#######################
2012-00-00
Purpose:
    a) 
Details:
1) main.c::fun()
    a)
#######################

2012-00-00
Purpose:
    a)fix bug for map, zombie process cause by fclose
Details:
1) readseq1by1.c::openFile4read()
    a)delete these lines
#if defined(SIGCHLD)
signal(SIGCHLD,SIG_IGN);
#elif defined(SIGCLD)
signal(SIGCLD,SIG_IGN);
#endif
    b)
    || ((lib_array[i].curr_type == 1) && feof (lib_array[i].fp1) && feof (lib_array[i].fp2))
    || ((lib_array[i].curr_type != 1) && (lib_array[i].curr_type != 4) && feof (lib_array[i].fp1)))
    -->
    || ((lib_array[i].curr_type == 1 || lib_array[i].curr_type == 2) && feof (lib_array[i].fp1) && feof (lib_array[i].fp2))
    || ((lib_array[i].curr_type != 1 && lib_array[i].curr_type != 2) && (lib_array[i].curr_type != 4) && feof (lib_array[i].fp1)))

2012-07-31
Purpose:
    a) fix bug for map, zombie process cause by fclose
Details:
1) readseq1by1.c::openFile4read()
    a)
        sprintf (cmd, "gzip -dc %s", fname);
        fp = popen (cmd, "r");
        free (cmd);
        return fp;
        -->
        sprintf (cmd, "gzip -dc %s", fname);
        fp = popen (cmd, "r");

#if defined(SIGCHLD) 
signal(SIGCHLD,SIG_IGN);        
#elif defined(SIGCLD) 
signal(SIGCLD,SIG_IGN); 
#endif

        free (cmd);
        return fp;

2012-07-13
Purpose:
    a) fix bugs for output, delete the empty line of the output(*.edge.gz, *.contig, *.scafSeq)
Details:
1) concatenateEdge.c::printEdgeSeq()
    a)
        if ((i + overlaplen + 1) % 100 == 0)
        -->
        if ((i + overlaplen + 1) % 100 == 0 && i < len - 1)
2) iterate.c::output_1contig()
    a)
        if ((i + overlaplen + 1) % 100 == 0)    
        -->
        if ((i + overlaplen + 1) % 100 == 0 && i < edge->length - 1)    
3) output_contig.c::output_1contig()
    a)
        if ((i + overlaplen + 1) % 100 == 0)
        -->    
        if ((i + overlaplen + 1) % 100 == 0 && i < edge->length - 1)    
4) output_pregraph.c::output_1edge()
    a)
        if ((i + 1) % 100 == 0)
        --> 
        if ((i + 1) % 100 == 0 && i < edge->length - 1) 
5) seq.c::printTightString()
    a)
        if ((i + 1) % 100 == 0)
        -->
        if ((i + 1) % 100 == 0 && i < len - 1)
6) prlReadFillGap.c
    a)outputTightStr2Visual()
            if ((++column) % 100 == 0)
            -->
            if ((++column) % 100 == 0 && i < end - 1)

            if ((++column) % 100 == 0)
            -->
            if ((++column) % 100 == 0 && i < length - 1 - start)

    b)outputTightStrLowerCase2Visual()
        if ((++column) % 100 == 0)
        -->
        if ((++column) % 100 == 0 && i < end - 1)

    c)outputScafSeq()
    fprintf (fo, "\n");
    -->
    if(column % 100 != 0)
        fprintf (fo, "\n");

            if(i != 0 && i % 100 == 0)  fprintf(fo3, "\n");
            -->
            if(i != 0 && i % 100 == 0 && i < ctg_start_pos - 1)  fprintf(fo3, "\n");

            if(i != 0 && i % 100 == 0)  fprintf(fo3, "\n");
            -->
            if(i != 0 && i % 100 == 0 && i < ctg_start_pos - 1)  fprintf(fo3, "\n");

    d)output_ctg()
    fprintf (fo, "\n");
    -->
    if(len % 100 != 0)
        fprintf (fo, "\n");

    e)output1gap()
    fprintf (fo, "\n");
    -->
    if(column % 100 != 0)
        fprintf (fo, "\n");

2012-07-13
Purpose:
    a) fix bugs for removeWeakEdges
Details:
1) cutTip_graph.c::removeWeakEdges()
    a)
    for (i = 1; i <= num_ed; i++)
    {
        if (edge_array[i].deleted || edge_array[i].length == 0 || edge_array[i].length > lenCutoff || EdSameAsTwin (i))
        {
            continue;
        }

        bal_ed = getTwinEdge (i);
        arcRight = arcCount (i, &arcRight_n);

        if (arcRight_n > 1 || !arcRight || arcRight->multiplicity > multiCutoff)
        {
            continue;
        }

        arcLeft = arcCount (bal_ed, &arcLeft_n);

        if (arcLeft_n > 1 || !arcLeft || arcLeft->multiplicity > multiCutoff)
        {
            continue;
        }

        destroyEdge (i);
        counter++;
    }
    -->
    while(counter)
    {
        counter = 0;
        
        for (i = 1; i <= num_ed; i++)
        {
            if (edge_array[i].deleted || edge_array[i].length == 0 || edge_array[i].length > lenCutoff || EdSameAsTwin (i))
            {
                continue;
            }

            bal_ed = getTwinEdge (i);
            arcRight = arcCount (i, &arcRight_n);

            if (arcRight_n > 1 || !arcRight || arcRight->multiplicity > multiCutoff)
            {
                continue;
            }

            arcLeft = arcCount (bal_ed, &arcLeft_n);

            if (arcLeft_n > 1 || !arcLeft || arcLeft->multiplicity > multiCutoff)
            {
                continue;
            }

            destroyEdge (i);
            counter++;
        }

        fprintf (stderr,"%d weak inner edges are destroyed.\n", counter);
    }

2012-07-13
Purpose:
    a) Make the output of SOAPdevovo the same when using different threads.
Details:
1) contig.c::call_heavygraph()
    a)
    if (repeatSolve)
    {
        ret = loadPathBin (graphfile);
    }
    -->
    if (repeatSolve)
    {
        ret = loadPathBin (graphfile);
    }
    swapedge();
    sortedge();
    freshArc();

