char* LytCtgData::readName(char* s, GHash<int>& names) {
char* p=strchrs(s, " \t");
if (p!=NULL) {
char* tmp;
char* tmp2;
GMALLOC(tmp, (p-s+30)*sizeof(char));
strncpy(tmp, s,p-s);
tmp[p-s]='\0';
GMALLOC(tmp2, (p-s+30)*sizeof(char));
strcpy(tmp2, tmp);
//make it unique (by simple versioning)
int v=0;
while (names.hasKey(tmp2)) {
v++;
sprintf(tmp2, "%s.%d", tmp, v);
}
name=Gstrdup(tmp2);
GFREE(tmp);
GFREE(tmp2);
names.shkAdd(name, new int(1));
p++;
}
else {
GMessage("LytCtgData::readName: Cannot find the token delimiter in:\n%s\n", s);
}
return p;
}
static void
gst_motioncells_update_motion_cells (GstMotioncells * filter)
{
int i = 0;
int cellscnt = 0;
int j = 0;
int newcellscnt;
motioncellidx *motioncellsidx;
for (i = 0; i < filter->motioncells_count; i++) {
if ((filter->gridx <= filter->motioncellsidx[i].columnidx) ||
(filter->gridy <= filter->motioncellsidx[i].lineidx)) {
cellscnt++;
}
}
newcellscnt = filter->motioncells_count - cellscnt;
motioncellsidx = g_new0 (motioncellidx, newcellscnt);
for (i = 0; i < filter->motioncells_count; i++) {
if ((filter->motioncellsidx[i].lineidx < filter->gridy) &&
(filter->motioncellsidx[i].columnidx < filter->gridx)) {
motioncellsidx[j].lineidx = filter->motioncellsidx[i].lineidx;
motioncellsidx[j].columnidx = filter->motioncellsidx[i].columnidx;
j++;
}
}
GFREE (filter->motioncellsidx);
filter->motioncells_count = newcellscnt;
filter->motioncellsidx = g_new0 (motioncellidx, filter->motioncells_count);
j = 0;
for (i = 0; i < filter->motioncells_count; i++) {
filter->motioncellsidx[i].lineidx = motioncellsidx[j].lineidx;
filter->motioncellsidx[i].columnidx = motioncellsidx[j].columnidx;
j++;
}
GFREE (motioncellsidx);
}
GArgs::~GArgs() {
int i;
for (i=0; i<fmtcount; i++)
GFREE(fmt[i].opt);
GFREE(fmt);
for (i=0; i<count; i++) {
GFREE(args[i].opt);
GFREE(args[i].value);
}
GFREE(args);
}
int GArgs::validLongOpt(char* o, char* to) {
char* pstr=Gstrdup(o,to);
for (int i=0; i<fmtcount; i++) {
if (fmt[i].longopt && strcmp(fmt[i].longopt, pstr)==0) {
GFREE(pstr);
return i;
}
}
GFREE(pstr);
return -1;
}
int Gmkdir(const char *path, bool recursive, int perms) {
if (path==NULL || path[0]==0) return -1;
mode_t process_mask = umask(0); //is this really needed?
if (!recursive) {
int r=G_mkdir(path, perms);
if (r!=0)
GMessage("Warning: G_mkdir(%s) failed: %s\n", path, strerror(errno));
umask(process_mask);
return r;
}
int plen=strlen(path);
char* gpath=NULL;
//make sure gpath ends with /
if (path[plen-1]=='/') {
gpath=Gstrdup(path);
}
else {
GMALLOC(gpath, plen+2);
strcpy(gpath,path);
strcat(gpath, "/");
++plen;
}
//char* ss=gpath+plen-1;
char* psep = gpath+plen-1; //start at the last /
GDynArray<char*> dirstack(4); // stack of directories that should be created
while (psep>gpath && *(psep-1)=='/') --psep; //skip double slashes
*psep='\0';
int fexists=0;
while ((fexists=fileExists(gpath))==0) {
dirstack.Push(psep);
do { --psep; } while (psep>gpath && *psep!='/');
if (psep<=gpath) { psep=NULL; break; }
while (psep>gpath && *(psep-1)=='/') --psep;
*psep='\0';
}
if (psep) *psep='/';
while (dirstack.Count()>0) {
psep=dirstack.Pop();
int mkdir_err=0;
if ((mkdir_err=G_mkdir(gpath, perms))!=0) {
GMessage("Warning: mkdir(%s) failed: %s\n", gpath, strerror(errno));
GFREE(gpath);
umask(process_mask);
return -1;
}
*psep='/';
}
GFREE(gpath);
umask(process_mask);
return 0;
}
/* Clean up */
static void
gst_motion_cells_finalize (GObject * obj)
{
GstMotioncells *filter = gst_motion_cells (obj);
motion_cells_free (filter->id);
//freeing previously allocated dynamic array
if (filter->motionmaskcoord_count > 0) {
GFREE (filter->motionmaskcoords);
}
if (filter->motionmaskcells_count > 0) {
GFREE (filter->motionmaskcellsidx);
}
if (filter->motioncells_count > 0) {
GFREE (filter->motioncellsidx);
}
if (filter->cvImage) {
cvReleaseImage (&filter->cvImage);
}
GFREE (filter->motioncellscolor);
GFREE (filter->prev_datafile);
GFREE (filter->cur_datafile);
GFREE (filter->basename_datafile);
GFREE (filter->datafile_extension);
g_mutex_free (filter->propset_mutex);
G_OBJECT_CLASS (parent_class)->finalize (obj);
}
char* replaceStr(char* &str, char* newvalue) {
if (str!=NULL) GFREE(str);
if (newvalue==NULL) { return NULL; }
GMALLOC(str, strlen(newvalue)+1);
strcpy(str,newvalue);
return str;
}
int GFastaIndex::storeIndex(const char* finame) { //write the hash to a file
if (records.Count()==0)
GError("Error at GFastaIndex:storeIndex(): no records found!\n");
FILE* fai=fopen(finame, "w");
if (fai==NULL) GError("Error creating fasta index file: %s\n",finame);
int rcount=storeIndex(fai);
GFREE(fai_name);
fai_name=Gstrdup(finame);
return rcount;
}
bool AceParser::loadContig(int ctgidx, fnLytSeq* seqfn, bool re_pos) {
bool forgetCtg = false;
if (ctgidx>=contigs.Count())
GError("LayoutParser: invalid contig index '%d'\n", ctgidx);
LytCtgData* ctgdata=contigs[ctgidx];
if (re_pos && currentContig!=NULL) { //free previously loaded contig data
currentContig->seqs.Clear(); // unless it was a parse() call
seqinfo.Clear();
}
currentContig=ctgdata;
int ctg_numSeqs=ctgdata->numseqs;
if (re_pos) {
seek(ctgdata->fpos); //position right where the contig definition starts
char *r = linebuf->getLine(f,f_pos);
if (r==NULL) return false;
}
if (seqfn!=NULL) { //process the contig sequence!
