How can I split a page in 4 smaller pages mosaicing?

[ErfolgreichCharismatisch]

edit: The following was originally reported as a bug. While it is not, some of the comments might be valuable to the community.

My goal:

An A4 PDF file has two columns. Read it with Python and output a new A5 PDF file where the width of the column corresponds to the width of the A5 page (enlarge/reduce if necessary) and copy the complete content into a new A5 PDF. If you do it right, there will be more pages in the A5 PDF file than in the A4 PDF file. Also take over chapters.

My code:

import PyPDF2
from reportlab.pdfgen import canvas
from reportlab.lib.units import mm
with open(r'c.pdf', 'rb') as a4_file:
    a4_reader = PyPDF2.PdfReader(a4_file)
    num_pages = len(a4_reader.pages)
    a5_width = 148 * mm
    a5_height = 210 * mm
    a5_canvas = canvas.Canvas(r'v.pdf', pagesize=(a5_width, a5_height))
    bookmarks = []
    for i in range(num_pages):
        a4_page = a4_reader.pages[i]
        a4_width = a4_page.mediabox.width
        a4_height = a4_page.mediabox.height
        column_width = a4_width / 2
        width_scale = a5_width / float(column_width)
        height_scale = a5_height / a4_height
        scale = min(width_scale, height_scale)
        a4_page.scale(scale, scale)
        
        column_positions = [0, a5_width / 2] * ((num_pages + 1) // 2)
        if num_pages % 2:
            column_positions.append(a5_width / 2)
        
        
        a5_canvas.saveState()
        a5_canvas.translate(column_positions[i], 0)
        a5_canvas.scale(scale, scale)
        a5_canvas.doForm(a4_page)
        a5_canvas.restoreState()
        
        if len(a4_page.get_contents().get_object()) > 0:
            bookmarks.append((a4_page.get_contents().get_object()['/Title'], i))
    for title, page_index in bookmarks:
        a5_canvas.addOutlineEntry(title, str(page_index), level=0, closed=None)
    a5_canvas.save()

My goal in pictures:

A4 with columns:

Goal:

The error I got

Traceback (most recent call last):File "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 877, in formatreturn pdfdocEnc("%s %s R" % document.idToObjectNumberAndVersion[self.name])KeyError: "FormXob.{'/CropBox': RectangleObject([0, 0, 420.80752613761768543554353527724742889404296875, 595.2755905511811533870059065520763397216796875]), '/Parent': IndirectObject(101, 0, 2653010734384), '/StructParents': 0, '/Contents': {}, '/Rotate': 0, '/MediaBox': RectangleObject([0, 0, 420.80752613761768543554353527724742889404296875, 595.2755905511811533870059065520763397216796875]), '/Resources': {'/ColorSpace': {'/CS0': IndirectObject(1030, 0, 2653010734384)}, '/Font': {'/TT0': IndirectObject(1027, 0, 2653010734384), '/TT1': IndirectObject(1028, 0, 2653010734384), '/TT2': IndirectObject(1029, 0, 2653010734384)}, '/ProcSet': ['/PDF', '/Text'], '/ExtGState': {'/GS0': IndirectObject(1033, 0, 2653010734384)}}, '/Type': '/Page', '/ArtBox': RectangleObject([0, 0, 297.50172046312587781358161009848117828369140625, 420.846827441873529096483252942562103271484375]), '/BleedBox': RectangleObject([0, 0, 297.50172046312587781358161009848117828369140625, 420.846827441873529096483252942562103271484375]), '/TrimBox': RectangleObject([0, 0, 297.50172046312587781358161009848117828369140625, 420.846827441873529096483252942562103271484375])}"



During handling of the above exception, another exception occurred:



