models.py

from pydantic import BaseModel, TypeAdapter, model_validator
from pathlib import Path
import json

class UnscriptedLine(BaseModel):
    """
    {
        "voice_id": "0010000793V",
        "text": "お昼頃か ギルドの研修の真っ最中ねー"
    }
    """
    voice_id: str
    text: str
    @property
    def scene_id(self):
        return self.voice_id[3:6]
    @property
    def scene_seq_id(self):
        return int(self.voice_id[6:10] or "-1")

class Line(UnscriptedLine):
    """
    {
        "character_id": "0xF",
        "voice_id": "0940010125V",
        "script_id": 49,
        "text": "おや、嬢ちゃんたちは……",
        "source_file": "C0100.txt",
        "context_prev": "",
        "context_next": "あなたが鉱山長さん？よかった、やっと見つけたわ。"
    }
    """
    character_id: str
    script_id: int
    source_file: str
    context_prev: str
    context_next: str

    @property
    def speaker_code(self):
        """voice_id 前三位的角色码，如 0010290822V -> 001"""
        return self.voice_id[:3] if len(self.voice_id) >= 3 else (self.voice_id or "")

class Conversation(BaseModel):
    lines: list[Line]
    def __len__(self) -> int:
        return len(self.lines)
    def __getitem__(self, index: int) -> Line:
        return self.lines[index]
    def __iter__(self):
        return iter(self.lines)

class RemakeCommand(BaseModel):
    """
    {
        "file": "F:\\code\\sora-script-test\\scena\\jp\\mp0000.py",
        "line": 9941,
        "column": 4,
        "type": "Command",
        "code": "Command('Cmd_text_00', [INT(10007), '<#E_0#M_0#B_0>', '何だ、エステル。', INT(10), 'どっか出かけんのか？'])",
        "normalized_args": "5,0,10007,<#E_0#M_0#B_0>,何だ、エステル。,10,どっか出かけんのか？",
        "command": "Cmd_text_00",
        "args": [
            10007,
            "何だ、エステル。どっか出かけんのか？"
        ],
        "line_corr": 9291
    }
    """
    file: str
    line: int
    column: int
    type: str
    code: str
    normalized_args: str
    command: str
    args: list
    line_corr: int | None

class RemakeLine(BaseModel):
    id: int 
    text: str
    remake_voice_id: int | None = None
    speaker: int | None = None
    function: str | None = None
    filebase: str
    lineno: int
    lineno_corr: int | None = None

    @model_validator(mode="before")
    @classmethod
    def handle_remake_commands(cls, data):
        if isinstance(data, dict):
            args = data.pop("args", None)
            if args:
                data["text"] = args[-1]
                if len(args) >= 3 and args[-3] == 11 and isinstance(args[-2], int):
                    data["remake_voice_id"] = args[-2]
                if len(args) >= 2 and args[-2] == 11 and isinstance(args[-1], int):
                    data["remake_voice_id"] = args[-1]
                    data["text"] = ""
                if isinstance(args[0], int):
                    data["speaker"] = args[0]
            function = data.pop("function", None)
            if function:
                data["function"] = function
            file = data.pop("file", None)
            if file:
                data["filebase"] = Path(file).stem
            line = data.pop("line", None)
            if line:
                data["lineno"] = line
            line_corr = data.pop("line_corr", None)
            if line_corr:
                data["lineno_corr"] = line_corr
        elif isinstance(data, RemakeCommand):
            args = data.args
            if args:
                data["text"] = args[-1]
                if len(args) >= 3 and args[-3] == 11 and isinstance(args[-2], int):
                    data["remake_voice_id"] = args[-2]
                if len(args) >= 2 and args[-2] == 11 and isinstance(args[-1], int):
                    data["remake_voice_id"] = args[-1]
                    data["text"] = ""
                if isinstance(args[0], int):
                    data["speaker"] = args[0]
            function = data.function
            if function:
                data["function"] = function
            file = data.file
            if file:
                data["filebase"] = Path(file).stem
            line = data.line
            if line:
                data["lineno"] = line
            line_corr = data.line_corr
            if line_corr:
                data["lineno_corr"] = line_corr
        return data

