static inline GtWord decoder_calc_start_of_encoded_data(FILE *fp)
{
bool is_not_at_pageborder;
GtUword pagesize = gt_pagesize();
is_not_at_pageborder = (ftell(fp) % pagesize) != 0;
if (is_not_at_pageborder)
return (ftell(fp) / pagesize + 1) * pagesize;
else
return ftell(fp);
}
static HcrHuffDataIterator *decoder_init_data_iterator(GtWord start_of_encoding,
GtWord end_of_encoding,
const GtStr *filename)
{
HcrHuffDataIterator *data_iter = gt_malloc(sizeof (*data_iter));
data_iter->path = gt_str_get(filename);
data_iter->pos = data_iter->start = (size_t) start_of_encoding;
data_iter->end = (size_t) end_of_encoding;
data_iter->pages_per_chunk = HCR_PAGES_PER_CHUNK;
data_iter->pagesize = gt_pagesize();
gt_assert(data_iter->start % data_iter->pagesize == 0);
data_iter->blocksize = data_iter->pagesize * data_iter->pages_per_chunk;
gt_safe_assign(data_iter->bitseq_per_chunk,
(data_iter->blocksize / sizeof (GtBitsequence)));
data_iter->data = NULL;
return data_iter;
}
GtSfxmappedrange *gt_Sfxmappedrange_new(const char *tablename,
GtUword numofentries,
GtSfxmappedrangetype type,
GtSfxmappedrangetransformfunc
transformfunc,
const void *transformfunc_data)
{
GtSfxmappedrange *sfxmappedrange;
sfxmappedrange = gt_malloc(sizeof (*sfxmappedrange));
sfxmappedrange->ptr = NULL;
sfxmappedrange->pagesize = gt_pagesize();
sfxmappedrange->usedptrptr = NULL;
sfxmappedrange->filename = NULL;
sfxmappedrange->writable = false;
sfxmappedrange->entire = NULL;
sfxmappedrange->transformfunc = transformfunc;
sfxmappedrange->transformfunc_data = transformfunc_data;
sfxmappedrange->type = type;
sfxmappedrange->tablename = gt_str_new_cstr(tablename);
sfxmappedrange->currentminindex = sfxmappedrange->currentmaxindex = 0;
sfxmappedrange->indexrange_defined = false;
switch (type)
{
case GtSfxGtBitsequence:
sfxmappedrange->sizeofunit = sizeof (GtBitsequence);
sfxmappedrange->numofunits = GT_NUMOFINTSFORBITS(numofentries);
break;
case GtSfxuint32_t:
sfxmappedrange->sizeofunit = sizeof (uint32_t);
sfxmappedrange->numofunits = (size_t) numofentries;
break;
case GtSfxunsignedlong:
sfxmappedrange->sizeofunit = sizeof (GtUword);
sfxmappedrange->numofunits = (size_t) numofentries;
break;
default:
gt_assert(false);
break;
}
return sfxmappedrange;
}
GtHcrEncoder *gt_hcr_encoder_new(GtStrArray *files, GtAlphabet *alpha,
bool descs, GtQualRange qrange, GtTimer *timer,
GtError *err)
{
GtBaseQualDistr *bqd;
GtHcrEncoder *hcr_enc;
GtSeqIterator *seqit;
GtStrArray *file;
int had_err = 0,
status;
GtUword len1,
len2,
i,
num_of_reads = 0;
const GtUchar *seq,
*qual;
char *desc;
gt_error_check(err);
gt_assert(alpha && files);
if (timer != NULL)
gt_timer_show_progress(timer, "get <base,qual> distr", stdout);
if (qrange.start != GT_UNDEF_UINT)
if (qrange.start == qrange.end) {
gt_error_set(err, "qrange.start must unequal qrange.end");
return NULL;
}
hcr_enc = gt_malloc(sizeof (GtHcrEncoder));
hcr_enc->files = files;
hcr_enc->num_of_files = gt_str_array_size(files);
hcr_enc->num_of_reads = 0;
hcr_enc->page_sampling = false;
hcr_enc->regular_sampling = false;
hcr_enc->sampling_rate = 0;
hcr_enc->pagesize = gt_pagesize();
if (descs) {
hcr_enc->encdesc_encoder = gt_encdesc_encoder_new();
if (timer != NULL)
gt_encdesc_encoder_set_timer(hcr_enc->encdesc_encoder, timer);
}
else
hcr_enc->encdesc_encoder = NULL;
hcr_enc->seq_encoder = gt_malloc(sizeof (GtHcrSeqEncoder));
hcr_enc->seq_encoder->alpha = alpha;
hcr_enc->seq_encoder->sampling = NULL;
hcr_enc->seq_encoder->fileinfos = gt_calloc((size_t) hcr_enc->num_of_files,
sizeof (*(hcr_enc->seq_encoder->fileinfos)));
hcr_enc->seq_encoder->qrange = qrange;
bqd = hcr_base_qual_distr_new(alpha, qrange);
/* check if reads in the same file are of same length and get
<base, quality> pair distribution */
for (i = 0; i < hcr_enc->num_of_files; i++) {
file = gt_str_array_new();
gt_str_array_add(file, gt_str_array_get_str(files, i));
seqit = gt_seq_iterator_fastq_new(file, err);
if (!seqit) {
gt_error_set(err, "cannot initialize GtSeqIteratorFastQ object");
had_err = -1;
}
if (!had_err) {
gt_seq_iterator_set_symbolmap(seqit, gt_alphabet_symbolmap(alpha));
gt_seq_iterator_set_quality_buffer(seqit, &qual);
status = gt_seq_iterator_next(seqit, &seq, &len1, &desc, err);
if (status == 1) {
num_of_reads = 1UL;
while (!had_err) {
status = gt_seq_iterator_next(seqit, &seq, &len2, &desc, err);
if (status == -1)
had_err = -1;
if (status != 1)
break;
if (len2 != len1) {
gt_error_set(err, "reads have to be of equal length");
had_err = -1;
break;
}
if (hcr_base_qual_distr_add(bqd, qual, seq, len1) != 0)
had_err = -1;
len1 = len2;
num_of_reads++;
}
}
else if (status == -1)
had_err = -1;
if (!had_err) {
if (i == 0)
hcr_enc->seq_encoder->fileinfos[i].readnum = num_of_reads;
else
hcr_enc->seq_encoder->fileinfos[i].readnum =
hcr_enc->seq_encoder->fileinfos[i - 1].readnum + num_of_reads;
hcr_enc->seq_encoder->fileinfos[i].readlength = len1;
}
}
hcr_enc->num_of_reads += num_of_reads;
gt_str_array_delete(file);
gt_seq_iterator_delete(seqit);
}
if (!had_err)
hcr_base_qual_distr_trim(bqd);
if (!had_err) {
if (timer != NULL)
gt_timer_show_progress(timer, "build huffman tree for sequences and"
" qualities", stdout);
hcr_enc->seq_encoder->huffman =
gt_huffman_new(bqd,
hcr_base_qual_distr_func,
(GtUword) bqd->ncols * bqd->nrows);
}
if (!had_err) {
hcr_enc->seq_encoder->qual_offset = bqd->qual_offset;
hcr_base_qual_distr_delete(bqd);
return hcr_enc;
}
return NULL;
}
