Fix Crowdin config & workflow

Also add new languages and update translation scripts

.github/workflows/crowdin_prep.yml

@@ -15,15 +15,18 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Setup Java JDK
-        uses: actions/setup-java@v1
+        uses: actions/setup-java@v3
         with:
-          java-version: 1.8
+          java-version: 18
+          distribution: zulu
 
       - name: Setup Python
-        uses: actions/setup-python@v2
+        uses: actions/setup-python@v4
+        with:
+          python-version: '3.10'
 
       - name: Checkout
-        uses: actions/checkout@v2
+        uses: actions/checkout@v3
 
       - name: Upload Source
         shell: bash

.github/workflows/crowdin_translate.yml

@@ -5,22 +5,25 @@ name: Crowdin Translation Integration
 on:
   schedule:
   # please choose a random time & weekday to avoid all repos synching at the same time
-    - cron: '15 21 * * 5'  # Fridays at 9:15 PM, UTC
+    - cron: '10 11 * * 5'  # Fridays at 11:10 AM, UTC
 
 jobs:
   create_intl_file:
     runs-on: ubuntu-latest
     steps:
       - name: Setup Java JDK
-        uses: actions/setup-java@v1
+        uses: actions/setup-java@v3
         with:
-          java-version: 1.8
+          java-version: 18
+          distribution: zulu
 
       - name: Setup Python
-        uses: actions/setup-python@v2
+        uses: actions/setup-python@v4
+        with:
+          python-version: '3.10'
 
       - name: Checkout
-        uses: actions/checkout@v2
+        uses: actions/checkout@v3
         with:
           persist-credentials: false # otherwise, the token used is the GITHUB_TOKEN, instead of your personal access token.
           fetch-depth: 0 # otherwise, there would be errors pushing refs to the destination repository.

intl/activate.py

@@ -62,7 +62,7 @@ if __name__ == '__main__':
    translate_txt = translate_txt.replace('<0-59>', f"{minutes}")
    translate_txt = translate_txt.replace('<0-23>', f"{hour}")
    translate_txt = translate_txt.replace('# Fridays at , UTC',
-                                         f"# Fridays at {hour%12}:{minutes} {'AM' if hour < 12 else 'PM'}, UTC")
+                                         f"# Fridays at {hour%12}:{minutes if minutes > 9 else '0' + str(minutes)} {'AM' if hour < 12 else 'PM'}, UTC")
    translate_txt = translate_txt.replace("<CORE_NAME>", CORE_NAME)
    translate_txt = translate_txt.replace('<PATH/TO>/libretro_core_options_intl.h',
                                          core_intl_file)

intl/core_option_regex.py

@@ -1,14 +1,14 @@
 import re
 
-# 0: full struct; 1: up to & including first []; 2: content between first {}
+# 0: full struct; 1: up to & including first []; 2 & 3: comments; 4: content between first {}
-p_struct = re.compile(r'(struct\s*[a-zA-Z0-9_\s]+\[])\s*'
+p_struct = re.compile(r'(\bstruct\b\s*[a-zA-Z0-9_\s]+\[])\s*'  # 1st capturing group
-                      r'(?:(?:\/\*(?:.|[\r\n])*?\*\/|\/\/.*[\r\n]+)\s*)*'
+                      r'(?:(?=(\/\*(?:.|[\r\n])*?\*\/|\/\/.*[\r\n]+))\2\s*)*'  # 2nd capturing group
                       r'=\s*'  # =
-                      r'(?:(?:\/\*(?:.|[\r\n])*?\*\/|\/\/.*[\r\n]+)\s*)*'
+                      r'(?:(?=(\/\*(?:.|[\r\n])*?\*\/|\/\/.*[\r\n]+))\3\s*)*'  # 3rd capturing group
                       r'{((?:.|[\r\n])*?)\{\s*NULL,\s*NULL,\s*NULL\s*(?:.|[\r\n])*?},?(?:.|[\r\n])*?};')  # captures full struct, it's beginning and it's content
 # 0: type name[]; 1: type; 2: name
-p_type_name = re.compile(r'(retro_core_option_[a-zA-Z0-9_]+)\s*'
+p_type_name = re.compile(r'(\bretro_core_option_[a-zA-Z0-9_]+)\s*'
-                         r'(option_cats([a-z_]{0,8})|option_defs([a-z_]*))\s*\[]')
+                         r'(\boption_cats([a-z_]{0,8})|\boption_defs([a-z_]*))\s*\[]')
 # 0: full option; 1: key; 2: description; 3: additional info; 4: key/value pairs
 p_option = re.compile(r'{\s*'  # opening braces
                       r'(?:(?:\/\*(?:.|[\r\n])*?\*\/|\/\/.*[\r\n]+|#.*[\r\n]+)\s*)*'
@@ -76,9 +76,9 @@ p_key_value = re.compile(r'{\s*'  # opening braces
 
