static int calcDigitCount(struct axt *axt, int tSize, int qSize)
/* Figure out how many digits needed for blast position display. */
{
int tDig, qDig;
if (axt->qStrand == '-')
qDig = digitsBaseTen(qSize - axt->qStart + 1);
else
qDig = digitsBaseTen(axt->qEnd);
if (axt->tStrand == '-')
tDig = digitsBaseTen(tSize - axt->tStart + 1);
else
tDig = digitsBaseTen(axt->tEnd);
return Blatmax(tDig, qDig);
}
void segWrite(FILE *f, struct segBlock *block)
/* Write next segment block to the file. */
{
struct segComp *comp;
int srcChars = 0, startChars = 0, sizeChars = 0, srcSizeChars = 0;
/* Write segment block header. */
fprintf(f, "b");
if (block->name != NULL)
fprintf(f, " name=%s", block->name);
if (block->val != 0)
fprintf(f, " val=%d", block->val);
fprintf(f, "\n");
/* Figure out length of each field. */
for (comp = block->components; comp != NULL; comp = comp->next)
{
int len = 0;
/* A name line '.' will break some tools, so replace it with a
* generic name */
if (sameString(comp->src, "."))
{
freeMem(&comp->src);
comp->src = cloneString("defaultName");
}
len = strlen(comp->src);
if (srcChars < len)
srcChars = len;
len = digitsBaseTen(comp->start);
if (startChars < len)
startChars = len;
len = digitsBaseTen(comp->size);
if (sizeChars < len)
sizeChars = len;
len = digitsBaseTen(comp->srcSize);
if (srcSizeChars < len)
srcSizeChars = len;
}
/* Write out each component. */
for (comp = block->components; comp != NULL; comp = comp->next)
{
fprintf(f, "s %-*s %*d %*d %c %*d\n",
srcChars, comp->src, startChars, comp->start, sizeChars, comp->size,
comp->strand, srcSizeChars, comp->srcSize);
}
/* Write out blank separator line. */
fprintf(f, "\n");
}
void splitByRecord(char *inName, int splitCount, char *outRoot, off_t estSize)
/* Split into a file base by base. */
{
struct dnaSeq seq;
struct lineFile *lf = lineFileOpen(inName, TRUE);
int digits = digitsBaseTen(splitCount);
off_t nextEnd = 0;
off_t curPos = 0;
int fileCount = 0;
FILE *f = NULL;
char outPath[PATH_LEN];
ZeroVar(&seq);
while (faMixedSpeedReadNext(lf, &seq.dna, &seq.size, &seq.name))
{
curPos += seq.size;
if (curPos > nextEnd)
{
carefulClose(&f);
mkOutPath(outPath, outRoot, digits, fileCount++);
verbose(2, "writing %s\n", outPath);
f = mustOpen(outPath, "w");
nextEnd = calcNextEnd(fileCount, splitCount, estSize);
}
faWriteNext(f, seq.name, seq.dna, seq.size);
}
carefulClose(&f);
lineFileClose(&lf);
}
static int maxDigits(int x, int y)
{
int xDigits = digitsBaseTen(x);
int yDigits = digitsBaseTen(y);
return (xDigits > yDigits ? xDigits : yDigits);
}
void splitByGap(char *inName, int pieceSize, char *outRoot, long long estSize)
/* Split up file into pieces at most pieceSize bases long, at gap boundaries
* if possible. */
{
off_t pieces = (estSize + pieceSize-1)/pieceSize;
int digits = digitsBaseTen(pieces);
int minGapSize = optionInt("minGapSize", 1000);
boolean noGapDrops = optionExists("noGapDrops");
int maxN = optionInt("maxN", pieceSize-1);
boolean oneFile = optionExists("oneFile");
char fileName[512];
char dirOnly[256], noPath[128];
int pos, pieceIx = 0, writeCount = 0;
struct dnaSeq seq;
struct lineFile *lf = lineFileOpen(inName, TRUE);
FILE *f = NULL;
Bits *bits = NULL;
int seqCount = 0;
