/* PSPP - a program for statistical analysis.
- Copyright (C) 2007, 2009 Free Software Foundation, Inc.
+ Copyright (C) 2007, 2009, 2010, 2011, 2012, 2013 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
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>. */
-
-
#include <config.h>
-#include <libpspp/message.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"
+
+#include "c-xvasprintf.h"
#if !GNM_SUPPORT
struct casereader *
-gnumeric_open_reader (struct gnumeric_read_info *gri, struct dictionary **dict)
+gnumeric_open_reader (struct spreadsheet_read_info *gri, struct spreadsheet_read_options *opts, struct dictionary **dict)
{
- msg (ME, _("Support for Gnumeric files was not compiled into this installation of PSPP"));
+ msg (ME, _("Support for %s files was not compiled into this installation of PSPP"), "Gnumeric");
return NULL;
}
#else
-#include <data/casereader-provider.h>
-#include <errno.h>
-#include <libpspp/str.h>
-#include <libpspp/i18n.h>
-#include <data/dictionary.h>
-#include <data/variable.h>
-#include <xalloc.h>
+#include "data/gnumeric-reader.h"
+#include <assert.h>
+#include <stdbool.h>
#include <errno.h>
#include <libxml/xmlreader.h>
#include <zlib.h>
-#include <stdbool.h>
-
-#include <data/case.h>
-#include <data/value.h>
-#include "gnumeric-reader.h"
-#include <data/identifier.h>
-#include <assert.h>
+#include "data/case.h"
+#include "data/casereader-provider.h"
+#include "data/dictionary.h"
+#include "data/identifier.h"
+#include "data/value.h"
+#include "data/variable.h"
+#include "libpspp/i18n.h"
+#include "libpspp/str.h"
+#include "gl/xalloc.h"
static void gnm_file_casereader_destroy (struct casereader *, void *);
NULL,
};
-/* Convert a string, which is an integer encoded in base26
- IE, A=0, B=1, ... Z=25 to the integer it represents.
- ... except that in this scheme, digits with an exponent
- greater than 1 are implicitly incremented by 1, so
- AA = 0 + 1*26, AB = 1 + 1*26,
- ABC = 2 + 2*26 + 1*26^2 ....
-*/
-static int
-pseudo_base26 (const char *str)
-{
- int i;
- int multiplier = 1;
- int result = 0;
- int len = strlen (str);
-
- for ( i = len - 1 ; i >= 0; --i)
- {
- int mantissa = (str[i] - 'A');
-
- if ( mantissa < 0 || mantissa > 25 )
- return -1;
-
- if ( i != len - 1)
- mantissa++;
-
- result += mantissa * multiplier;
-
- multiplier *= 26;
- }
-
- return result;
-}
-
-
-
-/* Convert a cell reference in the form "A1:B2", to
- integers. A1 means column zero, row zero.
- B1 means column 1 row 0. AA1 means column 26, row 0.
