/*
function ascii NCBI:align:ref_seq_id ( I64 ref_id );
*/
static
rc_t CC align_ref_seq_id ( void *data, const VXformInfo *info,
int64_t row_id, VRowResult *rslt, uint32_t argc, const VRowData argv[] )
{
rc_t rc;
RefSeqID const *self = ( void const * )data;
char const *name = NULL;
uint32_t name_len;
/* get start and length of reference segment */
int64_t const *ref_id = argv[REF_ID].u.data.base;
assert( argv[ REF_ID ].u.data.elem_bits == sizeof( *ref_id ) * 8) ;
if ( self->curs == NULL || argv[ REF_ID ].u.data.elem_count == 0 )
{
rslt->elem_count = 0;
return 0;
}
ref_id += argv[ REF_ID] .u.data.first_elem;
SUB_DEBUG( ( "SUB.Rd in 'align_ref_seq_id.c' at #%lu\n", ref_id[ 0 ] ) );
rc = VCursorCellDataDirect( self->curs, ref_id[ 0 ], self->seqID_idx, NULL, (void const **)&name, NULL, &name_len );
if ( GetRCState( rc ) == rcNotFound && GetRCObject( rc ) == rcRow )
{
name = "";
name_len = 0;
}
else if ( rc != 0 )
{
return rc;
}
rc = KDataBufferCast( rslt->data, rslt->data, sizeof( name[ 0 ] ) * 8, true );
if ( rc != 0 )
return rc;
rc = KDataBufferResize( rslt->data, name_len );
if ( rc != 0 )
return rc;
memmove( rslt->data->base, name, sizeof( name[ 0 ] ) * name_len );
rslt->elem_count = name_len;
rslt->elem_bits = sizeof( name[ 0 ] ) * 8;
return rc;
}
static
rc_t CC dynamic_read_desc_static ( void *self, const VXformInfo *info, int64_t row_id,
VRowResult *rslt, uint32_t argc, const VRowData argv [] )
{
rc_t rc;
dynamic_read_desc *p;
KDataBuffer *dst = rslt -> data;
/* severe error if adapter is longer than spot */
if ( argv [ 0 ] . u . data . elem_count < argv [ 1 ] . u . data . elem_count )
return RC ( rcSRA, rcFunction, rcExecuting, rcData, rcCorrupt );
/* the buffer should have already been given the correct element size */
if ( dst -> elem_bits != 32 * 3 )
{
rc = KDataBufferCast ( dst, dst, 32 * 3, true );
if ( rc != 0 )
return rc;
}
/* we always produce 2 reads */
if ( dst -> elem_count != 2 )
{
rc = KDataBufferResize ( dst, 2 );
if ( rc != 0 )
return rc;
}
p = dst -> base;
/* adapter */
p [ 0 ] [ dyn_read_type ] = SRA_READ_TYPE_TECHNICAL;
p [ 0 ] [ dyn_read_start ] = 0;
assert(argv [ 1 ] . u . data . elem_count >> 32 == 0);
p [ 0 ] [ dyn_read_len ] = (uint32_t)argv [ 1 ] . u . data . elem_count;
/* fragment */
p [ 1 ] [ dyn_read_type ] = SRA_READ_TYPE_BIOLOGICAL;
p [ 1 ] [ dyn_read_start ] = p [ 0 ] [ dyn_read_len ];
assert(argv [ 0 ] . u . data . elem_count >> 32 == 0);
p [ 1 ] [ dyn_read_len ] = (uint32_t)(argv [ 0 ] . u . data . elem_count) - p [ 0 ] [ dyn_read_len ];
rslt -> elem_count = 2;
return 0;
}
/*
function ascii NCBI:align:ref_name ( I64 ref_id );
*/
static
rc_t CC align_ref_name ( void *data, const VXformInfo *info,
int64_t row_id, VRowResult *rslt, uint32_t argc, const VRowData argv[] )
{
rc_t rc;
RefName const *self = (void const *)data;
char const *name = NULL;
uint32_t name_len;
/* get start and length of reference segment */
int64_t const *ref_id = argv[REF_ID].u.data.base;
if (argv[REF_ID].u.data.elem_count == 0)
rc = RC(rcAlign, rcFunction, rcExecuting, rcRow, rcNotFound);
else {
assert(argv[REF_ID].u.data.elem_bits == sizeof(*ref_id) * 8);
ref_id += argv[REF_ID].u.data.first_elem;
rc = VCursorCellDataDirect(self->curs, ref_id[0], self->name_idx, NULL, (void const **)&name, NULL, &name_len);
