if isinstance(element, _Comment):
    continue

import docx
from enum import Enum
from typing import List
from lxml.etree import _Comment
from docx.text.paragraph import Paragraph
from docx.document import Document as DocxDocument
from haystack.components.converters import DOCXToDocument


class DOCXTableFormat(Enum):
    """
    Supported formats for storing DOCX tabular data in a Document.
    """

    MARKDOWN = "markdown"
    CSV = "csv"

    def __str__(self):
        return self.value

    @staticmethod
    def from_str(string: str) -> "DOCXTableFormat":
        """
        Convert a string to a DOCXTableFormat enum.
        """
        enum_map = {e.value: e for e in DOCXTableFormat}
        table_format = enum_map.get(string.lower())
        if table_format is None:
            msg = f"Unknown table format '{string}'. Supported formats are: {list(enum_map.keys())}"
            raise ValueError(msg)
        return table_format



class DOCXToDocumentCustom(DOCXToDocument):
    def __init__(self, **kwargs):
        super().__init__(**kwargs)
    
    def _extract_elements(self, document: "DocxDocument") -> List[str]:
        """
        Extracts elements from a DOCX file.

        :param document: The DOCX Document object.
        :returns: List of strings (paragraph texts and table representations) with page breaks added as '\f' characters.
        """
        elements = []
        for element in document.element.body:
            if isinstance(element, _Comment):
                continue
            
            if element.tag.endswith("p"):
                paragraph = Paragraph(element, document)
                if paragraph.contains_page_break:
                    para_text = self._process_paragraph_with_page_breaks(paragraph)
                else:
                    para_text = paragraph.text
                elements.append(para_text)
            elif element.tag.endswith("tbl"):
                table = docx.table.Table(element, document)
                table_str = (
                    self._table_to_markdown(table)
                    if self.table_format == DOCXTableFormat.MARKDOWN
                    else self._table_to_csv(table)
                )
                elements.append(table_str)

        return elements