/* PSPP - a program for statistical analysis.
- Copyright (C) 2011, 2012 Free Software Foundation, Inc.
+ Copyright (C) 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
int stop_col;
int start_row;
int stop_row;
-
- int maxcol;
- int maxrow;
};
struct ods_reader
{
struct spreadsheet spreadsheet;
-
+ struct zip_reader *zreader;
xmlTextReaderPtr xtr;
enum reader_state state;
int row;
int col;
int node_type;
- int sheet_index;
- int max_col;
- int min_col;
+ int current_sheet;
const xmlChar *target_sheet;
int target_sheet_index;
+
+ int wanted_row_start;
+ int wanted_col_start;
+
+#if 0
int start_row;
int start_col;
int stop_row;
int stop_col;
+#endif
+
+ int col_span;
struct sheet_detail *sheets;
int n_allocated_sheets;
char *
ods_get_sheet_range (struct spreadsheet *s, int n)
{
- int ret;
+ int ret = -1;
struct ods_reader *or = (struct ods_reader *) s;
assert (n < s->n_sheets);
while (
- (or->n_allocated_sheets <= n || or->sheets[n].stop_col == -1)
+ (
+ (or->n_allocated_sheets <= n)
+ || (or->sheets[n].stop_row == -1) )
&&
(1 == (ret = xmlTextReaderRead (or->xtr)))
)
process_node (or);
}
+
return create_cell_ref (
or->sheets[n].start_col,
or->sheets[n].start_row,
if (r->xtr)
xmlFreeTextReader (r->xtr);
+ r->xtr = NULL;
if ( ! ds_is_empty (&r->ods_errs))
msg (ME, "%s", ds_cstr (&r->ods_errs));
caseproto_unref (r->proto);
- free (r);
+ // free (r);
}
static void
if (0 == xmlStrcasecmp (name, _xml("office:spreadsheet")) &&
XML_READER_TYPE_ELEMENT == r->node_type)
{
- printf ("%s:%d Start of Workbook %d: Rows %d\n", __FILE__, __LINE__,
- r->sheet_index, r->row);
-
r->state = STATE_SPREADSHEET;
+ r->current_sheet = -1;
}
break;
case STATE_SPREADSHEET:
(XML_READER_TYPE_ELEMENT == r->node_type))
{
xmlChar *value = xmlTextReaderGetAttribute (r->xtr, _xml ("table:name"));
- r->sheets = xrealloc (r->sheets, sizeof (*r->sheets) * ++r->n_allocated_sheets);
- r->sheets[r->n_allocated_sheets - 1].start_col = -1;
- r->sheets[r->n_allocated_sheets - 1].stop_col = -1;
- r->sheets[r->n_allocated_sheets - 1].start_row = -1;
- r->sheets[r->n_allocated_sheets - 1].stop_row = -1;
- r->sheets[r->n_allocated_sheets - 1].name = value;
- r->col = -1;
- r->row = -1;
- r->max_col = -1;
- r->min_col = INT_MAX;
- ++r->sheet_index;
-
- printf ("%s:%d Start of SHEET %d: Rows %d\n", __FILE__, __LINE__,
- r->sheet_index, r->row);
+
+ ++r->current_sheet;
+
+ if (r->current_sheet >= r->n_allocated_sheets)
+ {
+ r->sheets = xrealloc (r->sheets, sizeof (*r->sheets) * ++r->n_allocated_sheets);
+ r->sheets[r->n_allocated_sheets - 1].start_col = -1;
+ r->sheets[r->n_allocated_sheets - 1].stop_col = -1;
+ r->sheets[r->n_allocated_sheets - 1].start_row = -1;
+ r->sheets[r->n_allocated_sheets - 1].stop_row = -1;
+ r->sheets[r->n_allocated_sheets - 1].name = value;
+ }
+
+ r->col = 0;
+ r->row = 0;
if ( r->target_sheet != NULL)
{
r->sheet_found = true;
}
}
- else if (r->target_sheet_index == r->sheet_index)
+ else if (r->target_sheet_index == r->current_sheet)
{
r->sheet_found = true;
}
{
r->state = STATE_INIT;
printf ("%s:%d End of Workbook %d: Rows %d Cols %d\n", __FILE__, __LINE__,
- r->sheet_index, r->row, r->col);
+ r->current_sheet, r->row, r->col);
}
break;
case STATE_TABLE:
int row_span = value ? _xmlchar_to_int (value) : 1;