2012-07-13
Purpose:
    a) fix bugs for merge_linearV2, get the correct coverage of edge, l_links maybe change in the original process
Details:
1)node2edge.c:merge_linearV2()
    a)
    while (nStack->item_c > 1)
    {
        temp = (KMER_PT *) stackPop (nStack);
        del_node = temp->node;
        del_node->inEdge = 1;
        symbol += get_kmer_left_covs (*del_node);

        if (temp->isSmaller)
        {
            del_node->l_links = edge_c;
            del_node->twin = (unsigned char) (bal_edge + 1);
        }
        else
        {
            del_node->l_links = edge_c + bal_edge;
            del_node->twin = (unsigned char) (-bal_edge + 1);
        }

        tightSeq[char_index--] = lastCharInKmer (temp->kmer);
    }
    -->
    while (nStack->item_c > 1)
    {
        temp = (KMER_PT *) stackPop (nStack);
        del_node = temp->node;
        del_node->inEdge = 1;
        symbol += get_kmer_left_covs (*del_node);       

        tightSeq[char_index--] = lastCharInKmer (temp->kmer);
    }

    stackRecover(nStack);
    temp = (KMER_PT *) stackPop (nStack);
    while(nStack->item_c > 1)
    {
        temp = (KMER_PT *) stackPop (nStack);
        del_node = temp->node;
        del_node->inEdge = 1;
        if (temp->isSmaller)
        {
            del_node->l_links = edge_c;
            del_node->twin = (unsigned char) (bal_edge + 1);
        }
        else
        {
            del_node->l_links = edge_c + bal_edge;
            del_node->twin = (unsigned char) (-bal_edge + 1);
        }
    }

2012-07-13
Purpose:
    a) fix bugs for removeMinorTips
Details:
1)cutTipPreGraph.c:removeMinorTips ()
    a)
    for (i = 0; i < thrd_num; i++)
    {
        set = KmerSets[i];
        flag = 1;

        while (flag)
        {
            flag = 0;
            set->iter_ptr = 0;

            while (set->iter_ptr < set->size)
            {
                if (!is_kmer_entity_null (set->flags, set->iter_ptr))
                {
                    rs = set->array + set->iter_ptr;
                    flag += clipTipFromNode (rs, cut_len_tip, 0);
                }

                set->iter_ptr++;
            }
        }

        fprintf (stderr,"Remove minor tips in kmer set %d is done.\n", i);
    }
    -->
    while (flag)
    {
        flag = 0;
        
        for (i = 0; i < thrd_num; i++)
        {
            set = KmerSets[i];
            set->iter_ptr = 0;

            while (set->iter_ptr < set->size)
            {
                if (!is_kmer_entity_null (set->flags, set->iter_ptr))
                {
                    rs = set->array + set->iter_ptr;
                    if (!rs->linear && !rs->deleted)
                    {
                        flag += clipTipFromNode (rs, cut_len_tip, 0);
                    }
                }
                set->iter_ptr++;
            }
        }
        fprintf (stderr,"Remove minor tips in kmer set is done(round %d).\n", round++);
    }

2012-07-13
Purpose:
    a) merge 63mer and 127mer version
Details:
    1) Use #define MER127 to define 127mer version
    2) Use #define MER63 to define 63mer version

2012-02-15
Purpose:
	a) fix bugs of AIORead in reading fastq file which has "@" at the beginning of qulity line. Edited by panqi.
Details:
1) prlHashReads.c::AIORead()
	a) if((curr_type == 2) || (curr_type == 6))
	   {
		for (i = get - 1; (temp[i] != '@') || (temp[i-1] != '\n') || (temp[i-2] == '+')|| (temp[i-3] == '/'); i--);
		for (i2 = i - 2; temp[i2] != '\n'; i2--);
		if (temp[i2 + 1] == '+')
		{
			for(i3 = i2 - 1; temp[i3] != '\n'; i3--);
		        for(i2 = i3 - 1; temp[i2] != '\n'; i2--);
		        if(temp[i2 + 1] == '@') {i = i2 + 1;}
		}
	   }
	   -->
	   int num;
	   if((curr_type == 2) || (curr_type == 6))
           {
                num = 0;
                for (i = get - 1; (temp[i] != '@') || (temp[i-1] != '\n'); i--)
                {
                        if(temp[i] == '\n') {num++;}
                }

                if (num <= 1)
                {
                        for (i2 = i - 2; temp[i2] != '\n'; i2--);
                        if (temp[i2 + 1] == '+')
                        {
                                for(i2 = i2 - 1; temp[i2] != '\n'; i2--);
                                if(temp[i2 + 1] != '+') {for (i = i2 - 1; (temp[i] != '@') || (temp[i-1] != '\n'); i--);}
                        }
                }
           }


2012-02-14
Purpose:
	a) seperate codes related to visualization from other codes. Edited by panqi.
Details:
	1) prlReadFillGap.c
	   codes related to visualization

