feat: PDF-Export zusätzlich zu DOCX

Mac-Worker: - pdf_export.py mit reportlab (pure Python, kein externes Tool) - Nutzt dieselbe profile.docx-Konfig wie der DOCX-Renderer → strukturell identisch, nur andere Optik - POST /api/export/pdf - requirements: reportlab==4.2.5 LXC: - Job.pdf_path neu (Migration: ALTER TABLE job ADD COLUMN) - Pipeline-Step nach DOCX: PDF wird ebenfalls erzeugt (Progress: 90 DOCX → 95 PDF → 100 done) - /api/jobs/{id}/download/pdf - /api/jobs/{id}/retry löscht PDF mit, damit es neu erzeugt wird - Frontend: zusätzlicher Download-Button "PDF" (vorne im Array) - Tabelle: Download-Spalte 320px, 3-spaltiges Grid (5 Buttons → 3+2) - Cards (Mobile) bleiben 2-spaltig (5 Buttons → 3 Zeilen)
2026-05-14 04:48:47 +00:00
parent 7d4961933e
commit 354e4c57eb
10 changed files with 222 additions and 9 deletions
@@ -0,0 +1,165 @@
+"""Profil-getriebenes PDF-Rendering mit reportlab.
+
+Nutzt dieselbe `docx`-Konfig im Profil wie der DOCX-Renderer (meta + sections).
+Damit ist das PDF strukturell konsistent zum DOCX, nur in anderer Optik.
+"""
+from __future__ import annotations
+
+from io import BytesIO
+from typing import Any
+
+from reportlab.lib import colors
+from reportlab.lib.pagesizes import A4
+from reportlab.lib.styles import ParagraphStyle, getSampleStyleSheet
+from reportlab.lib.units import cm
+from reportlab.platypus import (
+    ListFlowable,
+    ListItem,
+    Paragraph,
+    SimpleDocTemplate,
+    Spacer,
+)
+
+from . import profiles
+
+
+def _build_styles():
+    s = getSampleStyleSheet()
+    s.add(ParagraphStyle(
+        name="DocTitle", parent=s["Title"],
+        fontSize=20, spaceAfter=10, textColor=colors.HexColor("#1a1a1a"),
+    ))
+    s.add(ParagraphStyle(
+        name="MetaLine", parent=s["Normal"],
+        fontSize=10, leading=13, textColor=colors.HexColor("#444444"),
+    ))
+    s.add(ParagraphStyle(
+        name="SectionHeading", parent=s["Heading2"],
+        fontSize=13, spaceBefore=14, spaceAfter=6,
+        textColor=colors.HexColor("#1a1a1a"),
+    ))
+    s.add(ParagraphStyle(
+        name="Body2", parent=s["BodyText"],
+        fontSize=10.5, leading=14,
+    ))
+    s.add(ParagraphStyle(
+        name="BulletItem", parent=s["BodyText"],
+        fontSize=10.5, leading=14, leftIndent=0,
+    ))
+    return s
+
+
+_STYLES = _build_styles()
+
+
+def _esc(value: Any) -> str:
+    """XML-Escape für reportlab Paragraph-Markup."""
+    if value is None:
+        return ""
+    return (
+        str(value)
+        .replace("&", "&amp;")
+        .replace("<", "&lt;")
+        .replace(">", "&gt;")
+    )
+
+
+def _para(text: str, style_name: str = "Body2") -> Paragraph:
+    """Paragraph mit Newline → <br/>."""
+    return Paragraph(_esc(text).replace("\n", "<br/>"), _STYLES[style_name])
+
+
+def build_pdf(data: dict[str, Any], profile_name: str | None = None) -> bytes:
+    profile = profiles.get(profile_name)
+    cfg = profile.docx or {}
+
+    title_field = cfg.get("title_field", "title")
+    title = (data.get(title_field) or cfg.get("default_title") or profile.display_name).strip()
+
+    buf = BytesIO()
+    doc = SimpleDocTemplate(
+        buf,
+        pagesize=A4,
+        leftMargin=2 * cm,
+        rightMargin=2 * cm,
+        topMargin=2 * cm,
+        bottomMargin=2 * cm,
+        title=title,
+        author="Voice-Agent",
+    )
+
+    story: list = [Paragraph(_esc(title), _STYLES["DocTitle"])]
+
+    # Meta-Block
+    meta_paras = []
+    for meta in cfg.get("meta", []) or []:
+        val = data.get(meta["key"])
+        if val:
+            label = meta.get("label", meta["key"])
+            meta_paras.append(Paragraph(
