int main(int argc, char **argv) {
int a = 1;
int format = TT_ANY;
mFILE *ofp = mstdout();
char *name = NULL, *oname = NULL;
int output_conf = 0;
read_sections(READ_BASES);
while (a < argc) {
if (strcasecmp(argv[a], "-abi") == 0)
format = TT_ABI;
else if (strcasecmp(argv[a], "-alf") == 0)
format = TT_ALF;
else if (strcasecmp(argv[a], "-scf") == 0)
format = TT_SCF;
else if (strcasecmp(argv[a], "-pln") == 0)
format = TT_PLN;
else if (strcasecmp(argv[a], "-name") == 0) {
if (a == argc)
usage();
name = argv[++a];
} else if (strcasecmp(argv[a], "-output") == 0) {
if (a == argc)
usage();
ofp = mfopen(oname = argv[++a], "wb");
if (NULL == ofp) {
perror(argv[a]);
return 1;
}
} else if (strcasecmp(argv[a], "-conf") == 0) {
output_conf = 1;
} else if (argv[a][0] == '-')
usage();
else
break;
a++;
}
if (a+1 != argc)
usage();
if (!name)
name = oname;
if (!name)
name = argv[a];
return convert(argv[a], format, ofp, name, output_conf);
}
bool Vfile::read(string filename) {
// open the file
ifstream v(filename.c_str());
if (!v) {
cout << "Couldn't open file " << filename << " for reading" << endl;
return false;
}
// read the static portions
read_ok = true;
read_header(v);
read_sections(v);
// check if global bodies were found and datatype is double
if (bodies.GroupID==-1) { cout << "==============================" << endl; read_ok = false; return read_ok; }
assert(bodies.Type==6);
read_dynamics(v);
v.close();
return read_ok;
}
int main(int argc, char *argv[])
{
if (argc != 3)
fail("usage: unself in.self out.elf");
self = mmap_file(argv[1]);
read_header();
read_sections();
if (key_ver != 0x8000)
self_decrypt();
out = fopen(argv[2], "w+");
write_elf();
check_elf();
fclose(out);
return 0;
}
/*
* Read the ALF format sequence from FILE *fp into a Read structure.
* All printing characters (as defined by ANSII C `isprint')
* are accepted, but `N's are translated to `-'s. In this respect we
* are adhering (more or less) to the CSET_DEFAULT uncertainty code set.
*
* Returns:
* Read * - Success, the Read structure read.
* NULLRead - Failure.
*/
Read *fread_alf(FILE *fp) {
Read *read = NULLRead;
int i;
int numPoints;
int sections = read_sections(0);
uint_4 data_size;
uint_4 dataO;
uint_4 header_size=396; /* size of the header of the processed data
section */
uint_2 actBaseDataSize; /* actual number of bytes of data of information
containing the base and basePos information */
int num_points; /* keeps track of the actual number of points,
rather than the early guess of numPoints */
off_t indexO; /* File offset where the index is */
uint_4 baseO; /* File offset where the bases are stored */
/*
* RMD lots of changes below here until end of data reading section
* Some are cosmetic.
* getIndexEntry calls in front of where they were needed, and made
* There is a substantive change to the inner loop of the sequence
* reading section. This now uses fscanf - much less rigid than the
* previous scheme. Note that it reads bp as a float. This is because
* it is a float in multiple trace data files! (bizarre Pharmacia
* programming!).
*/
/*************************************************************
* Read the various file offsets
*************************************************************/
/* indexO is the offset of the index.
* Or I could look for the first label, starting 'ALF'
* if I used 512 then none of the entries are on long
* word boundaries
*/
indexO = 522;
/* offset in file of first base of sequence */
if (! (getIndexEntryLW(fp,indexO,BaseEntryLabel,12,&baseO)) )
goto bail_out;
/* actual size of region containing this data */
if (! (getIndexEntryW(fp,indexO,BaseEntryLabel,10,&actBaseDataSize)) )
goto bail_out;
/* Look for Processed data first. If we fail to find it, then look for
* the Raw data (same format).
