static int seqorder_str_compare_num(const void *v1, const void *v2, void *data)
{
GtUword n1 = *(GtUword*) v1,
n2 = *(GtUword*) v2,
desclen1, desclen2,
anum, bnum;
int arval, brval, rval = 0;
const char *desc1, *desc2;
char buf[BUFSIZ];
desc1 = gt_encseq_description((GtEncseq*) data, &desclen1, n1);
desc2 = gt_encseq_description((GtEncseq*) data, &desclen2, n2);
(void) strncpy(buf, desc1, MIN(BUFSIZ, desclen1) * sizeof (char));
buf[desclen1] = '\0';
arval = gt_parse_uword(&anum, buf);
(void) strncpy(buf, desc2, MIN(BUFSIZ, desclen2) * sizeof (char));
buf[desclen2] = '\0';
brval = gt_parse_uword(&bnum, buf);
if (arval == 0 && brval == 0)
rval = anum-bnum;
else if (arval == 0)
return -1;
else if (brval == 0)
return 1;
else
rval = 0;
return rval;
}
static int seqorder_str_compare_lex(const void *v1, const void *v2, void *data)
{
GtUword n1 = *(GtUword*) v1,
n2 = *(GtUword*) v2,
desclen1, desclen2;
const char *desc1, *desc2;
int rval = 0;
desc1 = gt_encseq_description((GtEncseq*) data, &desclen1, n1);
desc2 = gt_encseq_description((GtEncseq*) data, &desclen2, n2);
rval = strncmp(desc1, desc2, MIN(desclen1, desclen2) * sizeof (char));
if (rval == 0)
rval = desclen1-desclen2;
return rval;
}
static void gt_seqorder_output(unsigned long seqnum, GtEncseq *encseq)
{
GtEncseqReader *esr;
unsigned long startpos, len, desclen = 0;
const char *desc = NULL;
unsigned long i;
startpos = gt_encseq_seqstartpos(encseq, seqnum);
len = gt_encseq_seqlength(encseq, seqnum);
gt_xfputc(GT_FASTA_SEPARATOR, stdout);
if (gt_encseq_has_description_support(encseq))
{
desc = gt_encseq_description(encseq, &desclen, seqnum);
gt_xfwrite(desc, (size_t)1, (size_t)desclen, stdout);
}
gt_xfputc('\n', stdout);
esr = gt_encseq_create_reader_with_readmode(encseq, GT_READMODE_FORWARD,
startpos);
for (i = 0; i < len; i++)
{
gt_xfputc(gt_encseq_reader_next_decoded_char(esr), stdout);
}
gt_encseq_reader_delete(esr);
gt_xfputc('\n', stdout);
}
static int encseq_lua_description(lua_State *L)
{
GtEncseq **encseq;
GtUword seqno, desclen;
const char *string;
encseq = check_encseq(L, 1);
seqno = luaL_checknumber(L, 2);
luaL_argcheck(L, seqno < gt_encseq_num_of_sequences(*encseq), 2,
"cannot exceed number of sequences");
string = gt_encseq_description(*encseq, &desclen, seqno);
lua_pushlstring(L, string, desclen);
return 1;
}
const char* gt_bioseq_get_description(GtBioseq *bs, GtUword idx)
{
const char *desc;
char *mydesc;
GtUword desclen;
gt_assert(bs && bs->encseq);
gt_assert(idx < gt_encseq_num_of_sequences(bs->encseq));
if (!(mydesc = bs->descriptions[idx])) {
desc = gt_encseq_description(bs->encseq, &desclen, idx);
mydesc = gt_calloc(desclen + 1, sizeof (char));
strncpy(mydesc, desc, desclen);
bs->descriptions[idx] = mydesc;
}
return (const char*) mydesc;
}
static char* gt_encseq_col_get_description(const GtSeqCol *sc,
GtUword filenum,
GtUword seqnum)
{
