if someone requests 'pl' and myspell has a 'pl_PL' dictionary, return that as a 'close enough' answer

git-svn-id: svn+ssh://svn.abisource.com/svnroot/enchant/trunk@23210 bcba8976-2d24-0410-9c9c-aab3bd5fdfd6

6 files changed, 123 insertions, 61 deletions

diff --git a/configure.in b/configure.in
index d1e5c75..095e8c7 100644
--- a/configure.in
+++ b/configure.in
@@ -140,7 +140,6 @@ AC_SUBST(CC_WARN_CFLAGS)
 CXX_MAYBE_WARN="-Wall -Wextra \
 -Wsign-compare \
 -Wpointer-arith -Wwrite-strings \
--Wmissing-prototypes \
 -Wpacked -Wswitch-enum -Wmissing-format-attribute \
 -Wstrict-aliasing=2 -Winit-self -Wunsafe-loop-optimizations \
 -Wno-missing-field-initializers -Wno-unused-parameter \
diff --git a/src/enchant-provider.h b/src/enchant-provider.h
index 50599d1..ea9730d 100644
--- a/src/enchant-provider.h
+++ b/src/enchant-provider.h
@@ -55,9 +55,6 @@ ENCHANT_MODULE_EXPORT(char *)
 typedef struct str_enchant_provider EnchantProvider;
 
 ENCHANT_MODULE_EXPORT (char *)
-	enchant_get_user_home_dir (void);
-
-ENCHANT_MODULE_EXPORT (char *)
 	enchant_get_user_config_dir (void);
 
 ENCHANT_MODULE_EXPORT (char *)
diff --git a/src/enchant.c b/src/enchant.c
index a92d479..2d80f18 100644
--- a/src/enchant.c
+++ b/src/enchant.c
@@ -104,6 +104,21 @@ typedef void             (*EnchantPreConfigureFunc) (EnchantProvider * provider,
 /********************************************************************************/
 /********************************************************************************/
 
+static char *
+_enchant_get_user_home_dir (void)
+{
+	const char* home_dir;
+
+	home_dir = enchant_get_registry_value_ex (1, "Config", "Home_Dir");
+	if (home_dir)
+		return (char *)home_dir;
+
+	home_dir = g_get_home_dir ();
+	if (home_dir)
+		return g_strdup (home_dir);
+	return NULL;
+}
+
 static void
 _enchant_ensure_private_datadir (void)
 {
@@ -131,13 +146,13 @@ enchant_get_user_dir (void)
 #endif
 
 	if (!base_dir)
-		base_dir = enchant_get_user_home_dir ();
+		base_dir = _enchant_get_user_home_dir ();
 
 	if(base_dir)
 	{
-		user_dir = 	g_build_filename (base_dir,
-										  ENCHANT_USER_PATH_EXTENSION,
-										  NULL);
+		user_dir = g_build_filename (base_dir,
+					     ENCHANT_USER_PATH_EXTENSION,
+					     NULL);
 		g_free(base_dir);
 		return user_dir;
 	}
@@ -315,29 +330,6 @@ enchant_get_registry_value (const char * const prefix, const char * const key)
 	return val;
 }
 
-/**
- * enchant_get_user_home_dir
- *
- * Returns: the user's home directory, or %null. Returned value
- * must be free'd.
- *
- * This API is private to the providers.
- */
-ENCHANT_MODULE_EXPORT (char *)
-enchant_get_user_home_dir (void)
-{
-	const char* home_dir;
-
-	home_dir = enchant_get_registry_value_ex (1, "Config", "Home_Dir");
-	if (home_dir)
-		return (char *)home_dir;
-
-	home_dir = g_get_home_dir ();
-	if (home_dir)
-		return g_strdup (home_dir);
-	return NULL;
-}
-
 /********************************************************************************/
 /********************************************************************************/
 
diff --git a/src/myspell/myspell_checker.cpp b/src/myspell/myspell_checker.cpp
index bca3ced..68ced13 100644
--- a/src/myspell/myspell_checker.cpp
+++ b/src/myspell/myspell_checker.cpp
@@ -240,22 +240,18 @@ MySpellChecker::suggestWord(const char* const utf8Word, size_t len, size_t *nsug
 }
 
 static void
-s_buildHashNames (std::vector<std::string> & names, const char * dict)
+s_buildDictionaryDirs (std::vector<std::string> & dirs)
 {
-	char * tmp, * private_dir, * config_dir, * myspell_prefix, * dict_dic;
-
-	names.clear ();
+	char * private_dir, * config_dir, * myspell_prefix;
 
