X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=src%2Fdata%2Fods-reader.c;h=1b828d1e36a186a9e532c4c88871881e347d23a6;hb=89d1188edd1f031d7043d922620a3e2e49e1db16;hp=6d70f69759acf2581965b97447a86518cc24bd92;hpb=d18b69e583213b0662b38c7a1eafc8a0fe68f45e;p=pspp

diff --git a/src/data/ods-reader.c b/src/data/ods-reader.c
index 6d70f69759..1b828d1e36 100644
--- a/src/data/ods-reader.c
+++ b/src/data/ods-reader.c
@@ -100,7 +100,7 @@ enum reader_state
     STATE_CELL_CONTENT     /* Found a the text within a cell */
   };
 
-struct foo
+struct state_data
 {
   xmlTextReaderPtr xtr;
   int node_type;
@@ -111,9 +111,18 @@ struct foo
   xmlChar *current_sheet_name;
 
   int col_span;
-
 };
 
+static void
+state_data_destroy (struct state_data *sd)
+{
+  xmlFree (sd->current_sheet_name);
+  sd->current_sheet_name = NULL;
+
+  xmlFreeTextReader (sd->xtr);
+  sd->xtr = NULL;
+}
+
 struct ods_reader
 {
   struct spreadsheet spreadsheet;
@@ -122,7 +131,11 @@ struct ods_reader
   int target_sheet_index;
   xmlChar *target_sheet_name;
   
-  struct foo foo;
+  /* State data for the meta data */
+  struct state_data msd;
+
+  /* State data for the reader */
+  struct state_data rsd;
 
   int start_row;
   int start_col;
@@ -139,56 +152,58 @@ struct ods_reader
   bool read_names;
 
   struct string ods_errs;
+
+  struct string zip_errs;
 };
 
 void
 ods_destroy (struct spreadsheet *s)
 {
-#if 0
   struct ods_reader *r = (struct ods_reader *) s;
 
   if (--r->ref_cnt == 0)
     {
       int i;
 
+      state_data_destroy (&r->msd);
       for (i = 0; i < r->n_allocated_sheets; ++i)
 	{
 	  xmlFree (r->sheets[i].name);
 	}
-
+	
       zip_reader_destroy (r->zreader);
       free (r->sheets);
+	
       free (r);
     }
-#endif
 }
 
 
 
 static bool
-reading_target_sheet (const struct ods_reader *r)
+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, r->foo.current_sheet_name))
+      if ( 0 == xmlStrcmp (r->target_sheet_name, msd->current_sheet_name))
 	return true;
     }
   
-  if (r->target_sheet_index == r->foo.current_sheet + 1)
+  if (r->target_sheet_index == msd->current_sheet + 1)
     return true;
 
   return false;
 }
 
 
-static void process_node (struct ods_reader *or, struct foo *r);
+static void process_node (struct ods_reader *or, struct state_data *r);
 
 
 const char *
 ods_get_sheet_name (struct spreadsheet *s, int n)
 {
   struct ods_reader *r = (struct ods_reader *) s;
-  struct foo *or = &r->foo;
+  struct state_data *or = &r->msd;
 
   assert (n < s->n_sheets);
 
@@ -211,7 +226,7 @@ char *
 ods_get_sheet_range (struct spreadsheet *s, int n)
 {
   struct ods_reader *r = (struct ods_reader *) s;
-  struct foo *or = &r->foo;
+  struct state_data *or = &r->msd;
   
   assert (n < s->n_sheets);
 
@@ -228,7 +243,7 @@ ods_get_sheet_range (struct spreadsheet *s, int n)
       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,
@@ -239,30 +254,29 @@ ods_get_sheet_range (struct spreadsheet *s, int n)
 static void
 ods_file_casereader_destroy (struct casereader *reader UNUSED, void *r_)
 {
-#if 0
   struct ods_reader *r = r_;
   if ( r == NULL)
     return ;
 
-  if (r->xtr)
-    xmlFreeTextReader (r->xtr);
-  r->xtr = NULL;
+  state_data_destroy (&r->rsd);
 
   if ( ! ds_is_empty (&r->ods_errs))
     msg (ME, "%s", ds_cstr (&r->ods_errs));
 
   ds_destroy (&r->ods_errs);
 
-  if ( ! r->used_first_case )
+  if ( r->first_case && ! r->used_first_case )
     case_unref (r->first_case);
 
+
   caseproto_unref (r->proto);
+  r->proto = NULL;
 
