boolean faFastReadNext(FILE *f, DNA **retDna, int *retSize, char **retName)
/* Read in next FA entry as fast as we can. Return FALSE at EOF.
* The returned DNA and name will be overwritten by the next call
* to this function. */
{
int c;
int bufIx = 0;
static char name[256];
int nameIx = 0;
boolean gotSpace = FALSE;
/* Seek to next '\n' and save first word as name. */
dnaUtilOpen();
name[0] = 0;
for (;;)
{
if ((c = fgetc(f)) == EOF)
{
*retDna = NULL;
*retSize = 0;
*retName = NULL;
return FALSE;
}
if (!gotSpace && nameIx < ArraySize(name)-1)
{
if (isspace(c))
gotSpace = TRUE;
else if (c != '>')
{
name[nameIx++] = c;
}
}
if (c == '\n')
break;
}
name[nameIx] = 0;
/* Read until next '>' */
for (;;)
{
c = fgetc(f);
if (c == EOF || c == '>')
c = 0;
else if (!isalpha(c))
continue;
else
{
c = ntChars[c];
if (c == 0) c = 'n';
}
if (bufIx >= faFastBufSize)
expandFaFastBuf(bufIx, 0);
faFastBuf[bufIx++] = c;
if (c == 0)
{
*retDna = faFastBuf;
*retSize = bufIx-1;
*retName = name;
return TRUE;
}
}
}
boolean faMixedSpeedReadNext(struct lineFile *lf, DNA **retDna, int *retSize, char **retName)
/* Read in DNA or Peptide FA record in mixed case. Allow any upper or lower case
* letter, or the dash character in. */
{
char c;
int bufIx = 0;
static char name[512];
int lineSize, i;
char *line;
dnaUtilOpen();
/* Read first line, make sure it starts with '>', and read first word
* as name of sequence. */
name[0] = 0;
if (!lineFileNext(lf, &line, &lineSize))
{
*retDna = NULL;
*retSize = 0;
return FALSE;
}
if (line[0] == '>')
{
line = firstWordInLine(skipLeadingSpaces(line+1));
if (line == NULL)
errAbort("Expecting sequence name after '>' line %d of %s", lf->lineIx, lf->fileName);
strncpy(name, line, sizeof(name));
name[sizeof(name)-1] = '\0'; /* Just to make sure name is NULL terminated. */
}
else
{
errAbort("Expecting '>' line %d of %s", lf->lineIx, lf->fileName);
}
/* Read until next '>' */
for (;;)
{
if (!lineFileNext(lf, &line, &lineSize))
break;
if (line[0] == '>')
{
lineFileReuse(lf);
break;
}
if (bufIx + lineSize >= faFastBufSize)
expandFaFastBuf(bufIx, lineSize);
for (i=0; i<lineSize; ++i)
{
c = line[i];
if (isalpha(c) || c == '-')
faFastBuf[bufIx++] = c;
}
}
if (bufIx >= faFastBufSize)
expandFaFastBuf(bufIx, 0);
faFastBuf[bufIx] = 0;
*retDna = faFastBuf;
*retSize = bufIx;
*retName = name;
if (bufIx == 0)
{
warn("Invalid fasta format: sequence size == 0 for element %s",name);
}
return TRUE;
}
boolean faSpeedReadNextKeepCase(struct lineFile *lf,
DNA **retDna, int *retSize, char **retName)
/* Read in next FA entry as fast as we can. Faster than that old,
* pokey faFastReadNext. Return FALSE at EOF.
* The returned DNA and name will be overwritten by the next call
* to this function. */
{
int c;
int bufIx = 0;
static char name[256];
int lineSize, i;
char *line;
dnaUtilOpen();
/* Read first line, make sure it starts wiht '>', and read first word
* as name of sequence. */
name[0] = 0;
if (!lineFileNext(lf, &line, &lineSize))
{
*retDna = NULL;
*retSize = 0;
return FALSE;
}
if (line[0] == '>')
{
line = firstWordInLine(skipLeadingSpaces(line+1));
if (line == NULL)
errAbort("Expecting sequence name after '>' line %d of %s", lf->lineIx, lf->fileName);
strncpy(name, line, sizeof(name));
}
else
{
errAbort("Expecting '>' line %d of %s", lf->lineIx, lf->fileName);
}
/* Read until next '>' */
for (;;)
{
if (!lineFileNext(lf, &line, &lineSize))
break;
if (line[0] == '>')
{
lineFileReuse(lf);
break;
}
if (bufIx + lineSize >= faFastBufSize)
expandFaFastBuf(bufIx);
for (i=0; i<lineSize; ++i)
{
c = line[i];
if (isalpha(c))
{
faFastBuf[bufIx++] = c;
}
}
}
if (bufIx >= faFastBufSize)
expandFaFastBuf(bufIx);
faFastBuf[bufIx] = 0;
*retDna = faFastBuf;
*retSize = bufIx;
*retName = name;
return TRUE;
}