Improve and simplify setting input methods and dictionaries using new…

… helper functions

Resolves: #318

New behaviour:

Let’s start the setup tool in Hindi locale:

    $ LC_ALL=hi_IN.UTF-8 python3 setup/main.py &

Setting 'dictionary' or 'inputmethod' to an empty string actually sets
the default for the current locale:

    $ dconf write /org/freedesktop/ibus/engine/typing-booster/dictionary '""'
    $ dconf read /org/freedesktop/ibus/engine/typing-booster/dictionary
    'hi_IN,en_GB'

    $ dconf write /org/freedesktop/ibus/engine/typing-booster/inputmethod '""'
    $ dconf read /org/freedesktop/ibus/engine/typing-booster/inputmethod
    'hi-inscript2,NoIME'

The typing booster engine does the same as the setup tool, it converts empty
strings to the locale defaults, but only at startup. So when the typing booster
engine runs in a different locale than the setup tool, setting empty strings
does not cause a conflict. The code to set the locale defaults is only
executed in the setup tool then because the typing booster engine is already
running. And when it next restarts, these keys are not empty strings anymore
because the setup tool already changed them.

The new code also does some normalization of the 'dictionary' and 'inputmethod'
values (removes encodings, changes to CLDR locale spellings, removes duplicates,
normalizes the spellings of the special dictionary 'None' and the special
input method 'NoIME'):

    $ dconf write /org/freedesktop/ibus/engine/typing-booster/dictionary '"nOnE, en_US.UTF8,en_US,en.UTF-8,en,C, nan_TW.utf8@latin,sr_RS.utf8@latin"'
    $ dconf read /org/freedesktop/ibus/engine/typing-booster/dictionary
    'None,en_US,en,en_US_POSIX,nan_Latn_TW,sr_Latn_RS'

    $ dconf write /org/freedesktop/ibus/engine/typing-booster/inputmethod '"nOiME,, t-latn-post,foobar, "'
    $ dconf read /org/freedesktop/ibus/engine/typing-booster/inputmethod
    'NoIME,t-latn-post,foobar'

engine/hunspell_table.py

@@ -384,34 +384,13 @@ def __init__(
         self._dictionary_names: List[str] = []
         dictionary = itb_util.variant_to_value(
             self._gsettings.get_value('dictionary'))
-        if dictionary:
-            # There is a dictionary setting in Gsettings, use that:
-            names = [x.strip() for x in dictionary.split(',')]
-            for name in names:
-                if name:
-                    self._dictionary_names.append(name)
-        else:
-            # There is no dictionary setting in Gsettings. Get the default
-            # dictionaries for the current effective value of
-            # LC_CTYPE and save it to Gsettings:
-            self._dictionary_names = itb_util.get_default_dictionaries(
-                itb_util.get_effective_lc_ctype())
+        self._dictionary_names = itb_util.dictionaries_str_to_list(dictionary)
+        if ','.join(self._dictionary_names) != dictionary:
+            # Value changed due to normalization or getting the locale
+            # defaults, save it back to settings:
             self._gsettings.set_value(
                 'dictionary',
                 GLib.Variant.new_string(','.join(self._dictionary_names)))
-        if (len(self._dictionary_names)
-            > itb_util.MAXIMUM_NUMBER_OF_DICTIONARIES):
-            LOGGER.warning(
-                'Trying to set more than the allowed maximum of '
-                '%s dictionaries.\n'
-                'Trying to set: %s\n'
-                'Really setting: %s\n',
-                itb_util.MAXIMUM_NUMBER_OF_DICTIONARIES,
-                self._dictionary_names,
-                self._dictionary_names[
-                    :itb_util.MAXIMUM_NUMBER_OF_DICTIONARIES])
-            self._dictionary_names = self._dictionary_names[
-                :itb_util.MAXIMUM_NUMBER_OF_DICTIONARIES]
         self.database.hunspell_obj.set_dictionary_names(
             self._dictionary_names[:])
 
