rodado
acfdcf97a0
Keep each vacancy's href when flattening the listing HTML, resolved to an absolute URL, so the Telegram alert links straight to the application page instead of the listing index. The extractor runs as python3 -c inside shell single quotes, so the anchor pattern's quote class is built with chr(39) rather than written inline.
127 lines
4.1 KiB
Bash
Executable File
127 lines
4.1 KiB
Bash
Executable File
#!/bin/bash
|
||
# Daily watch on sciCORE's open-positions page.
|
||
# Sends a Telegram message ONLY when a new line mentioning "software" appears.
|
||
# Other changes are logged silently so the snapshot stays current.
|
||
# Credentials come from .env: TELEGRAM_BOT_TOKEN, TELEGRAM_CHAT_ID
|
||
|
||
set -uo pipefail
|
||
|
||
REPO="/Users/polux/Projetos/linkme"
|
||
URL="${SCICORE_URL:-https://scicore.unibas.ch/about-scicore/open-positions/}" # override for testing
|
||
WANTED="software" # only changes mentioning this word are sent
|
||
STATE="$REPO/results/.scicore-watch"
|
||
SNAP="$STATE/snapshot.txt"
|
||
FAILS="$STATE/consecutive-failures"
|
||
LOG="$STATE/watch.log"
|
||
UA="Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36"
|
||
|
||
mkdir -p "$STATE"
|
||
|
||
# shellcheck disable=SC1091
|
||
# accepts either TELEGRAM_BOT_TOKEN= or the shorter TELEGRAM_BOT=
|
||
TELEGRAM_BOT_TOKEN=$(grep -E '^TELEGRAM_BOT(_TOKEN)?=' "$REPO/.env" 2>/dev/null | head -1 | cut -d= -f2-)
|
||
TELEGRAM_CHAT_ID=$(grep -E '^TELEGRAM_CHAT_ID=' "$REPO/.env" 2>/dev/null | cut -d= -f2-)
|
||
|
||
log() { echo "$(date '+%Y-%m-%d %H:%M:%S') $*" >> "$LOG"; }
|
||
|
||
notify() {
|
||
local text="$1"
|
||
if [ -z "$TELEGRAM_BOT_TOKEN" ] || [ -z "$TELEGRAM_CHAT_ID" ]; then
|
||
log "NOTIFY SKIPPED (no telegram creds in .env): $text"
|
||
return 1
|
||
fi
|
||
local code
|
||
code=$(curl -s -o "$STATE/last-telegram-response.json" -w '%{http_code}' \
|
||
"https://api.telegram.org/bot${TELEGRAM_BOT_TOKEN}/sendMessage" \
|
||
--data-urlencode "chat_id=${TELEGRAM_CHAT_ID}" \
|
||
--data-urlencode "text=${text}" \
|
||
--data-urlencode "disable_web_page_preview=true")
|
||
log "telegram http $code"
|
||
[ "$code" = "200" ]
|
||
}
|
||
|
||
HTML=$(curl -s --max-time 30 -A "$UA" "$URL")
|
||
if [ -z "$HTML" ]; then
|
||
n=$(( $(cat "$FAILS" 2>/dev/null || echo 0) + 1 ))
|
||
echo "$n" > "$FAILS"
|
||
log "FETCH FAILED (streak $n)"
|
||
[ "$n" -eq 3 ] && notify "⚠️ sciCORE watcher: 3 days in a row the page failed to load. Check the script.
|
||
$URL"
|
||
exit 0
|
||
fi
|
||
echo 0 > "$FAILS"
|
||
|
||
CURRENT=$(printf '%s' "$HTML" | SCICORE_BASE="$URL" python3 -c '
|
||
import sys, re, html, os
|
||
from urllib.parse import urljoin
|
||
BASE = os.environ.get("SCICORE_BASE", "")
|
||
h = sys.stdin.read()
|
||
m = re.search(r"<h1>Join the team!</h1>(.*?)<footer", h, re.S | re.I)
|
||
body = m.group(1) if m else h
|
||
body = re.sub(r"<(script|style).*?</\1>", "", body, flags=re.S | re.I)
|
||
body = re.sub(r"<li[^>]*>", "\n- ", body, flags=re.I)
|
||
|
||
# Keep the posting URL: <a href="X">Role</a> -> "Role -> X", so the alert can be
|
||
# tapped straight through to the application page instead of the listing index.
|
||
QUOTES = chr(34) + chr(39)
|
||
A_TAG = "<a\\s[^>]*href=([" + QUOTES + "])(.*?)\\1[^>]*>(.*?)</a>"
|
||
def keep_link(m):
|
||
href = html.unescape(m.group(2)).strip()
|
||
text = re.sub(r"<[^>]+>", "", m.group(3)).strip()
|
||
if not href or href.startswith(("#", "mailto:", "javascript:")):
|
||
return text
|
||
return "{} -> {}".format(text, urljoin(BASE, href))
|
||
body = re.sub(A_TAG, keep_link, body, flags=re.S | re.I)
|
||
body = re.sub(r"</(p|div|h[1-6]|ul|li|tr)>", "\n", body, flags=re.I)
|
||
body = re.sub(r"<[^>]+>", " ", body)
|
||
body = html.unescape(body)
|
||
lines = [re.sub(r"\s+", " ", l).strip() for l in body.split("\n")]
|
||
print("\n".join(l for l in lines if l and l not in {"-", " "}))
|
||
')
|
||
|
||
if [ -z "$CURRENT" ]; then
|
||
log "PARSE FAILED (empty extraction)"
|
||
exit 0
|
||
fi
|
||
|
||
if [ ! -f "$SNAP" ]; then
|
||
printf '%s\n' "$CURRENT" > "$SNAP"
|
||
log "BASELINE saved"
|
||
if printf '%s' "$CURRENT" | grep -qi "$WANTED"; then
|
||
notify "🚨 sciCORE listing mentions Software:
|
||
|
||
$CURRENT
|
||
|
||
$URL"
|
||
fi
|
||
exit 0
|
||
fi
|
||
|
||
if diff -q <(printf '%s\n' "$CURRENT") "$SNAP" >/dev/null; then
|
||
log "no change"
|
||
exit 0
|
||
fi
|
||
|
||
DIFF=$(diff "$SNAP" <(printf '%s\n' "$CURRENT") | sed -n 's/^> /+ /p;s/^< /- /p')
|
||
printf '%s\n' "$CURRENT" > "$SNAP"
|
||
log "CHANGED:
|
||
$DIFF"
|
||
|
||
# Only notify when a newly ADDED line contains the wanted word.
|
||
# Everything else (sysadmin roles, wording tweaks, removals) is tracked silently.
|
||
ADDED=$(printf '%s\n' "$DIFF" | grep '^+ ' || true)
|
||
if ! printf '%s' "$ADDED" | grep -qi "$WANTED"; then
|
||
log "change ignored (no \"$WANTED\" in added lines)"
|
||
exit 0
|
||
fi
|
||
|
||
notify "🚨 sciCORE posted a Software role
|
||
|
||
What changed:
|
||
$DIFF
|
||
|
||
Full listing now:
|
||
$CURRENT
|
||
|
||
$URL"
|