/* PSPP - a program for statistical analysis.
- Copyright (C) 2007, 2009, 2010, 2011, 2012, 2013 Free Software Foundation, Inc.
+ Copyright (C) 2007, 2009, 2010, 2011, 2012, 2013, 2016 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
#include <config.h>
-#include "libpspp/message.h"
-#include "libpspp/misc.h"
-
-#include "gl/minmax.h"
-#include "gl/c-strtod.h"
-
-#include "gettext.h"
-#define _(msgid) gettext (msgid)
-#define N_(msgid) (msgid)
-
-#include "spreadsheet-reader.h"
-
-#if !GNM_SUPPORT
-
-struct casereader *
-gnumeric_open_reader (const struct spreadsheet_read_options *opts, struct dictionary **dict)
-{
- msg (ME, _("Support for %s files was not compiled into this installation of PSPP"), "Gnumeric");
-
- return NULL;
-}
-
-#else
-
#include "data/gnumeric-reader.h"
+#include "spreadsheet-reader.h"
#include <assert.h>
#include <stdbool.h>
#include "data/case.h"
#include "data/casereader-provider.h"
+#include "data/data-in.h"
#include "data/dictionary.h"
+#include "data/format.h"
#include "data/identifier.h"
#include "data/value.h"
#include "data/variable.h"
#include "libpspp/i18n.h"
+#include "libpspp/message.h"
+#include "libpspp/misc.h"
#include "libpspp/str.h"
+#include "gl/c-strtod.h"
+#include "gl/minmax.h"
#include "gl/xalloc.h"
+#include "gettext.h"
+#define _(msgid) gettext (msgid)
+#define N_(msgid) (msgid)
+
+/* Shamelessly lifted from the Gnumeric sources:
+ https://git.gnome.org/browse/gnumeric/tree/src/value.h
+ */
+enum gnm_value_type
+{
+ VALUE_EMPTY = 10,
+ VALUE_BOOLEAN = 20,
+ VALUE_INTEGER = 30, /* Note, this was removed from gnumeric in 2006 - old versions may of
+ course still be around. New ones are supposed to use float.*/
+ VALUE_FLOAT = 40,
+ VALUE_ERROR = 50,
+ VALUE_STRING = 60,
+ VALUE_CELLRANGE = 70,
+ VALUE_ARRAY = 80
+};
+
+
+
static void gnm_file_casereader_destroy (struct casereader *, void *);
static struct ccase *gnm_file_casereader_read (struct casereader *, void *);
int maxrow;
};
-struct state_data
+struct state_data
{
/* The libxml reader for this instance */
xmlTextReaderPtr xtr;
struct gnumeric_reader
{
struct spreadsheet spreadsheet;
- int ref_cnt;
struct state_data rsd;
struct state_data msd;
int stop_col;
int start_row;
int stop_row;
-
+
struct sheet_detail *sheets;
const xmlChar *target_sheet;
struct dictionary *dict;
struct ccase *first_case;
bool used_first_case;
+
+ enum gnm_value_type vtype;
};
void
-gnumeric_destroy (struct spreadsheet *s)
+gnumeric_unref (struct spreadsheet *s)
{
struct gnumeric_reader *r = (struct gnumeric_reader *) s;
- if (0 == --r->ref_cnt)
+ if (0 == --s->ref_cnt)
{
int i;
{
xmlFree (r->sheets[i].name);
}
-
+
+
free (r->sheets);
state_data_destroy (&r->msd);
+ dict_unref (r->dict);
+
+ free (s->file_name);
+
free (r);
}
}
struct gnumeric_reader *gr = (struct gnumeric_reader *) s;
assert (n < s->n_sheets);
- return gr->sheets[n].name;
+ return gr->sheets[n].name;
}
{
int ret;
struct gnumeric_reader *gr = (struct gnumeric_reader *) s;
-
+
assert (n < s->n_sheets);
- while (
+ while (
(gr->sheets[n].stop_col == -1)
- &&
+ &&
(1 == (ret = xmlTextReaderRead (gr->msd.xtr)))
- )
+ )
{
process_node (gr, &gr->msd);
}
{
struct gnumeric_reader *r = r_;
- if ( r == NULL)
+ if (r == NULL)
return ;
state_data_destroy (&r->rsd);
- if (r->first_case && ! r->used_first_case )
+ if (r->first_case && ! r->used_first_case)
case_unref (r->first_case);
- if (r->proto)
+ if (r->proto)
caseproto_unref (r->proto);
- gnumeric_destroy (&r->spreadsheet);