Purpose:
	a) fix bugs of AIORead in reading fastq file which has "@" at the beginning of qulity line. Edited by panqi.
Details:
1) prlHashReads.c::AIORead()
	a) if((curr_type == 2) || (curr_type == 6))
		for (i = get - 1; (temp[i] != '@') || (temp[i-1] != '\n') || (temp[i-2] == '+')|| (temp[i-3] == '/'); i--);
	   else if((curr_type == 1) || (curr_type == 3) ||(curr_type == 5))
		for (i = get - 1; temp[i] != '>';i--) ;
	   -->
	   if((curr_type == 2) || (curr_type == 6))
	   {
		for (i = get - 1; (temp[i] != '@') || (temp[i-1] != '\n') || (temp[i-2] == '+')|| (temp[i-3] == '/'); i--);
		for (i2 = i - 2; temp[i2] != '\n'; i2--);
		if (temp[i2 + 1] == '+')
		{
			for(i3 = i2 - 1; temp[i3] != '\n'; i3--);
		        for(i2 = i3 - 1; temp[i2] != '\n'; i2--);
		        if(temp[i2 + 1] == '@') {i = i2 + 1;}
		}
	   }
	   else if((curr_type == 1) || (curr_type == 3) ||(curr_type == 5)
		for (i = get - 1; temp[i] != '>';i--) ;

Purpose:
	a) fix bug of AIORead to deal with the case that read length is shorter than Kmer size. Edited by panqi.
Details:
1) prlHashReads.c::prlRead2HashTable()
	a) while (start1 < offset1 || start2 < offset2)
	   {
		if (turn == 1)
		{
			turn = 2;
			readseqInLib (seqBuffer[read_c], next_name, &(lenBuffer[read_c]), readBuffer1, &start1, offset1, libNo);
			if ((++i) % 100000000 == 0)
				printf ("--- %lldth reads\n", i);
			if (lenBuffer[read_c] < 0)
				printf ("read len %d\n", lenBuffer[read_c]);
			if (lenBuffer[read_c] < overlaplen + 1)
				continue;
			......
		}
		if (turn == 2)
		{
			turn = 1;
			readseqInLib (seqBuffer[read_c], next_name, &(lenBuffer[read_c]), readBuffer2, &start2, offset2, libNo);
			if ((++i) % 100000000 == 0)
				printf ("--- %lldth reads\n", i);
			if (lenBuffer[read_c] < 0)
				printf ("read len %d\n", lenBuffer[read_c]);
			if (lenBuffer[read_c] < overlaplen + 1)
				continue;
			......
		}
	   }
	   -->
	   while (start1 < offset1 || start2 < offset2)
	   {
		if (turn == 1)
		{
			turn = 2;
			readseqInLib (seqBuffer[read_c], next_name, &(lenBuffer[read_c]), readBuffer1, &start1, offset1, libNo);
			if ((++i) % 100000000 == 0)
				printf ("--- %lldth reads\n", i);
			if (lenBuffer[read_c] < 0)
				printf ("read len %d\n", lenBuffer[read_c]);
			if (lenBuffer[read_c] < overlaplen + 1)
			{
				if(start1>=offset1)
				{
					start1=0;
					flag1=AIORead (&aio1, &offset1, readBuffer1, cach1, &rt1, lib_array[libNo].curr_type);
				}
				continue;
			}
		}
		if (turn == 2)
		{
			turn = 1;
			readseqInLib (seqBuffer[read_c], next_name, &(lenBuffer[read_c]), readBuffer2, &start2, offset2, libNo);
			if ((++i) % 100000000 == 0)
				printf ("--- %lldth reads\n", i);
			if (lenBuffer[read_c] < 0)
				printf ("read len %d\n", lenBuffer[read_c]);
			if (lenBuffer[read_c] < overlaplen + 1)
			{
				if((flag2 == 2) && (start2 >= offset2))
					break;
				if(start2 >= offset2)
				{
					start2=0;
					flag2 = AIORead (&aio2, &offset2, readBuffer2, cach2, &rt2, lib_array[libNo].curr_type);
				}
				continue;
			}
		}
	   }


2012-01-09
Purpose:
	a) add a parameter "-V" so that some information for visualization can be output by setting this parameter. Edited by panqi.

Details:
	a) prlReadFillGap.c
	   codes related to visualization

2011-12-16
Purpose:
	a) Fix a bug of scaffolding which may lead to segmentation fault or endless loop

Details:
1) Function: orderContig.c::linearC2C()
	a) new:
	   if (usCtgCounter != oldsize)
	   {
		unsigned int prev_c2 = upstreamCTG[usCtgCounter+1];
		unsigned int bal_prev_c2 = getTwinCtg (prev_c2);
		setConnectMask (bal_prev_c2, c2, 1);
		setConnectMask (bal_prev_c2, c2, 0);

		int i = usCtgCounter+1;
		for ( ; i <= oldsize; i++)
		{
			contig_array[upstreamCTG[i]].from_vt = prev_c2;
			contig_array[getTwinCtg(upstreamCTG[i])].to_vt = bal_prev_c2;
		}

		if ((cn_temp = getCntBetween(c1,c2)) != NULL)
		{
			setConnectMask (c1, c2, 0);
			setConnectDelete (c1, c2, 0, 0);
			return 1;
		}
	   }

2011-11-08
Purpose: 
	a) Change default value of '-b' from 0 to 1.5. Update relative statements.

	b) Change default value of '-B' from 0 to 0.6. Update relative statements.

	c) Update some code so it can read '*.readInGap' and '*.readOnContig' produced by V1.05.