-*/
-static bool
-convert_cell_ref (const char *ref,
- int *col0, int *row0,
- int *coli, int *rowi)
-{
- char startcol[5];
- char stopcol [5];
-
- int startrow;
- int stoprow;
-
- int n = sscanf (ref, "%4[a-zA-Z]%d:%4[a-zA-Z]%d",
- startcol, &startrow,
- stopcol, &stoprow);
- if ( n != 4)
- return false;
-
- str_uppercase (startcol);
- *col0 = pseudo_base26 (startcol);
- str_uppercase (stopcol);
- *coli = pseudo_base26 (stopcol);
- *row0 = startrow - 1;
- *rowi = stoprow - 1 ;
-
- return true;
-}
-
-
enum reader_state
{
- STATE_INIT = 0, /* Initial state */
+ STATE_PRE_INIT = 0, /* Initial state */
+ STATE_SHEET_COUNT, /* Found the sheet index */
+ STATE_INIT , /* Other Initial state */
STATE_SHEET_START, /* Found the start of a sheet */
STATE_SHEET_NAME, /* Found the sheet name */
STATE_MAXROW,
+ STATE_MAXCOL,
STATE_SHEET_FOUND, /* Found the sheet that we actually want */
STATE_CELLS_START, /* Found the start of the cell array */
STATE_CELL /* Found a cell */
};
+struct sheet_detail
+{
+ xmlChar *name;
+
+ int start_col;
+ int stop_col;
+ int start_row;
+ int stop_row;
+
+ int maxcol;
+ int maxrow;
+};
+
struct gnumeric_reader
{
+ struct spreadsheet spreadsheet;
+
+ /* The libxml reader for this instance */
xmlTextReaderPtr xtr;
+ /* An internal state variable */
enum reader_state state;
+
int row;
int col;
+ int min_col;
int node_type;
int sheet_index;
-
- const xmlChar *target_sheet;
- int target_sheet_index;
-
- int start_row;
int start_col;
- int stop_row;
int stop_col;
+ int start_row;
+ int stop_row;
+
+ struct sheet_detail *sheets;
+ const xmlChar *target_sheet;
+ int target_sheet_index;
- size_t value_cnt;
+ struct caseproto *proto;
struct dictionary *dict;
struct ccase *first_case;
bool used_first_case;
};
+
+const char *
+gnumeric_get_sheet_name (struct spreadsheet *s, int n)
+{
+ struct gnumeric_reader *gr = (struct gnumeric_reader *) s;
+ assert (n < s->n_sheets);
+
+ return gr->sheets[n].name;
+}
+
+
static void process_node (struct gnumeric_reader *r);
-#define _xml(X) (const xmlChar *)(X)
-#define _xmlchar_to_int(X) atoi((const char *)X)
+char *
+gnumeric_get_sheet_range (struct spreadsheet *s, int n)
+{
+ int ret;
+ struct gnumeric_reader *gr = (struct gnumeric_reader *) s;
+
+ assert (n < s->n_sheets);
+
+ while (
+ (gr->sheets[n].stop_col == -1)
+ &&
+ (1 == (ret = xmlTextReaderRead (gr->xtr)))
+ )
+ {
+ process_node (gr);
+ }
+
+
+ return create_cell_ref (
+ gr->sheets[n].start_col,
+ gr->sheets[n].start_row,
+ gr->sheets[n].stop_col,
+ gr->sheets[n].stop_row);
+}
+
static void
gnm_file_casereader_destroy (struct casereader *reader UNUSED, void *r_)
{
+ int i;
struct gnumeric_reader *r = r_;
if ( r == NULL)
return ;
if ( ! r->used_first_case )
case_unref (r->first_case);
+ caseproto_unref (r->proto);
+
+ for (i = 0; i < r->spreadsheet.n_sheets; ++i)
+ {
+ xmlFree (r->sheets[i].name);
+ }
+
+ free (r->sheets);
+
free (r);
}
switch ( r->state)
{
+ case STATE_PRE_INIT:
+ r->sheet_index = -1;
+ if (0 == xmlStrcasecmp (name, _xml("gnm:SheetNameIndex")) &&
+ XML_READER_TYPE_ELEMENT == r->node_type)
+ {
+ r->state = STATE_SHEET_COUNT;
+ r->spreadsheet.n_sheets = 0;
+ }
+ break;