*/
/* offset in file to start of processed data segment - there
* is then a header of size header_size (currently 396)
*/
if (! (getIndexEntryLW(fp,indexO,DataEntryLabel,12,&dataO)) ) {
if (! (getIndexEntryLW(fp,indexO,RawDataEntryLabel,12,&dataO)) )
goto bail_out;
/* actual size of region containing this data */
if (! (getIndexEntryLW(fp,indexO,RawDataEntryLabel,10,&data_size)) )
goto bail_out;
} else {
/* actual size of region containing this data */
if (! (getIndexEntryLW(fp,indexO,DataEntryLabel,10,&data_size)) )
goto bail_out;
}
/* Because each trace value is stored in a 2 byte
* integer, thus to store A C G T information
* it takes 8 bytes. So subtract off the header and
* divide by 8
*/
numPoints = (int)((data_size - header_size)/ 8);
/* Allocate the sequence */
if (NULLRead == (read = read_allocate(numPoints, BASELIMIT)))
goto bail_out;
/*************************************************************
* Read the bases information
*************************************************************/
if (sections & READ_BASES) {
/* new locals introduced by LFW and/or RMD for the ALF */
int numBases; /* number of nucleotides read in */
float bp;
char ch;
if (!(fseek(fp, (off_t)baseO, 0) == 0))
goto bail_out;
for (numBases = 0; (unsigned)ftell(fp) < baseO+(unsigned short)actBaseDataSize
&& numBases<BASELIMIT;) {
char line[200];
fgets(line, (int)sizeof(line), fp);
sscanf(line, "%c %*d %f", &ch, &bp);
/* we convert ch to Staden format here */
switch (ch) {
case 'A':
case 'C':
case 'G':
case 'T':
break;
default:
ch = '-';
/*
if (isupper(ch))
ch = '-';
else
ch = '\0';
*/
}
if (ch) {
read->base[numBases] = ch;
read->prob_A[numBases] = 0;
read->prob_C[numBases] = 0;
read->prob_G[numBases] = 0;
read->prob_T[numBases] = 0;
read->basePos[numBases] = bp;
++numBases;
}
}
read->base[numBases] = 0;
read->NBases = numBases;
}
/*************************************************************
* Read the trace information
*************************************************************/
if (sections & READ_SAMPLES) {
/*
* Traces are stored as 2 byte integers in records in the order of
* A C G T A C G T ...
*/
if (fseek(fp, (off_t)(dataO+header_size), 0) != 0)
goto bail_out;
num_points = 0;
for (i=0; i < read->NPoints; i++) {
if (!le_read_int_2(fp, &(read->traceA[i])))
goto bail_out;
if (read->maxTraceVal < read->traceA[i])
read->maxTraceVal = read->traceA[i];
if (!le_read_int_2(fp, &(read->traceC[i])))
goto bail_out;
if (read->maxTraceVal < read->traceC[i])
read->maxTraceVal = read->traceC[i];
if (!le_read_int_2(fp, &(read->traceG[i])))
goto bail_out;
if (read->maxTraceVal < read->traceG[i])
read->maxTraceVal = read->traceG[i];
if (!le_read_int_2(fp, &(read->traceT[i])))
goto bail_out;
if (read->maxTraceVal < read->traceT[i])
read->maxTraceVal = read->traceT[i];
if (read->traceA[i]==0 && read->traceT[i]==0 &&
read->traceC[i]==0 && read->traceG[i]==0 &&
i > (numPoints-64))
break;
num_points++;
}
}
/* SUCCESS */
read->format = TT_ALF;
return(read);
/* FAILURE */
bail_out:
if (read)
read_deallocate(read);
return NULLRead;
}
int main(int argc, char **argv) {
Read *r = NULL;
char *directory = NULL;
char *ident, *value;
int ident_len, value_len;
int i, j, found;
int *found_args = NULL;
char **FileList = NULL;
char trace_filename[FILENAME_MAX]="";
int num_traces, trace_iter, files_read = 0;
char *str;
if(argc != 2)
usage();
directory = argv[1];
// Get a list of all chromatogram files in the directory
num_traces = GetFileList(&FileList, directory);
if(num_traces == 0){
fprintf(stderr,"* Path %d yielded 0 files...exiting\n", num_traces);
exit(2);
}
/* step through all the sequences */
for (trace_iter = 0; trace_iter < num_traces; trace_iter++){
if(r){
read_deallocate(r);
}
//Get the file name from the iterator.