Details:
1) Variant: inc/globe.h
	a) float cvg4SNP = 0.0;
	   -->
	   float cvg4SNP = 0.6;

	b) float ins_var_idx=0;
	   -->
	   float ins_var_idx=1.5

2) Function: scaffold.c::display_scaff_usage()
	a) printf ("\nscaff -g inputGraph [-F -m -u -S -w] [-G gapLenDiff -L minContigLen -c minContigCvg -C maxContigCvg -b insertSizeUpperBound -N genomeSize -p n_cpu]\n");
	   -->
	   printf ("\nscaff -g inputGraph [-F -m -u -S -w] [-G gapLenDiff -L minContigLen -c minContigCvg -C maxContigCvg -b insertSizeUpperBound -B bubbleCoverage -N genomeSize -p n_cpu]\n");

	b) printf ("  -b <float>\tinsertSizeUpperBound: (b*avg_ins) will be used as upper bound of insert size for long insert size ( > 1000) when handling pair-end connections between contigs if b is set to larger than 1, [0]\n");
	   -->
	   printf ("  -b <float>\tinsertSizeUpperBound: (b*avg_ins) will be used as upper bound of insert size for long insert size ( > 1000) when handling pair-end connections between contigs if b is set to larger than 1, [1.5]\n");

	c) printf ("  -B <float>\tbubbleCoverage: remove contig with lower cvoerage in bubble structure if both contigs' coverage are smaller than bubbleCoverage*avgCvg, [0]\n");
	   -->
	   printf ("  -B <float>\tbubbleCoverage: remove contig with lower cvoerage in bubble structure if both contigs' coverage are smaller than bubbleCoverage*avgCvg, [0.6]\n");

3) Variant: main.c::pipeline()
	a) unsigned char getB is now related with "bubbleCoverage" instead of "insertSizeUpperBound".

	b) new:
	   unsigned char getb;	//related with "insertSizeUpperBound".

	c) new:
	   char bubble_coverage_s[16];

	d) new:
	   float bubble_coverage = 0.0;

	e) int insert_size_bound = 0;
	   -->
	   float insert_size_bound = 0.0;

4) Function: main.c::pipeline()
	a) while ((copt = getopt (argc, argv, "a:s:o:K:M:L:p:G:d:D:RuFk:fc:C:b:N:w")) != EOF)
	   -->
	   while ((copt = getopt (argc, argv, "a:s:o:K:M:L:p:G:d:D:RuFk:fc:C:b:B:N:w")) != EOF)

	b) new:
	   case 'B':
		getB = 1;
		sscanf (optarg, "\s", temp);
		bubble_coverage = atof (temp);
		break;

	c) new:
	   if (getB)
	   {
		options[index++] = "-B";
		sprintf (bubble_coverage_s, "%f", bubble_coverage);
		options[index++] = bubble_coverage_s;
	   }

5) Function: main.c::display_all_usage()
	a) printf ("\nSOAPdenovo all -s configFile -o outputGraph [-R -f -F -u -w] [-K kmer -p n_cpu -a initMemoryAssumption -d KmerFreqCutOff -D EdgeCovCutoff -M mergeLevel -k kmer_R2C, -G gapLenDiff -L minContigLen -c minContigCvg -C maxContigCvg -b insertSizeUpperBound -N genomeSize]\n");
	   -->
	   printf ("\nSOAPdenovo all -s configFile -o outputGraph [-R -f -F -u -w] [-K kmer -p n_cpu -a initMemoryAssumption -d KmerFreqCutOff -D EdgeCovCutoff -M mergeLevel -k kmer_R2C, -G gapLenDiff -L minContigLen -c minContigCvg -C maxContigCvg -b insertSizeUpperBound -B bubbleCoverage -N genomeSize]\n");

	b) printf ("  -b <float>\tinsertSizeUpperBound: (b*avg_ins) will be used as upper bound of insert size for large insert size ( > 1000) when handling pair-end connections between contigs if b is set to larger than 1, [0]\n");
	   -->
	   printf ("  -b <float>\tinsertSizeUpperBound: (b*avg_ins) will be used as upper bound of insert size for large insert size ( > 1000) when handling pair-end connections between contigs if b is set to larger than 1, [1.5]\n");

	c) new:
	   printf ("  -B <float>\tbubbleCoverage: remove contig with lower cvoerage in bubble structure if both contigs' coverage are smaller than bubbleCoverage*avgCvg, [0.6]\n");

6) Function: orderContig.c::PE2Links()
	a) gzFile *fp
	   -->
	   FILE *fp1;
	   gzFile *fp2;

	b) sprintf (name, "%s.readOnContig.gz", infile);
	   fp = gzopen (name, "r");
	   -->
	   if (COMPATIBLE_MODE == 1)
	   {
		sprintf(name,"%s.readOnContig",infile);
		fp1 = ckopen(name,"r");
	   }
	   else
	   {
		sprintf (name, "%s.readOnContig.gz", infile);
		fp2 = gzopen (name, "r");
	   }

	c) gzgets (fp, line, lineLen);
	   -->
	   if (COMPATIBLE_MODE == 1)
	   {
		fgets(line,lineLen,fp1);
	   }
	   else
	   {
		gzgets (fp2, line, lineLen);
	   }

	d) flag += connectByPE_grad(fp,i,line);
	   -->
	   if (COMPATIBLE_MODE == 1)
	   {
		flag += connectByPE_grad(fp1,i,line);
	   }
	   else
	   {
		flag += connectByPE_grad_gz(fp2,i,line);
	   }

	e) gzclose(fp);
	   -->
	   if (COMPATIBLE_MODE == 1)
	   {
		fclose(fp1);
	   }
	   else
	   {
		gzclose(fp2);
	   }

7) Function: inc/extfunc.h
	a) connectByPE_grad( gzFile* fp, int peGrad, char * line );
	   -->
	   extern int connectByPE_grad_gz ( gzFile * fp, int peGrad, char * line );

8) Function: attachPEinfo.c
	a) recover:
	   int connectByPE_grad (FILE * fp, int peGrad, char *line)

	b) int connectByPE_grad (gzFile * fp, int peGrad, char *line)
	   -->
	   int connectByPE_grad_gz (gzFile * fp, int peGrad, char *line)

9) Function: prlReadFillGap.c::loadReads4gap()
	a) update some code to read '*.readInGap'.

