Update app.py

app.py

@@ -27,10 +27,12 @@ except Exception:
 # CONFIG / FEATURE GATES
 # ======================================================
 
-ENABLE_SEMANTIC = False
+ENABLE_SEMANTIC = False          # user opt-in only
+ENABLE_PDF_EXPORT = True        # LIVE results only
 ENABLE_PDF_THUMBNAILS = True
 ENABLE_ENTITY_GRAPHS = True
 ENABLE_TIMELINES = True
+ENABLE_JOURNALIST_ZIP = True
 
 # ======================================================
 # BASE ADAPTER
@@ -45,6 +47,7 @@ class FOIAAdapter:
         start = time.time()
         url = self.search_url.format(q=quote_plus(query))
         latency = round(time.time() - start, 3)
+
         return [{
             "agency": self.agency,
             "title": f"{self.agency} FOIA Search Results",
@@ -86,17 +89,27 @@ class NSA(FOIAAdapter):
     agency = "NSA"
     search_url = "https://www.nsa.gov/resources/everyone/foia/reading-room/?q={q}"
 
-LIVE_ADAPTERS = [CIA(), FBI(), DOJ(), DHS(), STATE(), GSA(), NSA()]
+LIVE_ADAPTERS = [
+    CIA(),
+    FBI(),
+    DOJ(),
+    DHS(),
+    STATE(),
+    GSA(),
+    NSA()
+]
 
 # ======================================================
-# STUB ADAPTERS (NON-EXPORTABLE)
+# STUB ADAPTERS (CLEARLY LABELED)
 # ======================================================
 
 class StubAdapter(FOIAAdapter):
     is_live = False
+
     def __init__(self, agency):
         self.agency = agency
         self.search_url = ""
+
     def search(self, query):
         return [{
             "agency": self.agency,
@@ -131,6 +144,9 @@ def bluebook_full(r):
         f"{r['url']} (retrieved {datetime.utcnow().strftime('%b %d, %Y')})."
     )
 
+def bluebook_short(r):
+    return f"{r['agency']}, FOIA Reading Room, {r['url']}."
+
 # ======================================================
 # GLOBAL STATE
 # ======================================================
@@ -154,6 +170,7 @@ def run_search(query, include_stubs, semantic_mode):
             coverage[r["agency"]] += 1
             if r["is_live"]:
                 LAST_LIVE_RECORDS.append(r)
+
             rows.append([
                 r["agency"],
                 "LIVE" if r["is_live"] else "STUB",
@@ -169,119 +186,120 @@ def run_search(query, include_stubs, semantic_mode):
         if coverage.get(agency, 0) == 0:
             gap_md += f"- ❌ **{agency}**: no public results found\n"
 
-    return rows, gap_md, build_pdf_thumbnail_gallery()
+    return rows, gap_md
 
 # ======================================================
-# PDF THUMBNAIL GALLERY
+# SEMANTIC STATUS
 # ======================================================
 
-def build_pdf_thumbnail_gallery():
-    cards = []
-    for r in LAST_LIVE_RECORDS:
-        url = r["url"]
-        if not url.lower().endswith(".pdf"):
-            continue
-        cards.append(f"""
-        <div style="border:1px solid #ddd;border-radius:8px;padding:12px;margin-bottom:16px;">
-            <b>{r['agency']} — {r['title']}</b><br><br>
-            <iframe src="{url}" width="100%" height="220"></iframe>
-            <div style="margin-top:8px;">
-                <a href="{url}" target="_blank">View</a> |
-                <a href="{url}" download>Download</a> |
-                <a href="#" onclick="shareDoc('{url}')">Share</a> |
-                <a href="#" onclick="askAI('{r['agency']}', '{r['title']}', '{url}')">Ask AI</a>
-            </div>
-        </div>
-        """)
-    return "".join(cards) if cards else "<i>No PDF documents found.</i>"
+def semantic_status(enabled):
+    if enabled and not FAISS_AVAILABLE:
+        return "⚠ Semantic mode unavailable (optional dependencies missing)"
+    if enabled:
+        return "🧠 Semantic mode enabled (metadata only)"
+    return "Semantic mode off"
 
