATableOpenColumnRead(sratbl, &c, "PLATFORM", sra_platform_id_t)) != 0 ) {
return rc;
}
if( (rc = SRAColumnRead(c, 1, (const void **)&platform, &o, &z)) != 0 ) {
return rc;
}
if( *platform == SRA_PLATFORM_ABSOLID ) {
colorSpace = true;
}
SRAColumnRelease(c);
}}
if( (rc = FastqReaderMake(&reader, sratbl, g_accession,
colorSpace, origFormat, false, printLabel, printReadId,
!clipQuality, minReadLen, qualityOffset, colorSpaceKey[0],
obj->minSpotId, obj->maxSpotId)) != 0 ) {
return rc;
} else {
size_t written = 0;
uint32_t blk = 0, spots_per_block = 0, proj_id_qty = 0;
SIndexNode* inode = NULL;
size_t z_blk = 0;
size_t spots_buf_sz = g_file_block_sz * 100;
size_t zbuf_sz = spots_buf_sz + 100;
char* zbuf = malloc(zbuf_sz);
char* spots_buf = malloc(spots_buf_sz);
bool eof = false;
if( zbuf == NULL || spots_buf == NULL ) {
rc = RC(rcExe, rcIndex, rcConstructing, rcMemory, rcExhausted);
}
while( rc == 0 ) {
if( (rc = FastqReader_GetNextSpotSplitData(reader, buffer, buffer_sz, &written)) == 0 ) {
if( inode == NULL ) {
spotid_t spotid = 0;
if( (rc = FastqReaderCurrentSpot(reader, &spotid)) != 0 ) {
break;
}
inode = malloc(sizeof(SIndexNode));
if( inode == NULL ) {
rc = RC(rcExe, rcIndex, rcConstructing, rcMemory, rcExhausted);
break;
}
inode->key = obj->file_size;
inode->key_size = 0;
inode->id = spotid;
inode->id_qty = 0;
DEBUG_MSG(5, ("%s open key: spot %ld, offset %lu\n", obj->index, inode->id, inode->key));
}
if( blk + written > spots_buf_sz ) {
rc = RC(rcExe, rcIndex, rcConstructing, rcMemory, rcInsufficient);
break;
}
inode->id_qty++;
memmove(&spots_buf[blk], buffer, written);
blk += written;
if( g_dump ) {
fwrite(buffer, written, 1, stderr);
}
}
if( (eof = (GetRCObject(rc) == rcRow && GetRCState(rc) == rcExhausted)) ) {
rc = 0;
if( inode == NULL ) {
break;
}
}
if( rc == 0 && (eof ||
(proj_id_qty == 0 && inode->id_qty > (spots_per_block * 0.95)) ||
(proj_id_qty > 0 && inode->id_qty >= proj_id_qty) ) ) {
rc = ZLib_DeflateBlock(spots_buf, blk, zbuf, zbuf_sz, &z_blk);
if( z_blk < g_file_block_sz ) {
/* project needed id_qty */
proj_id_qty = g_file_block_sz * inode->id_qty / z_blk * 1.05;
DEBUG_MSG(5, ("%s: project id qty %u\n", obj->index, proj_id_qty));
} else {
DEBUG_MSG(10, ("%s: no projection %u > %u\n", obj->index, z_blk, g_file_block_sz));
}
}
if( rc == 0 && (eof || z_blk >= g_file_block_sz) ) {
obj->file_size += z_blk;
MD5StateAppend(&obj->md5, zbuf, z_blk);
inode->key_size = z_blk;
SLListPushTail(&obj->li, &inode->n);
DEBUG_MSG(5, ("%s close key: spots %lu, size %lu, ratio %hu%%, raw %u\n",
obj->index, inode->id_qty, inode->key_size, (uint16_t)(((float)(blk - z_blk)/blk)*100), blk ));
spots_per_block = inode->id_qty;
inode = NULL;
if( blk > obj->buffer_sz ) {
obj->buffer_sz = blk;
}
blk = 0;
z_blk = 0;
