+2007-08-12 Ben Pfaff <blp@gnu.org>
+
+ * dictionary.c (dict_compact_values): Don't delete scratch
+ variables as well as compacting case indexes. Update all callers.
+ (dict_get_compacted_value_cnt): Rename dict_count_values and
+ change interface. Update all callers.
+ (dict_get_compacted_value_cnt): Remove.
+ (dict_compacting_would_shrink): Remove.
+ (dict_compacting_would_change): Remove.
+ (dict_make_compactor): Add new parameter. Update all callers.
+
+ * procedure.c (proc_casereader_read): Use casewriter_get_value_cnt
+ instead of dict_count_values, changing an O(N) operation into
+ O(1).
+
2007-08-12 Ben Pfaff <blp@gnu.org>
* casereader.c (casereader_read): Don't require cases read by a
return sizeof (union value) * dict_get_next_value_idx (d);
}
-/* Deletes scratch variables in dictionary D and reassigns values
- so that fragmentation is eliminated. */
+/* Reassigns values in dictionary D so that fragmentation is
+ eliminated. */
void
dict_compact_values (struct dictionary *d)
{
size_t i;
d->next_value_idx = 0;
- for (i = 0; i < d->var_cnt; )
+ for (i = 0; i < d->var_cnt; i++)
{
struct variable *v = d->var[i];
-
- if (dict_class_from_id (var_get_name (v)) != DC_SCRATCH)
- {
- set_var_case_index (v, d->next_value_idx);
- d->next_value_idx += var_get_value_cnt (v);
- i++;
- }
- else
- dict_delete_var (d, v);
+ set_var_case_index (v, d->next_value_idx);
+ d->next_value_idx += var_get_value_cnt (v);
}
}
-
/*
Reassigns case indices for D, increasing each index above START by
the value PADDING.
}
-/* Returns the number of values that would be used by a case if
- dict_compact_values() were called. */
+/* Returns the number of values occupied by the variables in
+ dictionary D. All variables are considered if EXCLUDE_CLASSES
+ is 0, or it may contain one or more of (1u << DC_ORDINARY),
+ (1u << DC_SYSTEM), or (1u << DC_SCRATCH) to exclude the
+ corresponding type of variable.
+
+ The return value may be less than the number of values in one
+ of dictionary D's cases (as returned by
+ dict_get_next_value_idx) even if E is 0, because there may be
+ gaps in D's cases due to deleted variables. */
size_t
-dict_get_compacted_value_cnt (const struct dictionary *d)
+dict_count_values (const struct dictionary *d, unsigned int exclude_classes)
{
size_t i;
size_t cnt;
- cnt = 0;
- for (i = 0; i < d->var_cnt; i++)
- if (dict_class_from_id (var_get_name (d->var[i])) != DC_SCRATCH)
- cnt += var_get_value_cnt (d->var[i]);
- return cnt;
-}
-
-/* Creates and returns an array mapping from a dictionary index
- to the case index that the corresponding variable will have
- after calling dict_compact_values(). Scratch variables
- receive -1 for case index because dict_compact_values() will
- delete them. */
-int *
-dict_get_compacted_dict_index_to_case_index (const struct dictionary *d)
-{
- size_t i;
- size_t next_value_idx;
- int *map;
+ assert ((exclude_classes & ~((1u << DC_ORDINARY)
+ | (1u << DC_SYSTEM)
+ | (1u << DC_SCRATCH))) == 0);
- map = xnmalloc (d->var_cnt, sizeof *map);
- next_value_idx = 0;
+ cnt = 0;
for (i = 0; i < d->var_cnt; i++)
{
- struct variable *v = d->var[i];
-
- if (dict_class_from_id (var_get_name (v)) != DC_SCRATCH)
- {
- map[i] = next_value_idx;
- next_value_idx += var_get_value_cnt (v);
- }
- else
- map[i] = -1;
- }
- return map;
-}
-
-/* Returns true if a case for dictionary D would be smaller after
- compacting, false otherwise. Compacting a case eliminates
- "holes" between values and after the last value. Holes are
- created by deleting variables (or by scratch variables).
-
- The return value may differ from whether compacting a case
- from dictionary D would *change* the case: compacting could
- rearrange values even if it didn't reduce space
- requirements. */
-bool
-dict_compacting_would_shrink (const struct dictionary *d)
-{
- return dict_get_compacted_value_cnt (d) < dict_get_next_value_idx (d);
-}
-
-/* Returns true if a case for dictionary D would change after
- compacting, false otherwise. Compacting a case eliminates
- "holes" between values and after the last value. Holes are
- created by deleting variables (or by scratch variables).