2011-11-04
1) Variant: inc/globe.h
	a) new:
	   float cvg4SNP = 0.0;

2) Variant: inc/extvab.h
	a) new:
	   extern float cvg4SNP;

3) struct: inc/def.h::struct contig
	a) new:
	   unsigned char bubbleInScaff: 1;

4) Function: scaffold.c::initenv()
	a) while ((copt = getopt (argc, argv, "g:L:p:G:N:c:C:b:FmuSw")) != EOF)
	   -->
	   while ((copt = getopt (argc, argv, "g:L:p:G:N:c:C:b:B:FmuSw")) != EOF)

	b) new:
	   case 'B':
		sscanf (optarg, "%s", temp);
		cvg4SNP = atof (temp) > 0 ? atof (temp) : 0.0;
		break;

5) Function: scaffold.c::display_scaff_usage()
	a) new:
	   printf ("  -B <float>\tbubbleCoverage: remove contig with lower cvoerage in bubble structure if both contigs' coverage are smaller than bubbleCoverage*avgCvg, [0]\n");

6) Function: loadGraph.c::loadUpdatedEdges()
	a) new:
	   contig_array[newIndex].bubbleInScaff = 0;

7) Variant: orderContig.c
	a) new:
	   static FILE *snp_fp = NULL;

8) Fucntion: orderContig.c::Links2Scaf()
	a) cvg4SNP = ((double)(0.5*cvgAvg) + sqrt(2*cvgAvg));
	   -->
	   if (cvg4SNP > 0.001)
	   {
		sprintf(name, "%s.bubbleInScaff", infile);
		snp_fp = ckopen(name, "w");
	   }
	   cvg4SNP = (double)(cvg4SNP*cvgAvg);

	b) new:
	   if (cvg4SNP > 0.001)
	   {
		fclose (snp_fp);
	   }

9) Fucntion: inc/extfunc.h
	a) new:
	   extern void outputTightStr(FILE *fp,char *tightStr,int start,int length, int outputlen,int revS,int *col);

10) Fcuntion: prlReadFillGap.c::outputTightStr()
	a) static void outputTightStr (FILE * fp, char *tightStr, int start, int length, int outputlen, int revS, int *col)
	   -->
	   void outputTightStr (FILE * fp, char *tightStr, int start, int length, int outputlen, int revS, int *col)

11) Function: orderContig.c
	a) new:
	   static void output_ctg(unsigned int ctg,FILE *fo)
	   {
		if(contig_array[ctg].length<1)
			return;
		int len;
		unsigned int bal_ctg = getTwinCtg(ctg);

		len = contig_array[ctg].length + overlaplen;

		int col = 0;
		if(contig_array[ctg].seq){
			fprintf(fo,">C%d %4.1f\n",ctg,(double)contig_array[ctg].cvg);
			outputTightStr(fo,contig_array[ctg].seq,0,len,len,0,&col);
		}
		else if(contig_array[bal_ctg].seq){
			fprintf(fo,">C%d %4.1f\n",bal_ctg,(double)contig_array[ctg].cvg);
			outputTightStr(fo,contig_array[bal_ctg].seq,0,len,len,0,&col);
		}
		fprintf(fo,"\n");
	   }

12) Function: orderContig.c::removeSNPCtg()
	a) unsigned int node1, node2;
	   -->
	   unsigned int node1, node2, bal_node1, bal_node2;

	b) new:
	   bal_node1 = getTwinCtg(node1);
	   bal_node2 = getTwinCtg(node2);

	c) getEndKmers(contig_array[getTwinCtg(node1)].seq, len1, 1, firstKmer1, lastKmer1);
	   -->
	   getEndKmers(contig_array[bal_node1].seq, len1, 1, firstKmer1, lastKmer1);

	d) getEndKmers(contig_array[getTwinCtg(node2)].seq, len2, 1, firstKmer2, lastKmer2);
	   -->
	   getEndKmers(contig_array[bal_node2].seq, len2, 1, firstKmer2, lastKmer2);

	e) new:
	   if (contig_array[node1].bubbleInScaff == 0 || contig_array[node2].bubbleInScaff == 0)
	   {
		contig_array[node1].bubbleInScaff = 1;
		contig_array[bal_node1].bubbleInScaff = 1;
		contig_array[node2].bubbleInScaff = 1;
		contig_array[bal_node2].bubbleInScaff = 1;
		output_ctg(node1, snp_fp);
		output_ctg(node2, snp_fp);
	   }

	f) transferCnt2RemainNode(getTwinCtg(node2), getTwinCtg(node1));
	   -->
	   transferCnt2RemainNode(bal_node2, bal_node1);

	g) transferCnt2RemainNode(getTwinCtg(node1), getTwinCtg(node2));
	   -->
	   transferCnt2RemainNode(bal_node1, bal_node2);

2011-11-03
1) Function: orderContig.c::scaffolding()
	a) if(pre_score == 0)
	   {
		*(unsigned int *)darrayPut(tempArray,prev_p)=0;
		mask_num++;
	   }
	   else if(now_ctg_score == 0)
	   {
		*(unsigned int *)darrayPut(tempArray,j)=0;
		mask_num++;
		if(j < tempCounter -1)
			continue;
	   }
	   -->
	   if(pre_score == 0)
	   {
		*(unsigned int *)darrayPut(tempArray,prev_p)=0;
		contig_array[prev_id].flag = 0;
		contig_array[getTwinCtg(prev_id)].flag = 0;
		mask_num++;
	   }
	   else if(now_ctg_score == 0)
	   {
		*(unsigned int *)darrayPut(tempArray,j)=0;
		contig_array[nowid].flag = 0;
		contig_array[getTwinCtg(nowid)].flag = 0;
		mask_num++;
		if(j < tempCounter -1)
			continue;
	   }