proj_id_qty = 0;
}
if( eof ) {
break;
}
}
rc = rc ? rc : Quitting();
if( rc != 0 ) {
spotid_t spot = 0;
FastqReaderCurrentSpot(reader, &spot);
PLOGERR(klogErr, (klogErr, rc, "spot $(s)", PLOG_U32(s), spot));
}
free(zbuf);
free(spots_buf);
}
if( rc == 0 ) {
KMDataNode* opt = NULL, *nd = NULL;
if( (rc = KMDataNodeOpenNodeUpdate(obj->meta, &opt, "Format/Options")) != 0 ) {
return rc;
}
if( rc == 0 && (rc = KMDataNodeOpenNodeUpdate(opt, &nd, "ZlibVersion")) == 0 ) {
rc = KMDataNodeWriteB16(nd, &zlib_ver);
KMDataNodeRelease(nd);
}
if( rc == 0 && (rc = KMDataNodeOpenNodeUpdate(opt, &nd, "colorSpace")) == 0 ) {
rc = KMDataNodeWriteB8(nd, &colorSpace);
KMDataNodeRelease(nd);
}
if( rc == 0 && (rc = KMDataNodeOpenNodeUpdate(opt, &nd, "colorSpaceKey")) == 0 ) {
rc = KMDataNodeWrite(nd, colorSpaceKey, 1);
KMDataNodeRelease(nd);
}
if( rc == 0 && (rc = KMDataNodeOpenNodeUpdate(opt, &nd, "origFormat")) == 0 ) {
rc = KMDataNodeWriteB8(nd, &origFormat);
KMDataNodeRelease(nd);
}
if( rc == 0 && (rc = KMDataNodeOpenNodeUpdate(opt, &nd, "printLabel")) == 0 ) {
rc = KMDataNodeWriteB8(nd, &printLabel);
KMDataNodeRelease(nd);
}
if( rc == 0 && (rc = KMDataNodeOpenNodeUpdate(opt, &nd, "printReadId")) == 0 ) {
rc = KMDataNodeWriteB8(nd, &printReadId);
KMDataNodeRelease(nd);
}
if( rc == 0 && (rc = KMDataNodeOpenNodeUpdate(opt, &nd, "clipQuality")) == 0 ) {
rc = KMDataNodeWriteB8(nd, &clipQuality);
KMDataNodeRelease(nd);
}
if( rc == 0 && (rc = KMDataNodeOpenNodeUpdate(opt, &nd, "minReadLen")) == 0 ) {
rc = KMDataNodeWriteB32(nd, &minReadLen);
KMDataNodeRelease(nd);
}
if( rc == 0 && (rc = KMDataNodeOpenNodeUpdate(opt, &nd, "qualityOffset")) == 0 ) {
rc = KMDataNodeWriteB16(nd, &qualityOffset);
KMDataNodeRelease(nd);
}
KMDataNodeRelease(opt);
}
FastqReaderWhack(reader);
return rc;
}
static
rc_t Fastq_Idx(const SRATable* sratbl, SIndexObj* obj, char* buffer, const size_t buffer_sz)
{
rc_t rc = 0;
const FastqReader* reader = NULL;
uint8_t colorSpace = false;
char* colorSpaceKey = "\0";
uint8_t origFormat = false;
uint8_t printLabel = true;
uint8_t printReadId = true;
uint8_t clipQuality = true;
uint32_t minReadLen = 0;
uint16_t qualityOffset = 0;
{{
const SRAColumn* c = NULL;
const uint8_t *platform = SRA_PLATFORM_UNDEFINED;
bitsz_t o, z;
if( (rc = SRATableOpenColumnRead(sratbl, &c, "PLATFORM", sra_platform_id_t)) != 0 ) {
return rc;
}
if( (rc = SRAColumnRead(c, 1, (const void **)&platform, &o, &z)) != 0 ) {
return rc;
}
if( *platform == SRA_PLATFORM_ABSOLID ) {
colorSpace = true;
}
SRAColumnRelease(c);
}}
if( (rc = FastqReaderMake(&reader, sratbl, g_accession,
colorSpace, origFormat, false, printLabel, printReadId,
!clipQuality, minReadLen, qualityOffset, colorSpaceKey[0],
obj->minSpotId, obj->maxSpotId)) != 0 ) {
return rc;