sprintf(trace_filename, "%s/%s",directory, FileList[trace_iter]);
/* Read the file */
read_sections(READ_COMMENTS);
if (NULL == (r = read_reading(trace_filename, TT_ANY))) {
continue; // don't worry about it
}
files_read++;
if (!r->info)
return 1;
for(str = strtok(r->info,"\n"); str != NULL; str = strtok(NULL,"\n")){
char *name;
char *value;
int items = 0;
char *start,*end;
name = str;
value = strchr(name,'=');
*value = '\0'; // skip over the '='
value++;
if(!strcmp(name,"RUND")){
start = value;
end = strstr(value," - ");
*end = '\0'; // terminate the start string
end += 3; // skip over the " - "
fprintf(stdout,"%s.RUND=start=%s,end=%s\n", FileList[trace_iter], start,end);
}else if(! strcmp(name,"DATE")){
start = value;
end = strstr(value," to ");
*end = '\0'; // terminate the start string
end += 4; // skip over the " to "
fprintf(stdout,"%s.DATE=start=%s,end=%s\n", FileList[trace_iter],start,end);
}else{
fprintf(stdout,"%s.%s=%s\n", FileList[trace_iter],name, value);
}
}
}
return (files_read == 0); // is zero, unless we read nothing.
}
/*
* fread_ztr
*
* Reads a ZTR file from 'fp'. This checks for the correct magic number and
* major version number, but not minor version number.
*
* FIXME: Add automatic uncompression?
*
* Arguments:
* fp A readable FILE pointer
*
* Returns:
* Success: Pointer to a ztr_t structure (malloced)
* Failure: NULL
*/
ztr_t *fread_ztr(FILE *fp) {
ztr_t *ztr;
ztr_chunk_t *chunk;
int sections = read_sections(0);
/* Allocate */
if (NULL == (ztr = new_ztr()))
return NULL;
/* Read the header */
if (-1 == ztr_read_header(fp, &ztr->header))
return NULL;
/* Check magic number and version */
if (memcmp(ztr->header.magic, ZTR_MAGIC, 8) != 0)
return NULL;
if (ztr->header.version_major != ZTR_VERSION_MAJOR)
return NULL;
/* Load chunks */
while ((chunk = ztr_read_chunk_hdr(fp))) {
/*
char str[5];
fprintf(stderr, "Read chunk %.4s %08x length %d\n",
ZTR_BE2STR(chunk->type, str), chunk->type, chunk->dlength);
*/
switch(chunk->type) {
case ZTR_TYPE_HEADER:
/* End of file */
return ztr;
case ZTR_TYPE_SAMP:
case ZTR_TYPE_SMP4:
if (! (sections & READ_SAMPLES)) {
fseek(fp, chunk->dlength, SEEK_CUR);
xfree(chunk);
continue;
}
break;
case ZTR_TYPE_BASE:
case ZTR_TYPE_BPOS:
case ZTR_TYPE_CNF4:
case ZTR_TYPE_CNF1:
case ZTR_TYPE_CSID:
if (! (sections & READ_BASES)) {
fseek(fp, chunk->dlength, SEEK_CUR);
xfree(chunk);
continue;
}
break;
case ZTR_TYPE_TEXT:
if (! (sections & READ_COMMENTS)) {
fseek(fp, chunk->dlength, SEEK_CUR);
xfree(chunk);
continue;
}
break;
case ZTR_TYPE_CLIP:
case ZTR_TYPE_FLWO:
case ZTR_TYPE_FLWC:
break;
/*
default:
fprintf(stderr, "Unknown chunk type '%s': skipping\n",
ZTR_BE2STR(chunk->type,str));
fseek(fp, chunk->dlength, SEEK_CUR);
xfree(chunk);
continue;
*/
}
chunk->ztr_owns = 1;
chunk->data = (char *)xmalloc(chunk->dlength);
if (chunk->dlength != fread(chunk->data, 1, chunk->dlength, fp)) {
delete_ztr(ztr);
return NULL;
}
ztr->nchunks++;
ztr->chunk = (ztr_chunk_t *)xrealloc(ztr->chunk, ztr->nchunks *
sizeof(ztr_chunk_t));
memcpy(&ztr->chunk[ztr->nchunks-1], chunk, sizeof(*chunk));
xfree(chunk);
}
return ztr;
}
/*
* Read the SCF format sequence from FILE *fp into a 'scf' structure.
