feat: upgrade scraper - multi keyword, --once mode, GitHub Actions workflow

.env.example

@@ -1,2 +1,2 @@
 SUPABASE_URL=your_supabase_url_here
-SUPABASE_KEY=your_supabase_anon_key_here
+SUPABASE_ANON_KEY=your_supabase_anon_key_here

.github/workflows/scrape.yml

@@ -0,0 +1,28 @@
+name: scrape
+
+on:
+  schedule:
+    - cron: "0 3 * * *"
+  workflow_dispatch:
+
+jobs:
+  run:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+
+      - name: Install dependencies
+        run: pip install -r requirements.txt
+
+      - name: Install Playwright Chromium
+        run: python -m playwright install --with-deps chromium
+
+      - name: Run scraper
+        run: python run.py --once
+        env:
+          SUPABASE_URL: ${{ secrets.SUPABASE_URL }}
+          SUPABASE_ANON_KEY: ${{ secrets.SUPABASE_ANON_KEY }}

base.py

@@ -1,11 +1,29 @@
 import random
 import time
 
+
 def human_delay(min_sec=2, max_sec=6):
     time.sleep(random.uniform(min_sec, max_sec))
 
 
+def short_delay(min_sec=1, max_sec=2):
+    time.sleep(random.uniform(min_sec, max_sec))
+
+
 def jitter_scroll(page):
     page.evaluate("""
         window.scrollTo(0, document.body.scrollHeight / 2);
-    """)
+    """)
+
+
+def goto_with_retry(page, url, timeout=45000, attempts=3):
+    for i in range(attempts):
+        try:
+            page.goto(url, wait_until="commit", timeout=timeout)
+            return True
+        except Exception as e:
+            if "ERR_NAME_NOT_RESOLVED" in str(e) or "ERR_CONNECTION" in str(e):
+                time.sleep(2 * (i + 1))
+                continue
+            raise
+    return False

cleaner.py

@@ -23,4 +23,7 @@ def clean_job(job):
         "location": (job.get("location") or "").strip() or "Indonesia",
         "description": (job.get("description") or "").strip(),
         "skills": job.get("skills", []),
+        "type": job.get("type", "Full-time"),
+        "salary": (job.get("salary") or "").strip() or None,
+        "contact_info": (job.get("contact_info") or "").strip() or None,
     }

extract.py

@@ -0,0 +1,47 @@
+import re
+
+
+def extract_salary(text):
+    if not text:
+        return None
+
+    patterns = [
+        r"Rp\s?[\d.,]+\s?(rb|ribu|juta|jt|miliar|m)?",
+        r"\$\s?[\d.,]+\s?(k|m)?",
+        r"IDR\s?[\d.,]+",
+        r"[\d.,]+\s?(juta|jt|rb|ribu|miliar)\s?(per|/)?\s?(bulan|tahun|month|year)?",
+    ]
+    for pattern in patterns:
+        match = re.search(pattern, text, re.IGNORECASE)
+        if match:
+            return match.group(0).strip()
+    return None
+
+
+def extract_job_type(text):
+    if not text:
+        return "Full-time"
+
+    lowered = text.lower()
+    if re.search(r"\bintern(ship)?\b", lowered):
+        return "Internship"
+    if re.search(r"\bpart[- ]?time\b", lowered):
+        return "Part-time"
+    if re.search(r"\bcontract\b|\bfreelance\b", lowered):
+        return "Contract"
+    return "Full-time"
+
+
+def extract_contact(text):
+    if not text:
+        return None
+
+    email = re.search(r"[\w.+-]+@[\w-]+\.[\w.-]+", text)
+    if email:
+        return email.group(0)
+
+    phone = re.search(r"(?:\+62|0)\s?8[\d\s-]{8,12}", text)
+    if phone:
+        return phone.group(0).strip()
+
+    return None

jobstreet.py

@@ -1,122 +1,156 @@
+import logging
+
 from bs4 import BeautifulSoup
-from browser import get_browser
-from base import human_delay, jitter_scroll
 from urllib.parse import urljoin
+
+from base import goto_with_retry, human_delay, jitter_scroll, short_delay
+from browser import get_browser
+from extract import extract_contact, extract_job_type, extract_salary
 from skills_keyword import extract_skills
 
+log = logging.getLogger(__name__)
 