	b) if(contig_array[prev_id].cvg < contig_array[nowid].cvg)
	   {
		*(unsigned int *)darrayPut(tempArray,prev_p)=0;
	   }
	   else
	   {
		*(unsigned int *)darrayPut(tempArray,j)=0;
		if(j < tempCounter -1)
			continue;
	   }
	   -->
	   if(contig_array[prev_id].cvg < contig_array[nowid].cvg)
	   {
		*(unsigned int *)darrayPut(tempArray,prev_p)=0;
		contig_array[prev_id].flag = 0;
		contig_array[getTwinCtg(prev_id)].flag = 0;
	   }
	   else
	   {
		*(unsigned int *)darrayPut(tempArray,j)=0;
		contig_array[nowid].flag = 0;
		contig_array[getTwinCtg(nowid)].flag = 0;
		if(j < tempCounter -1)
			continue;
	   }

	c) if(score_count < 2)
	   {
		free((void *)score_array);
		continue;
	   }
	   -->
	   if(score_mask == 1 && num3 + num5 > 5 && score_count < 2)
	   {
		free((void *)score_array);
		--count;
		sum -= len;
		for (j=0; j<num3; j++)
		{
		    ctg = *(unsigned int *)darrayGet(scaf3,j);
		    contig_array[ctg].flag = 0;
		    contig_array[getTwinCtg(ctg)].flag = 0;
		}
		for (j=0; j<num5; j++)
		{
		    ctg = *(unsigned int *)darrayGet(scaf5,j);
		    contig_array[ctg].flag = 0;
		    contig_array[getTwinCtg(ctg)].flag = 0;
		}
		continue;
	   }

	d) after obtaining downstream contigs and upstream contigs
	   add:
	   if (num5+num3 == 1)
	   {
		contig_array[i].flag = 0;
		continue;
	   }
	   
	   
2011-10-27
1) Function: orderContig.c::checkScafConsist()
	a) if(!linkCount1||!linkCount2)
	   -->
	   if(!linkCount1)

2) Function: loadGraph.c::
	a) new:
	   int cut_len;
	   cut_len = COMPATIBLE_MODE == 0 ? overlaplen : 0;

	b) if(length != 0) contig_array[newIndex].length = length - overlaplen;
	   -->
	   if(length != 0) contig_array[newIndex].length = length - cut_len;

	c) counter += length - overlaplen;
	   cvgSum += cvg * (length - overlaplen);
	   -->
	   counter += length - cut_len;
	   cvgSum += cvg * (length - cut_len);

2011-10-24
1) Function: orderContig.c::checkScafConsist()
	a) boolean checkScafConsist(STACK *scafStack1, int len1, STACK *scafStack2)
	   -->
	   boolean checkScafConsist(STACK *scafStack1, int len1, STACK *scafStack2, int len2)

	b) int linkCount1 = getDSLink2Scaf(scafStack1,downwardTo1,downwardWt1);
	   int linkCount2 = getDSLink2Scaf(scafStack2,downwardTo2,downwardWt2);
	   -->
	   int linkCount1 = getDSLink2Scaf(scafStack1,downwardTo1,downwardWt1, len1);
	   int linkCount2 = getDSLink2Scaf(scafStack2,downwardTo2,downwardWt2, len2);

	c) annotate:
	   boolean flag2 = isLinkReliable(downwardWt2,linkCount2);

	d) if(!flag1||!flag2)
	   -->
	   if(!flag1)

	e) if (linkCount2 && flag2)
	   -->
	   if (linkCount2)

2) Function: orderContig.c::detectBreakScaff()
	a) int flag1 = checkScafConsist(scafStack1,scafStack2);
	   int flag2 = checkScafConsist(scafStack2,scafStack1);
	   -->
	   int flag1 = checkScafConsist(scafStack1, len1, scafStack2, len2);
	   int flag2 = checkScafConsist(scafStack2, len2, scafStack1, len1);

3) Function: 
	a) int getDSLink2Scaf(STACK *scafStack,DARRAY *SCAF,DARRAY *WT)
	   -->
	   int getDSLink2Scaf(STACK *scafStack,DARRAY *SCAF,DARRAY *WT, int total_len)

	b) new:
	   CONNECT *bind_cnt;
	   int len=0, gap;

	c) new:
	   bind_cnt = getBindCnt(ctg);
	   gap = bind_cnt ? bind_cnt->gapLen : 0;
	   len += contig_array[ctg].length + gap;

	d) if ((contig_array[ctg].mask && contig_array[ctg].length < 500) || !contig_array[ctg].downwardConnect)
	   -->
	   if ((contig_array[ctg].mask && contig_array[ctg].length < 500) || !contig_array[ctg].downwardConnect
		|| total_len - len > Insert_size)

2011-10-21
1) Function: orderContig.c::getDSLink2Scaf()
	a) new:
	   if (ite_cnt->newIns != 1)
	   {
		ite_cnt = ite_cnt->next;
		continue;
	   }

2011-10-19
1) Function: orderContig.c::getScaffold()
	a) len += contig_array[ctg].length;
	   -->
	   len += bindCnt->gapLen + contig_array[ctg].length;

2) Function: orderContig.c::getDSLink2Scaf()
	a) new:
	   unsigned int targetCtg,bal_targetCtg;

	b) new:
	   targetCtg = ite_cnt->contigID;
	   bal_targetCtg = getTwinCtg(targetCtg);

	c) if ((ite_cnt->mask && contig_array[ctg].length < 500) || ite_cnt->singleInScaf
	   -->
	   if ((ite_cnt->mask && contig_array[targetCtg].length < 500) || ite_cnt->singleInScaf

	d) if(contig_array[ctg].from_vt==contig_array[targetCtg].from_vt)
	   -->
	   if(contig_array[ctg].from_vt==contig_array[targetCtg].from_vt 
		|| (targetCtg==contig_array[targetCtg].from_vt && bal_targetCtg==contig_array[bal_targetCtg].from_vt))