-  xmlFree (r->current_sheet_name);
   xmlFree (r->target_sheet_name);
+  r->target_sheet_name = NULL;
+
 
   ods_destroy (&r->spreadsheet);
-#endif
 }
 
 
@@ -270,7 +284,7 @@ ods_file_casereader_destroy (struct casereader *reader UNUSED, void *r_)
 
 
 static void
-process_node (struct ods_reader *or, struct foo *r)
+process_node (struct ods_reader *or, struct state_data *r)
 {
   xmlChar *name = xmlTextReaderName (r->xtr);
   if (name == NULL)
@@ -463,7 +477,7 @@ 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);
 
@@ -489,12 +503,22 @@ 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);
 	}
     }
 }
@@ -553,82 +577,85 @@ ods_error_handler (void *ctx, const char *mesg,
 }
 
 
-static bool
+static xmlTextReaderPtr
 init_reader (struct ods_reader *r, bool report_errors)
 {
   struct zip_member *content = zip_member_open (r->zreader, "content.xml");
   xmlTextReaderPtr xtr;
 
   if ( content == NULL)
-    return false;
+    return NULL;
 
-  if (r->foo.xtr)
-    xmlFreeTextReader (r->foo.xtr);
-
-  zip_member_ref (content);
   xtr = xmlReaderForIO ((xmlInputReadCallback) zip_member_read,
-			(xmlInputCloseCallback) zip_member_finish,
+			(xmlInputCloseCallback) NULL,
 			content,   NULL, NULL,
 			report_errors ? 0 : (XML_PARSE_NOERROR | XML_PARSE_NOWARNING) );
 
   if ( xtr == NULL)
     return false;
 
-  r->foo.xtr = xtr;
+
   r->spreadsheet.type = SPREADSHEET_ODS;
-  r->foo.row = 0;
-  r->foo.col = 0;
-  r->foo.current_sheet = 0;
-  r->foo.state = STATE_INIT;
 
   if (report_errors) 
     xmlTextReaderSetErrorHandler (xtr, ods_error_handler, r);
 
-  return true;
+  return xtr;
 }
 
 
+
 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);
+  struct ods_reader *r = xzalloc (sizeof *r);
+  xmlTextReaderPtr xtr;  
+  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;
 
-  if (! init_reader (r, report_errors))
+  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.n_sheets = sheet_count;
   r->n_allocated_sheets = 0;
   r->sheets = NULL;
 
-  ds_destroy (&errs);
-
   r->spreadsheet.file_name = filename;
   return &r->spreadsheet;
 
  error:
+  ds_destroy (&r->zip_errs);
   zip_reader_destroy (r->zreader);
-  ds_destroy (&errs);
   free (r);
   return NULL;
 }
@@ -644,6 +671,7 @@ 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;
@@ -653,9 +681,19 @@ ods_make_reader (struct spreadsheet *spreadsheet,
   ds_init_empty (&r->ods_errs);
   ++r->ref_cnt;
 
-  if ( !init_reader (r, true))
+  xtr = init_reader (r, true);
+  if ( xtr == NULL)
     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,
@@ -675,20 +713,17 @@ ods_make_reader (struct spreadsheet *spreadsheet,
       r->stop_row = -1;
     }
 
-  r->foo.state = STATE_INIT;
   r->target_sheet_name = xmlStrdup (BAD_CAST opts->sheet_name);
   r->target_sheet_index = opts->sheet_index;
-  r->foo.row = r->foo.col = 0;
-
 
   /* Advance to the start of the cells for the target sheet */
-  while ( ! reading_target_sheet (r)  
-	  || r->foo.state != STATE_ROW || r->foo.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->foo.xtr)))
+      if (1 != (ret = xmlTextReaderRead (r->rsd.xtr)))
 	   break;
 
-      process_node (r, &r->foo);
+      process_node (r, &r->rsd);
     }
 
   if (ret < 1)
@@ -700,17 +735,17 @@ ods_make_reader (struct spreadsheet *spreadsheet,
 
   if ( opts->read_names)
     {
-      while (1 == (ret = xmlTextReaderRead (r->foo.xtr)))
+      while (1 == xmlTextReaderRead (r->rsd.xtr))
 	{
 	  int idx;
 
-	  process_node (r, &r->foo);
+	  process_node (r, &r->rsd);
 
 	  /* If the row is finished then stop for now */
-	  if (r->foo.state == STATE_TABLE && r->foo.row > r->start_row)
+	  if (r->rsd.state == STATE_TABLE && r->rsd.row > r->start_row)
 	    break;
 