-def scrape_jobstreet(keyword="developer"):
-    p, browser, context = get_browser()
-    page = context.new_page()
 
-    base_url = "https://id.jobstreet.com"
-    url = f"{base_url}/id/{keyword}-jobs"
+def scrape_jobstreet(keyword="developer", max_jobs=25):
+    p, browser, context = get_browser()
+    jobs = []
 
     try:
-        page.goto(url, wait_until="networkidle", timeout=30000)
-    except Exception:
-        page.goto(url, wait_until="domcontentloaded", timeout=30000)
+        page = context.new_page()
 
-    human_delay()
-    jitter_scroll(page)
-    human_delay()
+        base_url = "https://id.jobstreet.com"
+        slug = keyword.strip().lower().replace(" ", "-")
+        url = f"{base_url}/id/{slug}-jobs"
 
-    soup = BeautifulSoup(page.content(), "html.parser")
+        if not goto_with_retry(page, url):
+            log.warning("[JobStreet] Gagal memuat halaman pencarian: %s", url)
+            return jobs
 
-    jobs = []
-
-    # JobStreet job cards are wrapped in <article> elements
-    cards = soup.select("article[data-testid='job-card'], article[data-automation='job-card'], article")
+        human_delay()
+        jitter_scroll(page)
+        human_delay()
 
-    for card in cards:
         try:
-            # Job title link
-            title_el = card.select_one(
-                "a[data-automation='jobTitle'], "
-                "h1[data-automation='job-detail-title'], "
-                "a[data-testid='job-title'], "
-                "h3 a, h2 a, a[href*='/job/']"
-            )
-            # Company name
-            company_el = card.select_one(
-                "span[data-automation='jobCompany'], "
-                "a[data-automation='jobCompany'], "
-                "[data-testid='company-name'], "
-                "span[class*='company'], "
-                "a[class*='company']"
+            page.wait_for_selector(
+                "article[data-testid='job-card'], article[data-automation='job-card']",
+                timeout=30000,
             )
-            # Location
-            location_el = card.select_one(
-                "span[data-automation='jobLocation'], "
-                "span[class*='location'], "
-                "[data-testid='job-location']"
-            )
-
-            if not title_el:
-                continue
+        except Exception:
+            log.warning("[JobStreet] Kartu lowongan tidak muncul untuk keyword: %s", keyword)
 
-            title_text = title_el.text.strip()
-            company_text = company_el.text.strip() if company_el else ""
-            location_text = location_el.text.strip() if location_el else ""
+        soup = BeautifulSoup(page.content(), "html.parser")
 
-            href = title_el.get("href", "")
-            if not href:
-                continue
+        cards = soup.select("article[data-testid='job-card'], article[data-automation='job-card'], article")
 
-            if href.startswith("http"):
-                job_url = href.split("?")[0]
-            else:
-                job_url = urljoin(base_url, href.split("?")[0])
+        for card in cards:
+            if len(jobs) >= max_jobs:
+                break
 
-            # Visit detail page for description
-            description = ""
-            skills = []
             try:
-                detail_page = context.new_page()
-                detail_page.goto(job_url, wait_until="domcontentloaded", timeout=15000)
-                human_delay()
-                detail_soup = BeautifulSoup(detail_page.content(), "html.parser")
-
-                # JobStreet description sections
-                desc_el = detail_soup.select_one(
-                    "div[data-automation='jobAdDetails'], "
-                    "div[class*='description'], "
-                    "section[class*='description'], "
-                    "article"
+                title_el = card.select_one(
+                    "a[data-automation='jobTitle'], "
+                    "h1[data-automation='job-detail-title'], "
+                    "a[data-testid='job-title'], "
+                    "h3 a, h2 a, a[href*='/job/']"
+                )
+                company_el = card.select_one(
+                    "span[data-automation='jobCompany'], "
+                    "a[data-automation='jobCompany'], "
+                    "[data-testid='company-name'], "
+                    "span[class*='company'], "
+                    "a[class*='company']"
+                )
+                location_el = card.select_one(
+                    "span[data-automation='jobLocation'], "
+                    "span[class*='location'], "
+                    "[data-testid='job-location']"
                 )
-                if desc_el:
-                    description = desc_el.text.strip()
 