 # ======================================================
-# JOURNALIST ZIP EXPORT (LINKS + CITATIONS ONLY)
+# FOIA REQUEST GENERATOR (PDF)
 # ======================================================
 
-def generate_journalist_zip():
-    if not LAST_LIVE_RECORDS:
-        return None
-
-    mem = io.BytesIO()
-    with zipfile.ZipFile(mem, "w", zipfile.ZIP_DEFLATED) as z:
-        z.writestr(
-            "README.txt",
-            "Public FOIA sources only.\n"
-            "This ZIP contains links and citations only.\n"
-            "No documents are hosted or redistributed.\n"
-        )
-
-        z.writestr(
-            "citations.txt",
-            "\n\n".join(bluebook_full(r) for r in LAST_LIVE_RECORDS)
-        )
-
-        csv = "agency,title,url,retrieved\n"
-        for r in LAST_LIVE_RECORDS:
-            csv += f"{r['agency']},{r['title']},{r['url']},{r['timestamp']}\n"
-        z.writestr("links.csv", csv)
+def generate_foia_request(requester, description):
+    buffer = io.BytesIO()
+    doc = SimpleDocTemplate(buffer)
+    styles = getSampleStyleSheet()
+    story = []
 
-        pdf_links = "\n".join(
-            r["url"] for r in LAST_LIVE_RECORDS if r["url"].lower().endswith(".pdf")
-        )
-        z.writestr("pdf_links.txt", pdf_links)
+    story.append(Paragraph("<b>Freedom of Information Act Request</b>", styles["Title"]))
+    story.append(Spacer(1, 12))
 
-    mem.seek(0)
-    return mem
+    story.append(Paragraph(f"<b>Requester:</b> {requester}", styles["Normal"]))
+    story.append(Spacer(1, 8))
+
+    story.append(Paragraph("<b>Description of Records Requested:</b>", styles["Normal"]))
+    story.append(Paragraph(description, styles["Normal"]))
+    story.append(Spacer(1, 12))
+
+    agencies = ", ".join(sorted({r["agency"] for r in LAST_LIVE_RECORDS}))
+    story.append(Paragraph(f"<b>Agencies Referenced:</b> {agencies}", styles["Normal"]))
+
+    doc.build(story)
+    buffer.seek(0)
+    return buffer
 
 # ======================================================
-# PUBLIC SHAREABLE RESULT PAGE (STATIC HTML)
+# ENTITY GRAPH + TIMELINE
 # ======================================================
 
-def generate_share_page():
-    if not LAST_LIVE_RECORDS:
-        return None
+def build_entity_graph():
+    domains = Counter(urlparse(r["url"]).netloc for r in LAST_LIVE_RECORDS if r["url"])
+    return go.Figure([go.Bar(x=list(domains.keys()), y=list(domains.values()))])
+
+def build_timeline():
+    dates = Counter(r["timestamp"][:10] for r in LAST_LIVE_RECORDS if r["timestamp"])
+    return go.Figure([go.Bar(x=list(dates.keys()), y=list(dates.values()))])
+
+# ======================================================
+# PDF PREVIEW + ACTION BUTTONS
+# ======================================================
 
-    html = """
-    <html><head><title>FOIA Search Results</title></head><body>
-    <h1>Federal FOIA Search Results</h1>
-    <p>Public electronic reading rooms only.</p><hr>
+def preview_selected(row):
+    if not row:
+        return "<i>Select a result</i>"
+
+    url = row[3]
+    if not url:
+        return "<i>No preview available (STUB)</i>"
+
+    buttons = f"""
+    <div style="margin-bottom:8px">
+      <a href="{url}" target="_blank">View</a> |
+      <a href="{url}" download>Download</a> |
+      <a href="{url}" target="_blank">Share</a> |
+      <i>Ask AI (link-out only)</i>
+    </div>
     """
 
-    for r in LAST_LIVE_RECORDS:
-        html += f"""
-        <h3>{r['agency']} — {r['title']}</h3>
-        <p><a href="{r['url']}" target="_blank">{r['url']}</a></p>
-        <p><b>Citation:</b> {bluebook_full(r)}</p>
-        <p><b>Hash:</b> {citation_hash(r)}</p>
-        <hr>
-        """
+    if url.lower().endswith(".pdf"):
+        return buttons + f"<iframe src='{url}' width='100%' height='520'></iframe>"
+
+    return buttons + f"<a href='{url}' target='_blank'>Open link</a>"
 
