claw/diarization-ui
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix(pdf): sanitize text to Latin-1 before fpdf2, absorb per-line errors

Browse Source 
The fpdf2 built-in Helvetica font is Latin-1 only; characters outside
that range caused "Not enough horizontal space" errors which silently
aborted PDF generation. Fix:
- _l1() helper encodes/decodes via latin-1 with replacement before any
  multi_cell call
- _cell() wrapper catches per-line render errors so one bad character
  doesn't drop the whole document
- PDF generation errors are now logged as warnings instead of swallowed

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
Wolf G. Beckmann
2026-06-23 15:56:38 +02:00
parent aff166f89d
commit 4a81ef3673
1 changed files with 28 additions and 24 deletions
Download Patch File Download Diff File Expand all files Collapse all files

52
app.py
View File

@@ -262,11 +262,18 @@ def _split_thinking(text: str) -> tuple[str, str]:
def _md_to_pdf(title: str, content_md: str) -> bytes:
"""Render Markdown content to a PDF byte string."""
def _l1(text: str) -> str:
"""Sanitize text to Latin-1; fpdf2 built-in fonts require it."""
return text.encode("latin-1", errors="replace").decode("latin-1")
_title = _l1(title[:100])
class PDF(fpdflib.FPDF):
def header(self):
self.set_font("Helvetica", "B", 9)
self.set_text_color(100, 100, 100)
self.cell(0, 6, title, align="L")
self.cell(0, 6, _title, align="L")
self.ln(2)
self.set_draw_color(200, 200, 200)
self.line(self.l_margin, self.get_y(), self.w - self.r_margin, self.get_y())
@@ -282,41 +289,37 @@ def _md_to_pdf(title: str, content_md: str) -> bytes:
pdf.set_auto_page_break(auto=True, margin=15)
pdf.add_page()
# Render markdown line by line with minimal formatting
def _cell(text: str, size: int = 10, bold: bool = False, indent: float = 0, h: float = 5):
try:
pdf.set_font("Helvetica", "B" if bold else "", size)
pdf.set_text_color(20 if bold else 30, 20 if bold else 30, 20 if bold else 30)
if indent:
pdf.set_x(pdf.l_margin + indent)
pdf.multi_cell(pdf.w - pdf.l_margin - pdf.r_margin - indent, h, _l1(text))
except Exception:
pass
for raw_line in content_md.split("\n"):
line = raw_line.rstrip()
if line.startswith("# "):
pdf.set_font("Helvetica", "B", 14)
pdf.set_text_color(20, 20, 20)
pdf.multi_cell(0, 7, line[2:])
_cell(line[2:], size=14, bold=True, h=7)
pdf.ln(1)
elif line.startswith("## "):
pdf.set_font("Helvetica", "B", 12)
pdf.set_text_color(40, 40, 40)
pdf.multi_cell(0, 6, line[3:])
_cell(line[3:], size=12, bold=True, h=6)
pdf.ln(1)
elif line.startswith("### "):
pdf.set_font("Helvetica", "B", 11)
pdf.set_text_color(60, 60, 60)
pdf.multi_cell(0, 6, line[4:])
_cell(line[4:], size=11, bold=True, h=6)
elif line.startswith(("- ", "* ", "+ ")):
pdf.set_font("Helvetica", "", 10)
pdf.set_text_color(30, 30, 30)
pdf.multi_cell(0, 5, " • " + line[2:])
_cell("• " + line[2:], indent=5)
elif re.match(r"^\d+\. ", line):
pdf.set_font("Helvetica", "", 10)
pdf.set_text_color(30, 30, 30)
pdf.multi_cell(0, 5, " " + line)
elif line == "" or line == "---":
_cell(line, indent=5)
elif line in ("", "---"):
pdf.ln(3)
else:
# Strip inline markdown markers for bold/italic
clean = re.sub(r"\*\*(.+?)\*\*", r"\1", line)
clean = re.sub(r"\*(.+?)\*", r"\1", clean)
clean = re.sub(r"`(.+?)`", r"\1", clean)
pdf.set_font("Helvetica", "", 10)
pdf.set_text_color(30, 30, 30)
pdf.multi_cell(0, 5, clean)
_cell(clean)
return bytes(pdf.output())
@@ -775,8 +778,9 @@ def _process_analysis_job(job_id: int):
try:
pdf_bytes = _md_to_pdf(step_title, content)
attachments.append((f"{fname_base}.pdf", pdf_bytes, "application", "pdf"))
except Exception:
pass
except Exception as pdf_err:
import logging
logging.warning(f"PDF-Generierung fehlgeschlagen für '{step_title}': {pdf_err}")
_send_email(email_to, subject, answer, attachments)
except Exception as mail_err:
_job_set(job_id, error=f"[E-Mail-Fehler] {mail_err}")