lib/nib.c File Reference

#include "common.h"
#include "hash.h"
#include "dnautil.h"
#include "dnaseq.h"
#include "nib.h"
#include "sig.h"

Include dependency graph for nib.c:

Go to the source code of this file.

Data Structures
struct	nibStream
Functions
static char *	findNibSubrange (char *fileName)
static void	parseSubrange (char *subrange, char *name, unsigned *start, unsigned *end)
void	nibParseName (unsigned options, char *fileSpec, char *filePath, char *name, unsigned *start, unsigned *end)
void	nibOpenVerify (char *fileName, FILE **retFile, int *retSize)
static struct dnaSeq *	nibInput (int options, char *fileName, char *seqName, FILE *f, int seqSize, int start, int size)
static void	nibOutput (int options, struct dnaSeq *seq, char *fileName)
dnaSeq *	nibLdPartMasked (int options, char *fileName, FILE *f, int seqSize, int start, int size)
dnaSeq *	nibLdPart (char *fileName, FILE *f, int seqSize, int start, int size)
dnaSeq *	nibLoadPartMasked (int options, char *fileName, int start, int size)
dnaSeq *	nibLoadPart (char *fileName, int start, int size)
dnaSeq *	nibLoadAllMasked (int options, char *fileName)
dnaSeq *	nibLoadAll (char *fileName)
void	nibWriteMasked (int options, struct dnaSeq *seq, char *fileName)
void	nibWrite (struct dnaSeq *seq, char *fileName)
nibStream *	nibStreamOpen (char *fileName)
void	nibStreamClose (struct nibStream **pNs)
void	nibStreamOne (struct nibStream *ns, DNA base)
void	nibStreamMany (struct nibStream *ns, DNA *dna, int size)
boolean	nibIsFile (char *fileName)
boolean	nibIsRange (char *fileName)
nibInfo *	nibInfoNew (char *path)
void	nibInfoFree (struct nibInfo **pNib)
nibInfo *	nibInfoFromCache (struct hash *hash, char *nibDir, char *nibName)
int	nibGetSize (char *nibFile)
Variables
static char const	rcsid [] = "$Id: nib.c,v 1.24 2007/03/13 20:43:05 kent Exp $"

---

Function Documentation

static char* findNibSubrange

(

char *

fileName

)

[static]

Definition at line 15 of file nib.c.

Referenced by nibIsFile(), nibIsRange(), and nibParseName().

{
char *baseName = strrchr(fileName, '/');
baseName = (baseName == NULL) ? fileName : baseName+1;
return strchr(baseName, ':');
}

Here is the caller graph for this function:

int nibGetSize

(

char *

nibFile

)

Definition at line 507 of file nib.c.

References carefulClose(), nibOpenVerify(), and nibInfo::size.

{
FILE* fh;
int size;

nibOpenVerify(nibFile, &fh, &size);
carefulClose(&fh);
return size;
}

Here is the call graph for this function:

void nibInfoFree

(

struct nibInfo **

pNib

)

Definition at line 480 of file nib.c.

References carefulClose(), nibInfo::f, nibInfo::fileName, freeMem(), and freez().

Referenced by gfFileCacheFreeEl(), and nibTwoCacheFree().

{
struct nibInfo *nib = *pNib;
if (nib != NULL)
    {
    carefulClose(&nib->f);
    freeMem(nib->fileName);
    freez(pNib);
    }
}

Here is the call graph for this function:

Here is the caller graph for this function:

struct nibInfo* nibInfoFromCache	(	struct hash *	hash,
		char *	nibDir,
		char *	nibName
	)			[read]

Definition at line 492 of file nib.c.

References hashAdd(), hashFindVal(), nibInfoNew(), PATH_LEN, and safef().

Referenced by nibTwoCacheSeq(), and nibTwoCacheSeqPartExt().

{
struct nibInfo *nib;
char path[PATH_LEN];
safef(path, sizeof(path), "%s/%s.nib", nibDir, nibName);
nib = hashFindVal(hash, path);
if (nib == NULL)
    {
    nib = nibInfoNew(path);
    hashAdd(hash, path, nib);
    }
return nib;
}

Here is the call graph for this function:

Here is the caller graph for this function:

struct nibInfo* nibInfoNew

(

char *

path

)

[read]

Definition at line 470 of file nib.c.

References AllocVar, cloneString(), nibInfo::f, nibInfo::fileName, nibOpenVerify(), and nibInfo::size.

Referenced by gfiExpandAndLoadCached(), and nibInfoFromCache().

