examples/loadables/csv.c

   1 /* csv - process a line of csv data and populate an indexed array with the
   2          fields */
   3
   4 /*
   5    Copyright (C) 2020 Free Software Foundation, Inc.
   6
   7    Bush is free software: you can redistribute it and/or modify
   8    it under the terms of the GNU General Public License as published by
   9    the Free Software Foundation, either version 3 of the License, or
  10    (at your option) any later version.
  11
  12    Bush is distributed in the hope that it will be useful,
  13    but WITHOUT ANY WARRANTY; without even the implied warranty of
  14    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15    GNU General Public License for more details.
  16
  17    You should have received a copy of the GNU General Public License
  18    along with Bush.  If not, see <http://www.gnu.org/licenses/>.
  19 */
  20
  21 /* See Makefile for compilation details. */
  22
  23 #include <config.h>
  24
  25 #if defined (HAVE_UNISTD_H)
  26 #  include <unistd.h>
  27 #endif
  28 #include "bushansi.h"
  29 #include <stdio.h>
  30
  31 #include "loadables.h"
  32
  33 #define CSV_ARRAY_DEFAULT       "CSV"
  34
  35 #define NQUOTE  0
  36 #define DQUOTE  1
  37
  38 /* Split LINE into comma-separated fields, storing each field into a separate
  39    element of array variable CSV, starting at index 0. The format of LINE is
  40    as described in RFC 4180. */
  41 static int
  42 csvsplit (csv, line)
  43      SHELL_VAR *csv;
  44      char *line;
  45 {
  46   arrayind_t ind;
  47   char *field, *prev, *buf, *xbuf;
  48   int delim, qstate;
  49   int b, rval;
  50
  51   xbuf = 0;
  52   ind = 0;
  53   field = prev = line;
  54
  55   do
  56     {
  57       if (*prev == '"')
  58         {
  59           if (xbuf == 0)
  60             xbuf = xmalloc (strlen (prev) + 1);
  61           buf = xbuf;
  62           b = 0;
  63           qstate = DQUOTE;
  64           for (field = ++prev; *field; field++)
  65             {
  66               if (qstate == DQUOTE && *field == '"' && field[1] == '"')
  67                 buf[b++] = *field++;    /* skip double quote */
  68               else if (qstate == DQUOTE && *field == '"')
  69                 qstate = NQUOTE;
  70               else if (qstate == NQUOTE && *field == ',')
  71                 break;
  72               else
  73                 /* This copies any text between a closing double quote and the
  74                    delimiter. If you want to change that, make sure to do the
  75                    copy only if qstate == DQUOTE. */
  76                 buf[b++] = *field;
  77             }
  78           buf[b] = '\0';
  79         }
  80       else
  81         {
  82           buf = prev;
  83           field = prev + strcspn (prev, ",");
  84         }
  85
  86       delim = *field;
  87       *field = '\0';
  88
  89       bind_array_element (csv, ind, buf, 0);
  90       ind++;
  91
  92       *field = delim;
  93
  94       if (delim == ',')
  95         prev = field + 1;
  96     }
  97   while (delim == ',');
  98
  99   if (xbuf)
 100     free (xbuf);
 101
 102   return (rval = ind);                          /* number of fields */
 103 }
 104
 105 int
 106 csv_builtin (list)
 107      WORD_LIST *list;
 108 {
 109   int opt, rval;
 110   char *array_name, *csvstring;
 111   SHELL_VAR *v;
 112
 113   array_name = 0;
 114   rval = EXECUTION_SUCCESS;
 115
 116   reset_internal_getopt ();
 117   while ((opt = internal_getopt (list, "a:")) != -1)
 118     {
 119       switch (opt)
 120         {
 121         case 'a':
 122           array_name = list_optarg;
 123           break;
 124         CASE_HELPOPT;
 125         default:
 126           builtin_usage ();
 127           return (EX_USAGE);
 128         }
 129     }
 130   list = loptend;
 131
 132   if (array_name == 0)
 133     array_name = CSV_ARRAY_DEFAULT;
 134
 135   if (legal_identifier (array_name) == 0)
 136     {
 137       sh_invalidid (array_name);
 138       return (EXECUTION_FAILURE);
 139     }
 140
 141   if (list == 0)
 142     {
 143       builtin_error ("csv string argument required");
 144       return (EX_USAGE);
 145     }
 146
 147   v = find_or_make_array_variable (array_name, 1);
 148   if (v == 0 || readonly_p (v) || noassign_p (v))
 149     {
 150       if (v && readonly_p (v))
 151         err_readonly (array_name);
 152       return (EXECUTION_FAILURE);
 153     }
 154   else if (array_p (v) == 0)
 155     {
 156       builtin_error ("%s: not an indexed array", array_name);
 157       return (EXECUTION_FAILURE);
 158     }
 159   if (invisible_p (v))
 160     VUNSETATTR (v, att_invisible);
 161   array_flush (array_cell (v));
 162
 163   csvstring = list->word->word;
 164
 165   if (csvstring == 0 || *csvstring == 0)
 166     return (EXECUTION_SUCCESS);
 167
 168   opt = csvsplit (v, csvstring);
 169   /* Maybe do something with OPT here, it's the number of fields */
 170
 171   return (rval);
 172 }
 173
 174 /* Called when builtin is enabled and loaded from the shared object.  If this
 175    function returns 0, the load fails. */
 176 int
 177 csv_builtin_load (name)
 178      char *name;
 179 {
 180   return (1);
 181 }
 182
 183 /* Called when builtin is disabled. */
 184 void
 185 csv_builtin_unload (name)
 186      char *name;
 187 {
 188 }
 189
 190 char *csv_doc[] = {
 191         "Read comma-separated fields from a string.",
 192         "",
 193         "Parse STRING, a line of comma-separated values, into individual fields,",
 194         "and store them into the indexed array ARRAYNAME starting at index 0.",
 195         "If ARRAYNAME is not supplied, \"CSV\" is the default array name.",
 196         (char *)NULL
 197 };
 198
 199 struct builtin csv_struct = {
 200         "csv",                  /* builtin name */
 201         csv_builtin,            /* function implementing the builtin */
 202         BUILTIN_ENABLED,        /* initial flags for builtin */
 203         csv_doc,                /* array of long documentation strings. */
 204         "csv [-a ARRAY] string",        /* usage synopsis; becomes short_doc */
 205         0                       /* reserved for internal use */
 206 };