GtEncseqCol *esc;
const char *desc;
GtUword encseq_seqnum, desclen;
esc = gt_encseq_col_cast(sc);
gt_assert(esc && filenum < gt_encseq_num_of_files(esc->encseq));
encseq_seqnum = gt_encseq_filenum_first_seqnum(esc->encseq, filenum) + seqnum;
gt_assert(encseq_seqnum < gt_encseq_num_of_sequences(esc->encseq));
desc = gt_encseq_description(esc->encseq, &desclen, encseq_seqnum);
gt_assert(desc && desclen > 0);
return gt_cstr_dup_nt(desc, desclen);;
}
static int gt_encseq_col_do_grep_desc(GtEncseqCol *esc, GtUword *filenum,
GtUword *seqnum, GtStr *seqid,
GtError *err)
{
GtUword j;
const GtSeqInfo *seq_info_ptr;
GtSeqInfo seq_info;
bool match = false;
int had_err = 0;
gt_error_check(err);
gt_assert(esc && filenum && seqnum && seqid);
/* create cache */
if (!esc->grep_cache)
esc->grep_cache = gt_seq_info_cache_new();
/* try to read from cache */
seq_info_ptr = gt_seq_info_cache_get(esc->grep_cache, gt_str_get(seqid));
if (seq_info_ptr) {
*filenum = seq_info_ptr->filenum;
*seqnum = seq_info_ptr->seqnum;
return 0;
}
for (j = 0; !had_err && j < gt_encseq_num_of_sequences(esc->encseq); j++) {
const char *desc;
char *buf;
GtUword desc_len;
desc = gt_encseq_description(esc->encseq, &desc_len, j);
buf = gt_calloc(desc_len + 1, sizeof (char));
memcpy(buf, desc, desc_len * sizeof (char));
had_err = gt_grep(&match, gt_str_get(seqid), buf, err);
gt_free(buf);
if (!had_err && match) {
*filenum = seq_info.filenum =
gt_encseq_filenum(esc->encseq,
gt_encseq_seqstartpos(esc->encseq, j));
*seqnum = seq_info.seqnum =
j - gt_encseq_filenum_first_seqnum(esc->encseq, *filenum);
gt_seq_info_cache_add(esc->grep_cache, gt_str_get(seqid), &seq_info);
break;
}
}
if (!had_err && !match) {
gt_error_set(err, "no description matched sequence ID '%s'",
gt_str_get(seqid));
had_err = -1;
}
return had_err;
}
static int giextract_encodedseq2fasta(FILE *fpout,
const GtEncseq *encseq,
unsigned long seqnum,
const Fastakeyquery *fastakeyquery,
unsigned long linewidth,
GT_UNUSED GtError *err)
{
const char *desc;
unsigned long desclen;
bool haserr = false;
desc = gt_encseq_description(encseq, &desclen, seqnum);
gt_xfputc('>',fpout);
if (fastakeyquery != NULL && !COMPLETE(fastakeyquery))
{
printf("%s %lu %lu ",fastakeyquery->fastakey,
fastakeyquery->frompos,
fastakeyquery->topos);
}
gt_xfwrite(desc,sizeof *desc,(size_t) desclen,fpout);
if (!haserr)
{
unsigned long frompos, topos, seqstartpos, seqlength ;
gt_xfputc('\n',fpout);
seqstartpos = gt_encseq_seqstartpos(encseq, seqnum);
seqlength = gt_encseq_seqlength(encseq, seqnum);
if (fastakeyquery != NULL && !COMPLETE(fastakeyquery))
{
frompos = fastakeyquery->frompos-1;
topos = fastakeyquery->topos - fastakeyquery->frompos + 1;
} else
{
frompos = 0;
topos = seqlength;
}
gt_encseq2symbolstring(fpout,
encseq,
GT_READMODE_FORWARD,
seqstartpos + frompos,
topos,
linewidth);
}
return haserr ? -1 : 0;
}
