models.py

from typing import Optional, Dict, Any, List
from pydantic import BaseModel, Field, validator
from urllib.parse import urljoin


class ResponseStatus(BaseModel):
    status: str
    message: Optional[str] = Field(default=None, exclude_unset=True)

class MetadataModel(BaseModel):
    opengraph: Optional[Dict[str, str]] = None
    json_ld: Optional[Dict[str, Any]] = None
    oembed: Optional[Dict[str, Any]] = None
    spotify: Optional[Dict[str, Any]] = None

    @validator('opengraph', 'json_ld', 'oembed', 'spotify', pre=True)
    def validate_non_empty_dict(cls, v):
        if not v or not isinstance(v, dict) or not any(v.values()):
            return None
        return v

class LinkPreview(ResponseStatus):
    url: str
    title: Optional[str] = None
    description: Optional[str] = None
    image: Optional[str] = None
    favicon: Optional[str] = None
    author: Optional[str] = None
    keywords: Optional[List[str]] = Field(default_factory=list)
    language: Optional[str] = None
    metadata: Optional[MetadataModel] = None

    @validator('keywords', pre=True)
    def split_keywords(cls, v):
        if isinstance(v, str):
            return [k.strip() for k in v.split(',') if k.strip()]
        elif isinstance(v, list):
            return [k for k in v if k and isinstance(k, str)]
        return []

    @validator('image', 'favicon')
    def validate_urls(cls, v, values):
        if v and values.get('url'):
            try:
                return urljoin(values['url'], v)
            except Exception:
                return None
        return v


class SummarizedLinkPreview(LinkPreview):
    summary: Optional[str] = None