Function apply_specific_fixes

scripts/postprocess_fulltext.py:593–696

Apply document-specific fixes to clean up severe OCR and layout conversion issues seen in certain MDR and MDCG guidance PDFs (e.g. 2019-16).

(content: str)

    return content, n1 + n2

def apply_specific_fixes(content: str) -> tuple[str, int]:
    """
    Apply document-specific fixes to clean up severe OCR and layout conversion issues
    seen in certain MDR and MDCG guidance PDFs (e.g. 2019-16).
    """
    modified = 0
    orig = content
    
    # 1. Figure 1 diagram textual elements in MDCG 2019-16
    def repl_fig1(m):
        images = re.findall(r'(?:<!-- image -->|!\[.*?\]\(.*?\))', m.group(0))
        return '\n\n'.join(images) + '\n\n'
        
    content = re.sub(
        r'General safety and performance requirements with focus on cybersecurity.*?1nт 1\. Cahraanetar enmmlenmonta\n+',
        repl_fig1,
        content,
        flags=re.DOTALL
    )
    
    # 2. Fix the "Table 1: I" orphan header
    content = re.sub(
        r'^Table 1:\s*I\n+Correspondence table between sections, relevant for this guidance, in MDR Annex I and IVDR Annex',
        'Table 1: Correspondence table between sections, relevant for this guidance, in MDR Annex I and IVDR Annex I',
        content,
        flags=re.MULTILINE
    )
    
    # 3. Numbered lists with weird bullets e.g. "9. • A description..."
    content = re.sub(
        r'^(\d+\.)\s*(?:-\s*|[•➢o]\s+)+(.*)$',
        r'\1 \2',
        content,
        flags=re.MULTILINE
    )
    
    # Missing GDPR in MDCG-2019-16 abbreviations
    content = content.replace(
        "Field Safety Corrective actions\n\n\nGeneral Data Protection Regulation",
        "Field Safety Corrective actions\n\nGDPR\n\nGeneral Data Protection Regulation"
    )
    # the extra hyphen before IEC/TR
    content = content.replace("\n- IEC/TR\n", "\nIEC/TR\n")
    
    # 4. Turn spaced abbreviation items into table
    abbr_block_match = re.search(r'(### 1\.5\.\s*Abbreviations\n+)(.*?)(?=\n\||\n#)', content, re.DOTALL)
    if not abbr_block_match:
        abbr_block_match = re.search(r'(## 1\.5\.\s*Abbreviations\n+)(.*?)(?=\n\||\n#)', content, re.DOTALL)
        
    if abbr_block_match:
        prefix = abbr_block_match.group(1)
        block = abbr_block_match.group(2)
        lines = [l.strip() for l in block.split('\n') if l.strip()]
        out_table = ["| Abbreviation | Meaning |", "| --- | --- |"]
        
        # Only do this if it looks like alternating items without pipes
        if all('|' not in l for l in lines) and len(lines) % 2 == 0 and len(lines) > 2:
            for i in range(0, len(lines)-1, 2):