Sinhala phonetic layout (#965)

gen_sinhala_phonetic_layout.py

@@ -0,0 +1,524 @@
+#! /bin/env python3
+
+"""
+Script to generate a layout based on an existing.
+
+Tuned to create Sinhala phonetic layout based on qwerty (US), but may be adoped
+for other scripts. Look at dicts before the LayoutBuilder code.
+
+Usage:
+    python3 gen_sinhala_phonetic_layout [-h|--help] [-v|--verbose] [-o|--output]
+
+By default with no args will write to corresponding file in `srcs/layouts/`.
+
+Script uses central symbol (in direction "c") to identify a key, which may not
+be appropriate for base (reference) layouts other, than qwerty (US).
+
+Warning will be printed to stderr if new symbol overrides some symbol of the
+reference layout in directions other, than "c".
+
+Exception will be rised on other
+conflicts e. g. when trying to move a symbol into occupied position.
+
+  - Made with latn_qwerty_us.xml from commit `6b1551d`
+  - Made with Python 3.13
+  - Requires Python >= 3.11
+"""
+
+import argparse
+import logging
+
+from enum import StrEnum
+from pathlib import Path
+from xml.etree import ElementTree
+
+
+class Placement(StrEnum):
+    C = 'c'
+    NW = 'nw'
+    N = 'n'
+    NE = 'ne'
+    E = 'e'
+    SE = 'se'
+    S = 's'
+    SW = 'sw'
+    W = 'w'
+
+
+# Based on XKB Sinhala (phonetic)
+KEYS_MAP: dict[str, tuple[str, str, str, str]] = {
+    # Row 1 ###########################################
+    'q': ('ඍ', 'ඎ', '\u0DD8', '\u0DF2'),
+    'w': ('ඇ', 'ඈ', '\u0DD0', '\u0DD1'),
+    'e': ('එ', 'ඒ', '\u0DD9', '\u0DDA'),
+    'r': ('ර', '', '', ''),  # In XKB virama is on layer 2
+    't': ('ත', 'ථ', 'ට', 'ඨ'),
+    'y': ('ය', '', '', ''),  # In XKB virama is on layer 2
+    'u': ('උ', 'ඌ', '\u0DD4', '\u0DD6'),
+    'i': ('ඉ', 'ඊ', '\u0DD2', '\u0DD3'),
+    'o': ('ඔ', 'ඕ', '\u0DDC', '\u0DDD'),
+    'p': ('ප', 'ඵ', '', ''),
+    # Row 2 ###########################################
+    'a': ('අ', 'ආ', '\u0DCA', '\u0DCF'),
+    's': ('ස', 'ශ', 'ෂ', ''),
+    'd': ('ද', 'ධ', 'ඩ', 'ඪ'),
+    'f': ('ෆ', '\u0D93', '', '\u0DDB'),  # In XKB aiyanna placed otherwise
+    'g': ('ග', 'ඝ', 'ඟ', ''),
+    'h': ('හ', '\u0D83', '\u0DDE', 'ඖ'),
+    'j': ('ජ', 'ඣ', 'ඦ', ''),
+    'k': ('ක', 'ඛ', 'ඦ', 'ඐ'),
+    'l': ('ල', 'ළ', '\u0DDF', '\u0DF3'),
+    # Row 3 ###########################################
+    'z': ('ඤ', 'ඥ', '', ''),  # In XKB contains bar, broken bar
+    'x': ('ඳ', 'ඬ', '', ''),
+    'c': ('ච', 'ඡ', '', ''),
+    'v': ('ව', '', '', ''),
+    'b': ('බ', 'භ', '', ''),
+    'n': ('න', 'ණ', '\u0D82', 'ඞ'),
+    'm': ('ම', 'ඹ', '', ''),
+}
+
+# How to place four levels of Key.
+# Syntax: LEVEL: PLACEMENT | 'FROM_LEVEL+MODIFIER'
+# The last means symbol on level FROM_LEVEL with modifier key MODIFIER gives
+# key on level LEVEL
+#
+LEVELS_MAP = {
+    0: Placement.C,
+    1: Placement.NE,
+    2: '0+shift',
+    3: '1+shift',
+}
+
+# Additional modify keys combinations.