-	dict_dic = g_strconcat(dict, ".dic", NULL);
+	dirs.clear ();
 
 	config_dir = enchant_get_user_config_dir ();
 	if (config_dir) {
 		private_dir = g_build_filename (config_dir, 
 						"myspell", NULL);
 		
-		tmp = g_build_filename (private_dir, dict_dic, NULL);
-		names.push_back (tmp);
-		g_free (tmp);
+		dirs.push_back (private_dir);
 
 		g_free (private_dir);
 		g_free (config_dir);
@@ -263,9 +259,7 @@ s_buildHashNames (std::vector<std::string> & names, const char * dict)
 
 	myspell_prefix = myspell_checker_get_prefix ();
 	if (myspell_prefix) {
-		tmp = g_build_filename (myspell_prefix, dict_dic, NULL);
-		names.push_back (tmp);
-		g_free (tmp);
+		dirs.push_back (myspell_prefix);
 		g_free (myspell_prefix);
 	}
 
@@ -274,38 +268,71 @@ s_buildHashNames (std::vector<std::string> & names, const char * dict)
 	char* open_office_dicts_dir = myspell_checker_get_open_office_dicts_dir ();
 	if (open_office_dicts_dir) 
         {
-		    tmp = g_build_filename (open_office_dicts_dir, dict_dic, NULL);
-		    names.push_back (tmp);
-		    g_free (tmp);
-		    g_free (open_office_dicts_dir);
-	    }
+		dirs.push_back (open_office_dicts_dir);
+		g_free (open_office_dicts_dir);
+	}
     }
 #endif
+}
+
+static void
+s_buildHashNames (std::vector<std::string> & names, const char * dict)
+{
+	names.clear ();
+
+	std::vector<std::string> dirs;
+	s_buildDictionaryDirs (dirs);
 
-    g_free(dict_dic);
+	char *dict_dic = g_strconcat(dict, ".dic", NULL);
+	for (size_t i = 0; i < dirs.size(); i++)
+		{
+			char *tmp = g_build_filename (dirs[i].c_str(), dict_dic, NULL);
+			names.push_back (tmp);
+			g_free (tmp);
+		}
+
+	g_free(dict_dic);
 }
 
 static char *
 myspell_request_dictionary (const char * tag) 
 {
-	char * dic = NULL;
-
 	std::vector<std::string> names;
 
 	s_buildHashNames (names, tag);
 
-	for (size_t i = 0; i < names.size () && !dic; i++) {
+	for (size_t i = 0; i < names.size (); i++) {
 		if (g_file_test(names[i].c_str(), G_FILE_TEST_EXISTS))
-			dic = g_strdup (names[i].c_str());
+			return g_strdup (names[i].c_str());
 	}
 	
-	return dic;
+	std::vector<std::string> dirs;
+	s_buildDictionaryDirs (dirs);
+
+	for (size_t i = 0; i < dirs.size(); i++) {
+		GDir *dir = g_dir_open (dirs[i].c_str(), 0, NULL);
+		if (dir) {
+			const char *dir_entry;
+			while ((dir_entry = g_dir_read_name (dir)) != NULL) {
+				if (strncmp (dir_entry, tag, strlen(tag)) == 0 &&
+				    strstr (dir_entry, ".dic") != NULL) {
+					char *dict = g_build_filename (dirs[i].c_str(), 
+								       dir_entry, NULL);
+					g_dir_close (dir);
+					return dict;
+				}
+			}
+
+			g_dir_close (dir);
+		}
+	}
+
+	return NULL;
 }
 
 bool
 MySpellChecker::requestDictionary(const char *szLang)
 {
-	const char *dictBase = NULL;
 	char *dic = NULL, *aff = NULL;
 
