uninews.database Namespace Reference

Classes
class	Database

Functions
str	sort_key (Article article, set[tuple[str, str]] known_sources)
str	search_text (Article article)
tuple[str, list]	build_where (str|None category, str|None publisher, str|None source, str|None search)
str	escape_like (str text)

Variables
int	SCHEMA_VERSION = 1
list	ARTICLE_COLUMNS = [field.name for field in fields(Article)]
str	SCHEMA
str	INSERT_ARTICLE

Detailed Description

SQLite storage for articles.

Use it from one thread only (the UI thread). Fetch in a worker, save here.

Function Documentation

build_where()

tuple[str, list] uninews.database.build_where	(	str | None	category,
		str | None	publisher,
		str | None	source,
		str | None	search
	)

escape_like()

str uninews.database.escape_like

(

str

text

)

search_text()

str uninews.database.search_text

(

Article

article

)

sort_key()

str uninews.database.sort_key	(	Article	article,
		set[tuple[str, str]]	known_sources
	)

Where an article goes in the newest-first list.

- Has a date: its date.
- No date, source fetched before: fetch time, since it appeared since last refresh.
- No date, first fetch of this source: bottom, since its real age is unknown.

Variable Documentation

ARTICLE_COLUMNS

list uninews.database.ARTICLE_COLUMNS = [field.name for field in fields(Article)]

INSERT_ARTICLE

str uninews.database.INSERT_ARTICLE

Initial value:

=  f"""
INSERT OR IGNORE INTO articles ({", ".join(ARTICLE_COLUMNS)}, sort_at, search_text)
SELECT {", ".join("?" * (len(ARTICLE_COLUMNS) + 2))}
WHERE NOT EXISTS (SELECT 1 FROM removed_links WHERE link = ?)
"""

SCHEMA

str uninews.database.SCHEMA

Initial value:

=  f"""
CREATE TABLE articles (
    id           INTEGER PRIMARY KEY AUTOINCREMENT,
    publisher    TEXT NOT NULL,
    category     TEXT NOT NULL,
    source       TEXT NOT NULL,
    title        TEXT NOT NULL,
    link         TEXT NOT NULL UNIQUE,
    published    TEXT,
    published_at TEXT,
    summary      TEXT NOT NULL,
    image_url    TEXT NOT NULL,
    fetched_at   TEXT NOT NULL,
    sort_at      TEXT NOT NULL,   -- see sort_key()
    search_text  TEXT NOT NULL    -- lowercase, accent-free copy for search
);
 
CREATE INDEX articles_by_date ON articles (sort_at DESC, id DESC);
CREATE INDEX articles_by_source ON articles (category, publisher, source);
 
-- Links removed by trim(), so they don't come back as new on the next refresh.
CREATE TABLE removed_links (
    link TEXT PRIMARY KEY
);
 
PRAGMA user_version = {SCHEMA_VERSION};
"""

SCHEMA_VERSION

int uninews.database.SCHEMA_VERSION = 1