static int gt_encseq_col_md5_to_description(GtSeqCol *sc, GtStr *desc,
GtStr *md5_seqid, GtError *err)
{
GtUword seqnum = GT_UNDEF_UWORD;
char seqid[GT_MD5_SEQID_HASH_LEN + 1];
int had_err = 0;
GtEncseqCol *esc;
esc = gt_encseq_col_cast(sc);
gt_error_check(err);
gt_assert(esc && desc && md5_seqid && err);
gt_assert(gt_md5_seqid_has_prefix(gt_str_get(md5_seqid)));
if (gt_str_length(md5_seqid) >= GT_MD5_SEQID_TOTAL_LEN) {
const char *cstrseqid = gt_str_get(md5_seqid);
if (cstrseqid[GT_MD5_SEQID_TOTAL_LEN-1] != GT_MD5_SEQID_SEPARATOR) {
gt_error_set(err, "MD5 sequence id %s not terminated with '%c'",
gt_str_get(md5_seqid), GT_MD5_SEQID_SEPARATOR);
had_err = -1;
}
if (!had_err) {
strncpy(seqid, cstrseqid + GT_MD5_SEQID_PREFIX_LEN,
GT_MD5_SEQID_HASH_LEN);
seqid[GT_MD5_SEQID_HASH_LEN] = '\0';
}
}
seqnum = gt_md5_tab_map(esc->md5_tab, seqid);
if (seqnum != GT_UNDEF_UWORD) {
const char *cdesc;
GtUword desc_len;
gt_assert(seqnum < gt_encseq_num_of_sequences(esc->encseq));
cdesc = gt_encseq_description(esc->encseq, &desc_len, seqnum);
gt_str_append_cstr_nt(desc, cdesc, desc_len);
} else {
gt_error_set(err, "sequence %s not found", gt_str_get(md5_seqid));
had_err = -1;
}
return had_err;
}
static int output_sequence(GtEncseq *encseq, GtEncseqDecodeArguments *args,
const char *filename, GtError *err)
{
GtUword i, j, sfrom, sto;
int had_err = 0;
bool has_desc;
GtEncseqReader *esr;
gt_assert(encseq);
if (!(has_desc = gt_encseq_has_description_support(encseq)))
gt_warning("Missing description support for file %s", filename);
if (strcmp(gt_str_get(args->mode), "fasta") == 0) {
/* specify a single sequence to extract */
if (args->seq != GT_UNDEF_UWORD) {
if (args->seq >= gt_encseq_num_of_sequences(encseq)) {
gt_error_set(err,
"requested sequence "GT_WU" exceeds number of sequences "
"("GT_WU")", args->seq,
gt_encseq_num_of_sequences(encseq));
return -1;
}
sfrom = args->seq;
sto = args->seq + 1;
} else if (args->seqrng.start != GT_UNDEF_UWORD
&& args->seqrng.end != GT_UNDEF_UWORD) {
/* specify a sequence range to extract */
if (args->seqrng.start >= gt_encseq_num_of_sequences(encseq)
|| args->seqrng.end >= gt_encseq_num_of_sequences(encseq)) {
gt_error_set(err,
"range "GT_WU"-"GT_WU" includes a sequence number "
"exceeding the total number of sequences ("GT_WU")",
args->seqrng.start,
args->seqrng.end,
gt_encseq_num_of_sequences(encseq));
return -1;
}
sfrom = args->seqrng.start;
sto = args->seqrng.end + 1;
} else {
/* extract all sequences */
sfrom = 0;
sto = gt_encseq_num_of_sequences(encseq);
}
for (i = sfrom; i < sto; i++) {
GtUword desclen, startpos, len;
char buf[BUFSIZ];
const char *desc = NULL;
/* XXX: maybe make this distinction in the functions via readmode? */
if (!GT_ISDIRREVERSE(args->rm)) {
startpos = gt_encseq_seqstartpos(encseq, i);
len = gt_encseq_seqlength(encseq, i);
if (has_desc) {
desc = gt_encseq_description(encseq, &desclen, i);