+# Syntax:
+#   MODKEY: { A: B }
+#
+MODMAP_EXTRA: dict[str, dict[str, str]] = {
+    'shift': {
+        # Astrological numbers
+        '1': '෧',
+        '2': '෨',
+        '3': '෩',
+        '4': '෪',
+        '5': '෫',
+        '6': '෬',
+        '7': '෭',
+        '8': '෮',
+        '9': '෯',
+        '0': '෦',
+        # Kunddaliya
+        '.': '෴',
+        # Extra broken bar intead z key in XKB
+        '\u007C': '\u00A6',
+        # Special whitespaces
+        'zwj': 'zwnj',
+    },
+    'fn': {
+        # Sinhala archaic digits
+        'ඍ': '𑇡',  # 1
+        'ඇ': '𑇢',  # 2
+        'එ': '𑇣',  # 3
+        'ර': '𑇤',  # 4
+        'ත': '𑇥',  # 5
+        'ය': '𑇦',  # 6
+        'උ': '𑇧',  # 7
+        'ඉ': '𑇨',  # 8
+        'ඔ': '𑇩',  # 9
+        'ප': '𑇪',  # 10
+        'අ': '𑇫',  # 20
+        'ස': '𑇬',  # 30
+        'ද': '𑇭',  # 40
+        'ෆ': '𑇮',  # 50
+        'ග': '𑇯',  # 60
+        'හ': '𑇰',  # 70
+        'ජ': '𑇱',  # 80
+        'ක': '𑇲',  # 90
+        'ල': '𑇳',  # 100
+        'ළ': '𑇴',  # 1000
+        # Sinhala candrabindu for Sanskrit
+        'ණ': '\u0D81',
+    },
+}
+
+# Table to move additional characters in reference layout.
+# Format is (CENTRAL_CHAR, PLACEMENT): (CENTRAL_CHAR, PLACEMENT). E. g. to move
+# char from key with central character "q", direction "se" to key with central
+# character "w", direction "sw", add line:
+#   ('q', Placement.SE): ('w', Placement.SW),
+#
+# To delete a char, use None as destination placement. E.g.:
+#   ('q', Placment.SE): ('q', None)
+#
+# Moving of main char in central placement is not supported.
+#
+TRANSITIONS_MAP: dict[tuple[str, Placement], tuple[str, Placement | None]] = {
+    ('q', Placement.SE): ('q', Placement.SW),  # loc esc
+    ('q', Placement.NE): ('q', Placement.SE),  # 1
+
+    ('w', Placement.NE): ('w', Placement.SE),  # 2
+
+    ('e', Placement.SE): ('r', Placement.NW),  # loc €
+    ('e', Placement.NE): ('e', Placement.SE),  # 3
+
+    ('r', Placement.NE): ('r', Placement.SE),  # 4
+    ('t', Placement.NE): ('t', Placement.SE),  # 5
+    ('y', Placement.NE): ('y', Placement.SE),  # 6
+    ('u', Placement.NE): ('u', Placement.SE),  # 7
+    ('i', Placement.NE): ('i', Placement.SE),  # 8
+
+    ('o', Placement.SE): ('p', Placement.SW),  # )
+    ('o', Placement.NE): ('o', Placement.SE),  # 9
+
+    ('p', Placement.NE): ('p', Placement.SE),  # 0
+
+    ('a', Placement.NE): ('a', Placement.NW),  # `
+    ('a', Placement.NW): ('a', Placement.SW),  # loc tab
+
+    ('s', Placement.NE): ('s', Placement.NW),  # loc §
+
+    ('g', Placement.SW): ('g', Placement.NW),  # _
+    ('g', Placement.NE): ('g', Placement.SW),  # -
+
+    ('h', Placement.SW): ('h', Placement.NW),  # +
+    ('h', Placement.NE): ('h', Placement.SW),  # =
+
+    ('l', Placement.NE): ('l', Placement.NW),  # |
+
+    ('x', Placement.NE): ('x', Placement.NW),  # loc †
+    ('c', Placement.NE): ('c', Placement.NW),  # <
+    ('b', Placement.NE): ('b', Placement.NW),  # ?