char *outFile = optionVal("out", NULL);
char *liftFile = optionVal("lift", NULL);
FILE *lift = NULL;
ZeroVar(&seq);
if (minGapSize < 1)
errAbort("ERROR: minGapSize must be > 0");
splitPath(outRoot, dirOnly, noPath, NULL);
if (oneFile)
{
sprintf(fileName, "%s.fa", outRoot);
f = mustOpen(fileName, "w");
}
else
fileName[0] = '\0';
if (liftFile)
lift = mustOpen(liftFile, "w");
while (faMixedSpeedReadNext(lf, &seq.dna, &seq.size, &seq.name))
{
bits = bitAlloc(seq.size);
setBitsN(seq.dna, seq.size, bits);
++seqCount;
if (outFile != NULL)
{
if (seqCount > 1)
errAbort("Can only handle in files with one sequence using out option");
bitsForOut(outFile, seq.size, bits);
}
pos = 0;
while (pos < seq.size)
{
boolean gotGap = FALSE;
int gapStart = 0;
int gapSize = 0;
int endSize = seq.size - pos;
int thisSize = min(endSize, pieceSize);
int startGapLen = 0;
if (seq.dna[pos] == 'n' || seq.dna[pos] == 'N')
{
startGapLen = bitFindClear(bits, pos, endSize) - pos;
verbose(3,"#\tstarting gap at %d for length: %d\n", pos,
startGapLen );
}
/* if a block is all gap for longer than minGapSize, then
* keep it all together in one large piece
*/
if (startGapLen > minGapSize)
{
if (noGapDrops)
{
writeOneByGap(oneFile, outRoot, digits, &pieceIx,
f, noPath, pos, startGapLen, &seq, lift,
&writeCount, fileName);
}
else
verbose(3,"#\tbeginning gap of %d size skipped\n", startGapLen);
thisSize = startGapLen;
}
else if (thisSize > 0 && bitCountRange(bits, pos, thisSize) <= maxN)
{
if (endSize>pieceSize) /* otherwise chops tiny piece at very end */
{
gotGap = findLastGap(&(seq.dna[pos]), thisSize, endSize,
minGapSize, &gapStart, &gapSize);
if (gotGap)
thisSize = gapStart;
}
writeOneByGap(oneFile, outRoot, digits, &pieceIx,
f, noPath, pos, thisSize, &seq, lift, &writeCount, fileName);
}
pos += thisSize;
if (gotGap)
{
/* last block is all gap, write it all out */
/*if ((pos + gapSize) >= seq.size)*/
if (noGapDrops)
{
writeOneByGap(oneFile, outRoot, digits, &pieceIx,
f, noPath, pos, gapSize, &seq ,lift, &writeCount, fileName);
verbose(3,
"#\tadding gapSize %d to pos %d -> %d and writing gap\n",
gapSize, pos, pos+gapSize);
}
else
verbose(3,"#\tadding gapSize %d to pos %d -> %d\n",
gapSize, pos, pos+gapSize);
pos += gapSize;
}
}
bitFree(&bits);
}
carefulClose(&f);
carefulClose(&lift);
lineFileClose(&lf);
printf("%d pieces of %d written\n", writeCount, pieceIx);
}
void splitByCount(char *inName, int pieceSize, char *outRoot, off_t estSize, int extra)
/* Split up file into pieces pieceSize long. */
{
off_t pieces = (estSize + pieceSize-1)/pieceSize;
int digits = digitsBaseTen(pieces);
int maxN = optionInt("maxN", pieceSize-1);
boolean oneFile = optionExists("oneFile");
char fileName[PATH_LEN];
char dirOnly[PATH_LEN], noPath[128];
int pos, pieceIx = 0, writeCount = 0;
struct dnaSeq seq;
struct lineFile *lf = lineFileOpen(inName, TRUE);
FILE *f = NULL;
Bits *bits = NULL;
int seqCount = 0;
char *outFile = optionVal("out", NULL);
char *liftFile = optionVal("lift", NULL);
FILE *lift = NULL;
ZeroVar(&seq);
splitPath(outRoot, dirOnly, noPath, NULL);
if (oneFile)
{
sprintf(fileName, "%s.fa", outRoot);
f = mustOpen(fileName, "w");
}
if (liftFile)
lift = mustOpen(liftFile, "w");
/* Count number of N's from s[0] to s[size-1].
