/* 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
#include "libpspp/message.h"
#include "libpspp/misc.h"
+#include "libpspp/assertion.h"
#include "data/data-in.h"
#if !ODF_READ_SUPPORT
struct casereader *
-ods_open_reader (struct spreadsheet_read_info *gri, struct dictionary **dict)
+ods_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"), "OpenDocument");
NULL,
};
+struct sheet_detail
+{
+ /* The name of the sheet (utf8 encoding) */
+ char *name;
+
+ int start_col;
+ int stop_col;
+ int start_row;
+ int stop_row;
+};
+
+
enum reader_state
{
STATE_INIT = 0, /* Initial state */
struct ods_reader
{
+ struct spreadsheet spreadsheet;
+
xmlTextReaderPtr xtr;
enum reader_state state;
int stop_row;
int stop_col;
+ struct sheet_detail *sheets;
+ int n_allocated_sheets;
+
struct caseproto *proto;
struct dictionary *dict;
struct ccase *first_case;
bool read_names;
struct string ods_errs;
- int span;
};
+
static void process_node (struct ods_reader *r);
+
+const char *
+ods_get_sheet_name (struct spreadsheet *s, int n)
+{
+ int ret;
+ struct ods_reader *or = (struct ods_reader *) s;
+
+ assert (n < s->n_sheets);
+
+ while (
+ (or->n_allocated_sheets <= n)
+ &&
+ (1 == (ret = xmlTextReaderRead (or->xtr)))
+ )
+ {
+ process_node (or);
+ }
+
+ return or->sheets[n].name;
+}
+
+char *
+ods_get_sheet_range (struct spreadsheet *s, int n)
+{
+ int ret;
+ struct ods_reader *or = (struct ods_reader *) s;
+
+ assert (n < s->n_sheets);
+
+ while (
+ (or->n_allocated_sheets <= n || or->state != STATE_SPREADSHEET)
+ &&
+ (1 == (ret = xmlTextReaderRead (or->xtr)))
+ )
+ {
+ process_node (or);
+ }
+
+ return create_cell_ref (
+ or->sheets[n].start_col,
+ or->sheets[n].start_row,
+ or->sheets[n].stop_col,
+ or->sheets[n].stop_row);
+}
+
+
static void
ods_file_casereader_destroy (struct casereader *reader UNUSED, void *r_)
{
if (name == NULL)
name = xmlStrdup (_xml ("--"));
+
r->node_type = xmlTextReaderNodeType (r->xtr);
- switch ( r->state)
+ switch (r->state)
{
case STATE_INIT:
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;
}
break;
case STATE_SPREADSHEET:
- if (0 == xmlStrcasecmp (name, _xml("table:table")))
+ if (0 == xmlStrcasecmp (name, _xml("table:table"))
+ &&
+ (XML_READER_TYPE_ELEMENT == r->node_type))
{
- if (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 = 0;
+ r->row = 0;
+ ++r->sheet_index;
+
+ printf ("%s:%d Start of SHEET %d: Rows %d\n", __FILE__, __LINE__,
+ r->sheet_index, r->row);
+
+ if ( r->target_sheet != NULL)
{
- r->col = -1;
- r->row = -1;
- ++r->sheet_index;
- if ( r->target_sheet != NULL)
- {
- xmlChar *value = xmlTextReaderGetAttribute (r->xtr, _xml ("table:name"));
- if ( 0 == xmlStrcmp (value, r->target_sheet))
- {
- r->sheet_found = true;
- r->state = STATE_TABLE;
- }
- free (value);
- }
- else if (r->target_sheet_index == r->sheet_index)
+ if ( 0 == xmlStrcmp (value, r->target_sheet))
{
r->sheet_found = true;
- r->state = STATE_TABLE;
}
- else if ( r->target_sheet_index == -1)
- r->state = STATE_TABLE;
}
+ else if (r->target_sheet_index == r->sheet_index)
+ {
+ r->sheet_found = true;
+ }
+ r->state = STATE_TABLE;
}
