static int proc_any_char(GtIO *obo_file, GtStr *capture, bool be_permissive,
GtError *err)
{
gt_error_check(err);
gt_assert(obo_file && capture);
if (!any_char(obo_file, be_permissive)) {
if (gt_io_peek(obo_file) == GT_END_OF_FILE) {
gt_error_set(err, "file \"%s\": line %lu: unexpected end-of-file",
gt_io_get_filename(obo_file), gt_io_get_line_number(obo_file));
}
else if ((gt_io_peek(obo_file) == GT_CARRIAGE_RETURN) ||
(gt_io_peek(obo_file) == GT_END_OF_LINE)) {
gt_error_set(err, "file \"%s\": line %lu: unexpected newline",
gt_io_get_filename(obo_file), gt_io_get_line_number(obo_file));
}
else {
gt_error_set(err, "file \"%s\": line %lu: unexpected character '%c'",
gt_io_get_filename(obo_file), gt_io_get_line_number(obo_file),
gt_io_peek(obo_file));
}
return -1;
}
gt_str_append_char(capture, gt_io_next(obo_file));
return 0;
}
int gt_io_expect(GtIO *io, char expected_char, GtError *err)
{
char cc;
gt_error_check(err);
cc = gt_io_next(io);
if (cc != expected_char) {
if (expected_char == GT_END_OF_LINE && cc == GT_CARRIAGE_RETURN) {
if (gt_io_peek(io) == GT_END_OF_LINE)
gt_io_next(io);
return 0;
}
if (expected_char == GT_END_OF_FILE) {
gt_error_set(err, "file \"%s\": line %lu: expected end-of-file, got '%c'",
gt_io_get_filename(io), gt_io_get_line_number(io), cc);
}
else if ((cc == GT_CARRIAGE_RETURN) || (cc == GT_END_OF_LINE)) {
gt_error_set(err, "file \"%s\": line %lu: expected character '%c', got "
"newline", gt_io_get_filename(io), gt_io_get_line_number(io),
expected_char);
}
else {
gt_error_set(err, "file \"%s\": line %lu: expected character '%c', got "
"'%c'", gt_io_get_filename(io), gt_io_get_line_number(io),
expected_char, cc);
}
return -1;
}
return 0;
}
static int parse_bed_range(GtRange *range, GtStr *start, GtStr *end,
GtWord offset, GtIO *bed_file, bool thick,
GtError *err)
{
int had_err;
gt_error_check(err);
had_err = gt_parse_range(range, gt_str_get(start), gt_str_get(end),
gt_io_get_line_number(bed_file),
gt_io_get_filename(bed_file), err);
/* BED has a weird numbering scheme: positions are 0-based, but the end
position is not part of the feature. Transform to 1-based coordinates. */
range->start++;
/* Ranges defining a 'thick' region sometimes come with length 0 to
designate that there are no thick regions. So do not fail here and
handle that case later. */
if (!thick) {
if (!had_err && range->start > range->end) {
gt_error_set(err, "file \"%s\": line "GT_WU": BED feature has length 0",
gt_io_get_filename(bed_file),
gt_io_get_line_number(bed_file));
had_err = -1;
}
}
if (offset)
*range = gt_range_offset(range, offset);
return had_err;
}
static int process_blocks(GtBEDParser *bed_parser, GtFeatureNode *fn,
unsigned long block_count, GtStr *block_sizes,
GtStr *block_starts, GtIO *bed_file, GtError *err)
{
GtSplitter *size_splitter = NULL , *start_splitter = NULL;
int had_err = 0;
gt_error_check(err);
gt_assert(fn && block_count && block_sizes && block_starts);
if (!gt_str_length(block_sizes)) {
gt_error_set(err,
"file \"%s\": line %lu: blockCount given without blockSizes",
gt_io_get_filename(bed_file),
gt_io_get_line_number(bed_file));
had_err = -1;
}
if (!had_err && !gt_str_length(block_starts)) {
gt_error_set(err,
"file \"%s\": line %lu: blockCount given without blockStarts",
gt_io_get_filename(bed_file),
gt_io_get_line_number(bed_file));
had_err = -1;
}
if (!had_err) {
/* remove terminal commas found in real-world BED files */