} else {
(void) snprintf(buf, BUFSIZ, "sequence "GT_WU"", i);
desclen = strlen(buf);
desc = buf;
}
} else {
startpos = gt_encseq_seqstartpos(encseq, i);
len = gt_encseq_seqlength(encseq,
gt_encseq_num_of_sequences(encseq)-1-i);
startpos = gt_encseq_total_length(encseq)
- (gt_encseq_seqstartpos(encseq,
gt_encseq_num_of_sequences(
encseq)-1-i) + len);
if (has_desc) {
desc = gt_encseq_description(encseq,
&desclen,
gt_encseq_num_of_sequences(encseq)-1-i);
} else {
(void) snprintf(buf, BUFSIZ, "sequence "GT_WU"", i);
desclen = strlen(buf);
desc = buf;
}
}
gt_assert(desc);
/* output description */
gt_xfputc(GT_FASTA_SEPARATOR, stdout);
gt_xfwrite(desc, 1, desclen, stdout);
gt_xfputc('\n', stdout);
/* XXX: make this more efficient by writing in a buffer first and then
showing the result */
if (args->singlechars) {
for (j = 0; j < len; j++) {
gt_xfputc(gt_encseq_get_decoded_char(encseq,
startpos + j,
args->rm),
stdout);
}
} else {
esr = gt_encseq_create_reader_with_readmode(encseq, args->rm, startpos);
for (j = 0; j < len; j++) {
gt_xfputc(gt_encseq_reader_next_decoded_char(esr), stdout);
}
gt_encseq_reader_delete(esr);
}
gt_xfputc('\n', stdout);
}
}
if (strcmp(gt_str_get(args->mode), "concat") == 0) {
GtUword from = 0,
to = gt_encseq_total_length(encseq) - 1;
if (args->rng.start != GT_UNDEF_UWORD && args->rng.end != GT_UNDEF_UWORD) {
if (args->rng.end > to) {
had_err = -1;
gt_error_set(err,
"end of range ("GT_WU") exceeds encoded sequence length "
"("GT_WU")", args->rng.end, to);
}
if (!had_err) {
from = args->rng.start;
to = args->rng.end;
}
}
if (!had_err) {
if (args->singlechars) {
for (j = from; j <= to; j++) {
char cc = gt_encseq_get_decoded_char(encseq, j, args->rm);
if (cc == (char) SEPARATOR)
cc = gt_str_get(args->sepchar)[0];
gt_xfputc(cc, stdout);
}
} else {
esr = gt_encseq_create_reader_with_readmode(encseq, args->rm, from);
if (esr) {
for (j = from; j <= to; j++) {
char cc = gt_encseq_reader_next_decoded_char(esr);
if (cc == (char) SEPARATOR)
cc = gt_str_get(args->sepchar)[0];
gt_xfputc(cc, stdout);
}
gt_encseq_reader_delete(esr);
}
}
gt_xfputc('\n', stdout);
}
}
return had_err;
}
static void condenseq_process_descriptions(GtCondenseq *condenseq,
const GtEncseq *orig_es,
GtLogger *logger)
{
GtUword *dist;
const char *desc;
char *cur_id_startptr;
GtUword desclen,
dist_idx,
distsize = (GtUword) 128,
idlen,
idx,
maxendidx = 0,
maxlen = 0,
minlen = GT_UWORD_MAX,
wastedmem = 0,
sdssize,
cur_total_id_len = 0;
bool use_const_len;
condenseq->ids_total_len = 0;
dist = gt_calloc((size_t) distsize, sizeof (*dist));
for (idx = 0; idx < condenseq->orig_num_seq; ++idx) {
desc = gt_encseq_description(orig_es, &desclen, idx);
idlen = condenseq_idlen(desc, desclen);
if (distsize <= idlen) {
dist = gt_realloc(dist, (size_t) (idlen + 1) * sizeof (*dist));
for (dist_idx = distsize; dist_idx <= idlen; dist_idx++)
