Hey there! Let's work through your PDF generation issue together. It sounds like you're hitting output errors when using formatted strings from a database with FPDF, and you're wondering if fonts are the culprit—plus you need solid multi-language font recommendations. Let's break this down:

First, let’s narrow down why you might be seeing errors:

• Font Path Issues: Double-check that the path /data/mlo/fonts/DejaVuSansCondensed.ttf is correct. If FPDF can’t find the font file, it’ll fail to render text properly (or throw an error outright).
• Formatted String Conflicts: If your database strings include special formatting (like HTML tags, custom line breaks, or escape characters), FPDF might not parse them as expected. For example, FPDF doesn’t natively understand <br> tags—you’ll need to convert those to \n manually.
• Incomplete Unicode Support: Even with uni=True, DejaVu Sans Condensed has limited coverage for Asian languages. If your database text includes Chinese, Japanese, Korean, or other non-Latin scripts, the font might not have those glyphs, leading to missing characters or squares.

Here are some free, open-source fonts that handle both Asian and Latin languages flawlessly:

• Noto Sans Family: Google’s go-to for universal language support. Grab specific variants like Noto Sans SC (Simplified Chinese), Noto Sans JP (Japanese), or Noto Sans KR (Korean)—all include full Latin character sets too.
• Source Han Sans: Developed by Adobe and Google, this font is optimized for East Asian languages (CN/JP/KR) and pairs perfectly with Latin text. It’s great for professional-looking reports.
• WenQuanYi Zen Hei: A lightweight, reliable option with full GBK Chinese character support and natural Latin language rendering.
• DejaVu Sans Extended: If you want to stick with DejaVu, this extended version supports more Unicode characters than the condensed variant—though it still doesn’t cover all Asian scripts as well as the options above.

Here’s an updated version of your code with checks for font existence, better text handling, and multi-language font support:

from fpdf import FPDF
import os

def create_pdf(filename, text):
    pdf = FPDF()
    pdf.add_page()
    # Add margin to auto-page-break for better readability
    pdf.set_auto_page_break(auto=True, margin=15)
    
    # Verify font file exists first to avoid path errors
    font_path = '/data/mlo/fonts/NotoSansSC-Regular.ttf'
    if not os.path.isfile(font_path):
        raise FileNotFoundError(f"Font file missing at: {font_path}")
    
    # Load multi-language font with Unicode support
    pdf.add_font('NotoSans', '', font_path, uni=True)
    pdf.set_font('NotoSans', '', 10)
    
    # Clean up formatted text from database: convert custom line breaks to FPDF-compatible ones
    # Replace <br> tags or database-specific newlines with \n
    cleaned_text = text.replace('<br>', '\n').replace('\r\n', '\n')
    
    # Use multi_cell to handle automatic line wrapping for long text
    pdf.multi_cell(w=0, h=5, txt=cleaned_text)
    
    pdf.output(filename)

• Test with a simple multi-language string first (e.g., "Hello 你好 こんにちは 안녕하세요") to confirm the font renders all scripts correctly.
• Ensure your database strings are encoded in UTF-8—FPDF requires UTF-8 text for Unicode support.
• If you’re using complex formatting (like bold or italics), load the bold/italic variant of your font and switch between styles in your code.

内容的提问来源于stack exchange，提问作者mlo0424