adding summarization and restructure project

This commit is contained in:
2025-01-09 23:36:12 +01:00
parent 706ebe25a0
commit 38d4f1ad38
28 changed files with 579 additions and 209 deletions

View File

@@ -0,0 +1,17 @@
-- +goose Up
-- +goose StatementBegin
CREATE TABLE articles (
id SERIAL PRIMARY KEY,
title VARCHAR(255) NOT NULL,
sourceUrl VARCHAR(255) NOT NULL UNIQUE,
author VARCHAR(255) NOT NULL,
content TEXT NOT NULL,
publishDate TIMESTAMP NOT NULL,
fetchDate TIMESTAMP NOT NULL
);
-- +goose StatementEnd
-- +goose Down
-- +goose StatementBegin
DROP TABLE IF EXISTS articles;
-- +goose StatementEnd

View File

@@ -0,0 +1,15 @@
-- +goose Up
-- +goose StatementBegin
ALTER TABLE articles
ADD COLUMN fts_vector tsvector GENERATED ALWAYS AS (
to_tsvector('german', coalesce(title, '') || ' ' || coalesce(content, '') || ' ' || coalesce(author, ''))
) STORED;
CREATE INDEX articles_fts_idx ON articles USING gin(fts_vector);
-- +goose StatementEnd
-- +goose Down
-- +goose StatementBegin
DROP INDEX IF EXISTS articles_fts_idx;
ALTER TABLE articles DROP COLUMN IF EXISTS fts_vector;
-- +goose StatementEnd

View File

@@ -0,0 +1,13 @@
-- +goose Up
-- +goose StatementBegin
CREATE TABLE responses (
url VARCHAR(255) NOT NULL UNIQUE PRIMARY KEY,
content BYTEA NOT NULL,
fetchDate TIMESTAMP DEFAULT CURRENT_TIMESTAMP
);
-- +goose StatementEnd
-- +goose Down
-- +goose StatementBegin
DROP TABLE IF EXISTS responses;
-- +goose StatementEnd

View File

@@ -0,0 +1,9 @@
-- +goose Up
-- +goose StatementBegin
ALTER TABLE responses ADD COLUMN processed BOOLEAN DEFAULT false;
-- +goose StatementEnd
-- +goose Down
-- +goose StatementBegin
ALTER TABLE responses DROP COLUMN IF EXISTS processed;
-- +goose StatementEnd

View File

@@ -0,0 +1,14 @@
-- +goose Up
-- +goose StatementBegin
DROP TABLE IF EXISTS responses;
-- +goose StatementEnd
-- +goose Down
-- +goose StatementBegin
CREATE TABLE responses (
url VARCHAR(255) NOT NULL UNIQUE PRIMARY KEY,
content BYTEA NOT NULL,
fetchDate TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
processed BOOLEAN DEFAULT FALSE
);
-- +goose StatementEnd

View File

@@ -0,0 +1,36 @@
-- +goose Up
-- +goose StatementBegin
BEGIN;
DROP INDEX IF EXISTS articles_fts_idx;
ALTER TABLE articles DROP COLUMN IF EXISTS fts_vector;
ALTER TABLE articles DROP COLUMN IF EXISTS author;
ALTER TABLE articles
ADD COLUMN fts_vector tsvector GENERATED ALWAYS AS (
to_tsvector('german', coalesce(title, '') || ' ' || coalesce(content, ''))
) STORED;
CREATE INDEX articles_fts_idx ON articles USING gin(fts_vector);
COMMIT;
-- +goose StatementEnd
-- +goose Down
-- +goose StatementBegin
BEGIN;
ALTER TABLE articles ADD COLUMN author VARCHAR(255) DEFAULT '';
DROP INDEX IF EXISTS articles_fts_idx;
ALTER TABLE articles DROP COLUMN IF EXISTS fts_vector;
ALTER TABLE articles
ADD COLUMN fts_vector tsvector GENERATED ALWAYS AS (
to_tsvector('german', coalesce(title, '') || ' ' || coalesce(content, '') || ' ' || coalesce(author, ''))
) STORED;
CREATE INDEX articles_fts_idx ON articles USING gin(fts_vector);
COMMIT;
-- +goose StatementEnd

View File

@@ -0,0 +1,9 @@
-- +goose Up
-- +goose StatementBegin
ALTER TABLE articles ADD COLUMN aisummary TEXT DEFAULT '';
-- +goose StatementEnd
-- +goose Down
-- +goose StatementBegin
ALTER TABLE articles DROP COLUMN IF EXISTS aisummary;
-- +goose StatementEnd