{
struct nibInfo *nib;
AllocVar(nib);
nib->fileName = cloneString(path);
nibOpenVerify(path, &nib->f, &nib->size);
return nib;
}

Here is the call graph for this function:

Here is the caller graph for this function:

static struct dnaSeq* nibInput	(	int	options,
		char *	fileName,
		char *	seqName,
		FILE *	f,
		int	seqSize,
		int	start,
		int	size
	)			[static, read]

Definition at line 157 of file nib.c.

References AllocVar, bitAlloc(), bits32, bitSetOne(), cloneString(), dnaSeq::dna, errAbort(), dnaSeq::mask, mask, MASKED_BASE_BIT, dnaSeq::name, needLargeMem(), NIB_MASK_MIXED, SEEK_SET, dnaSeq::size, valToNt, and valToNtMasked.

Referenced by nibLdPartMasked(), and nibLoadAllMasked().

{
int end;
DNA *d;
int bVal;
DNA *valToNtTbl = ((options &  NIB_MASK_MIXED) ? valToNtMasked : valToNt);
struct dnaSeq *seq;
Bits* mask = NULL;
int bytePos, byteSize;
int maskIdx = 0;

assert(start >= 0);
assert(size >= 0);

end = start+size;
if (end > seqSize)
    errAbort("nib read past end of file (%d %d) in file: %s", 
             end, seqSize, (fileName != NULL ? fileName : "(NULL)"));

AllocVar(seq);
seq->size = size;
seq->name = cloneString(seqName);
seq->dna = d = needLargeMem(size+1);
if (options & NIB_MASK_MIXED)
    seq->mask = mask = bitAlloc(size);

bytePos = (start>>1);
fseek(f, bytePos + 2*sizeof(bits32), SEEK_SET);
if (start & 1)
    {
    bVal = getc_unlocked(f);
    if (bVal < 0)
        {
        errAbort("Read error 1 in %s", fileName);
        }
    *d++ = valToNtTbl[(bVal&0xf)];
    size -= 1;
    if (mask != NULL)
        {
        if ((bVal&0xf&MASKED_BASE_BIT) == 0)
            bitSetOne(mask, maskIdx);
        maskIdx++;
        }
    }
byteSize = (size>>1);
while (--byteSize >= 0)
    {
    bVal = getc_unlocked(f);
    if (bVal < 0)
        errAbort("Read error 2 in %s", fileName);
    d[0] = valToNtTbl[(bVal>>4)];
    d[1] = valToNtTbl[(bVal&0xf)];
    d += 2;
    if (mask != NULL)
        {
        if (((bVal>>4)&0xf) == 0)
            bitSetOne(mask, maskIdx);
        if ((bVal&0xf) == 0)
            bitSetOne(mask, maskIdx+1);
        maskIdx += 2;
        }
    }
if (size&1)
    {
    bVal = getc_unlocked(f);
    if (bVal < 0)
        errAbort("Read error 3 in %s", fileName);
    *d++ = valToNtTbl[(bVal>>4)];
    if (mask != NULL)
        {
        if ((bVal>>4) == 0)
            bitSetOne(mask, maskIdx);
        maskIdx++;
        }
    }
*d = 0;
return seq;
}

Here is the call graph for this function:

Here is the caller graph for this function:

boolean nibIsFile

(

char *

fileName

)

Definition at line 444 of file nib.c.

References endsWith(), and findNibSubrange().

Referenced by dnaLoadSingle(), gfClientFileArray(), gfClientSeqList(), gfFileCacheFreeEl(), gfiExpandAndLoadCached(), gfiGetSeqName(), gfIndexNibsAndTwoBits(), gfIndexTransNibsAndTwoBits(), gfMakeOoc(), readSeqFrag(), and searchOneIndex().

{
boolean isANib;
char *subrange = findNibSubrange(fileName);
if (subrange != NULL)
    *subrange = '\0';
isANib = endsWith(fileName, ".nib") || endsWith(fileName, ".NIB");
if (subrange != NULL)
    *subrange = ':';
return isANib;
}

Here is the call graph for this function:

Here is the caller graph for this function:

boolean nibIsRange

(

char *

fileName

)

Definition at line 457 of file nib.c.

References endsWith(), FALSE, and findNibSubrange().

{
boolean isANib;
char *subrange = findNibSubrange(fileName);;
if (subrange == NULL)
    return FALSE;
*subrange = '\0';
isANib = endsWith(fileName, ".nib") || endsWith(fileName, ".NIB");
*subrange = ':';
return isANib;
}

Here is the call graph for this function:

struct dnaSeq* nibLdPart	(	char *	fileName,
		FILE *	f,
		int	seqSize,
		int	start,
		int	size
	)			[read]

Definition at line 301 of file nib.c.