char* ctgseq=readSeq();
forgetCtg=(*seqfn)(numContigs, ctgdata, NULL, ctgseq);
GFREE(ctgseq); //obviously the caller should have made a copy
}
//now look for all the component sequences
if (fskipTo("AF ")<0) {
GMessage("AceParser: error finding sequence offsets (AF)"
" for contig '%s' (%d)\n", ctgdata->name, ctgdata->len);
return false;
}
int numseqs=0;
while (startsWith(linebuf->chars(), "AF ",3)) {
if (addSeq(linebuf->chars(), ctgdata)==NULL) {
GMessage("AceParser: error parsing AF entry:\n%s\n",linebuf->chars());
return false;
}
numseqs++;
//read next line:
linebuf->getLine(f,f_pos);
}
if (numseqs!=ctg_numSeqs) {
GMessage("Invalid number of AF entries found (%d) for contig '%s' "
"(length %d, numseqs %d)\n", numseqs,
ctgdata->name, ctgdata->len, ctg_numSeqs);
return false;
}
//now read each sequence entry
off_t seqpos=fskipTo("RD ");
numseqs=0; //count again, now the RD entries
if (seqpos<0) {
GMessage("AceParser: error locating first RD entry for contig '%s'\n",
ctgdata->name);
return false;
}
//int numseqs=0;
//reading the actual component sequence details
while (startsWith(linebuf->chars(), "RD ",3)) {
char* s=linebuf->chars()+3;
char* p=strchrs(s, " \t");
LytSeqInfo* seq;
if (p==NULL) {
GMessage("AceParser: Error parsing RD header line:\n%s\n", linebuf->chars());
return false;
}
*p='\0';
if ((seq=seqinfo.Find(s))==NULL) {
GMessage("AceParser: unknown RD encountered: '%s'\n", s);
return false;
}
p++; //now p is in linebuf after the RD name
seq->fpos=seqpos;
int len;
if (sscanf(p, "%d", &len)!=1) {
GMessage("AceParser: cannot parse RD length for '%s'\n", s);
return false;
}
seq->setLength(len);
//read the sequence data here if a callback fn was given:
char* sseq=NULL;
if (seqfn!=NULL)
sseq=readSeq(seq); //read full sequence here
if (fskipTo("QA ")<0) {
GMessage("AceParser: Error finding QA entry for read %s! (fpos=%llu)\n", seq->name, (unsigned long long)f_pos);
return false;
}
//parse QA entry:
int tmpa, tmpb;
if (sscanf(linebuf->chars()+3, "%d %d %d %d", &tmpa, &tmpb, &seq->left,&seq->right)!=4 ||
seq->left<=0 || seq->right<=0) {
GMessage("AceParser: Error parsing QA entry.\n");
return false;
}
/*
if (fskipTo("DS")<0) {
GMessage("AceParser: Error closing RD entry ('DS' not found).\n");
return false;
}
*/
seqpos=getFilePos()+1;
bool forgetSeq=false;
if (seqfn!=NULL) {
forgetSeq=(*seqfn)(numContigs, ctgdata, seq, sseq);
GFREE(sseq);
}
if (forgetSeq) { //parsing the whole stream -- aceconv)
ctg_numSeqs--;
seqinfo.Remove(seq->name);
ctgdata->seqs.RemovePtr(seq);
}
numseqs++;
if (numseqs<ctgdata->numseqs)
seqpos=fskipTo("RD ", "CO "); //more sequences left to read
}
if (numseqs!=ctgdata->numseqs) {
GMessage("Error: Invalid number of RD entries found (%d) for contig '%s' "
"(length %d, numseqs %d)\n", numseqs,
ctgdata->name, ctgdata->len, ctg_numSeqs);
return false;
}
if (forgetCtg) {
ctgIDs.Remove(ctgdata->name);
ctgdata->seqs.Clear();
seqinfo.Clear();
contigs.RemovePtr(ctgdata);
}
return true;
}
GPrivate void urbd_free(RBDataP_t pData)
{
GFREE(pData);
}
~SeqInfo() {
GFREE(descr);
}
~RefTran() {
GFREE(new_name);
}
int main(int argc, char * const argv[]) {
GArgs args(argc, argv, "hFCq:r:o:");
int e;
if ((e=args.isError())>0)
GError("%s\nInvalid argument: %s\n", USAGE, argv[e]);
if (args.getOpt('h')!=NULL){
GMessage("%s\n", USAGE);
exit(1);
}
args.startNonOpt();
GStr fadb(args.nextNonOpt());
if (fadb.is_empty()) GError("%s Error: multi-fasta file expected!\n",USAGE);
GStr fname(fadb);
fname.append(".fai");
bool createLocal=(args.getOpt('F')!=NULL);
const char* idxname=(createLocal)? NULL : fname.chars();
GFastaIndex faidx(fadb.chars(), idxname);
//also tried to load the index if exists in the current directory
GStr fnamecwd(fname); //name in current directory (without path)
int ip=-1;
if ((ip=fnamecwd.rindex(CHPATHSEP))>=0) {
fnamecwd.cut(0,ip+1);
}
if (!createLocal) { //look for existing indexes to load
//try the same directory as the fasta file first
if (!faidx.hasIndex() and fileExists(fnamecwd.chars())>1) { //try current working directory next
faidx.loadIndex(fnamecwd.chars());
}
if (!faidx.hasIndex()) {//could not load any index data
//try to create it in the same directory as the fasta file
GMessage("No fasta index found. Rebuilding..\n");
faidx.buildIndex();
if (faidx.getCount()==0) GError("Error: no fasta records to be indexed!\n");
GMessage("Fasta index rebuilt.\n");
//check if we can create a file there
FILE* fcreate=fopen(fname.chars(), "w");
if (fcreate==NULL)
GMessage("Warning: cannot create fasta index %s! (permissions?)\n", fname.chars());
else {
fclose(fcreate);
if (faidx.storeIndex(fname.chars())<faidx.getCount())
GMessage("Warning: error writing the index file %s!\n",fname.chars());
} //creating index file in the same directory as fasta file
}//trying to create the index file
}
if (createLocal || !faidx.hasIndex()) {
//simply rebuild the index in the current directory and use it:
//remove directories in path, if any
if (faidx.getCount()==0) {
faidx.buildIndex();
if (faidx.getCount()==0) GError("Error: no fasta records to be indexed!\n");
}
if (faidx.storeIndex(fnamecwd.chars())<faidx.getCount())
GMessage("Warning: error writing the index file %s!\n",fnamecwd.chars());
}
GStr qry(args.getOpt('q'));
if (qry.is_empty()) exit(0);
GFastaRec* farec=faidx.getRecord(qry.chars());
if (farec==NULL) {
GMessage("Error: couldn't find fasta record for '%s'!\n",qry.chars());
exit(1);
}
GFaSeqGet faseq(fadb.chars(),farec->seqlen, farec->fpos, farec->line_len, farec->line_blen);
//TODO: read these from -r option
uint qstart=0;
uint qend=0; //farec->seqlen
bool revCompl=(args.getOpt('C')!=NULL);
char* s=args.getOpt('r');
if (s!=NULL) {
char *p=s;
while (isdigit(*p)) p++;
if (*p=='-') {
sscanf(s,"%u-%u",&qstart, &qend);
if (qstart==0 || qend==0)
GError("Error parsing sequence range: %s\n",s);
}
else if (*p==':') {
int qlen=0;
sscanf(s,"%u:%d", &qstart, &qlen);
if (qstart==0 || qlen==0)
GError("Error parsing sequence range: %s\n",s);
qend=qstart+qlen-1;
}
else if (*p=='.') {
sscanf(s,"%u..%u",&qstart, &qend);
if (qstart==0 || qend==0)
GError("Error parsing sequence range: %s\n",s);
}
}
if (qstart==0) qstart=1;
if (qend==0) qend=farec->seqlen;
// call faseq.loadall() here if multiple ranges are to be extracted all
// over this genomic sequence
char* subseq=faseq.copyRange(qstart, qend, revCompl, true);
FILE* f_out=NULL;
openfwrite(f_out, args, 'o');
if (f_out==NULL) f_out=stdout;