dist[dist_idx] = 0;
distsize = idlen + 1;
}
dist[idlen]++;
if (idlen > maxlen)
maxlen = idlen;
if (idlen < minlen)
minlen = idlen;
maxendidx += idlen;
}
/* calculate memory we would waste if we assume equal length, and size if we
store actual descriptions */
for (dist_idx = minlen; dist_idx < maxlen; dist_idx++) {
wastedmem += dist[dist_idx] * (maxlen - dist_idx);
condenseq->ids_total_len += dist[dist_idx] * dist_idx;
}
condenseq->ids_total_len += dist_idx * dist[dist_idx];
sdssize = (GtUword) gt_intset_best_memory_size(maxendidx,
condenseq->orig_num_seq);
use_const_len = wastedmem < sdssize;
if (use_const_len) {
gt_logger_log(logger, "Condenseq descriptions will use const len, " GT_WU
", \"wasting\" " GT_WU " bytes. SDS would use "
GT_WU " bytes",
maxlen, wastedmem, sdssize);
condenseq->id_len = maxlen;
condenseq->ids_total_len = maxlen * condenseq->orig_num_seq;
}
else {
gt_logger_log(logger, "Condenseq descriptions will use sdstab with size "
GT_WU ". Const length would have wasted " GT_WU " bytes.",
sdssize, wastedmem);
condenseq->sdstab = gt_intset_best_new(maxendidx, condenseq->orig_num_seq);
}
condenseq->orig_ids = gt_calloc((size_t) condenseq->ids_total_len,
sizeof (*condenseq->orig_ids));
cur_id_startptr = condenseq->orig_ids;
for (idx = 0; idx < condenseq->orig_num_seq; ++idx) {
desc = gt_encseq_description(orig_es, &desclen, idx);
idlen = condenseq_idlen(desc, desclen);
gt_assert(idlen <= maxlen);
(void) memcpy(cur_id_startptr, desc, (size_t) idlen);
if (use_const_len) {
cur_id_startptr += maxlen;
cur_total_id_len += maxlen;
}
else {
cur_id_startptr += idlen;
cur_total_id_len += idlen;
gt_intset_add(condenseq->sdstab, cur_total_id_len);
}
}
gt_assert(cur_total_id_len == condenseq->ids_total_len);
gt_free(dist);
}
static GtMatchIteratorStatus gt_match_iterator_sw_next(GtMatchIterator *mi,
GT_UNUSED GtMatch **match,
GT_UNUSED GtError *err)
{
GtMatchIteratorSW *mis;
GtSeq *seq_a, *seq_b;
char *a, *b;
const char *adesc, *bdesc;
GtAlignment *ali = NULL;
unsigned long seqlen_a, seqlen_b, seqpos;
GtRange arng, brng;
gt_assert(mi && match);
mis = gt_match_iterator_sw_cast(mi);
while (true) {
if (!mis->pvt->firstali)
mis->pvt->seqno_es2++;
if (mis->pvt->seqno_es2 == gt_encseq_num_of_sequences(mis->pvt->es2)) {
mis->pvt->seqno_es1++;
if (mis->pvt->seqno_es1 == gt_encseq_num_of_sequences(mis->pvt->es1))
return GT_MATCHER_STATUS_END;
mis->pvt->seqno_es2 = 0;
}
seqlen_a = gt_encseq_seqlength(mis->pvt->es1, mis->pvt->seqno_es1);
seqlen_b = gt_encseq_seqlength(mis->pvt->es2, mis->pvt->seqno_es2);
/* XXX: reuse buffers for performance improvement */
a = gt_malloc(seqlen_a * sizeof (char));
seqpos = gt_encseq_seqstartpos(mis->pvt->es1, mis->pvt->seqno_es1);
gt_encseq_extract_decoded(mis->pvt->es1, a, seqpos, seqpos + seqlen_a - 1);
b = gt_malloc(seqlen_b * sizeof (char));
seqpos = gt_encseq_seqstartpos(mis->pvt->es2, mis->pvt->seqno_es2);
