void gfiGetSeqName(char *spec, char *name, char *file)
/* Extract sequence name and optionally file name from spec,
* which includes nib and 2bit files. (The file may be NULL
* if you don't care.) */
{
if (nibIsFile(spec))
{
splitPath(spec, NULL, name, NULL);
if (file != NULL)
strcpy(file, spec);
}
else
{
char *s = strchr(spec, ':');
if (s == NULL)
errAbort("Expecting colon in %s", spec);
strcpy(name, s+1);
if (file != NULL)
{
int fileNameSize = s - spec;
memcpy(file, spec, fileNameSize);
file[fileNameSize] = 0;
}
}
}
struct dnaSeq *gfiExpandAndLoadCached(struct gfRange *range,
struct hash *tFileCache, char *tSeqDir, int querySize,
int *retTotalSeqSize, boolean respectFrame, boolean isRc, int expansion)
/* Expand range to cover an additional expansion bases on either side.
* Load up target sequence and return. (Done together because don't
* know target size before loading.) */
{
struct dnaSeq *target = NULL;
char fileName[PATH_LEN+256];
safef(fileName, sizeof(fileName), "%s/%s", tSeqDir, range->tName);
if (nibIsFile(fileName))
{
struct nibInfo *nib = hashFindVal(tFileCache, fileName);
if (nib == NULL)
{
nib = nibInfoNew(fileName);
hashAdd(tFileCache, fileName, nib);
}
if (isRc)
reverseIntRange(&range->tStart, &range->tEnd, nib->size);
gfiExpandRange(range, querySize, nib->size, respectFrame, isRc, expansion);
target = nibLdPart(fileName, nib->f, nib->size,
range->tStart, range->tEnd - range->tStart);
if (isRc)
{
reverseComplement(target->dna, target->size);
reverseIntRange(&range->tStart, &range->tEnd, nib->size);
}
*retTotalSeqSize = nib->size;
}
else
{
struct twoBitFile *tbf = NULL;
char *tSeqName = strchr(fileName, ':');
int tSeqSize = 0;
if (tSeqName == NULL)
errAbort("No colon in .2bit response from gfServer");
*tSeqName++ = 0;
tbf = hashFindVal(tFileCache, fileName);
if (tbf == NULL)
{
tbf = twoBitOpen(fileName);
hashAdd(tFileCache, fileName, tbf);
}
tSeqSize = twoBitSeqSize(tbf, tSeqName);
if (isRc)
reverseIntRange(&range->tStart, &range->tEnd, tSeqSize);
gfiExpandRange(range, querySize, tSeqSize, respectFrame, isRc, expansion);
target = twoBitReadSeqFragLower(tbf, tSeqName, range->tStart, range->tEnd);
if (isRc)
{
reverseComplement(target->dna, target->size);
reverseIntRange(&range->tStart, &range->tEnd, tSeqSize);
}
*retTotalSeqSize = tSeqSize;
}
return target;
}
struct dnaSeq *dnaLoadSingle(char *fileName, int *retStart, int *retEnd, int *retParentSize)
/* Return sequence if it's a nib file or 2bit part, NULL otherwise. */
{
struct dnaSeq *seq = NULL;
unsigned start = 0, end = 0;
int parentSize = 0;
if (nibIsFile(fileName))
{
/* Save offset out of fileName for auto-lifting */
char filePath[PATH_LEN];
char name[PATH_LEN];
nibParseName(0, fileName, filePath, name, &start, &end);
if (end != 0) /* It's just a range. */
{
FILE *f;
int size;
nibOpenVerify(filePath, &f, &size);
parentSize = size;
}
seq = nibLoadAllMasked(NIB_MASK_MIXED, fileName);
if (end == 0)
parentSize = end = seq->size;
freez(&seq->name);
seq->name = cloneString(name);
}
else if (twoBitIsRange(fileName))
{
/* Save offset out of fileName for auto-lifting */
char *rangeSpec = cloneString(fileName);
