Want your good news closer to home?
{/if} +No highlights yet today β try a calmer filter, or check back soon.
{/if} {:else if feed.length} - {#if selected === 'browse'} + {#if selected === 'latest'} {#if homeEditing || (!homeValue && !homePromptDismissed)}Want good news closer to home?
@@ -971,6 +1008,11 @@ .linkish { background: none; border: none; color: var(--accent-deep); font: inherit; font-size: 0.86rem; cursor: pointer; text-decoration: underline; padding: 0; } .homebar { font-size: 0.86rem; color: var(--muted); margin: 0 0 16px; } + .briefscope { display: flex; gap: 8px; align-items: center; margin: 0 0 16px; } + .bs-btn { font: inherit; font-size: 0.88rem; font-weight: 600; padding: 7px 16px; border: 1px solid var(--line); + border-radius: 999px; background: var(--bg); color: var(--ink); cursor: pointer; } + .bs-btn.on { border-color: var(--accent); background: var(--accent-soft); color: var(--accent-deep); } + .bs-change { margin-left: auto; } .feed-section { grid-column: 1 / -1; margin: 8px 0 2px; font-family: var(--label); font-size: 0.78rem; text-transform: uppercase; letter-spacing: 0.06em; color: var(--muted); } .grid > .feed-section:first-child { margin-top: 0; } diff --git a/goodnews/api.py b/goodnews/api.py index 7a4c525..7740827 100644 --- a/goodnews/api.py +++ b/goodnews/api.py @@ -2186,25 +2186,42 @@ def create_app() -> FastAPI: limit: int = Query(10, ge=1, le=50), prefs: str | None = Query(None), exclude: str = Query("", description="comma-separated article ids the reader has dismissed"), + home: str | None = Query(None, max_length=8, description="local-first highlights: 'US' or 'US-NY'"), ) -> BriefResponse: # The default highlights are global (date-keyed, no session) β edge-cacheable # so a new visitor's "Gathering the good newsβ¦" resolves from their POP, not - # a pull to the residential origin. Personal filters stay private. - shareable = not prefs and not exclude.strip() + # a pull to the residential origin. Personal filters (incl. a home) stay private. + home_country = home_state = None + if home: + parts = home.upper().split("-", 1) + home_country = parts[0][:2] or None + if home_country == "US" and len(parts) > 1: + home_state = parts[1][:2] or None + shareable = not prefs and not exclude.strip() and not home_country response.headers["Cache-Control"] = _EDGE_FEED if shareable else _PRIVATE fp = prefs_from_json(prefs) now = datetime.now(timezone.utc) excl = {int(x) for x in exclude.split(",") if x.strip().lstrip("-").isdigit()} with get_conn() as conn: - data = queries.brief(conn, brief_date=date, limit=limit) + if home_country: + # The reader's home leads the landing: local good news first, blended out + # to country/world so it's always a full, sexy set. Over-fetch to survive + # dismissal/boundary filtering, then cap to limit. + meta = queries.brief(conn, brief_date=date, limit=1) + data = {"brief_date": meta["brief_date"], "title": "Close to home", "created_at": meta.get("created_at")} + pool = queries.home_brief(conn, home_country, home_state, limit=limit + 12) + else: + data = queries.brief(conn, brief_date=date, limit=limit) + pool = data["items"] # Drop dismissed (replaced-away) items and anything the reader's # boundaries hide; avoid-terms take precedence over curation. - items = [a for a in data["items"] if a["id"] not in excl] + items = [a for a in pool if a["id"] not in excl] if not fp.is_empty(): items = filter_articles(items, fp, now) + items = items[:limit] # home mode over-fetches to survive filtering; cap here # Keep the highlights full: if a boundary or a dismissal removed a # story, top up with other readable, boundary-respecting good news - # rather than show fewer. + # rather than show fewer. (Home mode's home_brief already blends to world.) if len(items) < limit: have = {a["id"] for a in items} | excl pool = queries.feed( diff --git a/goodnews/geo.py b/goodnews/geo.py index 71dbe88..0cbbfc2 100644 --- a/goodnews/geo.py +++ b/goodnews/geo.py @@ -213,10 +213,16 @@ def tag_articles(conn: sqlite3.Connection, client: LocalModelClient, limit: int for r in rows: try: store_geo(conn, r["id"], classify_geo(client, r)) + # Keep live auth/admin writes healthy while the scheduled cycle runs. + # Geo classification calls the LLM per article; if we batch commits, the + # first stored article opens a write transaction that can stay open while + # the next several LLM calls run. That starves login/session writes long + # enough to trip SQLite's busy timeout. Commit each successful article so + # the writer lock is held for milliseconds, not minutes. + conn.commit() tagged += 1 except Exception: # noqa: BLE001 β non-fatal, like other cycle steps + conn.rollback() errors += 1 - if (tagged + errors) % 25 == 0: - conn.commit() conn.commit() return {"candidates": len(rows), "tagged": tagged, "errors": errors} diff --git a/goodnews/queries.py b/goodnews/queries.py index a0a39ce..630b460 100644 --- a/goodnews/queries.py +++ b/goodnews/queries.py @@ -259,6 +259,51 @@ def reindex_search(conn: sqlite3.Connection) -> int: return conn.execute("SELECT COUNT(*) FROM article_search").fetchone()[0] +def home_brief(conn: sqlite3.Connection, home_country: str, home_state: str | None = None, + limit: int = 7, window_days: int = 3) -> list[dict]: + """Local-first landing highlights. Leads with high/medium-confidence local good news, + then blends out to your country and the world so the set is always full (never the + sad thin-local look), and prefers already-summarized stories so the calm read stays + rich. Brief-shaped rows (incl. summary) tagged with a section, best-first within tier. + """ + if home_state: + near = ("(g.confidence IN ('high','medium') AND EXISTS (SELECT 1 FROM article_places p " + "WHERE p.article_id = a.id AND p.country_code = ? AND p.state_code = ?))") + country = "EXISTS (SELECT 1 FROM article_places p WHERE p.article_id = a.id AND p.country_code = ?)" + section_case = f"CASE WHEN {near} THEN 0 WHEN {country} THEN 1 ELSE 2 END" + section_params = [home_country, home_state, home_country] + else: + near = ("(g.confidence IN ('high','medium') AND EXISTS (SELECT 1 FROM article_places p " + "WHERE p.article_id = a.id AND p.country_code = ?))") + section_case = f"CASE WHEN {near} THEN 0 ELSE 2 END" # no "country" tier without a state + section_params = [home_country] + rows = conn.execute( + f""" + SELECT {_ARTICLE_COLUMNS}, + sm.summary AS summary, + {section_case} AS section_rank, + (sm.summary IS NOT NULL) AS has_summary + FROM articles a + JOIN sources src ON src.id = a.source_id + JOIN article_scores s ON s.article_id = a.id + LEFT JOIN article_geo g ON g.article_id = a.id + LEFT JOIN article_summaries sm ON sm.article_id = a.id + WHERE a.duplicate_of IS NULL AND src.content_visible = 1 AND s.accepted = 1 + AND a.discovered_at >= datetime('now', ?) + ORDER BY section_rank ASC, has_summary DESC, rank_score DESC, + COALESCE(a.published_at, a.discovered_at) DESC + LIMIT ? + """, + section_params + [f"-{window_days} days", limit], + ).fetchall() + out = [] + for r in rows: + d = dict(r) + d["__section"] = {0: "near", 1: "country", 2: "world"}.get(d.pop("section_rank", 2), "world") + out.append(d) + return out + + def brief(conn: sqlite3.Connection, brief_date: str | None = None, limit: int = 10) -> dict: """Return a stored daily brief (latest if no date) with its ranked items.""" target_date = brief_date or _latest_brief_date(conn) diff --git a/scripts/substance_audit.py b/scripts/substance_audit.py new file mode 100644 index 0000000..b2cb226 --- /dev/null +++ b/scripts/substance_audit.py @@ -0,0 +1,141 @@ +#!/usr/bin/env python3 +"""PROTOTYPE substance audit (not production). + +The classifier scores emotional TONE (cortisol/ragebait/constructive) but not +SUBSTANCE, so pleasant-but-empty filler (evergreen how-tos, B2B SEO, product +listicles, recipes) slips through. Before adding a `not_newsworthy` rejection +dimension to the live classifier, measure whether the model can reliably tell +genuine news from filler against Codex's rubric, and what the reject rate would be. + +Read-only over a sample; writes a scratch JSON + prints a report. Does NOT change +the classifier or reject anything. + + .venv/bin/python scripts/substance_audit.py --limit 250 --base-url http://127.0.0.1:8080/v1 +""" +from __future__ import annotations + +import argparse +import json +from collections import Counter +from pathlib import Path + +from goodnews.cli import _default_db +from goodnews.db import connect +from goodnews.llm import LocalModelClient, parse_classifier_json + +# Codex's rubric. KEEP = real news; the rest are "positive but not news" filler. +KINDS = ("news_event", "finding", "announcement", "feature_human_interest", + "evergreen_advice", "marketing", "product_listicle", "opinion", "other") +FILLER = {"evergreen_advice", "marketing", "product_listicle"} + +SYSTEM = ( + "You judge whether a story is genuine NEWS or content-mill filler for a calm " + "good-news site. GOOD (keep): a specific event or achievement, a recent " + "development, a research finding, a credible announcement, or human/community/" + "science/environmental uplift tied to something that actually happened. FILLER " + "(not news): evergreen how-to/advice, marketing or B2B service explainers, generic " + "'why X matters' SEO pieces, product round-ups/listicles, recipes. Judge SUBSTANCE, " + "not tone β pleasant and non-negative is NOT the same as newsworthy. When genuinely " + "unsure, lean KEEP (don't reject real good news). Reply with ONLY a JSON object." +) +INSTRUCT = ( + "Return JSON exactly like:\n" + '{"kind": "