+
+ case STATE_SHEET_COUNT:
+ if (0 == xmlStrcasecmp (name, _xml("gnm:SheetName")) &&
+ XML_READER_TYPE_ELEMENT == r->node_type)
+ {
+ struct sheet_detail *sd ;
+ r->spreadsheet.n_sheets++;
+ r->sheets = xrealloc (r->sheets, r->spreadsheet.n_sheets * sizeof *r->sheets);
+ sd = &r->sheets[r->spreadsheet.n_sheets - 1];
+ sd->start_col = sd->stop_col = sd->start_row = sd->stop_row = -1;
+ }
+ else if (0 == xmlStrcasecmp (name, _xml("gnm:SheetNameIndex")) &&
+ XML_READER_TYPE_END_ELEMENT == r->node_type)
+ {
+ r->state = STATE_INIT;
+ }
+ else if (XML_READER_TYPE_TEXT == r->node_type)
+ {
+ r->sheets [r->spreadsheet.n_sheets - 1].name = xmlTextReaderValue (r->xtr);
+ }
+ break;
+
case STATE_INIT:
if (0 == xmlStrcasecmp (name, _xml("gnm:Sheet")) &&
XML_READER_TYPE_ELEMENT == r->node_type)
{
+ ++r->sheet_index;
r->state = STATE_SHEET_START;
}
break;
{
r->state = STATE_SHEET_NAME;
}
- else if (0 == xmlStrcasecmp (name, _xml("gnm:Name")) &&
- XML_READER_TYPE_END_ELEMENT == r->node_type)
- {
- r->state = STATE_INIT;
- }
break;
case STATE_SHEET_NAME:
if (0 == xmlStrcasecmp (name, _xml("gnm:Name")) &&
XML_READER_TYPE_END_ELEMENT == r->node_type)
{
- r->state = STATE_SHEET_START;
+ r->state = STATE_INIT;
+ }
+ else if (0 == xmlStrcasecmp (name, _xml("gnm:Sheet")) &&
+ XML_READER_TYPE_END_ELEMENT == r->node_type)
+ {
+ r->state = STATE_INIT;
}
else if (XML_READER_TYPE_TEXT == r->node_type)
{
- ++r->sheet_index;
- if ( r->target_sheet != NULL)
+ if ( r->target_sheet != NULL)
{
xmlChar *value = xmlTextReaderValue (r->xtr);
if ( 0 == xmlStrcmp (value, r->target_sheet))
r->state = STATE_SHEET_FOUND;
free (value);
}
- else if (r->target_sheet_index == r->sheet_index)
+ else if (r->target_sheet_index == r->sheet_index + 1)
+ {
+ r->state = STATE_SHEET_FOUND;
+ }
+ else if (r->target_sheet_index == -1)
{
r->state = STATE_SHEET_FOUND;
}
if (0 == xmlStrcasecmp (name, _xml("gnm:Cells")) &&
XML_READER_TYPE_ELEMENT == r->node_type)
{
+ r->min_col = INT_MAX;
if (! xmlTextReaderIsEmptyElement (r->xtr))
r->state = STATE_CELLS_START;
}
{
r->state = STATE_MAXROW;
}
+ else if (0 == xmlStrcasecmp (name, _xml("gnm:MaxCol")) &&
+ XML_READER_TYPE_ELEMENT == r->node_type)
+ {
+ r->state = STATE_MAXCOL;
+ }
else if (0 == xmlStrcasecmp (name, _xml("gnm:Sheet")) &&
XML_READER_TYPE_END_ELEMENT == r->node_type)
{
- r->state = STATE_INIT;
+ r->state = STATE_INIT;
}
break;
case STATE_MAXROW:
{
r->state = STATE_SHEET_FOUND;
}
+ else if (r->node_type == XML_READER_TYPE_TEXT)
+ {
+ xmlChar *value = xmlTextReaderValue (r->xtr);
+ r->sheets[r->sheet_index].maxrow = _xmlchar_to_int (value);
+ xmlFree (value);
+ }
+ break;
+ case STATE_MAXCOL:
+ if (0 == xmlStrcasecmp (name, _xml("gnm:MaxCol")) &&
+ XML_READER_TYPE_END_ELEMENT == r->node_type)
+ {
+ r->state = STATE_SHEET_FOUND;
+ }
+ else if (r->node_type == XML_READER_TYPE_TEXT)
+ {
+ xmlChar *value = xmlTextReaderValue (r->xtr);
+ r->sheets[r->sheet_index].maxcol = _xmlchar_to_int (value);
+ xmlFree (value);
+ }
+ break;
case STATE_CELLS_START:
if (0 == xmlStrcasecmp (name, _xml ("gnm:Cell")) &&
XML_READER_TYPE_ELEMENT == r->node_type)
r->col = _xmlchar_to_int (attr);
free (attr);
