Merge branch 'main' of https://github.com/microsoft/intelligence-toolkit

README.md

@@ -4,6 +4,14 @@ pip install -r requirements.txt
 
 streamlit run app/Home.py
 
+## PDF export
+Install wkhtmltopdf to be able to generate the final story in PDF:
+
+Windows: [Download wkhtmltopdf installer](https://wkhtmltopdf.org/downloads.html)
+
+Linux:  `sudo apt-get install wkhtmltopdf`
+
+MacOS: `brew install homebrew/cask/wkhtmltopdf`
 
 # Project
 

app/util/download_pdf.py

@@ -0,0 +1,28 @@
+import tempfile
+import markdown2
+import pdfkit
+from util.wkhtmltopdf import config_pdfkit, pdfkit_options
+import streamlit as st
+
+css = '''body {
+    font-family: 'helvetica';
+}
+'''
+#itk-label
+text_label = 'Report generated using Intelligence Toolkit (https://aka.ms/itk)'
+
+def add_download_pdf(name, text, button_text='Download PDF', is_markdown=True, disabled=False):
+    if not name.endswith('.pdf'):
+        name += '.pdf'
+    text = f'<style>{css}</style>{text}<hr>{text_label}'
+    # Convert text to HTML if it's in Markdown format
+    text = markdown2.markdown(text) if is_markdown else text
+
+    # Generate PDF from HTML string
+    config_pdf = config_pdfkit()
+    with tempfile.NamedTemporaryFile(suffix='.pdf', delete=False) as temp_file:
+        pdfkit.from_string(text, temp_file.name, options=pdfkit_options, configuration=config_pdf)
+
+        # Provide download button for the generated PDF
+        with open(temp_file.name, 'rb') as f:
+            st.download_button(button_text, f, file_name=name, mime='application/pdf', disabled=disabled)

app/util/wkhtmltopdf.py

@@ -0,0 +1,27 @@
+import os
+import pdfkit
+
+# Specify the name of the executable
+executable = 'wkhtmltopdf'
+
+# Check if the executable is in the system PATH
+def is_in_path(executable):
+    for path in os.environ["PATH"].split(os.pathsep):
+        if os.path.exists(os.path.join(path.strip('"'), executable)):
+            return True
+    return False
+
+def config_pdfkit():
+    path_wkhtmltopdf = 'C:\\Program Files\\wkhtmltopdf\\bin\\wkhtmltopdf.exe'
+    # Verify if wkhtmltopdf is in PATH
+    if is_in_path(executable):
+        path_wkhtmltopdf=''
+    else:
+        path_wkhtmltopdf = 'C:\\Program Files\\wkhtmltopdf\\bin\\wkhtmltopdf.exe'
+
+    return pdfkit.configuration(wkhtmltopdf=path_wkhtmltopdf)
+
+pdfkit_options = {
+    'encoding': 'UTF-8',
+    'enable-local-file-access': True,
+}

app/workflows/attribute_patterns/workflow.py

@@ -10,13 +10,10 @@
     GridUpdateMode,
     ColumnsAutoSizeMode
 )   
-
+from util.download_pdf import add_download_pdf
 import workflows.attribute_patterns.functions as functions
 import workflows.attribute_patterns.classes as classes
-import workflows.attribute_patterns.config as config
-import workflows.attribute_patterns.prompts as prompts
 import workflows.attribute_patterns.variables as vars
-
 import workflows.attribute_patterns.config as config
 
 import util.AI_API
@@ -176,5 +173,9 @@ def create():
                         prefix=''
                     )
                     sv.attribute_report.value = result
-                report_placeholder.markdown(sv.attribute_report.value)
-                st.download_button('Download AI pattern report', data=sv.attribute_report.value, file_name='attribute_pattern_report.md', mime='text/markdown', disabled=sv.attribute_report.value == '')
+                report_data = sv.attribute_report.value
+                report_placeholder.markdown(report_data)
+                is_download_disabled = report_data == ''
+                name = 'attribute_pattern_report'
+                add_download_pdf(f'{name}.pdf', report_data, 'Download PDF report', disabled=is_download_disabled)
+                st.download_button('Download markdown report', data=report_data, file_name=f'{name}.md', mime='text/markdown', disabled=is_download_disabled)

app/workflows/group_narratives/workflow.py

@@ -1,6 +1,7 @@
 import streamlit as st
 import pandas as pd
 
+from util.download_pdf import add_download_pdf
 import workflows.group_narratives.config as config
 import workflows.group_narratives.variables as vars
 
