ArchiTXT is a robust tool designed to convert unstructured textual data into structured formats that are ready for database storage. It automates the generation of database schemas and creates corresponding data instances, simplifying the integration of text-based information into database systems.
Working with unstructured text can be challenging when you need to store and query it in a structured database. ArchiTXT bridges this gap by transforming raw text into organized, query-friendly structures. By automating both schema generation and data instance creation, it streamlines the entire process of managing textual information in databases.
To install ArchiTXT, make sure you have Python 3.10+ and pip installed. Then, run:
pip install architxtArchiTXT is built to work seamlessly with BRAT-annotated corpora that include pre-labeled named entities. It also requires access to a CoreNLP server, which you can easily set up using the Docker configuration available in the source repository.
$ architxt --help
Usage: architxt [OPTIONS] COMMAND [ARGS]...
ArchiTXT is a tool for structuring textual data into a valid database model.
It is guided by a meta-grammar and uses an iterative process of tree rewriting.
╭─ Options ──────────────────────────────────────────────────────────────────────────────────────────────────────╮
│ --install-completion Install completion for the current shell. │
│ --show-completion Show completion for the current shell, to copy it or customize the installation. │
│ --help Show this message and exit. │
╰────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Commands ─────────────────────────────────────────────────────────────────────────────────────────────────────╮
│ run Extract a database schema form a corpus. │
│ ui Launch the web-based UI. │
╰────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯$ architxt run --help
Usage: architxt run [OPTIONS] CORPUS_PATH
Extract a database schema form a corpus.
╭─ Arguments ────────────────────────────────────────────────────────────────────────────────────────────────────╮
│ * corpus_path PATH Path to the input corpus. [default: None] [required] │
╰────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Options ──────────────────────────────────────────────────────────────────────────────────────────────────────╮
│ --tau FLOAT The similarity threshold. [default: 0.7] │
│ --epoch INTEGER Number of iteration for tree rewriting. [default: 100] │
│ --min-support INTEGER Minimum support for tree patterns. [default: 20] │
│ --corenlp-url TEXT URL of the CoreNLP server. [default: http://localhost:9000] │
│ --gen-instances INTEGER Number of synthetic instances to generate. [default: 0] │
│ --language TEXT Language of the input corpus. [default: French] │
│ --debug --no-debug Enable debug mode for more verbose output. [default: no-debug] │
│ --help Show this message and exit. │
╰────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯To deploy the CoreNLP server using the source repository, you can use Docker Compose with the following command:
docker compose up -dTo regenerate the meta-grammar parser/lexer, use the following :
$ poetry run antlr4 -Dlanguage=Python3 metagrammar.g4 -o architxt/grammar