src/data/short-names.c

   1 /* PSPP - a program for statistical analysis.
   2    Copyright (C) 2007, 2009, 2010, 2011, 2014 Free Software Foundation, Inc.
   3
   4    This program is free software: you can redistribute it and/or modify
   5    it under the terms of the GNU General Public License as published by
   6    the Free Software Foundation, either version 3 of the License, or
   7    (at your option) any later version.
   8
   9    This program is distributed in the hope that it will be useful,
  10    but WITHOUT ANY WARRANTY; without even the implied warranty of
  11    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12    GNU General Public License for more details.
  13
  14    You should have received a copy of the GNU General Public License
  15    along with this program.  If not, see <http://www.gnu.org/licenses/>. */
  16
  17 #include <config.h>
  18
  19 #include "data/short-names.h"
  20
  21 #include "data/dictionary.h"
  22 #include "data/sys-file-private.h"
  23 #include "data/variable.h"
  24 #include "libpspp/assertion.h"
  25 #include "libpspp/compiler.h"
  26 #include "libpspp/i18n.h"
  27 #include "libpspp/message.h"
  28 #include "libpspp/str.h"
  29 #include "libpspp/stringi-set.h"
  30
  31 #include "gettext.h"
  32 #define _(msgid) gettext (msgid)
  33
  34 static void
  35 claim_short_name (struct variable *v, size_t i,
  36                   struct stringi_set *short_names)
  37 {
  38   const char *short_name = var_get_short_name (v, i);
  39   if (short_name != NULL && !stringi_set_insert (short_names, short_name))
  40     var_set_short_name (v, i, NULL);
  41 }
  42
  43 /* Form initial short_name from the variable name, then try _A,
  44    _B, ... _AA, _AB, etc., if needed. */
  45 static void
  46 assign_short_name (struct variable *v, size_t i,
  47                    struct stringi_set *short_names)
  48 {
  49   int trial;
  50
  51   if (var_get_short_name (v, i) != NULL)
  52     return;
  53
  54   for (trial = 0; ; trial++)
  55     {
  56       char suffix[SHORT_NAME_LEN + 1];
  57       char *short_name;
  58
  59       /* Compose suffix. */
  60       if (trial == 0)
  61         suffix[0] = '\0';
  62       else
  63         {
  64           suffix[0] = '_';
  65           str_format_26adic (trial, true, &suffix[1], sizeof suffix - 1);
  66         }
  67
  68       /* Set name. */
  69       short_name = utf8_encoding_concat (var_get_name (v), suffix,
  70                                          var_get_encoding (v), SHORT_NAME_LEN);
  71       if (stringi_set_insert (short_names, short_name))
  72         {
  73           var_set_short_name (v, i, short_name);
  74           free (short_name);
  75           return;
  76         }
  77       free (short_name);
  78     }
  79 }
  80
  81 /* Assigns a valid, unique short_name[] to each variable in D.
  82    Each variable whose actual name is short has highest priority
  83    for that short name.  Otherwise, variables with an existing
  84    short_name[] have the next highest priority for a given short
  85    name; if it is already taken, then the variable is treated as
  86    if short_name[] had been empty.  Otherwise, long names are
  87    truncated to form short names.  If that causes conflicts,
  88    variables are renamed as PREFIX_A, PREFIX_B, and so on. */
  89 void
  90 short_names_assign (struct dictionary *d)
  91 {
  92   size_t n_vars = dict_get_n_vars (d);
  93   struct stringi_set short_names;
  94
  95   stringi_set_init (&short_names);
  96
  97   /* Clear short names that conflict with a variable name. */
  98   for (size_t i = 0; i < n_vars; i++)
  99     {
 100       struct variable *v = dict_get_var (d, i);
 101       int n_segments = sfm_width_to_segments (var_get_width (v));
 102       for (size_t j = 0; j < n_segments; j++)
 103         {
 104           const char *name = var_get_short_name (v, j);
 105           if (name != NULL)
 106             {
 107               struct variable *ov = dict_lookup_var (d, name);
 108               if (ov != NULL && (ov != v || j > 0))
 109                 var_set_short_name (v, j, NULL);
 110             }
 111         }
 112     }
 113
 114   /* Give variables whose names are short the corresponding short
 115      name. */
 116   for (size_t i = 0; i < n_vars; i++)
 117     {
 118       struct variable *v = dict_get_var (d, i);
 119       const char *name = var_get_name (v);
 120       int len = recode_string_len (var_get_encoding (v), "UTF-8", name, -1);
 121       if (len <= SHORT_NAME_LEN)
 122         var_set_short_name (v, 0, name);
 123     }
 124
 125   /* Each variable with an assigned short name for its first
 126      segment now gets it unless there is a conflict.  In case of
 127      conflict, the claimant earlier in dictionary order wins.
 128      Then similarly for additional segments of very long
 129      strings. */
 130   for (size_t i = 0; i < n_vars; i++)
 131     {
 132       struct variable *v = dict_get_var (d, i);
 133       claim_short_name (v, 0, &short_names);
 134     }
 135   for (size_t i = 0; i < n_vars; i++)
 136     {
 137       struct variable *v = dict_get_var (d, i);
 138       int n_segments = sfm_width_to_segments (var_get_width (v));
 139       for (size_t j = 1; j < n_segments; j++)
 140         claim_short_name (v, j, &short_names);
 141     }
 142
 143   /* Assign short names to first segment of remaining variables,
 144      then similarly for additional segments. */
 145   for (size_t i = 0; i < n_vars; i++)
 146     {
 147       struct variable *v = dict_get_var (d, i);
 148       assign_short_name (v, 0, &short_names);
 149     }
 150   for (size_t i = 0; i < n_vars; i++)
 151     {
 152       struct variable *v = dict_get_var (d, i);
 153       int n_segments = sfm_width_to_segments (var_get_width (v));
 154       for (size_t j = 1; j < n_segments; j++)
 155         assign_short_name (v, j, &short_names);
 156     }
 157
 158   stringi_set_destroy (&short_names);
 159 }