Dogcows Code - chaz/homebank/blob - src/hb-import-csv.c

   1 /*      HomeBank -- Free, easy, personal accounting for everyone.
   2  *      Copyright (C) 1995-2019 Maxime DOYEN
   3  *
   4  *      This file is part of HomeBank.
   5  *
   6  *      HomeBank is free software; you can redistribute it and/or modify
   7  *      it under the terms of the GNU General Public License as published by
   8  *      the Free Software Foundation; either version 2 of the License, or
   9  *      (at your option) any later version.
  10  *
  11  *      HomeBank is distributed in the hope that it will be useful,
  12  *      but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  *      MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.    See the
  14  *      GNU General Public License for more details.
  15  *
  16  *      You should have received a copy of the GNU General Public License
  17  *      along with this program.        If not, see <http://www.gnu.org/licenses/>.
  18  */
  19
  20 #include "homebank.h"
  21
  22
  23 #include "hb-import.h"
  24
  25
  26 /****************************************************************************/
  27 /* Debug macros                                                             */
  28 /****************************************************************************/
  29 #define MYDEBUG 0
  30
  31 #if MYDEBUG
  32 #define DB(x) (x);
  33 #else
  34 #define DB(x);
  35 #endif
  36
  37 /* our global datas */
  38 extern struct HomeBank *GLOBALS;
  39 extern struct Preferences *PREFS;
  40
  41
  42 static gint csvtype[7] = {
  43                                         CSV_DATE,
  44                                         CSV_INT,
  45                                         CSV_STRING,
  46                                         CSV_STRING,
  47                                         CSV_STRING,
  48                                         CSV_DOUBLE,
  49                                         CSV_STRING,
  50                                         };
  51
  52
  53
  54 static gchar *hb_csv_strndup (gchar *str, gsize n)
  55 {
  56 gchar *new_str;
  57 gchar *twoquote;
  58
  59         if (str)
  60         {
  61                 new_str = g_new (gchar, n + 1);
  62                 if(*str=='\"')
  63                 {
  64                         str++; n--;
  65                 }
  66                 if(str[n-1]=='\"')
  67                         n--;
  68
  69                 strncpy (new_str, str, n);
  70                 new_str[n] = '\0';
  71
  72                 // replace ""
  73                 twoquote = strstr(new_str, "\"\"");
  74                 if(twoquote)
  75                         strcpy (twoquote, twoquote+1);
  76
  77                 //todo: replace &amp; &lt; &gt; &apos; &quot; ??
  78
  79         }
  80         else
  81                 new_str = NULL;
  82
  83         return new_str;
  84 }
  85
  86
  87 static gchar *hb_csv_find_delimiter(gchar *string, gchar delimiter)
  88 {
  89  gchar *s = string;
  90 gboolean enclosed = FALSE;
  91
  92         while( *s != '\0' )
  93         {
  94                 if( (*s == delimiter) && (enclosed == FALSE) )
  95                         break;
  96
  97                 if( *s == '\"' )
  98                 {
  99                         enclosed = !enclosed;
 100                 }
 101
 102                 s++;
 103         }
 104
 105         return s;
 106 }
 107
 108
 109 static gboolean hb_csv_row_valid(gchar **str_array, guint nbcolumns, gint *csvtype)
 110 {
 111 gboolean valid = TRUE;
 112 guint i;
 113 extern int errno;
 114
 115 #if MYDEBUG == 1
 116 gchar *type[5] = { "string", "date", "int", "double" };
 117 gint lasttype;
 118 #endif
 119
 120         DB( g_print("\n** hb_string_csv_valid: init %d\n", valid) );
 121
 122         DB( g_print(" -> length %d, nbcolumns %d\n", g_strv_length( str_array ), nbcolumns) );
 123
 124         if( g_strv_length( str_array ) != nbcolumns )
 125         {
 126                 valid = FALSE;
 127                 goto csvend;
 128         }
 129
 130         for(i=0;i<nbcolumns;i++)
 131         {
 132 #if MYDEBUG == 1
 133                 lasttype = csvtype[i];
 134 #endif
 135
 136                 if(valid == FALSE)
 137                 {
 138                         DB( g_print(" -> fail on column %d, type: %s\n", i, type[lasttype]) );
 139                         break;
 140                 }
 141
 142                 DB( g_print(" -> control column %d, type: %d, valid: %d '%s'\n", i, lasttype, valid, str_array[i]) );
 143
 144                 switch( csvtype[i] )
 145                 {
 146                         case CSV_DATE:
 147                                 valid = hb_string_isdate(str_array[i]);
 148                                 break;
 149                         case CSV_STRING:
