New option --ignore-case.

15 files changed, 583 insertions, 39 deletions

diff --git a/ChangeLog b/ChangeLog
index f552e91..746d66e 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,49 @@
 2002-12-10  Bruno Haible  <bruno@clisp.org>
 
+	* src/options.h (UPPERLOWER): New enum value.
+	* src/options.cc (Options::long_usage): Document option --ignore-case.
+	(Options::~Options): Update.
+	(long_options): Add option --ignore-case.
+	(Options::parse_options): Handle option --ignore-case.
+	* src/input.cc (Input::read_input): Recognize option %ignore-case.
+	* src/keyword.h (KeywordExt::init_selchars_tuple,
+	KeywordExt::init_selchars_multiset, KeywordExt::init_selchars_low):
+	Add alpha_unify argument.
+	* src/keyword.cc (KeywordExt::init_selchars_low): Add alpha_unify
+	argument.
+	(KeywordExt::init_selchars_tuple): Add alpha_unify argument.
+	(KeywordExt::init_selchars_multiset): Add alpha_unify argument.
+	* src/search.h (Search::compute_alpha_size,
+	Search::compute_alpha_unify): New declarations.
+	(Search::init_selchars_multiset): Add alpha_unify argument.
+	(Search::_alpha_unify): New field.
+	* src/search.cc (Search::compute_alpha_size,
+	Search::compute_alpha_unify): New functions.
+	(Search::init_selchars_tuple): Update.
+	(Search::find_positions): Temporarily set _alpha_unify. Perform a
+	case insensitive comparison if needed.
+	(Search::init_selchars_multiset): Add alpha_unify argument.
+	(Search::count_duplicates_multiset): Call compute_alpha_unify.
+	(Search::find_alpha_inc): Temporarily set _alpha_unify. At the end,
+	set _alpha_size and _alpha_unify.
+	(Search::prepare): Update. Don't compute _alpga_size here.
+	(Search::optimize): Propagate unified asso_values.
+	(Search::~Search) Delete _alpha_unify.
+	* src/output.cc (output_upperlower_strcmp, output_upperlower_strncmp,
+	output_upperlower_memcmp): New functions.
+	(Output_Compare_Strcmp::output_comparison,
+	Output_Compare_Strncmp::output_comparison,
+	Output_Compare_Memcmp::output_comparison): Use the case-insensitive
+	comparison function if --ignore-case was given.
+	(Output::output): Emit the auxiliary case-insensitive comparison
+	function if needed.
+	* tests/permutc2.gperf, tests/permutc2.exp: New files.
+	* tests/Makefile.in (check-test): Also check permutc2.gperf.
+	* tests/test-6.exp: Update.
+	* doc/gperf.texi (Gperf Declarations): Document %ignore-case.
+	(Input Details): Document option --ignore-case.
+	* NEWS: Update.
+
 	* src/search.cc (Search::optimize): Fill unused asso_values[] entries
 	with a large value.
 	* src/output.h (Output::Output): Remove occurrences argument.
diff --git a/NEWS b/NEWS
index 2926113..74acb0f 100644
--- a/NEWS
+++ b/NEWS
@@ -10,6 +10,7 @@ New in 2.97:
 * The following options can now be specified inside the input file:
     %delimiters=DELIMITER-LIST
     %struct-type
+    %ignore-case
     %language=LANGUAGE-NAME
     %define slot-name NAME
     %define hash-function-name NAME
@@ -42,6 +43,7 @@ New in 2.97:
 * The options -f/--fast and -o/--occurrence-sort have no effect any more.
 * Added option -P/--pic that optimizes the generated code for use in shared
   libraries.
+* Added option --ignore-case that produces a case independent lookup function.
 * Bug fixes.
 
 New in 2.7.2:
diff --git a/doc/gperf.texi b/doc/gperf.texi
index b9ce792..56676ce 100644
--- a/doc/gperf.texi
+++ b/doc/gperf.texi
@@ -436,6 +436,12 @@ commas or newlines.
 Allows you to include a @code{struct} type declaration for generated
 code; see above for an example.
 
+@item %ignore-case
+@cindex @samp{%ignore-case}
+Consider upper and lower case ASCII characters as equivalent.  The string
+comparison will use a case insignificant character comparison.  Note that
+locale dependent case mappings are ignored.
+
 @item %language=@var{language-name}
 @cindex @samp{%language}
 Instructs @code{gperf} to generate code in the language specified by the
@@ -790,6 +796,17 @@ part of the type declaration.  Keywords and additional fields may follow
 this, one group of fields per line.  A set of examples for generating
 perfect hash tables and functions for Ada, C, C++, Pascal, Modula 2,
 Modula 3 and JavaScript reserved words are distributed with this release.
+
+@item --ignore-case
+Consider upper and lower case ASCII characters as equivalent.  The string
+comparison will use a case insignificant character comparison.  Note that
+locale dependent case mappings are ignored.  This option is therefore not
+suitable if a properly internationalized or locale aware case mapping
+should be used.  (For example, in a Turkish locale, the upper case equivalent
+of the lowercase ASCII letter @samp{i} is the non-ASCII character
+@samp{capital i with dot above}.)  For this case, it is better to apply
+an uppercase or lowercase conversion on the string before passing it to
+the @code{gperf} generated function.
 @end table
 
