Update app.py
Browse files
app.py
CHANGED
|
@@ -14,6 +14,14 @@ import json
|
|
| 14 |
import tempfile
|
| 15 |
import pytz # Import pytz for timezone handling
|
| 16 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
# --- Runtime / Environment Checks ---
|
| 18 |
print(f"Gradio version at runtime: {gr.__version__}", file=sys.stderr)
|
| 19 |
print(f"Python executable: {sys.executable}", file=sys.stderr)
|
|
@@ -821,33 +829,55 @@ def perform_ticket_analysis(file, user_query, selected_timezone, progress=gr.Pro
|
|
| 821 |
|
| 822 |
# --- Output Download Functions (using global_analysis_df_for_export) ---
|
| 823 |
|
| 824 |
-
# --- File Generation Libraries ---
|
| 825 |
-
from fpdf import FPDF
|
| 826 |
-
from docx import Document as DocxDocument
|
| 827 |
-
from docx.shared import Inches, Pt
|
| 828 |
-
from docx.enum.text import WD_ALIGN_PARAGRAPH, WD_PARAGRAPH_ALIGNMENT
|
| 829 |
|
| 830 |
def generate_pdf_report(selected_timezone):
|
| 831 |
df = global_analysis_df_for_export
|
| 832 |
if df.empty:
|
| 833 |
-
return gr.update(value=None, visible=False)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 834 |
|
| 835 |
-
pdf = FPDF()
|
| 836 |
pdf.add_page()
|
| 837 |
|
| 838 |
-
|
| 839 |
-
|
| 840 |
-
pdf.
|
|
|
|
|
|
|
|
|
|
|
|
|
| 841 |
|
| 842 |
now_utc = datetime.utcnow()
|
| 843 |
current_gen_time = format_time_for_display(now_utc, selected_timezone)
|
| 844 |
-
pdf.set_font(
|
| 845 |
-
pdf.cell(0,
|
| 846 |
-
pdf.ln(
|
| 847 |
|
| 848 |
-
pdf.set_font(
|
| 849 |
-
pdf.multi_cell(0,
|
| 850 |
-
pdf.set_font(
|
| 851 |
total_tickets_processed = df['Count'].sum() if 'Count' in df.columns else 0
|
| 852 |
total_categories_identified = len(df['Category'].unique()) if 'Category' in df.columns else 0
|
| 853 |
|
|
@@ -857,102 +887,98 @@ def generate_pdf_report(selected_timezone):
|
|
| 857 |
f"Total Tickets Summarized in Report: {total_tickets_processed}.\n"
|
| 858 |
f"Total Categories Identified: {total_categories_identified}."
|
| 859 |
)