remove_terminal_comma(block_sizes);
remove_terminal_comma(block_starts);
}
if (!had_err) {
size_splitter = gt_splitter_new();
gt_splitter_split(size_splitter, gt_str_get(block_sizes),
gt_str_length(block_sizes), ',');
if (gt_splitter_size(size_splitter) != block_count) {
gt_error_set(err, "file \"%s\": line %lu: blockSizes column does not "
"have blockCount=%lu many comma separated fields",
gt_io_get_filename(bed_file),
gt_io_get_line_number(bed_file), block_count);
had_err = -1;
}
}
if (!had_err) {
start_splitter = gt_splitter_new();
gt_splitter_split(start_splitter, gt_str_get(block_starts),
gt_str_length(block_starts), ',');
if (gt_splitter_size(start_splitter) != block_count) {
gt_error_set(err, "file \"%s\": line %lu: blockStarts column does not "
"have " "blockCount=%lu many comma separated fields",
gt_io_get_filename(bed_file),
gt_io_get_line_number(bed_file), block_count);
had_err = -1;
}
}
if (!had_err) {
had_err = create_block_features(bed_parser, fn, block_count, size_splitter,
start_splitter, bed_file, err);
}
gt_splitter_delete(start_splitter);
gt_splitter_delete(size_splitter);
return had_err;
}
static int create_block_features(GtBEDParser *bed_parser, GtFeatureNode *fn,
GtUword block_count,
GtSplitter *size_splitter,
GtSplitter *start_splitter, GtIO *bed_file,
GtError *err)
{
GtUword i;
int had_err = 0;
gt_assert(fn && block_count && size_splitter && start_splitter);
gt_assert(gt_splitter_size(size_splitter) == block_count);
gt_assert(gt_splitter_size(start_splitter) == block_count);
for (i = 0; !had_err && i < block_count; i++) {
GtUword block_size, block_start, start, end;
GtGenomeNode *block;
const char *name;
if (gt_parse_uword(&block_size, gt_splitter_get_token(size_splitter, i))) {
gt_error_set(err,
"file \"%s\": line "GT_WU": could not parse blockSize '%s'",
gt_io_get_filename(bed_file),
gt_io_get_line_number(bed_file),
gt_splitter_get_token(size_splitter, i));
had_err = -1;
}
if (!had_err && gt_parse_uword(&block_start,
gt_splitter_get_token(start_splitter, i))) {
gt_error_set(err, "file \"%s\": line "GT_WU": could not parse blockStart "
"'%s'", gt_io_get_filename(bed_file),
gt_io_get_line_number(bed_file),
gt_splitter_get_token(start_splitter, i));
had_err = -1;
}
if (!had_err) {
start = gt_genome_node_get_start((GtGenomeNode*) fn) + block_start;
end = start + block_size - 1;
block = gt_feature_node_new(gt_genome_node_get_seqid((GtGenomeNode*) fn),
bed_parser->block_type
? bed_parser->block_type
: BED_BLOCK_TYPE,
start, end, gt_feature_node_get_strand(fn));
if ((name = gt_feature_node_get_attribute(fn, GT_GFF_NAME))) {
gt_feature_node_add_attribute((GtFeatureNode*) block, GT_GFF_NAME,
name);
}
gt_feature_node_set_score((GtFeatureNode*) block,
gt_feature_node_get_score(fn));
gt_feature_node_set_strand((GtFeatureNode*) block,
gt_feature_node_get_strand(fn));
gt_feature_node_add_child(fn, (GtFeatureNode*) block);
}
}
return had_err;
}
static int header(GtOBOParseTree *obo_parse_tree, GtIO *obo_file, GtError *err)
{
GtStr *tag, *value;
int had_err;
gt_error_check(err);
gt_assert(obo_parse_tree && obo_file);
tag = gt_str_new();
value = gt_str_new();
do {
gt_str_reset(tag);
gt_str_reset(value);
had_err = tag_line(obo_file, tag, value, err);
if (!had_err) {
obo_header_add(obo_parse_tree->obo_header, gt_str_get(tag),
gt_str_get(value));
}
} while (!had_err && any_char(obo_file, false));
if (!had_err) {