-
- The return value may differ from whether compacting a case
- from dictionary D would *shrink* the case: compacting could
- rearrange values without reducing space requirements. */
-bool
-dict_compacting_would_change (const struct dictionary *d)
-{
- size_t case_idx;
- size_t i;
-
- case_idx = 0;
- for (i = 0; i < dict_get_var_cnt (d); i++)
- {
- struct variable *v = dict_get_var (d, i);
- if (var_get_case_index (v) != case_idx)
- return true;
- case_idx += var_get_value_cnt (v);
+ enum dict_class class = dict_class_from_id (var_get_name (d->var[i]));
+ if (!(exclude_classes & (1u << class)))
+ cnt += var_get_value_cnt (d->var[i]);
}
- return false;
+ return cnt;
}
\f
/* How to copy a contiguous range of values between cases. */
compact cases for dictionary D.
Compacting a case eliminates "holes" between values and after
- the last value. Holes are created by deleting variables (or
- by scratch variables). */
+ the last value. (Holes are created by deleting variables.)
+
+ All variables are compacted if EXCLUDE_CLASSES is 0, or it may
+ contain one or more of (1u << DC_ORDINARY), (1u << DC_SYSTEM),
+ or (1u << DC_SCRATCH) to cause the corresponding type of
+ variable to be deleted during compaction. */
struct dict_compactor *
-dict_make_compactor (const struct dictionary *d)
+dict_make_compactor (const struct dictionary *d, unsigned int exclude_classes)
{
struct dict_compactor *compactor;
struct copy_map *map;
size_t value_idx;
size_t i;
+ assert ((exclude_classes & ~((1u << DC_ORDINARY)
+ | (1u << DC_SYSTEM)
+ | (1u << DC_SCRATCH))) == 0);
+
compactor = xmalloc (sizeof *compactor);
compactor->maps = NULL;
compactor->map_cnt = 0;
for (i = 0; i < d->var_cnt; i++)
{
struct variable *v = d->var[i];
-
- if (dict_class_from_id (var_get_name (v)) == DC_SCRATCH)
+ enum dict_class class = dict_class_from_id (var_get_name (v));
+ if (exclude_classes & (1u << class))
continue;
+
if (map != NULL && map->src_idx + map->cnt == var_get_case_index (v))
map->cnt += var_get_value_cnt (v);
else
COMPACTOR.
Compacting a case eliminates "holes" between values and after
- the last value. Holes are created by deleting variables (or
- by scratch variables). */
+ the last value. (Holes are created by deleting variables.) */
void
dict_compactor_compact (const struct dict_compactor *compactor,
struct ccase *dst, const struct ccase *src)
int dict_get_next_value_idx (const struct dictionary *);
size_t dict_get_case_size (const struct dictionary *);
+size_t dict_count_values (const struct dictionary *,
+ unsigned int exclude_classes);
void dict_compact_values (struct dictionary *);
-size_t dict_get_compacted_value_cnt (const struct dictionary *);
-int *dict_get_compacted_dict_index_to_case_index (const struct dictionary *);
-bool dict_compacting_would_shrink (const struct dictionary *);
-bool dict_compacting_would_change (const struct dictionary *);
-struct dict_compactor *dict_make_compactor (const struct dictionary *);
+struct dict_compactor *dict_make_compactor (const struct dictionary *,
+ unsigned int exclude_classes);
void dict_compactor_compact (const struct dict_compactor *,
struct ccase *, const struct ccase *);
void dict_compactor_destroy (struct dict_compactor *);
/* Prepare sink. */
if (!ds->discard_output)
{
- ds->compactor = (dict_compacting_would_shrink (ds->permanent_dict)
- ? dict_make_compactor (ds->permanent_dict)
+ struct dictionary *pd = ds->permanent_dict;
+ size_t compacted_value_cnt = dict_count_values (pd, 1u << DC_SCRATCH);
+ bool should_compact = compacted_value_cnt < dict_get_next_value_idx (pd);
+ ds->compactor = (should_compact
+ ? dict_make_compactor (pd, 1u << DC_SCRATCH)
: NULL);
- ds->sink = autopaging_writer_create (dict_get_compacted_value_cnt (
- ds->permanent_dict));
+ ds->sink = autopaging_writer_create (compacted_value_cnt);
}
else
{
struct ccase tmp;
if (ds->compactor != NULL)
{
- case_create (&tmp, dict_get_compacted_value_cnt (ds->dict));
+ case_create (&tmp, casewriter_get_value_cnt (ds->sink));
dict_compactor_compact (ds->compactor, &tmp, c);
}
else
if (ds->compactor != NULL)
{
dict_compactor_destroy (ds->compactor);
- dict_compact_values (ds->dict);
ds->compactor = NULL;
+
+ dict_delete_scratch_vars (ds->dict);
+ dict_compact_values (ds->dict);
}
/* Old data sink becomes new data source. */
#include <data/dictionary.h>
#include <data/file-handle-def.h>
#include <data/scratch-handle.h>
+#include <data/variable.h>
#include <libpspp/compiler.h>
#include <libpspp/taint.h>
/* Copy the dictionary and compact if needed. */
scratch_dict = dict_clone (dictionary);
- if (dict_compacting_would_shrink (scratch_dict))
+ dict_delete_scratch_vars (scratch_dict);
+ if (dict_count_values (scratch_dict, 0)
+ < dict_get_next_value_idx (scratch_dict))
{
- compactor = dict_make_compactor (scratch_dict);
+ compactor = dict_make_compactor (scratch_dict, 0);
dict_compact_values (scratch_dict);
}
else
goto error;
}
+ dict_delete_scratch_vars (dict);
dict_compact_values (dict);
if (fh_get_referent (handle) == FH_REF_FILE)
|| !create_flag_var ("LAST", last_name, mtf.dict, &mtf.last))
goto error;
+ dict_delete_scratch_vars (mtf.dict);
dict_compact_values (mtf.dict);
mtf.output = autopaging_writer_create (dict_get_next_value_idx (mtf.dict));
taint = taint_clone (casewriter_get_taint (mtf.output));
+2007-08-12 Ben Pfaff <blp@gnu.org>
+
+ * flip.c: Drop use of dict_get_compacted_dict_index_to_case_index
+ and just use the ordinary case indexes. There seemed to be no
+ reason for the former method.
+
2007-08-03 Ben Pfaff <blp@gnu.org>
* rank.q (rank_cmd): Instead of sorting by SPLIT FILE vars, group
{
struct pool *pool; /* Pool containing FLIP data. */
const struct variable **var; /* Variables to transpose. */
- int *idx_to_fv; /* var[]->index to compacted sink case fv. */
size_t var_cnt; /* Number of elements in `var'. */
int case_cnt; /* Pre-flip case count. */
flip = pool_create_container (struct flip_pgm, pool);
flip->var = NULL;
- flip->idx_to_fv = dict_get_compacted_dict_index_to_case_index (dict);
- pool_register (flip->pool, free, flip->idx_to_fv);
flip->var_cnt = 0;
flip->case_cnt = 0;
flip->new_names = NULL;
flip->case_cnt = 1;
/* Read the active file into a flip_sink. */
- proc_make_temporary_transformations_permanent (ds);
proc_discard_output (ds);
input = proc_open (ds);
if (flip->new_names != NULL)
{
struct varname *v = pool_alloc (flip->pool, sizeof *v);
- int fv = flip->idx_to_fv[var_get_dict_index (flip->new_names)];
v->next = NULL;
if (var_is_numeric (flip->new_names))
{
- double f = case_num_idx (c, fv);
+ double f = case_num (c, flip->new_names);
if (f == SYSMIS)
strcpy (v->name, "VSYSMIS");
else
{
int width = MIN (var_get_width (flip->new_names), MAX_SHORT_STRING);
- memcpy (v->name, case_str_idx (c, fv), width);
+ memcpy (v->name, case_str (c, flip->new_names), width);
v->name[width] = 0;
}
/* Write to external file. */
for (i = 0; i < flip->var_cnt; i++)
{
- double out;
-
- if (var_is_numeric (flip->var[i]))
- {
- int fv = flip->idx_to_fv[var_get_dict_index (flip->var[i])];
- out = case_num_idx (c, fv);
- }
- else
- out = SYSMIS;
+ const struct variable *v = flip->var[i];
+ double out = var_is_numeric (v) ? case_num (c, v) : SYSMIS;
fwrite (&out, sizeof out, 1, flip->file);
}
return true;
+2007-08-12 Ben Pfaff <blp@gnu.org>
+
+ * psppire-dict.c (psppire_dict_dump): Don't use
+ dict_get_compacted_dict_index_to_case_index, as that function has
+ been deleted.
+
2007-08-13 John Darrington <john@darrington.wattle.id.au>
* psppire-case-file.c (psppire_case_file_append_case):
gint i;
const struct dictionary *d = dict->dict;
- int *map = dict_get_compacted_dict_index_to_case_index (d);
-
for (i = 0; i < dict_get_var_cnt (d); ++i)
{
const struct variable *v = psppire_dict_get_variable (dict, i);
int di = var_get_dict_index (v);
- g_print ("\"%s\" idx=%d, fv=%d(%d), size=%d\n",
+ g_print ("\"%s\" idx=%d, fv=%d, size=%d\n",
var_get_name(v),
di,
var_get_case_index(v),
- map[di],
value_cnt_from_width(var_get_width(v)));
}
projects / pspp-builds.git / commitdiff