}
if (GetRCState(rc) == rcNotFound && GetRCObject(rc) == rcRow) {
name = "";
name_len = 0;
}
else if (rc) return rc;
rc = KDataBufferCast(rslt->data, rslt->data, sizeof(name[0]) * 8, true);
if (rc) return rc;
rc = KDataBufferResize(rslt->data, name_len);
if (rc) return rc;
memcpy(rslt->data->base, name, sizeof(name[0]) * name_len);
rslt->elem_count = name_len;
rslt->elem_bits = sizeof(name[0]) * 8;
return rc;
}
static
rc_t CC environment_read_func(
void *Self,
const VXformInfo *info,
int64_t row_id,
VRowResult *rslt,
uint32_t argc,
const VRowData argv[]
) {
const KDataBuffer *value = Self;
rc_t rc = 0;
rslt->data->elem_bits = value->elem_bits;
rslt->data->elem_count = 0;
rc = KDataBufferResize(rslt->data, value->elem_count);
if (rc == 0) {
memcpy(rslt->data->base, value->base, KDataBufferBytes(value));
rc = KDataBufferCast(rslt->data, rslt->data, rslt->elem_bits, true);
if (rc == 0)
rslt->elem_count = rslt->data->elem_count;
}
return rc;
}
static
rc_t CC dynamic_read_desc_with_linker ( void *xself, const VXformInfo *info, int64_t row_id,
VRowResult *rslt, uint32_t argc, const VRowData argv [] )
{
rc_t rc;
dynamic_read_desc *p;
KDataBuffer *dst = rslt -> data;
const linker_agrep *self = ( const void* ) xself;
AgrepMatch match;
int32_t found;
AgrepFlags agrepflags;
Agrep *agrep;
/* AgrepCallArgs args; */
const char *agreppattern;
char buf[4096];
const char *text;
uint32_t textlen;
/* severe error if adapter is longer than spot */
if ( argv [ 0 ] . u . data . elem_count < argv [ 1 ] . u . data . elem_count )
return RC ( rcSRA, rcFunction, rcExecuting, rcData, rcCorrupt );
/* the buffer should have already been given the correct element size */
if ( dst -> elem_bits != 32 * 3 )
{
rc = KDataBufferCast ( dst, dst, 32 * 3, true );
if ( rc != 0 )
return rc;
}
/* we always produce 4 reads for when the linker is present */
if ( dst -> elem_count != 4 )
{
rc = KDataBufferResize ( dst, 4 );
if ( rc != 0 )
return rc;
}
/* TBD - a mechanism for detecting when this has not changed
since typically it will be identical for every row in a table
but not necessarily so */
agreppattern = argv[2].u.data.base;
agreppattern += argv[2].u.data.first_elem;
textlen = (uint32_t)string_copy(buf, sizeof buf, agreppattern, argv[2].u.data.elem_count);
if ( textlen >= sizeof buf )
return RC ( rcSRA, rcFunction, rcExecuting, rcData, rcExcessive );
text = argv[0].u.data.base;
text += argv[0].u.data.first_elem;
assert(argv[0].u.data.elem_count >> 32 == 0);
textlen = (uint32_t)argv[0].u.data.elem_count;
text += argv[1].u.data.elem_count;
textlen -= argv[1].u.data.elem_count;
agrepflags = AGREP_TEXT_EXPANDED_2NA
| AGREP_PATTERN_4NA
| AGREP_EXTEND_BETTER
| AGREP_LEFT_MAINTAIN_SCORE
| AGREP_ANYTHING_ELSE_IS_N;
/* This might fail due to size restrictions. */
rc = AgrepMake(&agrep, agrepflags | AGREP_ALG_MYERS, buf);
if (rc == 0) {
/* fprintf(stderr, "Using myers.\n"); */
} else {
rc = AgrepMake(&agrep, agrepflags | AGREP_ALG_MYERS_UNLTD, buf);
/* Try one more time. */
if (rc) {
rc = AgrepMake(&agrep, agrepflags | AGREP_ALG_DP, buf);
}
if (rc)
return rc;
}
found = AgrepFindBest(agrep, self->edit_distance, text, textlen, &match);
if (found) {
p = dst -> base;
/* adapter */