+ gnumeric_unref (&r->spreadsheet);
}
}
else if (XML_READER_TYPE_TEXT == sd->node_type)
{
- if ( r->sheets [r->spreadsheet.n_sheets - 1].name == NULL)
+ if (r->sheets [r->spreadsheet.n_sheets - 1].name == NULL)
r->sheets [r->spreadsheet.n_sheets - 1].name = CHAR_CAST (char *, xmlTextReaderValue (sd->xtr));
}
break;
}
else if (XML_READER_TYPE_TEXT == sd->node_type)
{
- if ( r->target_sheet != NULL)
+ if (r->target_sheet != NULL)
{
xmlChar *value = xmlTextReaderValue (sd->xtr);
- if ( 0 == xmlStrcmp (value, r->target_sheet))
+ if (0 == xmlStrcmp (value, r->target_sheet))
sd->state = STATE_SHEET_FOUND;
free (value);
}
if (! xmlTextReaderIsEmptyElement (sd->xtr))
sd->state = STATE_CELL;
}
- else if ( (0 == xmlStrcasecmp (name, _xml("gnm:Cells"))) && (XML_READER_TYPE_END_ELEMENT == sd->node_type) )
+ else if ((0 == xmlStrcasecmp (name, _xml("gnm:Cells"))) && (XML_READER_TYPE_END_ELEMENT == sd->node_type))
{
r->sheets[sd->current_sheet].stop_col = sd->col;
r->sheets[sd->current_sheet].stop_row = sd->row;
*/
static void
convert_xml_string_to_value (struct ccase *c, const struct variable *var,
- const xmlChar *xv)
+ const xmlChar *xv, enum gnm_value_type type, int col, int row)
{
union value *v = case_data_rw (c, var);
if (xv == NULL)
value_set_missing (v, var_get_width (var));
- else if ( var_is_alpha (var))
+ else if (var_is_alpha (var))
value_copy_str_rpad (v, var_get_width (var), xv, ' ');
- else
+ else if (type == VALUE_FLOAT || type == VALUE_INTEGER)
{
const char *text = CHAR_CAST (const char *, xv);
char *endptr;
errno = 0;
v->f = c_strtod (text, &endptr);
- if ( errno != 0 || endptr == text)
+ if (errno != 0 || endptr == text)
v->f = SYSMIS;
}
+ else
+ {
+ const char *text = CHAR_CAST (const char *, xv);
+
+ const struct fmt_spec *fmt = var_get_write_format (var);
+
+ char *m = data_in (ss_cstr (text), "UTF-8",
+ fmt->type,
+ v,
+ var_get_width (var),
+ "UTF-8");
+
+ if (m)
+ {
+ char buf [FMT_STRING_LEN_MAX + 1];
+ char *cell = create_cell_ref (col, row);
+
+ msg (MW, _("Cannot convert the value in the spreadsheet cell %s to format (%s): %s"),
+ cell, fmt_to_string (fmt, buf), m);
+ free (cell);
+ }
+ free (m);
+ }
}
struct var_spec
char *name;
int width;
xmlChar *first_value;
+ int first_type;
};
static void
gnumeric_error_handler (void *ctx, const char *mesg,
- UNUSED xmlParserSeverities sev, xmlTextReaderLocatorPtr loc)
+ xmlParserSeverities sev UNUSED,
+ xmlTextReaderLocatorPtr loc)
{
struct gnumeric_reader *r = ctx;
-
+
msg (MW, _("There was a problem whilst reading the %s file `%s' (near line %d): `%s'"),
"Gnumeric",
r->spreadsheet.file_name,
static struct gnumeric_reader *
gnumeric_reopen (struct gnumeric_reader *r, const char *filename, bool show_errors)
-{
+{
int ret = -1;
struct state_data *sd;
xtr = xmlReaderForIO ((xmlInputReadCallback) gzread,
(xmlInputCloseCallback) gzclose, gz,
NULL, NULL,
- show_errors ? 0 : (XML_PARSE_NOERROR | XML_PARSE_NOWARNING) );
+ show_errors ? 0 : (XML_PARSE_NOERROR | XML_PARSE_NOWARNING));
if (xtr == NULL)
{
{
r = xzalloc (sizeof *r);
r->spreadsheet.n_sheets = -1;
- r->spreadsheet.file_name = filename;
+ r->spreadsheet.file_name = strdup (filename);
sd = &r->msd;
}
else
{
sd = &r->rsd;
}
-
- if (show_errors)
+
+ if (show_errors)
xmlTextReaderSetErrorHandler (xtr, gnumeric_error_handler, r);
r->target_sheet = NULL;
sd->row = sd->col = -1;
sd->state = STATE_PRE_INIT;
sd->xtr = xtr;
- r->ref_cnt++;
+ r->spreadsheet.ref_cnt++;
+
/* Advance to the start of the workbook.