- printf ("%s:%d Start of Row %d Span %d\n", __FILE__, __LINE__, r->row, row_span);
r->row += row_span;
- r->col = -1;
+ r->col = 0;
if (! xmlTextReaderIsEmptyElement (r->xtr))
r->state = STATE_ROW;
else if (0 == xmlStrcasecmp (name, _xml("table:table")) &&
(XML_READER_TYPE_END_ELEMENT == r->node_type))
{
- printf ("%s:%d End of SHEET %d %d,%d\n", __FILE__, __LINE__, r->sheet_index, r->row, r->col);
- r->sheets[r->sheet_index].stop_row = r->row;
- r->sheets[r->sheet_index].stop_col = r->max_col - 1;
- r->sheets[r->sheet_index].start_col = r->min_col;
r->state = STATE_SPREADSHEET;
-
}
break;
case STATE_ROW:
xmlTextReaderGetAttribute (r->xtr,
_xml ("table:number-columns-repeated"));
- int col_span = value ? _xmlchar_to_int (value) : 0;
-
- r->col += col_span;
- r->col ++;
+ r->col_span = value ? _xmlchar_to_int (value) : 1;
+ r->col += r->col_span;
- if (r->min_col > r->col)
- r->min_col = r->col;
-
- printf ("%s:%d Start of Cell %d, %d\n", __FILE__, __LINE__, r->row, r->col);
if (! xmlTextReaderIsEmptyElement (r->xtr))
r->state = STATE_CELL;
}
&&
(XML_READER_TYPE_END_ELEMENT == r->node_type))
{
- /* Set the span back to the default */
- printf ("%s:%d End of Cell: %d, %d\n", __FILE__, __LINE__, r->row, r->col);
- if ( r->max_col < r->col)
- r->max_col = r->col;
r->state = STATE_TABLE;
}
break;
&&
( XML_READER_TYPE_ELEMENT == r->node_type))
{
- // printf ("%s:%d Start of Cell Contents %d\n", __FILE__, __LINE__, r->row);
if (! xmlTextReaderIsEmptyElement (r->xtr))
r->state = STATE_CELL_CONTENT;
}
(XML_READER_TYPE_END_ELEMENT == r->node_type)
)
{
- // printf ("%s:%d End of Cell contents: Rows %d\n", __FILE__, __LINE__, r->row);
r->state = STATE_ROW;
}
break;
case STATE_CELL_CONTENT:
- if (r->sheets[r->sheet_index].start_row == -1)
- r->sheets[r->sheet_index].start_row = r->row;
- // printf ("%s:%d Cell contents: Rows %d\n", __FILE__, __LINE__, r->row);
- /* if (XML_READER_TYPE_TEXT != r->node_type) */
- r->state = STATE_CELL;
+ assert (r->current_sheet >= 0);
+ assert (r->current_sheet < r->n_allocated_sheets);
+
+ if (r->sheets[r->current_sheet].start_row == -1)
+ r->sheets[r->current_sheet].start_row = r->row - 1;
+
+ if (
+ (r->sheets[r->current_sheet].start_col == -1)
+ ||
+ (r->sheets[r->current_sheet].start_col >= r->col - 1)
+ )
+ r->sheets[r->current_sheet].start_col = r->col - 1;
+
+ r->sheets[r->current_sheet].stop_row = r->row - 1;
+
+ if ( r->sheets[r->current_sheet].stop_col < r->col - 1)
+ r->sheets[r->current_sheet].stop_col = r->col - 1;
+
+ if (XML_READER_TYPE_END_ELEMENT == r->node_type)
+ r->state = STATE_CELL;
break;
default:
NOT_REACHED ();
}
-struct spreadsheet *
-ods_probe (const char *filename, bool report_errors)
+static bool
+init_reader (struct ods_reader *r, bool report_errors)
{
- struct ods_reader *r;
- struct string errs;
- xmlTextReaderPtr xtr ;
- int sheet_count;
- struct zip_member *content = NULL;
-
- struct zip_reader *zreader = NULL;
-
- ds_init_empty (&errs);
-
- zreader = zip_reader_create (filename, &errs);
-
- if (zreader == NULL)
- return NULL;
-
- content = zip_member_open (zreader, "content.xml");
+ struct zip_member *content = zip_member_open (r->zreader, "content.xml");
+ xmlTextReaderPtr xtr;