+    ('n', Placement.NE): ('n', Placement.NW),  # :
+    ('m', Placement.NE): ('m', Placement.NW),  # "
+}
+
+# Add additional characters to arbitrary places.
+# Syntax is CHAR: POSITION, where POSITION is a pari as in TRANSITIONS_MAP.
+#
+CHARS_EXTRA = {
+    # In XKB ZWJ is on `/` key, and ZWNJ is on spacebar
+    'zwj': ('m', Placement.SE),
+}
+
+
+# List of char unicode numbers and inclusive ranges of numbers to encode as XML
+# numeric character references.
+# Good for combining signs to not mess with quotes.
+# Characters in line of the keyboard tag will not be escaped.
+#
+ESCAPE_LIST: list[int | tuple[int, int]] = [
+    # Sinhalese diacritics
+    (0xD81, 0xD83),
+    (0xDCA, 0xDDF),
+]
+
+# Default filename. Output path can be overrided with `-o` flag also.
+LAYOUT_FILENAME = 'sinhala_phonetic.xml'
+
+# Will be placed after XML declaration. Need to have proper <!-- --> tags.
+COMMENT = '''
+<!-- This file defines Sinhala layout.
+
+Based on XKB Sinhala (phonetic) layout.
+-->
+'''
+
+BASE_DIR = Path(__file__).parent
+REFERENCE_LAYOUT_FILE = BASE_DIR / 'srcs/layouts/latn_qwerty_us.xml'
+
+LOGGER = logging.getLogger(__name__)
+KeysMapType = list[list[ElementTree.Element]]
+
+
+class LayoutGenError(RuntimeError):
+    ...
+
+
+def xml_elem_to_str(element: ElementTree.Element) -> str:
+    return ElementTree.tostring(
+        element,
+        xml_declaration=False,
+        encoding='unicode').strip()
+
+
+def keys_map_to_str(keys_map: KeysMapType) -> str:
+    """ Make laout rows map printable for debug purposes """
+    result = '[\n'
+    for row in keys_map:
+        result += ' ' * 4
+        for key in row:
+            result += str(key.attrib) + ', '
+        result += '\n'
+    result += ']'
+    return result
+
+
+def is_in_escape_list(char: str | int) -> bool:
+    if isinstance(char, str):
+        char = ord(char)
+    for item in ESCAPE_LIST:
+        if isinstance(item, tuple) and char >= item[0] and char <= item[1]:
+            return True
+        elif isinstance(item, int):
+            if char == item:
+                return True
+        else:
+            TypeError(f'Unexpected item {item} of ESCAPE_LIST')
+    return False
+
+
+def xml_encode_char(ch: str | int) -> str:
+    if isinstance(ch, str):
+        ch = ord(ch)
+    hex_val = hex(ch).split('x')[-1]
+    return f'&#x{hex_val.upper().zfill(4)};'
+
+
+class LayoutBuilder:
+    XML_DECLARATION = "<?xml version='1.0' encoding='utf-8'?>"
+
+    def __init__(
+        self,
+        name: str = '',
+        script: str = '',
+        numpad_script: str = '',
+        comment: str = '',
+    ) -> None:
+        """
+        :param comment: MUST be a valid XML comment wrapped in <!-- tags -->
+        """
+        attrs = {}
+        if name:
+            attrs['name'] = name
+        if script:
+            attrs['script'] = script
+        if numpad_script:
+            attrs['numpad_script'] = numpad_script
+        self._comment = None
+        if comment:
+            self._comment = comment.strip() or None
+        self._xml_keyboard = ElementTree.Element('keyboard', attrib=attrs)
+        self._modmap = ElementTree.Element('modmap')
+
+    @staticmethod
+    def _parse_reference_layout() -> list[ElementTree.Element]:
+        return ElementTree.parse(REFERENCE_LAYOUT_FILE).findall('row')