writeFasta(f_out, qry.chars(), NULL, subseq, 70, qend-qstart+1);
GFREE(subseq);
}
int GFastaIndex::buildIndex() {
//this parses the whole fasta file, so it could be slow
if (fa_name==NULL)
GError("Error: GFastaIndex::buildIndex() called with no fasta file!\n");
FILE* fa=fopen(fa_name,"rb");
if (fa==NULL) {
GMessage("Warning: cannot open fasta index file: %s!\n",fa_name);
return 0;
}
records.Clear();
GLineReader fl(fa);
char* s=NULL;
uint seqlen=0;
int line_len=0,line_blen=0;
bool newSeq=false; //set to true after defline
off_t newSeqOffset=0;
int prevOffset=0;
char* seqname=NULL;
int last_len=0;
bool mustbeLastLine=false; //true if the line length decreases
while ((s=fl.nextLine())!=NULL) {
if (s[0]=='>') {
if (seqname!=NULL) {
if (seqlen==0)
GError("Warning: empty FASTA record skipped (%s)!\n",seqname);
else { //seqlen!=0
addRecord(seqname, seqlen,newSeqOffset, line_len, line_blen);
}
}
char *p=s;
while (*p > 32) p++;
*p=0;
GFREE(seqname);
seqname=Gstrdup(&s[1]);
newSeq=true;
newSeqOffset=fl.getfpos();
last_len=0;
line_len=0;
line_blen=0;
seqlen=0;
mustbeLastLine=false;
} //defline parsing
else { //sequence line
int llen=fl.length();
int lblen=fl.getFpos()-prevOffset;
if (newSeq) { //first sequence line after defline
line_len=llen;
line_blen=lblen;
}
else {//next seq lines after first
if (mustbeLastLine || llen>last_len)
GError(ERR_FALINELEN);
if (llen<last_len) mustbeLastLine=true;
}
seqlen+=llen;
last_len=llen;
newSeq=false;
} //sequence line
prevOffset=fl.getfpos();
}//for each line of the fasta file
if (seqlen>0)
addRecord(seqname, seqlen, newSeqOffset, line_len, line_blen);
GFREE(seqname);
fclose(fa);
return records.Count();
}
bool process_transcript(GFastaDb& gfasta, GffObj& gffrec) {
//returns true if the transcript passed the filter
char* gname=gffrec.getGeneName();
if (gname==NULL) gname=gffrec.getGeneID();
GStr defline(gffrec.getID());
if (f_out && !fmtGTF) {
const char* tname=NULL;
if ((tname=gffrec.getAttr("transcript_name"))!=NULL) {
gffrec.addAttr("Name", tname);
gffrec.removeAttr("transcript_name");
}
}
if (ensembl_convert && startsWith(gffrec.getID(), "ENS")) {
const char* biotype=gffrec.getAttr("gene_biotype");
if (biotype) {
gffrec.addAttr("type", biotype);
gffrec.removeAttr("gene_biotype");
}
else { //old Ensembl files lacking gene_biotype
gffrec.addAttr("type", gffrec.getTrackName());
}
//bool is_gene=false;
bool is_pseudo=false;
if (strcmp(biotype, "protein_coding")==0 || gffrec.hasCDS())
gffrec.setFeatureName("mRNA");
else {
if (strcmp(biotype, "processed_transcript")==0)
gffrec.setFeatureName("proc_RNA");
else {
//is_gene=endsWith(biotype, "gene");
is_pseudo=strifind(biotype, "pseudo");
if (is_pseudo) {
gffrec.setFeatureName("pseudo_RNA");
}
else if (endsWith(biotype, "RNA")) {
gffrec.setFeatureName(biotype);
} else gffrec.setFeatureName("misc_RNA");
}
}
}
if (gname && strcmp(gname, gffrec.getID())!=0) {
int* isonum=isoCounter.Find(gname);
if (isonum==NULL) {
isonum=new int(1);
isoCounter.Add(gname,isonum);
}
else (*isonum)++;
defline.appendfmt(" gene=%s", gname);
}
int seqlen=0;
const char* tlabel=tracklabel;
if (tlabel==NULL) tlabel=gffrec.getTrackName();
//defline.appendfmt(" track:%s",tlabel);
char* cdsnt = NULL;
char* cdsaa = NULL;
int aalen=0;
for (int i=1;i<gffrec.exons.Count();i++) {
int ilen=gffrec.exons[i]->start-gffrec.exons[i-1]->end-1;
if (ilen>4000000)
GMessage("Warning: very large intron (%d) for transcript %s\n",
ilen, gffrec.getID());
if (ilen>maxintron) {
return false;
}
}
GList<GSeg> seglst(false,true);
GFaSeqGet* faseq=fastaSeqGet(gfasta, gffrec);
if (spliceCheck && gffrec.exons.Count()>1) {
//check introns for splice site consensi ( GT-AG, GC-AG or AT-AC )
if (faseq==NULL) GError("Error: no genomic sequence available!\n");
int glen=gffrec.end-gffrec.start+1;
const char* gseq=faseq->subseq(gffrec.start, glen);
bool revcompl=(gffrec.strand=='-');
bool ssValid=true;
for (int e=1;e<gffrec.exons.Count();e++) {
const char* intron=gseq+gffrec.exons[e-1]->end+1-gffrec.start;
int intronlen=gffrec.exons[e]->start-gffrec.exons[e-1]->end-1;
GSpliceSite acceptorSite(intron,intronlen,true, revcompl);
GSpliceSite donorSite(intron,intronlen, false, revcompl);
//GMessage("%c intron %d-%d : %s .. %s\n",
// gffrec.strand, istart, iend, donorSite.nt, acceptorSite.nt);
if (acceptorSite=="AG") { // GT-AG or GC-AG
if (!donorSite.canonicalDonor()) {
ssValid=false;break;
}
}
else if (acceptorSite=="AC") { //
if (donorSite!="AT") { ssValid=false; break; }
}
else { ssValid=false; break; }
}
//GFREE(gseq);
if (!ssValid) {
if (verbose)
GMessage("Invalid splice sites found for '%s'\n",gffrec.getID());
return false; //don't print this one!
}
}
bool trprint=true;
int stopCodonAdjust=0;
int mCDphase=0;
bool hasStop=false;
if (gffrec.CDphase=='1' || gffrec.CDphase=='2')
mCDphase = gffrec.CDphase-'0';
if (f_y!=NULL || f_x!=NULL || validCDSonly) {
if (faseq==NULL) GError("Error: no genomic sequence provided!\n");
//if (protmap && fullCDSonly) {
//if (protmap && (fullCDSonly || (gffrec.qlen>0 && gffrec.qend==gffrec.qlen))) {
if (validCDSonly) { //make sure the stop codon is always included
//adjust_stopcodon(gffrec,3);
stopCodonAdjust=adjust_stopcodon(gffrec,3);
}
int strandNum=0;
int phaseNum=0;
CDS_CHECK:
cdsnt=gffrec.getSpliced(faseq, true, &seqlen, NULL, NULL, &seglst);
if (cdsnt==NULL) trprint=false;
else { //has CDS
if (validCDSonly) {
cdsaa=translateDNA(cdsnt, aalen, seqlen);
char* p=strchr(cdsaa,'.');
hasStop=false;
if (p!=NULL) {
if (p-cdsaa>=aalen-2) { //stop found as the last codon
*p='0';//remove it
hasStop=true;
if (aalen-2==p-cdsaa) {
//previous to last codon is the stop codon
//so correct the CDS stop accordingly
adjust_stopcodon(gffrec,-3, &seglst);
stopCodonAdjust=0; //clear artificial stop adjustment
seqlen-=3;
cdsnt[seqlen]=0;
}
aalen=p-cdsaa;
}
else {//stop found before the last codon
trprint=false;
}
}//stop codon found
if (trprint==false) { //failed CDS validity check
//in-frame stop codon found
if (altPhases && phaseNum<3) {
phaseNum++;
gffrec.CDphase = '0'+((mCDphase+phaseNum)%3);
GFREE(cdsaa);
goto CDS_CHECK;
}
if (gffrec.exons.Count()==1 && bothStrands) {
strandNum++;
phaseNum=0;
if (strandNum<2) {
GFREE(cdsaa);
gffrec.strand = (gffrec.strand=='-') ? '+':'-';
goto CDS_CHECK; //repeat the CDS check for a different frame
}
}
if (verbose) GMessage("In-frame STOP found for '%s'\n",gffrec.getID());
} //has in-frame STOP
if (fullCDSonly) {
if (!hasStop || cdsaa[0]!='M') trprint=false;
}
} // CDS check requested
