int gt_parse_intermediate_output(GthInput *input,
GthSAProcessFunc saprocessfunc, void *data,
const char *outputfilename,
GtFile *intermediate_fp, GtError *err)
{
GtStr *line;
XML_Parser parser;
enum XML_Error error;
Parseinfo parseinfo;
int had_err = 0;
gt_error_check(err);
/* init */
line = gt_str_new();
/* create parser */
parser = XML_ParserCreate(NULL);
/* init parse info structure */
parseinfo.currentSA = NULL;
parseinfo.databuf = gt_str_new();
parseinfo.genomicfilename = gt_str_new();
parseinfo.referencefilename = gt_str_new();
parseinfo.linenumber = 0;
parseinfo.outputfilename = outputfilename;
parseinfo.input = input;
parseinfo.saprocessfunc = saprocessfunc;
parseinfo.data = data;
parseinfo.err = err;
/* set element handler */
XML_SetElementHandler(parser, start_element_handler, end_element_handler);
/* set character data handler */
XML_SetCharacterDataHandler(parser, character_data_handler);
/* register the parse info structure */
XML_SetUserData(parser, &parseinfo);
/* parse the intermediate output line by line */
while (gt_str_read_next_line_generic(line, intermediate_fp) != EOF) {
parseinfo.linenumber++;
if (XML_Parse(parser, gt_str_get(line), gt_str_length(line), false) ==
XML_STATUS_ERROR) {
error = XML_GetErrorCode(parser);
gt_error_set(err, "an error occurred parsing line "GT_WU
" of file \"%s\": %s",
parseinfo.linenumber, outputfilename,
XML_ErrorString(error));
had_err = -1;
}
/* reset line buffer */
gt_str_reset(line);
}
if (!had_err) {
/* finish parsing */
if (XML_Parse(parser, NULL, 0, true) == XML_STATUS_ERROR) {
error = XML_GetErrorCode(parser);
gt_error_set(err, "an error occurred while finishing the parsing of file "
"\"%s\": %s", outputfilename, XML_ErrorString(error));
had_err = -1;
}
}
/* free space */
XML_ParserFree(parser);
gt_str_delete(line);
gt_str_delete(parseinfo.databuf);
gt_str_delete(parseinfo.genomicfilename);
gt_str_delete(parseinfo.referencefilename);
return had_err;
}
int gt_gtf_parser_parse(GtGTFParser *parser, GtQueue *genome_nodes,
GtStr *filenamestr, GtFile *fpin, bool be_tolerant,
GtError *err)
{
GtStr *seqid_str, *source_str, *line_buffer;
char *line;
size_t line_length;
GtUword i, line_number = 0;
GtGenomeNode *gn;
GtRange range;
GtPhase phase_value;
GtStrand gt_strand_value;
GtSplitter *splitter, *attribute_splitter;
float score_value;
char *seqname,
*source,
*feature,
*start,
*end,
*score,
*strand,
*frame,
*attributes,
*token,
*gene_id,
*gene_name = NULL,
*transcript_id,
*transcript_name = NULL,
**tokens;
GtHashmap *transcript_id_hash; /* map from transcript id to array of genome
nodes */
GtArray *gt_genome_node_array;
ConstructionInfo cinfo;
GTF_feature_type gtf_feature_type;
GT_UNUSED bool gff_type_is_valid = false;
const char *type = NULL;
const char *filename;
bool score_is_defined;
int had_err = 0;
gt_assert(parser && genome_nodes);
gt_error_check(err);
filename = gt_str_get(filenamestr);
/* alloc */
line_buffer = gt_str_new();
splitter = gt_splitter_new(),
attribute_splitter = gt_splitter_new();
#define HANDLE_ERROR \
if (had_err) { \
if (be_tolerant) { \
fprintf(stderr, "skipping line: %s\n", gt_error_get(err)); \
gt_error_unset(err); \
gt_str_reset(line_buffer); \
had_err = 0; \
continue; \
} \
else { \
had_err = -1; \
break; \
} \
}
while (gt_str_read_next_line_generic(line_buffer, fpin) != EOF) {
line = gt_str_get(line_buffer);
line_length = gt_str_length(line_buffer);
line_number++;
had_err = 0;
if (line_length == 0) {
gt_warning("skipping blank line " GT_WU " in file \"%s\"", line_number,
filename);
}
else if (line[0] == '#') {
/* storing comment */
if (line_length >= 2 && line[1] == '#')
