SR_AnchorInfo* SR_AnchorInfoAlloc(uint32_t capacity)
{
SR_AnchorInfo* pNewInfo = (SR_AnchorInfo*) malloc(sizeof(SR_AnchorInfo));
if (pNewInfo == NULL)
SR_ErrQuit("ERROR: Not enough memory for an anchor information object.\n");
pNewInfo->pAnchors = (char**) malloc(capacity * sizeof(char*));
if (pNewInfo->pAnchors == NULL)
SR_ErrQuit("ERROR: Not enough memory for the storage of anchor names in the anchor information object.\n");
pNewInfo->pLength = (int32_t*) malloc(capacity * sizeof(int32_t));
if (pNewInfo->pLength == NULL)
SR_ErrQuit("ERROR: Not enough memory for the storage of anchor length in the anchor information object.\n");
pNewInfo->pMd5s = (char*) malloc(capacity * MD5_STR_LEN * sizeof(char));
if (pNewInfo->pMd5s == NULL)
SR_ErrQuit("ERROR: Not enough memory for the storage of md5 string in the anchor information object.\n");
pNewInfo->size = 0;
pNewInfo->capacity = capacity;
pNewInfo->pAnchorHash = kh_init(name);
kh_resize(name, pNewInfo->pAnchorHash, 2 * capacity);
return pNewInfo;
}
mrb_value
mrb_hash_new_capa(mrb_state *mrb, size_t capa)
{
struct RHash *h;
h = (struct RHash*)mrb_obj_alloc(mrb, MRB_TT_HASH, mrb->hash_class);
h->ht = kh_init(ht, mrb);
kh_resize(ht, h->ht, capa);
h->iv = 0;
return mrb_obj_value(h);
}
static void worker_post(void *g, long i, int tid)
{
int j, start_a, start_p, n, n_keys;
idxhash_t *h;
mm_idx_t *mi = (mm_idx_t*)g;
mm_idx_bucket_t *b = &mi->B[i];
if (b->a.n == 0) return;
// sort by minimizer
radix_sort_128x(b->a.a, b->a.a + b->a.n);
// count and preallocate
for (j = 1, n = 1, n_keys = 0, b->n = 0; j <= b->a.n; ++j) {
if (j == b->a.n || b->a.a[j].x != b->a.a[j-1].x) {
++n_keys;
if (n > 1) b->n += n;
n = 1;
} else ++n;
}
h = kh_init(idx);
kh_resize(idx, h, n_keys);
b->p = (uint64_t*)calloc(b->n, 8);
// create the hash table
for (j = 1, n = 1, start_a = start_p = 0; j <= b->a.n; ++j) {
if (j == b->a.n || b->a.a[j].x != b->a.a[j-1].x) {
khint_t itr;
int absent;
mm128_t *p = &b->a.a[j-1];
itr = kh_put(idx, h, p->x>>mi->b<<1, &absent);
assert(absent && j - start_a == n);
if (n == 1) {
kh_key(h, itr) |= 1;
kh_val(h, itr) = p->y;
} else {
int k;
for (k = 0; k < n; ++k)
b->p[start_p + k] = b->a.a[start_a + k].y;
kh_val(h, itr) = (uint64_t)start_p<<32 | n;
start_p += n;
}
start_a = j, n = 1;
} else ++n;
void TGM_FragLenHistArrayInit(TGM_FragLenHistArray* pHistArray, unsigned int newSize)
{
TGM_FragLenHistArrayClear(pHistArray);
if (newSize > pHistArray->capacity)
{
TGM_ARRAY_RESIZE(pHistArray, newSize * 2, TGM_FragLenHist);
memset(pHistArray->data + pHistArray->size, 0, (newSize * 2 - pHistArray->size) * sizeof(TGM_FragLenHist));
}
for (unsigned int i = 0; i != newSize; ++i)
{
if (pHistArray->data[i].rawHist == NULL)
{
pHistArray->data[i].rawHist = kh_init(fragLen);
kh_resize(fragLen, pHistArray->data[i].rawHist, DEFAULT_NUM_HIST_ELMNT);
}
}
pHistArray->size = newSize;
}
