X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=src%2Fdata%2Fods-reader.c;h=8f307200b052bb5aa2a8b8d83a6aa93590662d3d;hb=96994a54e60e9c95b8bba54c2281acf7059b1203;hp=fa364090c4564b583d5f4b3bee982ea63772a24b;hpb=41a628f070c7a0d2e9468308cb1bf4ff77de9900;p=pspp diff --git a/src/data/ods-reader.c b/src/data/ods-reader.c index fa364090c4..8f307200b0 100644 --- a/src/data/ods-reader.c +++ b/src/data/ods-reader.c @@ -1,5 +1,5 @@ /* PSPP - a program for statistical analysis. - Copyright (C) 2011, 2012, 2013 Free Software Foundation, Inc. + Copyright (C) 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 @@ -16,56 +16,38 @@ #include -#include "libpspp/message.h" -#include "libpspp/misc.h" -#include "libpspp/assertion.h" - -#include "data/data-in.h" - -#include "gl/c-strtod.h" -#include "gl/minmax.h" - -#include "gettext.h" -#define _(msgid) gettext (msgid) -#define N_(msgid) (msgid) - #include "ods-reader.h" #include "spreadsheet-reader.h" -#if !ODF_READ_SUPPORT - -struct casereader * -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"); - - return NULL; -} - -#else - -#include "libpspp/zip-reader.h" - - #include #include #include #include #include -#include "data/format.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/assertion.h" #include "libpspp/i18n.h" +#include "libpspp/message.h" +#include "libpspp/misc.h" #include "libpspp/str.h" +#include "libpspp/zip-reader.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) + static void ods_file_casereader_destroy (struct casereader *, void *); static struct ccase *ods_file_casereader_read (struct casereader *, void *); @@ -103,6 +85,7 @@ enum reader_state struct state_data { xmlTextReaderPtr xtr; + struct zip_member *zm; int node_type; enum reader_state state; int row; @@ -121,16 +104,19 @@ state_data_destroy (struct state_data *sd) xmlFreeTextReader (sd->xtr); sd->xtr = NULL; + + zip_member_finish (sd->zm); + sd->zm = NULL; } struct ods_reader { struct spreadsheet spreadsheet; struct zip_reader *zreader; - int ref_cnt; + int target_sheet_index; xmlChar *target_sheet_name; - + /* State data for the meta data */ struct state_data msd; @@ -152,14 +138,16 @@ struct ods_reader bool read_names; struct string ods_errs; + + struct string zip_errs; }; void -ods_destroy (struct spreadsheet *s) +ods_unref (struct spreadsheet *s) { struct ods_reader *r = (struct ods_reader *) s; - if (--r->ref_cnt == 0) + if (--s->ref_cnt == 0) { int i; @@ -168,10 +156,12 @@ ods_destroy (struct spreadsheet *s) { xmlFree (r->sheets[i].name); } - + + dict_unref (r->dict); + zip_reader_destroy (r->zreader); free (r->sheets); - + free (s->file_name); free (r); } } @@ -183,10 +173,10 @@ reading_target_sheet (const struct ods_reader *r, const struct state_data *msd) { if (r->target_sheet_name != NULL) { - if ( 0 == xmlStrcmp (r->target_sheet_name, msd->current_sheet_name)) + if (0 == xmlStrcmp (r->target_sheet_name, msd->current_sheet_name)) return true; } - + if (r->target_sheet_index == msd->current_sheet + 1) return true; @@ -205,13 +195,13 @@ ods_get_sheet_name (struct spreadsheet *s, int n) assert (n < s->n_sheets); - while ( + while ( (r->n_allocated_sheets <= n) || or->state != STATE_SPREADSHEET - ) + ) { int ret = xmlTextReaderRead (or->xtr); - if ( ret != 1) + if (ret != 1) break; process_node (r, or); @@ -225,23 +215,23 @@ ods_get_sheet_range (struct spreadsheet *s, int n) { struct ods_reader *r = (struct ods_reader *) s; struct