File size: 15,753 Bytes
9f4dbc4 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 | import os
import re
from datetime import datetime, timedelta, timezone
from typing import List, Dict
import psycopg2
from psycopg2.extras import Json
# ── Configuration ────────────────────────────────────────────────────────────
# Pull DATABASE_URL from the environment, falling back to a local .env file
# so the same Postgres credentials work for both this scraper and the
# Next.js / Drizzle side without having to re-export the variable.
try:
from dotenv import load_dotenv
_here = os.path.dirname(os.path.abspath(__file__))
for _candidate in (
os.path.join(_here, ".env"),
os.path.join(_here, "components", ".env"),
):
if os.path.exists(_candidate):
load_dotenv(_candidate, override=False)
except ImportError:
pass
DATABASE_URL = os.environ.get("DATABASE_URL")
if not DATABASE_URL:
raise RuntimeError(
"DATABASE_URL is not set. Define it in your environment or in a "
".env file at the project root (or in components/.env)."
)
def _connect():
"""Open a fresh Postgres connection.
We open/close per call instead of holding a global. With Neon's pooler
endpoint (the one in DATABASE_URL) each `connect` is effectively a
cheap checkout from their server-side pool, so this stays simple and
plays nicely with FastAPI's threadpool — no shared cursors between
request handlers.
"""
return psycopg2.connect(DATABASE_URL)
# ── Schema bootstrap ─────────────────────────────────────────────────────────
def init_database():
"""Initialize the database with required tables.
Idempotent: every statement uses IF NOT EXISTS, so it's safe to run on
every import and it won't fight with Drizzle's migrations if they've
already created the table.
"""
conn = _connect()
try:
with conn.cursor() as cur:
cur.execute(
"""
CREATE TABLE IF NOT EXISTS scraped_jobs (
id SERIAL PRIMARY KEY,
title TEXT,
company TEXT,
location TEXT,
link TEXT UNIQUE,
posted TEXT,
posted_datetime TIMESTAMP,
programs JSONB DEFAULT '[]'::jsonb,
schools JSONB DEFAULT '[]'::jsonb,
scraped_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
keywords TEXT,
freshness TEXT,
work_types JSONB DEFAULT '[]'::jsonb,
job_types JSONB DEFAULT '[]'::jsonb,
company_rating REAL,
contact_details JSONB DEFAULT '[]'::jsonb
)
"""
)
# Migrate existing tables that don't yet have company_rating or contact_details
cur.execute(
"""
ALTER TABLE scraped_jobs
ADD COLUMN IF NOT EXISTS company_rating REAL
"""
)
cur.execute(
"""
ALTER TABLE scraped_jobs
ADD COLUMN IF NOT EXISTS contact_details JSONB DEFAULT '[]'::jsonb
"""
)
cur.execute(
"CREATE INDEX IF NOT EXISTS idx_scraped_at "
"ON scraped_jobs(scraped_at)"
)
cur.execute(
"CREATE INDEX IF NOT EXISTS idx_link ON scraped_jobs(link)"
)
conn.commit()
finally:
conn.close()
# ── Internal helpers ─────────────────────────────────────────────────────────
def _coerce_posted_datetime(posted: str, raw_posted_dt, now_dt: datetime) -> datetime:
"""Normalize the scraper's posted_datetime into a real `datetime`.
The scraper sometimes only gives us a relative label like "3 hours
ago" or a bare date string. SQLite stored these as TEXT, but Postgres
has a real TIMESTAMP column, so we resolve everything to a `datetime`
here before insertion. Falls back to `now_dt` when neither source is
usable. Behavior matches the original SQLite version.