had_err = obo_header_validate(obo_parse_tree->obo_header,
gt_io_get_filename(obo_file), err);
}
gt_str_delete(value);
gt_str_delete(tag);
return had_err;
}
static int gt_fasta_reader_rec_run(GtFastaReader *fasta_reader,
GtFastaReaderProcDescription
proc_description,
GtFastaReaderProcSequencePart
proc_sequence_part,
GtFastaReaderProcSequenceLength
proc_sequence_length, void *data,
GtError *err)
{
GtFastaReaderRec *fr = gt_fasta_reader_rec_cast(fasta_reader);
GtStr *description, *sequence;
int had_err = 0;
gt_error_check(err);
/* at least one function has to be defined */
gt_assert(proc_description || proc_sequence_part || proc_sequence_length);
/* init */
description = gt_str_new();
sequence = gt_str_new();
/* make sure file is not empty */
if (!gt_io_has_char(fr->seqio)) {
gt_error_set(err, "sequence file \"%s\" is empty",
gt_io_get_filename(fr->seqio));
had_err = -1;
}
/* parse file */
while (!had_err && gt_io_has_char(fr->seqio)) {
/* reset */
gt_str_reset(description);
gt_str_reset(sequence);
/* parse entry */
had_err = parse_fasta_entry(description, sequence, fr->seqio, err);
/* process entry */
if (!had_err && proc_description) {
had_err = proc_description(gt_str_get(description),
gt_str_length(description), data, err);
}
if (!had_err && proc_sequence_part) {
had_err = proc_sequence_part(gt_str_get(sequence),
gt_str_length(sequence), data, err);
}
if (!had_err && proc_sequence_length)
had_err = proc_sequence_length(gt_str_length(sequence), data, err);
}
/* free */
gt_str_delete(description);
gt_str_delete(sequence);
return had_err;
}
static int parse_bed_range(GtRange *range, GtStr *start, GtStr *end,
long offset, GtIO *bed_file, GtError *err)
{
int had_err;
gt_error_check(err);
had_err = gt_parse_range(range, gt_str_get(start), gt_str_get(end),
gt_io_get_line_number(bed_file),
gt_io_get_filename(bed_file), err);
/* BED has a weird numbering scheme: positions are 0-based, but the end
position is not part of the feature. Transform to 1-based coordinates. */
range->start++;
if (!had_err && range->start > range->end) {
gt_error_set(err, "file \"%s\": line %lu: BED feature has length 0",
gt_io_get_filename(bed_file), gt_io_get_line_number(bed_file));
had_err = -1;
}
if (offset)
*range = gt_range_offset(range, offset);
return had_err;
}
static int skip_blanks(GtIO *bed_file, GtError *err)
{
gt_error_check(err);
if (!bed_separator(bed_file)) {
gt_error_set(err, "file \"%s\": line %lu: expected blank or tabulator, got "
"'%c'", gt_io_get_filename(bed_file),
gt_io_get_line_number(bed_file), gt_io_peek(bed_file));
return -1;
}
while (bed_separator(bed_file))
gt_io_next(bed_file);
return 0;
}
static int track_rest(GtBEDParser *bed_parser, GtIO *bed_file, GtError *err)
{
char cc;
int had_err = 0;
gt_error_check(err);
bed_parser->offset = 0; /* reset offset for new track line */
if (bed_separator(bed_file)) /* skip to first attribute=value pair */
had_err = skip_blanks(bed_file, err);
while (!had_err &&
(cc = gt_io_peek(bed_file)) != GT_END_OF_LINE &&
cc != GT_CARRIAGE_RETURN) {
/* parse attribute */
word(bed_parser->word, bed_file);
had_err = gt_io_expect(bed_file, PAIR_SEPARATOR, err);
/* parse value */
if (!had_err) {
if (gt_io_peek(bed_file) == QUOTE_CHAR)
had_err = quoted_word(bed_parser->another_word, bed_file, err);
else
word(bed_parser->another_word, bed_file);
}
/* process offset if necessary */
