feat: add PDF export generation via WeasyPrint with branded template

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
chihlasm
2026-03-17 00:28:22 -04:00
parent 2c11917b5a
commit 312024e143
6 changed files with 686 additions and 1 deletions

View File

@@ -6,6 +6,10 @@ WORKDIR /app
RUN apt-get update && apt-get install -y \
gcc \
libpq-dev \
libpango1.0-dev \
libcairo2-dev \
libgdk-pixbuf2.0-dev \
libffi-dev \
&& rm -rf /var/lib/apt/lists/*
# Install Python dependencies

View File

@@ -391,6 +391,22 @@ async def export_session(
detail="You don't have access to this session"
)
# PDF export — separate path with binary response
if export_options.format == "pdf":
from app.services.export_service import generate_pdf_export
from fastapi.responses import Response
pdf_bytes = await generate_pdf_export(session, export_options, db)
if session.completed_at:
session.exported = True
await db.commit()
return Response(
content=pdf_bytes,
media_type="application/pdf",
headers={"Content-Disposition": f'attachment; filename="session-export-{session_id}.pdf"'},
)
# Generate export based on format
if export_options.format == "markdown":
content = generate_markdown_export(session, export_options)

View File

@@ -1,11 +1,13 @@
"""
Session export generators for ResolutionFlow.
Provides markdown, plain text, HTML, and PSA/ticket note export formatters
Provides markdown, plain text, HTML, PDF, and PSA/ticket note export formatters
for troubleshooting sessions.
"""
import html
import os
from datetime import datetime
from pathlib import Path
from typing import Any
from app.models.session import Session
@@ -904,3 +906,188 @@ def _generate_procedural_psa(session: Session, options: SessionExport) -> str:
lines.append(f"Duration: {_format_duration(session.started_at, session.completed_at)}")
return "\n".join(lines)
async def generate_pdf_export(session: Session, options: SessionExport, db) -> bytes:
"""Generate PDF export using WeasyPrint and a Jinja2 HTML template.
Args:
session: The session to export.
options: Export options (redaction_mode, max_step_index, etc.).
db: Async database session for loading supporting data and branding.
Returns:
PDF file contents as bytes.
"""
from jinja2 import Environment, FileSystemLoader
import weasyprint
from sqlalchemy import select as sa_select
# Load Jinja2 template
template_dir = Path(__file__).resolve().parent.parent / "templates"
env = Environment(loader=FileSystemLoader(str(template_dir)), autoescape=True)
template = env.get_template("export_pdf.html")
# Tree snapshot data
tree_snapshot = session.tree_snapshot or {}
flow_title = tree_snapshot.get("name", "Session Export")
tree_type = tree_snapshot.get("tree_type", "troubleshooting")
is_procedural = tree_type == "procedural"
report_type = "Procedure Report" if is_procedural else "Troubleshooting Report"
# Branding — check team first, then user (solo pros)
logo_data = None
logo_content_type = None
company_name = None
from app.models.user import User
user_result = await db.execute(
sa_select(User).where(User.id == session.user_id)
)
user = user_result.scalar_one_or_none()
engineer_name = user.name if user else "Unknown"
if user and user.team_id:
from app.models.team import Team
team_result = await db.execute(
sa_select(Team).where(Team.id == user.team_id)
)
team = team_result.scalar_one_or_none()
if team:
logo_data = team.logo_data
logo_content_type = team.logo_content_type
company_name = team.company_display_name or team.name
elif user:
logo_data = user.logo_data
logo_content_type = user.logo_content_type
company_name = user.company_display_name
has_custom_logo = bool(logo_data)
# Build steps list from decisions
decisions = session.decisions or []
if options.max_step_index is not None:
decisions = decisions[:options.max_step_index]
steps = []
for decision in decisions:
title = decision.get("question") or decision.get("action_performed", "Step")
answer = decision.get("answer", "")
notes = decision.get("notes", "")
duration_seconds = _get_step_duration_seconds(decision)
duration_str = _format_step_duration(duration_seconds) if duration_seconds is not None else None
if is_procedural:
completed = answer == "completed"
decision_text = "Completed" if completed else ("Skipped" if answer else "")
else:
decision_text = answer
steps.append({
"title": title,
"decision": decision_text,
"notes": notes,
"duration": duration_str,
})
# Query supporting data
from app.models.supporting_data import SessionSupportingData
sd_result = await db.execute(
sa_select(SessionSupportingData)
.where(SessionSupportingData.session_id == session.id)
.order_by(SessionSupportingData.sort_order)
)
supporting_data_rows = sd_result.scalars().all()
supporting_data = [
{
"label": sd.label,
"data_type": sd.data_type,
"content": sd.content,
"content_type": sd.content_type,
}
for sd in supporting_data_rows
]
# Calculate duration and format outcome
duration = _format_duration(session.started_at, session.completed_at)
session_date = session.started_at.strftime("%Y-%m-%d %H:%M")
outcome_label = _get_outcome_label(session) or ("In Progress" if not session.completed_at else "Completed")
outcome_raw = getattr(session, "outcome", None) or ""
outcome_class = f"outcome-{outcome_raw}" if outcome_raw else ""
# Build summary text
summary_text = ""
if options.include_summary:
summary_fields = _build_summary_fields(session)
parts = []
for label, value in summary_fields.items():
if value:
parts.append(f"{label.replace('_', ' ').title()}: {value}")
summary_text = "\n".join(parts)
# Resolution / outcome notes as summary fallback
if not summary_text:
_raw_notes = getattr(session, "outcome_notes", None)
if isinstance(_raw_notes, str) and _raw_notes.strip():
summary_text = _raw_notes.strip()
generated_at = datetime.utcnow().strftime("%Y-%m-%d %H:%M UTC")
# Variable resolution
session_vars = getattr(session, "session_variables", None) or {}
if session_vars:
from app.services.variable_service import resolve_variables
flow_title = resolve_variables(flow_title, session_vars)
summary_text = resolve_variables(summary_text, session_vars)
for step in steps:
step["title"] = resolve_variables(step["title"], session_vars)
if step["decision"]:
step["decision"] = resolve_variables(step["decision"], session_vars)
if step["notes"]:
step["notes"] = resolve_variables(step["notes"], session_vars)
for sd in supporting_data:
if sd["data_type"] == "text_snippet":
sd["content"] = resolve_variables(sd["content"], session_vars)
# Apply redaction
if options.redaction_mode == "mask":
from app.services.redaction_service import apply_redaction_to_text
try:
flow_title, _ = apply_redaction_to_text(flow_title)
summary_text, _ = apply_redaction_to_text(summary_text)
for step in steps:
step["title"], _ = apply_redaction_to_text(step["title"])
if step["decision"]:
step["decision"], _ = apply_redaction_to_text(step["decision"])
if step["notes"]:
step["notes"], _ = apply_redaction_to_text(step["notes"])
for sd in supporting_data:
if sd["data_type"] == "text_snippet":
sd["content"], _ = apply_redaction_to_text(sd["content"])
except Exception:
pass # Redaction is best-effort for PDF
# Render HTML
html_content = template.render(
report_type=report_type,
flow_title=flow_title,
logo_data=logo_data,
logo_content_type=logo_content_type or "image/png",
has_custom_logo=has_custom_logo,
company_name=company_name,
engineer_name=engineer_name,
client_name=session.client_name,
ticket_number=session.ticket_number,
session_date=session_date,
duration=duration,
outcome_class=outcome_class,
outcome_display=outcome_label,
summary=summary_text,
steps=steps,
supporting_data=supporting_data,
generated_at=generated_at,
)
# Convert to PDF
pdf_bytes = weasyprint.HTML(string=html_content).write_pdf()
return pdf_bytes