+
+    @staticmethod
+    def _move_untransited_to_new_map(
+        ref_map: KeysMapType,
+        new_map: KeysMapType
+    ) -> None:
+        coordinates = [
+            (row_num, key_num)
+            for row_num in range(len(ref_map))
+            for key_num in range(len(ref_map[row_num]))
+        ]
+
+        for row_num, key_num in coordinates:
+            old_key = ref_map[row_num][key_num]
+            new_key = new_map[row_num][key_num]
+            for k, val in old_key.attrib.items():
+                if (transited := new_key.attrib.get(k)) is not None:
+                    msg = (
+                        f'Transition of {transited} to'
+                        f' {new_key.get(Placement.C)}:{k} conflictls with'
+                        f' existing value "{val}"')
+                    raise LayoutGenError(msg)
+                new_key.set(k, val)
+
+    @staticmethod
+    def _add_extra_chars_to_ref_map(
+        coord_map: dict[str, tuple[int, int]],
+        new_map: KeysMapType
+    ) -> None:
+        for char, (to_key_name, to_plc) in CHARS_EXTRA.items():
+            if not (to_coord := coord_map.get(to_key_name)):
+                msg = f'Trying to add "{char}" to missing key "{to_key_name}"'
+                raise LayoutGenError(msg)
+            row_num, key_num = to_coord
+            key = new_map[row_num][key_num]
+            if (existing := key.get(to_plc)) is not None:
+                msg = f'Trying to add char to <{to_key_name}:{to_plc}>, but already contains "{existing}"'
+                raise LayoutGenError(msg)
+            key.set(to_plc, char)
+            LOGGER.info(
+                'Added "%s" to <%s:%s>',
+                char, to_key_name, to_plc)
+
+    @classmethod
+    def _apply_transitions(cls, ref_map: list) -> list:
+        coord_map: dict[str, tuple[int, int]] = {}
+
+        coordinates = [
+            (row_num, key_num)
+            for row_num in range(len(ref_map))
+            for key_num in range(len(ref_map[row_num]))
+        ]
+
+        for row_num, key_num in coordinates:
+            row = ref_map[row_num]
+            key = row[key_num]
+            key_name = key.get(Placement.C)
+            if key_name in coord_map:
+                msg = f'Duplicated value "{key_name}" in central position'
+                raise LayoutGenError(msg)
+            coord_map[key_name] = (row_num, key_num)
+
+        # Make new map with empty keys
+        result_map = [[ElementTree.Element('key') for key in row] for row in ref_map]
+
+        # Place by transitions map on new places
+        for (from_key_name, from_plc), (to_key_name, to_plc) in TRANSITIONS_MAP.items():
+            if Placement.C in (from_plc, to_plc):
+                raise NotImplementedError('Transition from or to placment "c"')
+            if not (from_coord := coord_map.get(from_key_name)):
+                raise LayoutGenError(f'Transition from missing key {from_key_name}')
+            if not (to_coord := coord_map.get(to_key_name)):
+                raise LayoutGenError(f'Transition to missing key {to_key_name}')
+            from_key = ref_map[from_coord[0]][from_coord[1]]
+            to_key = result_map[to_coord[0]][to_coord[1]]
+            try:
+                val = from_key.attrib.pop(from_plc)