* Treat any parts past end of string as N's. */
while (faMixedSpeedReadNext(lf, &seq.dna, &seq.size, &seq.name))
{
bits = bitAlloc(seq.size);
setBitsN(seq.dna, seq.size, bits);
++seqCount;
if (outFile != NULL)
{
if (seqCount > 1)
errAbort("Can only handle in files with one sequence using out option");
bitsForOut(outFile, seq.size, bits);
}
for (pos = 0; pos < seq.size; pos += pieceSize)
{
char numOut[128];
int thisSize = seq.size - pos;
if (thisSize > (pieceSize + extra))
thisSize = pieceSize + extra;
if ((thisSize <= extra) && (pos > 0))
break; /* nobody wants duplicate smaller than extra overhang */
if (bitCountRange(bits, pos, thisSize) <= maxN)
{
if (!oneFile)
{
mkOutPath(fileName, outRoot, digits, pieceIx);
f = mustOpen(fileName, "w");
}
sprintf(numOut, "%s%0*d", noPath, digits, pieceIx);
faWriteNext(f, numOut, seq.dna + pos, thisSize);
if (lift)
fprintf(lift, "%d\t%s\t%d\t%s\t%d\n",
pos, numOut, thisSize, seq.name, seq.size);
++writeCount;
if (!oneFile)
carefulClose(&f);
}
pieceIx++;
}
bitFree(&bits);
}
carefulClose(&f);
carefulClose(&lift);
lineFileClose(&lf);
printf("%d pieces of %d written\n", writeCount, pieceIx);
}
void splitByBase(char *inName, int splitCount, char *outRoot, off_t estSize)
/* Split into a file base by base. */
{
struct lineFile *lf = lineFileOpen(inName, TRUE);
int lineSize;
char *line;
char c;
char dir[PATH_LEN], seqName[128], outFile[128], outPathName[PATH_LEN];
int digits = digitsBaseTen(splitCount);
boolean warnedMultipleRecords = FALSE;
int fileCount = 0;
off_t nextEnd = 0;
off_t curPos = 0;
FILE *f = NULL;
int linePos = 0;
int outLineSize = 50;
if (!lineFileNext(lf, &line, &lineSize))
errAbort("%s is empty", inName);
if (line[0] == '>')
{
line = firstWordInLine(line+1);
if (line == NULL)
errAbort("Empty initial '>' line in %s", inName);
strncpy(seqName, line, sizeof(seqName));
}
else
{
splitPath(inName, dir, seqName, NULL);
lineFileReuse(lf);
}
splitPath(outRoot, NULL, outFile, NULL);
while (lineFileNext(lf, &line, &lineSize))
{
if (line[0] == '>')
{
if (!warnedMultipleRecords)
{
warnedMultipleRecords = TRUE;
warn("More than one record in FA file line %d of %s",
lf->lineIx, lf->fileName);
continue;
}
}
while ((c = *line++) != 0)
{
if (isdigit(c) || isspace(c))
continue;
if (!isalpha(c))
errAbort("Weird %c (0x%x) line %d of %s", c, c, lf->lineIx, lf->fileName);
if (++curPos >= nextEnd)
{
if (f != NULL)
{
if (linePos != 0)
fputc('\n', f);
fclose(f);
}
mkOutPath(outPathName, outRoot, digits, fileCount);
verbose(2, "writing %s\n", outPathName);
f = mustOpen(outPathName, "w");
fprintf(f, ">%s%0*d\n", outFile, digits, fileCount);
++fileCount;
linePos = 0;
nextEnd = calcNextEnd(fileCount, splitCount, estSize);
}
fputc(c, f);
if (++linePos >= outLineSize)
{
fputc('\n', f);
linePos = 0;
}
}
}
if (f != NULL)
{
if (linePos != 0)
fputc('\n', f);
fclose(f);
}
lineFileClose(&lf);
}
void mafPrettyOut(FILE *f, struct mafAli *maf, int lineSize,
boolean onlyDiff, int blockNo)
{
int ii, ch;
int srcChars = 0;
struct mafComp *mc;
int lineStart, lineEnd;
char *summaryLine = needMem(lineSize+1);
char *referenceText;
int startChars, sizeChars, srcSizeChars;