View File

@@ -0,0 +1,378 @@
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<style>
{% if has_custom_logo %}
@page {
size: A4;
margin: 2cm;
@bottom-right {
content: "Powered by ResolutionFlow";
font-size: 8pt;
color: #999;
}
@bottom-left {
content: "Generated {{ generated_at }}";
font-size: 8pt;
color: #999;
}
}
{% else %}
@page {
size: A4;
margin: 2cm;
@bottom-left {
content: "Generated {{ generated_at }}";
font-size: 8pt;
color: #999;
}
}
{% endif %}
* {
margin: 0;
padding: 0;
box-sizing: border-box;
}
body {
font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, Helvetica, Arial, sans-serif;
color: #1a1a2e;
background: #ffffff;
font-size: 10pt;
line-height: 1.5;
}
/* --- Header --- */
.header {
display: flex;
justify-content: space-between;
align-items: flex-start;
margin-bottom: 24px;
padding-bottom: 16px;
border-bottom: 3px solid #06b6d4;
}
.header-left {
flex: 1;
}
.report-type {
font-size: 8pt;
text-transform: uppercase;
letter-spacing: 0.1em;
color: #06b6d4;
font-weight: 600;
margin-bottom: 4px;
}
.flow-title {
font-size: 18pt;
font-weight: 700;
color: #1a1a2e;
margin-bottom: 4px;
}
.company-name {
font-size: 10pt;
color: #666;
}
.header-logo {
flex-shrink: 0;
margin-left: 24px;
}
.header-logo img {
max-width: 120px;
max-height: 60px;
object-fit: contain;
}
/* --- Metadata Grid --- */
.metadata-grid {
display: grid;
grid-template-columns: 1fr 1fr 1fr;
gap: 12px;
margin-bottom: 24px;
padding: 16px;
background: #f8fafc;
border-radius: 8px;
border: 1px solid #e2e8f0;
}
.meta-item {
display: flex;
flex-direction: column;
}
.meta-label {
font-size: 7pt;
text-transform: uppercase;
letter-spacing: 0.08em;
color: #94a3b8;
font-weight: 600;
margin-bottom: 2px;
}
.meta-value {
font-size: 10pt;
color: #1a1a2e;
font-weight: 500;
}
.outcome-resolved {
color: #059669;
font-weight: 700;
}
.outcome-escalated {
color: #d97706;
font-weight: 700;
}
.outcome-workaround {
color: #2563eb;
font-weight: 700;
}
.outcome-unresolved {
color: #dc2626;
font-weight: 700;
}
/* --- Section Headers --- */
.section-title {
font-size: 12pt;
font-weight: 700;
color: #1a1a2e;
margin-bottom: 12px;
padding-bottom: 6px;
border-bottom: 2px solid #06b6d4;
}
/* --- Summary --- */
.summary {
margin-bottom: 24px;
}
.summary-text {
font-size: 10pt;
color: #334155;
line-height: 1.6;
white-space: pre-wrap;
}
/* --- Troubleshooting Path Timeline --- */
.timeline {
margin-bottom: 24px;
}
.timeline-step {
display: flex;
align-items: flex-start;
margin-bottom: 8px;
position: relative;
padding-left: 24px;
}
.timeline-step::before {
content: "";
position: absolute;
left: 5px;
top: 6px;
width: 10px;
height: 10px;
background: #06b6d4;
border-radius: 50%;
}
.timeline-step::after {
content: "";
position: absolute;
left: 9px;
top: 18px;
width: 2px;
height: calc(100% + 0px);
background: #e2e8f0;
}
.timeline-step:last-child::after {
display: none;
}
.step-content {
flex: 1;
padding-bottom: 8px;
}
.step-title {
font-size: 10pt;
font-weight: 600;
color: #1a1a2e;
}
.step-decision {
font-size: 9pt;
color: #06b6d4;
font-weight: 500;
margin-top: 2px;
}
.step-notes {
font-size: 9pt;
color: #64748b;
font-style: italic;
margin-top: 2px;
}
.step-duration {
font-size: 8pt;
color: #94a3b8;
margin-top: 2px;
}
/* --- Supporting Data --- */