p [ 0 ] [ dyn_read_type ] = SRA_READ_TYPE_TECHNICAL;
p [ 0 ] [ dyn_read_start ] = 0;
assert(argv [ 1 ] . u . data . elem_count >> 32 == 0);
p [ 0 ] [ dyn_read_len ] = (uint32_t)argv [ 1 ] . u . data . elem_count;
/* fragment */
p [ 1 ] [ dyn_read_type ] = SRA_READ_TYPE_BIOLOGICAL;
p [ 1 ] [ dyn_read_start ] = (uint32_t)argv [ 1 ] . u . data . elem_count;
p [ 1 ] [ dyn_read_len ] = match.position;
/* linker */
p [ 2 ] [ dyn_read_type ] = SRA_READ_TYPE_TECHNICAL;
p [ 2 ] [ dyn_read_start ] = match.position + (uint32_t)argv[1].u.data.elem_count;
p [ 2 ] [ dyn_read_len ] = match.length;
/* fragment */
p [ 3 ] [ dyn_read_type ] = SRA_READ_TYPE_BIOLOGICAL;
p [ 3 ] [ dyn_read_start ] = match.position + match.length + (uint32_t)argv[1].u.data.elem_count;
p [ 3 ] [ dyn_read_len ] = (uint32_t)argv [ 0 ] . u . data . elem_count - match.position - match.length - (uint32_t)argv[1].u.data.elem_count;
rslt -> elem_count = 4;
} else {
/* format_spot_name ( ascii name_fmt, I32 X, I32 Y * ascii spot_name );
* given a name format string, X, and Y
* produce a reconstructed spot name string
*
* "name_fmt" [ DATA ] - name format string ( see format explanation below )
*
* "X" [ DATA ] - X coordinate for spot
*
* "Y" [ DATA ] - Y coordinate for spot
*
* "spot_name" [ DATA, OPTIONAL ] - potential source of unformatted names
*
* SYNOPSIS:
* "name_fmt" may have any ASCII characters
* the special character '$' is an escape symbol
* when followed by a recognized format character,
* both the '$' and its format character will be
* replaced with a numeral generated from X and/or Y.
*
* when "spot_name" is present and the "name_fmt" row is empty,
* output is taken verbatim from "spot_name"
*
* FORMAT:
* 'X' ( or 'x' ) = substitute with a decimal representation for X
* 'Y' ( or 'y' ) = substitute with a decimal representation for Y
* 'Q' ( or 'q' ) = calculate 454-style base-36 representation
* for both X and Y
*/
static
rc_t CC format_spot_name ( void *self,
const VXformInfo *info, int64_t row_id, VRowResult *rslt,
uint32_t argc, const VRowData argv [] )
{
rc_t rc;
char *name;
uint32_t elem_count;
KDataBuffer *dst = rslt -> data;
if ( dst -> elem_bits != 8 )
{
rc = KDataBufferCast ( dst, dst, 8, true );
if ( rc != 0 )
return rc;
}
/* check for NAME_FMT */
if ( argv [ 0 ] . u. data . elem_count != 0 )
{
size_t num_writ;
char sname[1024]; /** name on stack **/
const char *name_fmt = ((char*)argv[0].u.data.base) + argv[0].u.data.first_elem;
uint32_t i, j, x, y;
const uint32_t fmt_size = argv [ 0 ] . u . data . elem_count;
/* the coordinates to substitute */
x = ( ( const int32_t* ) argv [ 1 ] . u . data . base )
[ argv [ 1 ] . u . data . first_elem ];
y = ( ( const int32_t* ) argv [ 2 ] . u . data . base )
[ argv [ 2 ] . u . data . first_elem ];
for ( i=j=0; i < fmt_size -1;) {
if( name_fmt [ i ] == '$' ) {
switch( name_fmt [ i+1 ]) {
case 'x':
case 'X':
if( j > sizeof(sname) - 11) {
return RC ( rcXF, rcFunction, rcDecoding, rcBuffer, rcInsufficient );
} else {
i+=2;
if( i < fmt_size -1 && name_fmt [ i ] == '%' && isdigit(name_fmt [ i+1 ])) {
x += 24*1024*(name_fmt [ i+1 ]-'0');
i+=2;
}
j+=sprintf(sname+j,"%d",x);
}
break;
case 'y':
case 'Y':
if( j > sizeof(sname) - 11) {
return RC ( rcXF, rcFunction, rcDecoding, rcBuffer, rcInsufficient );
