2fcc92dbd9
- Edition de contenu PDF (Annotations > Editer le contenu) : - Onglet Texte : liste le texte detecte, masquer des lignes - Onglet Images : liste les images, remplacer ou masquer - Actions en ligne de commande (--action booklet_a4|booklet_a3|nup_2x2|unbooklet) - Menu contextuel Windows enrichi : - Ouvrir, Livret A4, Livret A3, 2x2 par feuille, Delivretiser - Bouton Editer dans la barre d'outils droite Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
91 lines
3.1 KiB
Python
91 lines
3.1 KiB
Python
"""Extraction de texte et images depuis un PDF."""
|
|
|
|
from pypdf import PdfReader
|
|
from .pdf_utils import get_page_size
|
|
|
|
|
|
def extract_text_from_page(input_path, page_index):
|
|
"""Extraire le texte d'une page avec positions approximatives."""
|
|
reader = PdfReader(input_path)
|
|
if page_index >= len(reader.pages):
|
|
return []
|
|
page = reader.pages[page_index]
|
|
text = page.extract_text() or ""
|
|
lines = text.split("\n")
|
|
w, h = get_page_size(page)
|
|
# Positions approximatives (pypdf ne donne pas les coords exactes)
|
|
results = []
|
|
line_height = h / max(len(lines), 1)
|
|
for i, line in enumerate(lines):
|
|
if line.strip():
|
|
results.append({
|
|
"text": line.strip(),
|
|
"y_approx": h - (i + 1) * line_height,
|
|
"line_index": i,
|
|
})
|
|
return results
|
|
|
|
|
|
def extract_images_from_page(input_path, page_index):
|
|
"""Extraire les images d'une page."""
|
|
reader = PdfReader(input_path)
|
|
if page_index >= len(reader.pages):
|
|
return []
|
|
page = reader.pages[page_index]
|
|
images = []
|
|
if "/XObject" in page["/Resources"]:
|
|
xobjects = page["/Resources"]["/XObject"].get_object()
|
|
for obj_name in xobjects:
|
|
obj = xobjects[obj_name].get_object()
|
|
if obj["/Subtype"] == "/Image":
|
|
width = obj["/Width"]
|
|
height = obj["/Height"]
|
|
images.append({
|
|
"name": obj_name,
|
|
"width": width,
|
|
"height": height,
|
|
})
|
|
return images
|
|
|
|
|
|
def replace_image_in_page(input_path, output_path, page_index, image_name, new_image_path):
|
|
"""Remplacer une image dans une page (overlay avec masquage)."""
|
|
from pypdf import PdfWriter
|
|
from reportlab.pdfgen import canvas
|
|
from reportlab.lib.utils import ImageReader
|
|
import io
|
|
|
|
reader = PdfReader(input_path)
|
|
writer = PdfWriter()
|
|
|
|
for i, page in enumerate(reader.pages):
|
|
if i == page_index and new_image_path:
|
|
w, h = get_page_size(page)
|
|
# On ne peut pas vraiment remplacer l'image dans le flux PDF
|
|
# On ajoute la nouvelle image par-dessus (avec un masque blanc si besoin)
|
|
packet = io.BytesIO()
|
|
c = canvas.Canvas(packet, pagesize=(w, h))
|
|
try:
|
|
img = ImageReader(new_image_path)
|
|
img_w, img_h = img.getSize()
|
|
# Centrer l'image
|
|
scale = min(w / img_w, h / img_h) * 0.8
|
|
draw_w = img_w * scale
|
|
draw_h = img_h * scale
|
|
x = (w - draw_w) / 2
|
|
y = (h - draw_h) / 2
|
|
c.drawImage(new_image_path, x, y, width=draw_w, height=draw_h,
|
|
preserveAspectRatio=True, mask='auto')
|
|
except Exception:
|
|
pass
|
|
c.showPage()
|
|
c.save()
|
|
packet.seek(0)
|
|
overlay = PdfReader(packet)
|
|
page.merge_page(overlay.pages[0])
|
|
writer.add_page(page)
|
|
|
|
with open(output_path, "wb") as f:
|
|
writer.write(f)
|
|
return output_path
|