[Helper] Add -normalize-match option.

Decomposes string and remove accent characters before matching.
This makes o match ö, é match e and more.
It is not a perfect implementation but works.
Currently disables the match highlighting.

Fixes: #1119

2 files changed, 50 insertions, 2 deletions

diff --git a/source/helper.c b/source/helper.c
index a7bfa3e8..0a5e86ca 100644
--- a/source/helper.c
+++ b/source/helper.c
@@ -180,10 +180,43 @@ static gchar *fuzzy_to_regex ( const char * input )
     return retv;
 }
 
+static char *utf8_helper_simplify_string ( const char *s)
+{
+  gunichar buf2[G_UNICHAR_MAX_DECOMPOSITION_LENGTH] = {0,};
+  char buf[6] = {0,};
+  // Compose the string in maximally composed form.
+  char * str = g_malloc0((g_utf8_strlen(s,0)*6+2));
+  char *striter = str;
+  for ( const char *iter = s; iter && *iter; iter = g_utf8_next_char ( iter )) {
+    gunichar uc = g_utf8_get_char ( iter );
+    int l = 0;
+    gsize dl = g_unichar_fully_decompose ( uc, FALSE, buf2, G_UNICHAR_MAX_DECOMPOSITION_LENGTH) ;
+    if ( dl ) {
+      l = g_unichar_to_utf8 ( buf2[0], buf);
+    } else {
+      l = g_unichar_to_utf8 ( uc, buf);
+    }
+    memcpy(striter, buf, l);
+    striter+=l;
+  }
+
+  return str;
+}
+
 // Macro for quickly generating regex for matching.
 static inline GRegex * R ( const char *s, int case_sensitive  )
 {
+  if ( config.normalize_match ) {
+    char *str = utf8_helper_simplify_string ( s );
+
+    GRegex *r = g_regex_new ( str, G_REGEX_OPTIMIZE | ( ( case_sensitive ) ? 0 : G_REGEX_CASELESS ), 0, NULL );
+
+    g_free ( str );
+    return r;
+  } else {
     return g_regex_new ( s, G_REGEX_OPTIMIZE | ( ( case_sensitive ) ? 0 : G_REGEX_CASELESS ), 0, NULL );
+
+  }
 }
 
 static rofi_int_matcher * create_regex ( const char *input, int case_sensitive )
@@ -380,6 +413,10 @@ int find_arg_char ( const char * const key, char *val )
 
 PangoAttrList *helper_token_match_get_pango_attr ( RofiHighlightColorStyle th, rofi_int_matcher**tokens, const char *input, PangoAttrList *retv )
 {
+    // Disable highlighting for normalize match, not supported atm.
+    if ( config.normalize_match ) {
+      return retv;
+    }
     // Do a tokenized match.
     if ( tokens ) {
         for ( int j = 0; tokens[j]; j++ ) {
@@ -453,10 +490,19 @@ int helper_token_match ( rofi_int_matcher* const *tokens, const char *input )
     int match = TRUE;
     // Do a tokenized match.
     if ( tokens ) {
+      if ( config.normalize_match ) {
+        char *r = utf8_helper_simplify_string(input);
+        for ( int j = 0; match && tokens[j]; j++ ) {
+          match  = g_regex_match ( tokens[j]->regex, r, 0, NULL );
+          match ^= tokens[j]->invert;
+        }
+        g_free(r);
+      } else {
         for ( int j = 0; match && tokens[j]; j++ ) {
-            match  = g_regex_match ( tokens[j]->regex, input, 0, NULL );
-            match ^= tokens[j]->invert;
+          match  = g_regex_match ( tokens[j]->regex, input, 0, NULL );
+          match ^= tokens[j]->invert;
         }
+      }
     }
     return match;
 }
diff --git a/source/xrmoptions.c b/source/xrmoptions.c
index aff2ec9b..9e4289d0 100644
--- a/source/xrmoptions.c
+++ b/source/xrmoptions.c
@@ -231,6 +231,8 @@ static XrmOption xrmOptions[] = {
       "DRUN: build and use a cache with desktop file content.", CONFIG_DEFAULT },
     { xrm_Boolean, "drun-reload-desktop-cache", { .snum  = &config.drun_reload_desktop_cache            }, NULL,
       "DRUN: If enabled, reload the cache with desktop file content.", CONFIG_DEFAULT },
+    { xrm_Boolean, "normalize-match", 		      { .snum  = &config.normalize_match                      }, NULL,
+            "Normalize string when matching (implies -no-show-match).", CONFIG_DEFAULT },
 };
 
 /** Dynamic array of extra options */