} else {
i+=2;
if( i < fmt_size -1 && name_fmt [ i ] == '%' && isdigit(name_fmt [ i+1 ])) {
y += 24*1024*(name_fmt [ i+1 ]-'0');
i+=2;
}
j+=sprintf(sname+j,"%d",y);
}
break;
case 'q':
case 'Q':
if( j > sizeof(sname) - 5) {
return RC ( rcXF, rcFunction, rcDecoding, rcBuffer, rcInsufficient );
} else if( y > 0xFFF) {
return RC ( rcXF, rcFunction, rcDecoding, rcRange, rcTooBig);
} else {
uint32_t q= x << 12 | y;
sname[j+4]=Q_ALPHABET[q % 36];
q /= 36;
sname[j+3]=Q_ALPHABET[q % 36];
q /= 36;
sname[j+2]=Q_ALPHABET[q % 36];
q /= 36;
sname[j+1]=Q_ALPHABET[q % 36];
q /= 36;
sname[j] =Q_ALPHABET[q];
j+=5;
i+=2;
}
break;
case 'r':
case 'R':
rc = string_printf ( & sname [ j ], sizeof sname - j, & num_writ, "%ld", row_id );
assert ( rc == 0 );
j += ( uint32_t ) num_writ;
i+=2;
break;
default:
sname[j++]=name_fmt[i++];
break;
}
} else {
if(j < sizeof(sname)) {
sname[j++]=name_fmt[i++];
} else {
return RC ( rcXF, rcFunction, rcDecoding, rcBuffer, rcInsufficient );
}
}
if( j > sizeof(sname)-1) {
return RC ( rcXF, rcFunction, rcDecoding, rcBuffer, rcInsufficient );
}
}
if(i==fmt_size -1) {
if(j < sizeof(sname)) {
sname[j++]=name_fmt[i++];
} else {
return RC ( rcXF, rcFunction, rcDecoding, rcBuffer, rcInsufficient );
}
}
/* j is our new element count **/
elem_count = j;
/* resize output */
if ( dst -> elem_count <= elem_count )
{
rc = KDataBufferResize ( dst, elem_count);
if ( rc != 0 )
return rc;
}
/* the output name */
name = dst -> base;
memcpy ( name, sname, elem_count );
rslt -> elem_count = elem_count;
return 0;
}
/* check for NAME */
if( argc == 4 && argv[3].u.data.elem_count != 0 ) {
const char *sname = ((char*)argv[3].u.data.base) + argv[3].u.data.first_elem;
/* output size */
elem_count = argv[3].u.data.elem_count;
/* resize output */
if( dst -> elem_count <= elem_count ) {
rc = KDataBufferResize( dst, elem_count + 1 );
if( rc != 0 )
return rc;
}
name = dst->base;
memcpy(dst->base, sname, elem_count);
rslt->elem_count = elem_count;
name[elem_count] = 0;
return 0;
}
/* spot has no name */
return RC ( rcSRA, rcColumn, rcReading, rcRow, rcNull );
}
static
rc_t CC format_spot_name_no_coord ( void *self,
const VXformInfo *info, int64_t row_id, VRowResult *rslt,
uint32_t argc, const VRowData argv [] )
{
rc_t rc;
char *name;
uint32_t elem_count;
KDataBuffer *dst = rslt -> data;
if ( dst -> elem_bits != 8 )
{
rc = KDataBufferCast ( dst, dst, 8, true );
if ( rc != 0 )
return rc;
}
/* check for NAME_FMT */
if ( argv [ 0 ] . u. data . elem_count != 0 )
{
size_t num_writ;
char sname[1024]; /** name on stack **/
const char *name_fmt = ((char*)argv[0].u.data.base) + argv[0].u.data.first_elem;
uint32_t i, j;
const uint32_t fmt_size = argv [ 0 ] . u . data . elem_count;
for ( i=j=0; i < fmt_size -1;) {
if( name_fmt [ i ] == '$' ) {
switch( name_fmt [ i+1 ]) {
case 'r':
case 'R':
rc = string_printf ( & sname [ j ], sizeof sname - j, & num_writ, "%ld", row_id );
assert ( rc == 0 );
j += ( uint32_t ) num_writ;
i+=2;
break;
default:
sname[j++]=name_fmt[i++];
break;
}
} else {
if(j < sizeof(sname)) {
sname[j++]=name_fmt[i++];
} else {
return RC ( rcXF, rcFunction, rcDecoding, rcBuffer, rcInsufficient );
}
}
if( j > sizeof(sname)-1) {
return RC ( rcXF, rcFunction, rcDecoding, rcBuffer, rcInsufficient );