 p_masked = re.compile(r'([A-Z_][A-Z0-9_]+)\s*(\"(?:"\s*"|\\\s*|.)*\")')
 
-p_intl = re.compile(r'(struct retro_core_option_definition \*option_defs_intl\[RETRO_LANGUAGE_LAST]) = {'
+p_intl = re.compile(r'(\bstruct retro_core_option_definition \*option_defs_intl\[RETRO_LANGUAGE_LAST]) = {'
                     r'((?:.|[\r\n])*?)};')
-p_set = re.compile(r'static INLINE void libretro_set_core_options\(retro_environment_t environ_cb\)'
+p_set = re.compile(r'\bstatic INLINE void libretro_set_core_options\(retro_environment_t environ_cb\)'
                    r'(?:.|[\r\n])*?};?\s*#ifdef __cplusplus\s*}\s*#endif')
 
 p_yaml = re.compile(r'"project_id": "[0-9]+".*\s*'

intl/core_option_translation.py

@@ -134,13 +134,12 @@ def is_viable_non_dupe(text: str, comparison) -> bool:
 
 
 def is_viable_value(text: str) -> bool:
-   """text must be longer than 2 ('""'), not 'NULL' and text.lower() not in
+   """text must be longer than 2 ('""') and not 'NULL'.
-   {'"enabled"', '"disabled"', '"true"', '"false"', '"on"', '"off"'}.
 
    :param text: String to be tested.
    :return: bool
    """
-   return 2 < len(text) and text != 'NULL' and text.lower() not in ON_OFFS
+   return 2 < len(text) and text != 'NULL'
 
 
 def create_non_dupe(base_name: str, opt_num: int, comparison) -> str:
@@ -183,17 +182,17 @@ def get_texts(text: str) -> dict:
       if lang not in just_string:
          hash_n_string[lang] = {}
          just_string[lang] = set()
-
+      is_v2_definition = 'retro_core_option_v2_definition' == struct_type_name[0]
-      is_v2 = False
       pre_name = ''
+      # info texts format
       p = cor.p_info
-      if 'retro_core_option_v2_definition' == struct_type_name[0]:
+      if 'retro_core_option_v2_category' == struct_type_name[0]:
-         is_v2 = True
+         # prepend category labels, as they can be the same as option labels
-      elif 'retro_core_option_v2_category' == struct_type_name[0]:
          pre_name = 'CATEGORY_'
+         # categories have different info texts format
          p = cor.p_info_cat
 
-      struct_content = struct.group(2)
+      struct_content = struct.group(4)
       # 0: full option; 1: key; 2: description; 3: additional info; 4: key/value pairs
       struct_options = cor.p_option.finditer(struct_content)
       for opt, option in enumerate(struct_options):
@@ -219,7 +218,7 @@ def get_texts(text: str) -> dict:
          if option.group(3):
             infos = option.group(3)
             option_info = p.finditer(infos)
-            if is_v2:
+            if is_v2_definition:
                desc1 = next(option_info).group(1)
                if is_viable_non_dupe(desc1, just_string[lang]):
                   just_string[lang].add(desc1)
@@ -248,16 +247,21 @@ def get_texts(text: str) -> dict:
          else:
             raise ValueError(f'Too few arguments in struct {struct_type_name[1]} option {option.group(1)}!')
 