} //has CDS
} //translation or codon check/output was requested
if (!trprint) {
GFREE(cdsnt);
GFREE(cdsaa);
return false;
}
if (stopCodonAdjust>0 && !hasStop) {
//restore stop codon location
adjust_stopcodon(gffrec, -stopCodonAdjust, &seglst);
if (cdsnt!=NULL && seqlen>0) {
seqlen-=stopCodonAdjust;
cdsnt[seqlen]=0;
}
if (cdsaa!=NULL) aalen--;
}
if (f_y!=NULL) { //CDS translation fasta output requested
//char*
if (cdsaa==NULL) { //translate now if not done before
cdsaa=translateDNA(cdsnt, aalen, seqlen);
}
if (fullattr && gffrec.attrs!=NULL) {
//append all attributes found for each transcripts
for (int i=0;i<gffrec.attrs->Count();i++) {
defline.append(" ");
defline.append(gffrec.getAttrName(i));
defline.append("=");
defline.append(gffrec.getAttrValue(i));
}
}
printFasta(f_y, defline, cdsaa, aalen);
}
if (f_x!=NULL) { //CDS only
if (writeExonSegs) {
defline.append(" loc:");
defline.append(gffrec.getGSeqName());
defline.appendfmt("(%c)",gffrec.strand);
//warning: not CDS coordinates are written here, but the exon ones
defline+=(int)gffrec.start;
defline+=(char)'-';
defline+=(int)gffrec.end;
// -- here these are CDS substring coordinates on the spliced sequence:
defline.append(" segs:");
for (int i=0;i<seglst.Count();i++) {
if (i>0) defline.append(",");
defline+=(int)seglst[i]->start;
defline.append("-");
defline+=(int)seglst[i]->end;
}
}
if (fullattr && gffrec.attrs!=NULL) {
//append all attributes found for each transcript
for (int i=0;i<gffrec.attrs->Count();i++) {
defline.append(" ");
defline.append(gffrec.getAttrName(i));
defline.append("=");
defline.append(gffrec.getAttrValue(i));
}
}
printFasta(f_x, defline, cdsnt, seqlen);
}
GFREE(cdsnt);
GFREE(cdsaa);
if (f_w!=NULL) { //write spliced exons
uint cds_start=0;
uint cds_end=0;
seglst.Clear();
char* exont=gffrec.getSpliced(faseq, false, &seqlen, &cds_start, &cds_end, &seglst);
if (exont!=NULL) {
if (gffrec.CDstart>0) {
defline.appendfmt(" CDS=%d-%d", cds_start, cds_end);
}
if (writeExonSegs) {
defline.append(" loc:");
defline.append(gffrec.getGSeqName());
defline+=(char)'|';
defline+=(int)gffrec.start;
defline+=(char)'-';
defline+=(int)gffrec.end;
defline+=(char)'|';
defline+=(char)gffrec.strand;
defline.append(" exons:");
for (int i=0;i<gffrec.exons.Count();i++) {
if (i>0) defline.append(",");
defline+=(int)gffrec.exons[i]->start;
defline.append("-");
defline+=(int)gffrec.exons[i]->end;
}
defline.append(" segs:");
for (int i=0;i<seglst.Count();i++) {
if (i>0) defline.append(",");
defline+=(int)seglst[i]->start;
defline.append("-");
defline+=(int)seglst[i]->end;
}
}
if (fullattr && gffrec.attrs!=NULL) {
//append all attributes found for each transcripts
for (int i=0;i<gffrec.attrs->Count();i++) {
defline.append(" ");
defline.append(gffrec.getAttrName(i));
defline.append("=");
defline.append(gffrec.getAttrValue(i));
}
}
printFasta(f_w, defline, exont, seqlen);
GFREE(exont);
}
} //writing f_w (spliced exons)
return true;
}
static void
gst_motion_cells_set_property (GObject * object, guint prop_id,
const GValue * value, GParamSpec * pspec)
{
GstMotioncells *filter = gst_motion_cells (object);
//variables for overlay regions setup
gchar **strs, **colorstr, **motioncellsstr, **motionmaskcellsstr;
int i, ux, uy, lx, ly;
int r, g, b;
int cellscolorscnt = 0;
int linidx, colidx, masklinidx, maskcolidx;
int tmpux = -1;
int tmpuy = -1;
int tmplx = -1;
int tmply = -1;
GstStateChangeReturn ret;
g_mutex_lock (filter->propset_mutex);
switch (prop_id) {
case PROP_GRID_X:
ret = gst_element_get_state (GST_ELEMENT (filter),
&filter->state, NULL, 250 * GST_NSECOND);
filter->gridx = g_value_get_int (value);
if (filter->prevgridx != filter->gridx
&& ret == GST_STATE_CHANGE_SUCCESS
&& filter->state == GST_STATE_PLAYING) {
filter->changed_gridx = true;
}
filter->prevgridx = filter->gridx;
break;
case PROP_GRID_Y:
ret = gst_element_get_state (GST_ELEMENT (filter),
&filter->state, NULL, 250 * GST_NSECOND);
filter->gridy = g_value_get_int (value);
if (filter->prevgridy != filter->gridy
&& ret == GST_STATE_CHANGE_SUCCESS
&& filter->state == GST_STATE_PLAYING) {
filter->changed_gridy = true;
}
filter->prevgridy = filter->gridy;
break;
case PROP_GAP:
filter->gap = g_value_get_int (value);
break;
case PROP_POSTNOMOTION:
filter->postnomotion = g_value_get_int (value);
break;
case PROP_MINIMUNMOTIONFRAMES:
filter->minimum_motion_frames = g_value_get_int (value);
break;
case PROP_SENSITIVITY:
filter->sensitivity = g_value_get_double (value);
break;
case PROP_THRESHOLD:
filter->threshold = g_value_get_double (value);
break;
case PROP_DISPLAY:
filter->display = g_value_get_boolean (value);
break;
case PROP_POSTALLMOTION:
filter->postallmotion = g_value_get_boolean (value);
break;
case PROP_USEALPHA:
filter->usealpha = g_value_get_boolean (value);
break;
case PROP_CALCULATEMOTION:
filter->calculate_motion = g_value_get_boolean (value);
break;
case PROP_DATE:
ret = gst_element_get_state (GST_ELEMENT (filter),
&filter->state, NULL, 250 * GST_NSECOND);
if (ret == GST_STATE_CHANGE_SUCCESS && filter->state == GST_STATE_PLAYING) {
filter->changed_startime = true;
}
filter->starttime = g_value_get_long (value);
break;
case PROP_DATAFILE:
GFREE (filter->cur_datafile);
GFREE (filter->basename_datafile);
filter->basename_datafile = g_value_dup_string (value);
if (strlen (filter->basename_datafile) == 0) {
filter->cur_datafile = g_strdup (NULL);
break;
}
filter->cur_datafile =
g_strdup_printf ("%s-0.%s", filter->basename_datafile,
filter->datafile_extension);
if (g_strcmp0 (filter->prev_datafile, filter->basename_datafile) != 0) {
filter->changed_datafile = TRUE;
filter->sent_init_error_msg = FALSE;
filter->sent_save_error_msg = FALSE;
filter->datafileidx = 0;
motion_cells_free_resources (filter->id);
} else {
filter->changed_datafile = FALSE;
}
GFREE (filter->prev_datafile);
filter->prev_datafile = g_strdup (filter->basename_datafile);
break;
case PROP_DATAFILE_EXT:
GFREE (filter->datafile_extension);
filter->datafile_extension = g_value_dup_string (value);
break;
case PROP_MOTIONMASKCOORD:
strs = g_strsplit (g_value_get_string (value), ",", 255);
GFREE (filter->motionmaskcoords);
//setting number of regions
for (filter->motionmaskcoord_count = 0;
strs[filter->motionmaskcoord_count] != NULL;
++filter->motionmaskcoord_count);
if (filter->motionmaskcoord_count > 0) {
sscanf (strs[0], "%d:%d:%d:%d", &tmpux, &tmpuy, &tmplx, &tmply);
if (tmpux > -1 && tmpuy > -1 && tmplx > -1 && tmply > -1) {