 @node Output Language, Output Details, Input Details, Options
diff --git a/src/input.cc b/src/input.cc
index 3c81ba4..933911e 100644
--- a/src/input.cc
+++ b/src/input.cc
@@ -472,6 +472,10 @@ Input::read_input ()
                   option.set (TYPE);
                 else
 
+                if (is_declaration (line, line_end, lineno, "ignore-case"))
+                  option.set (UPPERLOWER);
+                else
+
                 if (is_declaration_with_arg (line, line_end, lineno,
                                              "language", &arg))
                   option.set_language (arg);
diff --git a/src/keyword.cc b/src/keyword.cc
index a784e07..2d7904e 100644
--- a/src/keyword.cc
+++ b/src/keyword.cc
@@ -59,7 +59,7 @@ static inline void sort_char_set (unsigned int *base, int len)
  */
 
 unsigned int *
-KeywordExt::init_selchars_low (bool use_all_chars, const Positions& positions, const unsigned int *alpha_inc)
+KeywordExt::init_selchars_low (bool use_all_chars, const Positions& positions, const unsigned int *alpha_unify, const unsigned int *alpha_inc)
 {
   const char *k = _allchars;
   unsigned int *key_set =
@@ -73,6 +73,8 @@ KeywordExt::init_selchars_low (bool use_all_chars, const Positions& positions, c
         unsigned int c = static_cast<unsigned char>(*k);
         if (alpha_inc)
           c += alpha_inc[k-_allchars];
+        if (alpha_unify)
+          c = alpha_unify[c];
         *ptr = c;
         ptr++;
       }
@@ -99,6 +101,8 @@ KeywordExt::init_selchars_low (bool use_all_chars, const Positions& positions, c
           else
             /* Out of range of KEY length, so we'll just skip it.  */
             continue;
+          if (alpha_unify)
+            c = alpha_unify[c];
           *ptr = c;
           ptr++;
         }
@@ -111,16 +115,16 @@ KeywordExt::init_selchars_low (bool use_all_chars, const Positions& positions, c
 }
 
 void
-KeywordExt::init_selchars_tuple (bool use_all_chars, const Positions& positions)
+KeywordExt::init_selchars_tuple (bool use_all_chars, const Positions& positions, const unsigned int *alpha_unify)
 {
-  init_selchars_low (use_all_chars, positions, NULL);
+  init_selchars_low (use_all_chars, positions, alpha_unify, NULL);
 }
 
 void
-KeywordExt::init_selchars_multiset (bool use_all_chars, const Positions& positions, const unsigned int *alpha_inc)
+KeywordExt::init_selchars_multiset (bool use_all_chars, const Positions& positions, const unsigned int *alpha_unify, const unsigned int *alpha_inc)
 {
   unsigned int *selchars =
-    init_selchars_low (use_all_chars, positions, alpha_inc);
+    init_selchars_low (use_all_chars, positions, alpha_unify, alpha_inc);
 
   /* Sort the selchars elements alphabetically.  */
   sort_char_set (selchars, _selchars_length);
diff --git a/src/keyword.h b/src/keyword.h
index 02e0364..9f09186 100644
--- a/src/keyword.h
+++ b/src/keyword.h
@@ -68,9 +68,9 @@ struct KeywordExt : public Keyword
 
   /* Methods depending on the keyposition list.  */
   /* Initializes selchars and selchars_length, without reordering.  */
-  void                  init_selchars_tuple (bool use_all_chars, const Positions& positions);
+  void                  init_selchars_tuple (bool use_all_chars, const Positions& positions, const unsigned int *alpha_unify);
   /* Initializes selchars and selchars_length, with reordering.  */
-  void                  init_selchars_multiset (bool use_all_chars, const Positions& positions, const unsigned int *alpha_inc);
+  void                  init_selchars_multiset (bool use_all_chars, const Positions& positions, const unsigned int *alpha_unify, const unsigned int *alpha_inc);
   /* Deletes selchars.  */
   void                  delete_selchars ();
 
@@ -81,7 +81,7 @@ struct KeywordExt : public Keyword
   int                   _final_index;
 
 private:
-  unsigned int *        init_selchars_low (bool use_all_chars, const Positions& positions, const unsigned int *alpha_inc);
+  unsigned int *        init_selchars_low (bool use_all_chars, const Positions& positions, const unsigned int *alpha_unify, const unsigned int *alpha_inc);
 };
 