-         # group 4:
+         # group 4: key/value pairs
          if option.group(4):
             for j, kv_set in enumerate(cor.p_key_value.finditer(option.group(4))):
                set_key, set_value = kv_set.group(1, 2)
                if not is_viable_value(set_value):
-                  if not is_viable_value(set_key):
+                  # use the key if value not available
-                     continue
                   set_value = set_key
+                  if not is_viable_value(set_value):
+                     continue
                # re.fullmatch(r'(?:[+-][0-9]+)+', value[1:-1])
-               if set_value not in just_string[lang] and not re.sub(r'[+-]', '', set_value[1:-1]).isdigit():
+
+               # add only if non-dupe, not translated by RetroArch directly & not purely numeric
+               if set_value not in just_string[lang]\
+                  and set_value.lower() not in ON_OFFS\
+                  and not re.sub(r'[+-]', '', set_value[1:-1]).isdigit():
                   clean_key = set_key[1:-1]
                   clean_key = remove_special_chars(clean_key).upper().replace(' ', '_')
                   m_h = create_non_dupe(re.sub(r'__+', '_', f"OPTION_VAL_{clean_key}"), opt, hash_n_string[lang])
@@ -298,8 +302,12 @@ def h2json(file_paths: dict) -> dict:
    for file_lang in file_paths:
       if not os.path.isfile(file_paths[file_lang]):
          continue
-
+      file_path = file_paths[file_lang]
-      jsons[file_lang] = file_paths[file_lang][:-2] + '.json'
+      try:
+         jsons[file_lang] = file_path[:file_path.rindex('.')] + '.json'
+      except ValueError:
+         print(f"File {file_path} has incorrect format! File ending missing?")
+         continue
 
       p = cor.p_masked
 
@@ -397,11 +405,11 @@ def get_crowdin_client(dir_path: str) -> str:
    return jar_path
 
 
-def create_intl_file(localisation_file_path: str, intl_dir_path: str, text: str, file_path: str) -> None:
+def create_intl_file(intl_file_path: str, localisations_path: str, text: str, file_path: str) -> None:
    """Creates 'libretro_core_options_intl.h' from Crowdin translations.
 
-   :param localisation_file_path: Path to 'libretro_core_options_intl.h'
+   :param intl_file_path: Path to 'libretro_core_options_intl.h'
-   :param intl_dir_path: Path to the intl/<core_name> directory.
+   :param localisations_path: Path to the intl/<core_name> directory.
    :param text: Content of the 'libretro_core_options.h' being translated.
    :param file_path: Path to the '_us.h' file, containing the original English texts.
    :return: None
@@ -497,10 +505,11 @@ def create_intl_file(localisation_file_path: str, intl_dir_path: str, text: str,
              'extern "C" {\n' \
              '#endif\n'
 
-   if os.path.isfile(localisation_file_path):
+   if os.path.isfile(intl_file_path):
       # copy top of the file for re-use
-      with open(localisation_file_path, 'r', encoding='utf-8') as intl:  # libretro_core_options_intl.h
+      with open(intl_file_path, 'r', encoding='utf-8') as intl:  # libretro_core_options_intl.h
          in_text = intl.read()
+         # attempt 1: find the distinct comment header
          intl_start = re.search(re.escape('/*\n'
                                           ' ********************************\n'
                                           ' * Core Option Definitions\n'
@@ -509,19 +518,22 @@ def create_intl_file(localisation_file_path: str, intl_dir_path: str, text: str,
          if intl_start:
             out_txt = in_text[:intl_start.end(0)]
          else:
+            # attempt 2: if no comment header present, find c++ compiler instruction (it is kind of a must)
             intl_start = re.search(re.escape('#ifdef __cplusplus\n'
                                              'extern "C" {\n'
                                              '#endif\n'), in_text)
             if intl_start:
                out_txt = in_text[:intl_start.end(0)]
+            # if all attempts fail, use default from above
 