2011-10-18
1) struct: inc/def.h::struct connection
	a) add new member: unsigned char newIns:1;
	   It's used to indicate whether the PE relationship between contigs is provided by the latest rank

2) Function: orderContig.c::inputLinks()
	a) add variants: CONNECT *cnt, *bal_cnt;

	b) add1Connect(ctg,toCtg,gap,wt,0);
	   add1Connect(bal_toCtg,bal_ctg,gap,wt,0);
	   -->
	   cnt = add1Connect(ctg,toCtg,gap,wt,0);
	   bal_cnt = add1Connect(bal_toCtg,bal_ctg,gap,wt,0);
	   if (cnt && insertS > 1000)
	   {
		cnt->newIns = bal_cnt->newIns = 1;
	   }

3) Function: orderContig.c::detectBreakScaff()
	a) new function, developed from detectBreakScaf().

4) Function: orderContig.c::clearNewInsFlag()
	a) new function.

5) Function: orderContig.c::Links2Scaf()
	a) if(i==gradsCounter-1&&!isPrevSmall&&smallPE)
		detectBreakScaf();
	   -->
	   if(Insert_size > 1000)
		detectBreakScaff();
	b) new:
	   if (Insert_size > 1000 && i != gradsCounter-1)
		clearNewInsFlag();

2011-10-11
1) Function: readseq1by1.c::openFile4read()
	a) fp = popen (cmd, "r");
           free (cmd);
           return fp;
	   -->
           do {
                   if ((fp = popen (cmd, "r")) != NULL)
                   {
                           free (cmd);
                           return fp;
                   }
           } while (1);

2011-09-28
1) Function: readseqfq()
	a) int i, n, strL, m, p;
	   -->
	   int i, n, strL, m, p=0;

	b) else if (buf[m] == '\n' && buf[p] == '\n')
	   -->
	   else if (buf[m] == '\n' && buf[p] == '\n' && m>p)
	   Note: Both modifications are provided by Pan Qi.

2011-09-27
1) Function: Links2Scaf() 
	a) cvg4SNP = ((double)(0.5*cvgAvg) + sqrt(2*cvgAvg));
	   -->
	   cvg4SNP = ((double)(0.5*cvgAvg) + 3.0*sqrt(cvgAvg/2.0));
	   Note: This change maybe not suit for all cases

2) Function: general_linearization()
	a) SNPCtgCounter += removeSNPCtg();
	   -->
	   if (Insert_size < 10000)
	   {
		SNPCtgCounter += removeSNPCtg();
	   }

2011-09-21
1) Function: arrangeNodes_general()
	a) if(temp_cnt)
	   {
		//recover connections
	   }
	   -->
	   if(temp_cnt && (bySmall || temp_cnt->bySmall || temp_cnt->smallIns || (!dh_cnt || !dh_cnt->bySmall || !dh_cnt->smallIns)))
	   {
		//recover connections
	   }

	b) else if (dh_cnt && ((dh_cnt->bySmall || dh_cnt->smallIns || bySmall)
				|| ((-gap > (int)contig_array[node1].length || -gap > (int)contig_array[node2].length)
					&& tmp_dis > 0 && tmp_dis < 0.2*Insert_size )))
	   {
		//exchange connections
	   }
	   -->
	   else if (dh_cnt && ((dh_cnt->bySmall || dh_cnt->smallIns || bySmall)
				|| ((-gap > (int)contig_array[node1].length || -gap > (int)contig_array[node2].length)
					&& tmp_dis > 0 && tmp_dis < 500 )))
	   {
		//exchange connections
	   }

2011-09-15
1) Function: createAnalogousCnt()
	a) if(gap<GapLowerBound)
	   {
		gapCounter++;
		return;
	   }
	   -->
	   if(gap<GapLowerBound)
	   {
		gapCounter++;
		originCnt->deleted = 1;
		temp_cnt = getCntBetween(balSourceStop,balSourceStart);
		return change_flag;
	   }

2011-09-07
1) Function: downSlide()
	a) annotate the codes related to "lastTop".

2) Function: scaffolding()
	a) replace the old codes related to the calculation of the score of contig's connection with the new codes from version 0729/63mer updated by Binghang Liu

	b) if(now_ctg_score == 0 && ((j == num3-1 && contig_array[nextid].length < 200 && num3 + num5 > 5)|| (now_cnt_weigth == 0&& j>0 )))
	   -->
	   if(score_mask == 1 && now_ctg_score == 0 
                && ((j == num3-1 && contig_array[nextid].length < 200 && num3 + num5 > 5)|| (now_cnt_weigth == 0&& j>0 )))

	c) if(now_ctg_score == 0 && ((j == num5-1 && contig_array[*(unsigned int *)darrayGet(scaf5,j-1)].length <200 && num3 + num5 > 5)|| (j != num5-1 && now_cnt_weigth == 0)))
	   -->
	   if(score_mask == 1 && now_ctg_score == 0   
                && ((j == num5-1 && contig_array[*(unsigned int *)darrayGet(scaf5,j-1)].length <200 && num3 + num5 > 5)
                        || (j != num5-1 && now_cnt_weigth == 0)))

3) Function: get_ctg_score2()
	a) if(dh_cnt && dh_cnt->weight >0)
                i++;
	   -->
	   if(dh_cnt && dh_cnt->weight >0)
                in++;