+            except KeyError:
+                msg = f'No value in key {from_key_name}, placement {from_plc} to move'
+                raise LayoutGenError(msg)
+            if to_plc is not None:
+                if to_key.get(to_plc):
+                    msg = f'Second transition to key {to_key_name}, placement {to_plc}'
+                    raise LayoutGenError(msg)
+                to_key.set(to_plc, val)
+                LOGGER.info(
+                    'Moved "%s" from <%s:%s> to <%s:%s>',
+                    val, from_key_name, from_plc, to_key_name, to_plc)
+            else:
+                LOGGER.info(
+                    'Deleted "%s" from <%s:%s>',
+                    val, from_key_name, from_plc)
+
+        # Fill new map with other values
+        cls._move_untransited_to_new_map(ref_map, new_map=result_map)
+
+        # Add additional characters
+        cls._add_extra_chars_to_ref_map(coord_map, new_map=result_map)
+
+        return result_map
+
+    @staticmethod
+    def _resolve_placement(
+        key: ElementTree.Element,
+        placement: Placement,
+        new_char: str
+    ) -> None:
+        if placement != Placement.C:
+            central_char = key.get(Placement.C)
+            existing = key.get(placement)
+            if existing:
+                LOGGER.warning(
+                    'Placement %s of key %s already occupied with %s',
+                    placement, central_char, existing)
+        key.set(placement, new_char)
+
+    def _process_key(self, key: ElementTree.Element) -> ElementTree.Element:
+        central_char = key.get(Placement.C)
+        if not central_char:
+            return key
+        new_key_entry = KEYS_MAP.get(central_char)
+        if new_key_entry is None:
+            return key
+
+        for level, placement_spec in LEVELS_MAP.items():
+            if not (new_char := new_key_entry[level]):
+                continue
+            if '+' in placement_spec:
+                pair = placement_spec.split('+')
+                from_level, modkey = int(pair[0]), pair[1]
+                key_a = new_key_entry[from_level]
+                key_b = new_char
+                if key_a is None:
+                    raise LayoutGenError(f'Tried to modife {key_a} to {key_b}')
+                ElementTree.SubElement(self._modmap, modkey, a=key_a, b=key_b)
+            else:
+                placement = Placement(placement_spec)
+                self._resolve_placement(key, placement=placement, new_char=new_char)
+        return key
+
+    @staticmethod
+    def _make_extra_modmap(modmap: ElementTree.Element) -> ElementTree.Element:
+        for modkey, ab_map in MODMAP_EXTRA.items():
+            for a, b in ab_map.items():
+                LOGGER.info('Adding modmap %s "%s" -> "%s"', modkey, a, b)
+                ElementTree.SubElement(modmap, modkey, a=a, b=b)
+        return modmap
+
+    @staticmethod
+    def _post_escape(data: str) -> str:
+        buf = ''
+        lines = data.splitlines(keepends=True)
+        for line in lines:
+            # Skip keyboard tag line to keep attributes
+            if '<keyboard ' in line:
+                buf += line
+                continue
+            for ch in line:
+                if is_in_escape_list(ch):
+                    ch = xml_encode_char(ch)