gn = gt_comment_node_new(line+2); /* store '##' line as '#' line */
else
gn = gt_comment_node_new(line+1);
gt_genome_node_set_origin(gn, filenamestr, line_number);
gt_queue_add(genome_nodes, gn);
}
else {
/* process tab delimited GTF line */
gt_splitter_reset(splitter);
gt_splitter_split(splitter, line, line_length, '\t');
if (gt_splitter_size(splitter) != 9UL) {
gt_error_set(err, "line " GT_WU " in file \"%s\" contains " GT_WU
" tab (\\t) " "separated fields instead of 9", line_number,
filename,
gt_splitter_size(splitter));
had_err = -1;
break;
}
tokens = gt_splitter_get_tokens(splitter);
seqname = tokens[0];
source = tokens[1];
feature = tokens[2];
start = tokens[3];
end = tokens[4];
score = tokens[5];
strand = tokens[6];
frame = tokens[7];
attributes = tokens[8];
/* parse feature */
if (GTF_feature_type_get(>f_feature_type, feature) == -1) {
/* we skip unknown features */
fprintf(stderr, "skipping line " GT_WU " in file \"%s\": unknown "
"feature: \"%s\"\n", line_number, filename, feature);
gt_str_reset(line_buffer);
continue;
}
/* translate into GFF3 feature type */
switch (gtf_feature_type) {
case GTF_CDS:
case GTF_stop_codon:
gff_type_is_valid = gt_type_checker_is_valid(parser->type_checker,
gt_ft_CDS);
type = gt_ft_CDS;
break;
case GTF_exon:
gff_type_is_valid = gt_type_checker_is_valid(parser->type_checker,
gt_ft_exon);
type = gt_ft_exon;
}
gt_assert(gff_type_is_valid);
/* parse the range */
had_err = gt_parse_range(&range, start, end, line_number, filename, err);
HANDLE_ERROR;
/* process seqname (we have to do it here because we need the range) */
gt_region_node_builder_add_region(parser->region_node_builder, seqname,
range);
/* parse the score */
had_err = gt_parse_score(&score_is_defined, &score_value, score,
line_number, filename, err);
HANDLE_ERROR;
/* parse the strand */
had_err = gt_parse_strand(>_strand_value, strand, line_number, filename,
err);
HANDLE_ERROR;
/* parse the frame */
had_err = gt_parse_phase(&phase_value, frame, line_number, filename, err);
HANDLE_ERROR;
/* parse the attributes */
gt_splitter_reset(attribute_splitter);
gene_id = NULL;
transcript_id = NULL;
gt_splitter_split(attribute_splitter, attributes, strlen(attributes),
';');
for (i = 0; i < gt_splitter_size(attribute_splitter); i++) {
token = gt_splitter_get_token(attribute_splitter, i);
/* skip leading blanks */
while (*token == ' ')
token++;
/* look for the two mandatory attributes */
if (strncmp(token, GENE_ID_ATTRIBUTE, strlen(GENE_ID_ATTRIBUTE)) == 0) {
if (strlen(token) + 2 < strlen(GENE_ID_ATTRIBUTE)) {
gt_error_set(err, "missing value to attribute \"%s\" on line "
GT_WU "in file \"%s\"", GENE_ID_ATTRIBUTE, line_number,
filename);
had_err = -1;
}
HANDLE_ERROR;
gene_id = token + strlen(GENE_ID_ATTRIBUTE) + 1;
}
else if (strncmp(token, TRANSCRIPT_ID_ATTRIBUTE,
strlen(TRANSCRIPT_ID_ATTRIBUTE)) == 0) {
if (strlen(token) + 2 < strlen(TRANSCRIPT_ID_ATTRIBUTE)) {
gt_error_set(err, "missing value to attribute \"%s\" on line "
GT_WU "in file \"%s\"", TRANSCRIPT_ID_ATTRIBUTE,
line_number, filename);
had_err = -1;
}
HANDLE_ERROR;
transcript_id = token + strlen(TRANSCRIPT_ID_ATTRIBUTE) + 1;
}
else if (strncmp(token, GENE_NAME_ATTRIBUTE,
strlen(GENE_NAME_ATTRIBUTE)) == 0) {
if (strlen(token) + 2 < strlen(GENE_NAME_ATTRIBUTE)) {
gt_error_set(err, "missing value to attribute \"%s\" on line "
GT_WU "in file \"%s\"", GENE_NAME_ATTRIBUTE,
line_number, filename);
had_err = -1;
}
HANDLE_ERROR;
gene_name = token + strlen(GENE_NAME_ATTRIBUTE) + 1;
/* for output we want to strip quotes */
if (*gene_name == '"')
gene_name++;
if (gene_name[strlen(gene_name)-1] == '"')