 	dic = myspell_request_dictionary (szLang);
diff --git a/src/pwl.c b/src/pwl.c
index 3816fa8..0ca6395 100644
--- a/src/pwl.c
+++ b/src/pwl.c
@@ -500,12 +500,40 @@ static int enchant_is_all_caps(const char*const word, size_t len)
 			GUnicodeType type = g_unichar_type(g_utf8_get_char(it));
 			switch(type)
 				{
-					case G_UNICODE_UPPERCASE_LETTER:
-						hasCap = 1;
-						break;
-					case G_UNICODE_TITLECASE_LETTER:
-					case G_UNICODE_LOWERCASE_LETTER:
-						return 0;
+				case G_UNICODE_UPPERCASE_LETTER:
+					hasCap = 1;
+					break;
+				case G_UNICODE_TITLECASE_LETTER:
+				case G_UNICODE_LOWERCASE_LETTER:
+					return 0;
+
+				case G_UNICODE_CONTROL:
+				case G_UNICODE_FORMAT:
+				case G_UNICODE_UNASSIGNED:
+				case G_UNICODE_PRIVATE_USE:
+				case G_UNICODE_SURROGATE:
+				case G_UNICODE_MODIFIER_LETTER:
+				case G_UNICODE_OTHER_LETTER:
+				case G_UNICODE_COMBINING_MARK:
+				case G_UNICODE_ENCLOSING_MARK:
+				case G_UNICODE_NON_SPACING_MARK:
+				case G_UNICODE_DECIMAL_NUMBER:
+				case G_UNICODE_LETTER_NUMBER:
+				case G_UNICODE_OTHER_NUMBER:
+				case G_UNICODE_CONNECT_PUNCTUATION:
+				case G_UNICODE_DASH_PUNCTUATION:
+				case G_UNICODE_CLOSE_PUNCTUATION:
+				case G_UNICODE_FINAL_PUNCTUATION:
+				case G_UNICODE_INITIAL_PUNCTUATION:
+				case G_UNICODE_OTHER_PUNCTUATION:
+				case G_UNICODE_OPEN_PUNCTUATION:
+				case G_UNICODE_CURRENCY_SYMBOL:
+				case G_UNICODE_MODIFIER_SYMBOL:
+				case G_UNICODE_MATH_SYMBOL:
+				case G_UNICODE_OTHER_SYMBOL:
+				case G_UNICODE_LINE_SEPARATOR:
+				case G_UNICODE_PARAGRAPH_SEPARATOR:
+				case G_UNICODE_SPACE_SEPARATOR:
 				default:
 					break;
 				}
@@ -946,7 +974,7 @@ static void enchant_trie_find_matches(EnchantTrie* trie,EnchantTrieMatcher *matc
 	if (trie == EOSTrie) {
 		size_t word_len = strlen(matcher->word);
 		errs = matcher->num_errors;
-		if(word_len > matcher->word_pos) {
+		if((ssize_t)word_len > matcher->word_pos) {
 			matcher->num_errors = errs + word_len - matcher->word_pos;
 		}
 		if (matcher->num_errors <= matcher->max_errors) {
diff --git a/tests/enchant-ispell.c b/tests/enchant-ispell.c
index ff12c09..f50daee 100644
--- a/tests/enchant-ispell.c
+++ b/tests/enchant-ispell.c
@@ -208,6 +208,25 @@ is_word_char (gunichar uc, size_t n)
 	case G_UNICODE_OTHER_NUMBER:
 	case G_UNICODE_CONNECT_PUNCTUATION:
                 return 1;     /* Enchant 1.3.0 defines word chars like this. */
+
+	case G_UNICODE_CONTROL:
+	case G_UNICODE_FORMAT:
+	case G_UNICODE_UNASSIGNED:
+	case G_UNICODE_PRIVATE_USE:
+	case G_UNICODE_SURROGATE:
+	case G_UNICODE_DASH_PUNCTUATION:
+	case G_UNICODE_CLOSE_PUNCTUATION:
+	case G_UNICODE_FINAL_PUNCTUATION:
+	case G_UNICODE_INITIAL_PUNCTUATION:
+	case G_UNICODE_OTHER_PUNCTUATION:
+	case G_UNICODE_OPEN_PUNCTUATION:
+	case G_UNICODE_CURRENCY_SYMBOL:
+	case G_UNICODE_MODIFIER_SYMBOL:
+	case G_UNICODE_MATH_SYMBOL:
+	case G_UNICODE_OTHER_SYMBOL:
+	case G_UNICODE_LINE_SEPARATOR:
+	case G_UNICODE_PARAGRAPH_SEPARATOR:
+	case G_UNICODE_SPACE_SEPARATOR:
 	default:
 		if ((n > 0) && (uc == g_utf8_get_char("'"))) {
 		        return 1;  /** Char ' is accepted only within a word. */