 /* An abstract factory for creating Keyword instances.
diff --git a/src/options.cc b/src/options.cc
index 29d0b7b..7e533cd 100644
--- a/src/options.cc
+++ b/src/options.cc
@@ -106,6 +106,10 @@ Options::long_usage (FILE * stream) const
            "                         is considered part of the type declaration. Key\n"
            "                         words and additional fields may follow this, one\n"
            "                         group of fields per line.\n");
+  fprintf (stream,
+           "      --ignore-case      Consider upper and lower case ASCII characters as\n"
+           "                         equivalent. Note that locale dependent case mappings\n"
+           "                         are ignored.\n");
   fprintf (stream, "\n");
   fprintf (stream,
            "Language for the output code:\n");
@@ -463,6 +467,7 @@ Options::~Options ()
                "\nENUM is........: %s"
                "\nINCLUDE is.....: %s"
                "\nSEVENBIT is....: %s"
+               "\nUPPERLOWER is..: %s"
                "\nlookup function name = %s"
                "\nhash function name = %s"
                "\nword list name = %s"
@@ -492,6 +497,7 @@ Options::~Options ()
                _option_word & ENUM ? "enabled" : "disabled",
                _option_word & INCLUDE ? "enabled" : "disabled",
                _option_word & SEVENBIT ? "enabled" : "disabled",
+               _option_word & UPPERLOWER ? "enabled" : "disabled",
                _function_name, _hash_name, _wordlist_name, _slot_name,
                _initializer_suffix, _asso_iterations, _jump, _size_multiple,
                _initial_asso_value, _delimiters, _total_switches);
@@ -605,6 +611,7 @@ Options::set_delimiters (const char *delimiters)
 static const struct option long_options[] =
 {
   { "output-file", required_argument, NULL, CHAR_MAX + 1 },
+  { "ignore-case", no_argument, NULL, CHAR_MAX + 2 },
   { "delimiters", required_argument, NULL, 'e' },
   { "struct-type", no_argument, NULL, 't' },
   { "language", required_argument, NULL, 'L' },
@@ -949,6 +956,11 @@ warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.\n\
             _output_file_name = /*getopt*/optarg;
             break;
           }
+        case CHAR_MAX + 2:      /* Case insignificant.  */
+          {
+            _option_word |= UPPERLOWER;
+            break;
+          }
         default:
           short_usage (stderr);
           exit (1);
diff --git a/src/options.h b/src/options.h
index 2a9567b..4dda513 100644
--- a/src/options.h
+++ b/src/options.h
@@ -98,7 +98,10 @@ enum Option_Type
   SEVENBIT     = 1 << 19,
 
   /* Optimize for position-independent code.  */
-  SHAREDLIB    = 1 << 20
+  SHAREDLIB    = 1 << 20,
+
+  /* Ignore case of ASCII characters.  */
+  UPPERLOWER   = 1 << 21
 };
 
 /* Class manager for gperf program Options. */
diff --git a/src/output.cc b/src/output.cc
index a92a9a3..6e0d688 100644
--- a/src/output.cc
+++ b/src/output.cc
@@ -232,6 +232,131 @@ Output::output_constants (struct Output_Constants& style) const
 