gene_name[strlen(gene_name)-1] = '\0';
}
else if (strncmp(token, TRANSCRIPT_NAME_ATTRIBUTE,
strlen(TRANSCRIPT_NAME_ATTRIBUTE)) == 0) {
if (strlen(token) + 2 < strlen(TRANSCRIPT_NAME_ATTRIBUTE)) {
gt_error_set(err, "missing value to attribute \"%s\" on line "
GT_WU "in file \"%s\"", TRANSCRIPT_NAME_ATTRIBUTE,
line_number, filename);
had_err = -1;
}
HANDLE_ERROR;
transcript_name = token + strlen(TRANSCRIPT_NAME_ATTRIBUTE) + 1;
/* for output we want to strip quotes */
if (*transcript_name == '"')
transcript_name++;
if (transcript_name[strlen(transcript_name)-1] == '"')
transcript_name[strlen(transcript_name)-1] = '\0';
}
}
/* check for the mandatory attributes */
if (!gene_id) {
gt_error_set(err, "missing attribute \"%s\" on line " GT_WU
" in file \"%s\"", GENE_ID_ATTRIBUTE, line_number,
filename);
had_err = -1;
}
HANDLE_ERROR;
if (!transcript_id) {
gt_error_set(err, "missing attribute \"%s\" on line " GT_WU
" in file \"%s\"", TRANSCRIPT_ID_ATTRIBUTE, line_number,
filename);
had_err = -1;
}
HANDLE_ERROR;
/* process the mandatory attributes */
if (!(transcript_id_hash = gt_hashmap_get(parser->gene_id_hash,
gene_id))) {
transcript_id_hash = gt_hashmap_new(GT_HASH_STRING, gt_free_func,
(GtFree) gt_array_delete);
gt_hashmap_add(parser->gene_id_hash, gt_cstr_dup(gene_id),
transcript_id_hash);
}
gt_assert(transcript_id_hash);
if (!(gt_genome_node_array = gt_hashmap_get(transcript_id_hash,
transcript_id))) {
gt_genome_node_array = gt_array_new(sizeof (GtGenomeNode*));
gt_hashmap_add(transcript_id_hash, gt_cstr_dup(transcript_id),
gt_genome_node_array);
}
gt_assert(gt_genome_node_array);
/* save optional gene_name and transcript_name attributes */
if (transcript_name
&& !gt_hashmap_get(parser->transcript_id_to_name_mapping,
transcript_id)) {
gt_hashmap_add(parser->transcript_id_to_name_mapping,
gt_cstr_dup(transcript_id),
gt_cstr_dup(transcript_name));
}
if (gene_name && !gt_hashmap_get(parser->gene_id_to_name_mapping,
gene_id)) {
gt_hashmap_add(parser->gene_id_to_name_mapping,
gt_cstr_dup(gene_id),
gt_cstr_dup(gene_name));
}
/* get seqid */
seqid_str = gt_hashmap_get(parser->seqid_to_str_mapping, seqname);
if (!seqid_str) {
seqid_str = gt_str_new_cstr(seqname);
gt_hashmap_add(parser->seqid_to_str_mapping, gt_str_get(seqid_str),
seqid_str);
}
gt_assert(seqid_str);
/* construct the new feature */
gn = gt_feature_node_new(seqid_str, type, range.start, range.end,
gt_strand_value);
gt_genome_node_set_origin(gn, filenamestr, line_number);
/* set source */
source_str = gt_hashmap_get(parser->source_to_str_mapping, source);
if (!source_str) {
source_str = gt_str_new_cstr(source);
gt_hashmap_add(parser->source_to_str_mapping, gt_str_get(source_str),
source_str);
}
gt_assert(source_str);
gt_feature_node_set_source((GtFeatureNode*) gn, source_str);
if (score_is_defined)
gt_feature_node_set_score((GtFeatureNode*) gn, score_value);
if (phase_value != GT_PHASE_UNDEFINED)
gt_feature_node_set_phase((GtFeatureNode*) gn, phase_value);
gt_array_add(gt_genome_node_array, gn);
}
gt_str_reset(line_buffer);
}
/* process all region nodes */
if (!had_err)
gt_region_node_builder_build(parser->region_node_builder, genome_nodes);
/* process all feature nodes */
cinfo.genome_nodes = genome_nodes;
cinfo.gene_id_to_name_mapping = parser->gene_id_to_name_mapping;
cinfo.transcript_id_to_name_mapping = parser->transcript_id_to_name_mapping;
if (!had_err) {
had_err = gt_hashmap_foreach(parser->gene_id_hash, construct_genes,
&cinfo, err);
}
/* free */
gt_splitter_delete(splitter);
gt_splitter_delete(attribute_splitter);
gt_str_delete(line_buffer);
return had_err;
}