* A NULL result indicates failure.
*/
Scf *fread_scf(FILE *fp) {
Scf *scf;
Header h;
int err;
float scf_version;
int sections = read_sections(0);
/* Read header */
if (read_scf_header(fp, &h) == -1) {
return NULL;
}
/* Allocate memory */
if (NULL == (scf = scf_allocate(h.samples, h.sample_size,
h.bases, h.comments_size,
h.private_size)))
return NULL;
/* fake things for older style SCF -- SD */
if (h.sample_size != 1 && h.sample_size != 2) h.sample_size = 1;
scf_version = scf_version_str2float(h.version);
memcpy(&scf->header, &h, sizeof(Header));
if (sections & READ_SAMPLES) {
/* Read samples */
if (fseek(fp, (off_t)h.samples_offset, 0 /* SEEK_SET */) != 0) {
scf_deallocate(scf);
return NULL;
}
if ( 2.9 > scf_version ) {
if (h.sample_size == 1) {
err= read_scf_samples1(fp, scf->samples.samples1, h.samples);
}
else {
err= read_scf_samples2(fp, scf->samples.samples2, h.samples);
}
}
else {
if (h.sample_size == 1) {
err= read_scf_samples31(fp, scf->samples.samples1, h.samples);
}
else {
err= read_scf_samples32(fp, scf->samples.samples2, h.samples);
}
}
if (-1 == err) {
scf_deallocate(scf);
return NULL;
}
}
if (sections & READ_BASES) {
/* Read bases */
if (fseek(fp, (off_t)h.bases_offset, 0 /* SEEK_SET */) != 0) {
scf_deallocate(scf);
return NULL;
}
if ( 2.9 > scf_version ) {
if (-1 == read_scf_bases(fp, scf->bases, h.bases)) {
scf_deallocate(scf);
return NULL;
}
}
else {
if (-1 == read_scf_bases3(fp, scf->bases, h.bases)) {
scf_deallocate(scf);
return NULL;
}
}
}
if (sections & READ_COMMENTS) {
/* Read comments */
if (scf->comments) {
if (fseek(fp,(off_t)(h.comments_offset), 0) != 0
|| -1 == read_scf_comment(fp, scf->comments,
h.comments_size)) {
/*
* Was: "scf_deallocate(scf); return NULL;".
* We now simply clear the comments and gracefully continue.
*/
fprintf(stderr, "Warning: SCF file had invalid comment field\n");
xfree(scf->comments);
scf->comments = NULL;
} else {
scf->comments[h.comments_size] = '\0';
}
}
}
/* Read private data */
if (h.private_size) {
if (-1 == fseek(fp, (off_t)(h.private_offset), 0) ||
h.private_size != fread(scf->private_data, 1, h.private_size, fp)){
scf_deallocate(scf);
return NULL;
}
}
return scf;
}
/*
* Translates an Scf structure into a Read structure.
* The Scf structure is left unchanged.
*
* Returns:
* A pointer to an allocated Read structure upon success.
* NULLRead upon failure.