filter->motionmaskcoords =
g_new0 (motionmaskcoordrect, filter->motionmaskcoord_count);
for (i = 0; i < filter->motionmaskcoord_count; ++i) {
sscanf (strs[i], "%d:%d:%d:%d", &ux, &uy, &lx, &ly);
ux = CLAMP (ux, 0, filter->width - 1);
uy = CLAMP (uy, 0, filter->height - 1);
lx = CLAMP (lx, 0, filter->width - 1);
ly = CLAMP (ly, 0, filter->height - 1);
filter->motionmaskcoords[i].upper_left_x = ux;
filter->motionmaskcoords[i].upper_left_y = uy;
filter->motionmaskcoords[i].lower_right_x = lx;
filter->motionmaskcoords[i].lower_right_y = ly;
}
} else {
filter->motionmaskcoord_count = 0;
}
}
if (strs)
g_strfreev (strs);
tmpux = -1;
tmpuy = -1;
tmplx = -1;
tmply = -1;
break;
case PROP_MOTIONMASKCELLSPOS:
motionmaskcellsstr = g_strsplit (g_value_get_string (value), ",", 255);
GFREE (filter->motionmaskcellsidx);
//setting number of regions
for (filter->motionmaskcells_count = 0;
motionmaskcellsstr[filter->motionmaskcells_count] != NULL;
++filter->motionmaskcells_count);
if (filter->motionmaskcells_count > 0) {
sscanf (motionmaskcellsstr[0], "%d:%d", &tmpux, &tmpuy);
if (tmpux > -1 && tmpuy > -1) {
filter->motionmaskcellsidx =
g_new0 (motioncellidx, filter->motionmaskcells_count);
for (i = 0; i < filter->motionmaskcells_count; ++i) {
sscanf (motionmaskcellsstr[i], "%d:%d", &masklinidx, &maskcolidx);
filter->motionmaskcellsidx[i].lineidx = masklinidx;
filter->motionmaskcellsidx[i].columnidx = maskcolidx;
}
} else {
filter->motionmaskcells_count = 0;
}
}
if (motionmaskcellsstr)
g_strfreev (motionmaskcellsstr);
tmpux = -1;
tmpuy = -1;
tmplx = -1;
tmply = -1;
break;
case PROP_CELLSCOLOR:
colorstr = g_strsplit (g_value_get_string (value), ",", 255);
for (cellscolorscnt = 0; colorstr[cellscolorscnt] != NULL;
++cellscolorscnt);
if (cellscolorscnt == 3) {
sscanf (colorstr[0], "%d", &r);
sscanf (colorstr[1], "%d", &g);
sscanf (colorstr[2], "%d", &b);
//check right RGB color format
r = CLAMP (r, 1, 255);
g = CLAMP (g, 1, 255);
b = CLAMP (b, 1, 255);
filter->motioncellscolor->R_channel_value = r;
filter->motioncellscolor->G_channel_value = g;
filter->motioncellscolor->B_channel_value = b;
}
if (colorstr)
g_strfreev (colorstr);
break;
case PROP_MOTIONCELLSIDX:
motioncellsstr = g_strsplit (g_value_get_string (value), ",", 255);
//setting number of regions
for (filter->motioncells_count = 0;
motioncellsstr[filter->motioncells_count] != NULL;
++filter->motioncells_count);
if (filter->motioncells_count > 0) {
sscanf (motioncellsstr[0], "%d:%d", &tmpux, &tmpuy);
if (tmpux > -1 && tmpuy > -1) {
GFREE (filter->motioncellsidx);
filter->motioncellsidx =
g_new0 (motioncellidx, filter->motioncells_count);
for (i = 0; i < filter->motioncells_count; ++i) {
sscanf (motioncellsstr[i], "%d:%d", &linidx, &colidx);
filter->motioncellsidx[i].lineidx = linidx;
filter->motioncellsidx[i].columnidx = colidx;
}
} else {
filter->motioncells_count = 0;
}
}
if (motioncellsstr)
g_strfreev (motioncellsstr);
tmpux = -1;
tmpuy = -1;
tmplx = -1;
tmply = -1;
break;
case PROP_MOTIONCELLTHICKNESS:
filter->thickness = g_value_get_int (value);
break;
default:
G_OBJECT_WARN_INVALID_PROPERTY_ID (object, prop_id, pspec);
break;
}
g_mutex_unlock (filter->propset_mutex);
}
int main(int argc, char * const argv[]) {
GArgs args(argc, argv,
"debug;merge;cluster-only;help;force-exons;no-pseudo;MINCOV=MINPID=hvOUNHWCVJMKQNSXTDAPRZFGLEm:g:i:r:s:t:a:b:o:w:x:y:d:");
args.printError(USAGE, true);
if (args.getOpt('h') || args.getOpt("help")) {
GMessage("%s",USAGE);
exit(1);
}
debugMode=(args.getOpt("debug")!=NULL);
decodeChars=(args.getOpt('D')!=NULL);
forceExons=(args.getOpt("force-exons")!=NULL);
NoPseudo=(args.getOpt("no-pseudo")!=NULL);
mRNAOnly=(args.getOpt('O')==NULL);
//sortByLoc=(args.getOpt('S')!=NULL);
addDescr=(args.getOpt('A')!=NULL);
verbose=(args.getOpt('v')!=NULL);
wCDSonly=(args.getOpt('C')!=NULL);
validCDSonly=(args.getOpt('V')!=NULL);
altPhases=(args.getOpt('H')!=NULL);
fmtGTF=(args.getOpt('T')!=NULL); //switch output format to GTF
bothStrands=(args.getOpt('B')!=NULL);
fullCDSonly=(args.getOpt('J')!=NULL);
spliceCheck=(args.getOpt('N')!=NULL);
bool matchAllIntrons=(args.getOpt('K')==NULL);
bool fuzzSpan=(args.getOpt('Q')!=NULL);
if (args.getOpt('M') || args.getOpt("merge")) {
doCluster=true;
doCollapseRedundant=true;
}
else {
if (!matchAllIntrons || fuzzSpan) {
GMessage("%s",USAGE);
GMessage("Error: -K or -Q options require -M/--merge option!\n");
exit(1);
}
}
if (args.getOpt("cluster-only")) {
doCluster=true;
doCollapseRedundant=false;
if (!matchAllIntrons || fuzzSpan) {
GMessage("%s",USAGE);
GMessage("Error: -K or -Q options have no effect with --cluster-only.\n");
exit(1);
}
}
if (fullCDSonly) validCDSonly=true;
if (verbose) {
fprintf(stderr, "Command line was:\n");
args.printCmdLine(stderr);
}
fullattr=(args.getOpt('F')!=NULL);
if (args.getOpt('G')==NULL)
noExonAttr=!fullattr;
else {
noExonAttr=true;
fullattr=true;
}
if (NoPseudo && !fullattr) {
noExonAttr=true;
fullattr=true;
}
ensembl_convert=(args.getOpt('L')!=NULL);
if (ensembl_convert) {
fullattr=true;
noExonAttr=false;
//sortByLoc=true;
}
mergeCloseExons=(args.getOpt('Z')!=NULL);
multiExon=(args.getOpt('U')!=NULL);
writeExonSegs=(args.getOpt('W')!=NULL);
tracklabel=args.getOpt('t');
GFastaDb gfasta(args.getOpt('g'));
//if (gfasta.fastaPath!=NULL)
// sortByLoc=true; //enforce sorting by chromosome/contig
GStr s=args.getOpt('i');
if (!s.is_empty()) maxintron=s.asInt();
FILE* f_repl=NULL;
s=args.getOpt('d');
if (!s.is_empty()) {
if (s=="-") f_repl=stdout;
else {
f_repl=fopen(s.chars(), "w");
if (f_repl==NULL) GError("Error creating file %s\n", s.chars());
}
}
rfltWithin=(args.getOpt('R')!=NULL);
s=args.getOpt('r');
if (!s.is_empty()) {
s.trim();
if (s[0]=='+' || s[0]=='-') {
rfltStrand=s[0];
s.cut(0,1);
}
int isep=s.index(':');
if (isep>0) { //gseq name given
if (rfltStrand==0 && (s[isep-1]=='+' || s[isep-1]=='-')) {
isep--;
rfltStrand=s[isep];
s.cut(isep,1);
}
if (isep>0)
rfltGSeq=Gstrdup((s.substr(0,isep)).chars());
s.cut(0,isep+1);
}
GStr gsend;
char slast=s[s.length()-1];
if (rfltStrand==0 && (slast=='+' || slast=='-')) {
s.chomp(slast);
rfltStrand=slast;
}
if (s.index("..")>=0) gsend=s.split("..");
else gsend=s.split('-');
if (!s.is_empty()) rfltStart=(uint)s.asInt();
if (!gsend.is_empty()) {
rfltEnd=(uint)gsend.asInt();
if (rfltEnd==0) rfltEnd=MAX_UINT;
}
} //gseq/range filtering
else {
if (rfltWithin)
GError("Error: option -R requires -r!\n");
//if (rfltWholeTranscript)
// GError("Error: option -P requires -r!\n");
