-/* PSPP - computes sample statistics.
- Copyright (C) 1997-9, 2000 Free Software Foundation, Inc.
- Written by Ben Pfaff <blp@gnu.org>.
+/* PSPP - a program for statistical analysis.
+ Copyright (C) 1997-9, 2000, 2006 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 <stdlib.h>
#include <data/case.h>
+#include <data/data-out.h>
#include <data/procedure.h>
#include <data/transformations.h>
#include <data/variable.h>
+#include <data/format.h>
#include <language/command.h>
#include <language/data-io/data-writer.h>
#include <language/data-io/file-handle.h>
#include <language/lexer/format-parser.h>
#include <language/lexer/lexer.h>
#include <language/lexer/variable-parser.h>
-#include <libpspp/alloc.h>
#include <libpspp/assertion.h>
#include <libpspp/compiler.h>
#include <libpspp/ll.h>
#include <output/manager.h>
#include <output/table.h>
+#include "xalloc.h"
+
#include "gettext.h"
#define _(msgid) gettext (msgid)
int first_column; /* 0-based first column. */
/* PRT_VAR only. */
- struct variable *var; /* Associated variable. */
+ const struct variable *var; /* Associated variable. */
struct fmt_spec format; /* Output spec. */
bool add_space; /* Add trailing space? */
+ bool sysmis_as_spaces; /* Output SYSMIS as spaces? */
/* PRT_LITERAL only. */
struct string string; /* String to output. */
};
static inline struct prt_out_spec *
-ll_to_prt_out_spec (struct ll *ll)
+ll_to_prt_out_spec (struct ll *ll)
{
return ll_data (ll, struct prt_out_spec, ll);
}
{
struct pool *pool; /* Stores related data. */
bool eject; /* Eject page before printing? */
- bool omit_new_lines; /* Omit new-line characters? */
+ bool include_prefix; /* Prefix lines with space? */
+ enum legacy_encoding encoding; /* Encoding to use for output. */
struct dfm_writer *writer; /* Output file, NULL=listing file. */
struct ll_list specs; /* List of struct prt_out_specs. */
size_t record_cnt; /* Number of records to write. */
enum which_formats
{
- PRINT,
+ PRINT,
WRITE
};
-static int internal_cmd_print (enum which_formats, bool eject);
+static int internal_cmd_print (struct lexer *, struct dataset *ds,
+ enum which_formats, bool eject);
static trns_proc_func print_trns_proc;
static trns_free_func print_trns_free;
-static bool parse_specs (struct pool *tmp_pool, struct print_trns *,
- enum which_formats);
+static bool parse_specs (struct lexer *, struct pool *tmp_pool, struct print_trns *,
+ struct dictionary *dict, enum which_formats);
static void dump_table (struct print_trns *, const struct file_handle *);
\f
/* Basic parsing. */
/* Parses PRINT command. */
int
-cmd_print (void)
+cmd_print (struct lexer *lexer, struct dataset *ds)
{
- return internal_cmd_print (PRINT, false);
+ return internal_cmd_print (lexer, ds, PRINT, false);
}
/* Parses PRINT EJECT command. */
int
-cmd_print_eject (void)
+cmd_print_eject (struct lexer *lexer, struct dataset *ds)
{
- return internal_cmd_print (PRINT, true);
+ return internal_cmd_print (lexer, ds, PRINT, true);
}
/* Parses WRITE command. */
int
-cmd_write (void)
+cmd_write (struct lexer *lexer, struct dataset *ds)
{
- return internal_cmd_print (WRITE, false);
+ return internal_cmd_print (lexer, ds, WRITE, false);
}
/* Parses the output commands. */
static int
-internal_cmd_print (enum which_formats which_formats, bool eject)
+internal_cmd_print (struct lexer *lexer, struct dataset *ds,
+ enum which_formats which_formats, bool eject)
{
bool print_table = 0;
struct print_trns *trns;
tmp_pool = pool_create_subpool (trns->pool);
/* Parse the command options. */
- while (token != '/')
+ while (lex_token (lexer) != '/' && lex_token (lexer) != '.')