 /* ------------------------------------------------------------------------- */
 
+/* Output gperf's ASCII-case insensitive strcmp replacement.  */
+
+static void
+output_upperlower_strcmp ()
+{
+  printf ("#ifndef GPERF_CASE_STRCMP\n"
+          "#define GPERF_CASE_STRCMP 1\n"
+          "static int\n"
+          "gperf_case_strcmp ");
+  printf (option[KRC] ?
+               "(s1, s2)\n"
+          "     register char *s1;\n"
+          "     register char *s2;\n" :
+          option[C] ?
+               "(s1, s2)\n"
+          "     register const char *s1;\n"
+          "     register const char *s2;\n" :
+          option[ANSIC] | option[CPLUSPLUS] ?
+               "(register const char *s1, register const char *s2)\n" :
+          "");
+  printf ("{\n"
+          "  for (;;)\n"
+          "    {\n"
+          "      unsigned char c1 = *s1++;\n"
+          "      unsigned char c2 = *s2++;\n"
+          "      if (c1 >= 'A' && c1 <= 'Z')\n"
+          "        c1 += 'a' - 'A';\n"
+          "      if (c2 >= 'A' && c2 <= 'Z')\n"
+          "        c2 += 'a' - 'A';\n"
+          "      if (c1 != 0 && c1 == c2)\n"
+          "        continue;\n"
+          "      return (int)c1 - (int)c2;\n"
+          "    }\n"
+          "}\n"
+          "#endif\n\n");
+}
+
+/* Output gperf's ASCII-case insensitive strncmp replacement.  */
+
+static void
+output_upperlower_strncmp ()
+{
+  printf ("#ifndef GPERF_CASE_STRNCMP\n"
+          "#define GPERF_CASE_STRNCMP 1\n"
+          "static int\n"
+          "gperf_case_strncmp ");
+  printf (option[KRC] ?
+               "(s1, s2, n)\n"
+          "     register char *s1;\n"
+          "     register char *s2;\n"
+          "     register unsigned int n;\n" :
+          option[C] ?
+               "(s1, s2, n)\n"
+          "     register const char *s1;\n"
+          "     register const char *s2;\n"
+          "     register unsigned int n;\n" :
+          option[ANSIC] | option[CPLUSPLUS] ?
+               "(register const char *s1, register const char *s2, register unsigned int n)\n" :
+          "");
+  printf ("{\n"
+          "  for (; n > 0;)\n"
+          "    {\n"
+          "      unsigned char c1 = *s1++;\n"
+          "      unsigned char c2 = *s2++;\n"
+          "      if (c1 >= 'A' && c1 <= 'Z')\n"
+          "        c1 += 'a' - 'A';\n"
+          "      if (c2 >= 'A' && c2 <= 'Z')\n"
+          "        c2 += 'a' - 'A';\n"
+          "      if (c1 != 0 && c1 == c2)\n"
+          "        {\n"
+          "          n--;\n"
+          "          continue;\n"
+          "        }\n"
+          "      return (int)c1 - (int)c2;\n"
+          "    }\n"
+          "  return 0;\n"
+          "}\n"
+          "#endif\n\n");
+}
+
+/* Output gperf's ASCII-case insensitive memcmp replacement.  */
+
+static void
+output_upperlower_memcmp ()
+{
+  printf ("#ifndef GPERF_CASE_MEMCMP\n"
+          "#define GPERF_CASE_MEMCMP 1\n"
+          "static int\n"
+          "gperf_case_memcmp ");
+  printf (option[KRC] ?
+               "(s1, s2, n)\n"
+          "     register char *s1;\n"
+          "     register char *s2;\n"
+          "     register unsigned int n;\n" :
+          option[C] ?
+               "(s1, s2, n)\n"
+          "     register const char *s1;\n"
+          "     register const char *s2;\n"
+          "     register unsigned int n;\n" :
+          option[ANSIC] | option[CPLUSPLUS] ?
+               "(register const char *s1, register const char *s2, register unsigned int n)\n" :
+          "");
+  printf ("{\n"
+          "  for (; n > 0;)\n"
+          "    {\n"
+          "      unsigned char c1 = *s1++;\n"
+          "      unsigned char c2 = *s2++;\n"
+          "      if (c1 >= 'A' && c1 <= 'Z')\n"
+          "        c1 += 'a' - 'A';\n"
+          "      if (c2 >= 'A' && c2 <= 'Z')\n"
+          "        c2 += 'a' - 'A';\n"
+          "      if (c1 == c2)\n"
+          "        {\n"
+          "          n--;\n"
+          "          continue;\n"
+          "        }\n"
+          "      return (int)c1 - (int)c2;\n"
+          "    }\n"
+          "  return 0;\n"
+          "}\n"
+          "#endif\n\n");
+}
+
+/* ------------------------------------------------------------------------- */
+
 /* Outputs a keyword, as a string: enclosed in double quotes, escaping
    backslashes, double quote and unprintable characters.  */
 
@@ -363,7 +488,10 @@ void Output_Compare_Strcmp::output_comparison (const Output_Expr& expr1,
   expr1.output_expr ();
   printf (" == *");
   expr2.output_expr ();
-  printf (" && !strcmp (");
+  printf (" && !");
+  if (option[UPPERLOWER])
+    printf ("gperf_case_");
+  printf ("strcmp (");
   expr1.output_expr ();
   printf (" + 1, ");
   expr2.output_expr ();
@@ -389,7 +517,10 @@ void Output_Compare_Strncmp::output_comparison (const Output_Expr& expr1,
   expr1.output_expr ();
   printf (" == *");
   expr2.output_expr ();
-  printf (" && !strncmp (");
+  printf (" && !");
+  if (option[UPPERLOWER])
+    printf ("gperf_case_");
+  printf ("strncmp (");
   expr1.output_expr ();
   printf (" + 1, ");
   expr2.output_expr ();
@@ -418,7 +549,10 @@ void Output_Compare_Memcmp::output_comparison (const Output_Expr& expr1,
   expr1.output_expr ();
   printf (" == *");
   expr2.output_expr ();
-  printf (" && !memcmp (");
+  printf (" && !");
+  if (option[UPPERLOWER])
+    printf ("gperf_case_");
+  printf ("memcmp (");
   expr1.output_expr ();
   printf (" + 1, ");
   expr2.output_expr ();
@@ -1522,6 +1656,19 @@ Output::output ()
   printf ("/* maximum key range = %d, duplicates = %d */\n\n",
           _max_hash_value - _min_hash_value + 1, _total_duplicates);
 
