Improve URL input handling (#3)

- Improve URL input handling (e36a58f8866fd569e6bd851e40041ba5ba0293d2)


Co-authored-by: anmol pamday <[email protected]>

src/app.py

@@ -1,17 +1,12 @@
 import streamlit as st
-from compliance_lib import fetch_text, run_check, generate_report, RULES
-import textwrap, json, datetime, os
-
+from compliance_lib import *
+import PyPDF2  
 st.set_page_config(page_title="Anupalan Karta – Compliance Checker",
                    layout="wide")
-
-st.title("🛡️ Anupalan Karta (अनुपालंकर्ता)")
-
-st.subheader("Unified compliance self-check Tool")
-
-# --- introduction and how-to guide ------------------------------------------
+st.title('🛡️ Anupalan Karta (अनुपालंकर्ता)')
+st.subheader('Unified compliance self-check Tool')
 st.markdown("""
-## Introduction
+            ## Introduction
 
 **Anupalan Karta** (अनुपालंकर्ता) is a simple tool to help business owners, data architects, and data owners quickly check if their policies and procedures meet important regulations like GDPR, EU AI Act, and ISO 27001.
 
@@ -45,58 +40,101 @@ Visit [anktechsol.com](https://anktechsol.com) for professional compliance consu
 
 ---
 """)
-# --- sidebar ---------------------------------------------------------------
+
+# default state
+if "check_run" not in st.session_state:
+    st.session_state.check_run = False
+    st.session_state.raw_text = ""
+
+# ------ side bar ----
 with st.sidebar:
-    st.header("📑 Input options")
-    mode = st.radio("Choose data source:",
-                    ("Paste text", "URL of public policy", "Upload file"))
-    if mode == "Paste text":
-        raw_text = st.text_area("Paste your policy / procedures here")
-    elif mode == "URL of public policy":
-        url = st.text_input("Public URL (HTTPS)")
-        raw_text = fetch_text(url) if url else ""
-    else:
-        up = st.file_uploader("Upload .txt, .md or .pdf", type=["txt", "md", "pdf"])
-        raw_text = up.read().decode("utf-8", errors="ignore") if up else ""
-
-    st.markdown("---")
-    selected_fw = st.multiselect(
-        "Frameworks to check",
-        list(RULES.keys()),
-        default=list(RULES.keys())
-    )
-    run_btn = st.button("Run compliance check")
-
-# --- main body --------------------------------------------------------------
-if run_btn and raw_text.strip():
-    with st.spinner("Running rule-based checks…"):
-        results = run_check(raw_text)
-    st.subheader("📊 Checklist results")
-    for fw in selected_fw:
-        passed = sum(1 for _, ok in results[fw] if ok)
-        total = len(results[fw])
-        st.write(f"**{fw}: {passed}/{total} items passed**")
-        st.progress(passed / total)
-        for label, ok in results[fw]:
-            st.write(("✅" if ok else "❌") + "  " + label)
+        st.subheader('📥 Input options')
+        option = st.radio('Choose data source:',['Paste text','URL of public policy','Upload file'])
+        raw_text=""
+        if(option=='Paste text'):
+           raw_text= st.text_area('Paste your policy / procedures here')
+            
+        
+        elif option == 'URL of public policy':
+            Url = st.text_area("Public URL (HTTPS)", key="input_text")
+            raw_text = fetchText(Url) if Url else st.warning
+            if Url:
+                raw_text, error = fetchText(Url)
+                if error:
+                    st.warning(error)
+            else:
+                raw_text = ""
+            # print(raw_text)
+        else:
+            fileUpload = st.file_uploader("Upload .txt, .md or .pdf", ['txt', 'md', 'pdf'])
+            if fileUpload:
+                file_type = fileUpload.name.split('.')[-1].lower()
+        
+                if file_type in ["txt", "md"]:
+                    raw_text = fileUpload.read().decode("utf-8", errors="ignore")
+        
+                elif file_type=="pdf":
+                    try:
+                        pdf_reader = PyPDF2.PdfReader(fileUpload)
+                        for page in pdf_reader.pages:
+                            raw_text += page.extract_text() or ""
+                    except Exception as e:
+                        st.error(f"Error reading PDF: {e}")
+                    if raw_text and st.session_state.get("last_input") != raw_text:
+                       st.session_state.check_run = False
+                       st.session_state.last_input = raw_text
+          
+        # "Framework to Check", ["GDPR", "EU_AI_ACT", "ISO_27001"]
+        selectedFw=st.multiselect('Framework to Check',list(RULES.keys()),default=list(RULES.keys()))
         st.markdown("---")
 
-    # --- AI report section --------------------------------------------------
-    st.subheader("📝 Generate narrative report")
-    if st.button("Generate AI report"):
-        with st.spinner("Calling model… this may take ~30 s"):
-            bullet = "\n".join(
-                f"- {fw}: {sum(ok for _, ok in results[fw])}/{len(results[fw])} passed"
-                for fw in selected_fw
-            )
-            from config import AI_REPORT_PROMPT  # <-- import the prompt template
-            prompt = AI_REPORT_PROMPT.format(bullet=bullet)
-            report = generate_report(prompt)
-        st.markdown("#### Draft report")
-        st.code(report, language="markdown")
-        st.download_button("⬇️ Download .md",
-                           report.encode("utf-8"),
-                           file_name="anupalan_karta_report.md",
+
+
+  # ------- side bar end -----
+
+# <====== main code =====>
+# strip removes all the leading and trailing whitespace characters from a stringif run_btn and raw_text.strip():
+if "results" not in st.session_state:
+    st.session_state.results = None
+    st.session_state.selectedFw=[]
+
+if st.sidebar.button("Run Compliance Check") and raw_text.strip():
+#  reset_results()   
+ with st.spinner("Running rule-based checks…."):
+    st.session_state.results=run_check(raw_text, selectedFw)
+    st.session_state.selectedFw=selectedFw
+    st.session_state.check_run=True     
+if st.session_state.get("check_run"):
+    results = st.session_state.results
+    selectedFw = st.session_state.selectedFw
+        # 📊 Checklist results
+    st.subheader('📊 Checklist results')
+    for fw in selectedFw:
+            # Count how many rules passed for that framework
+            passed = sum(1 for _, ok in results[fw] if ok)  # _ is unused, ok is True/False
+            total = len(results[fw])
+             
+            st.write(f"**{fw}: {passed}/{total} items passed**")  # Example: EU_AI_ACT: 0/2 items passed
+            st.progress(passed / total)
+
+            for label, ok in results[fw]:
+                st.write(("✅" if ok else "❌") + "  " + label)  # Example: ❌ Lawful basis documented
+            st.markdown("---")
+
+        # 📝 Generate narrative report
+        
+    st.subheader('📝 Generate narrative report')
+    if st.button('Generate AI report'):
+            with st.spinner("Calling model… this may take ~30 s"):
+                bullet = "\n".join(
+                    f"- {fw}: {sum(ok for _, ok in results[fw])}/{len(results[fw])} passed"
+                    for fw in selectedFw
+                )
+                prompt=AI_REPORT_PROMPT.format(bullet=bullet)
+                report=generate_report(prompt)
+                st.markdown("#### Draft report")
+                st.code(report, language="markdown")
+                st.download_button("⬇️ Download .md",report.encode('utf-8'),file_name="anupalan_karta_report.md",
                            mime="text/markdown")
-else:
-    st.info("Awaiting input…")
+    else:
+           st.info("Awaiting input…")