PDF cannot open in Adobe InDesign

[Kishanpatel96]

I am using PyMuPDF for writing a text on PDF. I am successfully able to write text on PDF with this library. If I open PDF normally or Other Adobe software then It is working fine.

Describe the bug

If I open an output PDF file in Adobe InDesign then it is giving an error that PDF cannot Open.
If I will not write text and save PDF file using PyMuPDF then it can easly open in Adobe InDesign

To Reproduce

1. I have tried with insertTextbox method.

pdf.insertTextbox(position,
                word,
                align=alignment,
                fontname=fontname,
                fontfile=slot_data['fontFile'],
                fontsize=slot_data['fontSize'],
                rotate=0,
                color=RGBColor,
                )

pdf_document.save(output_file,garbage=3, deflate=True, pretty=True)

Expected behavior

I want to open a PDF file the normal way in Adobe InDesign.
Can you tell me it is possible or not? If possible then can you guide me on how I can do it?

Thank you,
Kishan Patel

[JorjMcKie]

Problem not reproducible like that.
Please provide the full script or confirm that your insert_textbox() indeed uses the correct call pattern:
position = a rectangle?
word = a string or list of strings
etc.
Which fonts are you using?
Also an example PDF.

[JorjMcKie]

Ah and of course the rest of mandatory information: your environment and PyMuPDF version.

[Kishanpatel96]

Thank you for your fast reply.

PyMuPDF version is 1.18.6.
Python version is 3.6
Flask version is 1.0.2

position = fitz.Rect(50,200,150,300)
        print_siegel_service.siegel_file.insert_textbox(position,
        'Hello',
        align=0,
        fontname="F1",
        fontfile="font/Nobel-Regular.otf",
        fontsize=15,
        rotate=0,
        color=(0,0,0),
        )

Yes as you told I have tried with insert_textbox() still PDF is not opening in Adobe InDesign. For example, can use any PDF.

[JorjMcKie]

operating system?

[Kishanpatel96]

I am using macOS Mojave 10.14.6 Operating System.
My Project is running on Docker and the version of Docker is 2.3

I hope this is enough information.

[JorjMcKie]

How do you save the modified PDF: garbage options?, deflate=True?, page.clean_contents() after text insertion ...

Adobe InDesign is not a free product - it costs money - and I am not using it, nor will I buy it.
There is a grace period of 7 days after downloading a test version. So the issue must be resolved within a week ...

So before using up this precious timeframe:
Can you confirm, that a PDF with your inserted text causes no problems when opening it in Adobe Acrobat and all other PDF viewers you may be using? Outside docker, on your desktop?

[Kishanpatel96]

file_path = "filename.pdf"
pdf_document = fitz.open(file_path)
page = pdf_document[0]

position = fitz.Rect(50,200,150,300)
page.insert_textbox(position,
'Hello',
align=0,
fontname="F1",
fontfile="font/Nobel-Regular.otf",
fontsize=15,
rotate=0,
color=(0,0,0),
)
page.clean_contents()

pdf_document.save('outputfile.pdf', garbage=2, deflate=True, pretty=True)

I am using this code to write text on PDF. I have tried garbage values from 1 to 4 but the output is the same (Cannot open PDF in Adobe InDesign).

Yes, I am able to open a PDF file in Adobe Acrobat and Adobe illustrate also in google Crome and normal PDF viewers.
But the problem is coming with Adobe InDesign.

If you have other code suggestions then also you can tell me I will try here and check also.

Yes, I am also want to solve this issue soon.

[JorjMcKie]

Your code looks good! And thanks for the accompanying info.
Ok, I will now go ahead and check with InDesign.

[Kishanpatel96]

Okay, I am really waiting for your answer.
Please make sure you have place pdf (Not Open) in Adobe InDesign.

I hope we can solve this issue soon.

[JorjMcKie]

Installed Adobe InDesign.
It does not open any PDF! What is wrong?

[Kishanpatel96]

Do you mean printed PDF from PyMuPDF or normal PDF?

And one more time.
You don't have to open PDF in Adobe InDesign.

You have to select Place from the menu.
For Reference, you can use this link.

[JorjMcKie]

It works:
Took an existing PDF and did this:

>>> doc=fitz.open("v110-changes.pdf")
>>> page=doc[0]
>>> text = "** test ** test ** test ** test ** test ** test **"
>>> page.insert_textbox(page.rect, text)
830.0949823856354
>>> doc.save("x.pdf")
>>> 

Then placed x.pdf in an indesign document:

Same thing with a PDF create new in PyMuPDF - no issues.

[Kishanpatel96]

Can you tell me your full configuration? What version you are using for all software including InDesign?

[JorjMcKie]

Windows 10 64bit, Python 3.9 64bit, InDesign 16.1, PyMuPDF v1.18.8

By the way - as soon as I use a fontfile-based font, InDesign refuses to open the PDF.
Choosing one of the Base-14 font (Helvetica, Times-Roman or Courier and their respective font weights) all works fine.