+  if (option[UPPERLOWER])
+    {
+      if (option[LENTABLE])
+        output_upperlower_memcmp ();
+      else
+        {
+          if (option[COMP])
+            output_upperlower_strncmp ();
+          else
+            output_upperlower_strcmp ();
+        }
+    }
+
   if (option[CPLUSPLUS])
     printf ("class %s\n"
             "{\n"
diff --git a/src/search.cc b/src/search.cc
index 730a1ac..7ceb26c 100644
--- a/src/search.cc
+++ b/src/search.cc
@@ -146,12 +146,39 @@ Search::preprepare ()
 
 /* ====================== Finding good byte positions ====================== */
 
+/* Computes the upper bound on the indices passed to asso_values[],
+   assuming no alpha_increments.  */
+unsigned int
+Search::compute_alpha_size () const
+{
+  return (option[SEVENBIT] ? 128 : 256);
+}
+
+/* Computes the unification rules between different asso_values[c],
+   assuming no alpha_increments.  */
+unsigned int *
+Search::compute_alpha_unify () const
+{
+  if (option[UPPERLOWER])
+    {
+      unsigned int alpha_size = compute_alpha_size();
+      unsigned int *alpha_unify = new unsigned int[alpha_size];
+      for (unsigned int c = 0; c < alpha_size; c++)
+        alpha_unify[c] = c;
+      for (unsigned int c = 'A'; c <= 'Z'; c++)
+        alpha_unify[c] = c + ('a'-'A');
+      return alpha_unify;
+    }
+  else
+    return NULL;
+}
+
 /* Initializes each keyword's _selchars array.  */
 void
 Search::init_selchars_tuple (bool use_all_chars, const Positions& positions) const
 {
   for (KeywordExt_List *temp = _head; temp; temp = temp->rest())
-    temp->first()->init_selchars_tuple(use_all_chars, positions);
+    temp->first()->init_selchars_tuple(use_all_chars, positions, _alpha_unify);
 }
 
 /* Deletes each keyword's _selchars array.  */
@@ -202,6 +229,9 @@ Search::find_positions ()
       return;
     }
 
+  /* Compute preliminary value for _alpha_unify.  */
+  _alpha_unify = compute_alpha_unify ();
+
   /* 1. Find positions that must occur in order to distinguish duplicates.  */
   Positions mandatory;
 
@@ -222,17 +252,42 @@ Search::find_positions ()
                   int n = keyword1->_allchars_length;
                   int i;
                   for (i = 1; i < n; i++)
-                    if (keyword1->_allchars[i-1] != keyword2->_allchars[i-1])
-                      break;
-                  if (i < n
-                      && memcmp (&keyword1->_allchars[i],
-                                 &keyword2->_allchars[i],
-                                 n - i)
-                         == 0)
                     {
-                      /* Position i is mandatory.  */
-                      if (!mandatory.contains (i))
-                        mandatory.add (i);
+                      unsigned char c1 = keyword1->_allchars[i-1];
+                      unsigned char c2 = keyword2->_allchars[i-1];
+                      if (option[UPPERLOWER])
+                        {
+                          if (c1 >= 'A' && c1 <= 'Z')
+                            c1 += 'a' - 'A';
+                          if (c2 >= 'A' && c2 <= 'Z')
+                            c2 += 'a' - 'A';
+                        }
+                      if (c1 != c2)
+                        break;
+                    }
+                  if (i < n)
+                    {
+                      int j;
+                      for (j = i + 1; j <= n; j++)
+                        {
+                          unsigned char c1 = keyword1->_allchars[j-1];
+                          unsigned char c2 = keyword2->_allchars[j-1];
+                          if (option[UPPERLOWER])
+                            {
+                              if (c1 >= 'A' && c1 <= 'Z')
+                                c1 += 'a' - 'A';
+                              if (c2 >= 'A' && c2 <= 'Z')
+                                c2 += 'a' - 'A';
+                            }
+                          if (c1 != c2)
+                            break;
+                        }
+                      if (j > n)
+                        {
+                          /* Position i is mandatory.  */
+                          if (!mandatory.contains (i))
+                            mandatory.add (i);
+                        }
                     }
                 }
             }
@@ -379,16 +434,113 @@ Search::find_positions ()
         }
       fprintf (stderr, "\n");
     }
+
+  /* Free preliminary value for _alpha_unify.  */
+  delete[] _alpha_unify;
 }
 