-                # Extract location from detail if not found in listing
-                if not location_text:
-                    loc_el = detail_soup.select_one(
-                        "span[data-automation='jobLocation'], "
-                        "span[class*='location']"
+                if not title_el:
+                    continue
+
+                title_text = title_el.text.strip()
+                company_text = company_el.text.strip() if company_el else ""
+                location_text = location_el.text.strip() if location_el else ""
+
+                href = title_el.get("href", "")
+                if not href:
+                    continue
+
+                if href.startswith("http"):
+                    job_url = href.split("?")[0]
+                else:
+                    job_url = urljoin(base_url, href.split("?")[0])
+
+                description = ""
+                salary = None
+                job_type = "Full-time"
+                contact_info = None
+
+                try:
+                    detail_page = context.new_page()
+                    detail_page.goto(job_url, wait_until="domcontentloaded", timeout=15000)
+                    short_delay()
+                    detail_soup = BeautifulSoup(detail_page.content(), "html.parser")
+
+                    desc_el = detail_soup.select_one(
+                        "div[data-automation='jobAdDetails'], "
+                        "div[class*='description'], "
+                        "section[class*='description'], "
+                        "article"
                     )
-                    if loc_el:
-                        location_text = loc_el.text.strip()
-
-                detail_page.close()
-            except Exception:
-                pass
-
-            if not location_text:
-                location_text = "Indonesia"
-
-            skills = extract_skills(description)
-
-            jobs.append({
-                "title": title_text,
-                "company": company_text,
-                "location": location_text,
-                "url": job_url,
-                "description": description,
-                "skills": skills,
-            })
-        except Exception:
-            continue
+                    if desc_el:
+                        description = desc_el.text.strip()
 
-    browser.close()
-    p.stop()
+                    salary_el = detail_soup.select_one(
+                        "span[data-automation='jobSalary'], "
+                        "span[class*='salary']"
+                    )
+                    if salary_el:
+                        salary = salary_el.text.strip() or None
+                    if not salary:
+                        page_text = detail_soup.get_text(" ", strip=True)
+                        salary = extract_salary(description or page_text)
+
+                    page_text = detail_soup.get_text(" ", strip=True)
+                    job_type = extract_job_type(page_text)
+                    contact_info = extract_contact(description or page_text)
+
+                    if not location_text:
+                        loc_el = detail_soup.select_one(
+                            "span[data-automation='jobLocation'], "
+                            "span[class*='location']"
+                        )
+                        if loc_el:
+                            location_text = loc_el.text.strip()
+
+                    detail_page.close()
+                except Exception as e:
+                    log.debug("[JobStreet] Gagal ambil detail %s: %s", job_url, e)
+
+                if not location_text:
+                    location_text = "Indonesia"
+
+                jobs.append({
+                    "title": title_text,
+                    "company": company_text,
+                    "location": location_text,
+                    "type": job_type,
+                    "salary": salary,
+                    "url": job_url,
+                    "description": description,
+                    "skills": extract_skills(description),
+                    "contact_info": contact_info,
+                })
+            except Exception as e:
+                log.debug("[JobStreet] Gagal proses kartu: %s", e)
+                continue
+    finally:
+        try:
+            browser.close()
+        except Exception:
+            pass
+        p.stop()
 
+    log.info("[JobStreet] Keyword '%s' menghasilkan %d lowongan", keyword, len(jobs))
     return jobs

linkedin.py

@@ -1,110 +1,135 @@
+import logging
+
 from bs4 import BeautifulSoup
-from browser import get_browser
-from base import human_delay, jitter_scroll
 from urllib.parse import urljoin
+
+from base import goto_with_retry, human_delay, jitter_scroll, short_delay
+from browser import get_browser
+from extract import extract_contact, extract_job_type, extract_salary
 from skills_keyword import extract_skills
 
+log = logging.getLogger(__name__)
 
-def scrape_linkedin(keyword="software engineer"):
-    p, browser, context = get_browser()
-    page = context.new_page()
 
-    base_url = "https://www.linkedin.com"
-    url = f"{base_url}/jobs/search/?keywords={keyword}&location=Indonesia"
+def scrape_linkedin(keyword="software engineer", max_jobs=25):
+    p, browser, context = get_browser()
+    jobs = []
 
     try:
-        page.goto(url, wait_until="networkidle", timeout=30000)
-    except Exception:
-        page.goto(url, wait_until="domcontentloaded", timeout=30000)
-
-    human_delay()
-    jitter_scroll(page)
-    human_delay()
+        page = context.new_page()
 