{
- if (lex_match_id ("OUTFILE"))
+ if (lex_match_id (lexer, "OUTFILE"))
{
- lex_match ('=');
+ lex_match (lexer, '=');
- fh = fh_parse (FH_REF_FILE);
+ fh = fh_parse (lexer, FH_REF_FILE);
if (fh == NULL)
goto error;
}
- else if (lex_match_id ("RECORDS"))
+ else if (lex_match_id (lexer, "RECORDS"))
{
- lex_match ('=');
- lex_match ('(');
- if (!lex_force_int ())
+ lex_match (lexer, '=');
+ lex_match (lexer, '(');
+ if (!lex_force_int (lexer))
goto error;
- trns->record_cnt = lex_integer ();
- lex_get ();
- lex_match (')');
+ trns->record_cnt = lex_integer (lexer);
+ lex_get (lexer);
+ lex_match (lexer, ')');
}
- else if (lex_match_id ("TABLE"))
+ else if (lex_match_id (lexer, "TABLE"))
print_table = true;
- else if (lex_match_id ("NOTABLE"))
+ else if (lex_match_id (lexer, "NOTABLE"))
print_table = false;
else
{
- lex_error (_("expecting a valid subcommand"));
+ lex_error (lexer, _("expecting a valid subcommand"));
goto error;
}
}
+ /* When PRINT or PRINT EJECT writes to an external file, we
+ prefix each line with a space for compatibility. */
+ trns->include_prefix = which_formats == PRINT && fh != NULL;
+
/* Parse variables and strings. */
- if (!parse_specs (tmp_pool, trns, which_formats))
+ if (!parse_specs (lexer, tmp_pool, trns, dataset_dict (ds), which_formats))
goto error;
- if (lex_end_of_command () != CMD_SUCCESS)
+ if (lex_end_of_command (lexer) != CMD_SUCCESS)
goto error;
if (fh != NULL)
trns->writer = dfm_open_writer (fh);
if (trns->writer == NULL)
goto error;
-
- trns->omit_new_lines = (which_formats == WRITE
- && fh_get_mode (fh) == FH_MODE_BINARY);
+ trns->encoding = dfm_writer_get_legacy_encoding (trns->writer);
}
+ else
+ trns->encoding = LEGACY_NATIVE;
/* Output the variable table if requested. */
if (print_table)
dump_table (trns, fh);
/* Put the transformation in the queue. */
- add_transformation (current_dataset, print_trns_proc, print_trns_free, trns);
+ add_transformation (ds, print_trns_proc, print_trns_free, trns);
pool_destroy (tmp_pool);
+ fh_unref (fh);
return CMD_SUCCESS;
error:
print_trns_free (trns);
+ fh_unref (fh);
return CMD_FAILURE;
}
\f
-static bool parse_string_argument (struct print_trns *,
+static bool parse_string_argument (struct lexer *, struct print_trns *,
int record, int *column);
-static bool parse_variable_argument (struct print_trns *,
+static bool parse_variable_argument (struct lexer *, const struct dictionary *,
+ struct print_trns *,
struct pool *tmp_pool,
int *record, int *column,
enum which_formats);
PRINT, PRINT EJECT, or WRITE command into the prt structure.
Returns success. */
static bool
-parse_specs (struct pool *tmp_pool, struct print_trns *trns,
+parse_specs (struct lexer *lexer, struct pool *tmp_pool, struct print_trns *trns,
+ struct dictionary *dict,
enum which_formats which_formats)
{
int record = 0;
int column = 1;
- while (token != '.')
+ if (lex_token (lexer) == '.')
+ {
+ trns->record_cnt = 1;
+ return true;
+ }
+
+ while (lex_token (lexer) != '.')
{
bool ok;
- if (!parse_record_placement (&record, &column))
+ if (!parse_record_placement (lexer, &record, &column))
return false;
- if (token == T_STRING)
- ok = parse_string_argument (trns, record, &column);
+ if (lex_token (lexer) == T_STRING)
+ ok = parse_string_argument (lexer, trns, record, &column);
else
- ok = parse_variable_argument (trns, tmp_pool, &record, &column,
+ ok = parse_variable_argument (lexer, dict, trns, tmp_pool, &record, &column,
which_formats);
if (!ok)
return 0;
- lex_match (',');
+ lex_match (lexer, ',');
}
if (trns->record_cnt != 0 && trns->record_cnt != record)
- msg (SW, _("Output calls for %d records but %d specified on RECORDS "
+ msg (SW, _("Output calls for %d records but %zu specified on RECORDS "
"subcommand."),
record, trns->record_cnt);
trns->record_cnt = record;
/* Parses a string argument to the PRINT commands. Returns success. */
static bool
-parse_string_argument (struct print_trns *trns, int record, int *column)
+parse_string_argument (struct lexer *lexer, struct print_trns *trns, int record, int *column)
{
struct prt_out_spec *spec = pool_alloc (trns->pool, sizeof *spec);
spec->type = PRT_LITERAL;
spec->record = record;
spec->first_column = *column;
- ds_init_string (&spec->string, &tokstr);
+ ds_init_string (&spec->string, lex_tokstr (lexer));
ds_register_pool (&spec->string, trns->pool);
- lex_get ();
+ lex_get (lexer);
/* Parse the included column range. */
- if (lex_is_number ())
+ if (lex_is_number (lexer))
{
int first_column, last_column;
bool range_specified;
- if (!parse_column_range (&first_column, &last_column, &range_specified))
- return false;
+ if (!parse_column_range (lexer, 1,
+ &first_column, &last_column, &range_specified))
+ return false;
spec->first_column = first_column;
if (range_specified)
to fixed_parse_compatible() or fixed_parse_fortran() as appropriate.
