summaryrefslogtreecommitdiffstats
path: root/src/H5Tconv.c
diff options
context:
space:
mode:
Diffstat (limited to 'src/H5Tconv.c')
-rw-r--r--src/H5Tconv.c67
1 files changed, 28 insertions, 39 deletions
diff --git a/src/H5Tconv.c b/src/H5Tconv.c
index 803a6da..563de65 100644
--- a/src/H5Tconv.c
+++ b/src/H5Tconv.c
@@ -2962,7 +2962,7 @@ H5T__conv_enum_numeric(hid_t src_id, hid_t dst_id, H5T_cdata_t *cdata, size_t ne
src_parent = src->shared->parent;
if(NULL == (tpath = H5T_path_find(src_parent, dst))) {
- HGOTO_ERROR(H5E_DATASET, H5E_UNSUPPORTED, FAIL, "unable to convert between src and dest datatype")
+ HGOTO_ERROR(H5E_DATASET, H5E_UNSUPPORTED, FAIL, "unable to convert between src and dest datatype")
} else if(!H5T_path_noop(tpath)) {
if((src_parent_id = H5I_register(H5I_DATATYPE, H5T_copy(src_parent, H5T_COPY_ALL), FALSE)) < 0)
HGOTO_ERROR(H5E_DATASET, H5E_CANTREGISTER, FAIL, "unable to register types for conversion")
@@ -8295,14 +8295,6 @@ H5_GCC_DIAG_ON(float-equal)
* Programmer: Raymond Lu
* Wednesday, Jan 21, 2004
*
- * Raymond Lu
- * Wednesday, April 21, 2004
- * There is a new design for exception handling like overflow,
- * which is passed in as a transfer property.
- *
- * Raymond Lu
- * Monday, March 13, 2006
- * Added support for VAX floating-point types.
*-------------------------------------------------------------------------
*/
herr_t
@@ -9008,8 +9000,7 @@ H5T__conv_i_f(hid_t src_id, hid_t dst_id, H5T_cdata_t *cdata, size_t nelmts,
}
#endif
- /*
- * Put the data in little endian order so our loops aren't so
+ /* Put the data in little endian order so our loops aren't so
* complicated. We'll do all the conversion stuff assuming
* little endian and then we'll fix the order at the end.
*/
@@ -9022,26 +9013,23 @@ H5T__conv_i_f(hid_t src_id, hid_t dst_id, H5T_cdata_t *cdata, size_t nelmts,
}
}
- /*zero-set all destination bits*/
+ /* Zero-set all destination bits*/
H5T__bit_set (d, dst.offset, dst.prec, FALSE);
/* Copy source into a temporary buffer */
H5T__bit_copy(int_buf, (size_t)0, s, src.offset, src.prec);
- /*
- * Find the sign bit value of the source.
- */
+ /* Find the sign bit value of the source */
if(H5T_SGN_2 == src.u.i.sign)
sign = (size_t)H5T__bit_get_d(int_buf, src.prec - 1, (size_t)1);
- /*
- * What is the bit position(starting from 0 as first one) for the most significant
- * bit(MSB) of S which is set?
+ /* What is the bit position(starting from 0 as first one) for the most significant
+ * bit(MSB) of S which is set?
*/
if(H5T_SGN_2 == src.u.i.sign) {
sfirst = H5T__bit_find(int_buf, (size_t)0, src.prec - 1, H5T_BIT_MSB, TRUE);
if(sign && sfirst < 0)
- /* The case 0x80...00, which is negative with maximal value */
+ /* The case 0x80...00, which is negative with maximal value */
is_max_neg = 1;
} else if(H5T_SGN_NONE == src.u.i.sign)
sfirst = H5T__bit_find(int_buf, (size_t)0, src.prec, H5T_BIT_MSB, TRUE);
@@ -9050,9 +9038,7 @@ H5T__conv_i_f(hid_t src_id, hid_t dst_id, H5T_cdata_t *cdata, size_t nelmts,
if(!sign && sfirst < 0)
goto padding;
- /*
- * Convert source integer if it's negative
- */
+ /* Convert source integer if it's negative */
if(H5T_SGN_2 == src.u.i.sign && sign) {
if(!is_max_neg) {
/* Equivalent to ~(i - 1) */
@@ -9060,10 +9046,10 @@ H5T__conv_i_f(hid_t src_id, hid_t dst_id, H5T_cdata_t *cdata, size_t nelmts,
H5T__bit_neg(int_buf, (size_t)0, buf_size * 8);
sfirst = H5T__bit_find(int_buf, (size_t)0, src.prec - 1, H5T_BIT_MSB, TRUE);
} else {
- /* If it's maximal negative number 0x80...000, treat it as if it overflowed
- * (create a carry) to help conversion. i.e. a character type number 0x80
- * is treated as 0x100.
