MiniWeb-demo / src /miniweb /scripts /seed_news_comments.py
sparklabuofu's picture
Update the demo
22de727 verified
Raw History Blame Contribute Delete
7.6 kB
"""Seed realistic threaded comment discussions for the Lakeport Tribune (news).
Populates the base `news_comments` table so curated local articles carry real
threaded discussions (top-level comments + nested replies) with per-user
upvotes, instead of a bare denormalized comments_count with zero backing rows.
Idempotent: skips if the table already has rows. Re-run after a DB rebuild.
Run: python src/miniweb/scripts/seed_news_comments.py
"""
import json
import os
import sqlite3
from datetime import datetime, timedelta
DB = os.environ.get("MINIWEB_DB", "data/miniweb.db")
# Commenters (news_users): id -> display_name
USERS = {1: ("alex_rivera", "Alex Rivera"),
2: ("rachel_kim", "Rachel Kim"),
3: ("samantha_liu", "Samantha Liu"),
4: ("elena_vasquez", "Elena Vasquez")}
# Per-article threads. Each entry: (author_uid, body, upvoter_uids, [replies]).
# A reply is (author_uid, body, upvoter_uids).
THREADS = {
1: [ # Meridian Systems $42M Series B
(2, "Huge win for a Lakeport company. Curious whether the new hiring will be local or remote.", [1, 3, 4],
[(3, "They mentioned expanding the Cedar Blvd office, so hopefully local roles.", [2]),
(2, "That would be great for the downtown economy.", [])]),
(4, "$42M is a serious raise for a workflow automation play. Who led the round?", [1],
[(1, "Article says a coastal growth fund led it with the existing seed investors following on.", [4])]),
(3, "Automation is eating back-office work everywhere. Smart bet.", [2]),
],
2: [ # Farmers Market year-round
(1, "Finally! Winter Saturdays without the market have been rough.", [2, 3, 4],
[(4, "Agreed. Hope the root-vegetable and cheese vendors stay on through January.", [1, 3])]),
(3, "Will the year-round schedule keep the same Riverside Park location?", [],
[(2, "Yes, same spot, just moving under the pavilion for the cold months.", [3])]),
],
4: [ # Cedar Blvd protected bike lanes 5-2
(3, "5-2 is a clearer margin than I expected. Protected lanes save lives.", [1, 2, 4],
[(1, "The two no votes were worried about parking loss on the 400 block.", [3]),
(4, "Parking studies showed occupancy under 60% there, so the tradeoff seems fair.", [1, 3])]),
(2, "Great for kids biking to Lakeport High. That stretch was genuinely dangerous.", [3, 4]),
(4, "Any timeline on construction starting?", [],
[(3, "Council notes say striping begins in spring, full barriers by summer.", [2])]),
],
5: [ # Drama club state win
(4, "So proud of these students! An original play winning state is no small thing.", [1, 2, 3],
[(2, "The writing award on top of the overall win is incredible.", [4])]),
(1, "Anyone know if they'll do a community performance before the season ends?", [3]),
],
6: [ # Lake cleanup 200 volunteers
(2, "1.2 tons is staggering. Thank you to everyone who showed up.", [1, 3, 4]),
(3, "Would love a recurring monthly cleanup instead of just the annual one.", [4],
[(4, "The organizers hinted at a spring follow-up. Fingers crossed.", [2])]),
],
7: [ # Cedar & Sage best new restaurant
(1, "Well deserved. The mushroom toast alone is worth the wait for a table.", [2, 4],
[(3, "The wait is the only downside. Two weeks out for a Friday now.", [1])]),
(4, "Great to see a local spot beat out the coastal chains for this.", [3]),
],
11: [ # Record October rainfall flooding
(3, "The Maple Ln underpass floods every heavy rain. When does the drainage fix happen?", [1, 2],
[(2, "Public works said the culvert upgrade is in next year's budget.", [3])]),
(4, "Stay safe everyone, turn around don't drown on the low crossings.", [1, 2, 3]),
],
12: [ # Budget flat property tax
(2, "Third year flat is impressive given inflation. Curious what got cut to hold the line.", [4],
[(1, "Mostly deferred capital projects, per the finance director's summary.", [2])]),
(3, "Flat rate is nice but I'd rather they fully fund the library expansion.", [1]),
],
15: [ # Board game convention 400
(1, "400 people! The tabletop scene here is way bigger than folks realize.", [2, 3]),
(4, "The cooperative-games tournament was the highlight for me this year.", [1],
[(3, "Same. Hoping they add a legacy-campaign track next year.", [4])]),
],
18: [ # Library maker space
(3, "A 3D printer AND a laser cutter at the public library? Incredible resource.", [1, 2, 4],
[(2, "And it's free with a library card. Take that, expensive maker memberships.", [1, 3])]),
(4, "Hope they run beginner workshops so it's not just for the already-experienced.", [1]),
],
19: [ # Coffee roasters 10th anniversary
(2, "Ten years already! Their single-origin Ethiopian pour-over is a Lakeport staple.", [1, 3, 4]),
(1, "The block party was packed. Great to see Main Street that lively on a Saturday.", [4],
[(4, "The live music plus free samples was a perfect combo.", [1, 2])]),
],
}
def main():
con = sqlite3.connect(DB)
con.row_factory = sqlite3.Row
existing = con.execute("SELECT COUNT(*) FROM news_comments").fetchone()[0]
if existing:
print(f"news_comments already has {existing} rows — skipping seed.")
con.close()
return
# Anchor each article's comments just after its publish date (dates are
# intentionally static/past — do not shift them).
dates = {r["id"]: r["date"]
for r in con.execute("SELECT id, date FROM news_articles")}
rows = []
cid = 0
for aid in sorted(THREADS):
base_date = dates.get(aid) or "2025-11-01"
try:
base_dt = datetime.strptime(base_date[:10], "%Y-%m-%d") + timedelta(hours=6)
except ValueError:
base_dt = datetime(2025, 11, 1, 6)
offset = 0
for entry in THREADS[aid]:
author_uid, body, upvoters = entry[0], entry[1], entry[2]
replies = entry[3] if len(entry) > 3 else []
cid += 1
parent_cid = cid
uname, dname = USERS[author_uid]
ts = (base_dt + timedelta(hours=offset)).isoformat() + "Z"
offset += 3
rows.append((parent_cid, aid, 0, author_uid, uname, dname, body, ts,
len(upvoters), json.dumps(upvoters)))
for r_uid, r_body, r_upvoters in replies:
cid += 1
r_uname, r_dname = USERS[r_uid]
r_ts = (base_dt + timedelta(hours=offset)).isoformat() + "Z"
offset += 2
rows.append((cid, aid, parent_cid, r_uid, r_uname, r_dname,
r_body, r_ts, len(r_upvoters), json.dumps(r_upvoters)))
con.executemany(
"INSERT INTO news_comments "
"(id, article_id, parent_id, user_id, username, display_name, body, "
" posted_at, upvotes, upvoted_by) VALUES (?,?,?,?,?,?,?,?,?,?)",
rows,
)
con.commit()
total = con.execute("SELECT COUNT(*) FROM news_comments").fetchone()[0]
replies = con.execute("SELECT COUNT(*) FROM news_comments WHERE parent_id>0").fetchone()[0]
arts = con.execute("SELECT COUNT(DISTINCT article_id) FROM news_comments").fetchone()[0]
print(f"Seeded {total} comments ({replies} nested replies) across {arts} articles.")
con.close()
if __name__ == "__main__":
main()