GtStr* gt_get_gtdata_path(const char *prog, GtError *err)
{
GtStr *path;
const char **defaultpath;
int had_err = 0;
gt_error_check(err);
gt_assert(prog);
path = gt_str_new();
had_err = gt_file_find_exec_in_path(path, prog, err);
if (!had_err) {
gt_assert(gt_str_length(path));
gt_str_append_cstr(path, GTDATADIR);
if (gt_file_exists_and_is_dir(gt_str_get(path)))
return path;
gt_str_set_length(path, gt_str_length(path) - strlen(GTDATADIR));
gt_str_append_cstr(path, UPDIR);
gt_str_append_cstr(path, GTDATADIR);
if (gt_file_exists_and_is_dir(gt_str_get(path)))
return path;
for (defaultpath = GTDATA_DEFAULT_PATHS; *defaultpath; defaultpath++) {
gt_str_reset(path);
gt_str_append_cstr(path, *defaultpath);
if (gt_file_exists_and_is_dir(gt_str_get(path)))
return path;
}
if (!gt_file_exists_and_is_dir(gt_str_get(path))) {
gt_error_set(err, "could not find gtdata%c directory", GT_PATH_SEPARATOR);
had_err = -1;
}
}
if (had_err) {
gt_str_delete(path);
return NULL;
}
return path;
}
int gt_condenseq_output_to_gff3(const GtCondenseq *condenseq,
GtError *err)
{
int had_err = 0;
GtUword idx,
name_len,
seqnum = 0, seqstart = 0, seqend = 0,
desclen;
GtStr *filename = NULL,
*id = gt_str_new_cstr("U"),
*name = gt_str_new_cstr("unique"),
*parent_unique = gt_str_new_cstr("U"),
*seqid = gt_str_new(),
*source = gt_str_new_cstr("Condenseq");
GtFile *outfile = NULL;
GtGFF3Visitor *gffv = NULL;
GtNodeVisitor *nodev = NULL;
GtFeatureNode *fnode = NULL;
GtGenomeNode *node = NULL;
GtRange range;
gt_assert(condenseq != NULL);
filename = gt_str_new_cstr(gt_condenseq_basefilename(condenseq));
name_len = gt_str_length(name);
gt_str_append_cstr(filename, ".gff3");
outfile = gt_file_new(gt_str_get(filename), "w", err);
nodev = gt_gff3_visitor_new(outfile);
gffv = (GtGFF3Visitor *) nodev;
gt_gff3_visitor_retain_id_attributes(gffv);
node = gt_feature_node_new(seqid, "experimental_feature", (GtUword) 1,
(GtUword) 1, GT_STRAND_BOTH);
fnode = (GtFeatureNode*) node;
gt_feature_node_set_source(fnode, source);
for (idx = 0; !had_err && idx < condenseq->udb_nelems; ++idx) {
GtCondenseqUnique uq = condenseq->uniques[idx];
if (seqend <= uq.orig_startpos) {
const char *desc;
gt_genome_node_delete(node);
seqnum = gt_condenseq_pos2seqnum(condenseq, uq.orig_startpos);
seqstart = gt_condenseq_seqstartpos(condenseq, seqnum);
seqend = seqstart + condenseq_seqlength_help(condenseq, seqnum, seqstart);
desc = gt_condenseq_description(condenseq, &desclen, seqnum);
gt_str_reset(seqid);
gt_str_append_cstr_nt(seqid, desc, desclen);
node = gt_feature_node_new(seqid, "experimental_feature", (GtUword) 1,
(GtUword) 1, GT_STRAND_BOTH);
fnode = (GtFeatureNode*) node;
gt_feature_node_set_source(fnode, source);
}
gt_str_set_length(name, name_len);
gt_str_append_uword(name, idx);
gt_str_set_length(id, (GtUword) 1);
gt_str_append_uword(id, idx);
gt_feature_node_set_attribute(fnode, "Name", gt_str_get(name));
gt_feature_node_set_attribute(fnode, "ID", gt_str_get(id));
/* 1 Based coordinates! */
range.start = uq.orig_startpos + 1 - seqstart;
range.end = uq.orig_startpos + uq.len - seqstart;
gt_genome_node_set_range(node, &range);
had_err = gt_genome_node_accept(node, nodev, err);
}
gt_str_reset(name);
gt_str_append_cstr(name, "link");
gt_str_reset(id);
gt_str_append_cstr(id, "L");
name_len = gt_str_length(name);
seqend = 0;
for (idx = 0; !had_err && idx < condenseq->ldb_nelems; ++idx) {
GtCondenseqLink link = condenseq->links[idx];
if (seqend <= link.orig_startpos) {
const char *desc;
gt_genome_node_delete(node);
seqnum = gt_condenseq_pos2seqnum(condenseq, link.orig_startpos);
seqstart = gt_condenseq_seqstartpos(condenseq, seqnum);
seqend = seqstart + condenseq_seqlength_help(condenseq, seqnum, seqstart);
desc = gt_condenseq_description(condenseq, &desclen, seqnum);
gt_str_reset(seqid);
gt_str_append_cstr_nt(seqid, desc, desclen);
node = gt_feature_node_new(seqid, "experimental_feature", (GtUword) 1,
(GtUword) 1, GT_STRAND_BOTH);
fnode = (GtFeatureNode*) node;
gt_feature_node_set_source(fnode, source);