+                buf += ch
+        return buf
+
+    def build(self) -> None:
+        raw_ref_rows = self._parse_reference_layout()
+        ref_rows = self._apply_transitions(raw_ref_rows)
+        for row in ref_rows:
+            new_row = ElementTree.SubElement(self._xml_keyboard, 'row')
+            for key in row:
+                LOGGER.debug(
+                    'Processing reference entry %s',
+                    xml_elem_to_str(key))
+                new_row.append(self._process_key(key))
+        self._modmap = self._make_extra_modmap(self._modmap)
+        self._xml_keyboard.append(self._modmap)
+
+    def get_xml(self) -> str:
+        ElementTree.indent(self._xml_keyboard)
+        body_raw = xml_elem_to_str(self._xml_keyboard)
+        body = self._post_escape(body_raw)
+
+        result = self.XML_DECLARATION + '\n'
+        if self._comment:
+            result += self._comment + '\n'
+        result += body + '\n'
+
+        return result
+
+
+def get_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(
+        prog='gen_sinhala_phonetic_layout',
+        description='Generate XKB-based Sinhala layout',)
+    parser.add_argument(
+        '-o',
+        '--output',
+        default=BASE_DIR / f'srcs/layouts/{LAYOUT_FILENAME}',
+        help='File to write result, `-` for stdout')
+    parser.add_argument(
+        '-v',
+        '--verbose',
+        help='More verbose logging',
+        action='store_true')
+    return parser.parse_args()
+
+
+if __name__ == '__main__':
+    args = get_args()
+    logging.basicConfig(
+        level=logging.DEBUG if args.verbose else logging.WARNING,
+        format='%(levelname)s: %(message)s')
+    builder = LayoutBuilder(name='සිංහල', script='sinhala', comment=COMMENT)
+    builder.build()
+    content = builder.get_xml()
+    if args.output == '-':
+        print(content)
+    else:
+        with open(args.output, 'w') as file:
+            file.write(content)

res/values/layouts.xml

@@ -80,6 +80,7 @@
     <item>latn_qwertz_sq</item>
     <item>latn_workman_us</item>
     <item>shaw_imperial_en</item>
+    <item>sinhala_phonetic</item>
     <item>tamil_default</item>
     <item>urdu_phonetic_ur</item>
     <item>custom</item>
@@ -163,6 +164,7 @@
     <item>QWERTZ (Albanian)</item>
     <item>WORKMAN (US)</item>
     <item>Shaw Imperial</item>
+    <item>සිංහල</item>
     <item>தமிழ்</item>
     <item>Urdu Phonetic</item>
     <item>@string/pref_layout_e_custom</item>
@@ -246,6 +248,7 @@
     <item>@xml/latn_qwertz_sq</item>
     <item>@xml/latn_workman_us</item>
     <item>@xml/shaw_imperial_en</item>
+    <item>@xml/sinhala_phonetic</item>
     <item>@xml/tamil_default</item>
     <item>@xml/urdu_phonetic_ur</item>
     <item>-1</item>

res/xml/method.xml

@@ -51,6 +51,7 @@
   <subtype android:label="%s" android:languageTag="pt" android:imeSubtypeLocale="pt_BR" android:imeSubtypeMode="keyboard" android:isAsciiCapable="true" android:imeSubtypeExtraValue="script=latin,default_layout=latn_qwerty_pt,extra_keys=accent_aigu:á:é:í:ó:ú@d|accent_cedille:ç@c|accent_circonflexe:â:ê:ô@f|accent_grave:à:ò@f|accent_tilde:ã:õ@n|€|ª|º"/>
   <subtype android:label="%s" android:languageTag="ro" android:imeSubtypeLocale="ro_RO" android:imeSubtypeMode="keyboard" android:isAsciiCapable="true" android:imeSubtypeExtraValue="script=latin,default_layout=latn_qwerty_ro,extra_keys=ă|â|î|ș|ț|€|$"/>
