Is it possible to rewrite such pdf to a correct font, so that the text shown matches the text selected?

[nissansz]

1.pdf

Is it possible to rewrite such pdf to a correct font, so that the text shown matches the text selected?
Only a thin line is selected now.

[JorjMcKie]

This is more a weakness of Adobe's viewer than one of the font: other viewers display it nicer, e.g. Foxit

and others.
Rewriting text with a different font is possible (see recipes here) but remains to be problematic because you cannot re-establish the original appearance sequence, potentially Optional Content layers etc.

The font does have a weird bbox definition causing this effect.
But PyMuPDF allows you to ignore it by choosing the right extraction flags. Standard TEXTFLAGS_TEXT will indeed deliver bboxes with a small height:

But if you use this as flags: flags=pymupdf.TEXTFLAGS_TEXT | pymupdf.TEXT_ACCURATE_BBOXES you will get exactly fitting boundary boxes:

[nissansz]

I just want to use acrobat to convert pdf to pptx, but original pdf failed to export good pptx.
Any solution?

[nissansz]

Understood. Thank you.

[nissansz]

There are some strange font heights after conversion. How to improve?

（正式版）GB∕T 14848-2017 地下水质量标准.pdf
（正式版）GB∕T 14848-2017 地下水质量标准_1.pdf

[nissansz]

How to change to word-level dict to rewrite, also adding judgement for English / CJK languages to use different fonts?

[robinwatts]

Please, just ask one question per discussion item. If you ask multiple, unrelated questions, it's just confusing to know what has been answered and what has not.

[nissansz]

Ok. Will close this discussion.
I tried below, no improvement, so maybe it is problem of original pdf.

fontname = "cjk"
if is_english_sentence(s["text"]):
fontname = "helv"

[nissansz]

1output.pdf
I got above result, no text is shown, how to correct below code?

import fitz  # PyMuPDF
import sys
import os

INPUT_PDF = r"L:\1.pdf"
OUTPUT_PDF = r"F:\1output.pdf"
FONT_PATH = r"C:\Windows\Fonts\simhei.ttf"  # Make sure this font file exists in your working directory, or use an absolute path


def redact_all_text(page, dict_out):
    """Redact all text spans on a page using rectangles from the dict output."""
    for block in dict_out["blocks"]:
        if block["type"] != 0:
            continue
        for line in block["lines"]:
            for span in line["spans"]:
                r = fitz.Rect(span["bbox"])
                # Add a redaction annotation for the span
                page.add_redact_annot(r, fill=(1, 1, 1))  # white fill


def write_spans(npage, blocks, fontname, fontsize_add=0):
    """Write each span back with a better font."""
    helv = fitz.Font("helv")
    cjk = fitz.Font("cjk")

    # arial = fitz.Font(fontfile="C:/Windows/Fonts/arial.ttf")
    arial = fitz.Font(fontfile="F:/fonts/Arial Unicode MS.ttf")

    # blocks = page.get_text("rawdict")["blocks"]
    helv = fitz.Font("helv")
    cjk = fitz.Font("cjk")
    # arial = fitz.Font(fontfile="C:/Windows/Fonts/arial.ttf")
    arial = fitz.Font(fontfile="F:/fonts/Arial Unicode MS.ttf")
    for b in blocks:
        if "lines" not in b:
            continue

        for l in b["lines"]:
            print(f'l={l}')
            cos, sin = l["dir"]
            matrix = fitz.Matrix(cos, -sin, sin, cos, 0, 0)
            for s in l["spans"]:
                textBox = s['bbox']
                fsize = s["size"]
                fname = s["font"]
                for c in s["chars"]:
                    # if fname.lower().startswith("arial"):
                    #     font = arial
                    # else:
                    #     font = helv

                    font = cjk
                    ch = c["c"]
                    origin = fitz.Point(c["origin"])
                    tw = fitz.TextWriter(npage.rect)
                    tw.append(origin, ch, font=font, fontsize=fsize)
                    tw.write_text(npage, morph=(origin, matrix))

            shape = npage.new_shape()
            shape.draw_rect(textBox)
            shape.finish(
                fill=None,  # fill color
                color=(0, 0, 1),  # line color
            )
            shape.commit()


def main(input_pdf, output_pdf, font_path):
    doc = fitz.open(input_pdf)
    # Register the font under a custom name
    # fontname = "MyBetterFont"
    # fitz.Font(fontname=fontname, fontfile=font_path, set_simple=True)
    # fitz.Font(fontname="HT", fontfile=r"C:\Windows\Fonts\simhei.ttf", fontbuffer=None, set_simple=False)

    new_doc = fitz.open()  # output PDF

    for i, page in enumerate(doc):
        ff = page.insert_font(fontname="HT", fontfile=r"C:\Windows\Fonts\simhei.ttf", fontbuffer=None, set_simple=False)

        dict_out = page.get_text("dict")
        blocks = page.get_text("rawdict")["blocks"]
        # Redact all text
        redact_all_text(page, dict_out)
        page.apply_redactions(images=fitz.PDF_REDACT_IMAGE_NONE)  # keep images
        # Copy the redacted page to new PDF
        new_page = new_doc.new_page(width=page.rect.width, height=page.rect.height)
        new_page.show_pdf_page(page.rect, doc, i)
        # Now write the text spans
        write_spans(new_page, blocks, fontname="HT", fontsize_add=0)

    new_doc.save(output_pdf)
    print(f"Done. Output saved to {output_pdf}")


if __name__ == "__main__":
    # Usage: python rewrite_pdf_font.py input.pdf output.pdf
    import argparse

    parser = argparse.ArgumentParser()
    parser.add_argument("input_pdf", nargs="?", default=INPUT_PDF)
    parser.add_argument("output_pdf", nargs="?", default=OUTPUT_PDF)
    parser.add_argument("--font", default=FONT_PATH, help="Path to TTF font file")
    args = parser.parse_args()

    if not os.path.exists(args.input_pdf):
        print("Input PDF not found!")
        sys.exit(1)
    if not os.path.exists(args.font):
        print("Font file not found!")
        sys.exit(1)
    main(args.input_pdf, args.output_pdf, args.font)