|
| 860 |
-
pdf.multi_cell(0,
|
| 861 |
-
pdf.ln(
|
| 862 |
|
| 863 |
-
pdf.set_font(
|
| 864 |
-
pdf.multi_cell(0,
|
| 865 |
-
pdf.ln(
|
| 866 |
|
| 867 |
-
# Ensure required columns are present before iteration (handled in generate_summary_html_by_category)
|
| 868 |
-
# The dataframe should already be prepared from the analysis step.
|
| 869 |
-
|
| 870 |
for category, category_df in df.groupby('Category'):
|
| 871 |
total_count = category_df['Count'].sum()
|
| 872 |
|
| 873 |
pdf.set_fill_color(220, 237, 200) # Light Green
|
| 874 |
pdf.set_text_color(76, 175, 80) # Dark Green
|
| 875 |
-
pdf.set_font(
|
| 876 |
-
pdf.multi_cell(0,
|
| 877 |
-
pdf.ln(
|
| 878 |
|
| 879 |
pdf.set_text_color(0, 0, 0) # Black
|
| 880 |
-
pdf.set_font(
|
| 881 |
-
pdf.multi_cell(0,
|
| 882 |
-
pdf.set_font(
|
| 883 |
unique_issues = category_df['Issue Description (Summary)'].astype(str).drop_duplicates().apply(lambda x: x.replace('*', '')).tolist()
|
| 884 |
for issue in unique_issues:
|
| 885 |
-
pdf.multi_cell(0,
|
| 886 |
-
if not unique_issues: pdf.multi_cell(0,
|
| 887 |
-
pdf.ln(
|
| 888 |
|
| 889 |
-
pdf.set_font(
|
| 890 |
-
pdf.multi_cell(0,
|
| 891 |
-
pdf.set_font(
|
| 892 |
if 'Sub-Category' in category_df.columns:
|
| 893 |
sub_cat_counts = category_df.groupby('Sub-Category')['Count'].sum().sort_values(ascending=False)
|
| 894 |
for sub_cat, count in sub_cat_counts.items():
|
| 895 |
-
pdf.multi_cell(0,
|
| 896 |
-
if 'Sub-Category' not in category_df.columns or sub_cat_counts.empty: pdf.multi_cell(0,
|
| 897 |
-
pdf.ln(
|
| 898 |
|
| 899 |
-
pdf.set_font(
|
| 900 |
-
pdf.multi_cell(0,
|
| 901 |
-
pdf.set_font(
|
| 902 |
unique_root_causes = category_df['Root Cause (Inferred)'].astype(str).drop_duplicates().apply(lambda x: x.replace('*', '')).tolist()
|
| 903 |
for rc in unique_root_causes:
|
| 904 |
-
pdf.multi_cell(0,
|
| 905 |
-
if not unique_root_causes: pdf.multi_cell(0,
|
| 906 |
-
pdf.ln(
|
| 907 |
|
| 908 |
-
pdf.set_font(
|
| 909 |
-
pdf.multi_cell(0,
|
| 910 |
-
pdf.set_font(
|
| 911 |
unique_resolutions = category_df['Potential Resolution Steps'].astype(str).drop_duplicates().apply(lambda x: x.replace('*', '')).tolist()
|
| 912 |
for res in unique_resolutions:
|
| 913 |
-
pdf.multi_cell(0,
|
| 914 |
-
if not unique_resolutions: pdf.multi_cell(0,
|
| 915 |
-
pdf.ln(
|
| 916 |
|
| 917 |
-
pdf.set_font(
|
| 918 |
-
pdf.multi_cell(0,
|
| 919 |
-
pdf.set_font(
|
| 920 |
future_predictions = category_df['Future Prediction'].astype(str).drop_duplicates().apply(lambda x: x.replace('*', '')).tolist()
|
| 921 |
proactive_measures = category_df['Proactive Measures'].astype(str).drop_duplicates().apply(lambda x: x.replace('*', '')).tolist()
|
| 922 |
|
| 923 |
if future_predictions:
|
| 924 |
pdf.set_text_color(141, 110, 99) # Brownish
|
| 925 |
-
pdf.multi_cell(0,
|
| 926 |
if proactive_measures:
|
| 927 |
pdf.set_text_color(141, 110, 99) # Brownish
|
| 928 |
-
pdf.multi_cell(0,
|
| 929 |
if not future_predictions and not proactive_measures:
|
| 930 |
pdf.set_text_color(0, 0, 0) # Black
|
| 931 |
-
pdf.multi_cell(0,
|
| 932 |
-
pdf.ln(
|
| 933 |
|
| 934 |
pdf.set_text_color(0, 0, 0)
|
| 935 |
pdf.set_fill_color(241, 248, 233)
|
| 936 |
-
pdf.set_font(
|
| 937 |
pdf.multi_cell(0, 10, "Disclaimer", 0, 'L', 1)
|
| 938 |
-
pdf.ln(
|
| 939 |
-
pdf.set_font(
|
| 940 |
-
pdf.multi_cell(0,
|
| 941 |
-
pdf.ln(
|
| 942 |
-
pdf.multi_cell(0,
|
| 943 |
-
|
| 944 |
|
| 945 |
with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as tmp_file:
|
| 946 |
pdf.output(tmp_file.name)
|
| 947 |
temp_file_path = tmp_file.name
|
| 948 |
|
| 949 |
-
return gr.update(value=temp_file_path, visible=True, label="Download PDF Report")
|
| 950 |
|
| 951 |
|
| 952 |
def generate_docx_report(selected_timezone):
|
| 953 |
df = global_analysis_df_for_export
|
| 954 |
if df.empty:
|
| 955 |
-
return gr.update(value=None, visible=False)
|
| 956 |
|
| 957 |
doc = DocxDocument()
|