-    soup = BeautifulSoup(page.content(), "html.parser")
+        base_url = "https://www.linkedin.com"
+        url = f"{base_url}/jobs/search/?keywords={keyword}&location=Indonesia"
 
-    jobs = []
+        if not goto_with_retry(page, url):
+            log.warning("[LinkedIn] Gagal memuat halaman pencarian: %s", url)
+            return jobs
 
-    # LinkedIn public job cards use these class patterns
-    cards = soup.select("div.base-card")
+        human_delay()
+        jitter_scroll(page)
+        human_delay()
 
-    for card in cards:
         try:
-            title_el = card.select_one(
-                "h3.base-search-card__title, "
-                "span.screen-reader-text, "
-                "h3"
-            )
-            company_el = card.select_one(
-                "h4.base-search-card__subtitle, "
-                "a.hidden-nested-link, "
-                "h4"
-            )
-            location_el = card.select_one(
-                "span.job-search-card__location, "
-                "span[class*='location']"
-            )
-            link_el = card.select_one("a.base-card__full-link, a[href*='/jobs/view/']")
-
-            if not (title_el and company_el and link_el):
-                continue
+            page.wait_for_selector("div.base-card", timeout=30000)
+        except Exception:
+            log.warning("[LinkedIn] Kartu lowongan tidak muncul untuk keyword: %s", keyword)
+
+        soup = BeautifulSoup(page.content(), "html.parser")
 
-            href = link_el.get("href", "")
-            if href.startswith("http"):
-                job_url = href.split("?")[0]
-            else:
-                job_url = urljoin(base_url, href.split("?")[0])
+        cards = soup.select("div.base-card")
 
-            location_text = location_el.text.strip() if location_el else ""
+        for card in cards:
+            if len(jobs) >= max_jobs:
+                break
 
-            # Visit detail page for description
-            description = ""
-            skills = []
             try:
-                detail_page = context.new_page()
-                detail_page.goto(job_url, wait_until="domcontentloaded", timeout=15000)
-                human_delay()
-                detail_soup = BeautifulSoup(detail_page.content(), "html.parser")
-
-                # LinkedIn description sections
-                desc_el = detail_soup.select_one(
-                    "div.description__text, "
-                    "div[class*='description'], "
-                    "section[class*='description']"
+                title_el = card.select_one(
+                    "h3.base-search-card__title, "
+                    "span.screen-reader-text, "
+                    "h3"
                 )
-                if desc_el:
-                    description = desc_el.text.strip()
+                company_el = card.select_one(
+                    "h4.base-search-card__subtitle, "
+                    "a.hidden-nested-link, "
+                    "h4"
+                )
+                location_el = card.select_one(
+                    "span.job-search-card__location, "
+                    "span[class*='location']"
+                )
+                link_el = card.select_one("a.base-card__full-link, a[href*='/jobs/view/']")
+
+                if not (title_el and company_el and link_el):
+                    continue
+
+                href = link_el.get("href", "")
+                if href.startswith("http"):
+                    job_url = href.split("?")[0]
+                else:
+                    job_url = urljoin(base_url, href.split("?")[0])
+
+                location_text = location_el.text.strip() if location_el else ""
+
+                description = ""
+                salary = None
+                job_type = "Full-time"
+                contact_info = None
+
+                try:
+                    detail_page = context.new_page()
+                    detail_page.goto(job_url, wait_until="domcontentloaded", timeout=15000)
+                    short_delay()
+                    detail_soup = BeautifulSoup(detail_page.content(), "html.parser")
+
+                    desc_el = detail_soup.select_one(
+                        "div.description__text, "
+                        "div[class*='description'], "
+                        "section[class*='description']"
+                    )
+                    if desc_el:
+                        description = desc_el.text.strip()
+
+                    page_text = detail_soup.get_text(" ", strip=True)
+
+                    salary = extract_salary(description or page_text)
+                    job_type = extract_job_type(page_text)
+                    contact_info = extract_contact(description or page_text)
+
+                    if not location_text:
+                        loc_el = detail_soup.select_one(
+                            "div[class*='location'], "
+                            "span[class*='location']"
+                        )
+                        if loc_el:
+                            location_text = loc_el.text.strip()
+
+                    detail_page.close()
+                except Exception as e:
+                    log.debug("[LinkedIn] Gagal ambil detail %s: %s", job_url, e)
 