@@ -429,32 +408,13 @@ def __init__(
         # Try to get the selected input methods from Gsettings:
         inputmethod = itb_util.variant_to_value(
             self._gsettings.get_value('inputmethod'))
-        if inputmethod:
-            inputmethods = [re.sub(re.escape('noime'), 'NoIME', x.strip(),
-                                   flags=re.IGNORECASE)
-                            for x in inputmethod.split(',') if x]
-            for ime in inputmethods:
-                self._current_imes.append(ime)
-        if self._current_imes == []:
-            # There is no ime set in Gsettings, get a default list
-            # of input methods for the current effective value of LC_CTYPE
-            # and save it to Gsettings:
-            self._current_imes = itb_util.get_default_input_methods(
-                itb_util.get_effective_lc_ctype())
+        self._current_imes = itb_util.input_methods_str_to_list(inputmethod)
+        if ','.join(self._current_imes) != inputmethod:
+            # Value changed due to normalization or getting the locale
+            # defaults, save it back to settings:
             self._gsettings.set_value(
                 'inputmethod',
                 GLib.Variant.new_string(','.join(self._current_imes)))
-        if len(self._current_imes) > itb_util.MAXIMUM_NUMBER_OF_INPUT_METHODS:
-            LOGGER.warning(
-                'Trying to set more than the allowed maximum of '
-                '%s input methods.\n'
-                'Trying to set: %s\n'
-                'Really setting: %s\n',
-                itb_util.MAXIMUM_NUMBER_OF_INPUT_METHODS,
-                self._current_imes,
-                self._current_imes[:itb_util.MAXIMUM_NUMBER_OF_INPUT_METHODS])
-            self._current_imes = (
-                self._current_imes[:itb_util.MAXIMUM_NUMBER_OF_INPUT_METHODS])
 
         self._commit_happened_after_focus_in = False
 
@@ -1376,6 +1336,8 @@ def set_current_imes(
         :param imes: List of input methods
                      If a single string is used, it should contain
                      the names of the input methods separated by commas.
+                     If the string is empty, the default input
+                     methods for the current locale are set.
         :param update_gsettings: Whether to write the change to Gsettings.
                                  Set this to False if this method is
                                  called because the Gsettings key changed
@@ -1384,22 +1346,9 @@ def set_current_imes(
         '''
         LOGGER.debug('imes=%s type(imes)=%s', imes, type(imes))
         if isinstance(imes, str):
-            imes = [x.strip() for x in imes.split(',')]
-        imes = [re.sub(re.escape('noime'), 'NoIME', x.strip(),
-                       flags=re.IGNORECASE)
-                for x in imes if x]
+            imes = itb_util.input_methods_str_to_list(imes)
         if imes == self._current_imes: # nothing to do
             return
-        if len(imes) > itb_util.MAXIMUM_NUMBER_OF_INPUT_METHODS:
-            LOGGER.error(
-                'Trying to set more than the allowed maximum of '
-                '%s input methods.\n'
-                'Trying to set: %s\n'
-                'Really setting: %s\n',
-                itb_util.MAXIMUM_NUMBER_OF_INPUT_METHODS,
-                imes,
-                imes[:itb_util.MAXIMUM_NUMBER_OF_INPUT_METHODS])
-            imes = imes[:itb_util.MAXIMUM_NUMBER_OF_INPUT_METHODS]
         if set(imes) != set(self._current_imes):
             # Input methods have been added or removed from the list
             # of current input methods. Initialize the
@@ -1431,6 +1380,8 @@ def set_dictionary_names(
                                  If a single string is used, it should contain
                                  the names of the dictionaries separated
                                  by commas.
+                                 If the string is empty, the default
+                                 dictionaries for the current locale are set.
         :param update_gsettings: Whether to write the change to Gsettings.
                                  Set this to False if this method is
                                  called because the Gsettings key changed
@@ -1440,21 +1391,10 @@ def set_dictionary_names(
         LOGGER.debug('dictionary_names=%s type(dictionary_names)=%s',
                      dictionary_names, type(dictionary_names))
         if isinstance(dictionary_names, str):
-            dictionary_names = [x.strip() for x in dictionary_names.split(',')]
-        dictionary_names = [x for x in dictionary_names if x]
+            dictionary_names = itb_util.dictionaries_str_to_list(
+                dictionary_names)
         if dictionary_names == self._dictionary_names: # nothing to do
             return
-        if len(dictionary_names) > itb_util.MAXIMUM_NUMBER_OF_DICTIONARIES:
-            LOGGER.error(
-                'Trying to set more than the allowed maximum of '
-                '%s dictionaries.\n'
-                'Trying to set: %s\n'
-                'Really setting: %s\n',
-                itb_util.MAXIMUM_NUMBER_OF_DICTIONARIES,
-                dictionary_names,
-                dictionary_names[:itb_util.MAXIMUM_NUMBER_OF_DICTIONARIES])
-            dictionary_names = (
-                dictionary_names[:itb_util.MAXIMUM_NUMBER_OF_DICTIONARIES])
         self._dictionary_names = dictionary_names
         self.database.hunspell_obj.set_dictionary_names(dictionary_names)
         self._update_dictionary_menu_dicts()