 150                                 valid = hb_string_isprint(str_array[i]);
 151                                 break;
 152                         case CSV_INT:
 153                                 valid = hb_string_isdigit(str_array[i]);
 154                                 break;
 155                         case CSV_DOUBLE :
 156
 157                                 //todo: use strtod (to take care or . or ,)
 158                                 g_ascii_strtod(str_array[i], NULL);
 159                                 //todo : see this errno
 160                                 if( errno )
 161                                 {
 162                                         DB( g_print("errno: %d\n", errno) );
 163                                         valid = FALSE;
 164                                 }
 165                                 break;
 166                 }
 167         }
 168
 169 csvend:
 170
 171         DB( g_print(" --> return %d\n", valid) );
 172
 173         return valid;
 174 }
 175
 176
 177 static gchar **hb_csv_row_get(gchar *string, gchar delimiter, gint max_tokens)
 178 {
 179 GSList *string_list = NULL, *slist;
 180 gchar **str_array, *s;
 181 guint n = 0;
 182 gchar *remainder;
 183
 184         g_return_val_if_fail (string != NULL, NULL);
 185         g_return_val_if_fail (delimiter != '\0', NULL);
 186
 187         if (max_tokens < 1)
 188                 max_tokens = G_MAXINT;
 189
 190         remainder = string;
 191         s = hb_csv_find_delimiter (remainder, delimiter);
 192         if (s)
 193         {
 194                 while (--max_tokens && s && *s != '\0')
 195                 {
 196                 gsize len;
 197
 198                         len = s - remainder;
 199                         string_list = g_slist_prepend (string_list, hb_csv_strndup (remainder, len));
 200                         DB( g_print("   stored=[%s]\n", (gchar *)string_list->data) );
 201
 202                         n++;
 203                         remainder = s + 1;
 204                         s = hb_csv_find_delimiter (remainder, delimiter);
 205                 }
 206         }
 207         if (*string)
 208         {
 209         gsize len;
 210
 211                 len = s - remainder;
 212                 n++;
 213                 string_list = g_slist_prepend (string_list, hb_csv_strndup (remainder, len));
 214                 DB( g_print("   stored=[%s]\n", (gchar *)string_list->data) );
 215         }
 216
 217         str_array = g_new (gchar*, n + 1);
 218
 219         str_array[n--] = NULL;
 220         for (slist = string_list; slist; slist = slist->next)
 221                 str_array[n--] = slist->data;
 222
 223         g_slist_free (string_list);
 224
 225         return str_array;
 226 }
 227
 228
 229 static gchar hb_csv_get_separator(void)
 230 {
 231 static const gchar sep[] = PRF_DTEX_CSVSEP_BUFFER;
 232         return sep[PREFS->dtex_csvsep];
 233 }
 234
 235
 236 gboolean hb_csv_test_line(gchar *rawline)
 237 {
 238 gchar **str_array;
 239 gchar sep;
 240 gboolean isvalid = FALSE;
 241
 242         hb_string_strip_crlf(rawline);
 243         sep = hb_csv_get_separator();
 244         str_array = hb_csv_row_get(rawline, sep, 8);
 245         isvalid = hb_csv_row_valid(str_array, 8, csvtype);
 246
 247         g_strfreev (str_array);
 248
 249         return isvalid;
 250 }
 251
 252
 253 GList *homebank_csv_import(ImportContext *ictx, GenFile *genfile)
 254 {
 255 GIOChannel *io;
 256 //GList *list = NULL;
 257
 258         DB( g_print("\n[import] homebank csv\n") );
 259
 260         io = g_io_channel_new_file(genfile->filepath, "r", NULL);
 261         if(io != NULL)
 262         {
 263         gchar *tmpstr;
 264         gchar sep;
 265         gsize length;
 266         gint io_stat;
 267         gboolean isvalid;
 268         gint count = 0;
 269         gint error = 0;
 270         GenAcc *newacc;
 271         GError *err = NULL;
 272
 273
 274                 newacc = hb_import_gen_acc_get_next(ictx, FILETYPE_CSV_HB, NULL, NULL);
 275
 276                 if( genfile->encoding != NULL )
 277                 {
 278                         g_io_channel_set_encoding(io, genfile->encoding, NULL);
 279                 }
 280
 281                 for(;;)
 282                 {
 283                         io_stat = g_io_channel_read_line(io, &tmpstr, &length, NULL, &err);
 284                         if( io_stat == G_IO_STATUS_EOF)
 285                                 break;
 286                         if( io_stat == G_IO_STATUS_ERROR )
 287                         {
 288                                 DB (g_print(" + ERROR %s\n",err->message));
 289                                 break;
 290                         }
 291                         if( io_stat == G_IO_STATUS_NORMAL)
 292                         {
 293                                 if( *tmpstr != '\0' )