int start, end;
char *file, *seqName;
twoBitParseRange(rangeSpec, &file, &seqName, &start, &end);
/* Load sequence. */
{
struct twoBitFile *tbf = twoBitOpen(file);
parentSize = twoBitSeqSize(tbf, seqName);
seq = twoBitReadSeqFrag(tbf, seqName, start, end);
twoBitClose(&tbf);
}
if (end == 0)
end = seq->size;
freez(&rangeSpec);
}
if (retStart != NULL)
*retStart = start;
if (retEnd != NULL)
*retEnd = end;
if (retParentSize != NULL)
*retParentSize = parentSize;
return seq;
}
static void gfFileCacheFreeEl(struct hashEl *el)
/* Free up one file cache info. */
{
char *name = el->name;
if (nibIsFile(name))
{
struct nibInfo *nib = el->val;
nibInfoFree(&nib);
}
else
{
struct twoBitFile *tbf = el->val;
twoBitClose(&tbf);
}
el->val = NULL;
}
struct hash *loadGeno(char *genoFile)
/* load genome sequences into a hash. This supports the multi-sequence
* specs of twoBitLoadAll */
{
struct dnaSeq *genos = NULL, *geno;
struct hash *genoHash = hashNew(0);
if (nibIsFile(genoFile))
genos = nibLoadAllMasked(NIB_MASK_MIXED|NIB_BASE_NAME, genoFile);
else if (twoBitIsSpec(genoFile))
genos = twoBitLoadAll(genoFile);
else
genos = faReadDna(genoFile);
while ((geno = slPopHead(&genos)) != NULL)
{
tolowers(geno->dna);
hashAdd(genoHash, geno->name, geno);
}
return genoHash;
}
void searchOneIndex(int fileCount, char *files[], struct genoFind *gf, char *outName,
boolean isProt, struct hash *maskHash, FILE *outFile, boolean showStatus)
/* Search all sequences in all files against single genoFind index. */
{
int i;
char *fileName;
int count = 0;
long long totalSize = 0;
gfOutputHead(gvo, outFile);
for (i=0; i<fileCount; ++i)
{
fileName = files[i];
if (nibIsFile(fileName))
{
struct dnaSeq *seq;
if (isProt)
errAbort("%s: Can't use .nib files with -prot or d=prot option\n", fileName);
seq = nibLoadAllMasked(NIB_MASK_MIXED, fileName);
freez(&seq->name);
seq->name = cloneString(fileName);
searchOneMaskTrim(seq, isProt, gf, outFile,
maskHash, &totalSize, &count);
freeDnaSeq(&seq);
}
else if (twoBitIsSpec(fileName))
{
struct twoBitSpec *tbs = twoBitSpecNew(fileName);
struct twoBitFile *tbf = twoBitOpen(tbs->fileName);
if (isProt)
errAbort("%s is a two bit file, which doesn't work for proteins.",
fileName);
if (tbs->seqs != NULL)
{
struct twoBitSeqSpec *ss = NULL;
for (ss = tbs->seqs; ss != NULL; ss = ss->next)
{
struct dnaSeq *seq = twoBitReadSeqFrag(tbf, ss->name,
ss->start, ss->end);
searchOneMaskTrim(seq, isProt, gf, outFile,
maskHash, &totalSize, &count);
dnaSeqFree(&seq);
}
}
else
{
struct twoBitIndex *index = NULL;
for (index = tbf->indexList; index != NULL; index = index->next)
{
struct dnaSeq *seq = twoBitReadSeqFrag(tbf, index->name, 0, 0);
searchOneMaskTrim(seq, isProt, gf, outFile,
maskHash, &totalSize, &count);
dnaSeqFree(&seq);
}
}
twoBitClose(&tbf);
}
else
{
static struct dnaSeq seq;
struct lineFile *lf = lineFileOpen(fileName, TRUE);
while (faMixedSpeedReadNext(lf, &seq.dna, &seq.size, &seq.name))
{
searchOneMaskTrim(&seq, isProt, gf, outFile,
maskHash, &totalSize, &count);
}
lineFileClose(&lf);
}
}
carefulClose(&outFile);
if (showStatus)
printf("Searched %lld bases in %d sequences\n", totalSize, count);
}