gt_encseq_extract_decoded(mis->pvt->es1, b, seqpos, seqpos + seqlen_b - 1);
seq_a = gt_seq_new(a, seqlen_a, gt_encseq_alphabet(mis->pvt->es1));
seq_b = gt_seq_new(b, seqlen_b, gt_encseq_alphabet(mis->pvt->es2));
ali = gt_swalign(seq_a, seq_b, mis->pvt->sf);
mis->pvt->firstali = false;
if (ali && gt_alignment_get_length(ali) >= mis->pvt->min_len
&& gt_alignment_eval(ali) <= mis->pvt->max_edist) {
break;
}
gt_alignment_delete(ali);
gt_seq_delete(seq_a);
gt_seq_delete(seq_b);
gt_free(a);
gt_free(b);
}
arng = gt_alignment_get_urange(ali);
brng = gt_alignment_get_vrange(ali);
adesc = gt_encseq_description(mis->pvt->es1, &seqlen_a, mis->pvt->seqno_es1);
bdesc = gt_encseq_description(mis->pvt->es2, &seqlen_b, mis->pvt->seqno_es2);
*match = gt_match_sw_new("", "",
mis->pvt->seqno_es1,
mis->pvt->seqno_es2,
gt_alignment_get_length(ali),
gt_alignment_eval(ali),
arng.start, brng.start,
arng.end, brng.end,
GT_MATCH_DIRECT);
gt_match_set_seqid1_nt(*match, adesc, seqlen_a);
gt_match_set_seqid2_nt(*match, bdesc, seqlen_b);
gt_alignment_delete(ali);
gt_seq_delete(seq_a);
gt_seq_delete(seq_b);
gt_free(a);
gt_free(b);
return GT_MATCHER_STATUS_OK;
}
int gt_region_mapping_get_description(GtRegionMapping *rm, GtStr *desc,
GtStr *seqid, GtError *err)
{
int had_err = 0;
gt_error_check(err);
gt_assert(rm && desc && seqid);
if (rm->userawseq) {
gt_str_append_cstr(desc, "<rawseq>");
return 0;
}
had_err = update_seq_col_if_necessary(rm, seqid, err);
if (!had_err) {
if (gt_md5_seqid_has_prefix(gt_str_get(seqid))) {
had_err = gt_seq_col_md5_to_description(rm->seq_col, desc, seqid,
err);
}
return had_err;
}
if (!had_err) {
if (rm->usedesc) {
unsigned long filenum, seqnum;
gt_assert(rm->seqid2seqnum_mapping);
had_err = gt_seqid2seqnum_mapping_map(rm->seqid2seqnum_mapping,
gt_str_get(seqid), NULL, &seqnum,
&filenum, NULL, err);
if (!had_err) {
char *cdesc;
cdesc = gt_seq_col_get_description(rm->seq_col, filenum, seqnum);
gt_assert(cdesc);
gt_str_append_cstr(desc, cdesc);
gt_free(cdesc);
}
}
else if (rm->useseqno) {
unsigned long seqno = GT_UNDEF_ULONG;
gt_assert(rm->encseq);
if (1 != sscanf(gt_str_get(seqid), "seq%lu", &seqno)) {
gt_error_set(err, "seqid '%s' does not have the form 'seqX' "
"where X is a sequence number in the encoded "
"sequence", gt_str_get(seqid));
had_err = -1;
}
gt_assert(had_err || seqno != GT_UNDEF_ULONG);
if (!had_err && seqno >= gt_encseq_num_of_sequences(rm->encseq)) {
gt_error_set(err, "trying to access sequence %lu, but encoded"
"sequence contains only %lu sequences",
seqno, gt_encseq_num_of_sequences(rm->encseq));
had_err = -1;
}
if (!had_err) {
unsigned long desclen;
const char *edesc;
edesc = gt_encseq_description(rm->encseq, &desclen, seqno);
gt_str_append_cstr_nt(desc, edesc, desclen);
}
} else if (rm->matchdesc) {
const char *md5;
/* XXX: not beautiful, but works -- this may be LOTS faster */