SR_SampleInfo* SR_SampleInfoAlloc(uint32_t capacity)
{
SR_SampleInfo* pNewInfo = (SR_SampleInfo*) malloc(sizeof(SR_SampleInfo));
if (pNewInfo == NULL)
SR_ErrQuit("ERROR: Not enough memory for a sample information object.\n");
pNewInfo->pSamples = (char**) malloc(capacity * sizeof(char*));
if (pNewInfo->pSamples == NULL)
SR_ErrQuit("ERROR: Not enough memory for the storage of sample names in the sample information object.\n");
pNewInfo->pReadFraction = (double*) malloc(capacity * sizeof(double));
if (pNewInfo->pReadFraction == NULL)
SR_ErrQuit("ERROR: Not enough memory for the storage of read fraction in the sample information object.\n");
pNewInfo->pSampleHash = kh_init(name);
kh_resize(name, pNewInfo->pSampleHash, 2 * capacity);
pNewInfo->size = 0;
pNewInfo->capacity = capacity;
return pNewInfo;
}
TGM_Status TGM_FragLenHistArrayUpdate(TGM_FragLenHistArray* pHistArray, unsigned int backHistIndex, uint32_t fragLen)
{
if (backHistIndex > pHistArray->size)
return TGM_ERR;
TGM_FragLenHist* pCurrHist = pHistArray->data + (pHistArray->size - backHistIndex);
// if the pair mode is not valid
// we only updated the count of the invalid pair and return
if (fragLen == 0)
{
++(pCurrHist->modeCount[INVALID_PAIR_MODE_SET_INDEX]);
return TGM_OK;
}
// because we can have up to 2 different pair mode sets (4 differen pair modes)
// we should choose which histogram we should update
// the first one (with index 0 or 1) or the second one(2, 3)
khash_t(fragLen)* pCurrHash = pCurrHist->rawHist;
if (pCurrHash == NULL)
{
pCurrHash = kh_init(fragLen);
kh_resize(fragLen, pCurrHash, 20);
}
int ret = 0;
khiter_t khIter = kh_put(fragLen, pCurrHash, fragLen, &ret);
if (ret == 0)
kh_value(pCurrHash, khIter) += 1;
else
kh_value(pCurrHash, khIter) = 1;
++(pCurrHist->modeCount[0]);
pCurrHist->rawHist = pCurrHash;
return TGM_OK;
}
SR_LibInfoTable* SR_LibInfoTableAlloc(uint32_t capAnchor, uint32_t capSample, uint32_t capReadGrp)
{
SR_LibInfoTable* pNewTable = (SR_LibInfoTable*) malloc(sizeof(SR_LibInfoTable));
if (pNewTable == NULL)
SR_ErrQuit("ERROR: Not enough memory for a library information table object.\n");
pNewTable->pSampleInfo = SR_SampleInfoAlloc(capSample);
pNewTable->pLibInfo = (SR_LibInfo*) malloc(capReadGrp * sizeof(SR_LibInfo));
if (pNewTable->pLibInfo == NULL)
SR_ErrQuit("ERROR: Not enough memory for the storage of library information in an library table object.\n");
pNewTable->pReadGrps = (char**) malloc(capReadGrp * sizeof(char*));
if (pNewTable->pLibInfo == NULL)
SR_ErrQuit("ERROR: Not enough memory for the storage of read group names in an library table object.\n");
pNewTable->pSampleMap = (int32_t*) malloc(capReadGrp * sizeof(int32_t));
if (pNewTable->pSampleMap == NULL)
SR_ErrQuit("ERROR: Not enough memory for the storage of read-group-to-sample map in an library table object.\n");
pNewTable->pSeqTech = (int8_t*) malloc(capReadGrp * sizeof(int8_t));