-	  idx = r->foo.col - r->start_col -1 ;
+	  idx = r->rsd.col - r->start_col -1 ;
 
 	  if ( idx < 0)
 	    continue;
@@ -718,11 +753,11 @@ ods_make_reader (struct spreadsheet *spreadsheet,
 	  if (r->stop_col != -1 && idx > r->stop_col - r->start_col)
 	    continue;
 
-	  if (r->foo.state == STATE_CELL_CONTENT 
+	  if (r->rsd.state == STATE_CELL_CONTENT 
 	      &&
-	      XML_READER_TYPE_TEXT  == r->foo.node_type)
+	      XML_READER_TYPE_TEXT  == r->rsd.node_type)
 	    {
-	      xmlChar *value = xmlTextReaderValue (r->foo.xtr);
+	      xmlChar *value = xmlTextReaderValue (r->rsd.xtr);
 
 	      if ( idx >= n_var_specs)
 		{
@@ -746,35 +781,35 @@ ods_make_reader (struct spreadsheet *spreadsheet,
     }
 
   /* Read in the first row of data */
-  while (1 == xmlTextReaderRead (r->foo.xtr))
+  while (1 == xmlTextReaderRead (r->rsd.xtr))
     {
       int idx;
-      process_node (r, &r->foo);
+      process_node (r, &r->rsd);
 
-      if ( ! reading_target_sheet (r) )
+      if ( ! reading_target_sheet (r, &r->rsd) )
 	break;
 
       /* If the row is finished then stop for now */
-      if (r->foo.state == STATE_TABLE &&
-	  r->foo.row > r->start_row + (opts->read_names ? 1 : 0))
+      if (r->rsd.state == STATE_TABLE &&
+	  r->rsd.row > r->start_row + (opts->read_names ? 1 : 0))
 	break;
 
-      idx = r->foo.col - r->start_col - 1;
+      idx = r->rsd.col - r->start_col - 1;
       if (idx < 0)
 	continue;
 
       if (r->stop_col != -1 && idx > r->stop_col - r->start_col)
 	continue;
 
-      if ( r->foo.state == STATE_CELL &&
-	   XML_READER_TYPE_ELEMENT  == r->foo.node_type)
+      if ( r->rsd.state == STATE_CELL &&
+	   XML_READER_TYPE_ELEMENT  == r->rsd.node_type)
 	{
-	  type = xmlTextReaderGetAttribute (r->foo.xtr, _xml ("office:value-type"));
-	  val_string = xmlTextReaderGetAttribute (r->foo.xtr, _xml ("office:value"));
+	  type = xmlTextReaderGetAttribute (r->rsd.xtr, _xml ("office:value-type"));
+	  val_string = xmlTextReaderGetAttribute (r->rsd.xtr, _xml ("office:value"));
 	}
 
-      if ( r->foo.state == STATE_CELL_CONTENT &&
-	   XML_READER_TYPE_TEXT  == r->foo.node_type)
+      if ( r->rsd.state == STATE_CELL_CONTENT &&
+	   XML_READER_TYPE_TEXT  == r->rsd.node_type)
 	{
 	  if (idx >= n_var_specs)
 	    {
@@ -788,7 +823,7 @@ ods_make_reader (struct spreadsheet *spreadsheet,
 	    }
 
 	  var_spec [idx].firstval.type = type;
-	  var_spec [idx].firstval.text = xmlTextReaderValue (r->foo.xtr);
+	  var_spec [idx].firstval.text = xmlTextReaderValue (r->rsd.xtr);
 	  var_spec [idx].firstval.value = val_string;
 
 	  val_string = NULL;
@@ -799,7 +834,7 @@ ods_make_reader (struct spreadsheet *spreadsheet,
 
   /* Create the dictionary and populate it */
   r->spreadsheet.dict = r->dict = dict_create (
-    CHAR_CAST (const char *, xmlTextReaderConstEncoding (r->foo.xtr)));
+    CHAR_CAST (const char *, xmlTextReaderConstEncoding (r->rsd.xtr)));
 
   for (i = 0; i < n_var_specs ; ++i )
     {
@@ -824,9 +859,6 @@ ods_make_reader (struct spreadsheet *spreadsheet,
       var_set_both_formats (var, &fmt);
     }
 