    # only write to file, if there is anything worthwhile to write!
    overwrite = False
 
    # iterate through localisation files
    files = {}
-   for file in os.scandir(intl_dir_path):
+   for file in os.scandir(localisations_path):
       files[file.name] = {'is_file': file.is_file(), 'path': file.path}
+
    for file in sorted(files):  # intl/<core_name>/_*
       if files[file]['is_file'] \
          and file.startswith('_') \
@@ -532,6 +544,7 @@ def create_intl_file(localisation_file_path: str, intl_dir_path: str, text: str,
          struct_groups = cor.p_struct.finditer(text)
          lang_low = os.path.splitext(file)[0].lower()
          lang_up = lang_low.upper()
+         # mark each language's section with a comment, for readability
          out_txt = out_txt + f'/* RETRO_LANGUAGE{lang_up} */\n\n'  # /* RETRO_LANGUAGE_NM */
 
          # copy adjusted translations (makros)
@@ -544,22 +557,22 @@ def create_intl_file(localisation_file_path: str, intl_dir_path: str, text: str,
             if 3 > len(struct_type_name):  # no language specifier
                new_decl = re.sub(re.escape(struct_type_name[1]), struct_type_name[1] + lang_low, declaration)
             else:
-               new_decl = re.sub(re.escape(struct_type_name[2]), lang_low, declaration)
                if '_us' != struct_type_name[2]:
+                  # only use _us constructs - other languages present in the source file are not important
                   continue
+               new_decl = re.sub(re.escape(struct_type_name[2]), lang_low, declaration)
 
-            p = cor.p_info
+            p = (cor.p_info_cat if 'retro_core_option_v2_category' == struct_type_name[0] else cor.p_info)
-            if 'retro_core_option_v2_category' == struct_type_name[0]:
-               p = cor.p_info_cat
             offset_construct = construct.start(0)
+            # append localised construct name and ' = {'
             start = construct.end(1) - offset_construct
-            end = construct.start(2) - offset_construct
+            end = construct.start(4) - offset_construct
             out_txt = out_txt + new_decl + construct.group(0)[start:end]
-
+            # insert macros
-            content = construct.group(2)
+            content = construct.group(4)
             new_content = cor.p_option.sub(replace_option, content)
-
+            start = construct.end(4) - offset_construct
-            start = construct.end(2) - offset_construct
+            # append macro-filled content and close the construct
             out_txt = out_txt + new_content + construct.group(0)[start:] + '\n'
 
             # for v2
@@ -574,7 +587,7 @@ def create_intl_file(localisation_file_path: str, intl_dir_path: str, text: str,
 
    # only write to file, if there is anything worthwhile to write!
    if overwrite:
-      with open(localisation_file_path, 'w', encoding='utf-8') as intl:
+      with open(intl_file_path, 'w', encoding='utf-8') as intl:
          intl.write(out_txt + '\n#ifdef __cplusplus\n'
                               '}\n#endif\n'
                               '\n#endif')
@@ -585,7 +598,7 @@ def create_intl_file(localisation_file_path: str, intl_dir_path: str, text: str,
 
 if __name__ == '__main__':
    try:
-      if os.path.isfile(sys.argv[1]):
+      if os.path.isfile(sys.argv[1]) or sys.argv[1].endswith('.h'):
          _temp = os.path.dirname(sys.argv[1])
       else:
          _temp = sys.argv[1]

intl/crowdin.yaml

@@ -6,8 +6,8 @@
 "files":
   [
     {
-      "source": "/intl/_core_name_/_us.json",
+      "source": "/_core_name_/_us.json",
       "dest": "/_core_name_/_core_name_.json",
-      "translation": "/intl/_core_name_/_%two_letters_code%.json",
+      "translation": "/_core_name_/_%two_letters_code%.json",
     },
   ]