if ( content == NULL)
- goto error;
+ return false;
zip_member_ref (content);
- sheet_count = get_sheet_count (zreader);
xtr = xmlReaderForIO ((xmlInputReadCallback) zip_member_read,
(xmlInputCloseCallback) zip_member_finish,
report_errors ? 0 : (XML_PARSE_NOERROR | XML_PARSE_NOWARNING) );
if ( xtr == NULL)
- goto error;
+ return false;
- r = xzalloc (sizeof *r);
r->xtr = xtr;
r->spreadsheet.type = SPREADSHEET_ODS;
- r->spreadsheet.n_sheets = sheet_count;
- r->n_allocated_sheets = 0;
- r->sheet_index = -1;
- r->sheets = NULL;
if (report_errors)
xmlTextReaderSetErrorHandler (xtr, ods_error_handler, r);
+ return true;
+}
- ds_destroy (&errs);
- printf ("%s:%d\n", __FILE__, __LINE__);
+struct spreadsheet *
+ods_probe (const char *filename, bool report_errors)
+{
+ struct ods_reader *r;
+ struct string errs = DS_EMPTY_INITIALIZER;
+ int sheet_count;
+ struct zip_reader *zr = zip_reader_create (filename, &errs);
+
+ if (zr == NULL)
+ return NULL;
+
+ sheet_count = get_sheet_count (zr);
+
+ r = xzalloc (sizeof *r);
+ r->zreader = zr;
+
+ if (! init_reader (r, report_errors))
+ {
+ goto error;
+ }
+
+ r->spreadsheet.n_sheets = sheet_count;
+ r->n_allocated_sheets = 0;
+ r->sheets = NULL;
+
+ ds_destroy (&errs);
r->spreadsheet.file_name = filename;
return &r->spreadsheet;
error:
- zip_reader_destroy (zreader);
+ zip_reader_destroy (r->zreader);
ds_destroy (&errs);
+ free (r);
return NULL;
}
r->read_names = opts->read_names;
ds_init_empty (&r->ods_errs);
+
+ if ( !init_reader (r, true))
+ goto error;
+
+#if 0
if ( opts->cell_range )
{
if ( ! convert_cell_ref (opts->cell_range,
r->stop_col = -1;
r->stop_row = -1;
}
+#endif
r->state = STATE_INIT;
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->row = r->col = 0;
+#if 0
/* If CELLRANGE was given, then we know how many variables should be read */
if ( r->stop_col != -1 )
{
var_spec = xrealloc (var_spec, sizeof (*var_spec) * n_var_specs);
memset (var_spec, '\0', sizeof (*var_spec) * n_var_specs);
}
-
+#endif
/* Advance to the start of the cells for the target sheet */
- while ( (r->row < r->start_row ))
+ while ( r->current_sheet < r->target_sheet_index - 1 ||
+ r->state != STATE_TABLE
+ )
{
if (1 != (ret = xmlTextReaderRead (r->xtr)))
break;
process_node (r);
}
+
if (ret < 1)
{
msg (MW, _("Selected sheet or range of spreadsheet `%s' is empty."),
while (1 == (ret = xmlTextReaderRead (r->xtr)))
{
int idx;
+
process_node (r);
- if ( r->row > r->start_row)
- break;
- if (r->col == -1 && r->row == r->start_row)
+ /* If the row is finished then stop for now */
+ if (r->state == STATE_TABLE && r->row > r->wanted_row_start)
break;
- if ( r->col < r->start_col)
- continue;
-
- idx = r->col - r->start_col;
+ idx = r->col - r->wanted_col_start - 1;
if (r->state == STATE_CELL_CONTENT
&&
XML_READER_TYPE_TEXT == r->node_type)
{
xmlChar *value = xmlTextReaderValue (r->xtr);
+
if ( idx >= n_var_specs)
{
-
var_spec = xrealloc (var_spec, sizeof (*var_spec) * (idx + 1));
/* xrealloc (unlike realloc) doesn't initialise its memory to 0 */
var_spec[idx].firstval.type = 0;
var_spec [idx].name = strdup (CHAR_CAST (const char *, value));
- free (value);
- value = NULL;
+ xmlFree (value);
}
}
}
{
int idx;
process_node (r);
- if ( r->row >= r->start_row + 1 + opts->read_names)