- */
+ /* If it's maximal negative number 0x80...000, treat it as if it overflowed
+ * (create a carry) to help conversion. i.e. a character type number 0x80
+ * is treated as 0x100.
+ */
sfirst = (ssize_t)(src.prec - 1);
is_max_neg = 0;
}
@@ -9072,7 +9058,8 @@ H5T__conv_i_f(hid_t src_id, hid_t dst_id, H5T_cdata_t *cdata, size_t nelmts,
/* Sign bit has been negated if bit vector isn't 0x80...00. Set all bits in front of
* sign bit to 0 in the temporary buffer because they're all negated from the previous
- * step. */
+ * step.
+ */
H5T__bit_set(int_buf, src.prec, (buf_size * 8) - src.prec, 0);
/* Set sign bit in destination */
@@ -9081,14 +9068,14 @@ H5T__conv_i_f(hid_t src_id, hid_t dst_id, H5T_cdata_t *cdata, size_t nelmts,
first = (size_t)sfirst;
- /*
- * Calculate the true destination exponent by adjusting according to
+ /* Calculate the true destination exponent by adjusting according to
* the destination exponent bias. Implied and non-implied normalization
* should be the same.
*/
if (H5T_NORM_NONE==dst.u.f.norm || H5T_NORM_IMPLIED==dst.u.f.norm) {
expo = first + dst.u.f.ebias;
- } else {
+ }
+ else {
HGOTO_ERROR(H5E_DATATYPE, H5E_CANTCONVERT, FAIL, "normalization method not implemented yet")
}
@@ -9096,13 +9083,14 @@ H5T__conv_i_f(hid_t src_id, hid_t dst_id, H5T_cdata_t *cdata, size_t nelmts,
if (H5T_NORM_IMPLIED==dst.u.f.norm) {
/* Imply first bit */
H5T__bit_set(int_buf, first, (size_t)1, 0);
- } else if (H5T_NORM_NONE==dst.u.f.norm) {
- first++;
+ }
+ else if (H5T_NORM_NONE==dst.u.f.norm) {
+ first++;
}
/* Roundup for mantissa */
if(first > dst.u.f.msize) {
- /* If the bit sequence is bigger than the mantissa part, there'll be some
+ /* If the bit sequence is bigger than the mantissa part, there'll be some
* precision loss. Let user's handler deal with the case if it's present
*/
if(cb_struct.func) {
@@ -9114,14 +9102,15 @@ H5T__conv_i_f(hid_t src_id, hid_t dst_id, H5T_cdata_t *cdata, size_t nelmts,
if(except_ret == H5T_CONV_HANDLED) {
reverse = FALSE;
goto padding;
- } else if(except_ret == H5T_CONV_ABORT)
+ }
+ else if(except_ret == H5T_CONV_ABORT)
HGOTO_ERROR(H5E_DATATYPE, H5E_CANTCONVERT, FAIL, "can't handle conversion exception")
- /* If user's exception handler does deal with it, we do it by dropping off the
- * extra bits at the end and do rounding. If we have .50...0(decimal) after radix
- * point, we do roundup when the least significant digit before radix is odd, we do
- * rounddown if it's even.
- */
+ /* If user's exception handler does deal with it, we do it by dropping off the
+ * extra bits at the end and do rounding. If we have .50...0(decimal) after radix
+ * point, we do roundup when the least significant digit before radix is odd, we do
+ * rounddown if it's even.
+ */
/* Check 1st dropoff bit, see if it's set. */
if(H5T__bit_get_d(int_buf, ((first - dst.u.f.msize) - 1), (size_t)1)) {
generated by cgit v0.12 at 2024-11-09 13:32:01 (GMT)