+ if (r->col < r->min_col)
+ r->min_col = r->col;
+
attr = xmlTextReaderGetAttribute (r->xtr, _xml ("Row"));
r->row = _xmlchar_to_int (attr);
free (attr);
+ if (r->sheets[r->sheet_index].start_row == -1)
+ {
+ r->sheets[r->sheet_index].start_row = r->row;
+ }
+
+ if (r->sheets[r->sheet_index].start_col == -1)
+ {
+ r->sheets[r->sheet_index].start_col = r->col;
+ }
}
else if (0 == xmlStrcasecmp (name, _xml("gnm:Cells")) &&
XML_READER_TYPE_END_ELEMENT == r->node_type)
- r->state = STATE_SHEET_NAME;
-
+ {
+ r->sheets[r->sheet_index].stop_col = r->col;
+ r->sheets[r->sheet_index].stop_row = r->row;
+ r->state = STATE_SHEET_NAME;
+ }
break;
case STATE_CELL:
if (0 == xmlStrcasecmp (name, _xml("gnm:Cell")) &&
XML_READER_TYPE_END_ELEMENT == r->node_type)
- r->state = STATE_CELLS_START;
+ {
+ r->state = STATE_CELLS_START;
+ }
break;
default:
break;
convert_xml_string_to_value (struct ccase *c, const struct variable *var,
const xmlChar *xv)
{
- int n_bytes = 0;
union value *v = case_data_rw (c, var);
- const char *text = (const char *) xv;
-
- if ( text)
- n_bytes = MIN (var_get_width (var), strlen (text));
-
- if ( var_is_alpha (var))
- {
- memcpy (v->s, text, n_bytes);
- }
+ if (xv == NULL)
+ value_set_missing (v, var_get_width (var));
+ else if ( var_is_alpha (var))
+ value_copy_str_rpad (v, var_get_width (var), xv, ' ');
else
{
+ const char *text = CHAR_CAST (const char *, xv);
char *endptr;
+
errno = 0;
- v->f = strtod (text, &endptr);
+ v->f = c_strtod (text, &endptr);
if ( errno != 0 || endptr == text)
v->f = SYSMIS;
}
xmlChar *first_value;
};
-struct casereader *
-gnumeric_open_reader (struct gnumeric_read_info *gri, struct dictionary **dict)
+
+void
+gnumeric_destroy (struct spreadsheet *s)
{
- unsigned long int vstart = 0;
+ struct gnumeric_reader *r = (struct gnumeric *) s;
+ gnm_file_casereader_destroy (NULL, s);
+}
+
+
+static struct gnumeric_reader *
+gnumeric_reopen (struct gnumeric_reader *r, const char *filename)
+{
int ret;
- casenumber n_cases = CASENUMBER_MAX;
- int i;
- struct var_spec *var_spec = NULL;
- int n_var_specs = 0;
- struct gnumeric_reader *r = NULL;
+ xmlTextReaderPtr xtr;
+ gzFile gz;
+
+ assert (r == NULL || filename == NULL);
- gzFile gz = gzopen (gri->file_name, "r");
+ if (r && r->xtr)
+ xmlFreeTextReader (r->xtr);
+
+ if (filename)
+ gz = gzopen (filename, "r");
+ else
+ gz = gzopen ( r->spreadsheet.file_name, "r");
+
+ if (NULL == gz)
+ return NULL;
+
+ xtr = xmlReaderForIO ((xmlInputReadCallback) gzread,
+ (xmlInputCloseCallback) gzclose, gz,
+ NULL, NULL, 0);
- if ( NULL == gz)
+ if (xtr == NULL)
{
- msg (ME, _("Error opening \"%s\" for reading as a gnumeric file: %s."),
- gri->file_name, strerror (errno));
+ gzclose (gz);
+ return NULL;
+ }
- goto error;
+ if (r == NULL)
+ {
+ r = xzalloc (sizeof *r);
+ r->spreadsheet.n_sheets = -1;
+ r->spreadsheet.file_name = filename;
}
- r = xzalloc (sizeof *r);
+ r->target_sheet = NULL;
+ r->target_sheet_index = -1;
- r->xtr = xmlReaderForIO ((xmlInputReadCallback) gzread, gzclose, gz,
- NULL, NULL, 0);
+ r->row = r->col = -1;
+ r->state = STATE_PRE_INIT;
+ r->xtr = xtr;
+
+ /* Advance to the start of the workbook.