-                # Extract location from detail if not found in listing
                 if not location_text:
-                    loc_el = detail_soup.select_one(
-                        "div[class*='location'], "
-                        "span[class*='location']"
-                    )
-                    if loc_el:
-                        location_text = loc_el.text.strip()
-
-                detail_page.close()
-            except Exception:
-                pass
-
-            if not location_text:
-                location_text = "Indonesia"
-
-            skills = extract_skills(description)
-
-            jobs.append({
-                "title": title_el.text.strip(),
-                "company": company_el.text.strip(),
-                "location": location_text,
-                "url": job_url,
-                "description": description,
-                "skills": skills,
-            })
+                    location_text = "Indonesia"
+
+                jobs.append({
+                    "title": title_el.text.strip(),
+                    "company": company_el.text.strip(),
+                    "location": location_text,
+                    "type": job_type,
+                    "salary": salary,
+                    "url": job_url,
+                    "description": description,
+                    "skills": extract_skills(description),
+                    "contact_info": contact_info,
+                })
+            except Exception as e:
+                log.debug("[LinkedIn] Gagal proses kartu: %s", e)
+                continue
+    finally:
+        try:
+            browser.close()
         except Exception:
-            continue
-
-    browser.close()
-    p.stop()
+            pass
+        p.stop()
 
+    log.info("[LinkedIn] Keyword '%s' menghasilkan %d lowongan", keyword, len(jobs))
     return jobs

normalize.py

@@ -4,8 +4,10 @@ def normalize(job, source):
         "company": job.get("company"),
         "location": job.get("location", "Indonesia"),
         "type": job.get("type", "Full-time"),
+        "salary": job.get("salary"),
         "description": job.get("description") or job.get("title", ""),
         "skills": job.get("skills", []),
+        "contact_info": job.get("contact_info"),
         "url": job.get("url"),
         "source": source,
         "is_active": True

readme.md

@@ -5,15 +5,17 @@ Proyek ini adalah scraper lowongan kerja yang mengumpulkan data dari LinkedIn da
 ## Fitur
 
 - Scraping lowongan dari **LinkedIn** dan **JobStreet**
+- Multi-keyword scraping (daftar di `run.py`)
 - Automasi browser menggunakan Playwright
 - Pembersihan dan normalisasi data sebelum disimpan
+- Ekstraksi otomatis skill, tipe kerja, gaji, dan kontak dari deskripsi
 - Penyimpanan data ke Supabase dengan mekanisme upsert
-- Penjadwalan scraping setiap 24 jam
+- Penjadwalan scraping setiap 24 jam, atau mode `--once` untuk cron/CI
 
 ## Setup
 
 1. Clone repository ini
-2. Buat virtual environment (disarankan Python 3.10+):
+2. Buat virtual environment (Python 3.10+):
 
     python -m venv venv
     venv\Scripts\activate
@@ -26,34 +28,79 @@ Proyek ini adalah scraper lowongan kerja yang mengumpulkan data dari LinkedIn da
 
     playwright install chromium
 
-5. Buat file `.env` di direktori proyek:
+5. Salin file `.env.example` menjadi `.env`, lalu isi kredensial Supabase:
+
+    cp .env.example .env
 
     SUPABASE_URL=your_supabase_url
     SUPABASE_ANON_KEY=your_supabase_anon_key
 
-6. Jalankan scraper:
+6. Jalankan scraper (sekali jalan lalu keluar, cocok untuk cron/CI):
+
+    python run.py --once
+
+   Atau jalankan sebagai service berjalan terus (scrape tiap 24 jam):
 
     python run.py
 
 ## Struktur Modul
 
-- `run.py` - Entry point, menjadwalkan scraping dan menyimpan data
+- `run.py` - Entry point, multi-keyword, menjadwalkan scraping dan menyimpan data
 - `linkedin.py` - Scraping lowongan dari LinkedIn
 - `jobstreet.py` - Scraping lowongan dari JobStreet
 - `browser.py` - Konfigurasi Playwright (Chromium headless)
-- `base.py` - Helper delay dan scroll untuk simulasi manusia
+- `base.py` - Helper delay, scroll, dan retry navigasi
+- `extract.py` - Ekstraksi gaji, tipe kerja, dan kontak dari deskripsi
 - `normalize.py` - Normalisasi format data sebelum disimpan
 - `cleaner.py` - Validasi dan pembersihan data yang tidak valid
 - `supabase_client.py` - Koneksi ke Supabase
+- `skills_keyword.py` - Daftar keyword untuk ekstraksi skill
+- `schema.sql` - Skema tabel `jobs` di Supabase
 
