Improper placement of inserted objects on PDF pages

[thaotonto]

Page.insertImage and Page.showPDFpage inserted a image or pdf into pdf page but the image or pdf is not inserted at the right location (inside rect). Image or pdf is auto resized and displayed at the bottom left of the page

[JorjMcKie]

Very probably your PDF contains improper changes of the page geometry.
This occurs when eg. the page orientation is reversed (point (0,0) is no longer the PDF standard bottom-left) and that geometry change is not properly localized in the string pair q / Q and thus remains active for ever.
Any program (like PyMuPDF) adding more commands to the page's /Contents will then unknowingly be subject to that setting.

Try to repair that by cleaning the file as a whole (using the cli mutool utility) or in you script on a per page level using page._cleanContents() before inserting something.
In rare occasions this action may lead to other problems, however. If so, please tell me and I will show you a way out. It basically works by prefixing and appending thos two commands to the page's /Contents object.

[thaotonto]

wow after added page. _cleanContents() it worked. Thanks you very much !!!

[JorjMcKie]

glad to hear that!
Have fun with the package.

[JorjMcKie]

This or a similar issue is popping up once in a while.
And the solution is not rocket science. So I am wondering what can be done to point to it more visibly ...

A radical solution may of course be checking the presence of this q/Q encapsulation every time before inserting something.

Have an opinion?

[thaotonto]

It would be great if there are some note about this problem in the document. I spent quite a while to read the doc but still couldn't figure it out and have to go up here asked for help.

[JorjMcKie]

Ok thanks.
You are right - it should be in the doc. It actually is, but I am afraid just not prominently enough.

[JorjMcKie]

Just published a new docu version which hopefully better addresses your issue.

[thaotonto]

thank you for the information. Can i ask you one more thing: is there any function that insert the image but can still move the image around like annotation object ?

[JorjMcKie]

Interesting question! No, not yet. One reason is the PDF concept itself:

Annotations are designed to be "temporary dust" on the surface of a page image. Therefore their position and other characteristics remain accessible for inspection, change or deletion. This is also reflected by the fact, that there is a separate permission category for annotations: you might be authorized to add or change them, but not the document itself.

Technically, changing locations of images is very probably never ever possible. Deleting and then re-inserting one might be achievable.

Looks a little better for things inserted via showPDFpage because those are stored with more explicit location info.

[thaotonto]

thank you for the answer. But i noticed that there are some tool that let user insert image on top of pdf's page but can still move it around freely like PDF-XChange Editor. I would hope to find a python library that can do the same as this tool.

[JorjMcKie]

This may be an approach to change an image location ... doable even today:

• Find out the image's reference name. Document.getPageImageList(pno) items each describe one image on a page. Example

>>> doc.getPageImageList(0)
[[316, 0, 261, 115, 8, 'DeviceRGB', '', 'Im1', 'DCTDecode']]

• Read (all of the) /Contents object(s) and search for b" /Im1 Do " (remember that multiple spaces are treated as one and may be interchanged with line breaks). Delete this string where found. Rewrite the contents object(s). Now the image will no longer be shown on that page (but physically it is still there).

• Extract the image stored under xref=316 and insert it at a new location.

This procedure generates a few left over orphans of information, which should be cleaned up on save using the garbage option.

[JorjMcKie]

thank you for the answer. But i noticed that there are some tool that let user insert image on top of pdf's page but can still move it around freely like PDF-XChange Editor. I would hope to find a python library that can do the same as this tool.

Does this work for arbitrary PDF images, ie. those not inserted by that tool?
Does this work after you have left the tool session and started over with again opening the PDF from before?