-    html += "</body></html>"
+# ======================================================
+# JOURNALIST ZIP EXPORT
+# ======================================================
+
+def journalist_zip():
+    buffer = io.BytesIO()
+    with zipfile.ZipFile(buffer, "w", zipfile.ZIP_DEFLATED) as z:
+        citations = []
+        links = []
+
+        for r in LAST_LIVE_RECORDS:
+            citations.append(bluebook_full(r))
+            links.append(f"{r['agency']},{r['title']},{r['url']},{r['timestamp']}")
 
-    buf = io.BytesIO(html.encode("utf-8"))
-    return buf
+        z.writestr("README.txt",
+                   "Public FOIA links only.\nNo documents are included.\n")
+        z.writestr("citations.txt", "\n".join(citations))
+        z.writestr("links.csv", "agency,title,url,timestamp\n" + "\n".join(links))
+        z.writestr("pdf_links.txt",
+                   "\n".join(r["url"] for r in LAST_LIVE_RECORDS if r["url"].lower().endswith(".pdf")))
+
+    buffer.seek(0)
+    return buffer
 
 # ======================================================
 # UI
 # ======================================================
 
-JS_HELPERS = """
-<script>
-function shareDoc(url) {
-    if (navigator.share) {
-        navigator.share({ title: "FOIA Document", url: url });
-    } else {
-        navigator.clipboard.writeText(url);
-        alert("Link copied to clipboard");
-    }
-}
-function askAI(a,t,u){
-    alert("AI analysis placeholder for: " + t);
-}
-</script>
-"""
-
 with gr.Blocks(title="Federal FOIA Intelligence Search") as app:
-    gr.HTML(JS_HELPERS)
+    gr.Markdown("""
+    # 🏛️ Federal FOIA Intelligence Search
+    **Public Electronic Reading Rooms Only**
 
-    gr.Markdown("# 🏛️ Federal FOIA Intelligence Search")
+    ✔ LIVE results are exportable  
+    ⚠ STUB results are informational only  
+    """)
 
     query = gr.Textbox(label="Search FOIA Libraries")
     include_stubs = gr.Checkbox(label="Include Extended Coverage (STUB)", value=False)
@@ -295,19 +313,34 @@ with gr.Blocks(title="Federal FOIA Intelligence Search") as app:
     )
 
     gap_panel = gr.Markdown()
-    pdf_gallery = gr.HTML()
+    preview_panel = gr.HTML()
+    semantic_status_md = gr.Markdown()
 
     search_btn.click(
         run_search,
         inputs=[query, include_stubs, semantic_toggle],
-        outputs=[results, gap_panel, pdf_gallery]
+        outputs=[results, gap_panel]
+    )
+
+    semantic_toggle.change(semantic_status, semantic_toggle, semantic_status_md)
+
+    results.select(lambda e: preview_selected(e.value), outputs=preview_panel)
+
+    gr.Markdown("## FOIA Request Generator")
+    requester = gr.Textbox(label="Your Name / Organization")
+    description = gr.Textbox(label="Describe the records requested", lines=4)
+    gr.Button("Generate FOIA Request PDF").click(
+        generate_foia_request,
+        inputs=[requester, description],
+        outputs=gr.File()
     )
 
-    gr.Markdown("## 📄 PDF Document Previews")
-    pdf_gallery.render()
+    gr.Markdown("## Analysis Tools")
+    gr.Button("Show Entity Graph").click(build_entity_graph, outputs=gr.Plot())
+    gr.Button("Show Timeline").click(build_timeline, outputs=gr.Plot())
 
-    gr.Markdown("## 🗂 Journalist Tools")
-    gr.Button("Download Journalist ZIP").click(generate_journalist_zip, outputs=gr.File())
-    gr.Button("Generate Shareable Result Page").click(generate_share_page, outputs=gr.File())
+    if ENABLE_JOURNALIST_ZIP:
+        gr.Markdown("## Journalist Export")
+        gr.Button("Download Journalist ZIP").click(journalist_zip, outputs=gr.File())
 
 app.launch()