engine/itb_util.py

@@ -633,7 +633,7 @@ def input_methods_str_to_list(imes_str: str) -> List[str]:
 
     If the input contains duplicates, the duplicates are removed:
 
-    >>> input_methods_str_to_list('a ,a,b , c , a ,d,e,f')
+    >>> input_methods_str_to_list('a ,a,b , c ,, a ,d,e,f, ')
     ['a', 'b', 'c', 'd', 'e', 'f']
 
     Non-standard spellings of the special input method ['NoIME']
@@ -647,7 +647,7 @@ def input_methods_str_to_list(imes_str: str) -> List[str]:
     imes: List[str] = []
     for ime in  [re.sub(re.escape('noime'), 'NoIME', x.strip(),
                         flags=re.IGNORECASE)
-                 for x in imes_str.split(',') if x]:
+                 for x in imes_str.split(',') if x.strip()]:
         if ime not in imes:
             imes.append(ime)
     if len(imes) > MAXIMUM_NUMBER_OF_INPUT_METHODS:
@@ -721,7 +721,7 @@ def dictionaries_str_to_list(dictionaries_str: str) -> List[str]:
     dictionaries: List[str] = []
     for dictionary in [re.sub(re.escape('none'), 'None', x.strip(),
                               flags=re.IGNORECASE)
-                       for x in dictionaries_str.split(',') if x]:
+                       for x in dictionaries_str.split(',') if x.strip()]:
         if dictionary != 'None':
             dictionary = locale_normalize(dictionary)
         if dictionary and dictionary not in dictionaries:

setup/main.py

@@ -2120,35 +2120,10 @@ def _fill_dictionaries_listbox(self) -> None:
         self._dictionary_names = []
         dictionary = itb_util.variant_to_value(
             self._gsettings.get_value('dictionary'))
-        if dictionary:
-            names = [x.strip() for x in dictionary.split(',')]
-            for name in names:
-                if name:
-                    self._dictionary_names.append(name)
-        if self._dictionary_names == []:
-            # There are no dictionaries set in dconf, get a default list of
-            # dictionaries from the current effective value of LC_CTYPE:
-            self._dictionary_names = itb_util.get_default_dictionaries(
-                itb_util.get_effective_lc_ctype())
-            # And save the default dictionaries to settings:
-            self._gsettings.set_value(
-                'dictionary',
-                GLib.Variant.new_string(','.join(self._dictionary_names)))
-        if (len(self._dictionary_names)
-            > itb_util.MAXIMUM_NUMBER_OF_DICTIONARIES):
-            LOGGER.error(
-                'Trying to set more than the allowed maximum of '
-                '%s dictionaries.\n'
-                'Trying to set: %s\n'
-                'Really setting: %s\n',
-                itb_util.MAXIMUM_NUMBER_OF_DICTIONARIES,
-                self._dictionary_names,
-                self._dictionary_names[
-                    :itb_util.MAXIMUM_NUMBER_OF_DICTIONARIES])
-            self._dictionary_names = (
-                self._dictionary_names[
-                    :itb_util.MAXIMUM_NUMBER_OF_DICTIONARIES])
-            # Save reduced list of dictionaries back to settings:
+        self._dictionary_names = itb_util.dictionaries_str_to_list(dictionary)
+        if ','.join(self._dictionary_names) != dictionary:
+            # Value changed due to normalization or getting the locale
+            # defaults, save it back to settings:
             self._gsettings.set_value(
                 'dictionary',
                 GLib.Variant.new_string(','.join(self._dictionary_names)))
@@ -2228,34 +2203,10 @@ def _fill_input_methods_listbox(self) -> None:
         self._current_imes = []
         inputmethod = itb_util.variant_to_value(
             self._gsettings.get_value('inputmethod'))
-        if inputmethod:
-            inputmethods = [re.sub(re.escape('noime'), 'NoIME', x.strip(),
-                                   flags=re.IGNORECASE)
-                            for x in inputmethod.split(',') if x]
-            for ime in inputmethods:
-                self._current_imes.append(ime)
-        if self._current_imes == []:
-            # There is no ime set in dconf, get a default list of
-            # input methods for the current effective value of LC_CTYPE:
-            self._current_imes = itb_util.get_default_input_methods(
-                itb_util.get_effective_lc_ctype())
-            # And save the default input methods to settings:
-            self._gsettings.set_value(
-                'inputmethod',
-                GLib.Variant.new_string(','.join(self._current_imes)))
-        if len(self._current_imes) > itb_util.MAXIMUM_NUMBER_OF_INPUT_METHODS:
-            LOGGER.error(
-                'Trying to set more than the allowed maximum of '
-                '%s input methods.\n'
-                'Trying to set: %s\n'
-                'Really setting: %s\n',
-                itb_util.MAXIMUM_NUMBER_OF_INPUT_METHODS,
-                self._current_imes,
-                self._current_imes[
-                    :itb_util.MAXIMUM_NUMBER_OF_INPUT_METHODS])
-            self._current_imes = (
-                self._current_imes[:itb_util.MAXIMUM_NUMBER_OF_INPUT_METHODS])
-            # Save reduced list of input methods back to settings:
+        self._current_imes = itb_util.input_methods_str_to_list(inputmethod)
+        if ','.join(self._current_imes) != inputmethod:
+            # Value changed due to normalization or getting the locale
+            # defaults, save it back to settings:
             self._gsettings.set_value(
                 'inputmethod',
                 GLib.Variant.new_string(','.join(self._current_imes)))
@@ -2477,19 +2428,11 @@ def _on_restore_all_defaults_button_clicked(
             schema = gsettings.get_property('settings-schema')
             for key in schema.list_keys():
                 if key in ('googleapplicationcredentials',
-                           'dictionaryinstalltimestamp',
-                           'inputmethod',
-                           'dictionary'):
+                           'dictionaryinstalltimestamp'):
                     LOGGER.info('Skipping reset of gsettings key=%s', key)
                     continue
                 LOGGER.info('Resetting gsettings key=%s', key)
                 gsettings.reset(key)
-            LOGGER.info('Setting input methods to default for current locale:')
-            self.set_current_imes(itb_util.get_default_input_methods(
-                itb_util.get_effective_lc_ctype()), update_gsettings=True)
-            LOGGER.info('Setting dictionaries to default for current locale:')
-            self.set_dictionary_names(itb_util.get_default_dictionaries(
-                itb_util.get_effective_lc_ctype()), update_gsettings=True)
         else:
             LOGGER.info('Restore all defaults cancelled.')
         self._restore_all_defaults_button.set_sensitive(True)
@@ -3209,8 +3152,7 @@ def _on_dictionaries_default_button_clicked(self, *_args: Any) -> None:
         Sets the dictionaries to the default for the current locale.
 
