from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain_core.output_parsers import StrOutputParser
from langchain_core.prompts import ChatPromptTemplate
import re
from langchain_core.runnables import RunnablePassthrough

from common.extensions import minio_client
from common.utils.model_utils import create_language_template
from .base_processor import BaseProcessor
from common.utils.business_event_context import current_event
from .processor_registry import ProcessorRegistry


def _find_first_h1(markdown: str) -> str:
    # Look for # Header (allowing spaces after #)
    match = re.search(r'^#\s+(.+)$', markdown, re.MULTILINE)
    return match.group(1).strip() if match else ""


class MarkdownProcessor(BaseProcessor):
    def __init__(self, tenant, model_variables, document_version, catalog, processor):
        super().__init__(tenant, model_variables, document_version, catalog, processor)

        self.chunk_size = catalog.max_chunk_size
        self.chunk_overlap = 0
        self.tuning = self.processor.tuning

    def process(self):
        self._log("Starting Markdown processing")
        try:
            file_data = minio_client.download_document_file(
                self.tenant.id,
                self.document_version.bucket_name,
                self.document_version.object_name,
            )

            markdown = file_data.decode('utf-8')
            title = _find_first_h1(markdown)

            self._save_markdown(markdown)
            self._log("Finished processing Markdown")
            return markdown, title
        except Exception as e:
            self._log(f"Error processing Markdown: {str(e)}", level='error')
            raise


ProcessorRegistry.register("MARKDOWN_PROCESSOR", MarkdownProcessor)