Update app.py

app.py

@@ -3,7 +3,7 @@ import pandas as pd
 import json
 import smtplib
 from email.message import EmailMessage
-from typing import Dict
+from typing import Dict, List
 
 from jobspy import scrape_jobs
 import groq
@@ -23,6 +23,12 @@ def remove_duplicates(df: pd.DataFrame) -> pd.DataFrame:
     return df.drop_duplicates("__dedup__").drop(columns="__dedup__")
 
 
+def compute_keyword_score(text: str, keywords: List[str]) -> int:
+    """Simple keyword ranking: count of keyword hits (case-insensitive)."""
+    text_l = (text or "").lower()
+    return sum(text_l.count(k.lower()) for k in keywords if k.strip())
+
+
 # ======================================================
 # Optional Email Helper
 # ======================================================
@@ -49,12 +55,8 @@ def send_email_with_csv(recipient_email: str, df: pd.DataFrame):
     msg["Subject"] = "Your Job Search Results"
     msg["From"] = email_from
     msg["To"] = recipient_email
-
     msg.set_content(
-        "Hello,\n\n"
-        "Attached is the CSV file containing your job search results.\n\n"
-        "Regards,\n"
-        "Private Job Search Tool"
+        "Hello,\n\nAttached is the CSV file containing your job search results.\n\nRegards,\nPrivate Job Search Tool"
     )
 
     csv_data = df.to_csv(index=False)
@@ -67,7 +69,7 @@ def send_email_with_csv(recipient_email: str, df: pd.DataFrame):
 
 
 # ======================================================
-# AI helper
+# AI helper (intent extraction)
 # ======================================================
 
 def extract_search_parameters(client, prompt: str) -> Dict[str, str]:
@@ -98,77 +100,141 @@ def extract_search_parameters(client, prompt: str) -> Dict[str, str]:
 
 
 # ======================================================
-# Job scraping
+# Job scraping (Per-board toggles)
 # ======================================================
 
 @st.cache_data(ttl=3600)
-def get_indeed_jobs(search_term: str, location: str) -> pd.DataFrame:
+def get_indeed_jobs(
+    search_term: str,
+    location: str,
+    radius_km: int,
+    posted_within_days: int
+) -> pd.DataFrame:
     try:
         jobs = scrape_jobs(
             site_name=["indeed"],
             search_term=search_term,
             location=location,
             results_wanted=100,
-            hours_old=72,
-            country_indeed="Canada"
+            hours_old=posted_within_days * 24,
+            country_indeed="Canada",
+            radius=radius_km
         )
         return pd.DataFrame(jobs)
     except Exception:
         return pd.DataFrame()
 
 
+def get_other_board_stub(board_name: str) -> pd.DataFrame:
+    """Stub for future boards (toggle-safe)."""
+    return pd.DataFrame()
+
+
 # ======================================================
 # Streamlit App
 # ======================================================
 
 def main():
     st.set_page_config(page_title="Private Job Search", layout="centered")
-    st.title("📄 Private Job Search & Download")
+    st.title("📄 Private Job Search, Rank & Download")
 
+    # --- Inputs ---
     job_prompt = st.text_area(
         "Describe the job you are looking for",
-        placeholder="e.g. Civil Engineer, Planner, Water Resources Engineer in Alberta",
+        placeholder="e.g. Civil Engineer, Water Resources, Transportation in Alberta",
         height=120
     )
 
     api_key = st.text_input("Groq API Key", type="password")
 
-    send_email = st.checkbox("📧 Send results by email (optional)")
+    # --- Per-job-board toggles ---
+    st.subheader("Job Boards")
+    colb1, colb2, colb3 = st.columns(3)
+    with colb1:
+        use_indeed = st.checkbox("Indeed", value=True)
+    with colb2:
+        use_glassdoor = st.checkbox("Glassdoor (coming soon)", value=False, disabled=True)
+    with colb3:
+        use_linkedin = st.checkbox("LinkedIn (coming soon)", value=False, disabled=True)
+
+    # --- Filters ---
+    st.subheader("Filters")
+    posted_within_days = st.slider(
+        "Posted within last (days)",
+        min_value=1, max_value=30, value=7
+    )
 
-    email_address = None
-    if send_email:
-        email_address = st.text_input("Email address")
+    radius_km = st.slider(
+        "Search radius (km)",
+        min_value=5, max_value=100, value=25, step=5
+    )
 