 /* ===================== Finding good alpha increments ===================== */
 
+/* Computes the upper bound on the indices passed to asso_values[].  */
+unsigned int
+Search::compute_alpha_size (const unsigned int *alpha_inc) const
+{
+  unsigned int max_alpha_inc = 0;
+  for (int i = 0; i < _max_key_len; i++)
+    if (max_alpha_inc < alpha_inc[i])
+      max_alpha_inc = alpha_inc[i];
+  return (option[SEVENBIT] ? 128 : 256) + max_alpha_inc;
+}
+
+/* Computes the unification rules between different asso_values[c].  */
+unsigned int *
+Search::compute_alpha_unify (const Positions& positions, const unsigned int *alpha_inc) const
+{
+  if (option[UPPERLOWER])
+    {
+      /* Without alpha increments, we would simply unify
+           'A' -> 'a', ..., 'Z' -> 'z'.
+         But when a keyword contains at position i a character c,
+         we have the constraint
+            asso_values[tolower(c) + alpha_inc[i]] ==
+            asso_values[toupper(c) + alpha_inc[i]].
+         This introduces a unification
+           toupper(c) + alpha_inc[i] -> tolower(c) + alpha_inc[i].
+         Note that this unification can extend outside the range of
+         ASCII letters!  But still every unified character pair is at
+         a distance of 'a'-'A' = 32, or (after chained unification)
+         at a multiple of 32.  So in the end the alpha_unify vector has
+         the form    c -> c + 32 * f(c)   where f(c) is a nonnegative
+         integer.  */
+      unsigned int alpha_size = compute_alpha_size (alpha_inc);
+
+      unsigned int *alpha_unify = new unsigned int[alpha_size];
+      for (unsigned int c = 0; c < alpha_size; c++)
+        alpha_unify[c] = c;
+
+      for (KeywordExt_List *temp = _head; temp; temp = temp->rest())
+        {
+          KeywordExt *keyword = temp->first();
+
+          if (option[ALLCHARS])
+            /* Iterate through all character positions.  */
+            for (int i = 0; i < keyword->_allchars_length; i++)
+              {
+                unsigned int c = static_cast<unsigned char>(keyword->_allchars[i]);
+                if (c >= 'A' && c <= 'Z')
+                  c += 'a' - 'A';
+                if (c >= 'a' && c <= 'z')
+                  {
+                    c += alpha_inc[i];
+                    /* Unify c with c - ('a'-'A').  */
+                    unsigned int d = alpha_unify[c];
+                    unsigned int b = c - ('a'-'A');
+                    for (int a = b; a >= 0 && alpha_unify[a] == b; a -= ('a'-'A'))
+                      alpha_unify[a] = d;
+                  }
+              }
+          else
+            {
+              /* Iterate through the selected character positions.  */
+              PositionIterator iter (positions);
+
+              for (int i; (i = iter.next ()) != PositionIterator::EOS; )
+                {
+                  unsigned int c;
+                  if (i == Positions::LASTCHAR)
+                    c = static_cast<unsigned char>(keyword->_allchars[keyword->_allchars_length - 1]);
+                  else if (i <= keyword->_allchars_length)
+                    c = static_cast<unsigned char>(keyword->_allchars[i - 1]);
+                  else
+                    continue;
+                  if (c >= 'A' && c <= 'Z')
+                    c += 'a' - 'A';
+                  if (c >= 'a' && c <= 'z')
+                    {
+                      if (i != Positions::LASTCHAR)
+                        c += alpha_inc[i - 1];
+                      /* Unify c with c - ('a'-'A').  */
+                      unsigned int d = alpha_unify[c];
+                      unsigned int b = c - ('a'-'A');
+                      for (int a = b; a >= 0 && alpha_unify[a] == b; a -= ('a'-'A'))
+                        alpha_unify[a] = d;
+                    }
+                }
+            }
+        }
+      return alpha_unify;
+    }
+  else
+    /* Identity mapping.  */
+    return NULL;
+}
+
 /* Initializes each keyword's _selchars array.  */
 void
-Search::init_selchars_multiset (bool use_all_chars, const Positions& positions, const unsigned int *alpha_inc) const
+Search::init_selchars_multiset (bool use_all_chars, const Positions& positions, const unsigned int *alpha_unify, const unsigned int *alpha_inc) const
 {
   for (KeywordExt_List *temp = _head; temp; temp = temp->rest())
-    temp->first()->init_selchars_multiset(use_all_chars, positions, alpha_inc);
+    temp->first()->init_selchars_multiset(use_all_chars, positions, alpha_unify, alpha_inc);
 }
 
 /* Count the duplicate keywords that occur with the given set of positions
@@ -402,7 +554,9 @@ Search::count_duplicates_multiset (const unsigned int *alpha_inc) const
   /* Run through the keyword list and count the duplicates incrementally.
      The result does not depend on the order of the keyword list, thanks to
      the formula above.  */
-  init_selchars_multiset (option[ALLCHARS], _key_positions, alpha_inc);
+  init_selchars_multiset (option[ALLCHARS], _key_positions,
+                          compute_alpha_unify (_key_positions, alpha_inc),
+                          alpha_inc);
 
   unsigned int count = 0;
   {
@@ -428,7 +582,9 @@ Search::find_alpha_inc ()
   /* The goal is to choose _alpha_inc[] such that it doesn't introduce
      artificial duplicates.
      In other words, the goal is  # proj2 (proj1 (K)) = # proj1 (K).  */
+  _alpha_unify = compute_alpha_unify ();
   unsigned int duplicates_goal = count_duplicates_tuple (_key_positions);
+  delete[] _alpha_unify;
 
