script.py

import fitz

INPUT_PDF = "input.pdf"
OUTPUT_PDF = "output.pdf"

REPLACEMENTS = [
    {"old_text": "11/10/2025", "new_text": "22/10/2025", "max_replace": 1},
    {"old_text": "10/10/2025", "new_text": "22/10/2025", "max_replace": 1},
]

BUILTIN_FONTS = {"helv", "cour", "tiro", "times", "symbol", "zapfdingbats"}

FONT_ALIASES = {
    "Arial": "helv",
    "Calibri": "helv",
    "Helvetica": "helv",
    "Times-Roman": "times",
    "TimesNewRoman": "times",
    "Courier": "cour",
}


def sanitize_font_name(font_name: str) -> str:
    if not font_name:
        return "helv"

    clean_name = font_name.split("+")[-1]

    for alias, builtin in FONT_ALIASES.items():
        if clean_name.lower().startswith(alias.lower()):
            return builtin

    if clean_name not in BUILTIN_FONTS:
        return "helv"

    return clean_name


def find_font_and_size(page, target_text: str) -> tuple[str, float]:
    blocks = page.get_text("dict")["blocks"]
    for block in blocks:
        for line in block.get("lines", []):
            for span in line.get("spans", []):
                if target_text in span["text"]:
                    font = sanitize_font_name(span["font"])
                    return font, span["size"]
    return "helv", 10  # fallback


def replace_text_in_page(page, old_text: str, new_text: str, max_replace=None):
    matches = page.search_for(old_text)
    if not matches:
        return 0, None, None

    if max_replace is not None:
        matches = matches[:max_replace]

    font_name, font_size = find_font_and_size(page, old_text)

    for rect in matches:
        page.add_redact_annot(rect, fill=(1, 1, 1))
    page.apply_redactions()

    for rect in matches:
        x, y = rect.x0, rect.y1 - 2
        page.insert_text(
            (x, y),
            new_text,
            fontname=font_name,
            fontsize=font_size,
            color=(0, 0, 0),
        )

    return len(matches), font_name, font_size


def process_pdf(input_path: str, output_path: str, replacements: list[dict]):
    doc = fitz.open(input_path)

    for page_num, page in enumerate(doc, start=1):
        print(f"🔹 Processing page {page_num}...")

        for rep in replacements:
            old_text = rep["old_text"]
            new_text = rep["new_text"]
            max_replace = rep.get("max_replace")

            count, font, size = replace_text_in_page(page, old_text, new_text, max_replace)
            if count > 0:
                print(f"   ✅ '{old_text}' → '{new_text}' ({count}x, font={font}, size={size})")

    doc.save(output_path)
    doc.close()
    print(f"\n🎉 Done! Saved edited PDF → {output_path}")


if __name__ == "__main__":
    process_pdf(INPUT_PDF, OUTPUT_PDF, REPLACEMENTS)