 294                                 {
 295                                 gchar **str_array;
 296
 297                                         count++;
 298
 299                                         sep = hb_csv_get_separator();
 300
 301                                         hb_string_strip_crlf(tmpstr);
 302                                         str_array = hb_csv_row_get(tmpstr, sep, 8);
 303
 304                                         // 0:date; 1:paymode; 2:info; 3:payee, 4:wording; 5:amount; 6:category; 7:tags
 305                                         isvalid = hb_csv_row_valid(str_array, 8, csvtype);
 306
 307                                         DB( g_print("\n (row-%04d) ->|%s|<-\n", count, tmpstr) );
 308                                         DB( g_print(" valid %d, '%s'\n", isvalid, tmpstr) );
 309
 310                                         if( !isvalid )
 311                                         {
 312                                                 g_warning ("csv parse: line %d, invalid column count or data", count);
 313                                                 error++;
 314                                                 //todo log line in error to report user
 315                                         }
 316                                         else
 317                                         {
 318                                         GenTxn *newope = da_gen_txn_malloc();;
 319
 320                                                 DB( g_print(" adding txn\n" ) );
 321
 322                                                 /* convert to generic transaction */
 323                                                 newope->date            = g_strdup(str_array[0]);
 324                                                 newope->paymode         = atoi(str_array[1]);
 325                                                 //todo: reinforce controls here
 326                                                 // csv file are standalone, so no way to link a target txn
 327                                                 //added 5.1.8 forbid to import 5=internal xfer
 328                                                 if(newope->paymode == PAYMODE_INTXFER)
 329                                                         newope->paymode = PAYMODE_XFER;
 330                                                 newope->rawinfo         = g_strdup(str_array[2]);
 331                                                 newope->rawpayee        = g_strdup(g_strstrip(str_array[3]));
 332                                                 newope->rawmemo         = g_strdup(str_array[4]);
 333                                                 newope->amount          = hb_qif_parser_get_amount(str_array[5]);
 334                                                 newope->category        = g_strdup(g_strstrip(str_array[6]));
 335                                                 newope->tags            = g_strdup(str_array[7]);
 336                                                 newope->account         = g_strdup(newacc->name);
 337
 338                                                 /* todo: move this eval date valid */
 339                                                 //guint32 juliantmp = hb_date_get_julian(str_array[0], ictx->datefmt);
 340                                                 ///if( juliantmp == 0 )
 341                                                 //      ictx->cnt_err_date++;
 342
 343                                                 /*
 344                                                 DB( g_print(" storing %s : %s : %s :%s : %s : %s : %s : %s\n",
 345                                                         str_array[0], str_array[1], str_array[2],
 346                                                         str_array[3], str_array[4], str_array[5],
 347                                                         str_array[6], str_array[7]
 348                                                         ) );
 349                                                 */
 350
 351                                                 da_gen_txn_append(ictx, newope);
 352
 353                                                 g_strfreev (str_array);
 354                                         }
 355                                 }
 356                                 g_free(tmpstr);
 357                         }
 358
 359                 }
 360                 g_io_channel_unref (io);
 361
 362         /*
 363                 ui_dialog_msg_infoerror(data->window, error > 0 ? GTK_MESSAGE_ERROR : GTK_MESSAGE_INFO,
 364                         _("Transaction CSV import result"),
 365                         _("%d transactions inserted\n%d errors in the file"),
 366                         count, error);
 367                 */
 368         }
 369
 370
 371         return ictx->gen_lst_txn;
 372 }
 373
 374