-/* PSPP - computes sample statistics.
- Copyright (C) 2006 Free Software Foundation, Inc.
- Written by Ben Pfaff <blp@gnu.org>.
+/* PSPP - a program for statistical analysis.
+ Copyright (C) 2006, 2010 Free Software Foundation, Inc.
- This program is free software; you can redistribute it and/or
- modify it under the terms of the GNU General Public License as
- published by the Free Software Foundation; either version 2 of the
- License, or (at your option) any later version.
+ This program is free software: you can redistribute it and/or modify
+ it under the terms of the GNU General Public License as published by
+ the Free Software Foundation, either version 3 of the License, or
+ (at your option) any later version.
- This program is distributed in the hope that it will be useful, but
- WITHOUT ANY WARRANTY; without even the implied warranty of
- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- General Public License for more details.
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
You should have received a copy of the GNU General Public License
- along with this program; if not, write to the Free Software
- Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
- 02110-1301, USA. */
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
#include <config.h>
#include <libpspp/pool.h>
#include <libpspp/str.h>
+#include <data/format.h>
+
#include "xalloc.h"
#include "xsize.h"
/* Extensions to the format specifiers used only for
placement. */
-enum
+enum
{
PRS_TYPE_T = SCHAR_MAX - 3, /* Tab to absolute column. */
PRS_TYPE_X, /* Skip columns. */
interpreting *FORMATS, POOL may be destroyed. */
bool
parse_var_placements (struct lexer *lexer, struct pool *pool, size_t var_cnt, bool for_input,
- struct fmt_spec **formats, size_t *format_cnt)
+ struct fmt_spec **formats, size_t *format_cnt)
{
assert (var_cnt > 0);
if (lex_is_number (lexer))
return fixed_parse_columns (lexer, pool, var_cnt, for_input, formats, format_cnt);
- else if (lex_match (lexer, '('))
+ else if (lex_match (lexer, T_LPAREN))
{
size_t assignment_cnt;
size_t i;
if (!fixed_parse_fortran (lexer, pool, for_input, formats, format_cnt))
- return false;
+ return false;
assignment_cnt = 0;
for (i = 0; i < *format_cnt; i++)
if (assignment_cnt != var_cnt)
{
- msg (SE, _("Number of variables specified (%d) "
- "differs from number of variable formats (%d)."),
- (int) var_cnt, (int) assignment_cnt);
+ msg (SE, _("Number of variables specified (%zu) "
+ "differs from number of variable formats (%zu)."),
+ var_cnt, assignment_cnt);
return false;
}
int fc, lc;
size_t i;
- if ( !parse_column_range (lexer, &fc, &lc, NULL) )
+ if ( !parse_column_range (lexer, 1, &fc, &lc, NULL) )
return false;
- /* Divide columns evenly. */
+ /* Divide columns evenly. */
format.w = (lc - fc + 1) / var_cnt;
if ((lc - fc + 1) % var_cnt)
{
msg (SE, _("The %d columns %d-%d "
- "can't be evenly divided into %d fields."),
+ "can't be evenly divided into %zu fields."),
lc - fc + 1, fc, lc, var_cnt);
return false;
}
/* Format specifier. */
- if (lex_match (lexer, '('))
+ if (lex_match (lexer, T_LPAREN))
{
/* Get format type. */
if (lex_token (lexer) == T_ID)
{
if (!parse_format_specifier_name (lexer, &format.type))
return false;
- lex_match (lexer, ',');
+ lex_match (lexer, T_COMMA);
}
else
format.type = FMT_F;
else
format.d = 0;
- if (!lex_force_match (lexer, ')'))
+ if (!lex_force_match (lexer, T_RPAREN))
return false;
}
else
size_t formats_used = 0;
*formats = NULL;
- while (!lex_match (lexer, ')'))
+ while (!lex_match (lexer, T_RPAREN))
{
struct fmt_spec f;
struct fmt_spec *new_formats;
size_t new_format_cnt;
size_t count;
size_t formats_needed;
-
+
/* Parse count. */
if (lex_is_integer (lexer))
{
count = 1;
/* Parse format specifier. */
- if (lex_match (lexer, '('))
+ if (lex_match (lexer, T_LPAREN))
{
/* Call ourselves recursively to handle parentheses. */
if (!fixed_parse_fortran (lexer, pool, for_input,
{
new_formats = &f;
new_format_cnt = 1;
- if (lex_match (lexer, '/'))
+ if (lex_match (lexer, T_SLASH))
f.type = PRS_TYPE_NEW_REC;
else
{
char type[FMT_TYPE_LEN_MAX + 1];
-
+
if (!parse_abstract_format_specifier (lexer, type, &f.w, &f.d))
return false;
- if (!strcasecmp (type, "T"))
+ if (!strcasecmp (type, "T"))
f.type = PRS_TYPE_T;
- else if (!strcasecmp (type, "X"))