[Kishanpatel96]

Yes, you are right I am also able to open PDF in InDesign if I am using a font from Base-14 fonts.

Can you give me any idea why it is not working with fontfile-based font?
And it could be possible to solve with fontfile or something?

Because I want to use fontfile.

also can you tell me that is there a way to vectorize our font before export it or save file?

[JorjMcKie]

The problem is larger than around fontfile. And it is InDesign-specific.
Even if a convert a previously accepted PDF to a new PDF using mutool draw -o test.pdf v110-changes.pdf then the new file test.pdf will again not be accepted.

So in summary, InDesign only accepts some subset of otherwise perfectly valid PDFs -- PDFs that can be opened in Adobe Acrobat and all other PDF viewers without restrictions

I have no idea how this comes about and probably must be addressed / interrogated by InDesign. For you immediate needs you may want to convert your modified PDF to some image file -- page by page -- and them import these.
I see no way to help.
And it definitely does not look like a PyMuPDF or MuPDF error.

[Kishanpatel96]

Yes, I got your point but Can you give me a way to make my font vectorized while writing it on PDF or on Saving time?

[JorjMcKie]

font vectorized

Don't understand that.

You could convert a PDF page to an SVG image and then import this one. This is a vector format and by defaut font-independent.

[Kishanpatel96]

So for your clear understanding.
I want to write text on PDF. And If I write text as we are doing currently then it can change or edit with Software like Adobe or and others.

So I want to make functionality like the text written by PyMuPDF will be not changed by any software.
So with making the font vectorized it become Shape and it will be not changeable.

So is there any way to make text not editable by other software?

[JorjMcKie]

So is there any way to make text not editable by other software?

Well, you can always password-protect a PDF and encrypt it such, that it can be read, but not changed by anyone without the password.

What you meant amounts to not writing the text by using a font. Instead paint each character by elementary PDF drawing commands like a capital letter "D" is painted as a closed semi-circle.
This is not (directly) supported by MuPDF. But the default SVG output (text_as_path=True) does exactly that as explained above.

[JorjMcKie]

BTW: the correct term is "text as path". Text written using a font, by definition is vectorized.

[JorjMcKie]

I want to write text on PDF. And If I write text as we are doing currently then it can change or edit with Software like Adobe or and others.

Changing text no matter what you do is always possible. Even when written as paths.
Use an OCR software and re-insert the interpreted, but modified text.

[Kishanpatel96]

I feel Well, you can always password-protect a PDF and encrypt it such, that it can be read, but not changed by anyone without the password. this is a good way to go and I want to try it.

Can you give me an example code for this feature? Because we have permission, user_pw, and owner_pw
So I will example code will be good.

[JorjMcKie]

Use an owner password and encrypt the file with a method of your liking.
If you omit the user password, anyone can open the document without entering a password and will automatically be identified as a user - otherwise opening requires a password.
After opening, the actions allowed are controlled by the permissions integer - which is a bit-field actually.
To configure the permissions do this

perm = fitz.PDF_PERM_ACCESSIBILITY  # always start with this
perm |= fitz.PDF_PERM_PRINT  # then OR additional allowed actions into it
perm |= ...
doc.save("encrypted.pdf", owner_pw="owner", user_pw="user", permissions=perm, encryption=fitz.PDF_ENCRYPT_AES_256)

The maximum password length is 50 bytes.
When opening requires a password, the entered password value will automatically identify you as owner or user (or even both if the passwords are equal).

[Kishanpatel96]

Hello, I need your small suggestion.
As I want to use font outside of Base 14 font of pdf.

I embedded this fond file inside our library then I want to access it by name in insert_textbox method.

https://pymupdf.readthedocs.io/en/latest/font.html

currently, I am not able to use that font in insert_textbox. Can you give me a code snippet for it?

[JorjMcKie]

The Font objects are (primarily) intended for use with TextWriter objects.
Both are relatively new, whereas insert_text and insert_textbox exist since a long time.
So, admittedly we have some duplication of functionality here ... but allow the author to also walk through a learning curve 😎.

Anyway, it is still possible to use fitz.Font objects with the "old" text insertions. This works using the buffer of a fitz.Font:

>>> import fitz
>>> font=fitz.Font("figo")  # any supported font, incl. font files or font buffers
>>> doc=fitz.open()
>>> page=doc.new_page()
>>> page.insert_font(fontname="F1",fontbuffer=font.buffer)  # insert font in page, return its xref
5
>>> page.insert_text((100,100),"some text",fontname="F1")  # refer to font via F1
1
>>> doc.subset_fonts()  # just for fun: subset embedded font to the used characters
Subset built for 'FiraGO-Regular'.
773256
>>> doc.save("x.pdf",garbage=3,deflate=True)
>>> 

The buffer keyword is not supported in insert_textbox/insert_text, so you have to first insert the font separately in the page.
It is important not using one of the Base-14 reserved names or their abbreviations in this case - instead use generic names like indicated. This is true even if you have created the Font object e.g. like fitz.Font("helv").