+ This gives us some confidence that we are actually dealing with a gnumeric
+ spreadsheet.
+ */
+ while ( (r->state != STATE_INIT )
+ && 1 == (ret = xmlTextReaderRead (r->xtr)))
+ {
+ process_node (r);
+ }
+
+ r->spreadsheet.type = SPREADSHEET_GNUMERIC;
+
+ return r;
+}
- if ( r->xtr == NULL)
- goto error;
- if ( gri->cell_range )
+struct spreadsheet *
+gnumeric_probe (const char *filename)
+{
+ struct gnumeric_reader *r = gnumeric_reopen (NULL, filename);
+
+ return &r->spreadsheet;
+}
+
+
+struct casereader *
+gnumeric_make_reader (struct spreadsheet *spreadsheet,
+ const struct spreadsheet_read_info *gri,
+ struct spreadsheet_read_options *opts)
+{
+ struct gnumeric_reader *r = NULL;
+ unsigned long int vstart = 0;
+ int ret;
+ casenumber n_cases = CASENUMBER_MAX;
+ int i;
+ struct var_spec *var_spec = NULL;
+ int n_var_specs = 0;
+
+ r = (struct gnumeric_reader *) (spreadsheet);
+
+ if (r->row != -1)
+ r = gnumeric_reopen (r, NULL);
+
+ if ( opts->cell_range )
{
- if ( ! convert_cell_ref (gri->cell_range,
+ if ( ! convert_cell_ref (opts->cell_range,
&r->start_col, &r->start_row,
&r->stop_col, &r->stop_row))
{
- msg (SE, _("Invalid cell range \"%s\""),
- gri->cell_range);
+ msg (SE, _("Invalid cell range `%s'"),
+ opts->cell_range);
goto error;
}
}
else
{
- r->start_col = 0;
+ r->start_col = -1;
r->start_row = 0;
r->stop_col = -1;
r->stop_row = -1;
}
- r->state = STATE_INIT;
- r->target_sheet = BAD_CAST gri->sheet_name;
- r->target_sheet_index = gri->sheet_index;
+ r->target_sheet = BAD_CAST opts->sheet_name;
+ r->target_sheet_index = opts->sheet_index;
r->row = r->col = -1;
- r->sheet_index = 0;
+ r->sheet_index = -1;
/* Advance to the start of the cells for the target sheet */
while ( (r->state != STATE_CELL || r->row < r->start_row )
free (value);
}
-
/* If a range has been given, then use that to calculate the number
of cases */
- if ( gri->cell_range)
+ if ( opts->cell_range)
{
n_cases = MIN (n_cases, r->stop_row - r->start_row + 1);
}
if ( idx >= n_var_specs )
{
+ int i;
+ var_spec = xrealloc (var_spec, sizeof (*var_spec) * (idx + 1));
+ for (i = n_var_specs; i <= idx; ++i)
+ {
+ var_spec [i].name = NULL;
+ var_spec [i].width = -1;
+ var_spec [i].first_value = NULL;
+ }
n_var_specs = idx + 1 ;
- var_spec = realloc (var_spec, sizeof (*var_spec) * n_var_specs);
- var_spec [idx].name = NULL;
- var_spec [idx].width = -1;
- var_spec [idx].first_value = NULL;
}
if ( r->node_type == XML_READER_TYPE_TEXT )
{
xmlChar *value = xmlTextReaderValue (r->xtr);
- const char *text = (const char *) value;
+ const char *text = CHAR_CAST (const char *, value);
if ( r->row < r->start_row)
{
if ( gri->read_names )
{
- var_spec [idx].name = strdup (text);
+ var_spec [idx].name = xstrdup (text);
}
}
else
if (-1 == var_spec [idx].width )
var_spec [idx].width = (gri->asw == -1) ?