}
s=args.getOpt('m');
if (!s.is_empty()) {
FILE* ft=fopen(s,"r");
if (ft==NULL) GError("Error opening reference table: %s\n",s.chars());
loadRefTable(ft, reftbl);
fclose(ft);
}
s=args.getOpt('s');
if (!s.is_empty()) {
FILE* fsize=fopen(s,"r");
if (fsize==NULL) GError("Error opening info file: %s\n",s.chars());
loadSeqInfo(fsize, seqinfo);
fclose(fsize);
}
openfw(f_out, args, 'o');
//if (f_out==NULL) f_out=stdout;
if (gfasta.fastaPath==NULL && (validCDSonly || spliceCheck || args.getOpt('w')!=NULL || args.getOpt('x')!=NULL || args.getOpt('y')!=NULL))
GError("Error: -g option is required for options -w, -x, -y, -V, -N, -M !\n");
openfw(f_w, args, 'w');
openfw(f_x, args, 'x');
openfw(f_y, args, 'y');
if (f_y!=NULL || f_x!=NULL) wCDSonly=true;
//useBadCDS=useBadCDS || (fgtfok==NULL && fgtfbad==NULL && f_y==NULL && f_x==NULL);
int numfiles = args.startNonOpt();
//GList<GffObj> gfkept(false,true); //unsorted, free items on delete
int out_counter=0; //number of records printed
while (true) {
GStr infile;
if (numfiles) {
infile=args.nextNonOpt();
if (infile.is_empty()) break;
if (infile=="-") { f_in=stdin; infile="stdin"; }
else
if ((f_in=fopen(infile, "r"))==NULL)
GError("Error: cannot open input file %s!\n",infile.chars());
}
else
infile="-";
GffLoader gffloader(infile.chars());
gffloader.transcriptsOnly=mRNAOnly;
gffloader.fullAttributes=fullattr;
gffloader.noExonAttrs=noExonAttr;
gffloader.mergeCloseExons=mergeCloseExons;
gffloader.showWarnings=(args.getOpt('E')!=NULL);
gffloader.noPseudo=NoPseudo;
gffloader.load(g_data, &validateGffRec, doCluster, doCollapseRedundant,
matchAllIntrons, fuzzSpan, forceExons);
if (doCluster)
collectLocusData(g_data);
if (numfiles==0) break;
}
GStr loctrack("gffcl");
if (tracklabel) loctrack=tracklabel;
g_data.setSorted(&gseqCmpName);
GffPrintMode exonPrinting;
if (fmtGTF) {
exonPrinting = pgtfAny;
} else {
exonPrinting = forceExons ? pgffBoth : pgffAny;
}
bool firstGff3Print=!fmtGTF;
if (doCluster) {
//grouped in loci
for (int g=0;g<g_data.Count();g++) {
GenomicSeqData* gdata=g_data[g];
int gfs_i=0;
for (int l=0;l<gdata->loci.Count();l++) {
GffLocus& loc=*(gdata->loci[l]);
//check all non-replaced transcripts in this locus:
int numvalid=0;
int idxfirstvalid=-1;
for (int i=0;i<loc.rnas.Count();i++) {
GffObj& t=*(loc.rnas[i]);
if (f_out) {
while (gfs_i<gdata->gfs.Count() && gdata->gfs[gfs_i]->start<=t.start) {
GffObj& gfst=*(gdata->gfs[gfs_i]);
if ((gfst.udata&4)==0) { //never printed
gfst.udata|=4;
if (firstGff3Print) { printGff3Header(f_out, args);firstGff3Print=false; }
if (gfst.exons.Count()==0 && gfst.children.Count()==0 && forceExons)
gfst.addExon(gfst.start,gfst.end);
gfst.printGxf(f_out, exonPrinting, tracklabel, NULL, decodeChars);
}
++gfs_i;
}
}
GTData* tdata=(GTData*)(t.uptr);
if (tdata->replaced_by!=NULL) {
if (f_repl && (t.udata & 8)==0) {
//t.udata|=8;
fprintf(f_repl, "%s", t.getID());
GTData* rby=tdata;
while (rby->replaced_by!=NULL) {
fprintf(f_repl," => %s", rby->replaced_by->getID());
rby->rna->udata|=8;
rby=(GTData*)(rby->replaced_by->uptr);
}
fprintf(f_repl, "\n");
}
continue;
}
if (process_transcript(gfasta, t)) {
t.udata|=4; //tag it as valid
numvalid++;
if (idxfirstvalid<0) idxfirstvalid=i;
}
}
if (f_out && numvalid>0) {
GStr locname("RLOC_");
locname.appendfmt("%08d",loc.locus_num);
if (!fmtGTF) {
if (firstGff3Print) { printGff3Header(f_out, args);firstGff3Print=false; }
fprintf(f_out,"%s\t%s\tlocus\t%d\t%d\t.\t%c\t.\tID=%s;locus=%s",
loc.rnas[0]->getGSeqName(), loctrack.chars(), loc.start, loc.end, loc.strand,
locname.chars(), locname.chars());
//const char* loc_gname=loc.getGeneName();
if (loc.gene_names.Count()>0) { //print all gene names associated to this locus
fprintf(f_out, ";genes=%s",loc.gene_names.First()->name.chars());
for (int i=1;i<loc.gene_names.Count();i++) {
fprintf(f_out, ",%s",loc.gene_names[i]->name.chars());
}
}
if (loc.gene_ids.Count()>0) { //print all GeneIDs names associated to this locus
fprintf(f_out, ";geneIDs=%s",loc.gene_ids.First()->name.chars());
for (int i=1;i<loc.gene_ids.Count();i++) {
fprintf(f_out, ",%s",loc.gene_ids[i]->name.chars());
}
}
fprintf(f_out, ";transcripts=%s",loc.rnas[idxfirstvalid]->getID());
for (int i=idxfirstvalid+1;i<loc.rnas.Count();i++) {
fprintf(f_out, ",%s",loc.rnas[i]->getID());
}
fprintf(f_out, "\n");
}
//now print all valid, non-replaced transcripts in this locus:
for (int i=0;i<loc.rnas.Count();i++) {
GffObj& t=*(loc.rnas[i]);
GTData* tdata=(GTData*)(t.uptr);
if (tdata->replaced_by!=NULL || ((t.udata & 4)==0)) continue;
t.addAttr("locus", locname.chars());
out_counter++;
if (fmtGTF) t.printGxf(f_out, exonPrinting, tracklabel, NULL, decodeChars);
else {
if (firstGff3Print) { printGff3Header(f_out, args);firstGff3Print=false; }
//print the parent first, if any
if (t.parent!=NULL && ((t.parent->udata & 4)==0)) {
GTData* pdata=(GTData*)(t.parent->uptr);
if (pdata && pdata->geneinfo!=NULL)
pdata->geneinfo->finalize();
t.parent->addAttr("locus", locname.chars());
t.parent->printGxf(f_out, exonPrinting, tracklabel, NULL, decodeChars);
t.parent->udata|=4;
}
t.printGxf(f_out, exonPrinting, tracklabel, NULL, decodeChars);
}
}
} //have valid transcripts to print
}//for each locus
//print the rest of the isolated pseudo/gene/region features not printed yet
if (f_out) {
while (gfs_i<gdata->gfs.Count()) {
GffObj& gfst=*(gdata->gfs[gfs_i]);
if ((gfst.udata&4)==0) { //never printed
gfst.udata|=4;
if (firstGff3Print) { printGff3Header(f_out, args);firstGff3Print=false; }
if (gfst.exons.Count()==0 && gfst.children.Count()==0 && forceExons)
gfst.addExon(gfst.start,gfst.end);
gfst.printGxf(f_out, exonPrinting, tracklabel, NULL, decodeChars);
}
++gfs_i;
}
}
} //for each genomic sequence
}
else {
//not grouped into loci, print the rnas with their parents, if any
int numvalid=0;
for (int g=0;g<g_data.Count();g++) {
GenomicSeqData* gdata=g_data[g];
int gfs_i=0;
for (int m=0;m<gdata->rnas.Count();m++) {
GffObj& t=*(gdata->rnas[m]);
if (f_out) {
while (gfs_i<gdata->gfs.Count() && gdata->gfs[gfs_i]->start<=t.start) {
GffObj& gfst=*(gdata->gfs[gfs_i]);
if ((gfst.udata&4)==0) { //never printed
gfst.udata|=4;
if (firstGff3Print) { printGff3Header(f_out, args);firstGff3Print=false; }
if (gfst.exons.Count()==0 && gfst.children.Count()==0 && forceExons)
gfst.addExon(gfst.start,gfst.end);
gfst.printGxf(f_out, exonPrinting, tracklabel, NULL, decodeChars);
}
++gfs_i;
}
}
GTData* tdata=(GTData*)(t.uptr);