}
}
if(i==fmt_size -1) {
if(j < sizeof(sname)) {
sname[j++]=name_fmt[i++];
} else {
return RC ( rcXF, rcFunction, rcDecoding, rcBuffer, rcInsufficient );
}
}
/* j is our new element count **/
elem_count = j;
/* resize output */
if ( dst -> elem_count <= elem_count )
{
rc = KDataBufferResize ( dst, elem_count);
if ( rc != 0 )
return rc;
}
/* the output name */
name = dst -> base;
memcpy ( name, sname, elem_count );
rslt -> elem_count = elem_count;
return 0;
}
/* check for NAME */
if( argc == 2 && argv[1].u.data.elem_count != 0 ) {
const char *sname = ((char*)argv[1].u.data.base) + argv[1].u.data.first_elem;
/* output size */
elem_count = argv[1].u.data.elem_count;
/* resize output */
if( dst -> elem_count <= elem_count ) {
rc = KDataBufferResize( dst, elem_count + 1 );
if( rc != 0 )
return rc;
}
name = dst->base;
memcpy(dst->base, sname, elem_count);
rslt->elem_count = elem_count;
name[elem_count] = 0;
return 0;
}
/* spot has no name */
return RC ( rcSRA, rcColumn, rcReading, rcRow, rcNull );
}
static
rc_t CC illumina_rewrite_spot_name ( void *data, const VXformInfo *info, int64_t row_id,
VRowResult *rslt, uint32_t argc, const VRowData argv [] )
{
rc_t rc;
char buffer [ 64];
uint32_t coord_len;
uint32_t prefix_len;
unsigned int a, b, c, d;
KDataBuffer *dst = rslt -> data;
const char *prefix;
const char *skey = argv [ 0 ] . u . data . base;
uint64_t i, j, count = argv [ 0 ] . u . data . elem_count;
skey += argv [ 0 ] . u . data . first_elem;
/* find last hex portion */
for ( i = count; i > 0; )
{
if ( ! isxdigit ( skey [ -- i ] ) )
break;
}
if ( count - i < ( SLX_COORD_LEN - 1 ) )
{
const char *end = skey + count;
/* new format */
for ( d = 0, j = 0, i = count; i > 0; )
{
if ( ! isdigit ( skey [ -- i ] ) )
{
j = i + 1;
break;
}
}
d = non_braindead_atoi ( & skey [ j ], end );
for ( c = 0, j = 0; i > 0; )
{
if ( ! isdigit ( skey [ -- i ] ) )
{
j = i + 1;
break;
}
}
c = atoi ( & skey [ j ] );
for ( b = 0, j = 0; i > 0; )
{
if ( ! isdigit ( skey [ -- i ] ) )
{
j = i + 1;
break;
}
}
b = atoi ( & skey [ j ] );
for ( a = 0, j = 0; i > 0; )
{
if ( ! isdigit ( skey [ -- i ] ) )
{
j = i + 1;
break;
}
}
a = atoi ( & skey [ j ] );
if ( j > 0 )
{
if ( i > 0 )
-- i;
while ( isalpha ( skey [ i ] ) )
++ i;
}
}
else
{
a = scan_hex ( skey, 1 );
b = scan_hex ( & skey [ 1 ], 3 );
c = scan_hex ( & skey [ 4 ], 3 );
d = scan_hex ( & skey [ 7 ], 3 );
if ( count > SLX_COORD_LEN )
{
i = count - SLX_COORD_LEN;
}
}
/* generate coordinates */
coord_len = sprintf ( buffer, ":%d:%d:%d:%d", a, b, c, d );
/* get size of prefix */
if ( argc == 1 )
{
prefix = "";
prefix_len = 0;
}
else
{
prefix = argv [ 1 ] . u . data . base;
assert(argv [ 1 ] . u . data . elem_count >> 32 == 0);
prefix_len = (uint32_t)argv [ 1 ] . u . data . elem_count;
prefix += argv [ 1 ] . u . data . first_elem;
}
/* resize output buffer for prefix, name stuff, coordinates */
if ( dst -> elem_bits != 8 )
{
rc = KDataBufferCast ( dst, dst, 8, true );
if ( rc != 0 )
return rc;
}
rc = KDataBufferResize ( dst, prefix_len + i + coord_len + 1 );
if ( rc != 0 )
return rc;
/* copy in prefix, name prefix, coordinates */
rslt -> elem_count = sprintf ( dst -> base, "%.*s%.*s%s"
, ( int ) prefix_len, prefix
, ( int ) i, skey
, buffer );
return 0;
}