} else {
KMDataNode* opt = NULL, *nd = NULL;
if( (rc = KMDataNodeOpenNodeUpdate(obj->meta, &opt, "Format/Options")) != 0 ) {
return rc;
}
if( rc == 0 && (rc = KMDataNodeOpenNodeUpdate(opt, &nd, "colorSpace")) == 0 ) {
rc = KMDataNodeWriteB8(nd, &colorSpace);
KMDataNodeRelease(nd);
}
if( rc == 0 && (rc = KMDataNodeOpenNodeUpdate(opt, &nd, "colorSpaceKey")) == 0 ) {
rc = KMDataNodeWrite(nd, colorSpaceKey, 1);
KMDataNodeRelease(nd);
}
if( rc == 0 && (rc = KMDataNodeOpenNodeUpdate(opt, &nd, "origFormat")) == 0 ) {
rc = KMDataNodeWriteB8(nd, &origFormat);
KMDataNodeRelease(nd);
}
if( rc == 0 && (rc = KMDataNodeOpenNodeUpdate(opt, &nd, "printLabel")) == 0 ) {
rc = KMDataNodeWriteB8(nd, &printLabel);
KMDataNodeRelease(nd);
}
if( rc == 0 && (rc = KMDataNodeOpenNodeUpdate(opt, &nd, "printReadId")) == 0 ) {
rc = KMDataNodeWriteB8(nd, &printReadId);
KMDataNodeRelease(nd);
}
if( rc == 0 && (rc = KMDataNodeOpenNodeUpdate(opt, &nd, "clipQuality")) == 0 ) {
rc = KMDataNodeWriteB8(nd, &clipQuality);
KMDataNodeRelease(nd);
}
if( rc == 0 && (rc = KMDataNodeOpenNodeUpdate(opt, &nd, "minReadLen")) == 0 ) {
rc = KMDataNodeWriteB32(nd, &minReadLen);
KMDataNodeRelease(nd);
}
if( rc == 0 && (rc = KMDataNodeOpenNodeUpdate(opt, &nd, "qualityOffset")) == 0 ) {
rc = KMDataNodeWriteB16(nd, &qualityOffset);
KMDataNodeRelease(nd);
}
KMDataNodeRelease(opt);
}
if( rc == 0 ) {
size_t written = 0;
uint32_t blk = 0;
SIndexNode* inode = NULL;
while( rc == 0 ) {
rc = FastqReader_GetNextSpotSplitData(reader, buffer, buffer_sz, &written);
if( blk >= g_file_block_sz || (GetRCObject(rc) == rcRow && GetRCState(rc) == rcExhausted) ) {
inode->key_size = blk;
SLListPushTail(&obj->li, &inode->n);
DEBUG_MSG(5, ("Fastq index closed spots %lu, offset %lu, block size %lu\n",
inode->id_qty, inode->key, inode->key_size));
inode = NULL;
if( blk > obj->buffer_sz ) {
obj->buffer_sz = blk;
}
blk = 0;
}
if( GetRCObject(rc) == rcRow && GetRCState(rc) == rcExhausted ) {
rc = 0;
break;
}
if( inode == NULL ) {
spotid_t spotid = 0;
if( (rc = FastqReaderCurrentSpot(reader, &spotid)) != 0 ) {
break;
}
inode = malloc(sizeof(SIndexNode));
if( inode == NULL ) {
rc = RC(rcExe, rcIndex, rcConstructing, rcMemory, rcExhausted);
break;
}
inode->key = obj->file_size;
inode->key_size = 0;
inode->id = spotid;
inode->id_qty = 0;
DEBUG_MSG(5, ("Fastq index opened spot %ld, offset %lu\n", inode->id, inode->key));
}
inode->id_qty++;
obj->file_size += written;
blk += written;
MD5StateAppend(&obj->md5, buffer, written);
if( g_dump ) {
fwrite(buffer, written, 1, stderr);
}
}
rc = rc ? rc : Quitting();
if( rc != 0 ) {
spotid_t spot = 0;
FastqReaderCurrentSpot(reader, &spot);
PLOGERR(klogErr, (klogErr, rc, "spot $(s)", PLOG_U32(s), spot));
}
}
FastqReaderWhack(reader);
return rc;
}