class RemakeConversation(BaseModel):
    lines: list[RemakeLine]
    def __len__(self) -> int:
        return len(self.lines)
    def __getitem__(self, index: int) -> RemakeLine:
        return self.lines[index]
    def __iter__(self):
        return iter(self.lines)

class UnscriptedConversation:
    lines: list[UnscriptedLine]
    def __init__(self, file: str | None = None) -> None:
        self.lines = []
        if file:
            with open(file, "r", encoding="utf-8") as f:
                adapter = TypeAdapter(list[UnscriptedLine])
                self.lines = adapter.validate_json(f.read())
        self.texts = [line.text for line in self.lines]
    def __len__(self) -> int:
        return len(self.lines)
    def __getitem__(self, index: int) -> UnscriptedLine:
        return self.lines[index]
    def __iter__(self):
        return iter(self.lines)

class Script :
  def __init__(self, file:str) -> None:
    with open(file, "r", encoding="utf-8") as f:
      adapter = TypeAdapter(list[Line])
      self.lines = adapter.validate_json(f.read())
    self.texts = [line.text for line in self.lines]
  def __len__(self) -> int:
    return len(self.lines)
  def __getitem__(self, index: int) -> Line:
    return self.lines[index]
  def __iter__(self):
    return iter(self.lines)

class RemakeScript :
  NEW_ID_START=50001
  lines: list[RemakeLine]
  def __init__(self, file:str, new_id_start:int = 50001) -> None:
    self.NEW_ID_START = new_id_start
    self.lines = []
    try:
      with open(file, "r", encoding="utf-8") as f:
        commands: list[dict] = json.load(f)
        for i, entry in enumerate(commands):
          remake_line = RemakeLine(id=self.NEW_ID_START + i, **entry)
          self.lines.append(remake_line)
    except FileNotFoundError:
      print(f"File {file} not found.")
    self.texts = [line.text for line in self.lines]
  def __len__(self) -> int:
    return len(self.lines)
  def __getitem__(self, index: int) -> RemakeLine:
    return self.lines[index]
  def __iter__(self):
    return iter(self.lines)
    

def test_lines():
    with open("script_data.json", "r", encoding="utf-8") as f:
        # The JSON file is a list of objects, so we use TypeAdapter to validate it as a list[Line]
        adapter = TypeAdapter(list[Line])
        lines = adapter.validate_json(f.read())
        for line in lines:
            print(line)

def test_unscriptedline():
    with open("additional_voice_fc.json", "r", encoding="utf-8") as f:
        adapter = TypeAdapter(list[UnscriptedLine])
        lines = adapter.validate_json(f.read())
        for line in lines:
            print(line)

def test_voice_id():
    voice_id = VoiceId(voice_id="0940010125V")
    print(voice_id.scene_id)
    print(voice_id.scene_seq_id)
    voice_empty = VoiceId(voice_id="")
    print(voice_empty.scene_id)
    print(voice_empty.scene_seq_id)

def test_remake_command():
    NEW_ID_START=50001
    with open("scena_data_jp_Command_sample.json", "r", encoding="utf-8") as f:
        adapter = TypeAdapter(list[RemakeCommand])
        lines = adapter.validate_json(f.read())
        for line in lines:
            print(line)

def test_remake_line():
    NEW_ID_START=50001
    with open("scena_data_jp_Command.json", "r", encoding="utf-8") as f:
        commands: list[dict] = json.load(f)
        for i, entry in enumerate(commands):
            remake_line = RemakeLine(id=NEW_ID_START + i, **entry)
            print(remake_line)


if __name__ == "__main__":
    # test_lines()
    # test_voice_id()
    # test_remake_command()
    # test_remake_line()
    test_unscriptedline()