# David learns how to create a pdf with python

## The project

The project is to automate to some extent creating pdfs with the photos of my dog Moncho.

They come also with some text

:::info
La foto de hoy es un modelo 3D de perrito que puedes descargar en la asset store por el módico precio de 5.99 pollitos. Aquí se puede ver en un fondo de prueba. Oferta disponible por tiempo limitado. #fotoMoncho
:::

This is an idea of the layout done in inkscape in 2 minutes


## pypdf

Let's try pypdf, I have 2 sources of information

The docs https://pypdf.readthedocs.io/en/stable/index.html

And this website, since is from 2025 is suspicious of be AI generated
https://realpython.com/creating-modifying-pdf/

### Read an existing pdf

I created a pdf with inkscape, so let's see if I can use it as a template.

I used the instructions from the docs. I actually didn't use the Pathlib since the pdf was in the same folder

from pypdf import PdfReader


pdf_path = "prueba maquetación inkscape.pdf"

pdf_reader = PdfReader(pdf_path)

print(len(pdf_reader.pages)) #1

Since it only has one page, then is only 1 page. But at least it _reads_ the pdf.

### Extract the text

Reference: https://realpython.com/creating-modifying-pdf/#extracting-text-from-a-page

Welp, this gets more complicated. The example was plain simple but it doesn't work

The example extracts the text simple and neat, all I get is a complex function like this:


And the mistake was that I didn't wrote properly the method with the end parenthesis (I wrote in the line 10: ` print(page.extractext)`)

from pypdf import PdfReader


pdf_path = "prueba maquetación inkscape.pdf"

pdf_reader = PdfReader(pdf_path)


for page in pdf_reader.pages:
    print(page.extract_text())

And the output:

06 jul 2025, 11:54
La foto de hoy es un modelo 3D de perrito que    
puedes descargar en la asset store por el módico 
precio de 5.99 pollitos. Aquí se puede ver en un 
fondo de prueba. Oferta disponible por tiempo    
limitado. #fotoMoncho

### Extract image from a pdf

:::warning
For this we need to also instat pillow with the command `pip install pypdf[image]`
:::

Reference: https://pypdf.readthedocs.io/en/stable/user/extract-images.html

With this snippet I can create a png with the Moncho's image extracted from the pdf.

from pypdf import PdfReader

reader = PdfReader("prueba maquetación inkscape.pdf")

page = reader.pages[0]

for count, image_file_object in enumerate(page.images):
    with open(str(count) + image_file_object.name, "wb") as fp:
        fp.write(image_file_object.data)

The name of the file is 0x9.png written in the same folder of the script


### Copying a pdf

We need to use the pdf writer for this, so first I'm going to try just to copy the existing pdf.

from pypdf import PaperSize, PdfReader, PdfWriter, Transformation

# Read source file
reader = PdfReader("prueba maquetación inkscape.pdf")
page = reader.pages[0]

# Write the new file 
writer = PdfWriter()
writer.add_page(page)
writer.write("write.pdf")

works! = D

### Substituting the image

For this I need to use PIL (Python Image Library) that it seems included but I need to add that import.

:::info
You need to use the replace in the writer, not in the reader.

:::

from pypdf import PdfReader, PdfWriter
from PIL import Image 

# Read source file
reader = PdfReader("prueba maquetación inkscape.pdf")
page = reader.pages[0]



# Write the new file 
writer = PdfWriter()
writer.add_page(page)

writer.pages[0].images[0].replace(Image.open("PruebaMonchoEntrada.jpg"), quality=100)

writer.write("write.pdf")

And it works, now I have a copy pdf with a different image!


Reference:

https://pypdf.readthedocs.io/en/stable/modules/PageObject.html#pypdf._page.PageObject.images

### Substituting the text

It seems that extract_text() it just... well, extract the text but doesn't modify it.

I've had also tried page.get_contents() but returns {}


Buuuut we can try to use "getObject" and try to find the specific objects.

In this case we're looking for 2 text boxes so I write this code

from pypdf import PdfReader
# Read source file
reader = PdfReader("prueba maquetación inkscape.pdf")
page = reader.pages[0]

for key in page:
    print(key, page[key])


print("----")

for x in range(0,9):
    print(x, reader.get_object(x))

print("----")


Seems like a tree.

I couldn't access to the specifics. And the very documentation points to somewhere else.

## Borb

https://github.com/jorisschellekens/borb

This has a really good documentation in borb-examples, pretty useful!

https://github.com/jorisschellekens/borb-examples

This would be the next step, later.

### Notas en castellano

Como la documentación de Borb es tan buena, voy a pasar a hacerla en castellano como notas mías.

#### Importar

hay que sacar las cosas de borp.pdf, es curioso que tenga _extensión_

En el ejemplito de hola mundo saca solo unas pocas clases (intuyo para hacerlo todo más veloz).

`Document`, representa un documento
`Page`, una página en el documento.
`PageLayout` es una clase abstracta que define como eñes van las cosas en una página.
`SingleColumnLayout` es una implmentación de `PageLayout`
`Paragraph` es un cacharro par añadir texto en un pdf.
Y luego `PDF` se dedica a convertir el `Document` en un archivo de verdad.

:::info

Algo de OOP, aquí se ve que utiliza cosas de herencia y clases porque especifica intencionadamente qué es cada variable. Se nota cuando utilza la lína

`l: PageLayout = SingleColumnLayout(p)`

Donde PageLayout es la clase abstracta y la SingleColumnLayout es el constructor de la clase concreta.

:::

#### Fuentes

Parece que solo acepta ttfs, hay que importarlas en el script.

Para hacerlo, además hay que instalarse "fonttools" con un pip install fonttools.

#### Colores

También hay que importar la wea. Los básicos

RGBColor, HexColor, HSVColor, CMYKColor

Los otros:

PantoneColor, FarrowAndBallColor, X11Color https://www.w3schools.com/colors/colors_x11.asp

https://www.farrow-ball.com/eu/paint/all-paint-colours

la diferencia entre hexcolor y rgb es que el rgb le enchifas red green y blue y el otro solo el hexadecimal en string con el hashtag (por ejemplo "#FF003D" )

`ColorScheme` es una clase guapa que permite utilizando los otros colores generar complementarios y análogos.

:::info
Methods available in `ColorScheme`:

- **`analogous_colors`**: Generates colors adjacent to the base color on the color wheel.
- **`complementary_color`**: Provides the color opposite to the base color on the color wheel.
- **`monochromatic`**: Creates a palette based on variations in the lightness or darkness of a single color.
- **`split_complementary_color`**: Returns colors that are adjacent to the complementary color of the base color.
- **`square_colors`**: Generates four colors that are evenly spaced around the color wheel.
- **`tetradic_colors`**: Creates a palette of four colors using two complementary color pairs.
- **`tints`**: Creates lighter variations of the base color by adding white.
- **`triadic_colors`**: Generates three colors that are evenly spaced around the color wheel.
- **`shades`**: Creates darker variations of the base color by adding black.

:::

Espaciado de letras y palabras

https://github.com/borb-pdf/borb-examples/blob/master/02/README.md#23-setting-word_spacing-and-character_spacing