if (pNewTable->pSeqTech == NULL)
SR_ErrQuit("ERROR: Not enough memory for the storage of sequencing technologies in an library table object.\n");
pNewTable->pAnchorInfo = SR_AnchorInfoAlloc(capAnchor);
pNewTable->pReadGrpHash = kh_init(name);
kh_resize(name, pNewTable->pReadGrpHash, 2 * capReadGrp);
pNewTable->size = 0;
pNewTable->capacity = capReadGrp;
pNewTable->fragLenMax = 0;
pNewTable->cutoff = 0.0;
pNewTable->trimRate = 0.0;
return pNewTable;
}
SR_BamInStream* SR_BamInStreamAlloc(const char* bamFilename, uint32_t binLen, unsigned int numThreads, unsigned int buffCapacity,
unsigned int reportSize, const SR_StreamMode* pStreamMode)
{
SR_BamInStream* pBamInStream = (SR_BamInStream*) calloc(1, sizeof(SR_BamInStream));
if (pBamInStream == NULL)
SR_ErrQuit("ERROR: Not enough memory for a bam input stream object.");
pBamInStream->bam_cur_status = -1;
pBamInStream->fpBamInput = bam_open(bamFilename, "r");
if (pBamInStream->fpBamInput == NULL)
SR_ErrQuit("ERROR: Cannot open bam file %s for reading.\n", bamFilename);
if ((pStreamMode->controlFlag & SR_USE_BAM_INDEX) != 0)
{
pBamInStream->pBamIndex = bam_index_load(bamFilename);
if (pBamInStream->pBamIndex == NULL) {
SR_ErrMsg("WARNING: Cannot open bam index file for reading. Creating it......");
bam_index_build(bamFilename);
SR_ErrMsg(" The bam index is created.");
pBamInStream->pBamIndex = bam_index_load(bamFilename);
}
}
pBamInStream->filterFunc = pStreamMode->filterFunc;
pBamInStream->filterData = pStreamMode->filterData;
pBamInStream->numThreads = numThreads;
pBamInStream->reportSize = reportSize;
pBamInStream->currRefID = NO_QUERY_YET;
pBamInStream->currBinPos = NO_QUERY_YET;
pBamInStream->binLen = binLen;
pBamInStream->pNewNode = NULL;
pBamInStream->pBamIterator = NULL;
if (numThreads > 0)
{
pBamInStream->pRetLists = (SR_BamList*) calloc(numThreads, sizeof(SR_BamList));
if (pBamInStream->pRetLists == NULL)
SR_ErrQuit("ERROR: Not enough memory for the storage of retrun alignment lists in the bam input stream object.\n");
pBamInStream->pAlgnTypes = (SR_AlgnType*) malloc(numThreads * reportSize * sizeof(SR_AlgnType));
if (pBamInStream->pAlgnTypes == NULL)
SR_ErrQuit("ERROR: Not enough memory for the storage of pair alignment type in the bam input stream object.\n");
}
else
{
pBamInStream->pRetLists = NULL;
pBamInStream->pAlgnTypes = NULL;
pBamInStream->reportSize = 0;
}
if ((pStreamMode->controlFlag & SR_PAIR_GENOMICALLY) == 0)
{
pBamInStream->pNameHashes[PREV_BIN] = kh_init(queryName);
kh_resize(queryName, pBamInStream->pNameHashes[PREV_BIN], reportSize);
}
else
{
pBamInStream->pNameHashes[PREV_BIN] = NULL;
pBamInStream->binLen = SR_MAX_BIN_LEN;
}
pBamInStream->pNameHashes[CURR_BIN] = kh_init(queryName);
kh_resize(queryName, pBamInStream->pNameHashes[CURR_BIN], reportSize);
pBamInStream->pMemPool = SR_BamMemPoolAlloc(buffCapacity);
pBamInStream->bam_cur_status = 1;
return pBamInStream;
}