-    if st.button(
-        "🔍 Search Jobs",
-        disabled=not job_prompt or not api_key
-    ):
+    # --- Keyword ranking ---
+    keywords_raw = st.text_input(
+        "Keyword ranking (comma-separated)",
+        placeholder="water, wastewater, stormwater, EPANET, WNTR"
+    )
+    keywords = [k.strip() for k in keywords_raw.split(",") if k.strip()]
+
+    # --- Optional email ---
+    send_email = st.checkbox("📧 Send results by email (optional)")
+    email_address = st.text_input("Email address") if send_email else None
+
+    # --- Action ---
+    if st.button("🔍 Search Jobs", disabled=not job_prompt or not api_key):
         client = groq.Client(api_key=api_key)
 
         with st.spinner("Understanding your request..."):
             params = extract_search_parameters(client, job_prompt)
 
-        with st.spinner("Searching jobs..."):
-            jobs_df = get_indeed_jobs(
-                params["search_term"],
-                params["location"]
-            )
-
-            if jobs_df.empty:
-                st.warning("No jobs found.")
-                return
+        all_jobs = []
 
-            jobs_df.fillna("", inplace=True)
-            jobs_df = remove_duplicates(jobs_df)
+        with st.spinner("Searching job boards..."):
+            if use_indeed:
+                indeed_df = get_indeed_jobs(
+                    params["search_term"],
+                    params["location"],
+                    radius_km,
+                    posted_within_days
+                )
+                if not indeed_df.empty:
+                    indeed_df["source"] = "Indeed"
+                    all_jobs.append(indeed_df)
+
+            # Future boards (toggle-safe)
+            if use_glassdoor:
+                all_jobs.append(get_other_board_stub("Glassdoor"))
+            if use_linkedin:
+                all_jobs.append(get_other_board_stub("LinkedIn"))
+
+        if not all_jobs:
+            st.warning("No jobs found.")
+            return
+
+        jobs_df = pd.concat(all_jobs, ignore_index=True)
+        jobs_df.fillna("", inplace=True)
+        jobs_df = remove_duplicates(jobs_df)
+
+        # --- Keyword ranking ---
+        if keywords:
+            jobs_df["keyword_score"] = jobs_df.apply(
+                lambda r: compute_keyword_score(
+                    f"{r.get('title','')} {r.get('description','')}",
+                    keywords
+                ),
+                axis=1
+            )
+        else:
+            jobs_df["keyword_score"] = 0
 
-        st.success(f"✅ Found {len(jobs_df)} unique jobs")
+        # Sort by keyword score (desc) then date if available
+        sort_cols = ["keyword_score"]
+        jobs_df = jobs_df.sort_values(sort_cols, ascending=[False])
 
-        # --------------------------------------------------
-        # DOWNLOAD OPTION (ALWAYS AVAILABLE)
-        # --------------------------------------------------
+        st.success(f"✅ Found {len(jobs_df)} jobs")
 
+        # --- Download always available ---
         csv_data = jobs_df.to_csv(index=False).encode("utf-8")
-
         st.download_button(
             label="⬇️ Download Jobs (CSV)",
             data=csv_data,
@@ -176,18 +242,12 @@ def main():
             mime="text/csv"
         )
 
-        # --------------------------------------------------
-        # OPTIONAL EMAIL
-        # --------------------------------------------------
-
+        # --- Optional email ---
         if send_email:
             if not email_address:
                 st.warning("Please enter an email address.")
             elif not email_secrets_available():
-                st.warning(
-                    "Email configuration not found. "
-                    "Download is still available."
-                )
+                st.warning("Email not configured. Download is still available.")
             else:
                 with st.spinner("Sending email..."):
                     try:
@@ -196,6 +256,16 @@ def main():
                     except Exception as e:
                         st.error(f"Failed to send email: {e}")
 
+        # --- Preview ---
+        st.subheader("Preview (Top Results)")
+        preview_cols = [
+            c for c in [
+                "source", "title", "company", "location",
+                "keyword_score", "date_posted", "job_url"
+            ] if c in jobs_df.columns
+        ]
+        st.dataframe(jobs_df[preview_cols].head(20), use_container_width=True)
+
 
 if __name__ == "__main__":
     main()