.supporting-data {
margin-bottom: 24px;
}
.supporting-item {
margin-bottom: 16px;
break-inside: avoid;
border: 1px solid #e2e8f0;
border-radius: 8px;
overflow: hidden;
}
.supporting-item-label {
font-size: 9pt;
font-weight: 600;
color: #1a1a2e;
padding: 8px 12px;
background: #f8fafc;
border-bottom: 1px solid #e2e8f0;
}
.supporting-item-content {
padding: 12px;
}
.code-block {
font-family: "JetBrains Mono", "Fira Code", "Consolas", monospace;
font-size: 8pt;
line-height: 1.5;
background: #f1f5f9;
padding: 12px;
border-radius: 4px;
white-space: pre-wrap;
word-break: break-all;
color: #334155;
}
.screenshot-img {
max-width: 100%;
height: auto;
border-radius: 4px;
}
</style>
</head>
<body>
<!-- Header -->
<div class="header">
<div class="header-left">
<div class="report-type">{{ report_type }}</div>
<div class="flow-title">{{ flow_title }}</div>
{% if company_name %}
<div class="company-name">{{ company_name }}</div>
{% endif %}
</div>
{% if logo_data %}
<div class="header-logo">
<img src="data:{{ logo_content_type }};base64,{{ logo_data }}" alt="Logo">
</div>
{% endif %}
</div>
<!-- Metadata Grid -->
<div class="metadata-grid">
<div class="meta-item">
<span class="meta-label">Engineer</span>
<span class="meta-value">{{ engineer_name or "N/A" }}</span>
</div>
<div class="meta-item">
<span class="meta-label">Client</span>
<span class="meta-value">{{ client_name or "N/A" }}</span>
</div>
<div class="meta-item">
<span class="meta-label">Ticket #</span>
<span class="meta-value">{{ ticket_number or "N/A" }}</span>
</div>
<div class="meta-item">
<span class="meta-label">Date</span>
<span class="meta-value">{{ session_date }}</span>
</div>
<div class="meta-item">
<span class="meta-label">Duration</span>
<span class="meta-value">{{ duration }}</span>
</div>
<div class="meta-item">
<span class="meta-label">Outcome</span>
<span class="meta-value {{ outcome_class }}">{{ outcome_display }}</span>
</div>
</div>
<!-- Summary -->
{% if summary %}
<div class="summary">
<div class="section-title">Summary</div>
<div class="summary-text">{{ summary }}</div>
</div>
{% endif %}
<!-- Troubleshooting Path -->
{% if steps %}
<div class="timeline">
<div class="section-title">Troubleshooting Path</div>
{% for step in steps %}
<div class="timeline-step">
<div class="step-content">
<div class="step-title">{{ loop.index }}. {{ step.title }}</div>
{% if step.decision %}
<div class="step-decision">{{ step.decision }}</div>
{% endif %}
{% if step.notes %}
<div class="step-notes">{{ step.notes }}</div>
{% endif %}
{% if step.duration %}
<div class="step-duration">{{ step.duration }}</div>
{% endif %}
</div>
</div>
{% endfor %}
</div>
{% endif %}
<!-- Supporting Data -->
{% if supporting_data %}
<div class="supporting-data">
<div class="section-title">Supporting Data</div>
{% for item in supporting_data %}
<div class="supporting-item">
<div class="supporting-item-label">{{ item.label }}</div>
<div class="supporting-item-content">
{% if item.data_type == "screenshot" %}
<img class="screenshot-img" src="data:{{ item.content_type or 'image/png' }};base64,{{ item.content }}" alt="{{ item.label }}">
{% else %}
<div class="code-block">{{ item.content }}</div>
{% endif %}
</div>
</div>
{% endfor %}
</div>
{% endif %}
</body>
</html>