   <subtype android:label="%s" android:languageTag="ru" android:imeSubtypeLocale="ru_RU" android:imeSubtypeMode="keyboard" android:isAsciiCapable="true" android:imeSubtypeExtraValue="script=latin,default_layout=cyrl_jcuken_ru"/>
+  <subtype android:label="%s" android:languageTag="si" android:imeSubtypeLocale="si_LK" android:imeSubtypeMode="keyboard" android:isAsciiCapable="true" android:imeSubtypeExtraValue="script=sinhala,default_layout=sinhala_phonetic,extra_keys=₨"/>
   <subtype android:label="%s" android:languageTag="sk" android:imeSubtypeLocale="sk_SK" android:imeSubtypeMode="keyboard" android:isAsciiCapable="true" android:imeSubtypeExtraValue="script=latin,default_layout=latn_qwertz_sk,extra_keys=accent_caron:ě:ř:ž:š:č:ň:ď:ľ:ť@f|accent_ring:ů@s|accent_circonflexe:ô@f|accent_trema:ä:ü:ö@u|accent_aigu:á:é:í:ó:ú:ŕ:ś:ĺ:ý@d"/>
   <subtype android:label="%s" android:languageTag="sq" android:imeSubtypeLocale="sq_AL" android:imeSubtypeMode="keyboard" android:isAsciiCapable="true" android:imeSubtypeExtraValue="script=latin,default_layout=latn_qwertz_sq"/>
   <subtype android:label="%s" android:languageTag="sr" android:imeSubtypeLocale="sr_" android:imeSubtypeMode="keyboard" android:isAsciiCapable="true" android:imeSubtypeExtraValue="script=latin,default_layout=cyrl_lynyertz_sr"/>

srcs/juloo.keyboard2/KeyValue.java

@@ -769,6 +769,38 @@ public final class KeyValue implements Comparable<KeyValue>
       case "௲": case "௳":
         return makeStringKey(name, FLAG_SMALLER_FONT);
 
+      /* Sinhala letters to reduced size */
+      case "අ": case "ආ": case "ඇ": case "ඈ": case "ඉ":
+      case "ඊ": case "උ": case "ඌ": case "ඍ": case "ඎ":
+      case "ඏ": case "ඐ": case "එ": case "ඒ": case "ඓ":
+      case "ඔ": case "ඕ": case "ඖ": case "ක": case "ඛ":
+      case "ග": case "ඝ": case "ඞ": case "ඟ": case "ච":
+      case "ඡ": case "ජ": case "ඣ": case "ඤ": case "ඥ":
+      case "ඦ": case "ට": case "ඨ": case "ඩ": case "ඪ":
+      case "ණ": case "ඬ": case "ත": case "ථ": case "ද":
+      case "ධ": case "න": case "ඳ": case "ප": case "ඵ":
+      case "බ": case "භ": case "ම": case "ඹ": case "ය":
+      case "ර": case "ල": case "ව": case "ශ": case "ෂ":
+      case "ස": case "හ": case "ළ": case "ෆ":
+      /* Astrological numbers */
+      case "෦": case "෧": case "෨": case "෩": case "෪":
+      case "෫": case "෬": case "෭": case "෮": case "෯":
+      case "ෲ": case "ෳ":
+      /* Diacritics */
+      case "\u0d81": case "\u0d82": case "\u0d83": case "\u0dca":
+      case "\u0dcf": case "\u0dd0": case "\u0dd1": case "\u0dd2":
+      case "\u0dd3": case "\u0dd4": case "\u0dd6": case "\u0dd8":
+      case "\u0dd9": case "\u0dda": case "\u0ddb": case "\u0ddc":
+      case "\u0ddd": case "\u0dde": case "\u0ddf":
+      /* Archaic digits */
+      case "𑇡": case "𑇢": case "𑇣": case "𑇤": case "𑇥":
+      case "𑇦": case "𑇧": case "𑇨": case "𑇩": case "𑇪":
+      case "𑇫": case "𑇬": case "𑇭": case "𑇮": case "𑇯":
+      case "𑇰": case "𑇱": case "𑇲": case "𑇳": case "𑇴":
+      /* Exta */
+      case "෴": case "₨":  // Rupee is not exclusively Sinhala sign
+        return makeStringKey(name, FLAG_SMALLER_FONT);
+
       /* Internal keys */
       case "selection_mode": return makeInternalModifier(Modifier.SELECTION_MODE);
 