boolean haveInserts = FALSE;
struct mafComp *masterMc = maf->components;
startChars = sizeChars = srcSizeChars = 0;
for (mc = maf->components; mc != NULL; mc = mc->next)
{
/* Figure out length of source (species) field. */
/*if (mc->size != 0)*/
{
char dbOnly[128];
int len;
char *org;
memset(dbOnly, 0, sizeof(dbOnly));
safef(dbOnly, sizeof(dbOnly), "%s", mc->src);
chopPrefix(dbOnly);
if ((org = hOrganism(dbOnly)) == NULL)
len = strlen(dbOnly);
else
len = strlen(org);
if (srcChars < len)
srcChars = len;
len = digitsBaseTen(mc->start);
if (startChars < len)
startChars = len;
len = digitsBaseTen(mc->size);
if (sizeChars < len)
sizeChars = len;
len = digitsBaseTen(mc->srcSize);
if (srcSizeChars < len)
srcSizeChars = len;
if (mc->text && (mc->rightStatus == MAF_INSERT_STATUS) && (masterMc->start + masterMc->size < winEnd))
haveInserts = TRUE;
#ifdef REVERSESTRAND
/* complement bases if hgTracks is on reverse strand */
if (mc->size && cartCgiUsualBoolean(cart, COMPLEMENT_BASES_VAR, FALSE))
complement(mc->text, maf->textSize);
#endif
}
}
/* first sequence in the alignment */
referenceText = maf->components->text;
for (lineStart = 0; lineStart < maf->textSize; lineStart = lineEnd)
{
int size;
lineEnd = lineStart + lineSize;
if (lineEnd >= maf->textSize)
lineEnd = maf->textSize;
size = lineEnd - lineStart;
initSummaryLine(summaryLine, size, '*');
for (mc = maf->components; mc != NULL; mc = mc->next)
{
char dbOnly[128], *chrom;
int s = mc->start;
int e = s + mc->size;
char *org;
char *revComp = "";
char strand = mc->strand;
struct dyString *dy = newDyString(512);
#ifdef REVERSESTRAND
if (cartCgiUsualBoolean(cart, COMPLEMENT_BASES_VAR, FALSE))
strand = (strand == '+') ? '-' : '+';
#endif
if (strand == '-') revComp = "&hgSeq.revComp=on";
dyStringClear(dy);
safef(dbOnly, sizeof(dbOnly), "%s", mc->src);
chrom = chopPrefix(dbOnly);
if ((org = hOrganism(dbOnly)) == NULL)
org = dbOnly;
if (mc->strand == '-')
reverseIntRange(&s, &e, mc->srcSize);
if (mc->text != NULL)
{
if (lineStart == 0)
{
if (hDbIsActive(dbOnly))
{
dyStringPrintf(dy, "%s Browser %s:%d-%d %c %*dbps",hOrganism(dbOnly),chrom, s+1, e, mc->strand,sizeChars, mc->size);
linkToOtherBrowserTitle(dbOnly, chrom, s, e, dy->string);
dyStringClear(dy);
fprintf(f, "B</A> ");
}
else
fprintf(f, " ");
if (hDbExists(dbOnly))
{
dyStringPrintf(dy, "Get %s DNA %s:%d-%d %c %*dbps",hOrganism(dbOnly),chrom, s+1, e, mc->strand,sizeChars, mc->size);
printf("<A TITLE=\"%s\" TARGET=\"_blank\" HREF=\"%s?o=%d&g=getDna&i=%s&c=%s&l=%d&r=%d&db=%s%s\">D</A> ", dy->string,hgcName(),
s, cgiEncode(chrom),
chrom, s, e, dbOnly, revComp);
}
else
fprintf(f, " ");
}
else
{
fprintf(f, " ");
}
dyStringClear(dy);
dyStringPrintf(dy, "%s:%d-%d %c %*dbps",chrom, s+1, e, mc->strand,sizeChars, mc->size);
fprintf(f, "<A TITLE=\"%s\"> %*s </A> ", dy->string, srcChars, org);
updateSummaryLine(summaryLine, referenceText + lineStart, mc->text + lineStart, size);
blueCapWrite(f, mc->text + lineStart, size,
(onlyDiff && mc != maf->components) ? referenceText + lineStart : NULL);
fprintf(f, "\n");
}
else
{