         '''
-        self.set_dictionary_names(itb_util.get_default_dictionaries(
-            itb_util.get_effective_lc_ctype()))
+        self.set_dictionary_names(itb_util.dictionaries_str_to_list(''))
 
     def _on_dictionary_selected(
             self, _listbox: Gtk.ListBox, listbox_row: Gtk.ListBoxRow) -> None:
@@ -3527,8 +3469,7 @@ def _on_input_methods_default_button_clicked(self, *_args: Any) -> None:
 
         Sets the input methods to the default for the current locale.
         '''
-        self.set_current_imes(itb_util.get_default_input_methods(
-            itb_util.get_effective_lc_ctype()))
+        self.set_current_imes(itb_util.input_methods_str_to_list(''))
 
     def _on_input_method_selected(
             self, _listbox: Gtk.ListBox, listbox_row: Gtk.ListBoxRow) -> None:
@@ -5293,30 +5234,20 @@ def set_current_imes(
         :param imes: List of input methods
                      If a single string is used, it should contain
                      the names of the input methods separated by commas.
+                     If the string is empty, the default input
+                     methods for the current locale are set.
         :param update_gsettings: Whether to write the change to Gsettings.
                                  Set this to False if this method is
                                  called because the Gsettings key changed
                                  to avoid endless loops when the Gsettings
                                  key is changed twice in a short time.
         '''
-        LOGGER.debug('imes=%s type(imes)=%s', imes, type(imes))
+        LOGGER.debug('imes=“%s” type(imes)=%s update_gsettings=%s',
+                     imes, type(imes), update_gsettings)
         if isinstance(imes, str):
-            imes = [x.strip() for x in imes.split(',')]
-        imes = [re.sub(re.escape('noime'), 'NoIME', x.strip(),
-                       flags=re.IGNORECASE)
-                for x in imes if x]
+            imes = itb_util.input_methods_str_to_list(imes)
         if imes == self._current_imes: # nothing to do
             return
-        if len(imes) > itb_util.MAXIMUM_NUMBER_OF_INPUT_METHODS:
-            LOGGER.error(
-                'Trying to set more than the allowed maximum of '
-                '%s input methods.\n'
-                'Trying to set: %s\n'
-                'Really setting: %s\n',
-                itb_util.MAXIMUM_NUMBER_OF_INPUT_METHODS,
-                imes,
-                imes[:itb_util.MAXIMUM_NUMBER_OF_INPUT_METHODS])
-            imes = imes[:itb_util.MAXIMUM_NUMBER_OF_INPUT_METHODS]
         self._current_imes = imes
         self._fill_input_methods_listbox()
         if update_gsettings:
@@ -5343,6 +5274,8 @@ def set_dictionary_names(
                                  If a single string is used, it should contain
                                  the names of the dictionaries separated
                                  by commas.
+                                 If the string is empty, the default
+                                 dictionaries for the current locale are set.
         :param update_gsettings: Whether to write the change to Gsettings.
                                  Set this to False if this method is
                                  called because the Gsettings key changed
@@ -5352,21 +5285,10 @@ def set_dictionary_names(
         LOGGER.debug('dictionary_names=%s type(dictionary_names)=%s',
                      dictionary_names, type(dictionary_names))
         if isinstance(dictionary_names, str):
-            dictionary_names = [x.strip() for x in dictionary_names.split(',')]
-        dictionary_names = [x for x in dictionary_names if x]
+            dictionary_names = itb_util.dictionaries_str_to_list(
+                dictionary_names)
         if dictionary_names == self._dictionary_names: # nothing to do
             return
-        if len(dictionary_names) > itb_util.MAXIMUM_NUMBER_OF_DICTIONARIES:
-            LOGGER.error(
-                'Trying to set more than the allowed maximum of '
-                '%s dictionaries.\n'
-                'Trying to set: %s\n'
-                'Really setting: %s\n',
-                itb_util.MAXIMUM_NUMBER_OF_DICTIONARIES,
-                dictionary_names,
-                dictionary_names[:itb_util.MAXIMUM_NUMBER_OF_DICTIONARIES])
-            dictionary_names = dictionary_names[
-                :itb_util.MAXIMUM_NUMBER_OF_DICTIONARIES]
         self._dictionary_names = dictionary_names
         self._fill_dictionaries_listbox()
         if update_gsettings: