Update pages/01_Insights.py

pages/01_Insights.py

@@ -1,92 +1,92 @@
-import streamlit as st
-import tempfile
-import fitz  # PyMuPDF
-from langchain_openai import ChatOpenAI
-from langchain.prompts import PromptTemplate
-import os
-
-# Set your OpenAI API key here
-os.environ["OPENAI_API_KEY"] = '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'
-
-# Function to generate insights based on context
-def generate_insights(resume_text, resume_links):
-    prompt_template = f"""
-    You are an experienced Technical Recruiter specializing in AI roles. 
-    Review the candidate's resume and extract key insights relevant to an AI engineering position.
-
-    Resume Details:
-    - Parsed Resume Text Content: {resume_text}
-    - Hyperlinked Information: {resume_links}
-
-    Please extract and organize the following information:
-    1. **Candidate Name**
-    2. **Educational Qualifications**
-    3. **Total Work Experience**
-    4. **Relevant AI Experience**
-    5. **Gaps in Employment**
-    6. **Publications**
-    7. **Professional Links**
-    8. **Other Noteworthy Details**
-
-    Output the insights in a clear, summarized format.
-    """
-    prompt = PromptTemplate.from_template(prompt_template)
-    chat_model = ChatOpenAI(temperature=0, model="gpt-3.5-turbo")
-    response = chat_model.generate([prompt.format(resume_text=resume_text, resume_links=resume_links)])
-    insights = response.generations[0][0].text.strip()
-    return insights
-
-# Function to parse PDF with text and hyperlinks
-def parse_pdf_with_links(uploaded_file):
-    with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as temp_file:
-        temp_file.write(uploaded_file.read())
-        temp_file_path = temp_file.name
-
-    text_content = []
-    links_with_display_names = []
-
-    with fitz.open(temp_file_path) as doc:
-        for page_num, page in enumerate(doc):
-            text_content.append(page.get_text("text"))
-            for link in page.get_links():
-                if link.get("uri"):
-                    rect = fitz.Rect(link["from"])
-                    display_name = page.get_text("text", clip=rect).strip() or "Unnamed Link"
-                    links_with_display_names.append({
-                        "page": page_num + 1,
-                        "display_name": display_name,
-                        "uri": link["uri"],
-                        "rect": rect
-                    })
-
-    full_text = "\n".join(text_content)
-    return full_text, links_with_display_names
-
-# Initialize session states for page control
-if "page" not in st.session_state:
-    st.session_state.page = "upload"  # Set default page to "upload"
-
-# Display upload page if "upload" page is active
-if st.session_state.page == "upload":
-    st.header("Resume Insights Generator")
-    resume_file = st.file_uploader("Upload Resume (PDF only)", type="pdf")
-    jd_file = st.file_uploader("Upload Job Description (PDF only)", type="pdf")
-
-    if resume_file:
-        resume_text, resume_links = parse_pdf_with_links(resume_file)
-        formatted_links = [f"Page {link['page']}: [{link['display_name']}]({link['uri']}) at {link['rect']}" for link in resume_links]
-        st.session_state.resume_text = resume_text
-        st.session_state.resume_links = "\n".join(formatted_links)
-
-    if jd_file:
-        jd_text, jd_links = parse_pdf_with_links(jd_file)
-        st.session_state.jd_text = jd_text
-
-    if st.button("Generate Insights from Resume"):
-        if "resume_text" in st.session_state and "resume_links" in st.session_state:
-            insights = generate_insights(st.session_state.resume_text, st.session_state.resume_links)
-            st.session_state.insights = insights
-            st.write("### Extracted Insights from the Resume")
-            st.write(insights)
-        else:
-            st.error("Please upload a resume before generating insights.")
+import streamlit as st
+import tempfile
+import fitz  # PyMuPDF
+from langchain_openai import ChatOpenAI
+from langchain.prompts import PromptTemplate
+import os
+
+# Set your OpenAI API key here
+os.environ["OPENAI_API_KEY"] = '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'
+
+# Function to generate insights based on context
+def generate_insights(resume_text, resume_links):
+    prompt_template = f"""
+    You are an experienced Technical Recruiter specializing in AI roles. 
+    Review the candidate's resume and extract key insights relevant to an AI engineering position.
+
+    Resume Details:
+    - Parsed Resume Text Content: {resume_text}
+    - Hyperlinked Information: {resume_links}
+
+    Please extract and organize the following information:
+    1. **Candidate Name**
+    2. **Educational Qualifications**
+    3. **Total Work Experience**
+    4. **Relevant AI Experience**
+    5. **Gaps in Employment**
+    6. **Publications**
+    7. **Professional Links**
+    8. **Other Noteworthy Details**
+
+    Output the insights in a clear, summarized format.
+    """
+    prompt = PromptTemplate.from_template(prompt_template)
+    chat_model = ChatOpenAI(temperature=0, model="gpt-3.5-turbo")
+    response = chat_model.generate([prompt.format(resume_text=resume_text, resume_links=resume_links)])
+    insights = response.generations[0][0].text.strip()
+    return insights
+
+# Function to parse PDF with text and hyperlinks
+def parse_pdf_with_links(uploaded_file):
+    with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as temp_file:
+        temp_file.write(uploaded_file.read())
+        temp_file_path = temp_file.name
+
+    text_content = []
+    links_with_display_names = []
+
+    with fitz.open(temp_file_path) as doc:
+        for page_num, page in enumerate(doc):
+            text_content.append(page.get_text("text"))
+            for link in page.get_links():
+                if link.get("uri"):
+                    rect = fitz.Rect(link["from"])
+                    display_name = page.get_text("text", clip=rect).strip() or "Unnamed Link"
+                    links_with_display_names.append({
+                        "page": page_num + 1,
+                        "display_name": display_name,
+                        "uri": link["uri"],
+                        "rect": rect
+                    })
+
+    full_text = "\n".join(text_content)
+    return full_text, links_with_display_names
+
+# Initialize session states for page control
+if "page" not in st.session_state:
+    st.session_state.page = "upload"  # Set default page to "upload"
+
+# Display upload page if "upload" page is active
+if st.session_state.page == "upload":
+    st.header("Resume Insights Generator")
+    resume_file = st.file_uploader("Upload Resume (PDF only)", type="pdf")
+    jd_file = st.file_uploader("Upload Job Description (PDF only)", type="pdf")
+
+    if resume_file:
+        resume_text, resume_links = parse_pdf_with_links(resume_file)
+        formatted_links = [f"Page {link['page']}: [{link['display_name']}]({link['uri']}) at {link['rect']}" for link in resume_links]
+        st.session_state.resume_text = resume_text
+        st.session_state.resume_links = "\n".join(formatted_links)
+
+    if jd_file:
+        jd_text, jd_links = parse_pdf_with_links(jd_file)
+        st.session_state.jd_text = jd_text
+
+    if st.button("Generate Insights from Resume"):
+        if "resume_text" in st.session_state and "resume_links" in st.session_state:
+            insights = generate_insights(st.session_state.resume_text, st.session_state.resume_links)
+            st.session_state.insights = insights
+            st.write("### Extracted Insights from the Resume")
+            st.write(insights)
+        else:
+            st.error("Please upload a resume before generating insights.")