This gives us some confidence that we are actually dealing with a gnumeric
spreadsheet.
*/
- while ( (sd->state != STATE_INIT )
+ while ((sd->state != STATE_INIT)
&& 1 == (ret = xmlTextReaderRead (sd->xtr)))
{
process_node (r, sd);
}
- if ( ret != 1)
+ if (ret != 1)
{
/* Does not seem to be a gnumeric file */
- xmlFreeTextReader (sd->xtr);
- free (r);
+ gnumeric_unref (&r->spreadsheet);
return NULL;
}
const xmlChar *enc = xmlTextReaderConstEncoding (sd->xtr);
xmlCharEncoding xce = xmlParseCharEncoding (CHAR_CAST (const char *, enc));
- if ( XML_CHAR_ENCODING_UTF8 != xce)
+ if (XML_CHAR_ENCODING_UTF8 != xce)
{
- /* I have been told that ALL gnumeric files are UTF8 encoded. If that is correct, this
+ /* I have been told that ALL gnumeric files are UTF8 encoded. If that is correct, this
can never happen. */
msg (MW, _("The gnumeric file `%s' is encoded as %s instead of the usual UTF-8 encoding. "
"Any non-ascii characters will be incorrectly imported."),
gnumeric_make_reader (struct spreadsheet *spreadsheet,
const struct spreadsheet_read_options *opts)
{
+ int type = 0;
int x = 0;
struct gnumeric_reader *r = NULL;
unsigned long int vstart = 0;
r = gnumeric_reopen (r, NULL, true);
- if ( opts->cell_range )
+ if (opts->cell_range)
{
- if ( ! convert_cell_ref (opts->cell_range,
+ if (! convert_cell_ref (opts->cell_range,
&r->start_col, &r->start_row,
&r->stop_col, &r->stop_row))
{
r->proto = NULL;
/* Advance to the start of the cells for the target sheet */
- while ( (r->rsd.state != STATE_CELL || r->rsd.row < r->start_row )
+ while ((r->rsd.state != STATE_CELL || r->rsd.row < r->start_row)
&& 1 == (ret = xmlTextReaderRead (r->rsd.xtr)))
{
xmlChar *value ;
process_node (r, &r->rsd);
value = xmlTextReaderValue (r->rsd.xtr);
- if ( r->rsd.state == STATE_MAXROW && r->rsd.node_type == XML_READER_TYPE_TEXT)
+ if (r->rsd.state == STATE_MAXROW && r->rsd.node_type == XML_READER_TYPE_TEXT)
{
n_cases = 1 + _xmlchar_to_int (value) ;
}
/* If a range has been given, then use that to calculate the number
of cases */
- if ( opts->cell_range)
+ if (opts->cell_range)
{
n_cases = MIN (n_cases, r->stop_row - r->start_row + 1);
}
- if ( opts->read_names )
+ if (opts->read_names)
{
r->start_row++;
n_cases --;
}
+
/* Read in the first row of cells,
including the headers if read_names was set */
while (
- (( r->rsd.state == STATE_CELLS_START && r->rsd.row <= r->start_row) || r->rsd.state == STATE_CELL )
+ ((r->rsd.state == STATE_CELLS_START && r->rsd.row <= r->start_row) || r->rsd.state == STATE_CELL)
&& (ret = xmlTextReaderRead (r->rsd.xtr))
- )
+ )
{
int idx;
+
+ if (r->rsd.state == STATE_CELL && r->rsd.node_type == XML_READER_TYPE_TEXT)
+ {
+ xmlChar *attr =
+ xmlTextReaderGetAttribute (r->rsd.xtr, _xml ("ValueType"));