   /* Start with zero increments.  This is sufficient in most cases.  */
   unsigned int *current = new unsigned int [_max_key_len];
@@ -545,6 +701,8 @@ Search::find_alpha_inc ()
     }
 
   _alpha_inc = current;
+  _alpha_size = compute_alpha_size (_alpha_inc);
+  _alpha_unify = compute_alpha_unify (_key_positions, _alpha_inc);
 }
 
 /* ======================= Finding good asso_values ======================== */
@@ -555,7 +713,8 @@ Search::prepare ()
   KeywordExt_List *temp;
 
   /* Initialize each keyword's _selchars array.  */
-  init_selchars_multiset(option[ALLCHARS], _key_positions, _alpha_inc);
+  init_selchars_multiset(option[ALLCHARS], _key_positions,
+                         _alpha_unify, _alpha_inc);
 
   /* Check for duplicates, i.e. keywords with the same _selchars array
      (and - if !option[NOLENGTH] - also the same length).
@@ -634,14 +793,6 @@ Search::prepare ()
         }
     }
 
-  /* Compute _alpha_size, the upper bound on the indices passed to
-     asso_values[].  */
-  unsigned int max_alpha_inc = 0;
-  for (int i = 0; i < _max_key_len; i++)
-    if (max_alpha_inc < _alpha_inc[i])
-      max_alpha_inc = _alpha_inc[i];
-  _alpha_size = (option[SEVENBIT] ? 128 : 256) + max_alpha_inc;
-
   /* Compute the occurrences of each character in the alphabet.  */
   _occurrences = new int[_alpha_size];
   memset (_occurrences, 0, _alpha_size * sizeof (_occurrences[0]));
@@ -1492,6 +1643,12 @@ Search::optimize ()
   for (unsigned int c = 0; c < _alpha_size; c++)
     if (_occurrences[c] == 0)
       _asso_values[c] = max_hash_value + 1;
+
+  /* Propagate unified asso_values.  */
+  if (_alpha_unify)
+    for (unsigned int c = 0; c < _alpha_size; c++)
+      if (_alpha_unify[c] != c)
+        _asso_values[c] = _asso_values[_alpha_unify[c]];
 }
 
 /* Prints out some diagnostics upon completion.  */
@@ -1533,5 +1690,6 @@ Search::~Search ()
     }
   delete[] _asso_values;
   delete[] _occurrences;
+  delete[] _alpha_unify;
   delete[] _alpha_inc;
 }
diff --git a/src/search.h b/src/search.h
index ada14d2..0cd2829 100644
--- a/src/search.h
+++ b/src/search.h
@@ -41,6 +41,14 @@ public:
 private:
   void                  preprepare ();
 
+  /* Computes the upper bound on the indices passed to asso_values[],
+     assuming no alpha_increments.  */
+  unsigned int          compute_alpha_size () const;
+
+  /* Computes the unification rules between different asso_values[c],
+     assuming no alpha_increments.  */
+  unsigned int *        compute_alpha_unify () const;
+
   /* Initializes each keyword's _selchars array.  */
   void                  init_selchars_tuple (bool use_all_chars, const Positions& positions) const;
   /* Deletes each keyword's _selchars array.  */
@@ -52,8 +60,14 @@ private:
   /* Find good key positions.  */
   void                  find_positions ();
 
+  /* Computes the upper bound on the indices passed to asso_values[].  */
+  unsigned int          compute_alpha_size (const unsigned int *alpha_inc) const;
+
+  /* Computes the unification rules between different asso_values[c].  */
+  unsigned int *        compute_alpha_unify (const Positions& positions, const unsigned int *alpha_inc) const;
+
   /* Initializes each keyword's _selchars array.  */
-  void                  init_selchars_multiset (bool use_all_chars, const Positions& positions, const unsigned int *alpha_inc) const;
+  void                  init_selchars_multiset (bool use_all_chars, const Positions& positions, const unsigned int *alpha_unify, const unsigned int *alpha_inc) const;
 
   /* Count the duplicate keywords that occur with the given set of positions
      and a given alpha_inc[] array.  */
@@ -115,13 +129,17 @@ public:
   /* Adjustments to add to bytes add specific key positions.  */
   unsigned int *        _alpha_inc;
 
+  /* Size of alphabet.  */
+  unsigned int          _alpha_size;
+
+  /* Alphabet character unification, either the identity or a mapping from
+     upper case characters to lower case characters (and maybe more).  */
+  unsigned int *        _alpha_unify;
+
   /* Total number of duplicates that have been moved to _duplicate_link lists
      (not counting their representatives which stay on the main list).  */
   int                   _total_duplicates;
 