- else if (XML_READER_TYPE_END_ELEMENT == r->node_type
- && r->sheet_found)
+ else if (0 == xmlStrcasecmp (name, _xml("office:spreadsheet")) &&
+ XML_READER_TYPE_ELEMENT == r->node_type)
{
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);
}
- break;
+ break;
case STATE_TABLE:
- if (0 == xmlStrcasecmp (name, _xml("table:table-row")) )
+ if (0 == xmlStrcasecmp (name, _xml("table:table-row")) &&
+ (XML_READER_TYPE_ELEMENT == r->node_type))
{
- if ( XML_READER_TYPE_ELEMENT == r->node_type)
- {
- if (! xmlTextReaderIsEmptyElement (r->xtr))
- {
- r->state = STATE_ROW;
- }
- r->row++;
- r->span = 1;
- }
+ xmlChar *value =
+ xmlTextReaderGetAttribute (r->xtr,
+ _xml ("table:number-rows-repeated"));
+
+ 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 = 0;
+
+ if (! xmlTextReaderIsEmptyElement (r->xtr))
+ r->state = STATE_ROW;
}
- else if (XML_READER_TYPE_END_ELEMENT == r->node_type)
+ 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->state = STATE_SPREADSHEET;
+
}
break;
case STATE_ROW:
- if (0 == xmlStrcasecmp (name, _xml ("table:table-cell")))
+ if ( (0 == xmlStrcasecmp (name, _xml ("table:table-cell")))
+ &&
+ (XML_READER_TYPE_ELEMENT == r->node_type))
{
- if ( XML_READER_TYPE_ELEMENT == r->node_type)
- {
- xmlChar *value =
- xmlTextReaderGetAttribute (r->xtr,
- _xml ("table:number-columns-repeated"));
- r->col += r->span;
- r->span = value ? _xmlchar_to_int (value) : 1;
- free (value);
- if (! xmlTextReaderIsEmptyElement (r->xtr))
- {
- r->state = STATE_CELL;
- }
- }
+ xmlChar *value =
+ xmlTextReaderGetAttribute (r->xtr,
+ _xml ("table:number-columns-repeated"));
+
+ int col_span = value ? _xmlchar_to_int (value) : 1;
+
+ r->col += col_span;
+
+ printf ("%s:%d (span %d) Start of Cell %d, %d\n", __FILE__, __LINE__,
+ col_span,
+ r->row, r->col);
+ if (! xmlTextReaderIsEmptyElement (r->xtr))
+ r->state = STATE_CELL;
}
- else if (XML_READER_TYPE_END_ELEMENT == r->node_type)
+ else if ( (0 == xmlStrcasecmp (name, _xml ("table:table-row")))
+ &&
+ (XML_READER_TYPE_END_ELEMENT == r->node_type))
{
- r->state = STATE_TABLE;
- r->col = -1;
/* Set the span back to the default */
- r->span = 1;
+ printf ("%s:%d End of Cell: %d, %d\n", __FILE__, __LINE__, r->row, r->col);
+ r->state = STATE_TABLE;
}
break;
case STATE_CELL:
- if (0 == xmlStrcasecmp (name, _xml("text:p")))
+ if ( (0 == xmlStrcasecmp (name, _xml("text:p")))
+ &&
+ ( XML_READER_TYPE_ELEMENT == r->node_type))
{
- if ( XML_READER_TYPE_ELEMENT == r->node_type)
- {
- r->state = STATE_CELL_CONTENT;
- }
+ // printf ("%s:%d Start of Cell Contents %d\n", __FILE__, __LINE__, r->row);
+ if (! xmlTextReaderIsEmptyElement (r->xtr))
+ r->state = STATE_CELL_CONTENT;
}
- else if (XML_READER_TYPE_END_ELEMENT == r->node_type)
+ else if
+ ( (0 == xmlStrcasecmp (name, _xml("table:table-cell")))
+ &&
+ (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 (XML_READER_TYPE_TEXT != r->node_type)
- r->state = STATE_CELL;
+ if (r->sheets[r->sheet_index].start_row == -1)
+ r->sheets[r->sheet_index].start_row = r->row - 1;
+
+ if (
+ (r->sheets[r->sheet_index].start_col == -1)
+ ||
+ (r->sheets[r->sheet_index].start_col >= r->col - 1)