+
+ type = _xmlchar_to_int (attr);
+
+ xmlFree (attr);
+ }
+
process_node (r, &r->rsd);
- if ( r->rsd.row > r->start_row ) break;
+ if (r->rsd.row > r->start_row)
+ {
+ xmlChar *attr =
+ xmlTextReaderGetAttribute (r->rsd.xtr, _xml ("ValueType"));
+
+ r->vtype = _xmlchar_to_int (attr);
+
+ xmlFree (attr);
+ break;
+ }
- if ( r->rsd.col < r->start_col ||
+ if (r->rsd.col < r->start_col ||
(r->stop_col != -1 && r->rsd.col > r->stop_col))
continue;
idx = r->rsd.col - r->start_col;
- if ( idx >= n_var_specs )
+ if (idx >= n_var_specs)
{
int i;
var_spec = xrealloc (var_spec, sizeof (*var_spec) * (idx + 1));
var_spec [i].name = NULL;
var_spec [i].width = -1;
var_spec [i].first_value = NULL;
+ var_spec [i].first_type = -1;
}
n_var_specs = idx + 1 ;
}
- if ( r->rsd.node_type == XML_READER_TYPE_TEXT )
+ var_spec [idx].first_type = type;
+
+ if (r->rsd.node_type == XML_READER_TYPE_TEXT)
{
xmlChar *value = xmlTextReaderValue (r->rsd.xtr);
const char *text = CHAR_CAST (const char *, value);
- if ( r->rsd.row < r->start_row)
+ if (r->rsd.row < r->start_row)
{
- if ( opts->read_names )
+ if (opts->read_names)
{
var_spec [idx].name = xstrdup (text);
}
{
var_spec [idx].first_value = xmlStrdup (value);
- if (-1 == var_spec [idx].width )
+ if (-1 == var_spec [idx].width)
var_spec [idx].width = (opts->asw == -1) ?
ROUND_UP (strlen(text), SPREADSHEET_DEFAULT_WIDTH) : opts->asw;
}
free (value);
}
- else if ( r->rsd.node_type == XML_READER_TYPE_ELEMENT
+ else if (r->rsd.node_type == XML_READER_TYPE_ELEMENT
&& r->rsd.state == STATE_CELL)
{
- if ( r->rsd.row == r->start_row )
+ if (r->rsd.row == r->start_row)
{
xmlChar *attr =
xmlTextReaderGetAttribute (r->rsd.xtr, _xml ("ValueType"));
- if ( NULL == attr || 60 != _xmlchar_to_int (attr))
+ if (NULL == attr || VALUE_STRING != _xmlchar_to_int (attr))
var_spec [idx].width = 0;
free (attr);
{
const xmlChar *enc = xmlTextReaderConstEncoding (r->rsd.xtr);
- if ( enc == NULL)
+ if (enc == NULL)
goto error;
/* Create the dictionary and populate it */
spreadsheet->dict = r->dict = dict_create (CHAR_CAST (const char *, enc));
}
- for (i = 0 ; i < n_var_specs ; ++i )
+ for (i = 0 ; i < n_var_specs ; ++i)
{
char *name;
- if ( (var_spec[i].name == NULL) && (var_spec[i].first_value == NULL))
+ if ((var_spec[i].name == NULL) && (var_spec[i].first_value == NULL))
continue;
/* Probably no data exists for this variable, so allocate a
default width */
- if ( var_spec[i].width == -1 )
+ if (var_spec[i].width == -1)
var_spec[i].width = SPREADSHEET_DEFAULT_WIDTH;
name = dict_make_unique_var_name (r->dict, var_spec[i].name, &vstart);
/* Create the first case, and cache it */
r->used_first_case = false;
- if ( n_var_specs == 0 )
