Is there any way to copy and replace texts between two pdfs?

[WorstCodeWay]

Hi, guys. I don't know if this can be done with PyMuPDF. I want to copy one area texts(only texts within given rectangle area) within A.pdf and replace these texts into specific area on another B.pdf, and adapt the format of text exactly as B.pdf format, and without changing watermarks in B.pdf.

In my case, the scenario is limited to receipt-type PDFs, which have a fixed format. I want extract texts within green box(in A.pdf) and replace texts within red box(in B.pdf) with those texts showing in below image. Leaving content of B.pdf appears as untouched as possible, and the watermark remains intact(in out.pdf), referring to blue box(editing manually using commercial software WSP)

[JorjMcKie]

You can do this:

1. Identify the rectangle coordinates of Reprinted (green)
2. Identify the Cash Invoice rectangle (red).
3. Empty the area of Cash Invoice
4. Empty the area outside the green rectangle
5. Copy green to red area content.

All of the above points 3 to five can be done using PyMuPDF. Rectangle (green, red) identification is up to you, but I assume this won't be difficult: you mentioned a "fixed layout".

What will be the result of the above:
The former red area will look exactly like the green area. The green area content will automatically be scaled (up or down) to fit, so you have no worries in this respect.

If you can let us have 2 PDF examples (left / right) we could make a demo script.

[WorstCodeWay]

@JorjMcKie Thanks for your quick reply, and I have tried as your steps, they work, but the layout is not same as the source exactly, somehow messed up.

If you can let us have 2 PDF examples (left / right) we could make a demo script.

Sure, I have upload two examples pdf, feel free to use them.

A.pdf
B.pdf

[WorstCodeWay]

If you can let us have 2 PDF examples (left / right) we could make a demo script.

@JorjMcKie Hi, could you please show me your demo script? I think I have problem to get my things work.

[JorjMcKie]

The following is about the best we can do. Turns out that the watermarks cannot be maintained, but erased, respectively copied over too.
As I wrote before, regenerating the text in the target document is a hopeless undertaking. The only way we can think of is this.

import pymupdf

doca = pymupdf.open("A.pdf")
docb = pymupdf.open("B.pdf")
pagea = doca[0]
# find "Reprinted" rectangle on source page
r0 = pagea.search_for("Reprinted")[0]
r1 = pagea.search_for("FORM OF PAYMENT")[0]
r0.x0 = 0
r0.x1 = pagea.rect.x1
r0 |= r1.tl - 5
recta = +r0
# remove everything outside the rectangle (temporarily)
pagea.clip_to_rect(recta)

pageb = docb[0]
# find target rectangle on B.pdf
# we cannot use the same rectangle because they have different dimensions
r0 = pageb.search_for("VAT REG")[0]
r1 = pageb.search_for("FORM OF PAYMENT")[0]
bottom = r1.y0 - 5
r1.y0 = r0.y1 + 10
r1.y1 = bottom
r1.x0 = 0
r1.x1 = pageb.rect.x1
rectb = +r1
# empty target rectangle
pageb.add_redact_annot(rectb)
pageb.apply_redactions()
# copy source rectangle content
# we do not keep aspect ratio to fill the complete target rectangle
pageb.show_pdf_page(rectb, doca, 0, clip=recta, keep_proportion=False)
docb.ez_save("output.pdf")

[JorjMcKie]

Ah, here is an even better way: turns out that the watermark in the target can be retained and overlaid with the "Reprint" text.

import pymupdf

doca = pymupdf.open("A.pdf")
docb = pymupdf.open("B.pdf")
pagea = doca[0]
# find "Reprinted" rectangle on source page
r0 = pagea.search_for("Reprinted")[0]
r1 = pagea.search_for("FORM OF PAYMENT")[0]
r0.x0 = 0
r0.x1 = pagea.rect.x1
r0 |= r1.tl - 5
recta = +r0
# remove everything outside the rectangle (temporarily)
pagea.clip_to_rect(recta)
pagea.add_redact_annot(recta)
pagea.apply_redactions(  # remove images only
    images=pymupdf.PDF_REDACT_IMAGE_REMOVE,
    graphics=pymupdf.PDF_REDACT_LINE_ART_NONE,
    text=pymupdf.PDF_REDACT_TEXT_NONE,
)

pageb = docb[0]
# find target rectangle on B.pdf
# we cannot use the same rectangle because they have different dimensions
r0 = pageb.search_for("VAT REG")[0]
r1 = pageb.search_for("FORM OF PAYMENT")[0]
bottom = r1.y0 - 5
r1.y0 = r0.y1 + 10
r1.y1 = bottom
r1.x0 = 0
r1.x1 = pageb.rect.x1
rectb = +r1
# empty target rectangle
pageb.add_redact_annot(rectb)
pageb.apply_redactions(  # remove text only
    images=pymupdf.PDF_REDACT_IMAGE_NONE,
    graphics=pymupdf.PDF_REDACT_LINE_ART_NONE,
    text=pymupdf.PDF_REDACT_TEXT_REMOVE,
)
# copy source rectangle content
# we do not keep aspect ratio to fill the complete target rectangle
pageb.show_pdf_page(rectb, doca, 0, clip=recta, keep_proportion=True)
docb.ez_save("output.pdf")
``