if (!had_err && !strcmp(gt_str_get(bed_parser->word), OFFSET_KEYWORD)) {
if (gt_parse_word(&bed_parser->offset,
gt_str_get(bed_parser->another_word))) {
gt_error_set(err,
"file \"%s\": line "GT_WU": could not parse offset value "
"'%s'", gt_io_get_filename(bed_file),
gt_io_get_line_number(bed_file),
gt_str_get(bed_parser->another_word));
had_err = -1;
}
}
/* skip blanks up to next attribute or end-of-line */
if (!had_err && bed_separator(bed_file))
had_err = skip_blanks(bed_file, err);
}
/* the end of the line should now be reached */
if (!had_err)
had_err = gt_io_expect(bed_file, GT_END_OF_LINE, err);
return had_err;
}
static int parse_fasta_description(GtStr *description, GtIO *seqio,
GtError *err)
{
int rval;
char cc;
gt_error_check(err);
gt_assert(description && seqio);
rval = gt_io_get_char(seqio, &cc);
gt_assert(!rval); /* was checked earlier */
/* make sure we got a proper fasta description */
if (cc != FASTA_SEPARATOR) {
gt_error_set(err, "the first character of fasta file \"%s\" has to be '%c'",
gt_io_get_filename(seqio), FASTA_SEPARATOR);
return -1;
}
/* read description */
while (!gt_io_get_char(seqio, &cc) && cc != '\n')
gt_str_append_char(description, cc);
return 0;
}
static int gt_xrf_abbr_parse_tree_tag_line(GtIO *xrf_abbr_file, GtStr *tag,
GtStr *value, GtError *err)
{
int had_err = 0;
gt_error_check(err);
gt_log_log("tag");
gt_assert(xrf_abbr_file && tag && value);
do {
had_err = gt_xrf_abbr_parse_tree_proc_any_char(xrf_abbr_file, tag,
false, err);
} while (!had_err && gt_xrf_abbr_parse_tree_any_char(xrf_abbr_file, false));
if (!had_err)
had_err = gt_io_expect(xrf_abbr_file, XRF_SEPARATOR_CHAR, err);
while (!had_err && gt_io_peek(xrf_abbr_file) == XRF_BLANK_CHAR)
gt_io_next(xrf_abbr_file);
if (!had_err) {
do {
had_err = gt_xrf_abbr_parse_tree_proc_any_char(xrf_abbr_file, value,
true, err);
} while (!had_err && gt_xrf_abbr_parse_tree_any_char(xrf_abbr_file, true));
}
if (!had_err) {
if (gt_io_peek(xrf_abbr_file) == XRF_COMMENT_CHAR)
had_err = gt_xrf_abbr_parse_tree_comment_line(xrf_abbr_file, err);
else
had_err = gt_io_expect(xrf_abbr_file, GT_END_OF_LINE, err);
}
if (!had_err && !gt_xrf_abbr_parse_tree_valid_label(gt_str_get(tag))) {
gt_warning("file \"%s\": line "GT_WU": unknown label \"%s\"",
gt_io_get_filename(xrf_abbr_file),
gt_io_get_line_number(xrf_abbr_file),
gt_str_get(tag));
}
gt_log_log("parsed line %s/%s", gt_str_get(tag), gt_str_get(value));
return had_err;
}
static int bed_rest(GtBEDParser *bed_parser, GtIO *bed_file, GtError *err)
{
GtUword block_count = 0;
GtGenomeNode *gn = NULL;
GtRange range;
GtStr *seqid;
int had_err;
gt_error_check(err);
/* column 1.: chrom */
seqid = get_seqid(bed_parser);
had_err = skip_blanks(bed_file, err);
/* column 2.: chromStart */
if (!had_err) {
word(bed_parser->word, bed_file);
had_err = skip_blanks(bed_file, err);
}
/* column 3.: chromEnd */
if (!had_err) {
word(bed_parser->another_word, bed_file);
had_err = parse_bed_range(&range, bed_parser->word,
bed_parser->another_word, bed_parser->offset,
bed_file, false, err);
}
if (!had_err) {
/* add region */
gt_region_node_builder_add_region(bed_parser->region_node_builder,
gt_str_get(seqid), range);
/* create feature */
gn = gt_feature_node_new(seqid,
bed_parser->feature_type
? bed_parser->feature_type
: BED_FEATURE_TYPE,
range.start, range.end, GT_STRAND_BOTH);
gt_queue_add(bed_parser->feature_nodes, gn);
if (bed_separator(bed_file))