| 958 |
|
|
@@ -1045,7 +1071,7 @@ def generate_docx_report(selected_timezone):
|
|
| 1045 |
doc.save(tmp_file)
|
| 1046 |
temp_file_path = tmp_file.name
|
| 1047 |
|
| 1048 |
-
return gr.update(value=temp_file_path, visible=True, label="Download Word Report")
|
| 1049 |
|
| 1050 |
|
| 1051 |
# --- Gradio Interface Setup ---
|
|
|
|
| 14 |
import tempfile
|
| 15 |
import pytz # Import pytz for timezone handling
|
| 16 |
|
| 17 |
+
# --- Import fpdf2 instead of fpdf ---
|
| 18 |
+
from fpdf2 import FPDF
|
| 19 |
+
# --- End fpdf2 import ---
|
| 20 |
+
|
| 21 |
+
from docx import Document as DocxDocument
|
| 22 |
+
from docx.shared import Inches, Pt
|
| 23 |
+
from docx.enum.text import WD_ALIGN_PARAGRAPH, WD_PARAGRAPH_ALIGNMENT
|
| 24 |
+
|
| 25 |
# --- Runtime / Environment Checks ---
|
| 26 |
print(f"Gradio version at runtime: {gr.__version__}", file=sys.stderr)
|
| 27 |
print(f"Python executable: {sys.executable}", file=sys.stderr)
|
|
|
|
| 829 |
|
| 830 |
# --- Output Download Functions (using global_analysis_df_for_export) ---
|
| 831 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 832 |
|
| 833 |
def generate_pdf_report(selected_timezone):
|
| 834 |
df = global_analysis_df_for_export
|
| 835 |
if df.empty:
|
| 836 |
+
return gr.update(value=None, visible=False)
|
| 837 |
+
|
| 838 |
+
# Initialize FPDF with UTF-8 encoding (fpdf2 handles this implicitly better)
|
| 839 |
+
pdf = FPDF(unit="pt", format="A4")
|
| 840 |
+
|
| 841 |
+
# Add a Unicode font (DejaVuSans.ttf is a good choice for broad character support)
|
| 842 |
+
# Make sure 'DejaVuSans.ttf' is in the same directory as your script
|
| 843 |
+
# Or provide the full path to the .ttf file
|
| 844 |
+
try:
|
| 845 |
+
pdf.add_font('DejaVu', '', 'DejaVuSans.ttf', uni=True)
|
| 846 |
+
# Verify font was added
|
| 847 |
+
if 'DejaVu' not in pdf.fonts:
|
| 848 |
+
raise RuntimeError("DejaVu font not registered properly.")
|
| 849 |
+
except RuntimeError as e:
|
| 850 |
+
print(f"Warning: Problem loading DejaVuSans.ttf: {e}. Falling back to Arial for PDF generation.", file=sys.stderr)
|
| 851 |
+
# Fallback for font if DejaVu fails
|
| 852 |
+
def set_default_font():
|
| 853 |
+
pdf.set_font("Arial", size=12)
|
| 854 |
+
pdf.set_font_method = set_default_font # Override set_font to use Arial fallback
|
| 855 |
+
except Exception as e:
|
| 856 |
+
print(f"An unexpected error occurred while adding font: {e}. Falling back to Arial.", file=sys.stderr)
|
| 857 |
+
def set_default_font():
|
| 858 |
+
pdf.set_font("Arial", size=12)
|
| 859 |
+
pdf.set_font_method = set_default_font # Override set_font to use Arial fallback
|
| 860 |
+
|
| 861 |
|
|
|
|
| 862 |
pdf.add_page()
|
| 863 |
|
| 864 |
+
# Use the Unicode font if it was successfully loaded, otherwise Arial
|
| 865 |
+
font_name_bold = "DejaVu" if 'DejaVu' in pdf.fonts else "Arial"
|
| 866 |
+
font_name_regular = "DejaVu" if 'DejaVu' in pdf.fonts else "Arial"
|
| 867 |
+
|
| 868 |
+
pdf.set_font(font_name_bold, 'B', 16)
|
| 869 |
+
pdf.cell(0, 20, "AI Customer Ticket Analysis Report", ln=True, align='C')
|
| 870 |
+
pdf.ln(10)
|
| 871 |
|
| 872 |
now_utc = datetime.utcnow()
|
| 873 |
current_gen_time = format_time_for_display(now_utc, selected_timezone)
|
| 874 |
+
pdf.set_font(font_name_regular, size=10)
|
| 875 |
+
pdf.cell(0, 15, f"Generated on: {current_gen_time}", ln=True, align='R')
|
| 876 |
+
pdf.ln(15)
|
| 877 |
|
| 878 |
+
pdf.set_font(font_name_bold, 'B', 14)
|
| 879 |
+
pdf.multi_cell(0, 15, "Overview & Key Insights")
|
| 880 |
+
pdf.set_font(font_name_regular, size=10)
|
| 881 |
total_tickets_processed = df['Count'].sum() if 'Count' in df.columns else 0
|
| 882 |
total_categories_identified = len(df['Category'].unique()) if 'Category' in df.columns else 0
|
| 883 |
|
|
|
|
| 887 |
f"Total Tickets Summarized in Report: {total_tickets_processed}.\n"
|
| 888 |
f"Total Categories Identified: {total_categories_identified}."
