import datetime
import io
from dataclasses import dataclass, field
from typing import IO, Any, Dict, Final, Optional

from unstructured.utils import requires_dependencies

# NOTE(robison) - ref: https://www.media.mit.edu/pia/Research/deepview/exif.html
EXIF_DATETIME_FMT: Final[str] = "%Y:%m:%d %H:%M:%S"


@dataclass
class Metadata:
    author: str = ""
    category: str = ""
    comments: str = ""
    content_status: str = ""
    created: Optional[datetime.datetime] = None
    identifier: str = ""
    keywords: str = ""
    language: str = ""
    last_modified_by: str = ""
    last_printed: Optional[datetime.datetime] = None
    modified: Optional[datetime.datetime] = None
    revision: Optional[int] = 0
    subject: str = ""
    title: str = ""
    version: str = ""
    description: str = ""
    namespace: str = ""

    # NOTE(robinson) - Metadata for use with image files
    exif_data: Dict[str, Any] = field(default_factory=dict)

    def to_dict(self):
        return self.__dict__


@requires_dependencies("docx")
def get_docx_metadata(
    filename: str = "",
    file: Optional[IO[bytes]] = None,
) -> Metadata:
    """Extracts document metadata from a Microsoft .docx document."""
    import docx

    if filename:
        doc = docx.Document(filename)
    elif file:
        doc = docx.Document(file)
    else:
        raise FileNotFoundError("No filename nor file were specified")

    metadata = Metadata(
        author=getattr(doc.core_properties, "author", ""),
        category=getattr(doc.core_properties, "category", ""),
        comments=getattr(doc.core_properties, "comments", ""),
        content_status=getattr(doc.core_properties, "content_status", ""),
        created=getattr(doc.core_properties, "created", None),
        identifier=getattr(doc.core_properties, "identifier", ""),
        keywords=getattr(doc.core_properties, "keywords", ""),
        language=getattr(doc.core_properties, "language", ""),
        last_modified_by=getattr(doc.core_properties, "last_modified_by", ""),
        last_printed=getattr(doc.core_properties, "last_printed", None),
        modified=getattr(doc.core_properties, "modified", None),
        revision=getattr(doc.core_properties, "revision", None),
        subject=getattr(doc.core_properties, "subject", ""),
        title=getattr(doc.core_properties, "title", ""),
        version=getattr(doc.core_properties, "version", ""),
    )

    return metadata


@requires_dependencies("openpyxl")
def get_xlsx_metadata(
    filename: str = "",
    file: Optional[IO[bytes]] = None,
) -> Metadata:
    """Extracts document metadata from a Microsoft .xlsx document."""
    import openpyxl

    if filename:
        workbook = openpyxl.load_workbook(filename)
    elif file:
        workbook = openpyxl.load_workbook(file)
    else:
        raise FileNotFoundError("No filename nor file were specified")

    metadata = Metadata(
        author=getattr(workbook.properties, "creator", ""),
        category=getattr(workbook.properties, "category", ""),
        content_status=getattr(workbook.properties, "contentStatus", ""),
        created=getattr(workbook.properties, "created", None),
        description=getattr(workbook.properties, "description", ""),
        identifier=getattr(workbook.properties, "identifier", ""),
        keywords=getattr(workbook.properties, "keywords", ""),
        language=getattr(workbook.properties, "language", ""),
        last_modified_by=getattr(workbook.properties, "lastModifiedBy", ""),
        last_printed=getattr(workbook.properties, "lastPrinted", None),
        modified=getattr(workbook.properties, "modified", None),
        namespace=getattr(workbook.properties, "namespace", ""),
        revision=getattr(workbook.properties, "revision", None),
        subject=getattr(workbook.properties, "subject", ""),
        title=getattr(workbook.properties, "title", ""),
        version=getattr(workbook.properties, "version", ""),
    )

    return metadata


@requires_dependencies("PIL")
def get_jpg_metadata(
    filename: str = "",
    file: Optional[IO[bytes]] = None,
) -> Metadata:
    """Extracts metadata from a JPG image, including EXIF metadata."""
    from PIL import Image
    from PIL.ExifTags import TAGS

    if filename:
        image = Image.open(filename)
    elif file:
        image = Image.open(io.BytesIO(file.read()))
    else:
        raise FileNotFoundError("No filename nor file were specified")

    exif_data = image.getexif()
    exif_dict: Dict[str, Any] = {}
    for tag_id in exif_data:
        tag = TAGS.get(tag_id, tag_id)
        data = exif_data.get(tag_id)
        exif_dict[tag] = data

    metadata = Metadata(
        author=exif_dict.get("Artist", ""),
        comments=exif_dict.get("UserComment", ""),
        created=_get_exif_datetime(exif_dict, "DateTimeOriginal"),
        # NOTE(robinson) - Per EXIF docs, DateTime is the last modified data
        # ref: https://www.media.mit.edu/pia/Research/deepview/exif.html
        modified=_get_exif_datetime(exif_dict, "DateTime"),
        exif_data=exif_dict,
    )

    return metadata


def _get_exif_datetime(exif_dict: Dict[str, Any], key: str) -> Optional[datetime.datetime]:
    """Converts a datetime string from the EXIF data to a Python datetime object."""
    date = exif_dict.get(key)
    if not date:
        return None

    try:
        return datetime.datetime.strptime(date, EXIF_DATETIME_FMT)
    # NOTE(robinson) - An exception could occur if the datetime is not formatted
    # using the standard EXIF datetime format
    except ValueError:
        return None