@@ -200,4 +201,8 @@ def create():
                     )
                     sv.narrative_report.value = result
                 narrative_placeholder.markdown(sv.narrative_report.value)
-                st.download_button('Download AI group report', data=sv.narrative_report.value, file_name='narrative_report.md', mime='text/markdown', disabled=sv.narrative_report.value == '')
+                report_data = sv.narrative_report.value
+                is_download_disabled = report_data == ''
+                reports_name = 'narrative_report'
+                add_download_pdf(f'{reports_name}.pdf', report_data, button_text='Download PDF report', disabled=is_download_disabled)
+                st.download_button('Download markdown report', data=report_data, file_name=f'{reports_name}.md', mime='text/markdown', disabled=is_download_disabled)

app/workflows/question_answering/workflow.py

@@ -5,6 +5,7 @@
 import json
 import scipy.spatial.distance
 
+from util.download_pdf import add_download_pdf
 import workflows.question_answering.functions as functions
 import workflows.question_answering.classes as classes
 import workflows.question_answering.config as config
@@ -217,6 +218,10 @@ def create():
                     )
                     sv.answering_lazy_answer_text.value = result
                 report_placeholder.markdown(sv.answering_lazy_answer_text.value)
-
+                report_data = sv.answering_lazy_answer_text.value
+                is_download_disabled = report_data == ''
+                name = sv.answering_lazy_answer_text.value.split('\n')[0].replace('#','').strip().replace(' ', '_')
                 full_text = sv.answering_lazy_answer_text.value + '\n\n## Supporting FAQ\n\n' + re.sub(r' Q[\d]+: ', ' ', '\n\n'.join(sv.answering_matches.value.split('\n\n')[2:]), re.MULTILINE).replace('###### ', '### ')
-                st.download_button('Download AI answer report', data=full_text, file_name=sv.answering_lazy_answer_text.value.split('\n')[0].replace('#','').strip().replace(' ', '_')+'.md', mime='text/markdown', disabled=sv.answering_lazy_answer_text.value == '', key='lazy_download_button')      
+
+                add_download_pdf(f'{name}.pdf', full_text, 'Download PDF report', disabled=is_download_disabled)
+                st.download_button('Download markdown report', data=full_text, file_name=f'{name}.md', mime='text/markdown', disabled=sv.answering_lazy_answer_text.value == '', key='lazy_download_button')      

app/workflows/risk_networks/workflow.py

@@ -1,7 +1,6 @@
 import streamlit as st
 import pandas as pd
 import networkx as nx
-import numpy as np
 
 from collections import defaultdict
 from sklearn.neighbors import NearestNeighbors
@@ -15,9 +14,9 @@
     DataReturnMode,
     GridOptionsBuilder,
     GridUpdateMode,
-    ColumnsAutoSizeMode
 )   
 
+from util.download_pdf import add_download_pdf
 import workflows.risk_networks.functions as functions
 import workflows.risk_networks.classes as classes
 import workflows.risk_networks.config as config
@@ -508,4 +507,10 @@ def create():
                     )
                     sv.network_report.value = result
                 report_placeholder.markdown(sv.network_report.value)
-                st.download_button('Download AI network report', data=sv.network_report.value, file_name='network_report.md', mime='text/markdown', disabled=sv.network_report.value == '')
+
+                report_data = sv.network_report.value
+                is_download_disabled = report_data == ''
+                name = 'network_report'
+
+                add_download_pdf(f'{name}.pdf', report_data, 'Download PDF report', disabled=is_download_disabled)
+                st.download_button('Download AI network report', data=report_data, file_name=f'{name}.md', mime='text/markdown', disabled=is_download_disabled)