state_data *or = &r->msd; - + assert (n < s->n_sheets); - while ( + while ( (r->n_allocated_sheets <= n) - || (r->sheets[n].stop_row == -1) + || (r->sheets[n].stop_row == -1) || or->state != STATE_SPREADSHEET - ) + ) { int ret = xmlTextReaderRead (or->xtr); - if ( ret != 1) + if (ret != 1) break; process_node (r, or); } - return create_cell_ref ( + return create_cell_range ( r->sheets[n].start_col, r->sheets[n].start_row, r->sheets[n].stop_col, @@ -253,17 +243,17 @@ static void ods_file_casereader_destroy (struct casereader *reader UNUSED, void *r_) { struct ods_reader *r = r_; - if ( r == NULL) + if (r == NULL) return ; state_data_destroy (&r->rsd); - if ( ! ds_is_empty (&r->ods_errs)) + if (! ds_is_empty (&r->ods_errs)) msg (ME, "%s", ds_cstr (&r->ods_errs)); ds_destroy (&r->ods_errs); - if ( r->first_case && ! r->used_first_case ) + if (r->first_case && ! r->used_first_case) case_unref (r->first_case); @@ -274,7 +264,7 @@ ods_file_casereader_destroy (struct casereader *reader UNUSED, void *r_) r->target_sheet_name = NULL; - ods_destroy (&r->spreadsheet); + ods_unref (&r->spreadsheet); } @@ -304,7 +294,7 @@ process_node (struct ods_reader *or, struct state_data *r) break; case STATE_SPREADSHEET: if (0 == xmlStrcasecmp (name, _xml("table:table")) - && + && (XML_READER_TYPE_ELEMENT == r->node_type)) { xmlFree (r->current_sheet_name); @@ -335,38 +325,38 @@ process_node (struct ods_reader *or, struct state_data *r) } 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)) { xmlChar *value = xmlTextReaderGetAttribute (r->xtr, _xml ("table:number-rows-repeated")); - + int row_span = value ? _xmlchar_to_int (value) : 1; r->row += row_span; r->col = 0; - + if (! xmlTextReaderIsEmptyElement (r->xtr)) r->state = STATE_ROW; xmlFree (value); } - else if (0 == xmlStrcasecmp (name, _xml("table:table")) && + else if (0 == xmlStrcasecmp (name, _xml("table:table")) && (XML_READER_TYPE_END_ELEMENT == r->node_type)) { 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)) { xmlChar *value = xmlTextReaderGetAttribute (r->xtr, _xml ("table:number-columns-repeated")); - + r->col_span = value ? _xmlchar_to_int (value) : 1; r->col += r->col_span; @@ -375,7 +365,7 @@ process_node (struct ods_reader *or, struct state_data *r) xmlFree (value); } - else if ( (0 == xmlStrcasecmp (name, _xml ("table:table-row"))) + else if ((0 == xmlStrcasecmp (name, _xml ("table:table-row"))) && (XML_READER_TYPE_END_ELEMENT == r->node_type)) { @@ -383,18 +373,18 @@ process_node (struct ods_reader *or, struct state_data *r) } 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)) + (XML_READER_TYPE_ELEMENT == r->node_type)) { if (! xmlTextReaderIsEmptyElement (r->xtr)) r->state = STATE_CELL_CONTENT; } else if - ( (0 == xmlStrcasecmp (name, _xml("table:table-cell"))) + ((0 == xmlStrcasecmp (name, _xml("table:table-cell"))) && (XML_READER_TYPE_END_ELEMENT == r->node_type) - ) + ) { r->state = STATE_ROW; } @@ -406,16 +396,16 @@ process_node (struct ods_reader *or, struct state_data *r) if (or->sheets[r->current_sheet].start_row == -1) or->sheets[r->current_sheet].start_row = r->row - 1; - if ( + if ( (or->sheets[r->current_sheet].start_col == -1) || (or->sheets[r->current_sheet].start_col >= r->col - 1) - ) + ) or->sheets[r->current_sheet].start_col = r->col - 1; or->sheets[r->current_sheet].stop_row = r->row - 1; - if ( or->sheets[r->current_sheet].stop_col < r->col - 1) + if (or->sheets[r->current_sheet].stop_col < r->col - 1) or->sheets[r->current_sheet].stop_col = r->col - 1; if (XML_READER_TYPE_END_ELEMENT == r->node_type) @@ -429,8 +419,8 @@ process_node (struct ods_reader *or, struct state_data *r) xmlFree (name); } -/* - A struct containing the parameters of a cell's value +/* + A struct containing the parameters of a cell's value parsed from the xml */ struct xml_value @@ -457,13 +447,13 @@ xmv_to_width (const struct xml_value *xmv, int fallback) if (xmv->type != NULL && 0 != xmlStrcmp (xmv->type, _xml("string"))) return 0; - if ( fallback != -1) + if (fallback != -1) return fallback; - if ( xmv->value ) + if (xmv->value) width = ROUND_UP (xmlStrlen (xmv->value), SPREADSHEET_DEFAULT_WIDTH); - else if ( xmv->text) + else if (xmv->text) width = ROUND_UP (xmlStrlen (xmv->text), SPREADSHEET_DEFAULT_WIDTH); @@ -475,13 +465,13 @@ xmv_to_width (const struct xml_value *xmv, int fallback) */ static void convert_xml_to_value (struct ccase *c, const struct variable *var, - const struct xml_value *xmv) + const struct xml_value *xmv, int col, int row) { union value *v = case_data_rw (c, var); if (xmv->value == NULL && xmv->text == NULL) value_set_missing (v, var_get_width (var)); - else if ( var_is_alpha (var)) + else if (var_is_alpha (var)) /* Use the text field, because it seems that there is no value field for strings */ value_copy_str_rpad (v, var_get_width (var), xmv->text, ' '); @@ -490,9 +480,9 @@ convert_xml_to_value (struct ccase *c, const struct variable *var, const struct fmt_spec *fmt = var_get_write_format (var); enum fmt_category fc = fmt_get_category (fmt->type); - assert ( fc != FMT_CAT_STRING); + assert (fc != FMT_CAT_STRING); - if ( 0 == xmlStrcmp (xmv->type, _xml("float"))) + if (0 == xmlStrcmp (xmv->type, _xml("float"))) { v->f = c_strtod (CHAR_CAST (const char *, xmv->value), NULL); } @@ -501,16 +491,32 @@ convert_xml_to_value (struct ccase *c, const struct variable *var, const char *text = xmv->value ? CHAR_CAST (const char *, xmv->value) : CHAR_CAST (const char *, xmv->text); - - free (data_in (ss_cstr (text), "UTF-8", + char *m = data_in (ss_cstr (text), "UTF-8", fmt->type, v, var_get_width (var), - "UTF-8")); + "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); } } } +static int +xml_reader_for_zip_member (void *zm_, char *buffer, int len) +{ + struct zip_member *zm = zm_; + return zip_member_read (zm, buffer, len); +} /* Try to find out how many sheets there are in the "workbook" */ static int @@ -520,43 +526,44 @@ get_sheet_count (struct zip_reader *zreader) struct zip_member *meta = NULL; meta = zip_member_open (zreader, "meta.xml"); - if ( meta == NULL) + if (meta == NULL) return -1; - mxtr = xmlReaderForIO ((xmlInputReadCallback) zip_member_read, - (xmlInputCloseCallback) NULL, - meta, NULL, NULL, 0); + mxtr = xmlReaderForIO (xml_reader_for_zip_member, NULL, meta, NULL, NULL, 0); while (1 == xmlTextReaderRead (mxtr)) { xmlChar *name = xmlTextReaderName (mxtr); - if ( 0 == xmlStrcmp (name, _xml("meta:document-statistic"))) + if (0 == xmlStrcmp (name, _xml("meta:document-statistic"))) { xmlChar *attr = xmlTextReaderGetAttribute (mxtr, _xml ("meta:table-count")); - if ( attr != NULL) + if (attr != NULL) { int s = _xmlchar_to_int (attr); xmlFreeTextReader (mxtr); + zip_member_finish (meta); xmlFree (name); - xmlFree (attr); + xmlFree (attr); return s; } - xmlFree (attr); + xmlFree (attr); } - xmlFree (name); + xmlFree (name); } xmlFreeTextReader (mxtr); + zip_member_finish (meta); return -1; } static void ods_error_handler (void *ctx, const char *mesg, - UNUSED xmlParserSeverities sev, xmlTextReaderLocatorPtr loc) + xmlParserSeverities sev UNUSED, + 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, @@ -565,88 +572,83 @@ ods_error_handler (void *ctx, const char *mesg, } -static xmlTextReaderPtr -init_reader (struct ods_reader *r, bool report_errors) +static bool +init_reader (struct ods_reader *r, bool report_errors, + struct state_data *state) { struct zip_member *content = zip_member_open (r->zreader, "content.xml"); xmlTextReaderPtr xtr; - if ( content == NULL) + if (content == NULL) return NULL; - xtr = xmlReaderForIO ((xmlInputReadCallback) zip_member_read, - (xmlInputCloseCallback) NULL, - content, NULL, NULL, - report_errors ? 0 : (XML_PARSE_NOERROR | XML_PARSE_NOWARNING) ); + xtr = xmlReaderForIO (xml_reader_for_zip_member, NULL, content, NULL, NULL, + report_errors ? 0 : (XML_PARSE_NOERROR | XML_PARSE_NOWARNING)); - if ( xtr == NULL) - return false; + if (xtr == NULL) + return false; + *state = (struct state_data) { .xtr = xtr, + .zm = content, + .state = STATE_INIT }; r->spreadsheet.type = SPREADSHEET_ODS; - if (report_errors) + if (report_errors) xmlTextReaderSetErrorHandler (xtr, ods_error_handler, r); - return xtr; + return true; } + 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); - xmlTextReaderPtr xtr; + struct ods_reader *r = xzalloc (sizeof *r); + struct zip_reader *zr; + + ds_init_empty (&r->zip_errs); + + zr = zip_reader_create (filename, &r->zip_errs); if (zr == NULL) { if (report_errors) { msg (ME, _("Cannot open %s as a OpenDocument file: %s"), - filename, ds_cstr (&errs)); + filename, ds_cstr (&r->zip_errs)); } + ds_destroy (&r->zip_errs); + free (r); return NULL; } sheet_count = get_sheet_count (zr); - r = xzalloc (sizeof *r); r->zreader = zr; - r->ref_cnt = 1; - - xtr = init_reader (r, report_errors); - if (xtr == NULL) - { - goto error; - } - r->msd.xtr = xtr; - r->msd.row = 0; - r->msd.col = 0; - r->msd.current_sheet = 0; - r->msd.state = STATE_INIT; + r->spreadsheet.ref_cnt = 1; + if (!init_reader (r, report_errors, &r->msd)) + goto error; r->spreadsheet.n_sheets = sheet_count; r->n_allocated_sheets = 0; r->sheets = NULL; - ds_destroy (&errs); - - r->spreadsheet.file_name = filename; + r->spreadsheet.file_name = strdup (filename); return &r->spreadsheet; error: + ds_destroy (&r->zip_errs); zip_reader_destroy (r->zreader); - ds_destroy (&errs); free (r); return NULL; } struct casereader * -ods_make_reader (struct spreadsheet *spreadsheet, +ods_make_reader (struct spreadsheet *spreadsheet, const struct spreadsheet_read_options *opts) { intf ret = 0; @@ -656,7 +658,6 @@ ods_make_reader (struct spreadsheet *spreadsheet, int i; struct var_spec *var_spec = NULL; int n_var_specs = 0; - xmlTextReaderPtr xtr; struct ods_reader *r = (struct ods_reader *) spreadsheet; xmlChar *val_string = NULL; @@ -664,24 +665,17 @@ ods_make_reader (struct spreadsheet *spreadsheet, assert (r); r->read_names = opts->read_names; ds_init_empty (&r->ods_errs); - ++r->ref_cnt; + ++r->spreadsheet.ref_cnt; - xtr = init_reader (r, true); - if ( xtr == NULL) + if (!init_reader (r, true, &r->rsd)) goto error; - r->rsd.xtr = xtr; - r->rsd.row = 0; - r->rsd.col = 0; - r->rsd.current_sheet = 0; - r->rsd.state = STATE_INIT; - r->used_first_case = false; r->first_case = NULL; 