had_err = skip_blanks(bed_file, err);
}
/* optional column 4.: name */
if (!had_err) {
word(bed_parser->word, bed_file);
if (gt_str_length(bed_parser->word)) {
gt_feature_node_add_attribute((GtFeatureNode*) gn, GT_GFF_NAME,
gt_str_get(bed_parser->word));
}
if (bed_separator(bed_file))
had_err = skip_blanks(bed_file, err);
}
/* optional column 5.: score */
if (!had_err) {
word(bed_parser->word, bed_file);
if (gt_str_length(bed_parser->word)) {
bool score_is_defined;
float score_value;
had_err = gt_parse_score(&score_is_defined, &score_value,
gt_str_get(bed_parser->word),
gt_io_get_line_number(bed_file),
gt_io_get_filename(bed_file), err);
if (!had_err && score_is_defined)
gt_feature_node_set_score((GtFeatureNode*) gn, score_value);
}
}
if (!had_err && bed_separator(bed_file))
had_err = skip_blanks(bed_file, err);
/* optional column 6.: strand */
if (!had_err) {
word(bed_parser->word, bed_file);
if (gt_str_length(bed_parser->word)) {
GtStrand strand;
had_err = gt_parse_strand(&strand, gt_str_get(bed_parser->word),
gt_io_get_line_number(bed_file),
gt_io_get_filename(bed_file), err);
if (!had_err)
gt_feature_node_set_strand((GtFeatureNode*) gn, strand);
}
}
if (!had_err && bed_separator(bed_file))
had_err = skip_blanks(bed_file, err);
/* optional column 7.: thickStart */
if (!had_err) {
word(bed_parser->word, bed_file);
if (bed_separator(bed_file))
had_err = skip_blanks(bed_file, err);
}
/* optional column 8.: thickEnd */
if (!had_err) {
word(bed_parser->another_word, bed_file);
if (gt_str_length(bed_parser->another_word)) {
gt_assert(gt_str_length(bed_parser->word));
/* got a thickStart and a thickEnd -> construct corresponding feature */
had_err = parse_bed_range(&range, bed_parser->word,
bed_parser->another_word, bed_parser->offset,
bed_file, true, err);
if (!had_err && range.start <= range.end)
construct_thick_feature(bed_parser, (GtFeatureNode*) gn, range);
}
}
if (!had_err && bed_separator(bed_file))
had_err = skip_blanks(bed_file, err);
/* optional column 9.: itemRgb */
if (!had_err) {
word(bed_parser->word, bed_file);
/* we do not use the RGB values */
if (bed_separator(bed_file))
had_err = skip_blanks(bed_file, err);
}
/* optional column 10.: blockCount */
if (!had_err) {
word(bed_parser->word, bed_file);
if (gt_str_length(bed_parser->word)) {
if (gt_parse_uword(&block_count, gt_str_get(bed_parser->word))) {
gt_error_set(err,
"file \"%s\": line "GT_WU": could not parse blockCount",
gt_io_get_filename(bed_file),
gt_io_get_line_number(bed_file));
had_err = -1;
}
else {
/* reset to parse/process blockSizes and blockStarts properly */
gt_str_reset(bed_parser->word);
gt_str_reset(bed_parser->another_word);
}
}
}
if (!had_err && bed_separator(bed_file))
had_err = skip_blanks(bed_file, err);
/* optional column 11.: blockSizes */
if (!had_err) {
word(bed_parser->word, bed_file);
if (bed_separator(bed_file))
had_err = skip_blanks(bed_file, err);
}
/* optional column 12.: blockStarts */
if (!had_err) {
word(bed_parser->another_word, bed_file);
if (bed_separator(bed_file))
had_err = skip_blanks(bed_file, err);
}
/* process blocks if necessary */
if (!had_err && block_count) {
had_err = process_blocks(bed_parser, (GtFeatureNode*) gn, block_count,
bed_parser->word, bed_parser->another_word,
bed_file, err);
}
/* the end of the line should now be reached */
if (!had_err)
had_err = gt_io_expect(bed_file, GT_END_OF_LINE, err);
return had_err;
}