*/
Read *scf2read(Scf *scf) {
Read *read;
register int i, i_end;
TRACE max_val = 0;
int sections = read_sections(0);
int nsamples = 0;
int nbases = 0;
/* allocate */
if (sections & READ_SAMPLES)
nsamples = scf->header.samples;
if (sections & READ_BASES)
nbases = scf->header.bases;
read = read_allocate(nsamples, nbases);
if (NULLRead == read)
return NULLRead;
if (sections & READ_SAMPLES) {
/* copy the samples */
i_end = scf->header.samples;
read->NPoints = i_end;
if (scf->header.sample_size == 1) {
for (i = 0; i < i_end; i++) {
read->traceA[i] = scf->samples.samples1[i].sample_A;
read->traceC[i] = scf->samples.samples1[i].sample_C;
read->traceG[i] = scf->samples.samples1[i].sample_G;
read->traceT[i] = scf->samples.samples1[i].sample_T;
if (read->traceA[i] > max_val) max_val = read->traceA[i];
if (read->traceC[i] > max_val) max_val = read->traceC[i];
if (read->traceG[i] > max_val) max_val = read->traceG[i];
if (read->traceT[i] > max_val) max_val = read->traceT[i];
}
} else { /* sample_size == 2 */
for (i = 0; i < i_end; i++) {
read->traceA[i] = scf->samples.samples2[i].sample_A;
read->traceC[i] = scf->samples.samples2[i].sample_C;
read->traceG[i] = scf->samples.samples2[i].sample_G;
read->traceT[i] = scf->samples.samples2[i].sample_T;
if (read->traceA[i] > max_val) max_val = read->traceA[i];
if (read->traceC[i] > max_val) max_val = read->traceC[i];
if (read->traceG[i] > max_val) max_val = read->traceG[i];
if (read->traceT[i] > max_val) max_val = read->traceT[i];
}
}
read->maxTraceVal = max_val;
}
if (sections & READ_BASES) {
/* copy the bases */
i_end = scf->header.bases;
read->NBases = i_end;
for (i = 0; i < i_end; i++) {
read->basePos[i] = scf->bases[i].peak_index;
read->prob_A[i] = scf->bases[i].prob_A;
read->prob_C[i] = scf->bases[i].prob_C;
read->prob_G[i] = scf->bases[i].prob_G;
read->prob_T[i] = scf->bases[i].prob_T;
read->base[i] = scf->bases[i].base;
}
read->base[i] = 0;
}
if (sections & READ_COMMENTS) {
/* allocate and copy the comments */
if (scf->header.comments_size > 0 && scf->comments) {
read->info = (char *)xmalloc(scf->header.comments_size+1);
if (NULL == read->info) {
read_deallocate(read);
return NULLRead;
}
memcpy(read->info, scf->comments, scf->header.comments_size);
read->info[scf->header.comments_size] = '\0';
}
}
/* other bits and pieces */
read->leftCutoff = scf->header.bases_left_clip;
read->rightCutoff = read->NBases - scf->header.bases_right_clip + 1;
read->format = TT_SCF;
return read;
}
/*
* Allocate a new sequence, with the given sizes.
* Returns:
* "Read *" for success
* "NULLRead" for failure
*/
Read *read_allocate(int num_points, int num_bases) {
Read *seq = NULLRead;
int sections = read_sections(0);
/* Allocate the body of the sequence */
if ((seq = (Read *)xmalloc(sizeof(Read))) == NULL)
return(NULLRead);
seq->NPoints = num_points;
seq->NBases = num_bases;
/*
* Initialise the body, all pointers are set to NULL so we can
* happily call `read_deallocate()`.
*/
seq->leftCutoff = 0;
seq->rightCutoff = 0;
seq->maxTraceVal = 0;
seq->baseline = 0;
seq->traceC = NULL;
seq->traceA = NULL;
seq->traceG = NULL;
seq->traceT = NULL;
seq->base = NULL;
seq->basePos = NULL;
seq->info = NULL;
seq->format = TT_ANY;
seq->trace_name = NULL;
seq->prob_A = NULL;
seq->prob_C = NULL;
seq->prob_G = NULL;
seq->prob_T = NULL;
seq->orig_trace_format = TT_ANY;
seq->orig_trace = NULL;
seq->orig_trace_free = NULL;
seq->ident = NULL;
/* Allocate space for the bases - 1 extra for the ->base field so
* that we can treat it as a NULL terminated string.