+ )
+ r->sheets[r->sheet_index].start_col = r->col - 1;
+
+ r->sheets[r->sheet_index].stop_row = r->row - 1;
+
+ if ( r->sheets[r->sheet_index].stop_col < r->col - 1)
+ r->sheets[r->sheet_index].stop_col = r->col - 1;
+
+ r->state = STATE_CELL;
break;
default:
+ NOT_REACHED ();
break;
};
}
-struct casereader *
-ods_open_reader (struct spreadsheet_read_info *gri, struct dictionary **dict)
+/* Try to find out how many sheets there are in the "workbook" */
+static int
+get_sheet_count (struct zip_reader *zreader)
{
- int ret = 0;
- xmlChar *type = NULL;
- unsigned long int vstart = 0;
- casenumber n_cases = CASENUMBER_MAX;
- int i;
- struct var_spec *var_spec = NULL;
- int n_var_specs = 0;
+ xmlTextReaderPtr mxtr;
+ struct zip_member *meta = NULL;
+ meta = zip_member_open (zreader, "meta.xml");
+
+ if ( meta == NULL)
+ return -1;
+
+ mxtr = xmlReaderForIO ((xmlInputReadCallback) zip_member_read,
+ (xmlInputCloseCallback) zip_member_finish,
+ meta, NULL, NULL, 0);
+
+ while (1 == xmlTextReaderRead (mxtr))
+ {
+ xmlChar *name = xmlTextReaderName (mxtr);
+ if ( 0 == xmlStrcmp (name, _xml("meta:document-statistic")))
+ {
+ xmlChar *attr = xmlTextReaderGetAttribute (mxtr, _xml ("meta:table-count"));
+
+ if ( attr != NULL)
+ {
+ int s = _xmlchar_to_int (attr);
+ return s;
+ }
+ }
+ }
+ return -1;
+}
- struct ods_reader *r = xzalloc (sizeof *r);
+static void
+ods_error_handler (void *ctx, const char *mesg,
+ UNUSED xmlParserSeverities sev, xmlTextReaderLocatorPtr loc)
+{
+ struct ods_reader *r = ctx;
+
+ msg (MW, _("There was a problem whilst reading the %s file `%s' (near line %d): `%s'"),
+ "ODF",
+ r->spreadsheet.file_name,
+ xmlTextReaderLocatorLineNumber (loc),
+ mesg);
+}
+
+
+struct spreadsheet *
+ods_probe (const char *filename, bool report_errors)
+{
+ struct ods_reader *r;
+ struct string errs;
+ xmlTextReaderPtr xtr ;
+ int sheet_count;
struct zip_member *content = NULL;
- struct zip_reader *zreader ;
- xmlChar *val_string = NULL;
- r->read_names = gri->read_names;
- ds_init_empty (&r->ods_errs);
+ struct zip_reader *zreader = NULL;
- zreader = zip_reader_create (gri->file_name, &r->ods_errs);
+ ds_init_empty (&errs);
- if ( NULL == zreader)
- {
- msg (ME, _("Error opening `%s' for reading as a OpenDocument spreadsheet file: %s."),
- gri->file_name, ds_cstr (&r->ods_errs));
+ zreader = zip_reader_create (filename, &errs);
- goto error;
- }
+ if (zreader == NULL)
+ return NULL;
content = zip_member_open (zreader, "content.xml");
- if ( NULL == content)
- {
- msg (ME, _("Could not extract OpenDocument spreadsheet from file `%s': %s."),
- gri->file_name, ds_cstr (&r->ods_errs));
- goto error;
- }
+ if ( content == NULL)
+ goto error;
zip_member_ref (content);
- r->xtr = xmlReaderForIO ((xmlInputReadCallback) zip_member_read,
- (xmlInputCloseCallback) zip_member_finish,
- content, NULL, NULL, XML_PARSE_RECOVER);
+ sheet_count = get_sheet_count (zreader);
- if ( r->xtr == NULL)
- {
- goto error;
- }
+ xtr = xmlReaderForIO ((xmlInputReadCallback) zip_member_read,
+ (xmlInputCloseCallback) zip_member_finish,
+ content, NULL, NULL,