- ROUND_UP (strlen(text), MAX_SHORT_STRING) : gri->asw;
+ ROUND_UP (strlen(text), SPREADSHEET_DEFAULT_WIDTH) : gri->asw;
}
free (value);
}
}
-
- /* Create the dictionary and populate it */
- *dict = r->dict = dict_create ();
-
- dict_set_encoding (r->dict, (const char *) xmlTextReaderConstEncoding (r->xtr));
-
- r->value_cnt = 0;
+ {
+ const xmlChar *enc = xmlTextReaderConstEncoding (r->xtr);
+ 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 )
{
- char name[VAR_NAME_LEN + 1];
+ char *name;
+
+ 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 )
- var_spec[i].width = MAX_SHORT_STRING;
-
- r->value_cnt += value_cnt_from_width (var_spec[i].width);
-
- if ( ! dict_make_unique_var_name (r->dict, var_spec[i].name,
- &vstart, name))
- {
- msg (ME, _("Cannot create variable name from %s"), var_spec[i].name);
- goto error;
- }
+ var_spec[i].width = SPREADSHEET_DEFAULT_WIDTH;
+ name = dict_make_unique_var_name (r->dict, var_spec[i].name, &vstart);
dict_create_var (r->dict, name, var_spec[i].width);
+ free (name);
}
/* Create the first case, and cache it */
if ( n_var_specs == 0 )
{
- msg (MW, _("Selected sheet or range of spreadsheet \"%s\" is empty."),
- gri->file_name);
+ msg (MW, _("Selected sheet or range of spreadsheet `%s' is empty."),
+ spreadsheet->file_name);
goto error;
}
- r->first_case = case_create (r->value_cnt);
- memset (case_data_rw_idx (r->first_case, 0)->s,
- ' ', MAX_SHORT_STRING * r->value_cnt);
+ r->proto = caseproto_ref (dict_get_proto (r->dict));
+ r->first_case = case_create (r->proto);
+ case_set_missing (r->first_case);
+ int x = 0;
for ( i = 0 ; i < n_var_specs ; ++i )
{
- const struct variable *var = dict_get_var (r->dict, i);
+ if ( (var_spec[i].name == NULL) && (var_spec[i].first_value == NULL))
+ continue;
+
+ const struct variable *var = dict_get_var (r->dict, x++);
convert_xml_string_to_value (r->first_case, var,
var_spec[i].first_value);
}
free (var_spec);
-
+
+
+ if (opts->cell_range == NULL)
+ {
+ opts->cell_range = c_xasprintf ("%c%d:%c%ld",
+ r->start_col + 'A',
+ r->start_row,
+ r->stop_col + 'A' + caseproto_get_n_widths (r->proto),
+ r->start_row + n_cases);
+ }
+
return casereader_create_sequential
(NULL,
- r->value_cnt,
+ r->proto,
n_cases,
&gnm_file_casereader_class, r);
}
free (var_spec);
- dict_destroy (*dict);
+ dict_destroy (spreadsheet->dict);
+ spreadsheet->dict = NULL;
gnm_file_casereader_destroy (NULL, r);
return r->first_case;
}
- c = case_create (r->value_cnt);
+ c = case_create (r->proto);
+ case_set_missing (c);
- memset (case_data_rw_idx (c, 0)->s, ' ', MAX_SHORT_STRING * r->value_cnt);
+ if (r->start_col == -1)
+ r->start_col = r->min_col;
while ((r->state == STATE_CELL || r->state == STATE_CELLS_START )
&& r->row == current_row && (ret = xmlTextReaderRead (r->xtr)))
r->col > r->stop_col))
continue;
- if ( r->col - r->start_col >= r->value_cnt)
+ if ( r->col - r->start_col >= caseproto_get_n_widths (r->proto))
continue;
if ( r->stop_row != -1 && r->row > r->stop_row)
projects / pspp / blobdiff