*/
if (sections & READ_BASES &&
(((seq->base = (char *)xcalloc(num_bases+1,1)) == NULL) ||
((seq->basePos = (uint_2 *)xcalloc(num_bases+1,2)) == NULL) ||
((seq->prob_A = (char *)xcalloc(num_bases+1,1)) == NULL) ||
((seq->prob_C = (char *)xcalloc(num_bases+1,1)) == NULL) ||
((seq->prob_G = (char *)xcalloc(num_bases+1,1)) == NULL) ||
((seq->prob_T = (char *)xcalloc(num_bases+1,1)) == NULL))
)
{
read_deallocate(seq);
return NULLRead;
}
if (sections & READ_SAMPLES &&
(((seq->traceC =(TRACE *)xcalloc(num_points+1, 2)) == NULL)||
((seq->traceA =(TRACE *)xcalloc(num_points+1, 2)) == NULL)||
((seq->traceG =(TRACE *)xcalloc(num_points+1, 2)) == NULL)||
((seq->traceT =(TRACE *)xcalloc(num_points+1, 2)) == NULL))
)
{
read_deallocate(seq);
return NULLRead;
}
return seq;
}
/*
* ---------------------------------------------------------------------------
* Loads confidence values from the trace file and averages them.
* 'opos' is optional - if not known then set to NULL.
*
* Returns 0 for success
* -1 for failure
*/
int get_read_conf(Exp_info *e, int length, int2 *opos, int1 *conf) {
int ttype, i;
FILE *fp;
uint_1 *prob_A, *prob_C, *prob_G, *prob_T;
char *seq;
float scf_version;
int nbases = 0;
/* Sanity check */
if (!(exp_Nentries(e,EFLT_LT) && exp_Nentries(e,EFLT_LN)))
return -1;
/* Find and load trace file */
ttype = trace_type_str2int(exp_get_entry(e, EFLT_LT));
if (ttype != TT_SCF &&
ttype != TT_ZTR)
return -1;
/*
* We only support direct reading accuracy values from SCF files.
* Otherwise we have to take a slower approach.
*/
if (ttype != TT_SCF) {
Read *r;
int sec = read_sections(0);
read_sections(READ_BASES);
if (NULL == (r = read_reading(exp_get_entry(e,EFLT_LN), TT_ANYTR))) {
read_sections(sec);
return -1;
}
prob_A = (int1 *)xmalloc(r->NBases);
prob_C = (int1 *)xmalloc(r->NBases);
prob_G = (int1 *)xmalloc(r->NBases);
prob_T = (int1 *)xmalloc(r->NBases);
seq = (char *)xmalloc(r->NBases);
memcpy(prob_A, r->prob_A, r->NBases);
memcpy(prob_C, r->prob_C, r->NBases);
memcpy(prob_G, r->prob_G, r->NBases);
memcpy(prob_T, r->prob_T, r->NBases);
memcpy(seq, r->base, r->NBases);
nbases = r->NBases;
read_deallocate(r);
read_sections(sec);
} else {
Header h;
/* For SCF files we read directly - the above code would also do. */
if (NULL == (fp = open_trace_file(exp_get_entry(e,EFLT_LN), NULL)))
return -1;
/* Read the SCF header */
if (-1 == read_scf_header(fp, &h))
return -1;
scf_version = scf_version_str2float(h.version);
nbases = h.bases;
/* Alloc memory */
prob_A = (uint_1 *)xmalloc(h.bases * sizeof(*prob_A));
prob_C = (uint_1 *)xmalloc(h.bases * sizeof(*prob_A));
prob_G = (uint_1 *)xmalloc(h.bases * sizeof(*prob_A));
prob_T = (uint_1 *)xmalloc(h.bases * sizeof(*prob_A));
seq = (char *)xmalloc(h.bases * sizeof(*seq));
if (NULL == prob_A ||
NULL == prob_C ||
NULL == prob_G ||
NULL == prob_T ||
NULL == seq)
return -1;
/* Load base scores */
if (scf_version >= 3.0) {
/*
* Version 3 base format:
* num_bases * 4byte peak index
* num_bases * prob_A