 ## Database
 
-Pastikan tabel `lowongan_kerja` sudah ada di Supabase dengan kolom:
-- `nama_perusahaan`
-- `posisi_jabatan`
-- `deskripsi_kualifikasi`
-- `link_pendaftaran`
-- `sumber_data`
+Jalankan `schema.sql` di Supabase untuk membuat tabel `jobs`, atau pastikan sudah ada dengan skema berikut:
+
+- `title`
+- `company`
+- `location`
+- `type` (Full-time / Part-time / Contract / Internship)
+- `salary`
+- `description`
+- `skills` (jsonb)
+- `contact_info`
+- `url`
+- `source`
 - `is_active`
 
-Kolom `link_pendaftaran` digunakan sebagai konflik untuk upsert.
+Kolom `url` bersifat unik dan digunakan sebagai konflik untuk upsert.
+
+## Deployment
+
+### GitHub Actions (gratis, untuk demo skripsi)
+
+Jalankan scraping setiap 24 jam lewat cron workflow, contoh `.github/workflows/scrape.yml`:
+
+```yaml
+name: scrape
+on:
+  schedule:
+    - cron: "0 3 * * *"
+  workflow_dispatch:
+jobs:
+  run:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+      - run: pip install -r requirements.txt
+      - run: python -m playwright install --with-deps chromium
+      - run: python run.py --once
+        env:
+          SUPABASE_URL: ${{ secrets.SUPABASE_URL }}
+          SUPABASE_ANON_KEY: ${{ secrets.SUPABASE_ANON_KEY }}
+```
+
+Atur `SUPABASE_URL` dan `SUPABASE_ANON_KEY` di *Settings → Secrets and variables → Actions*. Alternatif lain: Railway/Render (cron bawaan) atau VPS murah.

requirements.txt

@@ -1,5 +1,5 @@
 beautifulsoup4==4.12.3
-playwright==1.40.0
+playwright==1.62.0
 python-dotenv==1.0.1
 schedule==1.2.2
-supabase==2.3.4
+supabase==2.31.0

run.py

@@ -1,12 +1,33 @@
-from linkedin import scrape_linkedin
+import argparse
+import logging
+import signal
+import sys
+import time
+
+import schedule
+
+from cleaner import clean_job
 from jobstreet import scrape_jobstreet
+from linkedin import scrape_linkedin
 from normalize import normalize
-from cleaner import clean_job
 from supabase_client import supabase
-import time
-import schedule
-import signal
-import sys
+
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s [%(levelname)s] %(message)s",
+)
+log = logging.getLogger("scraper")
+
+KEYWORDS = [
+    "software engineer",
+    "backend developer",
+    "frontend developer",
+    "data scientist",
+    "full stack developer",
+    "devops engineer",
+    "ui ux designer",
+    "mobile developer",
+]
 
 
 def push(jobs, source):
@@ -33,23 +54,27 @@ def push(jobs, source):
             .upsert(batch, on_conflict="url") \
             .execute()
     except Exception as e:
-        print(f"[ERROR] Supabase upsert gagal ({source}): {e}")
+        log.error("[Supabase] Upsert gagal (%s): %s", source, e)
         return 0
 
     return len(batch)
 
 
-def main():
+def scrape_source(scraper, source, keyword):
     try:
-        print("Scraping LinkedIn...")
-        linkedin_jobs = scrape_linkedin()
-        print(f"LinkedIn ditemukan: {len(linkedin_jobs)}, dimasukkan: {push(linkedin_jobs, 'LinkedIn')}")
-
-        print("Scraping JobStreet...")
-        jobstreet_jobs = scrape_jobstreet()
-        print(f"JobStreet ditemukan: {len(jobstreet_jobs)}, dimasukkan: {push(jobstreet_jobs, 'JobStreet')}")
+        jobs = scraper(keyword)
     except Exception as e:
-        print(f"[ERROR] Scraping gagal: {e}")
+        log.error("[%s] Scraping gagal (keyword=%s): %s", source, keyword, e)
+        return
+
+    inserted = push(jobs, source)
+    log.info("[%s] %s: %d ditemukan, %d dimasukkan", source, keyword, len(jobs), inserted)
+
+
+def main():
+    for keyword in KEYWORDS:
+        scrape_source(scrape_linkedin, "LinkedIn", keyword)
+        scrape_source(scrape_jobstreet, "JobStreet", keyword)
 
