#!/usr/bin/env -S uv run --quiet
# /// script
# requires-python = ">=3.12"
# dependencies = [
#     "pymarkdownlnt",
# ]
# ///

import xml.etree.ElementTree as ET
from datetime import datetime
import zipfile
import os
import re
import pymarkdown

import xml.etree.ElementTree as ET


class Location:
    def __init__(self, uid, title, address, lat, lng, zoom):
        self.uid = uid
        self.title = title
        self.address = address
        self.lat = lat
        self.lng = lng
        self.zoom = zoom

    def __repr__(self):
        return (
            f"Location(uid='{self.uid}', title='{self.title}', "
            f"address='{self.address}', lat={self.lat}, lng={self.lng}, zoom={self.zoom})"
        )


class Attachment:
    def __init__(self, uid, entry_uid, type, filename, position):
        self.uid = uid
        self.entry_uid = entry_uid
        self.type = type
        self.filename = filename
        self.position = position

    def __repr__(self):
        return (
            f"Attachment(uid='{self.uid}', entry_uid='{self.entry_uid}', "
            f"type='{self.type}', filename='{self.filename}', position={self.position})"
        )


def parse_locations_xml(xml_file_path):
    """
    Parses an XML file containing location data into a list of Location objects.

    Args:
        xml_file_path (str): The path to the XML file.

    Returns:
        list: A list of Location objects.
    """
    tree = ET.parse(xml_file_path)
    root = tree.getroot()

    locations = []
    for r_element in root.findall("r"):
        uid = r_element.find("uid").text if r_element.find("uid") is not None else ""
        title = (
            r_element.find("title").text if r_element.find("title") is not None else ""
        )
        address = (
            r_element.find("address").text
            if r_element.find("address") is not None
            else ""
        )

        lat_text = (
            r_element.find("lat").text if r_element.find("lat") is not None else "0.0"
        )
        lng_text = (
            r_element.find("lng").text if r_element.find("lng") is not None else "0.0"
        )
        zoom_text = (
            r_element.find("zoom").text if r_element.find("zoom") is not None else "0"
        )

        try:
            lat = float(lat_text)
        except ValueError:
            lat = 0.0

        try:
            lng = float(lng_text)
        except ValueError:
            lng = 0.0

        try:
            zoom = int(zoom_text)
        except ValueError:
            zoom = 0

        location = Location(uid, title, address, lat, lng, zoom)
        locations.append(location)
    return locations


def parse_attachments_xml(xml_file_path):
    """
    Parses an XML file containing attachment data into a list of Attachment objects.

    Args:
        xml_file_path (str): The path to the XML file.

    Returns:
        list: A list of Attachment objects.
    """
    tree = ET.parse(xml_file_path)
    root = tree.getroot()

    attachments = []
    for r_element in root.findall("r"):
        uid = r_element.find("uid").text if r_element.find("uid") is not None else ""
        entry_uid = r_element.find("entry_uid").text if r_element.find("entry_uid") is not None else ""
        type = r_element.find("type").text if r_element.find("type") is not None else ""
        filename = r_element.find("filename").text if r_element.find("filename") is not None else ""
        position_text = r_element.find("position").text if r_element.find("position") is not None else "0"

        try:
            position = int(position_text)
        except ValueError:
            position = 0

        attachment = Attachment(uid, entry_uid, type, filename, position)
        attachments.append(attachment)
    return attachments


def write_files():
    """
    Parses the Diaro XML data from a string, creates markdown files,
    and zips them up.
    """
    data = open('DiaroBackup.xml').read()
    root = ET.fromstring(data)

    folders = {
        folder.find('uid').text: folder.find('title').text
        for folder in root.findall(".//table[@name='diaro_folders']/r")
    }
    # locations = {
    #     location.find('uid').text: location.find('name').text
    #     for location in root.findall(".//table[@name='diaro_locations']/r")
    # }
    # moods = {
    #     mood.find('uid').text: mood.find('name').text
    #     for mood in root.findall(".//table[@name='diaro_moods']/r")
    # }

    entries = root.findall(".//table[@name='diaro_entries']/r")

    output_dir = "diaro_markdown_entries"
    if not os.path.exists(output_dir):
        os.makedirs(output_dir)

    markdown_files = []

    xml_path = "./locations.xml"
    parsed_locations = parse_locations_xml(xml_path)

    xml_path_attach = "./attachments.xml"
    parsed_attach = parse_attachments_xml(xml_path_attach)

    for entry in entries:
        date_ms = int(entry.find('date').text)
        dt_object = datetime.fromtimestamp(date_ms / 1000)
        date_str = dt_object.strftime("%Y-%m-%d %H:%M:%S %z")

        text_element = entry.find('text')
        text = text_element.text if text_element is not None and text_element.text else "No content"


        title_element = entry.find('title')
        title = title_element.text if title_element is not None and title_element.text else "Untitled"

        if title == "Untitled":
            title = re.sub(r"[^\w]", " ", text.split('.')[0][:30])
            # Use the first X characters of text if title is "Untitled"

        mood_element = entry.find('mood')
        mood = title_element.text if mood_element is not None and mood_element.text else "3"

        if mood == "1":
            mood = "5-🥰"
        if mood == "2":
            mood = "4-🙂"
        if mood == "3":
            mood = "3-😑"
        if mood == "4":
            mood = "2-🙁"
        if mood == "5":
            mood = "1-😢"


        media_path = "../media/"
        attachments = ""
        for i in parsed_attach:
            if i.entry_uid == entry.find('uid').text:
                attachments = "\n\n---------------------------------------------------------------"
                if i.type == "image":
                    attachments += f"\n\n![{i.filename}]({media_path}{i.filename})\n\n"
                elif i.type == "video":
                    attachments += f"\n\n[Video: {i.filename}]({media_path}{i.filename})\n\n"
                elif i.type == "audio":
                    attachments += f"\n\n[Audio: {i.filename}]({media_path}{i.filename})\n\n"
                else:
                    attachments += f"\n\n![{i.filename}]({media_path}{i.filename})\n\n"