Returns success. */
static bool
-parse_variable_argument (struct print_trns *trns, struct pool *tmp_pool,
+parse_variable_argument (struct lexer *lexer, const struct dictionary *dict,
+ struct print_trns *trns, struct pool *tmp_pool,
int *record, int *column,
enum which_formats which_formats)
{
- struct variable **vars;
+ const struct variable **vars;
size_t var_cnt, var_idx;
struct fmt_spec *formats, *f;
size_t format_cnt;
bool add_space;
-
- if (!parse_variables_pool (tmp_pool,
- dataset_dict (current_dataset), &vars, &var_cnt, PV_DUPLICATE))
+
+ if (!parse_variables_const_pool (lexer, tmp_pool, dict,
+ &vars, &var_cnt, PV_DUPLICATE))
return false;
- if (lex_is_number () || token == '(')
+ if (lex_is_number (lexer) || lex_token (lexer) == '(')
{
- if (!parse_var_placements (tmp_pool, var_cnt, &formats, &format_cnt))
+ if (!parse_var_placements (lexer, tmp_pool, var_cnt, false,
+ &formats, &format_cnt))
return false;
add_space = false;
}
{
size_t i;
- lex_match ('*');
-
+ lex_match (lexer, '*');
+
formats = pool_nmalloc (tmp_pool, var_cnt, sizeof *formats);
format_cnt = var_cnt;
- for (i = 0; i < var_cnt; i++)
+ for (i = 0; i < var_cnt; i++)
{
- struct variable *v = vars[i];
- formats[i] = which_formats == PRINT ? v->print : v->write;
+ const struct variable *v = vars[i];
+ formats[i] = (which_formats == PRINT
+ ? *var_get_print_format (v)
+ : *var_get_write_format (v));
}
- add_space = true;
+ add_space = which_formats == PRINT;
}
var_idx = 0;
for (f = formats; f < &formats[format_cnt]; f++)
if (!execute_placement_format (f, record, column))
{
- struct variable *var;
+ const struct variable *var;
struct prt_out_spec *spec;
var = vars[var_idx++];
- if (!check_specifier_width (f, var->width, true))
+ if (!fmt_check_width_compat (f, var_get_width (var)))
return false;
spec = pool_alloc (trns->pool, sizeof *spec);
spec->var = var;
spec->format = *f;
spec->add_space = add_space;
+
+ /* This is a completely bizarre twist for compatibility:
+ WRITE outputs the system-missing value as a field
+ filled with spaces, instead of using the normal format
+ that usually contains a period. */
+ spec->sysmis_as_spaces = (which_formats == WRITE
+ && var_is_numeric (var)
+ && (fmt_get_category (spec->format.type)
+ != FMT_CAT_BINARY));
+
ll_push_tail (&trns->specs, &spec->ll);
*column += f->w + add_space;
tab_text (t, 3, 0, TAB_CENTER | TAT_TITLE, _("Format"));
tab_dim (t, tab_natural_dimensions);
row = 1;
- ll_for_each (spec, struct prt_out_spec, ll, &trns->specs)
+ ll_for_each (spec, struct prt_out_spec, ll, &trns->specs)
{
+ char fmt_string[FMT_STRING_LEN_MAX + 1];
int width;
switch (spec->type)
{
width = ds_length (&spec->string);
break;
case PRT_VAR:
- tab_text (t, 0, row, TAB_LEFT, spec->var->name);
+ tab_text (t, 0, row, TAB_LEFT, var_get_name (spec->var));
tab_text (t, 3, row, TAB_LEFT | TAB_FIX,
- fmt_to_string (&spec->format));
+ fmt_to_string (&spec->format, fmt_string));
width = spec->format.w;
break;
default:
\f
/* Transformation. */
-static void flush_records (struct print_trns *,
- int target_record, int *record);
+static void flush_records (struct print_trns *, int target_record,
+ bool *eject, int *record);