* num_bases * prob_C
* num_bases * prob_G
* num_bases * prob_T
* num_bases * base
* num_bases * spare (x3)
*/
fseek(fp, (off_t)h.bases_offset + 4 * h.bases, SEEK_SET);
if (h.bases != fread(prob_A, 1, h.bases, fp))
return -1;
if (h.bases != fread(prob_C, 1, h.bases, fp))
return -1;
if (h.bases != fread(prob_G, 1, h.bases, fp))
return -1;
if (h.bases != fread(prob_T, 1, h.bases, fp))
return -1;
if (h.bases != fread(seq, 1, h.bases, fp))
return -1;
} else {
int i;
uint_1 buf[12];
/*
* Version 2 base format
* num_bases * base_struct, where base_struct is 12 bytes:
* 0-3 peak_index
* 4-7 prob_A/C/G/T
* 8 base
* 9- spare
*/
fseek(fp, (off_t)h.bases_offset, SEEK_SET);
for (i = 0; (unsigned)i < h.bases; i++) {
if (1 != fread(buf, 12, 1, fp))
return -1;
prob_A[i] = buf[4];
prob_C[i] = buf[5];
prob_G[i] = buf[6];
prob_T[i] = buf[7];
seq[i] = buf[8];
}
}
fclose(fp);
}
/* Determine confidence values */
if (opos) {
for (i=0; i<length; i++) {
if (opos[i] == 0) {
/* Inserted base, change to 0% */
conf[i] = 0;
} else {
switch(seq[opos[i]-1]) {
case 'a':
case 'A':
conf[i] = prob_A[opos[i]-1];
break;
case 'c':
case 'C':
conf[i] = prob_C[opos[i]-1];
break;
case 'g':
case 'G':
conf[i] = prob_G[opos[i]-1];
break;
case 't':
case 'T':
conf[i] = prob_T[opos[i]-1];
break;
default:
conf[i] = 2;
}
}
}
} else {
int mlength = MIN(length, nbases);
for (i=0; i < mlength; i++) {
switch(seq[i]) {
case 'a':
case 'A':
conf[i] = prob_A[i];
break;
case 'c':
case 'C':
conf[i] = prob_C[i];
break;
case 'g':
case 'G':
conf[i] = prob_G[i];
break;
case 't':
case 'T':
conf[i] = prob_T[i];
break;
case 'n':
case 'N':
case '-':
conf[i] = (prob_A[i] + prob_C[i] + prob_G[i] + prob_T[i]) / 4;
break;
default:
conf[i] = 2;
}
}
for (; i < length; i++)
conf[i] = 2;
}
xfree(prob_A);
xfree(prob_C);
xfree(prob_G);
xfree(prob_T);
xfree(seq);
return 0;
}
void
elf_mod_symload(int strtablen)
{
Elf_Ehdr ehdr;
char *shstrtab;
struct elf_section *head, *s;
char *symbuf, *strbuf;
/*
* Seek to the text offset to start loading...
*/
if (lseek(modfd, 0, SEEK_SET) == -1)
err(12, "lseek");
if (read_elf_header(modfd, &ehdr) < 0)
return;
shstrtab = read_shstring_table(modfd, &ehdr);
read_sections(modfd, &ehdr, shstrtab, &head);
for (s = head; s; s = s->next) {
struct elf_section *p = s;
if ((p->type == SHT_SYMTAB) || (p->type == SHT_DYNSYM)) {
if (debug)
fprintf(stderr, "loading `%s': addr = %p, "
"size = %#lx\n",
s->name, s->addr, (u_long)s->size);
/*
* Seek to the file offset to start loading it...
*/
if (lseek(modfd, p->offset, SEEK_SET) == -1)
err(12, "lseek");
symbuf = malloc(p->size);
if (symbuf == 0)
err(13, "malloc");
if (read(modfd, symbuf, p->size) != p->size)
err(14, "read");
loadsym(symbuf, p->size);
free(symbuf);
}
}
for (s = head; s; s = s->next) {
struct elf_section *p = s;
if ((p->type == SHT_STRTAB) &&
(strcmp(p->name, ".strtab") == 0 )) {
if (debug)
fprintf(stderr, "loading `%s': addr = %p, "
"size = %#lx\n",
s->name, s->addr, (u_long)s->size);
/*
* Seek to the file offset to start loading it...
