#!/usr/bin/env python3

import sys
import requests
import os

# Optional: caches for tag name -> ID and correspondent name -> ID
TAG_CACHE = {}
CORRESPONDENT_CACHE = {}

BASE_URL = "http://10.1.135.227:8777"  # Adjust if needed

def get_or_create_tag_id(tag_name, token):
    """
    Check if a tag named `tag_name` exists in Paperless.
    If not found, create it.
    Return the tag ID as an integer.
    """
    # If we already resolved this tag, return from cache
    if tag_name in TAG_CACHE:
        return TAG_CACHE[tag_name]
    
    headers = {"Authorization": f"Token {token}"}

    # 1) Check if tag already exists by exact name
    response = requests.get(
        f"{BASE_URL}/api/tags/",
        headers=headers,
        params={"name": tag_name}  
    )
    response.raise_for_status()
    
    results = response.json().get("results", [])
    found_id = None
    for item in results:
        if "name" in item and item["name"].lower() == tag_name.lower():
            found_id = item["id"]
            break
    
    if found_id is not None:
        # We have an existing tag
        TAG_CACHE[tag_name] = found_id
        return found_id
    else:
        # 2) Create a new tag via POST /api/tags/
        create_resp = requests.post(
            f"{BASE_URL}/api/tags/",
            headers=headers,
            json={"name": tag_name}
        )
        create_resp.raise_for_status()
        new_tag_data = create_resp.json()
        new_id = new_tag_data["id"]
        
        # Cache it
        TAG_CACHE[tag_name] = new_id
        return new_id

def get_or_create_correspondent_id(corr_name, token):
    """
    Check if a Correspondent named `corr_name` exists in Paperless.
    If not found, create it.
    Return the Correspondent ID as an integer.
    """
    # If we already resolved this correspondent, return from cache
    if corr_name in CORRESPONDENT_CACHE:
        return CORRESPONDENT_CACHE[corr_name]
    
    headers = {"Authorization": f"Token {token}"}

    # 1) Check if correspondent already exists by exact name
    response = requests.get(
        f"{BASE_URL}/api/correspondents/",
        headers=headers,
        params={"name": corr_name}
    )
    response.raise_for_status()
    
    results = response.json().get("results", [])
    found_id = None
    for item in results:
        if "name" in item and item["name"].lower() == corr_name.lower():
            found_id = item["id"]
            break
    
    if found_id is not None:
        # We have an existing correspondent
        CORRESPONDENT_CACHE[corr_name] = found_id
        return found_id
    else:
        # 2) Create a new correspondent via POST /api/correspondents/
        create_resp = requests.post(
            f"{BASE_URL}/api/correspondents/",
            headers=headers,
            json={"name": corr_name}
        )
        create_resp.raise_for_status()
        new_corr_data = create_resp.json()
        new_id = new_corr_data["id"]
        
        # Cache it
        CORRESPONDENT_CACHE[corr_name] = new_id
        return new_id

def upload_file_to_paperless(file_path, token, doc_title, devon_tags):
    """
    1) Resolve each DEVONthink tag to a Paperless tag ID (creating if necessary).
    2) Get/create the Correspondent = "KHCN".
    3) Upload the file to Paperless, including the final list of tag IDs, the doc_title, 
       and the corresponding correspondent ID.
    """
    headers = {"Authorization": f"Token {token}"}

    # Convert each DEVONthink tag to a tag ID in Paperless
    paperless_tag_ids = []
    for t in devon_tags:
        if t.strip():
            tag_id = get_or_create_tag_id(t.strip(), token)
            paperless_tag_ids.append(str(tag_id))

    # Always set this document's Correspondent to "KHCN"
    correspondent_id = get_or_create_correspondent_id("KHCN", token)

    # Build the form data
    form_data = []
    form_data.append(("title", doc_title))
    form_data.append(("correspondent", str(correspondent_id)))  # The key part
    for tid in paperless_tag_ids:
        form_data.append(("tags", tid))

    # Prepare the file for multipart/form-data
    with open(file_path, "rb") as f:
        files = {"document": (os.path.basename(file_path), f)}
        resp = requests.post(
            f"{BASE_URL}/api/documents/post_document/",
            headers=headers,
            files=files,
            data=form_data
        )

    resp.raise_for_status()
    return resp.json()

def main():
    """
    Usage:
      python3 upload.py /path/to/file PAPERLESS_TOKEN "TITLE" "TAG1||TAG2||TAG3"
    """
    if len(sys.argv) < 5:
        print("Usage: python3 upload.py /path/to/file TOKEN \"TITLE\" \"TAG1||TAG2||...\"")
        sys.exit(1)

    file_path = sys.argv[1]
    token = sys.argv[2]
    doc_title = sys.argv[3]
    raw_tags = sys.argv[4]

    # Split the passed tag string on '||' to get list of tags
    devon_tags = raw_tags.split("||") if raw_tags else []

    try:
        result = upload_file_to_paperless(file_path, token, doc_title, devon_tags)
        print("Upload started. Paperless returned:\n", result)
    except Exception as e:
        print("Error uploading document:", e)
        sys.exit(1)

if __name__ == "__main__":
    main()