had_err = gt_seq_col_grep_desc_md5(rm->seq_col, &md5, seqid, err);
if (!had_err) {
GtStr *md5_seqid = gt_str_new_cstr(md5);
had_err = gt_seq_col_md5_to_description(rm->seq_col, desc, md5_seqid,
err);
gt_str_delete(md5_seqid);
}
} else {
if (!had_err) {
char *cdesc;
cdesc = gt_seq_col_get_description(rm->seq_col, 0, 0);
gt_assert(cdesc);
gt_str_append_cstr(desc, cdesc);
gt_free(cdesc);
}
}
}
return had_err;
}
static int process_feature(GtLTRClusterStream *lcs,
const char *feature,
GtError *err)
{
GtArray *matches;
GtMatchIterator *mi = NULL;
GtMatch *match = NULL;
GtMatchIteratorStatus status;
GtEncseq *encseq;
unsigned long i;
int had_err = 0;
if (lcs->current_state != NULL) {
char tmp[BUFSIZ];
gt_free(*lcs->current_state);
(void) snprintf(tmp, BUFSIZ, "Clustering feature: %s", feature);
*lcs->current_state = gt_cstr_dup(tmp);
}
matches = gt_array_new(sizeof(GtMatch*));
encseq = (GtEncseq*) gt_hashmap_get(lcs->feat_to_encseq, feature);
gt_log_log("found encseq %p for feature %s", encseq, feature);
if (!had_err) {
mi = gt_match_iterator_last_new(encseq, encseq, lcs->match_score,
lcs->mismatch_cost,
lcs->gap_open_cost,
lcs->gap_ext_cost,
lcs->xdrop,
lcs->ydrop,
lcs->zdrop,
lcs->k,
lcs->mscoregapped,
lcs->mscoregapless, err);
if (mi != NULL) {
while ((status = gt_match_iterator_next(mi, &match, err))
!= GT_MATCHER_STATUS_END) {
if (status == GT_MATCHER_STATUS_OK) {
gt_array_add(matches, match);
} else {
gt_assert(status == GT_MATCHER_STATUS_ERROR);
had_err = -1;
break;
}
}
} else
had_err = -1;
}
if (!had_err) {
GtClusteredSet *cs;
GtHashmap *seqdesc2seqnum;
GtMatch *tmp_match;
const char *description;
char *output;
unsigned long desclen,
num_of_seq;
seqdesc2seqnum = gt_hashmap_new(GT_HASH_STRING, free_hash, NULL);
num_of_seq = gt_encseq_num_of_sequences(encseq);
for (i = 0; i < num_of_seq; i++) {
description = gt_encseq_description(encseq, &desclen, i);
output = gt_calloc((size_t) (desclen + 1), sizeof (char));
strncpy(output, description, (size_t) desclen);
output[desclen] = '\0';
gt_hashmap_add(seqdesc2seqnum, (void*) gt_cstr_dup(output),
(void*) (i + 1));
gt_free(output);
}
cs = gt_clustered_set_union_find_new(num_of_seq, err);
if (cs != NULL) {
if (cluster_sequences(matches, cs, seqdesc2seqnum, (unsigned) lcs->psmall,
(unsigned) lcs->plarge, encseq, err) != 0) {
had_err = -1;
}
if (!had_err) {
(void) cluster_annotate_nodes(cs, encseq, feature, lcs->nodes, err);
}
} else
had_err = -1;
for (i = 0; i < gt_array_size(matches); i++) {
tmp_match = *(GtMatch**) gt_array_get(matches, i);
gt_match_delete(tmp_match);
}
gt_array_delete(matches);
matches = NULL;
gt_hashmap_delete(seqdesc2seqnum);
gt_clustered_set_delete(cs, err);
}
gt_match_iterator_delete(mi);
return had_err;
}
static int cluster_annotate_nodes(GtClusteredSet *cs, GtEncseq *encseq,
const char *feature, GtArray *nodes,
GtError *err)
{
