Update patch script

iinkai.py

@@ -48,6 +48,18 @@ def extract_script(script_path: str, output_path: str):
         json.dump(result, f, ensure_ascii=False, indent=2)
 
 
+def extract_dict_terms(script_path: str):
+    terms = {}
+    overrides = {}
+    with open(script_path, "r", encoding="utf-8-sig") as f:
+        for row in DictReader(f):
+            term = row['term']
+            terms[term] = row
+            if '　' in term:
+                overrides[term.split('　')[0]] = term
+    return terms, overrides
+
+
 def extract_dict(script_path: str, output_path: str):
     with open(script_path, "r", encoding="utf-8") as f:
         script_text = f.read()
@@ -133,13 +145,67 @@ def parse_message(message: str) -> List[ParsedLine]:
     return result
 
 
-def patch_script(script_path: str, m3t_path: str, output_path: str, names):
+def patch_dict(script_path: str, dict_path: str, output_path: str):
+    with open(script_path, "r", encoding="utf-8") as f:
+        script_text = f.read()
+    in_dict = False
+    dict_data = '\n'
+    start_line = None
+    end_line = None
+    for (i, line) in enumerate(script_text.splitlines()):
+        if line == "var text = '":
+            in_dict = True
+            start_line = i + 1
+        elif line == "';":
+            in_dict = False
+            end_line = i - 1
+        elif in_dict:
+            dict_data += line + "\n"
+    script = KAGScriptParser(dict_data).parse(True)
+    dicts = {}
+    with open(dict_path, "r", encoding="utf-8-sig") as f:
+        for row in DictReader(f):
+            dicts[row['term']] = row
+    new_script = []
+    term = None
+    for line in script:
+        if isinstance(line, LabelNode):
+            term = line.name[1:]
+            new_script.append(LabelNode('.' + dicts[term]['translation']))
+        elif isinstance(line, CommentNode):
+            new_script.append([TextNode(f";{line.text}")])
+        elif isinstance(line, EmptyLineNode):
+            pass
+        elif isinstance(line, list):
+            pass
+        elif isinstance(line, CommandNode):
+            if line.name == "return":
+                data = dicts[term]
+                new_script.append([TextNode(data['translation'])])
+                desc: str = data['description']
+                for d in desc.splitlines():
+                    new_script.append([TextNode(d)])
+                new_script.append(EmptyLineNode())
+                new_script.append(line)
+            else:
+                raise ValueError("Unsupported command", line.name)
+    dict_data = KAGScriptParser.serialize(new_script)
+    origin_lines = script_text.splitlines(True)
+    with open(output_path, 'w', encoding='UTF-8') as f:
+        f.writelines(origin_lines[:start_line])
+        f.write(dict_data)
+        f.write("\n")
+        f.writelines(origin_lines[end_line + 1:])
+
+
+def patch_script(script_path: str, m3t_path: str, output_path: str, names, term):
     with open(script_path, "r", encoding="utf-8") as f:
         script_text = f.read()
     script = KAGScriptParser(script_text).parse(True)
     new_script: ParsedScript = []
     messages = []
     name = None
+    terms, overrides = term
     with open(m3t_path, "r", encoding="utf-8") as f:
         for line in f:
             line = line.strip()
@@ -151,6 +217,11 @@ def patch_script(script_path: str, m3t_path: str, output_path: str, names):
                 if name is not None:
                     d['name'] = name
                 d['message'] = message.replace('\\n', '\n')
+                if name:
+                    if not d['message'].startswith('「'):
+                        d['message'] = '「' + d['message']
+                    if not d['message'].endswith('」'):
+                        d['message'] += '」'
                 messages.append(d)
                 name = None
     i = 0
@@ -159,7 +230,19 @@ def patch_script(script_path: str, m3t_path: str, output_path: str, names):
             if line.name == "page":
                 message = messages[i]['message']
                 i += 1
-                new_script.extend(parse_message(message))
+                nws = parse_message(message)
+                for nw in nws:
+                    for node in nw:
+                        if isinstance(node, TagNode):
+                            if node.name == 'wd':
+                                target = node.attributes['s']
+                                if target not in terms:
+                                    if target not in overrides:
+                                        print(message)
+                                        raise ValueError('unknown wd target', target)
+                                    node.attributes['s'] = overrides[target]
+                                node.attributes['s'] = terms[node.attributes['s']]['translation']
+                new_script.extend(nws)
             elif line.name.startswith("【") and line.name.endswith("】"):
                 name = line.name[1:-1]
                 if name in names:
@@ -168,7 +251,7 @@ def patch_script(script_path: str, m3t_path: str, output_path: str, names):
             continue
         new_script.append(line)
     if i != len(messages):
-        print(f"WARNING: processed message not matched. expected {len(message)}, actual {i}")
+        print(f"WARNING: processed message not matched. expected {len(messages)}, actual {i}, {script_path}")
     script_data = KAGScriptParser.serialize(new_script)
     with open(output_path, 'w', encoding='utf-8') as f:
         f.write(script_data)
@@ -211,8 +294,9 @@ def read_names(name_dict_path: str):
     return names
 
 
-def patch_script_auto(script_path: str, m3t_path: str, output_path: str, name_dict_path: str):
+def patch_script_auto(script_path: str, m3t_path: str, output_path: str, name_dict_path: str, dict_path: str):
     names = read_names(name_dict_path)
+    term = extract_dict_terms(dict_path)
     if isdir(script_path):
         for file in listdir(script_path):
             if not file.lower().endswith(".scn"):
@@ -228,18 +312,18 @@ def patch_script_auto(script_path: str, m3t_path: str, output_path: str, name_di
                 continue
             output_full_path = join(output_path, basename(file))
             if file == "dict.scn":
-                pass
+                patch_dict(full_path, m3t_full_path, output_full_path)
             else:
-                patch_script(full_path, m3t_full_path, output_full_path, names)
+                patch_script(full_path, m3t_full_path, output_full_path, names, term)
     else:
         pdir = dirname(output_path)
         if pdir and not isdir(pdir):
             makedirs(pdir, exist_ok=True)
         base_name = basename(script_path)
         if base_name == "dict.scn":
-            pass
+            patch_dict(script_path, m3t_path, output_path)
         else:
-            patch_script(script_path, m3t_path, output_path, names)
+            patch_script(script_path, m3t_path, output_path, names, term)
 
 
 if __name__ == "__main__":
@@ -254,8 +338,9 @@ if __name__ == "__main__":
     patch_parser.add_argument("m3t_path", help="Path to m3t file or directory")
     patch_parser.add_argument("output_path", help="Path to output KAG script file or directory")
     patch_parser.add_argument("name_dict_path", help="Path to name dict")
+    patch_parser.add_argument("dict_path", help="path to dict.csv")
     args = parser.parse_args()
     if args.command == "extract":
         extract_script_auto(args.script_path, args.output_path)
     elif args.command == "patch":
-        patch_script_auto(args.script_path, args.m3t_path, args.output_path, args.name_dict_path)
+        patch_script_auto(args.script_path, args.m3t_path, args.output_path, args.name_dict_path, args.dict_path)