+ if (n_var_specs == 0)
{
msg (MW, _("Selected sheet or range of spreadsheet `%s' is empty."),
spreadsheet->file_name);
case_set_missing (r->first_case);
- for ( i = 0 ; i < n_var_specs ; ++i )
+ for (i = 0 ; i < n_var_specs ; ++i)
{
const struct variable *var;
- if ( (var_spec[i].name == NULL) && (var_spec[i].first_value == NULL))
+ if ((var_spec[i].name == NULL) && (var_spec[i].first_value == NULL))
continue;
var = dict_get_var (r->dict, x++);
convert_xml_string_to_value (r->first_case, var,
- var_spec[i].first_value);
+ var_spec[i].first_value,
+ var_spec[i].first_type,
+ r->rsd.col + i - 1,
+ r->rsd.row - 1);
}
- for ( i = 0 ; i < n_var_specs ; ++i )
+ for (i = 0 ; i < n_var_specs ; ++i)
{
free (var_spec[i].first_value);
free (var_spec[i].name);
}
free (var_spec);
-
+
return casereader_create_sequential
(NULL,
error:
- for ( i = 0 ; i < n_var_specs ; ++i )
+ for (i = 0 ; i < n_var_specs ; ++i)
{
free (var_spec[i].first_value);
free (var_spec[i].name);
}
free (var_spec);
- dict_destroy (spreadsheet->dict);
- spreadsheet->dict = NULL;
gnm_file_casereader_destroy (NULL, r);
struct gnumeric_reader *r = r_;
int current_row = r->rsd.row;
- if ( !r->used_first_case )
+ if (!r->used_first_case)
{
r->used_first_case = true;
return r->first_case;
if (r->start_col == -1)
r->start_col = r->rsd.min_col;
- while ((r->rsd.state == STATE_CELL || r->rsd.state == STATE_CELLS_START )
+
+ while ((r->rsd.state == STATE_CELL || r->rsd.state == STATE_CELLS_START)
&& r->rsd.row == current_row && (ret = xmlTextReaderRead (r->rsd.xtr)))
{
process_node (r, &r->rsd);
- if ( r->rsd.col < r->start_col || (r->stop_col != -1 &&
+ if (r->rsd.state == STATE_CELL && r->rsd.node_type == XML_READER_TYPE_ELEMENT)
+ {
+ xmlChar *attr =
+ xmlTextReaderGetAttribute (r->rsd.xtr, _xml ("ValueType"));
+
+ r->vtype = _xmlchar_to_int (attr);
+
+ xmlFree (attr);
+ }
+
+ if (r->rsd.col < r->start_col || (r->stop_col != -1 &&
r->rsd.col > r->stop_col))
continue;
- if ( r->rsd.col - r->start_col >= caseproto_get_n_widths (r->proto))
+ if (r->rsd.col - r->start_col >= caseproto_get_n_widths (r->proto))
continue;
- if ( r->stop_row != -1 && r->rsd.row > r->stop_row)
+ if (r->stop_row != -1 && r->rsd.row > r->stop_row)
break;
- if ( r->rsd.node_type == XML_READER_TYPE_TEXT )
+
+ if (r->rsd.node_type == XML_READER_TYPE_TEXT)
{
xmlChar *value = xmlTextReaderValue (r->rsd.xtr);
-
const int idx = r->rsd.col - r->start_col;
-
const struct variable *var = dict_get_var (r->dict, idx);
- convert_xml_string_to_value (c, var, value);
+ convert_xml_string_to_value (c, var, value, r->vtype,
+ r->rsd.col, r->rsd.row);
- free (value);
+ xmlFree (value);
}
}
return NULL;
}
}
-
-
-#endif /* GNM_SUPPORT */
projects / pspp / blobdiff