Traceback (most recent call last):File "E:\Gesichert\Desktop\SpaltenPDF in A5PDF.py", line 77, in <module>a5_canvas.save()File "E:\anaconda\lib\site-packages\reportlab\pdfgen\canvas.py", line 1280, in saveself._doc.SaveToFile(self._filename, self)File "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 212, in SaveToFiledata = self.GetPDFData(canvas)File "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 238, in GetPDFDatareturn self.format()File "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 420, in formatIOf = IO.format(self)File "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 867, in formatfcontent = format(self.content, document, toplevel=1)   # yes this is at top levelFile "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 67, in formatf = element.format(document)File "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 1023, in formatreturn format(dict, document)File "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 67, in formatf = element.format(document)File "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 680, in formatL = [(format(PDFName(k),document)+b" "+format(dict[k],document)) for k in keys]File "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 680, in <listcomp>L = [(format(PDFName(k),document)+b" "+format(dict[k],document)) for k in keys]File "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 67, in formatf = element.format(document)File "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 1902, in formatreturn format(DD, document)File "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 67, in formatf = element.format(document)File "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 680, in formatL = [(format(PDFName(k),document)+b" "+format(dict[k],document)) for k in keys]File "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 680, in <listcomp>L = [(format(PDFName(k),document)+b" "+format(dict[k],document)) for k in keys]File "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 67, in formatf = element.format(document)File "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 680, in formatL = [(format(PDFName(k),document)+b" "+format(dict[k],document)) for k in keys]File "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 680, in <listcomp>L = [(format(PDFName(k),document)+b" "+format(dict[k],document)) for k in keys]File "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 67, in formatf = element.format(document)File "E:\anaconda\lib\site-packages\reportlab\pdfbase\pdfdoc.py", line 879, in formatraise KeyError("forward reference to %s not resolved upon final formatting" % repr(self.name))KeyError: 'forward reference to "FormXob.{'/CropBox': RectangleObject([0, 0, 420.80752613761768543554353527724742889404296875, 595.2755905511811533870059065520763397216796875]), '/Parent': IndirectObject(101, 0, 2653010734384), '/StructParents': 0, '/Contents': {}, '/Rotate': 0, '/MediaBox': RectangleObject([0, 0, 420.80752613761768543554353527724742889404296875, 595.2755905511811533870059065520763397216796875]), '/Resources': {'/ColorSpace': {'/CS0': IndirectObject(1030, 0, 2653010734384)}, '/Font': {'/TT0': IndirectObject(1027, 0, 2653010734384), '/TT1': IndirectObject(1028, 0, 2653010734384), '/TT2': IndirectObject(1029, 0, 2653010734384)}, '/ProcSet': ['/PDF', '/Text'], '/ExtGState': {'/GS0': IndirectObject(1033, 0, 2653010734384)}}, '/Type': '/Page', '/ArtBox': RectangleObject([0, 0, 297.50172046312587781358161009848117828369140625, 420.846827441873529096483252942562103271484375]), '/BleedBox': RectangleObject([0, 0, 297.50172046312587781358161009848117828369140625, 420.846827441873529096483252942562103271484375]), '/TrimBox': RectangleObject([0, 0, 297.50172046312587781358161009848117828369140625, 420.846827441873529096483252942562103271484375])}" not resolved upon final formatting'

[pubpub-zz]

first you are using PyPDF2 which is an old version, you have to upgrade to the latests pypdf 3.5.1 (you will have to uninstall PyPDF2 and then install pypdf)

as a test sample here is a code I've just tested:

import pypdf
r = pypdf.PdfReader("crazyones.pdf")
w = pypdf.PdfWriter()
#add the pages 4 times using append to get annots....
w.append(r,[0]);w.append(r,[0]);w.append(r,[0]);w.append(r,[0])
#readjust the mediabox
w.pages[0].mediabox.bottom = r.pages[0].mediabox.top - r.pages[0].mediabox.height/2
w.pages[0].mediabox.right = r.pages[0].mediabox.left + r.pages[0].mediabox.width/2
w.pages[1].mediabox.right = w.pages[0].mediabox.right
w.pages[1].mediabox.top = w.pages[0].mediabox.bottom
w.pages[2].mediabox.left = w.pages[0].mediabox.right
w.pages[2].mediabox.bottom = w.pages[0].mediabox.bottom
w.pages[3].mediabox.left = w.pages[2].mediabox.left
w.pages[3].mediabox.top = w.pages[2].mediabox.bottom
w.write("out.pdf")

[wwwald]

Asking for my own understanding: this code will separate out the 4 "subpages" as requested - but it will duplicate the same image data 4 times, right? So the resulting PDF would be 4 times as big?

Is there a way to actually delete the image content outside the MediaBox for these pages, and keep the resulting PDF as lean as possible?

[pubpub-zz]

The image analysis is a tough work that a viewer is doing. pypdf is not designed for this. The solution I would propose would be to "print" the file using a "pdf printer" driver into a new pdf.