GtFeatureNodeIterator *fni;
GtFeatureNode *curnode = NULL, *tmp;
GtClusteredSetIterator *csi = NULL;
GtGenomeNode *gn;
GtHashmap *desc2node;
GtStr *seqid = NULL;
int had_err = 0;
unsigned long num_of_clusters, i, elm;
const char *fnt = NULL;
char buffer[BUFSIZ], *real_feature;
gt_error_check(err);
if ((strcmp(feature, "lLTR") == 0) || (strcmp(feature, "rLTR") == 0))
real_feature = gt_cstr_dup(gt_ft_long_terminal_repeat);
else
real_feature = gt_cstr_dup(feature);
desc2node = gt_hashmap_new(GT_HASH_STRING, free_hash, NULL);
for (i = 0; i < gt_array_size(nodes); i++) {
gn = *(GtGenomeNode**) gt_array_get(nodes, i);
if (gt_feature_node_try_cast(gn) == NULL)
continue;
fni = gt_feature_node_iterator_new((GtFeatureNode*) gn);
while ((curnode = gt_feature_node_iterator_next(fni)) != NULL) {
char header[BUFSIZ];
fnt = gt_feature_node_get_type(curnode);
if (strcmp(fnt, gt_ft_repeat_region) == 0) {
const char *rid;
unsigned long id;
seqid = gt_genome_node_get_seqid((GtGenomeNode*) curnode);
rid = gt_feature_node_get_attribute(curnode, "ID");
(void) sscanf(rid, "repeat_region%lu", &id);
(void) snprintf(buffer, BUFSIZ, "%s_%lu", gt_str_get(seqid), id);
} else if (strcmp(fnt, gt_ft_protein_match) == 0) {
GtRange range;
const char *attr;
attr = gt_feature_node_get_attribute(curnode, "name");
if (!attr)
continue;
if (strcmp(feature, attr) != 0)
continue;
range = gt_genome_node_get_range((GtGenomeNode*) curnode);
if ((range.end - range.start + 1) < 10UL)
continue;
(void) snprintf(header, BUFSIZ, "%s_%lu_%lu", buffer, range.start,
range.end);
gt_hashmap_add(desc2node, (void*) gt_cstr_dup(header), (void*) curnode);
} else if (strcmp(fnt, real_feature) == 0) {
GtRange range;
range = gt_genome_node_get_range((GtGenomeNode*) curnode);
if ((range.end - range.start + 1) < 10UL)
continue;
(void) snprintf(header, BUFSIZ, "%s_%lu_%lu", buffer, range.start,
range.end);
gt_hashmap_add(desc2node, (void*) gt_cstr_dup(header), (void*) curnode);
}
}
gt_feature_node_iterator_delete(fni);
}
gt_free(real_feature);
num_of_clusters = gt_clustered_set_num_of_clusters(cs, err);
for (i = 0; i < num_of_clusters; i++) {
csi = gt_clustered_set_get_iterator(cs, i ,err);
if (csi != NULL) {
while (!had_err && (gt_clustered_set_iterator_next(csi, &elm, err)
!= GT_CLUSTERED_SET_ITERATOR_STATUS_END)) {
char clid[BUFSIZ];
const char *encseqdesc;
char *encseqid;
unsigned long desclen;
encseqdesc = gt_encseq_description(encseq, &desclen, elm);
encseqid = gt_calloc((size_t) (desclen + 1), sizeof (char));
(void) strncpy(encseqid, encseqdesc, (size_t) desclen);
encseqid[desclen] = '\0';
tmp = (GtFeatureNode*) gt_hashmap_get(desc2node, (void*) encseqid);
(void) snprintf(clid, BUFSIZ, "%lu", i);
gt_feature_node_set_attribute(tmp, "clid", clid);
gt_free(encseqid);
}
}
gt_clustered_set_iterator_delete(csi, err);
csi = NULL;
}
gt_hashmap_delete(desc2node);
return had_err;
}