        # Create a safe filename
        safe_title = re.sub(r'[^\w\s-]', '', title).strip().lower()
        safe_title = re.sub(r'[-\s]+', '-', safe_title) if safe_title else "entry"
        filename = f"{dt_object.strftime('%Y-%m-%d')}-{safe_title}.md"



        # Front Matter
        front_matter = ""
        front_matter += "---\n"
        if title:
            front_matter += f"title: {title}\n"
        else:
            front_matter += f"title: Entry on {dt_object.strftime('%Y-%m-%d')}\n"


        front_matter += f"date: {date_str.rstrip()}\n"

        folder_uid = entry.find('folder_uid').text
        if folder_uid and folder_uid in folders:
            front_matter += f"folder: {folders[folder_uid]}\n"

        # location_uid = entry.find('location_uid').text
        # if location_uid and location_uid in locations:
        #     front_matter += f"location: {locations[location_uid]}\n"

        location_element = entry.find('location_uid')
        location = location_element.text if location_element is not None and location_element.text else "N/A"

        if location != "N/A":
            # Find the location in the parsed locations
            location_obj = next((loc for loc in parsed_locations if loc.uid == location), None)
            if location_obj:
                location_obj = location_obj
            else:
                location_obj = "Unknown Location"

        if location != "N/A":
            front_matter += f"location: {location_obj.lat},{location_obj.lng}\n"
            front_matter += f"location_title: {location_obj.title}\n"
            front_matter += f"location_address: {location_obj.address}\n"
        else:
            front_matter += "location: Unknown\n"

        if mood:
            front_matter += f"mood: {mood}\n"

        tags_text = entry.find('tags').text
        if tags_text and tags_text.strip(',') :
            tags = [tag.strip() for tag in tags_text.split(',') if tag.strip()]
            front_matter += f"tags: {tags}\n"

        # mood_uid = entry.find('mood').text
        # if mood_uid and mood_uid in moods:
        #     front_matter += f"mood: {moods[mood_uid]}\n"

        weather_temp = entry.find('weather_temperature').text
        weather_icon = entry.find('weather_icon').text
        weather_desc = entry.find('weather_description').text

        if weather_temp:
            front_matter += f"temp_c: {weather_temp}\n"
        if weather_icon:
            weather_data = weather_icon.split("-")

            time_of_day = weather_data[0]
            if time_of_day == "day":
                front_matter += "time_of_day: day\n"
            elif time_of_day == "night":
                front_matter += "time_of_day: night\n"
            else:
                front_matter += "time_of_day: unknown\n"

            weather_icon = weather_data[1:]
            if weather_icon:
                if 'alt' in weather_icon:
                    weather_icon.remove('alt')
                front_matter += f"weather: [{', '.join(weather_icon)}]\n"

        if weather_desc:
            front_matter += f"forcast: {weather_desc}\n"

        if "-----BEGIN PGP MESSAGE-----" in text:
            front_matter += "encrypted: true\n"
        else:
            front_matter += "encrypted: false\n"

        front_matter += "---\n\n"

        converted_entry = front_matter + text + attachments + "\n"

        # Write markdown file
        filepath = os.path.join(output_dir, filename)
        with open(filepath, 'w', encoding='utf-8') as f:
            f.write(converted_entry)
        markdown_files.append(filepath)


if __name__ == "__main__":
    write_files()

I recently made the switch from [Diaro](https://diaroapp.com) to [Obsidian](https://obsidian.md/) for my personal diary. While I had a great experience with Diaro - their lifetime subscription was a steal at around $20, and their Dropbox sync never lost a single entry over three or four years. Diaro's feature to automatically tag location, time, and date from uploaded photos was incredibly convenient, and I'm hoping to find a similar workflow in Obsidian.

## Extracting the Data

Getting my data out of Diaro proved to be the trickiest part. Diaro offered various export options (CSV, text, PDF), but these varied by platform (mobile vs. web) and none provided an easy way to export photos alongside the entries. Ripping images from PDFs was an absolute nightmare.

Fortunately, I dug into the Dropbox backend Diaro used for syncing. There, I discovered XML files containing all my entries and, more importantly, all my photos as JPEGs. The XML structure was relatively sane, with attachments, locations, and the diary entries themselves all contained within one large document, which made parsing manageable.

## Custom Scripting for Conversion

To get everything into a standardized Markdown format with YAML front matter and embedded images, I cooked up some Python magic. This conversion process took me a solid two to three days over a weekend. While not "crazy" in terms of effort, I do wish Diaro offered more robust export options – perhaps JSON or CSVs with easily linked attachments, or even a dedicated photo dump.

That said, for a $20 lifetime license and the peace of mind that came with never losing a single post (kudos to Dropbox for the reliable backend!), I can't really complain too much. The Diaro app itself was consistently reliable across all my devices, and the website was solid.

I'll include a link to the Python script, or perhaps embed it directly in this post, just in case it can help someone else migrate their notes from Diaro to Obsidian or another Markdown-based diary application down the road.

Thanks for reading!
--- over From this service to either obsidian Or another Mark down based Diary application. Thanks!