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)) { @@ -702,8 +696,8 @@ ods_make_reader (struct spreadsheet *spreadsheet, r->target_sheet_index = opts->sheet_index; /* Advance to the start of the cells for the target sheet */ - while ( ! reading_target_sheet (r, &r->rsd) - || r->rsd.state != STATE_ROW || r->rsd.row <= r->start_row ) + while (! reading_target_sheet (r, &r->rsd) + || r->rsd.state != STATE_ROW || r->rsd.row <= r->start_row) { if (1 != (ret = xmlTextReaderRead (r->rsd.xtr))) break; @@ -718,47 +712,47 @@ ods_make_reader (struct spreadsheet *spreadsheet, goto error; } - if ( opts->read_names) + if (opts->read_names) { - while (1 == (ret = xmlTextReaderRead (r->rsd.xtr))) + while (1 == xmlTextReaderRead (r->rsd.xtr)) { - int idx; - process_node (r, &r->rsd); /* If the row is finished then stop for now */ if (r->rsd.state == STATE_TABLE && r->rsd.row > r->start_row) break; - idx = r->rsd.col - r->start_col -1 ; + int idx = r->rsd.col - r->start_col - 1; - if ( idx < 0) + if (idx < 0) continue; if (r->stop_col != -1 && idx > r->stop_col - r->start_col) continue; - if (r->rsd.state == STATE_CELL_CONTENT + if (r->rsd.state == STATE_CELL_CONTENT && XML_READER_TYPE_TEXT == r->rsd.node_type) { xmlChar *value = xmlTextReaderValue (r->rsd.xtr); - - if ( idx >= n_var_specs) + 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 */ memset (var_spec + n_var_specs, - 0, + 0, (idx - n_var_specs + 1) * sizeof (*var_spec)); n_var_specs = idx + 1; } - var_spec[idx].firstval.text = 0; - var_spec[idx].firstval.value = 0; - var_spec[idx].firstval.type = 0; - - var_spec [idx].name = strdup (CHAR_CAST (const char *, value)); + for (int i = 0; i < r->rsd.col_span; ++i) + { + var_spec[idx - i].firstval.text = 0; + var_spec[idx - i].firstval.value = 0; + var_spec[idx - i].firstval.type = 0; + var_spec[idx - i].name = + strdup (CHAR_CAST (const char *, value)); + } xmlFree (value); } @@ -771,7 +765,7 @@ ods_make_reader (struct spreadsheet *spreadsheet, int idx; process_node (r, &r->rsd); - if ( ! reading_target_sheet (r, &r->rsd) ) + if (! reading_target_sheet (r, &r->rsd)) break; /* If the row is finished then stop for now */ @@ -786,21 +780,21 @@ ods_make_reader (struct spreadsheet *spreadsheet, if (r->stop_col != -1 && idx > r->stop_col - r->start_col) continue; - if ( r->rsd.state == STATE_CELL && + if (r->rsd.state == STATE_CELL && XML_READER_TYPE_ELEMENT == r->rsd.node_type) { type = xmlTextReaderGetAttribute (r->rsd.xtr, _xml ("office:value-type")); val_string = xmlTextReaderGetAttribute (r->rsd.xtr, _xml ("office:value")); } - if ( r->rsd.state == STATE_CELL_CONTENT && + if (r->rsd.state == STATE_CELL_CONTENT && XML_READER_TYPE_TEXT == r->rsd.node_type) { if (idx >= n_var_specs) { var_spec = xrealloc (var_spec, sizeof (*var_spec) * (idx + 1)); memset (var_spec + n_var_specs, - 0, + 0, (idx - n_var_specs + 1) * sizeof (*var_spec)); var_spec [idx].name = NULL; @@ -821,7 +815,7 @@ ods_make_reader (struct spreadsheet *spreadsheet, r->spreadsheet.dict = r->dict = dict_create ( CHAR_CAST (const char *, xmlTextReaderConstEncoding (r->rsd.