-  /* Size of alphabet.  */
-  unsigned int          _alpha_size;
-
   /* Counts occurrences of each key set character.
      _occurrences[c] is the number of times that c occurs among the _selchars
      of a keyword.  */
diff --git a/tests/Makefile.in b/tests/Makefile.in
index c857fc4..c7c3c37 100644
--- a/tests/Makefile.in
+++ b/tests/Makefile.in
@@ -143,6 +143,8 @@ check-test:
 	diff $(srcdir)/permut2.exp permut2.out
 	$(GPERF) -m5 < $(srcdir)/permut3.gperf > permut3.out
 	diff $(srcdir)/permut3.exp permut3.out
+	$(GPERF) -m5 --ignore-case < $(srcdir)/permutc2.gperf > permutc2.out
+	diff $(srcdir)/permutc2.exp permutc2.out
 	$(GPERF) -C -E -G -I -t < $(srcdir)/charsets.gperf > charsets.out
 	diff $(srcdir)/charsets.exp charsets.out
 	$(GPERF) -C -E -G -I -t < $(srcdir)/languages.gperf > languages.out
diff --git a/tests/permutc2.exp b/tests/permutc2.exp
new file mode 100644
index 0000000..02632d5
--- /dev/null
+++ b/tests/permutc2.exp
@@ -0,0 +1,116 @@
+/* C code produced by gperf version 2.7.2 */
+/* Command-line: ../src/gperf -m5 --ignore-case  */
+/* Computed positions: -k'1-2' */
+
+/* Test of a hash function which has to deal with permutation and
+   case-independence.  Without case-independence, the alpha_inc is 1.
+   With case-independence, the alpha_inc is 3.  */
+
+#define TOTAL_KEYWORDS 8
+#define MIN_WORD_LENGTH 2
+#define MAX_WORD_LENGTH 2
+#define MIN_HASH_VALUE 2
+#define MAX_HASH_VALUE 9
+/* maximum key range = 8, duplicates = 0 */
+
+#ifndef GPERF_CASE_STRCMP
+#define GPERF_CASE_STRCMP 1
+static int
+gperf_case_strcmp (s1, s2)
+     register const char *s1;
+     register const char *s2;
+{
+  for (;;)
+    {
+      unsigned char c1 = *s1++;
+      unsigned char c2 = *s2++;
+      if (c1 >= 'A' && c1 <= 'Z')
+        c1 += 'a' - 'A';
+      if (c2 >= 'A' && c2 <= 'Z')
+        c2 += 'a' - 'A';
+      if (c1 != 0 && c1 == c2)
+        continue;
+      return (int)c1 - (int)c2;
+    }
+}
+#endif
+
+#ifdef __GNUC__
+__inline
+#else
+#ifdef __cplusplus
+inline
+#endif
+#endif
+static unsigned int
+hash (str, len)
+     register const char *str;
+     register unsigned int len;
+{
+  static unsigned char asso_values[] =
+    {
+     10,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10, 1,10,10, 3,10,
+     10,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10,10,10,10, 3, 1,
+      0, 7, 1, 0, 3,10,10, 1,10,10,
+      3,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10,10,10,10,10,10,
+      3, 1, 0, 0, 1, 0, 2,10,10,10,
+     10,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10,10,10,10,10,10,
+     10,10,10,10,10,10,10,10,10
+    };
+  return len + asso_values[(unsigned char)str[1]+3] + asso_values[(unsigned char)str[0]];
+}
+
+#ifdef __GNUC__
+__inline
+#endif
+const char *
+in_word_set (str, len)
+     register const char *str;
+     register unsigned int len;
+{
+  static const char * wordlist[] =
+    {
+      "", "",
+      "{w",
+      "az",
+      "ay",
+      "za",
+      "ya",
+      "x{",
+      "x[",
+      "[w"
+    };
+
+  if (len <= MAX_WORD_LENGTH && len >= MIN_WORD_LENGTH)
+    {
+      register int key = hash (str, len);
+
+      if (key <= MAX_HASH_VALUE && key >= 0)
+        {
+          register const char *s = wordlist[key];
+
+          if (*str == *s && !gperf_case_strcmp (str + 1, s + 1))
+            return s;
+        }
+    }
+  return 0;
+}
diff --git a/tests/permutc2.gperf b/tests/permutc2.gperf
new file mode 100644
index 0000000..223a193
--- /dev/null
+++ b/tests/permutc2.gperf
@@ -0,0 +1,14 @@
+%{
+/* Test of a hash function which has to deal with permutation and
+   case-independence.  Without case-independence, the alpha_inc is 1.
+   With case-independence, the alpha_inc is 3.  */
+%}
+%%
+az
+za
+ay
+ya
+x{
+x[
+{w
+[w
diff --git a/tests/test-6.exp b/tests/test-6.exp
index 14cd1d7..cabbe9c 100644
--- a/tests/test-6.exp
+++ b/tests/test-6.exp
@@ -20,6 +20,9 @@ Input file interpretation:
                          is considered part of the type declaration. Key
                          words and additional fields may follow this, one
                          group of fields per line.
+      --ignore-case      Consider upper and lower case ASCII characters as
+                         equivalent. Note that locale dependent case mappings
+                         are ignored.
 
 Language for the output code:
   -L, --language=LANGUAGE-NAME