intl/crowdin_prep.py

@@ -4,7 +4,7 @@ import core_option_translation as t
 
 if __name__ == '__main__':
    try:
-      if t.os.path.isfile(t.sys.argv[1]):
+      if t.os.path.isfile(t.sys.argv[1]) or t.sys.argv[1].endswith('.h'):
          _temp = t.os.path.dirname(t.sys.argv[1])
       else:
          _temp = t.sys.argv[1]

intl/crowdin_translate.py

@@ -4,7 +4,7 @@ import core_option_translation as t
 
 if __name__ == '__main__':
    try:
-      if t.os.path.isfile(t.sys.argv[1]):
+      if t.os.path.isfile(t.sys.argv[1]) or t.sys.argv[1].endswith('.h'):
          _temp = t.os.path.dirname(t.sys.argv[1])
       else:
          _temp = t.sys.argv[1]

intl/v1_to_v2_converter.py

@@ -9,7 +9,6 @@ The original files will be preserved as *.v1
 """
 import core_option_regex as cor
 import os
-import sys
 import glob
 
 
@@ -410,11 +409,14 @@ def create_v2_code_file(struct_text, file_name):
                       '   &options_ar,      /* RETRO_LANGUAGE_ARABIC */\n' \
                       '   &options_el,      /* RETRO_LANGUAGE_GREEK */\n' \
                       '   &options_tr,      /* RETRO_LANGUAGE_TURKISH */\n' \
-                      '   &options_sv,      /* RETRO_LANGUAGE_SLOVAK */\n' \
+                      '   &options_sk,      /* RETRO_LANGUAGE_SLOVAK */\n' \
                       '   &options_fa,      /* RETRO_LANGUAGE_PERSIAN */\n' \
                       '   &options_he,      /* RETRO_LANGUAGE_HEBREW */\n' \
                       '   &options_ast,     /* RETRO_LANGUAGE_ASTURIAN */\n' \
                       '   &options_fi,      /* RETRO_LANGUAGE_FINNISH */\n' \
+                      '   &options_id,      /* RETRO_LANGUAGE_INDONESIAN */\n' \
+                      '   &options_sv,      /* RETRO_LANGUAGE_SWEDISH */\n' \
+                      '   &options_uk,      /* RETRO_LANGUAGE_UKRAINIAN */\n' \
                     + out_text[intl.end(2):]
          out_text = p_set.sub(new_set, new_intl)
       else:

libretro-common/include/libretro.h

@@ -286,6 +286,11 @@ enum retro_language
    RETRO_LANGUAGE_INDONESIAN          = 24,
    RETRO_LANGUAGE_SWEDISH             = 25,
    RETRO_LANGUAGE_UKRAINIAN           = 26,
+   RETRO_LANGUAGE_CZECH               = 27,
+   RETRO_LANGUAGE_CATALAN_VALENCIA    = 28,
+   RETRO_LANGUAGE_CATALAN             = 29,
+   RETRO_LANGUAGE_BRITISH_ENGLISH     = 30,
+   RETRO_LANGUAGE_HUNGARIAN           = 31,
    RETRO_LANGUAGE_LAST,
 
    /* Ensure sizeof(enum) == sizeof(int) */

libretro_core_options.h

@@ -744,6 +744,11 @@ struct retro_core_options_v2 *options_intl[RETRO_LANGUAGE_LAST] = {
    &options_id,      /* RETRO_LANGUAGE_INDONESIAN */
    &options_sv,      /* RETRO_LANGUAGE_SWEDISH */
    &options_uk,      /* RETRO_LANGUAGE_UKRAINIAN */
+   &options_cs,      /* RETRO_LANGUAGE_CZECH */
+   &options_val,     /* RETRO_LANGUAGE_CATALAN_VALENCIA */
+   &options_ca,      /* RETRO_LANGUAGE_CATALAN */
+   &options_en,      /* RETRO_LANGUAGE_BRITISH_ENGLISH */
+   &options_hu,      /* RETRO_LANGUAGE_HUNGARIAN */
 };
 #endif