References nibLdPartMasked().

Referenced by gfAddTilesInNib(), gfCountTilesInNib(), gfiExpandAndLoadCached(), and nibTwoCacheSeq().

{
return nibLdPartMasked(0, fileName, f, seqSize, start, size);
}

Here is the call graph for this function:

Here is the caller graph for this function:

struct dnaSeq* nibLdPartMasked	(	int	options,
		char *	fileName,
		FILE *	f,
		int	seqSize,
		int	start,
		int	size
	)			[read]

Definition at line 292 of file nib.c.

References nibInput(), and safef().

Referenced by nibLdPart(), nibLoadPartMasked(), and nibTwoCacheSeqPartExt().

{
char nameBuf[512];
safef(nameBuf, sizeof(nameBuf), "%s:%d-%d", fileName, start, start+size);
return nibInput(options, fileName, nameBuf, f, seqSize, start, size);
}

Here is the call graph for this function:

Here is the caller graph for this function:

struct dnaSeq* nibLoadAll

(

char *

fileName

)

[read]

Definition at line 352 of file nib.c.

References nibLoadAllMasked().

Referenced by gfMakeOoc(), and readMaskedNib().

{
return nibLoadAllMasked(0, fileName);
}

Here is the call graph for this function:

Here is the caller graph for this function:

struct dnaSeq* nibLoadAllMasked	(	int	options,
		char *	fileName
	)			[read]

Definition at line 325 of file nib.c.

References name, nibInput(), nibOpenVerify(), nibParseName(), and PATH_LEN.

Referenced by dnaLoadSingle(), gfClientSeqList(), nibLoadAll(), nibTwoLoadOne(), readMaskedNib(), and searchOneIndex().

                   :
 *    /path/file.nib:seqid:start-end
 * or\n"
 *    /path/file.nib:start-end
 * With the first form, seqid becomes the id of the subrange, with the second
 * form, a sequence id of file:start-end will be used.
 */
{
struct dnaSeq *seq;
FILE *f;
int seqSize;
char filePath[PATH_LEN];
char name[PATH_LEN];
unsigned start, end;

nibParseName(options, fileName, filePath, name, &start, &end);
nibOpenVerify(filePath, &f, &seqSize);
if (end == 0)
    end = seqSize;
seq = nibInput(options, fileName, name, f, seqSize, start, end-start);
fclose(f);
return seq;
}

Here is the call graph for this function:

Here is the caller graph for this function:

struct dnaSeq* nibLoadPart	(	char *	fileName,
		int	start,
		int	size
	)			[read]

Definition at line 319 of file nib.c.

References nibLoadPartMasked().

Referenced by readSeqFrag().

{
return nibLoadPartMasked(0, fileName, start, size);
}

Here is the call graph for this function:

Here is the caller graph for this function:

struct dnaSeq* nibLoadPartMasked	(	int	options,
		char *	fileName,
		int	start,
		int	size
	)			[read]

Definition at line 307 of file nib.c.

References nibLdPartMasked(), and nibOpenVerify().

Referenced by nibLoadPart().

{
struct dnaSeq *seq;
FILE *f;
int seqSize;
nibOpenVerify(fileName, &f, &seqSize);
seq = nibLdPartMasked(options, fileName, f, seqSize, start, size);
fclose(f);
return seq;
}

Here is the call graph for this function:

Here is the caller graph for this function:

void nibOpenVerify	(	char *	fileName,
		FILE **	retFile,
		int *	retSize
	)

Definition at line 84 of file nib.c.

References bits32, byteSwap32(), dnaUtilOpen(), errAbort(), errno, mustReadOne, nibSig, and safef().

Referenced by dnaLoadSingle(), gfAddTilesInNib(), gfCountTilesInNib(), nibGetSize(), nibInfoNew(), nibLoadAllMasked(), and nibLoadPartMasked().