 
 running = True
@@ -57,23 +82,31 @@ running = True
 
 def shutdown(signum, frame):
     global running
-    print("\n[INFO] Ctrl+C ditekan. Menghentikan scraper...")
+    log.info("Ctrl+C ditekan. Menghentikan scraper...")
     running = False
 
 
 if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="Job scraper LinkedIn & JobStreet")
+    parser.add_argument("--once", action="store_true",
+                        help="Jalankan sekali lalu keluar (untuk cron/CI)")
+    args = parser.parse_args()
+
     signal.signal(signal.SIGINT, shutdown)
     signal.signal(signal.SIGTERM, shutdown)
 
     main()  # jalankan langsung saat pertama kali start
 
+    if args.once:
+        sys.exit(0)
+
     schedule.every(24).hours.do(main)
-    print("Job scraper berjalan setiap 24 jam...")
-    print("Tekan Ctrl+C untuk menghentikan.")
+    log.info("Job scraper berjalan setiap 24 jam...")
+    log.info("Tekan Ctrl+C untuk menghentikan.")
 
     while running:
         schedule.run_pending()
         time.sleep(60)
 
-    print("[INFO] Scraper dihentikan. Sampai jumpa!")
-    sys.exit(0)
+    log.info("Scraper dihentikan. Sampai jumpa!")
+    sys.exit(0)

skills_keyword.py

@@ -3,13 +3,13 @@ import re
 
 SKILLS_DB = [
     # Programming Languages
-    "python", "javascript", "typescript", "java", "c++", "c#", "go", "golang",
-    "rust", "kotlin", "swift", "php", "ruby", "scala", "r", "dart",
+    "python", "javascript", "typescript", "java", "c++", "c#", "golang",
+    "rust", "kotlin", "swift", "php", "ruby", "scala", "dart",
     "shell", "bash", "powershell",
 
     # Web Frontend
     "react", "react native", "vue", "vue.js", "angular", "next.js", "nextjs",
-    "nuxt", "svelte", "html", "css", "scss", "sass", "less", "tailwind",
+    "nuxt", "svelte", "html", "css", "scss", "sass", "tailwind",
     "bootstrap", "jquery",
 
     # Web Backend
@@ -36,7 +36,7 @@ SKILLS_DB = [
     "natural language processing", "nlp", "computer vision",
 
     # Mobile
-    "flutter", "react native", "ios", "android", "swift", "kotlin",
+    "flutter", "ios", "android",
 
     # Testing
     "selenium", "jest", "pytest", "mocha", "cypress", "unit testing",
@@ -51,6 +51,35 @@ SKILLS_DB = [
 ]
 
 
+# Kata ambigu yang hanya dianggap skill jika muncul dengan konteks teknis.
+# Pola kecocokan disimpan lowercase.
+AMBIGUOUS = {
+    "go": [
+        r"\bgo\s+(language|lang|development|developer|programming)\b",
+        r"\bwritten?\s+in\s+go\b",
+        r"\bgo\s+golang\b",
+    ],
+    "r": [
+        r"\br\s+(language|programming|studio)\b",
+    ],
+    "less": [
+        r"\bless\s+(css|preprocessor)\b",
+        r"\bscss\s*/\s*less\b",
+        r"\bsass\s*/\s*less\b",
+    ],
+}
+
+
+def _match_ambiguous(text):
+    found = []
+    for skill, patterns in AMBIGUOUS.items():
+        for pattern in patterns:
+            if re.search(pattern, text, re.IGNORECASE):
+                found.append(skill)
+                break
+    return found
+
+
 def extract_skills(text):
     """Extract skills from job description text using keyword matching."""
     if not text:
@@ -64,4 +93,6 @@ def extract_skills(text):
         if re.search(pattern, text_lower):
             found.append(skill)
 
+    found += _match_ambiguous(text_lower)
+
     return list(set(found))