- break;
-
- if ( r->col < r->start_col)
- continue;
- if ( r->col - r->start_col + 1 > n_var_specs)
- continue;
+ /* If the row is finished then stop for now */
+ if (r->state == STATE_TABLE && r->row > r->wanted_row_start + (opts->read_names ? 1 : 0))
+ break;
- idx = r->col - r->start_col;
+ idx = r->col - r->wanted_col_start - 1;
if ( r->state == STATE_CELL &&
XML_READER_TYPE_ELEMENT == r->node_type)
if ( r->state == STATE_CELL_CONTENT &&
XML_READER_TYPE_TEXT == r->node_type)
{
+ if ( idx >= n_var_specs)
+ {
+ var_spec = xrealloc (var_spec, sizeof (*var_spec) * (idx + 1));
+ var_spec [idx].name = NULL;
+ n_var_specs = idx + 1;
+ }
+
var_spec [idx].firstval.type = type;
var_spec [idx].firstval.text = xmlTextReaderValue (r->xtr);
var_spec [idx].firstval.value = val_string;
+
val_string = NULL;
type = NULL;
}
r->spreadsheet.dict = r->dict = dict_create (
CHAR_CAST (const char *, xmlTextReaderConstEncoding (r->xtr)));
- for (i = 0 ; i < n_var_specs ; ++i )
+ for (i = 0; i < n_var_specs ; ++i )
{
struct fmt_spec fmt;
struct variable *var = NULL;
r->first_case = case_create (r->proto);
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 = dict_get_var (r->dict, i);
convert_xml_to_value (r->first_case, var, &var_spec[i].firstval);
}
+ /* Read in the first row of data */
+ while (1 == xmlTextReaderRead (r->xtr))
+ {
+ process_node (r);
+
+ if (r->state == STATE_ROW)
+ break;
+ }
+
// zip_reader_destroy (zreader);
+#if 0
for ( i = 0 ; i < n_var_specs ; ++i )
{
free (var_spec[i].firstval.type);
}
free (var_spec);
+#endif
return casereader_create_sequential
(NULL,
struct ods_reader *r = r_;
int current_row = r->row;
- if ( r->row == -1)
- return NULL;
-
if ( !r->used_first_case )
{
r->used_first_case = true;
case_set_missing (c);
}
+ if (r->state == STATE_SPREADSHEET
+ &&
+ r->current_sheet == r->target_sheet_index - 1
+ )
+ {
+ return NULL;
+ }
+
while (1 == xmlTextReaderRead (r->xtr))
{
process_node (r);
- if ( r->row > current_row)
- {
- break;
- }
- if ( r->col < r->start_col || (r->stop_col != -1 && r->col > r->stop_col))
- {
- continue;
- }
- if ( r->col - r->start_col >= caseproto_get_n_widths (r->proto))
- {
- continue;
- }
- if ( r->stop_row != -1 && r->row > r->stop_row)
- {
- continue;
- }
+
+ if (r->row > current_row && r->state == STATE_ROW)
+ break;
+
if ( r->state == STATE_CELL &&
r->node_type == XML_READER_TYPE_ELEMENT )
{
val_string = xmlTextReaderGetAttribute (r->xtr, _xml ("office:value"));
}
- if ( r->state == STATE_CELL_CONTENT && r->node_type == XML_READER_TYPE_TEXT )
+ if ( r->state == STATE_CELL_CONTENT &&
+ r->node_type == XML_READER_TYPE_TEXT )
{
int col;
struct xml_value *xmv = xzalloc (sizeof *xmv);
xmv->value = val_string;
val_string = NULL;
- /*
- for (col = 0; col < r->span ; ++col)
+ for (col = 0; col < r->col_span; ++col)
{
- const int idx = r->col + col - r->start_col;
-
+ const int idx = r->col + col - r->wanted_col_start - 1;
const struct variable *var = dict_get_var (r->dict, idx);
-
convert_xml_to_value (c, var, xmv);
}
- */
+
free (xmv->text);
free (xmv->value);
free (xmv);
break;
}
- if (NULL == c || (r->stop_row != -1 && r->row > r->stop_row + 1))
+ if (NULL == c)
{
case_unref (c);
return NULL;
projects / pspp / blobdiff