A self-contained booru-like web server for imgbrd-grabber

The interface based on a stripped down Gelbooru 1.0 with autocompletion.

Workflow

• Save images with imgbrd-grabber to a specific path and filename and store metadata with sqlite
• Copy/Symlink image directory to localbooru/data/images and metadata.sqlite3 database to localbooru/data/db/
• On startup grabber-viewer reads all metadata, associates it with saved images and is ready to serve a booru interface

Grabber setup

Settings > Save > Filename

• Default: %md5:maxlength=2%/%md5%.%ext%

Settings > Commands > Database Credentials

• Driver: QSQLITE
• Database: Put the metadata.sqlite3 filename here Settings > Commands > Database Commands
• Image: INSERT INTO posts(website, origid, creation_date, hash, image, height, width, ext, rating, tags) VALUES("%websitename%", "%id%", (DATETIME('now')), "%md5%", "%md5%.%ext%", %height%, %width%, ".%ext%", "%rating%", " %allo% ");

Usage

simply run python3 local_booru.py

Dependencies

• python3
• python3-cherrypy3
• python3-jinja2
• python3-pil

Advanced usage

• It supports importing tag lists generated by grabber's tag loader, copy tags.txt and tag-types.txt to a directory inside localbooru/data/tags, multiple directories can be used to join tags from different sites.
• With some work one could modify it to use alternative filenames