|
| 889 |
)
|
| 890 |
+
pdf.multi_cell(0, 10, overview_text)
|
| 891 |
+
pdf.ln(15)
|
| 892 |
|
| 893 |
+
pdf.set_font(font_name_bold, 'B', 14)
|
| 894 |
+
pdf.multi_cell(0, 15, "Aggregated Issue Summaries")
|
| 895 |
+
pdf.ln(10)
|
| 896 |
|
|
|
|
|
|
|
|
|
|
| 897 |
for category, category_df in df.groupby('Category'):
|
| 898 |
total_count = category_df['Count'].sum()
|
| 899 |
|
| 900 |
pdf.set_fill_color(220, 237, 200) # Light Green
|
| 901 |
pdf.set_text_color(76, 175, 80) # Dark Green
|
| 902 |
+
pdf.set_font(font_name_bold, 'B', 12)
|
| 903 |
+
pdf.multi_cell(0, 12, f"{category} ({total_count})", 0, 'L', 1)
|
| 904 |
+
pdf.ln(5)
|
| 905 |
|
| 906 |
pdf.set_text_color(0, 0, 0) # Black
|
| 907 |
+
pdf.set_font(font_name_bold, 'B', 10)
|
| 908 |
+
pdf.multi_cell(0, 10, "Summary:")
|
| 909 |
+
pdf.set_font(font_name_regular, size=9)
|
| 910 |
unique_issues = category_df['Issue Description (Summary)'].astype(str).drop_duplicates().apply(lambda x: x.replace('*', '')).tolist()
|
| 911 |
for issue in unique_issues:
|
| 912 |
+
pdf.multi_cell(0, 9, f"\u2022 {issue}", 0, 'L')
|
| 913 |
+
if not unique_issues: pdf.multi_cell(0, 9, "\u2022 No specific summary available.", 0, 'L')
|
| 914 |
+
pdf.ln(5)
|
| 915 |
|
| 916 |
+
pdf.set_font(font_name_bold, 'B', 10)
|
| 917 |
+
pdf.multi_cell(0, 10, "Sub-Categories:")
|
| 918 |
+
pdf.set_font(font_name_regular, size=9)
|
| 919 |
if 'Sub-Category' in category_df.columns:
|
| 920 |
sub_cat_counts = category_df.groupby('Sub-Category')['Count'].sum().sort_values(ascending=False)
|
| 921 |
for sub_cat, count in sub_cat_counts.items():
|
| 922 |
+
pdf.multi_cell(0, 9, f"\u2022 {sub_cat} (Count: {count})", 0, 'L')
|
| 923 |
+
if 'Sub-Category' not in category_df.columns or sub_cat_counts.empty: pdf.multi_cell(0, 9, "\u2022 No sub-categories identified.", 0, 'L')
|
| 924 |
+
pdf.ln(5)
|
| 925 |
|
| 926 |
+
pdf.set_font(font_name_bold, 'B', 10)
|
| 927 |
+
pdf.multi_cell(0, 10, "Root Cause:")
|
| 928 |
+
pdf.set_font(font_name_regular, size=9)
|
| 929 |
unique_root_causes = category_df['Root Cause (Inferred)'].astype(str).drop_duplicates().apply(lambda x: x.replace('*', '')).tolist()
|
| 930 |
for rc in unique_root_causes:
|
| 931 |
+
pdf.multi_cell(0, 9, f"\u2022 {rc}", 0, 'L')
|
| 932 |
+
if not unique_root_causes: pdf.multi_cell(0, 9, "\u2022 Undetermined.", 0, 'L')
|
| 933 |
+
pdf.ln(5)
|
| 934 |
|
| 935 |
+
pdf.set_font(font_name_bold, 'B', 10)
|
| 936 |
+
pdf.multi_cell(0, 10, "Resolution Steps:")
|
| 937 |
+
pdf.set_font(font_name_regular, size=9)
|
| 938 |
unique_resolutions = category_df['Potential Resolution Steps'].astype(str).drop_duplicates().apply(lambda x: x.replace('*', '')).tolist()
|
| 939 |
for res in unique_resolutions:
|
| 940 |
+