if (tdata->replaced_by!=NULL) continue;
if (process_transcript(gfasta, t)) {
t.udata|=4; //tag it as valid
numvalid++;
if (f_out) {
if (tdata->geneinfo) tdata->geneinfo->finalize();
out_counter++;
if (fmtGTF) t.printGxf(f_out, exonPrinting, tracklabel, NULL, decodeChars);
else {
if (firstGff3Print) { printGff3Header(f_out, args);firstGff3Print=false; }
//print the parent first, if any
if (t.parent!=NULL && ((t.parent->udata & 4)==0)) {
GTData* pdata=(GTData*)(t.parent->uptr);
if (pdata && pdata->geneinfo!=NULL)
pdata->geneinfo->finalize();
t.parent->printGxf(f_out, exonPrinting, tracklabel, NULL, decodeChars);
t.parent->udata|=4;
}
t.printGxf(f_out, exonPrinting, tracklabel, NULL, decodeChars);
}
}//GFF/GTF output requested
} //valid transcript
} //for each rna
//print the rest of the isolated pseudo/gene/region features not printed yet
if (f_out) {
while (gfs_i<gdata->gfs.Count()) {
GffObj& gfst=*(gdata->gfs[gfs_i]);
if ((gfst.udata&4)==0) { //never printed
gfst.udata|=4;
if (firstGff3Print) { printGff3Header(f_out, args);firstGff3Print=false; }
if (gfst.exons.Count()==0 && gfst.children.Count()==0 && forceExons)
gfst.addExon(gfst.start,gfst.end);
gfst.printGxf(f_out, exonPrinting, tracklabel, NULL, decodeChars);
}
++gfs_i;
}
}
} //for each genomic seq
} //not clustered
if (f_repl && f_repl!=stdout) fclose(f_repl);
seqinfo.Clear();
//if (faseq!=NULL) delete faseq;
//if (gcdb!=NULL) delete gcdb;
GFREE(rfltGSeq);
FRCLOSE(f_in);
FWCLOSE(f_out);
FWCLOSE(f_w);
FWCLOSE(f_x);
FWCLOSE(f_y);
}
/* chain function
* this function does the actual processing
*/
static GstFlowReturn
gst_motion_cells_chain (GstPad * pad, GstBuffer * buf)
{
GstMotioncells *filter;
filter = gst_motion_cells (GST_OBJECT_PARENT (pad));
if (filter->calculate_motion) {
double sensitivity;
int framerate, gridx, gridy, motionmaskcells_count, motionmaskcoord_count,
motioncells_count, i;
int thickness, success, motioncellsidxcnt, numberOfCells,
motioncellsnumber, cellsOfInterestNumber;
int mincellsOfInterestNumber, motiondetect;
char *datafile;
bool display, changed_datafile, useAlpha;
gint64 starttime;
motionmaskcoordrect *motionmaskcoords;
motioncellidx *motionmaskcellsidx;
cellscolor motioncellscolor;
motioncellidx *motioncellsidx;
g_mutex_lock (filter->propset_mutex);
buf = gst_buffer_make_writable (buf);
filter->cvImage->imageData = (char *) GST_BUFFER_DATA (buf);
if (filter->firstframe) {
setPrevFrame (filter->cvImage, filter->id);
filter->firstframe = FALSE;
}
sensitivity = filter->sensitivity;
framerate = filter->framerate;
gridx = filter->gridx;
gridy = filter->gridy;
display = filter->display;
motionmaskcoord_count = filter->motionmaskcoord_count;
motionmaskcoords =
g_new0 (motionmaskcoordrect, filter->motionmaskcoord_count);
for (i = 0; i < filter->motionmaskcoord_count; i++) { //we need divide 2 because we use gauss pyramid in C++ side
motionmaskcoords[i].upper_left_x =
filter->motionmaskcoords[i].upper_left_x / 2;
motionmaskcoords[i].upper_left_y =
filter->motionmaskcoords[i].upper_left_y / 2;
motionmaskcoords[i].lower_right_x =
filter->motionmaskcoords[i].lower_right_x / 2;
motionmaskcoords[i].lower_right_y =
filter->motionmaskcoords[i].lower_right_y / 2;
}
motioncellscolor.R_channel_value =
filter->motioncellscolor->R_channel_value;
motioncellscolor.G_channel_value =
filter->motioncellscolor->G_channel_value;
motioncellscolor.B_channel_value =
filter->motioncellscolor->B_channel_value;
if ((filter->changed_gridx || filter->changed_gridy
|| filter->changed_startime)) {
if ((g_strcmp0 (filter->cur_datafile, NULL) != 0)) {
GFREE (filter->cur_datafile);
filter->datafileidx++;
filter->cur_datafile =
g_strdup_printf ("%s-%d.%s", filter->basename_datafile,
filter->datafileidx, filter->datafile_extension);
filter->changed_datafile = TRUE;
motion_cells_free_resources (filter->id);
}
if (filter->motioncells_count > 0)
gst_motioncells_update_motion_cells (filter);
if (filter->motionmaskcells_count > 0)
gst_motioncells_update_motion_masks (filter);
filter->changed_gridx = FALSE;
filter->changed_gridy = FALSE;
filter->changed_startime = FALSE;
}
datafile = g_strdup (filter->cur_datafile);
filter->cur_buff_timestamp = (GST_BUFFER_TIMESTAMP (buf) / GST_MSECOND);
filter->starttime +=
(filter->cur_buff_timestamp - filter->prev_buff_timestamp);
starttime = filter->starttime;
if (filter->changed_datafile || filter->diff_timestamp < 0)
filter->diff_timestamp =
(gint64) (GST_BUFFER_TIMESTAMP (buf) / GST_MSECOND);
changed_datafile = filter->changed_datafile;
motionmaskcells_count = filter->motionmaskcells_count;
motionmaskcellsidx = g_new0 (motioncellidx, filter->motionmaskcells_count);
for (i = 0; i < filter->motionmaskcells_count; i++) {
motionmaskcellsidx[i].lineidx = filter->motionmaskcellsidx[i].lineidx;
motionmaskcellsidx[i].columnidx = filter->motionmaskcellsidx[i].columnidx;
}
motioncells_count = filter->motioncells_count;
motioncellsidx = g_new0 (motioncellidx, filter->motioncells_count);
for (i = 0; i < filter->motioncells_count; i++) {
motioncellsidx[i].lineidx = filter->motioncellsidx[i].lineidx;
motioncellsidx[i].columnidx = filter->motioncellsidx[i].columnidx;
}
useAlpha = filter->usealpha;
thickness = filter->thickness;
success =
perform_detection_motion_cells (filter->cvImage, sensitivity, framerate,
gridx, gridy,
(gint64) (GST_BUFFER_TIMESTAMP (buf) / GST_MSECOND) -
filter->diff_timestamp, display, useAlpha, motionmaskcoord_count,
motionmaskcoords, motionmaskcells_count, motionmaskcellsidx,
motioncellscolor, motioncells_count, motioncellsidx, starttime,
datafile, changed_datafile, thickness, filter->id);
if ((success == 1) && (filter->sent_init_error_msg == false)) {
char *initfailedreason;
int initerrorcode;
GstStructure *s;
GstMessage *m;
initfailedreason = getInitDataFileFailed (filter->id);
initerrorcode = getInitErrorCode (filter->id);
s = gst_structure_new ("motion", "init_error_code", G_TYPE_INT,
initerrorcode, "details", G_TYPE_STRING, initfailedreason, NULL);
m = gst_message_new_element (GST_OBJECT (filter), s);
gst_element_post_message (GST_ELEMENT (filter), m);
filter->sent_init_error_msg = TRUE;
}
if ((success == -1) && (filter->sent_save_error_msg == false)) {
char *savefailedreason;
int saveerrorcode;
GstStructure *s;
GstMessage *m;
savefailedreason = getSaveDataFileFailed (filter->id);
saveerrorcode = getSaveErrorCode (filter->id);
s = gst_structure_new ("motion", "save_error_code", G_TYPE_INT,
saveerrorcode, "details", G_TYPE_STRING, savefailedreason, NULL);