if (((mc->leftStatus == MAF_CONTIG_STATUS) && (mc->rightStatus == MAF_CONTIG_STATUS) )
|| ((mc->leftStatus == MAF_TANDEM_STATUS) && (mc->rightStatus == MAF_TANDEM_STATUS) )
|| ((mc->leftStatus == MAF_INSERT_STATUS) && (mc->rightStatus == MAF_INSERT_STATUS) )
|| ((mc->leftStatus == MAF_MISSING_STATUS) && (mc->rightStatus == MAF_MISSING_STATUS) ))
{
if (lineStart == 0)
{
int s = mc->start;
int e = s + mc->rightLen;
struct dyString *dy = newDyString(512);
if (mc->strand == '-')
reverseIntRange(&s, &e, mc->srcSize);
if ( hDbIsActive(dbOnly))
{
dyStringPrintf(dy, "%s Browser %s:%d-%d %c %d bps Unaligned",hOrganism(dbOnly),chrom, s+1, e, mc->strand, e-s);
linkToOtherBrowserTitle(dbOnly, chrom, s, e, dy->string);
fprintf(f,"B</A> ");
dyStringClear(dy);
}
else
fprintf(f," ");
if (hDbExists(dbOnly))
{
dyStringPrintf(dy, "Get %s DNA %s:%d-%d %c %d bps Unaligned",hOrganism(dbOnly),chrom, s+1, e, mc->strand, e-s);
printf("<A TITLE=\"%s\" TARGET=\"_blank\" HREF=\"%s?o=%d&g=getDna&i=%s&c=%s&l=%d&r=%d&db=%s%s\">D</A> ", dy->string, hgcName(),
s, cgiEncode(chrom),
chrom, s, e, dbOnly,revComp);
}
else
fprintf(f, " ");
}
else
fprintf(f, " ");
initSummaryLine(summaryLine, size, ' ');
dyStringClear(dy);
dyStringPrintf(dy, "%s:%d-%d %c %*dbps",chrom, s+1, e, mc->strand,sizeChars, mc->size);
fprintf(f, "<A TITLE=\"%s\">%*s</A> ", dy->string, srcChars, org);
ch = '-';
switch(mc->rightStatus)
{
case MAF_INSERT_STATUS:
ch = '=';
break;
case MAF_MISSING_STATUS:
ch = 'N';
break;
case MAF_TANDEM_STATUS:
case MAF_CONTIG_STATUS:
ch = '-';
break;
}
for(ii=lineStart; ii < lineEnd ; ii++)
fputc(ch,f);
fprintf(f,"\n");
}
}
}
#ifdef ADDMATCHLINE
if (lineStart == 0)
fprintf(f, " %-*s %s\n", srcChars, "", summaryLine);
else
fprintf(f, "%-*s %s\n", srcChars, "", summaryLine);
#else
fprintf(f, "\n");
#endif
}
if (haveInserts)
{
fprintf(f, "<B>Inserts between block %d and %d in window</B>\n",blockNo, blockNo+1);
for (mc = maf->components; mc != NULL; mc = mc->next)
{
char dbOnly[128], *chrom;
int s = mc->start + mc->size;
int e = s + mc->rightLen;
char *org;
if (mc->text == NULL)
continue;
if (mc->strand == '-')
reverseIntRange(&s, &e, mc->srcSize);
safef(dbOnly, sizeof(dbOnly), "%s", mc->src);
chrom = chopPrefix(dbOnly);
if ((org = hOrganism(dbOnly)) == NULL)
org = dbOnly;
if (mc->rightStatus == MAF_INSERT_STATUS)
{
char *revComp = "";
if (hDbIsActive(dbOnly))
{
char strand = mc->strand;
#ifdef REVERSESTRAND
if (cartCgiUsualBoolean(cart, COMPLEMENT_BASES_VAR, FALSE))
strand = (strand == '+') ? '-' : '+';
#endif
if (strand == '-') revComp = "&hgSeq.revComp=on";
linkToOtherBrowser(dbOnly, chrom, s, e);
fprintf(f,"B");
fprintf(f, "</A>");
fprintf(f, " ");
}
else
fprintf(f, " ");
if (hDbExists(dbOnly))
{
printf("<A TARGET=\"_blank\" HREF=\"%s?o=%d&g=getDna&i=%s&c=%s&l=%d&r=%d"
"&db=%s%s\">D</A> ", hgcName(), s, cgiEncode(chrom), chrom, s,
e, dbOnly,revComp);
}
else
fprintf(f, " ");
fprintf(f, "%*s %dbp\n", srcChars, org,mc->rightLen);
}
}
fprintf(f, "\n");
}
freeMem(summaryLine);
}
void segSplitPairwise(char *inSeg, char *outSeg)
/* segSplitPairwise - Split a segment file pairwise. */
{
struct segFile *sf = segOpen(inSeg);