pdf.multi_cell(0, 9, f"\u2022 {res}", 0, 'L')
|
| 941 |
+
if not unique_resolutions: pdf.multi_cell(0, 9, "\u2022 No specific steps identified.", 0, 'L')
|
| 942 |
+
pdf.ln(5)
|
| 943 |
|
| 944 |
+
pdf.set_font(font_name_bold, 'B', 10)
|
| 945 |
+
pdf.multi_cell(0, 10, "Future Prediction & Proactive Measures:")
|
| 946 |
+
pdf.set_font(font_name_regular, size=9)
|
| 947 |
future_predictions = category_df['Future Prediction'].astype(str).drop_duplicates().apply(lambda x: x.replace('*', '')).tolist()
|
| 948 |
proactive_measures = category_df['Proactive Measures'].astype(str).drop_duplicates().apply(lambda x: x.replace('*', '')).tolist()
|
| 949 |
|
| 950 |
if future_predictions:
|
| 951 |
pdf.set_text_color(141, 110, 99) # Brownish
|
| 952 |
+
pdf.multi_cell(0, 9, f"\u2022 Future Prediction: {'; '.join(future_predictions)}", 0, 'L')
|
| 953 |
if proactive_measures:
|
| 954 |
pdf.set_text_color(141, 110, 99) # Brownish
|
| 955 |
+
pdf.multi_cell(0, 9, f"\u2022 Proactive Measures: {'; '.join(proactive_measures)}", 0, 'L')
|
| 956 |
if not future_predictions and not proactive_measures:
|
| 957 |
pdf.set_text_color(0, 0, 0) # Black
|
| 958 |
+
pdf.multi_cell(0, 9, "\u2022 No future prediction or proactive measures identified.", 0, 'L')
|
| 959 |
+
pdf.ln(10)
|
| 960 |
|
| 961 |
pdf.set_text_color(0, 0, 0)
|
| 962 |
pdf.set_fill_color(241, 248, 233)
|
| 963 |
+
pdf.set_font(font_name_bold, 'B', 12)
|
| 964 |
pdf.multi_cell(0, 10, "Disclaimer", 0, 'L', 1)
|
| 965 |
+
pdf.ln(5)
|
| 966 |
+
pdf.set_font(font_name_regular, size=9)
|
| 967 |
+
pdf.multi_cell(0, 8, "*Note: AI-generated insights, especially 'Root Cause', 'Future Prediction', and 'Proactive Measures', are inferred based on the provided ticket descriptions. These are educated guesses by the AI and should be used for guidance only. For critical business decisions, human expert validation is highly recommended.")
|
| 968 |
+
pdf.ln(5)
|
| 969 |
+
pdf.multi_cell(0, 8, f"Report generated on: {current_gen_time}")
|
|
|
|
| 970 |
|
| 971 |
with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as tmp_file:
|
| 972 |
pdf.output(tmp_file.name)
|
| 973 |
temp_file_path = tmp_file.name
|
| 974 |
|
| 975 |
+
return gr.update(value=temp_file_path, visible=True, label="Download PDF Report")
|
| 976 |
|
| 977 |
|
| 978 |
def generate_docx_report(selected_timezone):
|
| 979 |
df = global_analysis_df_for_export
|
| 980 |
if df.empty:
|
| 981 |
+
return gr.update(value=None, visible=False)
|
| 982 |
|
| 983 |
doc = DocxDocument()
|
| 984 |
|
|
|
|
| 1071 |
doc.save(tmp_file)
|
| 1072 |
temp_file_path = tmp_file.name
|
| 1073 |
|
| 1074 |
+
return gr.update(value=temp_file_path, visible=True, label="Download Word Report")
|
| 1075 |
|
| 1076 |
|
| 1077 |
# --- Gradio Interface Setup ---
|