+ report_errors ? 0 : (XML_PARSE_NOERROR | XML_PARSE_NOWARNING) );
+
+ if ( xtr == NULL)
+ goto error;
+
+ 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);
+
+
+ ds_destroy (&errs);
+
+ printf ("%s:%d\n", __FILE__, __LINE__);
+
+ r->spreadsheet.file_name = filename;
+ return &r->spreadsheet;
- if ( gri->cell_range )
+ error:
+ zip_reader_destroy (zreader);
+ ds_destroy (&errs);
+ return NULL;
+}
+
+struct casereader *
+ods_make_reader (struct spreadsheet *spreadsheet,
+ const struct spreadsheet_read_options *opts)
+{
+ intf ret = 0;
+ xmlChar *type = NULL;
+ unsigned long int vstart = 0;
+ casenumber n_cases = CASENUMBER_MAX;
+ int i;
+ struct var_spec *var_spec = NULL;
+ int n_var_specs = 0;
+
+ struct ods_reader *r = (struct ods_reader *) spreadsheet;
+ xmlChar *val_string = NULL;
+
+ assert (r);
+ r->read_names = opts->read_names;
+ ds_init_empty (&r->ods_errs);
+
+ 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);
+ opts->cell_range);
goto error;
}
}
}
r->state = STATE_INIT;
- r->target_sheet = BAD_CAST gri->sheet_name;
- r->target_sheet_index = gri->sheet_index;
- r->row = r->col = -1;
+ r->target_sheet = BAD_CAST opts->sheet_name;
+ r->target_sheet_index = opts->sheet_index;
+ r->row = r->col = 0;
r->sheet_index = 0;
if (ret < 1)
{
msg (MW, _("Selected sheet or range of spreadsheet `%s' is empty."),
- gri->file_name);
+ spreadsheet->file_name);
goto error;
}
- if ( gri->read_names)
+ if ( opts->read_names)
{
while (1 == (ret = xmlTextReaderRead (r->xtr)))
{
{
int idx;
process_node (r);
- if ( r->row >= r->start_row + 1 + gri->read_names)
+ if ( r->row >= r->start_row + 1 + opts->read_names)
break;
if ( r->col < r->start_col)
}
/* Create the dictionary and populate it */
- *dict = r->dict = dict_create (
+ r->spreadsheet.dict = r->dict = dict_create (
CHAR_CAST (const char *, xmlTextReaderConstEncoding (r->xtr)));
for (i = 0 ; i < n_var_specs ; ++i )
struct fmt_spec fmt;
struct variable *var = NULL;
char *name = dict_make_unique_var_name (r->dict, var_spec[i].name, &vstart);
- int width = xmv_to_width (&var_spec[i].firstval, gri->asw);
+ int width = xmv_to_width (&var_spec[i].firstval, opts->asw);
dict_create_var (r->dict, name, width);
free (name);
if ( n_var_specs == 0 )
{
msg (MW, _("Selected sheet or range of spreadsheet `%s' is empty."),
- gri->file_name);
+ spreadsheet->file_name);
goto error;
}
convert_xml_to_value (r->first_case, var, &var_spec[i].firstval);
}
- zip_reader_destroy (zreader);
+ // zip_reader_destroy (zreader);
for ( i = 0 ; i < n_var_specs ; ++i )
{
error:
- zip_reader_destroy (zreader);
+ // zip_reader_destroy (zreader);
for ( i = 0 ; i < n_var_specs ; ++i )
{
free (var_spec);
+ dict_destroy (r->spreadsheet.dict);
+ r->spreadsheet.dict = NULL;
+ ods_file_casereader_destroy (NULL, r);
+
+
return NULL;
}
return r->first_case;
}
-
if ( r->state > STATE_INIT)
{
c = case_create (r->proto);
xmv->value = val_string;
val_string = NULL;
+ /*
for (col = 0; col < r->span ; ++col)
{
const int idx = r->col + col - r->start_col;
convert_xml_to_value (c, var, xmv);
}
+ */
free (xmv->text);
free (xmv->value);
free (xmv);