4) Function: arrangeNodes_general()
	a) else if(dh_cnt && ((dh_cnt->weight > weakPE && (-gap > (int)contig_array[node1].length || -gap > (int)contig_array[node2].length))|| (dh_cnt->bySmall || dh_cnt->smallIns || bySmall)))
	   -->
	   else if (dh_cnt && ((dh_cnt->bySmall || dh_cnt->smallIns || bySmall)
                                || ((-gap > (int)contig_array[node1].length || -gap > (int)contig_array[node2].length)
                                        && tmp_dis > 0 && tmp_dis < 0.2*Insert_size )))

5) Function: removeSNPCtg()
	a) add:
	   CONNECT *cnt, *bal_cnt;
	   cnt = getCntBetween(node1, node2);
	   dh_cnt = getCntBetween(node2, node1);
	   change:
	   if (gap >= 0 || contig_array[node1].cvg > cvg4SNP || contig_array[node2].cvg > cvg4SNP)
	   -->
	   if (gap >= 0 || contig_array[node1].cvg > cvg4SNP || contig_array[node2].cvg > cvg4SNP || cnt || bal_cnt)

6) Function: outputScafSeq()
	a) if (prevCtg && actg->scaftig_start)
                {
                        if(!fillGap)
                        {
                                gapN = actg->start - prevCtg->start - contig_array[prevCtg->ctgID].length - overlaplen;
                        }
                        else     gapN = actg->start - prevCtg->start - contig_array[prevCtg->ctgID].length;

//                      gapN = actg->start - prevCtg->start - contig_array[prevCtg->ctgID].length;      //lzy 0907
                        gapN = gapN > 0 ? gapN : 1;
                        outputNs (fo, scaffBuffer, gapN, &column);
                        ctg_start_pos += gapN;
                        //outputGapInfo(prevCtg->ctgID,ctg);
                        Ncounter++;
                }
	   if(fillGap)
           {
                        if (!prevCtg)
                        {
                                start = 0;
                        }
                        else
                        {
                                start = actg->cutHead;
                        }
           }
           else start = 0;
	   -->
	   if (prevCtg && actg->scaftig_start)
                {
/*<<lzy 0907
                        if(!fillGap)
                        {
                                gapN = actg->start - prevCtg->start - contig_array[prevCtg->ctgID].length - overlaplen;
                        }
                        else     gapN = actg->start - prevCtg->start - contig_array[prevCtg->ctgID].length;
>>*/
                        gapN = actg->start - prevCtg->start - contig_array[prevCtg->ctgID].length;      //lzy 0907
                        gapN = gapN > 0 ? gapN : 1;
                        outputNs (fo, scaffBuffer, gapN, &column);
                        ctg_start_pos += gapN;
                        //outputGapInfo(prevCtg->ctgID,ctg);
                        Ncounter++;
                }

           if (!prevCtg)
           {
                 start = 0;
           }
           else
           {
                 start = actg->cutHead;
           }


2011-09-06
1) Function: removeSNPCtg()
	a) maskNodeCnt()
	   -->
	   transferCnt2RemainNode()
	   reason: not only mask SNP contig, but also transfer the PE information from masked SNP contig to remain SNP contig

2011-09-02
1) Function: Links2Scaf()
	a) cvg4SNP = 0.6*cvgAvg;
	   -->
	   cvg4SNP = ((double)(0.5*cvgAvg) + sqrt(2*cvgAvg));

2011-08-31
1) recover the function removeSNPCtg()

2) Function: removeSNPCtg()
	a) if (len1 > len2 || (len1 == len2 && contig_array[node1].cvg > contig_array[node2].cvg))
	   -->
	   if (contig_array[node1].cvg > contig_array[node2].cvg || (len1 > len2 && contig_array[node1].cvg == contig_array[node2].cvg))
	   REASON: to keep accordance with the operation of merging bubble that remains the edge with higher coverage

2011-08-29
1) Function: arrangeNodes_general()
	a) if (comCount > 0 && abs(adjustedGap/comCount) > abs(gap))
	   -->
	   if (comCount > 0)

	b) if (bySmall > 0 ||(-gap > (int)contig_array[node1].length || -gap > (int)contig_array[node2].length)
                                        && (i != nodeCounter-1))
	   -->
	   if (bySmall > 0 && gap < 0 
		||((-gap > (int)contig_array[node1].length || -gap > (int)contig_array[node2].length)
	                && (i != nodeCounter-1)))

2) Function: getCntNodes()
	a) if (cnt->weight < 3)
	   -->
	   if (0 == bySmall && cnt->weight < 3 && !cnt->smallIns && !cnt->bySmall)

3) Function: calGapLen()
	a) if (cnt && cnt->weight>=3)
	   -->
	   if (cnt && (cnt->weight>=3 || bySmall || cnt->smallIns || cnt->bySmall))

2011-08-26
1) Function: arrangeNodes_general()
	a) else if(dh_cnt && ((dh_cnt->weight > weakPE && (-gap > (int)contig_array[node1].length || -gap > (int)contig_array[node2].length))|| (dh_cnt->bySmall && gap < -overlaplen ))) 
	   -->
	   else if(dh_cnt && ((dh_cnt->weight > weakPE && (-gap > (int)contig_array[node1].length || -gap > (int)contig_array[node2].length))|| (dh_cnt->bySmall || dh_cnt->smallIns || bySmall)))

	b) if ((-gap > (int)contig_array[node1].length || -gap > (int)contig_array[node2].length) && (i != nodeCounter-1))
	   -->
	   if ((bySmall > 0)
           ||(-gap > (int)contig_array[node1].length || -gap > (int)contig_array[node2].length) && (i != nodeCounter-1))

	c) if (comCount > 0 && adjustedGap/comCount) > gap)
	   -->
	   if (comCount > 0 && abs(adjustedGap/comCount) > abs(gap))

2) Function: Links2Scaf()
	a) bySmall = Insert_size > 1000 ? 2 : 0;
	   -->
	   bySmall = Insert_size > 1000 ? 0 : 1;