/* Performs the transformation inside print_trns T on case C. */
static int
-print_trns_proc (void *trns_, struct ccase *c, casenum_t case_num UNUSED)
+print_trns_proc (void *trns_, struct ccase *c, casenumber case_num UNUSED)
{
struct print_trns *trns = trns_;
+ bool eject = trns->eject;
+ char encoded_space = legacy_from_native (trns->encoding, ' ');
+ int record = 1;
struct prt_out_spec *spec;
- int record;
- if (trns->eject)
- som_eject_page ();
-
- record = 1;
ds_clear (&trns->line);
- ll_for_each (spec, struct prt_out_spec, ll, &trns->specs)
+ ds_put_char (&trns->line, ' ');
+ ll_for_each (spec, struct prt_out_spec, ll, &trns->specs)
{
- flush_records (trns, spec->record, &record);
-
- ds_set_length (&trns->line, spec->first_column - 1, ' ');
+ flush_records (trns, spec->record, &eject, &record);
+
+ ds_set_length (&trns->line, spec->first_column, encoded_space);
if (spec->type == PRT_VAR)
{
- data_out (ds_put_uninit (&trns->line, spec->format.w),
- &spec->format, case_data (c, spec->var->fv));
+ const union value *input = case_data (c, spec->var);
+ char *output = ds_put_uninit (&trns->line, spec->format.w);
+ if (!spec->sysmis_as_spaces || input->f != SYSMIS)
+ data_out_legacy (input, trns->encoding, &spec->format, output);
+ else
+ memset (output, encoded_space, spec->format.w);
if (spec->add_space)
- ds_put_char (&trns->line, ' ');
+ ds_put_char (&trns->line, encoded_space);
+ }
+ else
+ {
+ ds_put_substring (&trns->line, ds_ss (&spec->string));
+ if (trns->encoding != LEGACY_NATIVE)
+ {
+ size_t length = ds_length (&spec->string);
+ char *data = ss_data (ds_tail (&trns->line, length));
+ legacy_recode (LEGACY_NATIVE, data,
+ trns->encoding, data, length);
+ }
}
- else
- ds_put_substring (&trns->line, ds_ss (&spec->string));
}
- flush_records (trns, trns->record_cnt + 1, &record);
-
+ flush_records (trns, trns->record_cnt + 1, &eject, &record);
+
if (trns->writer != NULL && dfm_write_error (trns->writer))
return TRNS_ERROR;
return TRNS_CONTINUE;
}
+/* Advance from *RECORD to TARGET_RECORD, outputting records
+ along the way. If *EJECT is true, then the first record
+ output is preceded by ejecting the page (and *EJECT is set
+ false). */
static void
-flush_records (struct print_trns *trns, int target_record, int *record)
+flush_records (struct print_trns *trns, int target_record,
+ bool *eject, int *record)
{
- while (target_record > *record)
+ for (; target_record > *record; (*record)++)
{
+ char *line = ds_cstr (&trns->line);
+ size_t length = ds_length (&trns->line);
+ char leader = ' ';
+
+ if (*eject)
+ {
+ *eject = false;
+ if (trns->writer == NULL)
+ som_eject_page ();
+ else
+ leader = '1';
+ }
+ line[0] = legacy_from_native (trns->encoding, leader);
+
if (trns->writer == NULL)
- tab_output_text (TAB_FIX | TAT_NOWRAP, ds_cstr (&trns->line));
+ tab_output_text (TAB_FIX | TAT_NOWRAP, &line[1]);
else
{
- if (!trns->omit_new_lines)
- ds_put_char (&trns->line, '\n');
-
- dfm_put_record (trns->writer,
- ds_data (&trns->line), ds_length (&trns->line));
+ if (!trns->include_prefix)
+ {
+ line++;
+ length--;
+ }
+ dfm_put_record (trns->writer, line, length);
}
- ds_clear (&trns->line);
- (*record)++;
+ ds_truncate (&trns->line, 1);
}
}