FILE *of = mustOpen(outSeg, "w");
struct segBlock *sb;
struct segBlock newSb;
struct segComp *refComp, *sc, *refSc, *newSc;
char *splitName = NULL;
int sNum, bNum = 1;
segWriteStart(of);
while ((sb = segNext(sf)) != NULL)
{
/* Set ref if it wasn't set on the commandline. */
if (ref == NULL && ((ref = segFirstCompSpecies(sb, '.')) == NULL))
errAbort("ref is not set and the first segment block has no "
"components.");
/* Find and clone the reference species component. */
refComp = segFindCompSpecies(sb, ref, '.');
refSc = cloneSegComp(refComp);
sNum = 1;
for (sc = sb->components; sc != NULL; sc = sc->next)
{
if (sc == refComp)
continue;
/* Clone the current component. */
newSc = cloneSegComp(sc);
newSc->next = NULL;
/* Point the reference component to the newly cloned component. */
refSc->next = newSc;
/* Set up the segment block to hold the two components above. */
if (number)
{
if (sb->name == NULL)
{
AllocArray(sb->name, 6 + digitsBaseTen(bNum));
sprintf(sb->name, "Block%d", bNum);
}
AllocArray(splitName, strlen(sb->name) + digitsBaseTen(sNum) + 2);
sprintf(splitName, "%s.%d", sb->name, sNum++);
newSb.name = splitName;
}
else
newSb.name = sb->name;
newSb.val = sb->val;
newSb.components = refSc;
/* Output this new block. */
segWrite(of, &newSb);
if (number)
freeMem(splitName);
/* Free the cloned component. */
segCompFree(&newSc);
}
/* Free the cloned reference component. */
segCompFree(&refSc);
segBlockFree(&sb);
bNum++;
}
segWriteEnd(of);
carefulClose(&of);
segFileFree(&sf);
}
void mafWrite(FILE *f, struct mafAli *ali)
/* Write next alignment to file. */
{
struct mafComp *comp;
int srcChars = 0, startChars = 0, sizeChars = 0, srcSizeChars = 0;
/* Write out alignment header */
fprintf(f, "a score=%f\n", ali->score);
/* Figure out length of each field. */
for (comp = ali->components; comp != NULL; comp = comp->next)
{
int len = 0;
/* a name like '.' will break some tools, so replace it
* with a generic name */
if (sameString(comp->src,"."))
comp->src=cloneString("defaultName");
len = strlen(comp->src);
if (srcChars < len)
srcChars = len;
len = digitsBaseTen(comp->start);
if (startChars < len)
startChars = len;
len = digitsBaseTen(comp->size);
if (sizeChars < len)
sizeChars = len;
len = digitsBaseTen(comp->srcSize);
if (srcSizeChars < len)
srcSizeChars = len;
}
/* Write out each component. */
for (comp = ali->components; comp != NULL; comp = comp->next)
{
if ((comp->size == 0) && (comp->leftStatus))
fprintf(f, "e %-*s %*d %*d %c %*d %c\n",
srcChars, comp->src, startChars, comp->start,
sizeChars, comp->leftLen, comp->strand,
srcSizeChars, comp->srcSize, comp->leftStatus);
else
{
fprintf(f, "s %-*s %*d %*d %c %*d %s\n",
srcChars, comp->src, startChars, comp->start,
sizeChars, comp->size, comp->strand,
srcSizeChars, comp->srcSize, comp->text);
if (comp->quality)
fprintf(f, "q %-*s %s\n",
srcChars + startChars + sizeChars + srcSizeChars + 5,
comp->src, comp->quality);
if (comp->leftStatus)
fprintf(f,"i %-*s %c %d %c %d\n",srcChars,comp->src,
comp->leftStatus,comp->leftLen,comp->rightStatus,comp->rightLen);
}
}
/* Write out blank separator line. */
fprintf(f, "\n");
}