+                f"<b>{_esc(label)}:</b> {_esc(val)}",
+                _STYLES["MetaLine"],
+            ))
+    story.extend(meta_paras)
+    if meta_paras:
+        story.append(Spacer(1, 8))
+
+    # Sektionen
+    for section in cfg.get("sections", []) or []:
+        _render_section(story, data, section)
+
+    # Parse-Error-Fallback
+    if data.get("_parse_error") and data.get("_raw"):
+        story.append(Paragraph("Roh-Antwort des Modells", _STYLES["SectionHeading"]))
+        story.append(_para(str(data["_raw"])))
+
+    doc.build(story)
+    return buf.getvalue()
+
+
+def _render_section(story: list, data: dict, section: dict) -> None:
+    val = data.get(section["key"])
+    if val in (None, "", [], {}):
+        return
+    label = section.get("label", section["key"])
+    sec_type = section.get("type", "text")
+
+    story.append(Paragraph(_esc(label), _STYLES["SectionHeading"]))
+
+    if sec_type == "text":
+        story.append(_para(str(val)))
+    elif sec_type == "list":
+        items = val if isinstance(val, list) else [val]
+        bullets = [ListItem(_para(str(i), "BulletItem"), leftIndent=12) for i in items]
+        story.append(ListFlowable(bullets, bulletType="bullet", leftIndent=15))
+    elif sec_type == "tasks":
+        items = val if isinstance(val, list) else [val]
+        bullets = []
+        for t in items:
+            if isinstance(t, dict):
+                owner = t.get("owner") or "n/a"
+                task = t.get("task") or t.get("description") or ""
+                bullets.append(ListItem(
+                    Paragraph(f"<b>{_esc(owner)}:</b> {_esc(task)}", _STYLES["BulletItem"]),
+                    leftIndent=12,
+                ))
+            else:
+                bullets.append(ListItem(_para(str(t), "BulletItem"), leftIndent=12))
+        story.append(ListFlowable(bullets, bulletType="bullet", leftIndent=15))
+    elif sec_type == "transcript":
+        items = val if isinstance(val, list) else [val]
+        for item in items:
+            if isinstance(item, dict):
+                t = item.get("time", "")
+                spk = item.get("speaker", "")
+                txt = item.get("text", "")
+                story.append(Paragraph(
+                    f"<b>[{_esc(t)}] {_esc(spk)}:</b> {_esc(txt)}",
+                    _STYLES["Body2"],
+                ))
+            else:
+                story.append(_para(str(item)))
+    else:
+        story.append(_para(str(val)))
+
+    story.append(Spacer(1, 4))
@@ -9,6 +9,7 @@ from pydantic import BaseModel
 from app.core import ollama_client, profiles
 from app.core.config import settings
 from app.core.docx_export import build_docx
+from app.core.pdf_export import build_pdf
 from app.core.whisper_engine import engine as whisper

 logging.basicConfig(level=logging.INFO, format="%(asctime)s [%(levelname)s] %(name)s: %(message)s")
@@ -35,6 +36,11 @@ class DocxIn(BaseModel):
    profile: str | None = None


+class PdfIn(BaseModel):
+    data: dict
+    profile: str | None = None
+
+
@app.get("/health")
 async def health():
    info = {
@@ -99,3 +105,13 @@ async def export_docx(payload: DocxIn):
        media_type="application/vnd.openxmlformats-officedocument.wordprocessingml.document",
        headers={"Content-Disposition": 'attachment; filename="protocol.docx"'},
    )
+
+
+@app.post("/api/export/pdf")
+async def export_pdf(payload: PdfIn):
+    data = build_pdf(payload.data, profile_name=payload.profile)
+    return Response(
+        content=data,
+        media_type="application/pdf",
+        headers={"Content-Disposition": 'attachment; filename="protocol.pdf"'},
+    )