"""
needs_recompute = (
not raw_posted_dt
or (isinstance(raw_posted_dt, str) and len(raw_posted_dt) <= 10)
)
if posted and needs_recompute:
posted_lower = posted.lower()
match = re.search(r"(\d+)", posted_lower)
if match:
num = int(match.group(1))
if "minute" in posted_lower:
return now_dt - timedelta(minutes=num)
if "hour" in posted_lower:
return now_dt - timedelta(hours=num)
if "day" in posted_lower:
return now_dt - timedelta(days=num)
if "week" in posted_lower:
return now_dt - timedelta(weeks=num)
if "month" in posted_lower:
return now_dt - timedelta(days=num * 30)
return now_dt
if isinstance(raw_posted_dt, datetime):
return raw_posted_dt
if isinstance(raw_posted_dt, str):
try:
return datetime.fromisoformat(raw_posted_dt)
except ValueError:
pass
return now_dt
def _row_to_job(row) -> Dict:
"""Translate a SELECT row tuple into the dict shape main.py / the UI expect.
Notes:
* `programs` / `schools` come back as native Python lists because
psycopg2 auto-parses JSONB — no `json.loads` needed.
* `posted_datetime` / `scraped_at` come back as `datetime` objects;
we ISO-format them so the result survives `json.dumps` in FastAPI
without extra encoders.
"""
return {
"title": row[0],
"company": row[1],
"location": row[2],
"link": row[3],
"posted": row[4],
"posted_datetime": row[5].replace(tzinfo=timezone.utc).isoformat() if row[5] else None,
"programs": row[6] if row[6] is not None else [],
"schools": row[7] if row[7] is not None else [],
"scraped_at": row[8].replace(tzinfo=timezone.utc).isoformat() if row[8] else None,
"keywords": row[9],
"company_rating": row[10], # float or None
"contact_details": row[11] if len(row) > 11 and row[11] is not None else [],
}
# ── Public API (signatures preserved from the SQLite version) ────────────────
def save_scraped_jobs(jobs: List[Dict], search_params: Dict):
"""Save scraped jobs to database."""
conn = _connect()
saved_count = 0
try:
with conn.cursor() as cur:
now_dt = datetime.now(timezone.utc)
for job in jobs:
posted = job.get("posted", "")
posted_dt = _coerce_posted_datetime(
posted, job.get("posted_datetime"), now_dt
)
cur.execute(
"""
INSERT INTO scraped_jobs (
title, company, location, link, posted,
posted_datetime, programs, schools, keywords,
freshness, work_types, job_types, scraped_at, contact_details
)
VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s)
ON CONFLICT (link) DO UPDATE SET
title = EXCLUDED.title,
company = EXCLUDED.company,
location = EXCLUDED.location,
posted = EXCLUDED.posted,
posted_datetime = EXCLUDED.posted_datetime,
programs = EXCLUDED.programs,
schools = EXCLUDED.schools,
keywords = EXCLUDED.keywords,
freshness = EXCLUDED.freshness,
work_types = EXCLUDED.work_types,
job_types = EXCLUDED.job_types,
scraped_at = EXCLUDED.scraped_at,
contact_details = EXCLUDED.contact_details
""",
(
job.get("title"),
job.get("company"),
job.get("location"),
job.get("link"),
posted,
posted_dt,
Json(job.get("programs", []) or []),
Json(job.get("schools", []) or []),
search_params.get("keywords"),
search_params.get("freshness"),
Json(search_params.get("work_types", []) or []),
Json(search_params.get("job_types", []) or []),
now_dt,
Json(job.get("contact_details", []) or []),
),
)
saved_count += 1
conn.commit()
except Exception:
conn.rollback()
raise
finally:
conn.close()
return saved_count
def get_jobs_by_timeframe(hours_ago: int) -> List[Dict]:
"""Get jobs scraped within the specified timeframe."""
conn = _connect()
try:
with conn.cursor() as cur:
time_threshold = datetime.now(timezone.utc) - timedelta(hours=hours_ago)
cur.execute(
"""
SELECT title, company, location, link, posted, posted_datetime,
programs, schools, scraped_at, keywords, company_rating, contact_details
FROM scraped_jobs
WHERE posted_datetime >= %s
ORDER BY posted_datetime DESC
""",
(time_threshold,),
)
rows = cur.fetchall()
finally:
conn.close()
return [_row_to_job(row) for row in rows]
def get_jobs_in_timeframe(start_hours: int, end_hours: int) -> List[Dict]:
"""Get jobs scraped between start_hours and end_hours ago."""