{
bits32 size;
bits32 sig;
FILE *f = fopen(fileName, "rb");
char buffer[512];
char buffer2[512];
char buffer3[512];

if (f == NULL)
    {
    /* see if nib is down a few directories ala faSplit -outDirDepth */
    char *ptr = NULL;
    char *dir, *file;
    struct stat statBuf;

    /* divide fileName into file and directory components */
    safef(buffer, sizeof(buffer), "%s", fileName);
    if ((ptr = strrchr(buffer, '/')) != NULL)
        {
        *ptr++ = 0;
        dir = buffer;
        file = ptr;
        }
    else
        {
        dir = "";
        file = buffer;
        }
    
    buffer3[0] = 0;
    /* start at the end of the fileName (minus .nib) */
    for(ptr = &file[strlen(file) - 5]; ; )
        {
        strcpy(buffer2, buffer3);
        if (isdigit(*ptr))
            {
            /* if we have a digit in the fileName, see if there is a directory with this name */
            safef(buffer3, sizeof(buffer3), "%c/%s",*ptr,buffer2);
            ptr--;
            }
        else
            /* we've run out of digits in the fileName, just add 0's */
            safef(buffer3, sizeof(buffer3), "0/%s",buffer2);

        /* check to see if this directory exists */
        safef(buffer2, sizeof(buffer2), "%s/%s", dir, buffer3);
        if (stat(buffer2, &statBuf) < 0)
            break;

        /* directory exists, see if our file is down there */
        safef(buffer2, sizeof(buffer2), "%s/%s/%s", dir, buffer3, file);
        if  ((f = fopen(buffer2, "rb")) != NULL)
            break;
        }
    if (f == NULL)
        errAbort("Can't open %s to read: %s", fileName,  strerror(errno));
    }
dnaUtilOpen();
mustReadOne(f, sig);
mustReadOne(f, size);
if (sig != nibSig)
    {
    sig = byteSwap32(sig);
    size = byteSwap32(size);
    if (sig != nibSig)
        errAbort("%s is not a good .nib file.",  fileName);
    }
*retSize = size;
*retFile = f;
}

Here is the call graph for this function:

Here is the caller graph for this function:

static void nibOutput	(	int	options,
		struct dnaSeq *	seq,
		char *	fileName
	)			[static]

Definition at line 238 of file nib.c.

References bitReadOne(), bits32, carefulClose(), dnaSeq::dna, errAbort(), dnaSeq::mask, mask, MASKED_BASE_BIT, mustOpen(), NIB_MASK_MAP, NIB_MASK_MIXED, nibSig, ntVal5, ntValMasked, dnaSeq::size, UBYTE, and writeOne.

Referenced by nibWriteMasked().

{
UBYTE byte;
DNA *dna = seq->dna;
int dVal1, dVal2;
bits32 size = seq->size;
int byteCount = (size>>1);
bits32 sig = nibSig;
int *ntValTbl = ((options & NIB_MASK_MIXED) ? ntValMasked : ntVal5);
Bits* mask = ((options & NIB_MASK_MAP) ? seq->mask : NULL);
int maskIdx = 0;
FILE *f = mustOpen(fileName, "w");

assert(sizeof(bits32) == 4);

writeOne(f, sig);
writeOne(f, seq->size);

printf("Writing %d bases in %d bytes\n", seq->size, ((seq->size+1)/2) + 8);
while (--byteCount >= 0)
    {
    dVal1 = ntValTbl[(int)dna[0]];
    dVal2 = ntValTbl[(int)dna[1]];
    /* Set from mask, remember bit in character is opposite sense of bit
     * in mask. */
    if (mask != NULL)
        {
        if (!bitReadOne(mask, maskIdx))
            dVal1 |= MASKED_BASE_BIT;
        if (!bitReadOne(mask, maskIdx+1))
            dVal2 |= MASKED_BASE_BIT;
        maskIdx += 2;
        }
    byte = (dVal1<<4) | dVal2;
    if (putc(byte, f) < 0)
        {
        perror("");
        errAbort("Couldn't write all of %s", fileName);
        }
    dna += 2;
    }
if (size & 1)
    {
    dVal1 = ntValTbl[(int)dna[0]];
    if ((mask != NULL) && !bitReadOne(mask, maskIdx))
        dVal1 |= MASKED_BASE_BIT;
    byte = (dVal1<<4);
    putc(byte, f);
    }
carefulClose(&f);
}

Here is the call graph for this function:

Here is the caller graph for this function:

void nibParseName	(	unsigned	options,
		char *	fileSpec,
		char *	filePath,
		char *	name,
		unsigned *	start,
		unsigned *	end
	)

Definition at line 48 of file nib.c.

References findNibSubrange(), NIB_BASE_NAME, parseSubrange(), and splitPath().

Referenced by dnaLoadSingle(), and nibLoadAllMasked().