srcs/layouts/sinhala_phonetic.xml

@@ -0,0 +1,107 @@
+<?xml version='1.0' encoding='utf-8'?>
+<!-- This file defines Sinhala layout.
+
+Based on XKB Sinhala (phonetic) layout.
+-->
+<keyboard name="සිංහල" script="sinhala">
+  <row>
+    <key sw="loc esc" se="1" c="ඍ" ne="ඎ" />
+    <key se="2" c="ඇ" nw="~" sw="\@" ne="ඈ" />
+    <key se="3" c="එ" nw="!" sw="\#" ne="ඒ" />
+    <key nw="loc €" se="4" c="ර" sw="$" />
+    <key se="5" c="ත" sw="%" ne="ථ" />
+    <key se="6" c="ය" sw="^" />
+    <key se="7" c="උ" sw="&amp;" ne="ඌ" />
+    <key se="8" c="ඉ" sw="*" ne="ඊ" />
+    <key se="9" c="ඔ" sw="(" ne="ඕ" />
+    <key sw=")" se="0" c="ප" ne="ඵ" />
+  </row>
+  <row>
+    <key nw="`" sw="loc tab" shift="0.5" c="අ" ne="ආ" />
+    <key nw="loc §" c="ස" sw="loc ß" ne="ශ" />
+    <key c="ද" ne="ධ" />
+    <key c="ෆ" ne="ඓ" />
+    <key nw="_" sw="-" c="ග" ne="ඝ" />
+    <key nw="+" sw="=" c="හ" ne="&#x0D83;" />
+    <key c="ජ" se="}" sw="{" ne="ඣ" />
+    <key c="ක" sw="[" se="]" ne="ඛ" />
+    <key nw="|" c="ල" sw="\\" ne="ළ" />
+  </row>
+  <row>
+    <key width="1.5" c="shift" ne="loc capslock" />
+    <key c="ඤ" ne="ඥ" />
+    <key nw="loc †" c="ඳ" ne="ඬ" />
+    <key nw="&lt;" c="ච" sw="." ne="ඡ" />
+    <key c="ව" ne="&gt;" sw="," />
+    <key nw="\?" c="බ" sw="/" ne="භ" />
+    <key nw=":" c="න" sw=";" ne="ණ" />
+    <key nw="&quot;" c="ම" sw="'" se="zwj" ne="ඹ" />
+    <key width="1.5" c="backspace" ne="delete" />
+  </row>
+  <modmap>
+    <shift a="ඍ" b="&#x0DD8;" />
+    <shift a="ඎ" b="ෲ" />
+    <shift a="ඇ" b="&#x0DD0;" />
+    <shift a="ඈ" b="&#x0DD1;" />
+    <shift a="එ" b="&#x0DD9;" />
+    <shift a="ඒ" b="&#x0DDA;" />
+    <shift a="ත" b="ට" />
+    <shift a="ථ" b="ඨ" />
+    <shift a="උ" b="&#x0DD4;" />
+    <shift a="ඌ" b="&#x0DD6;" />
+    <shift a="ඉ" b="&#x0DD2;" />
+    <shift a="ඊ" b="&#x0DD3;" />
+    <shift a="ඔ" b="&#x0DDC;" />
+    <shift a="ඕ" b="&#x0DDD;" />
+    <shift a="අ" b="&#x0DCA;" />
+    <shift a="ආ" b="&#x0DCF;" />
+    <shift a="ස" b="ෂ" />
+    <shift a="ද" b="ඩ" />
+    <shift a="ධ" b="ඪ" />
+    <shift a="ඓ" b="&#x0DDB;" />
+    <shift a="ග" b="ඟ" />
+    <shift a="හ" b="&#x0DDE;" />
+    <shift a="&#x0D83;" b="ඖ" />
+    <shift a="ජ" b="ඦ" />
+    <shift a="ක" b="ඦ" />
+    <shift a="ඛ" b="ඐ" />
+    <shift a="ල" b="&#x0DDF;" />
+    <shift a="ළ" b="ෳ" />
+    <shift a="න" b="&#x0D82;" />
+    <shift a="ණ" b="ඞ" />
+    <shift a="1" b="෧" />
+    <shift a="2" b="෨" />
+    <shift a="3" b="෩" />
+    <shift a="4" b="෪" />
+    <shift a="5" b="෫" />
+    <shift a="6" b="෬" />
+    <shift a="7" b="෭" />
+    <shift a="8" b="෮" />
+    <shift a="9" b="෯" />
+    <shift a="0" b="෦" />
+    <shift a="." b="෴" />
+    <shift a="|" b="¦" />
+    <shift a="zwj" b="zwnj" />
+    <fn a="ඍ" b="𑇡" />
+    <fn a="ඇ" b="𑇢" />
+    <fn a="එ" b="𑇣" />
+    <fn a="ර" b="𑇤" />
+    <fn a="ත" b="𑇥" />
+    <fn a="ය" b="𑇦" />
+    <fn a="උ" b="𑇧" />
+    <fn a="ඉ" b="𑇨" />
+    <fn a="ඔ" b="𑇩" />
+    <fn a="ප" b="𑇪" />
+    <fn a="අ" b="𑇫" />
+    <fn a="ස" b="𑇬" />
+    <fn a="ද" b="𑇭" />
+    <fn a="ෆ" b="𑇮" />
+    <fn a="ග" b="𑇯" />
+    <fn a="හ" b="𑇰" />
+    <fn a="ජ" b="𑇱" />
+    <fn a="ක" b="𑇲" />
+    <fn a="ල" b="𑇳" />
+    <fn a="ළ" b="𑇴" />
+    <fn a="ණ" b="&#x0D81;" />
+  </modmap>
+</keyboard>