+ else if (!strcasecmp (type, "X"))
{
f.type = PRS_TYPE_X;
f.w = count;
count = 1;
}
- else
+ else
{
- if (!fmt_from_name (type, &f.type))
+ if (!fmt_from_name (type, &f.type))
{
- msg (SE, _("Unknown format type \"%s\"."), type);
+ msg (SE, _("Unknown format type `%s'."), type);
return false;
}
if (!fmt_check (&f, for_input))
return false;
}
- }
+ }
}
/* Add COUNT copies of the NEW_FORMAT_CNT formats in
sizeof *formats)))
xalloc_die ();
formats_needed = count * new_format_cnt;
- if (formats_used + formats_needed > formats_allocated)
+ if (formats_used + formats_needed > formats_allocated)
{
formats_allocated = formats_used + formats_needed;
*formats = pool_2nrealloc (pool, *formats, &formats_allocated,
sizeof **formats);
}
- for (; count > 0; count--)
+ for (; count > 0; count--)
{
memcpy (&(*formats)[formats_used], new_formats,
sizeof **formats * new_format_cnt);
formats_used += new_format_cnt;
}
- lex_match (lexer, ',');
+ lex_match (lexer, T_COMMA);
}
*format_cnt = formats_used;
without any side effects. */
bool
execute_placement_format (const struct fmt_spec *format,
- int *record, int *column)
+ int *record, int *column)
{
- switch (format->type)
+ switch (format->type)
{
case PRS_TYPE_X:
*column += format->w;
return true;
-
+
case PRS_TYPE_T:
*column = format->w;
return true;
-
+
case PRS_TYPE_NEW_REC:
(*record)++;
*column = 1;
}
}
+/* Parses a BASE-based column using LEXER. Returns true and
+ stores a 1-based column number into *COLUMN if successful,
+ otherwise emits an error message and returns false. */
+static bool
+parse_column (int value, int base, int *column)
+{
+ assert (base == 0 || base == 1);
+ *column = value - base + 1;
+ if (*column < 1)
+ {
+ if (base == 1)
+ msg (SE, _("Column positions for fields must be positive."));
+ else
+ msg (SE, _("Column positions for fields must not be negative."));
+ return false;
+ }
+ return true;
+}
+
/* Parse a column or a range of columns, specified as a single
- integer or two integer delimited by a dash. Stores the range
+ integer or two integers delimited by a dash. Stores the range
in *FIRST_COLUMN and *LAST_COLUMN. (If only a single integer
is given, it is stored in both.) If RANGE_SPECIFIED is
non-null, then *RANGE_SPECIFIED is set to true if the syntax
contained a dash, false otherwise. Returns true if
successful, false if the syntax was invalid or the values
- specified did not make sense. */
+ specified did not make sense.
+
+ If BASE is 0, zero-based column numbers are parsed; if BASE is
+ 1, 1-based column numbers are parsed. Regardless of BASE, the
+ values stored in *FIRST_COLUMN and *LAST_COLUMN are
+ 1-based. */
bool
-parse_column_range (struct lexer *lexer, int *first_column, int *last_column,
- bool *range_specified)
+parse_column_range (struct lexer *lexer, int base,
+ int *first_column, int *last_column,
+ bool *range_specified)
{
/* First column. */
- if (!lex_force_int (lexer))
+ if (!lex_force_int (lexer)
+ || !parse_column (lex_integer (lexer), base, first_column))
return false;
- *first_column = lex_integer (lexer);
- if (*first_column < 1)
- {
- msg (SE, _("Column positions for fields must be positive."));
- return false;
- }
lex_get (lexer);
/* Last column. */
- lex_negative_to_dash (lexer);
- if (lex_match (lexer, '-'))
+ if (lex_is_integer (lexer) && lex_integer (lexer) < 0)
{
- if (!lex_force_int (lexer))
- return false;
- *last_column = lex_integer (lexer);
- if (*last_column < 1)
- {
- msg (SE, _("Column positions for fields must be positive."));
- return false;
- }
- else if (*last_column < *first_column)
+ if (!parse_column (-lex_integer (lexer), base, last_column))
+ return false;
+ lex_get (lexer);
+
+ if (*last_column < *first_column)
{
msg (SE, _("The ending column for a field must be "
"greater than the starting column."));
if (range_specified)
*range_specified = true;
- lex_get (lexer);
}
- else
+ else
{
*last_column = *first_column;
if (range_specified)
Returns true if successful, false on syntax error. */
bool
-parse_record_placement (struct lexer *lexer, int *record, int *column)
+parse_record_placement (struct lexer *lexer, int *record, int *column)
{
- while (lex_match (lexer, '/'))
+ while (lex_match (lexer, T_SLASH))
{
if (lex_is_integer (lexer))
{
*column = 1;
}
assert (*record >= 1);
-
+
return true;
}