View File

@@ -42,6 +42,10 @@ voyageai>=0.3.0
# Monitoring
sentry-sdk[fastapi]>=2.54.0
# PDF Export
weasyprint>=62.0
jinja2>=3.1.0
# Utilities
python-dotenv==1.0.1
croniter>=2.0.0

View File

@@ -0,0 +1,96 @@
"""Tests for PDF export via WeasyPrint."""
import pytest
from httpx import AsyncClient
@pytest.mark.asyncio
class TestPDFExport:
"""Test PDF export endpoint."""
async def test_export_pdf_returns_pdf_content(
self, client: AsyncClient, auth_headers: dict, test_tree: dict
):
"""Test that PDF export returns application/pdf content starting with %PDF."""
# Create a session
create_response = await client.post(
"/api/v1/sessions",
json={"tree_id": test_tree["id"], "ticket_number": "PDF-001"},
headers=auth_headers,
)
assert create_response.status_code in (200, 201)
session_id = create_response.json()["id"]
# Add a decision so there's content
await client.put(
f"/api/v1/sessions/{session_id}",
json={
"decisions": [
{
"node_id": "root",
"question": "Is this a test?",
"answer": "Yes",
"notes": "PDF export test",
"timestamp": "2026-03-17T10:00:00Z",
}
]
},
headers=auth_headers,
)
# Export as PDF
response = await client.post(
f"/api/v1/sessions/{session_id}/export",
json={"format": "pdf", "include_tree_info": True},
headers=auth_headers,
)
assert response.status_code == 200
assert response.headers["content-type"] == "application/pdf"
assert "session-export-" in response.headers.get("content-disposition", "")
# PDF files start with %PDF
assert response.content[:5] == b"%PDF-"
async def test_export_pdf_with_no_supporting_data(
self, client: AsyncClient, auth_headers: dict, test_tree: dict
):
"""Test PDF export works when session has no supporting data."""
create_response = await client.post(
"/api/v1/sessions",
json={"tree_id": test_tree["id"]},
headers=auth_headers,
)
assert create_response.status_code in (200, 201)
session_id = create_response.json()["id"]
response = await client.post(
f"/api/v1/sessions/{session_id}/export",
json={"format": "pdf"},
headers=auth_headers,
)
assert response.status_code == 200
assert response.headers["content-type"] == "application/pdf"
assert response.content[:5] == b"%PDF-"
async def test_existing_markdown_export_still_works(
self, client: AsyncClient, auth_headers: dict, test_tree: dict
):
"""Verify markdown export is unaffected by PDF addition."""
create_response = await client.post(
"/api/v1/sessions",
json={"tree_id": test_tree["id"], "ticket_number": "MD-001"},
headers=auth_headers,
)
assert create_response.status_code in (200, 201)
session_id = create_response.json()["id"]
response = await client.post(
f"/api/v1/sessions/{session_id}/export",
json={"format": "markdown", "include_tree_info": True},
headers=auth_headers,
)
assert response.status_code == 200
assert "text/markdown" in response.headers["content-type"]
assert "MD-001" in response.text