m = gst_message_new_element (GST_OBJECT (filter), s);
gst_element_post_message (GST_ELEMENT (filter), m);
filter->sent_save_error_msg = TRUE;
}
if (success == -2) { //frame dropped
filter->prev_buff_timestamp = filter->cur_buff_timestamp;
//free
GFREE (datafile);
GFREE (motionmaskcoords);
GFREE (motionmaskcellsidx);
GFREE (motioncellsidx);
g_mutex_unlock (filter->propset_mutex);
return gst_pad_push (filter->srcpad, buf);
}
filter->changed_datafile = getChangedDataFile (filter->id);
motioncellsidxcnt = getMotionCellsIdxCnt (filter->id);
numberOfCells = filter->gridx * filter->gridy;
motioncellsnumber = motioncellsidxcnt / MSGLEN;
cellsOfInterestNumber = (filter->motioncells_count > 0) ? //how many cells interest for us
(filter->motioncells_count) : (numberOfCells);
mincellsOfInterestNumber =
floor ((double) cellsOfInterestNumber * filter->threshold);
motiondetect = (motioncellsnumber >= mincellsOfInterestNumber) ? 1 : 0;
if ((motioncellsidxcnt > 0) && (motiondetect == 1)) {
char *detectedmotioncells;
filter->last_motion_timestamp = GST_BUFFER_TIMESTAMP (buf);
detectedmotioncells = getMotionCellsIdx (filter->id);
if (detectedmotioncells) {
filter->consecutive_motion++;
if ((filter->previous_motion == false)
&& (filter->consecutive_motion >= filter->minimum_motion_frames)) {
GstStructure *s;
GstMessage *m;
filter->previous_motion = true;
filter->motion_begin_timestamp = GST_BUFFER_TIMESTAMP (buf);
s = gst_structure_new ("motion", "motion_cells_indices",
G_TYPE_STRING, detectedmotioncells, "motion_begin", G_TYPE_UINT64,
filter->motion_begin_timestamp, NULL);
m = gst_message_new_element (GST_OBJECT (filter), s);
gst_element_post_message (GST_ELEMENT (filter), m);
} else if (filter->postallmotion) {
GstStructure *s;
GstMessage *m;
filter->motion_timestamp = GST_BUFFER_TIMESTAMP (buf);
s = gst_structure_new ("motion", "motion_cells_indices",
G_TYPE_STRING, detectedmotioncells, "motion", G_TYPE_UINT64,
filter->motion_timestamp, NULL);
m = gst_message_new_element (GST_OBJECT (filter), s);
gst_element_post_message (GST_ELEMENT (filter), m);
}
} else {
GstStructure *s;
GstMessage *m;
s = gst_structure_new ("motion", "motion_cells_indices", G_TYPE_STRING,
"error", NULL);
m = gst_message_new_element (GST_OBJECT (filter), s);
gst_element_post_message (GST_ELEMENT (filter), m);
}
} else {
filter->consecutive_motion = 0;
if ((((GST_BUFFER_TIMESTAMP (buf) -
filter->last_motion_timestamp) / 1000000000l) >=
filter->gap)
&& (filter->last_motion_timestamp > 0)) {
GST_DEBUG ("POST MOTION FINISHED MSG\n");
if (filter->previous_motion) {
GstStructure *s;
GstMessage *m;
filter->previous_motion = false;
s = gst_structure_new ("motion", "motion_finished", G_TYPE_UINT64,
filter->last_motion_timestamp, NULL);
m = gst_message_new_element (GST_OBJECT (filter), s);
gst_element_post_message (GST_ELEMENT (filter), m);
}
}
}
if (filter->postnomotion > 0) {
guint64 last_buf_timestamp = GST_BUFFER_TIMESTAMP (buf) / 1000000000l;
if ((last_buf_timestamp -
(filter->last_motion_timestamp / 1000000000l)) >=
filter->postnomotion) {
GST_DEBUG ("POST NO MOTION MSG\n");
if ((last_buf_timestamp -
(filter->last_nomotion_notified / 1000000000l)) >=
filter->postnomotion) {
GstStructure *s;
GstMessage *m;
filter->last_nomotion_notified = GST_BUFFER_TIMESTAMP (buf);
s = gst_structure_new ("motion", "no_motion", G_TYPE_UINT64,
filter->last_motion_timestamp, NULL);
m = gst_message_new_element (GST_OBJECT (filter), s);
gst_element_post_message (GST_ELEMENT (filter), m);
}
}
}
filter->prev_buff_timestamp = filter->cur_buff_timestamp;
//free
GFREE (datafile);
GFREE (motionmaskcoords);
GFREE (motionmaskcellsidx);
GFREE (motioncellsidx);
g_mutex_unlock (filter->propset_mutex);
}
return gst_pad_push (filter->srcpad, buf);
}
int main(int argc, char * const argv[]) {
//GArgs args(argc, argv, "hg:c:s:t:o:p:help;genomic-fasta=COV=PID=seq=out=disable-flag;test=");
GArgs args(argc, argv, opts);
fprintf(stderr, "Command line was:\n");
args.printCmdLine(stderr);
args.printError(USAGE, true);
//if (args.getOpt('h') || args.getOpt("help"))
if (args.getOpt(OPT_HELP))
{
GMessage("%s\n", USAGE);
exit(1);
}
if (args.getOpt(OPT_NUM)) {
GStr snum(args.getOpt(OPT_NUM));
int num=snum.asInt();
char* numstr=commaprintnum(num);
GMessage("Number %d written with commas: %s\n", num, numstr);
GFREE(numstr);
}
//---
GHash<GVec<int> > ends;
/*
testGPVec();
//exit(0);
//uint pos=3;
//GStr spos((int)pos);
//GVec<int> *ev=ends[spos.chars()];
GPVec<Gint> v;
int r(5);
int rr=v.Add(new Gint(3));
//if (rr<0) {
// GMessage("Error adding 0! (code %d)\n",rr);
// }
v.Add(new Gint(r));
v.Add(new Gint(2));
v.Add(new Gint(1));
v.Add(new Gint(4));
rr=v.Add(new Gint(0));
v[rr]->v=-1;
v.Sort(cmpGint);
GMessage("collection has %d elements:\n",v.Count());
for (int i=0;i<v.Count();i++) {
GMessage("v[%d]=%d;\n",i,v[i]->v);
}
exit(0);
*/
//---
int numopts=args.startOpt();
if (numopts)
GMessage("#### Recognized %d option arguments:\n", numopts);
int optcode=0;
while ((optcode=args.nextCode())) {
char* r=args.getOpt(optcode);
GMessage("%14s\t= %s\n", args.getOptName(optcode), (r[0]==0)?"True":r);
}
int numargs=args.startNonOpt();
if (numargs>0) {
GMessage("\n#### Found %d non-option arguments given:\n", numargs);
char* a=NULL;
while ((a=args.nextNonOpt())) {
GMessage("%s\n",a);
}
}
GStr s=args.getOpt('t');
if (!s.is_empty()) {
GStr token;
GMessage("Tokens in \"%s\" :\n",s.chars());
s.startTokenize(";,: \t");
int c=1;
while (s.nextToken(token)) {
GMessage("token %2d : \"%s\"\n",c,token.chars());
c++;
}
}
if (args.getOpt(OPT_BITVEC)) {
uint numbits=4156888234;
GBitVec bits(numbits);
GMessage(">>> -- BitVec(%u) created (size=%u, mem=%lu) -- \n", numbits, bits.size(),
bits.getMemorySize());
bits[405523342]=true;
GMessage(" memory size: %lu , size()=%u, count()=%d \n", bits.getMemorySize(), bits.size(), bits.count());
/*
//GMessage(">>> -- Start BitVec Test -- \n");
if (bits[1092]) bitError(1092);
bits.resize(2049);
if (bits[2048]) bitError(2048);
bits[2048]=true;
if (!bits[2048]) bitError(2048);
bits.resize(4097);
if (!bits[2048]) bitError(2048);
if (bits[4096]) bitError(4096);
bits[4096]=true;
if (!bits[4096]) bitError(4096);
GBitVec bits2(64);
Gswap(bits, bits2);
if (!bits2[2048]) bitError(2048);
if (!bits2[4096]) bitError(4096);
*/
//GMessage("<<< -- End BitVec Test (size: %d, count: %d, bits2 size=%d, count=%d) --\n",
/// bits.size(), bits.count(), bits2.size(), bits2.count());
}
}
