Bash script to export bookmarks from Shiori as obsidian notes.

convert_shiori_to_obsidian.sh

#!/bin/bash
# Shiori is a good way to bookmark and archive webpage: https://github.com/go-shiori/shiori
# You can self-host it and use this script if you want your bookmarks as obsidian notes.
# This script assumes that shiori is running using docker.
#
# Usage: ./convert_shiori_to_obsidian.sh output_directory
# Example: ./convert_shiori_to_obsidian.sh ./obsidian_notes/

output_dir="$1"

if [ $# -ne 1 ]; then
    echo "Usage: $0 <output_directory>"
    echo "Example: $0 ./obsidian_notes/"
    exit 1
fi

echo "Fetching bookmarks from Shiori..."
# Get bookmarks from Shiori Docker container
if ! docker exec shiori shiori list > /tmp/shiori_bookmarks.txt 2>/dev/null; then
    echo "Error: Failed to fetch bookmarks from Shiori container"
    echo "Make sure Docker is running and the 'shiori' container exists"
    exit 1
fi

input_file="/tmp/shiori_bookmarks.txt"

# Database path
DB_PATH="/home/brunoamaral/containers/Shiori/data/shiori.db"

# Create output directory if it doesn't exist
mkdir -p "$output_dir"

# Function to get existing bookmark IDs from output directory
get_existing_ids() {
    if [ -d "$output_dir" ]; then
        # Find all .md files and extract id from frontmatter
        find "$output_dir" -name "*.md" -type f -exec awk '/^---$/{frontmatter=!frontmatter; next} frontmatter && /^id:/{gsub(/^id:[[:space:]]*/, ""); print; exit}' {} \; 2>/dev/null | grep -v '^$'
    fi
}

# Function to check if bookmark ID already exists
id_exists() {
    local id_to_check="$1"

    if [ -f "$existing_ids_file" ]; then
        grep -q "^${id_to_check}$" "$existing_ids_file"
        return $?
    fi

    return 1  # ID doesn't exist if file doesn't exist
}

# Get existing bookmark IDs to avoid duplicates
echo "Scanning existing notes for bookmark IDs..."
existing_ids_file="/tmp/existing_bookmark_ids.txt"
get_existing_ids > "$existing_ids_file"
existing_count=$(wc -l < "$existing_ids_file" 2>/dev/null || echo "0")
echo "Found $existing_count existing bookmark(s)"

# Initialize variables
bookmark_id=""
title=""
url=""
description=""
tags=""
new_bookmarks_count=0
skipped_bookmarks_count=0

# Function to create obsidian note
create_note() {
    if [ -n "$bookmark_id" ] && [ -n "$title" ]; then
        # Check if this bookmark ID already exists
        if id_exists "$bookmark_id"; then
            # echo "Skipping bookmark ID $bookmark_id (already exists): $title"
            ((skipped_bookmarks_count++))
            return
        fi

        # Clean title for filename (remove special characters and truncate)
        filename=$(echo "$title" | sed 's/[^a-zA-Z0-9 ]//g' | sed 's/  */ /g' | sed 's/^ *//;s/ *$//')
        filename="${filename// /_}"

        # Truncate filename to max 120 characters (leaving room for .md extension)
        if [ ${#filename} -gt 116 ]; then
            filename="${filename:0:116}"
        fi
        filename="${filename}.md"

        # Get content from SQLite database
        content=$(sqlite3 "$DB_PATH" "select bc.content from bookmark b left join bookmark_content bc on b.id = bc.docid where b.id=${bookmark_id};" 2>/dev/null)

        # Trim leading and trailing whitespace from content
        content=$(echo "$content" | sed 's/^[[:space:]]*//;s/[[:space:]]*$//')

        # Create note content
        cat > "$output_dir/$filename" << EOF
---
id: $bookmark_id
kind: bookmark
url: $url
tags: [$tags]
created: $(date -Iseconds)
---

# $title

$description

## Link
[$title]($url)

## Tags
$tags

## Content
$content
EOF

        echo "Created: $filename (ID: $bookmark_id)"
        ((new_bookmarks_count++))
    fi
}

# Process the input file
while IFS= read -r line; do
    if [[ $line =~ ^[0-9]+\. ]]; then
        # If we have a previous bookmark, create its note
        create_note

        # Extract bookmark ID and title
        bookmark_id=$(echo "$line" | sed 's/^\([0-9]*\)\..*$/\1/')
        title=$(echo "$line" | sed 's/^[0-9]*\. *\(.*\)$/\1/')

        # Reset other variables
        url=""
        description=""
        tags=""

    elif [[ $line =~ ^[[:space:]]*\> ]]; then
        # Extract URL
        url=$(echo "$line" | sed 's/^[[:space:]]*> *\(.*\)$/\1/')

    elif [[ $line =~ ^[[:space:]]*\+ ]]; then
        # Extract description
        description=$(echo "$line" | sed 's/^[[:space:]]*+ *\(.*\)$/\1/')

    elif [[ $line =~ ^[[:space:]]*# ]]; then
        # Extract tags
        tags=$(echo "$line" | sed 's/^[[:space:]]*# *\(.*\)$/\1/')
    fi
done < "$input_file"

# Don't forget the last bookmark
create_note

# Clean up temporary files
rm -f "/tmp/shiori_bookmarks.txt"
rm -f "$existing_ids_file"

echo "Conversion complete! Notes saved to: $output_dir"
echo "Statistics:"
echo "  - Existing bookmarks: $existing_count"
echo "  - New bookmarks created: $new_bookmarks_count"
echo "  - Bookmarks skipped (already exist): $skipped_bookmarks_count"