{
char *subrange = findNibSubrange(fileSpec);
if (subrange != NULL)
    {
    *subrange = '\0';
    parseSubrange(subrange, name, start, end);
    strcpy(filePath, fileSpec);
    *subrange = ':';
    if (strlen(name) == 0)
        {
        /* no name in spec, generate one */
        if (options & NIB_BASE_NAME)
            splitPath(filePath, NULL, name, NULL);
        else
            strcpy(name, filePath);
        sprintf(name+strlen(name), ":%u-%u", *start, *end);
        }
    }
else
    {
    *start = 0;
    *end = 0;
    strcpy(filePath, fileSpec);
    if (options & NIB_BASE_NAME)
        splitPath(fileSpec, NULL, name, NULL);
    else
        strcpy(name, fileSpec);
    }
}

Here is the call graph for this function:

Here is the caller graph for this function:

void nibStreamClose

(

struct nibStream **

pNs

)

Definition at line 401 of file nib.c.

References bits32, nibStream::byte, nibStream::f, nibStream::fileName, freeMem(), freez(), nibSig, SEEK_SET, nibStream::size, and writeOne.

{
struct nibStream *ns = *pNs;
FILE *f;
bits32 sig = nibSig;
if (ns == NULL)
    return;
f = ns->f;
if (ns->size&1)
    writeOne(f, ns->byte);
fseek(f,  0L, SEEK_SET);
writeOne(f, sig);
writeOne(f, ns->size);
fclose(f);
freeMem(ns->fileName);
freez(pNs);
}

Here is the call graph for this function:

void nibStreamMany	(	struct nibStream *	ns,
		DNA *	dna,
		int	size
	)

Definition at line 436 of file nib.c.

References nibStreamOne().

{
int i;
for (i=0; i<size; ++i)
    nibStreamOne(ns, *dna++);
}

Here is the call graph for this function:

void nibStreamOne	(	struct nibStream *	ns,
		DNA	base
	)

Definition at line 420 of file nib.c.

References nibStream::byte, nibStream::f, ntVal5, nibStream::size, UBYTE, and writeOne.

Referenced by nibStreamMany().

{
UBYTE ub = ntVal5[(int)base];

if ((++ns->size&1) == 0)
    {
    ub += ns->byte;
    writeOne(ns->f, ub);
    }
else
    {
    ns->byte = (ub<<4);
    }
}

Here is the caller graph for this function:

struct nibStream* nibStreamOpen

(

char *

fileName

)

[read]

Definition at line 383 of file nib.c.

References AllocVar, cloneString(), dnaUtilOpen(), nibStream::f, nibStream::fileName, mustOpen(), nibStream::size, and writeOne.

{
struct nibStream *ns;
FILE *f;

dnaUtilOpen();
AllocVar(ns);
ns->f = f = mustOpen(fileName, "wb");
ns->fileName = cloneString(fileName);

/* Write header - initially zero.  Will fix it up when we close. */
writeOne(f, ns->size);
writeOne(f, ns->size);

return ns;
}

Here is the call graph for this function:

void nibWrite	(	struct dnaSeq *	seq,
		char *	fileName
	)

Definition at line 365 of file nib.c.

References nibWriteMasked().

{
    nibWriteMasked(0, seq, fileName);
}

Here is the call graph for this function:

void nibWriteMasked	(	int	options,
		struct dnaSeq *	seq,
		char *	fileName
	)

Definition at line 358 of file nib.c.

References nibOutput().

Referenced by nibWrite().

{
    nibOutput(options, seq, fileName);
}

Here is the call graph for this function:

Here is the caller graph for this function:

static void parseSubrange	(	char *	subrange,
		char *	name,
		unsigned *	start,
		unsigned *	end
	)			[static]

Definition at line 24 of file nib.c.

References errAbort().

Referenced by nibParseName().

{
char *rangePart = strchr(subrange+1, ':');
if (rangePart != NULL)
    {
    /* :seqId:start-end form */
    *rangePart = '\0';
    strcpy(name, subrange+1);
    *rangePart = ':';
    rangePart++;
    }
else
    {
    /* :start-end form */
    rangePart = subrange+1;
    strcpy(name, ""); 
    }
if ((sscanf(rangePart, "%u-%u", start, end) != 2) || (*start > *end))
    errAbort("can't parse nib file subsequence specification: %s",
             subrange);
}

Here is the call graph for this function:

Here is the caller graph for this function:

---

Variable Documentation

char const rcsid[] = "$Id: nib.c,v 1.24 2007/03/13 20:43:05 kent Exp $" [static]

Definition at line 13 of file nib.c.

---