-  /* Create the first case, and cache it */
-  r->used_first_case = false;
-
   if ( n_var_specs ==  0 )
     {
       msg (MW, _("Selected sheet or range of spreadsheet `%s' is empty."),
@@ -834,6 +866,7 @@ ods_make_reader (struct spreadsheet *spreadsheet,
       goto error;
     }
 
+  /* Create the first case, and cache it */
   r->proto = caseproto_ref (dict_get_proto (r->dict));
   r->first_case = case_create (r->proto);
   case_set_missing (r->first_case);
@@ -842,15 +875,17 @@ 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 */
-  while (1 == xmlTextReaderRead (r->foo.xtr))
+  while (1 == xmlTextReaderRead (r->rsd.xtr))
     {
-      process_node (r, &r->foo);
+      process_node (r, &r->rsd);
 
-      if (r->foo.state == STATE_ROW)
+      if (r->rsd.state == STATE_ROW)
 	break;
     }
 
@@ -898,9 +933,10 @@ static struct ccase *
 ods_file_casereader_read (struct casereader *reader UNUSED, void *r_)
 {
   struct ccase *c = NULL;
+  struct ods_reader *r = r_;
+
   xmlChar *val_string = NULL;
   xmlChar *type = NULL;
-  struct ods_reader *r = r_;
 
   if (!r->used_first_case)
     {
@@ -910,17 +946,17 @@ ods_file_casereader_read (struct casereader *reader UNUSED, void *r_)
 
 
   /* Advance to the start of a row. (If there is one) */
-  while (r->foo.state != STATE_ROW 
-	 && 1 == xmlTextReaderRead (r->foo.xtr)
+  while (r->rsd.state != STATE_ROW 
+	 && 1 == xmlTextReaderRead (r->rsd.xtr)
 	 )
     {
-      process_node (r, &r->foo);
+      process_node (r, &r->rsd);
     }
 
 
-  if ( ! reading_target_sheet (r)  
-       ||  r->foo.state < STATE_TABLE
-       ||  (r->stop_row != -1 && r->foo.row > r->stop_row + 1)
+  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;
@@ -929,34 +965,35 @@ ods_file_casereader_read (struct casereader *reader UNUSED, void *r_)
   c = case_create (r->proto);
   case_set_missing (c);
   
-  while (1 == xmlTextReaderRead (r->foo.xtr))
+  while (1 == xmlTextReaderRead (r->rsd.xtr))
     {
-      process_node (r, &r->foo);
+      process_node (r, &r->rsd);
 
-      if ( r->stop_row != -1 && r->foo.row > r->stop_row + 1)
+      if ( r->stop_row != -1 && r->rsd.row > r->stop_row + 1)
 	break;
 
-      if (r->foo.state == STATE_CELL &&
-	   r->foo.node_type == XML_READER_TYPE_ELEMENT)
+      if (r->rsd.state == STATE_CELL &&
+	   r->rsd.node_type == XML_READER_TYPE_ELEMENT)
 	{
-	  type = xmlTextReaderGetAttribute (r->foo.xtr, _xml ("office:value-type"));
-	  val_string = xmlTextReaderGetAttribute (r->foo.xtr, _xml ("office:value"));
+	  type = xmlTextReaderGetAttribute (r->rsd.xtr, _xml ("office:value-type"));
+	  val_string = xmlTextReaderGetAttribute (r->rsd.xtr, _xml ("office:value"));
 	}
 
-      if (r->foo.state == STATE_CELL_CONTENT && 
-	   r->foo.node_type == XML_READER_TYPE_TEXT)
+      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->foo.xtr);
+	  xmv->text = xmlTextReaderValue (r->rsd.xtr);
 	  xmv->value = val_string;	 
-	  xmv->type = type;
 	  val_string = NULL;
+	  xmv->type = type;
+	  type = NULL;
 
-	  for (col = 0; col < r->foo.col_span; ++col)
+	  for (col = 0; col < r->rsd.col_span; ++col)
 	    {
 	      const struct variable *var;
-	      const int idx = r->foo.col - col - r->start_col - 1;
+	      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 )
@@ -965,7 +1002,7 @@ ods_file_casereader_read (struct casereader *reader UNUSED, void *r_)
 		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);
@@ -973,7 +1010,7 @@ ods_file_casereader_read (struct casereader *reader UNUSED, void *r_)
 	  xmlFree (xmv->type);
 	  free (xmv);
 	}
-      if ( r->foo.state <= STATE_TABLE)
+      if ( r->rsd.state <= STATE_TABLE)
 	break;
     }