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; @@ -832,7 +826,7 @@ ods_make_reader (struct spreadsheet *spreadsheet, var = dict_get_var (r->dict, i); - if ( 0 == xmlStrcmp (var_spec[i].firstval.type, _xml("date"))) + if (0 == xmlStrcmp (var_spec[i].firstval.type, _xml("date"))) { fmt.type = FMT_DATE; fmt.d = 0; @@ -844,7 +838,7 @@ ods_make_reader (struct spreadsheet *spreadsheet, var_set_both_formats (var, &fmt); } - if ( n_var_specs == 0 ) + if (n_var_specs == 0) { msg (MW, _("Selected sheet or range of spreadsheet `%s' is empty."), spreadsheet->file_name); @@ -860,7 +854,9 @@ ods_make_reader (struct spreadsheet *spreadsheet, { const struct variable *var = dict_get_var (r->dict, i); - convert_xml_to_value (r->first_case, var, &var_spec[i].firstval); + convert_xml_to_value (r->first_case, var, &var_spec[i].firstval, + r->rsd.col - n_var_specs + i, + r->rsd.row - 1); } /* Read in the first row of data */ @@ -873,7 +869,7 @@ ods_make_reader (struct spreadsheet *spreadsheet, } - for ( i = 0 ; i < n_var_specs ; ++i ) + for (i = 0 ; i < n_var_specs ; ++i) { free (var_spec[i].firstval.type); free (var_spec[i].firstval.value); @@ -891,8 +887,8 @@ ods_make_reader (struct spreadsheet *spreadsheet, &ods_file_casereader_class, r); error: - - for ( i = 0 ; i < n_var_specs ; ++i ) + + for (i = 0 ; i < n_var_specs ; ++i) { free (var_spec[i].firstval.type); free (var_spec[i].firstval.value); @@ -902,8 +898,6 @@ ods_make_reader (struct spreadsheet *spreadsheet, free (var_spec); - dict_destroy (r->spreadsheet.dict); - r->spreadsheet.dict = NULL; ods_file_casereader_destroy (NULL, r); return NULL; @@ -929,30 +923,30 @@ ods_file_casereader_read (struct casereader *reader UNUSED, void *r_) /* Advance to the start of a row. (If there is one) */ - while (r->rsd.state != STATE_ROW + while (r->rsd.state != STATE_ROW && 1 == xmlTextReaderRead (r->rsd.xtr) - ) + ) { process_node (r, &r->rsd); } - if ( ! reading_target_sheet (r, &r->rsd) + if (! reading_target_sheet (r, &r->rsd) || r->rsd.state < STATE_TABLE || (r->stop_row != -1 && r->rsd.row > r->stop_row + 1) - ) +) { return NULL; } c = case_create (r->proto); case_set_missing (c); - + while (1 == xmlTextReaderRead (r->rsd.xtr)) { process_node (r, &r->rsd); - if ( r->stop_row != -1 && r->rsd.row > r->stop_row + 1) + if (r->stop_row != -1 && r->rsd.row > r->stop_row + 1) break; if (r->rsd.state == STATE_CELL && @@ -962,15 +956,16 @@ ods_file_casereader_read (struct casereader *reader UNUSED, void *r_) val_string = xmlTextReaderGetAttribute (r->rsd.xtr, _xml ("office:value")); } - if (r->rsd.state == STATE_CELL_CONTENT && + if (r->rsd.state == STATE_CELL_CONTENT && r->rsd.node_type == XML_READER_TYPE_TEXT) { int col; struct xml_value *xmv = xzalloc (sizeof *xmv); xmv->text = xmlTextReaderValue (r->rsd.xtr); - xmv->value = val_string; - xmv->type = type; + xmv->value = val_string; val_string = NULL; + xmv->type = type; + type = NULL; for (col = 0; col < r->rsd.col_span; ++col) { @@ -978,13 +973,13 @@ ods_file_casereader_read (struct casereader *reader UNUSED, void *r_) const int idx = r->rsd.col - col - r->start_col - 1; if (idx < 0) continue; - if (r->stop_col != -1 && idx > r->stop_col - r->start_col ) + if (r->stop_col != -1 && idx > r->stop_col - r->start_col) break; if (idx >= dict_get_var_cnt (r->dict)) break; var = dict_get_var (r->dict, idx); - convert_xml_to_value (c, var, xmv); + convert_xml_to_value (c, var, xmv, idx + r->start_col, r->rsd.row - 1); } xmlFree (xmv->text); @@ -992,10 +987,12 @@ ods_file_casereader_read (struct casereader *reader UNUSED, void *r_) xmlFree (xmv->type); free (xmv); } - if ( r->rsd.state <= STATE_TABLE) + if (r->rsd.state <= STATE_TABLE) break; } + xmlFree (type); + xmlFree (val_string); + return c; } -#endif