}
gt_str_set_length(name, name_len);
gt_str_append_uword(name, idx);
gt_str_set_length(id, (GtUword) 1);
gt_str_append_uword(id, idx);
gt_feature_node_set_attribute(fnode, "Name", gt_str_get(name));
gt_feature_node_set_attribute(fnode, "ID", gt_str_get(id));
gt_str_set_length(parent_unique, (GtUword) 1);
gt_str_append_uword(parent_unique, link.unique_id);
gt_feature_node_set_attribute(fnode, "Derives_from",
gt_str_get(parent_unique));
/* 1 Based coordinates! */
range.start = link.orig_startpos + 1 - seqstart;
range.end = link.orig_startpos + link.len - seqstart;
gt_genome_node_set_range(node, &range);
had_err = gt_genome_node_accept(node, nodev, err);
}
gt_file_delete(outfile);
gt_genome_node_delete(node);
gt_node_visitor_delete(nodev);
gt_str_delete(filename);
gt_str_delete(id);
gt_str_delete(name);
gt_str_delete(parent_unique);
gt_str_delete(seqid);
gt_str_delete(source);
return had_err;
}
static void remove_terminal_comma(GtStr *str)
{
gt_assert(str && gt_str_length(str));
if (gt_str_get(str)[gt_str_length(str)-1] == ',')
gt_str_set_length(str, gt_str_length(str)-1);
}
static int hmmsearch_process_coarse_hits(
char *table_filename,
GtCondenseq *ces,
GtCondenseqHmmsearchArguments *arguments,
GtLogger *logger,
GtError *err) {
int had_err = 0;
GtStr *line = gt_str_new();
FILE *table = NULL;
GtSplitter *splitter = gt_splitter_new();
GtStr *query = gt_str_new(),
*fine_fasta_filename = gt_str_new_cstr("condenseq");
GtRBTree *sequences = NULL;
GtUword filecount = (GtUword) 1;
unsigned int querycount = 0;
const GtUword fine_fasta_name_length = gt_str_length(fine_fasta_filename);
const GtUword table_name_length = gt_str_length(arguments->outtable_filename);
table = gt_xfopen(table_filename, "r");
sequences = gt_rbtree_new(hmmsearch_cmp_seqnum,
hmmsearch_tree_free_node, NULL);
while (!had_err && gt_str_read_next_line(line, table) == 0) {
char *c_line = gt_str_get(line);
GtUword uid;
const GtUword target_column = 0,
query_column = (GtUword) 3;
if (c_line[0] != '#') {
gt_splitter_split_non_empty(splitter, c_line, gt_str_length(line), ' ');
gt_assert(gt_splitter_size(splitter) == (GtUword) 23);
if (sscanf(gt_splitter_get_token(splitter, target_column),
GT_WU, &uid) != 1) {
gt_error_set(err, "couldn't parse target number: %s",
gt_splitter_get_token(splitter, target_column));
had_err = -1;
}
if (gt_str_length(query) == 0 ||
strcmp(gt_str_get(query),
gt_splitter_get_token(splitter, query_column)) != 0) {
gt_str_set(query, gt_splitter_get_token(splitter, query_column));
gt_logger_log(logger, "new query: %s", gt_str_get(query));
querycount++;
}
if (!had_err && querycount == arguments->max_queries) {
hmmsearch_create_fine_fas(fine_fasta_filename, sequences, ces);
if (table_name_length != 0)
gt_str_append_uword(arguments->outtable_filename, filecount++);
had_err =
hmmsearch_call_fine_search(table_name_length != 0 ?
arguments->outtable_filename :
NULL,
gt_str_get(fine_fasta_filename),
gt_str_get(arguments->hmmsearch_path),
gt_str_get(arguments->hmm),
logger, err);
gt_rbtree_clear(sequences);
gt_str_set_length(fine_fasta_filename, fine_fasta_name_length);
if (table_name_length != 0)
gt_str_set_length(arguments->outtable_filename, table_name_length);
querycount = 0;
}
if (!had_err) {
if (gt_condenseq_each_redundant_seq(ces, uid,
hmmsearch_process_seq,
sequences, err) == 0) {
had_err = -1;
}
}
gt_splitter_reset(splitter);
}
gt_str_reset(line);
}
gt_splitter_delete(splitter);
gt_str_delete(line);
gt_str_delete(query);
gt_xfclose(table);
if (!had_err) {
hmmsearch_create_fine_fas(fine_fasta_filename, sequences, ces);
if (table_name_length != 0)
gt_str_append_uword(arguments->outtable_filename, filecount++);
had_err =
hmmsearch_call_fine_search(table_name_length != 0 ?
arguments->outtable_filename :
NULL,
gt_str_get(fine_fasta_filename),
gt_str_get(arguments->hmmsearch_path),
gt_str_get(arguments->hmm),
logger, err);
}
gt_log_log("created " GT_WU " files", filecount);
gt_rbtree_delete(sequences);
gt_str_delete(fine_fasta_filename);
return had_err;
}
