/* PSPP - a program for statistical analysis.
- Copyright (C) 2007, 2009, 2010, 2011, 2012 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
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 min_col;
int node_type;
int sheet_index;
+
+ int start_col;
+ int stop_col;
+ int start_row;
+ int stop_row;
- xmlChar **sheet_names;
+ struct sheet_detail *sheets;
const xmlChar *target_sheet;
int target_sheet_index;
- int start_row;
- int start_col;
- int stop_row;
- int stop_col;
-
struct caseproto *proto;
struct dictionary *dict;
struct ccase *first_case;
};
-char *
+const char *
gnumeric_get_sheet_name (struct spreadsheet *s, int n)
{
struct gnumeric_reader *gr = (struct gnumeric_reader *) s;
- assert (n < s->sheets);
+ assert (n < s->n_sheets);
- return gr->sheet_names[n];
+ return gr->sheets[n].name;
}
+
static void process_node (struct gnumeric_reader *r);
+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_)
{
caseproto_unref (r->proto);
- for (i = 0; i < r->spreadsheet.sheets; ++i)
+ for (i = 0; i < r->spreadsheet.n_sheets; ++i)
{
- xmlFree (r->sheet_names[i]);
+ xmlFree (r->sheets[i].name);
}
- free (r->sheet_names);
+ 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.sheets = 0;
+ r->spreadsheet.n_sheets = 0;
}
break;
if (0 == xmlStrcasecmp (name, _xml("gnm:SheetName")) &&
XML_READER_TYPE_ELEMENT == r->node_type)
{
- r->spreadsheet.sheets++;
- r->sheet_names = xrealloc (r->sheet_names, r->spreadsheet.sheets * sizeof *r->sheet_names);
+ 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)
}
else if (XML_READER_TYPE_TEXT == r->node_type)
{
- r->sheet_names [r->spreadsheet.sheets - 1] = xmlTextReaderValue (r->xtr);
+ r->sheets [r->spreadsheet.n_sheets - 1].name = xmlTextReaderValue (r->xtr);
}
break;
{
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)
{
- 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;
}
{
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)
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")) &&
xmlChar *first_value;
};
-struct spreadsheet *
-gnumeric_probe (const char *filename)
+
+void
+gnumeric_destroy (struct spreadsheet *s)
{
+ 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;
- struct gnumeric_reader *r = NULL;
+
xmlTextReaderPtr xtr;
+ gzFile gz;
+
+ assert (r == NULL || filename == NULL);
- gzFile gz = gzopen (filename, "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;
NULL, NULL, 0);
if (xtr == NULL)
- return NULL;
+ {
+ gzclose (gz);
+ return NULL;
+ }
- r = xzalloc (sizeof *r);
-
- r->xtr = xtr;
- r->spreadsheet.sheets = -1;
- r->state = STATE_PRE_INIT;
+ if (r == NULL)
+ {
+ r = xzalloc (sizeof *r);
+ r->spreadsheet.n_sheets = -1;
+ r->spreadsheet.file_name = filename;
+ }
+
+ r->target_sheet = NULL;
+ r->target_sheet_index = -1;
+ 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
process_node (r);
}
- if (ret != 1)
- {
- /* Not a gnumeric spreadsheet */
- free (r);
- gzclose (gz);
- return NULL;
- }
-
r->spreadsheet.type = SPREADSHEET_GNUMERIC;
- r->spreadsheet.file_name = filename;
-
+
+ return r;
+}
+
+
+struct spreadsheet *
+gnumeric_probe (const char *filename)
+{
+ struct gnumeric_reader *r = gnumeric_reopen (NULL, filename);
+
return &r->spreadsheet;
}
r = (struct gnumeric_reader *) (spreadsheet);
+ if (r->row != -1)
+ r = gnumeric_reopen (r, NULL);
+
if ( opts->cell_range )
{
if ( ! convert_cell_ref (opts->cell_range,
r->stop_row = -1;
}
-
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 ( opts->cell_range)
projects / pspp / blobdiff