conn = _connect()
try:
with conn.cursor() as cur:
now = datetime.now(timezone.utc)
start_time = now - timedelta(hours=start_hours)
end_time = now - timedelta(hours=end_hours)
cur.execute(
"""
SELECT title, company, location, link, posted, posted_datetime,
programs, schools, scraped_at, keywords, company_rating, contact_details
FROM scraped_jobs
WHERE posted_datetime >= %s AND posted_datetime <= %s
ORDER BY posted_datetime DESC
""",
(start_time, end_time),
)
rows = cur.fetchall()
finally:
conn.close()
return [_row_to_job(row) for row in rows]
def cleanup_old_jobs(days_old: int = 7):
"""Remove jobs older than specified days."""
conn = _connect()
try:
with conn.cursor() as cur:
cutoff_time = datetime.now(timezone.utc) - timedelta(days=days_old)
cur.execute(
"DELETE FROM scraped_jobs WHERE posted_datetime < %s",
(cutoff_time,),
)
deleted_count = cur.rowcount
conn.commit()
except Exception:
conn.rollback()
raise
finally:
conn.close()
return deleted_count
def get_binned_jobs() -> Dict[str, Dict]:
"""Fetch jobs from the last 30 days in one query and bin them by timeframe."""
conn = _connect()
try:
with conn.cursor() as cur:
now = datetime.now(timezone.utc)
start_time = now - timedelta(hours=720)
cur.execute(
"""
SELECT title, company, location, link, posted, posted_datetime,
programs, schools, scraped_at, keywords, company_rating, contact_details
FROM scraped_jobs
WHERE posted_datetime >= %s
ORDER BY posted_datetime DESC
""",
(start_time,),
)
rows = cur.fetchall()
t_1, t_2, t_5, t_24, t_168, t_720 = [], [], [], [], [], []
for row in rows:
dt = row[5]
job = _row_to_job(row)
if not dt:
continue
dt_aware = dt.replace(tzinfo=timezone.utc)
hours = (now - dt_aware).total_seconds() / 3600
if hours <= 1:
t_1.append(job)
elif hours <= 2:
t_2.append(job)
elif hours <= 5:
t_5.append(job)
elif hours <= 24:
t_24.append(job)
elif hours <= 168:
t_168.append(job)
elif hours <= 720:
t_720.append(job)
finally:
conn.close()
return {
"1_hour": {"jobs": t_1, "count": len(t_1)},
"2_hours": {"jobs": t_2, "count": len(t_2)},
"5_hours": {"jobs": t_5, "count": len(t_5)},
"24_hours": {"jobs": t_24, "count": len(t_24)},
"1_week": {"jobs": t_168, "count": len(t_168)},
"1_month": {"jobs": t_720, "count": len(t_720)},
}
init_database()
def get_rated_companies() -> set:
"""Return a set of lowercased company names that already have a non-NULL rating.
Used to skip re-fetching Serper ratings for companies we have already scored.
"""
conn = _connect()
try:
with conn.cursor() as cur:
cur.execute(
"""
SELECT DISTINCT LOWER(company)
FROM scraped_jobs
WHERE company_rating IS NOT NULL
AND company IS NOT NULL
"""
)
return {row[0] for row in cur.fetchall()}
finally:
conn.close()
def update_company_ratings(ratings: Dict[str, float]) -> int:
"""Bulk-update company_rating on all scraped_jobs rows by company name.
Uses a single multi-row UPDATE so the number of DB round-trips equals one
per unique company (not one per job row). Matching is case-insensitive.
Returns the total number of rows updated.
"""
if not ratings:
return 0
conn = _connect()
updated = 0
try:
with conn.cursor() as cur:
for company, rating in ratings.items():
cur.execute(
"""
UPDATE scraped_jobs
SET company_rating = %s
WHERE LOWER(company) = LOWER(%s)
""",
(float(rating), company),
)
updated += cur.rowcount
conn.commit()
except Exception:
conn.rollback()
raise
finally:
conn.close()
return updated
|