*/
if (lseek(modfd, p->offset, SEEK_SET) == -1)
err(12, "lseek");
strbuf = malloc(p->size);
if (strbuf == 0)
err(13, "malloc");
if (read(modfd, strbuf, p->size) != p->size)
err(14, "read");
loadsym(strbuf, p->size);
free(strbuf);
}
}
free(shstrtab);
free_sections(head);
return;
}
/* load a prelinked module; returns entry point */
void *
elf_mod_load(int fd)
{
Elf_Ehdr ehdr;
size_t zero_size = 0;
size_t b;
ssize_t n;
char *shstrtab;
struct elf_section *head, *s;
char buf[10 * BUFSIZ];
void *addr = NULL;
if (read_elf_header(fd, &ehdr) < 0)
return NULL;
shstrtab = read_shstring_table(fd, &ehdr);
read_sections(fd, &ehdr, shstrtab, &head);
for (s = head; s; s = s->next) {
if (s->type != SHT_STRTAB && s->type != SHT_SYMTAB &&
s->type != SHT_DYNSYM) {
if (debug)
fprintf(stderr, "loading `%s': addr = %p, "
"size = %#lx\n",
s->name, s->addr, (u_long)s->size);
if (s->type == SHT_NOBITS) {
/* skip some space */
zero_size += s->size;
} else {
if (addr != NULL) {
/*
* if there is a gap in the prelinked
* module, transfer some empty space.
*/
zero_size += (char*)s->addr -
(char*)addr;
}
if (zero_size) {
loadspace(zero_size);
zero_size = 0;
}
b = s->size;
if (lseek(fd, s->offset, SEEK_SET) == -1)
err(1, "lseek");
while (b) {
n = read(fd, buf, MIN(b, sizeof(buf)));
if (n == 0)
errx(1, "unexpected EOF");
if (n < 0)
err(1, "read");
loadbuf(buf, n);
b -= n;
}
addr = (char*)s->addr + s->size;
}
}
}
if (zero_size)
loadspace(zero_size);
free_sections(head);
free(shstrtab);
return (void *)ehdr.e_entry;
}
/* return size needed by the module */
int
elf_mod_sizes(int fd, size_t *modsize, int *strtablen,
struct lmc_resrv *resrvp, struct stat *sp)
{
Elf_Ehdr ehdr;
ssize_t off = 0;
size_t data_hole = 0;
char *shstrtab, *strtab;
struct elf_section *head, *s, *stab;
if (read_elf_header(fd, &ehdr) < 0)
return -1;
shstrtab = read_shstring_table(fd, &ehdr);
read_sections(fd, &ehdr, shstrtab, &head);
for (s = head; s; s = s->next) {
/* XXX impossible! */
if (s->type == SHT_STRTAB && s->type == SHT_SYMTAB &&
s->type == SHT_DYNSYM)
continue;
if (debug)
fprintf(stderr,
"%s: addr = %p size = %#lx align = %#lx\n",
s->name, s->addr, (u_long)s->size, (u_long)s->align);
/*
* XXX try to get rid of the hole before the data
* section that GNU-ld likes to put there
*/
if (strcmp(s->name, ".data") == 0 && s->addr > (void *)off) {
data_offset = roundup(off, s->align);
if (debug)
fprintf(stderr, ".data section forced to "
"offset %p (was %p)\n",
(void *)data_offset, s->addr);
/* later remove size of compressed hole from off */
data_hole = (ssize_t)s->addr - data_offset;
}
off = (ssize_t)s->addr + s->size;
}
off -= data_hole;
/* XXX round to pagesize? */
*modsize = roundup(off, sysconf(_SC_PAGESIZE));
/* get string table length */
strtab = read_string_table(fd, head, strtablen);
free(shstrtab);
free(strtab);
/* get symbol table sections */
get_symtab(&head);
stab = head;
resrvp->sym_symsize = 0;
while (stab) {
resrvp->sym_symsize += stab->size;
stab = stab->next;
}
resrvp->sym_size = resrvp->sym_symsize + *strtablen;
free_sections(head);
return (0);
}
