Studymaker / ppt_parser.py
g0th's picture
Update ppt_parser.py
b55bd9b verified
import json
import os
from pptx import Presentation
from pptx.util import Inches
from pptx.shapes.group import GroupShape
from pptx.shapes.picture import Picture
from PIL import Image
import io
def print_json(item):
return json.dumps(item, ensure_ascii=False, indent=4)
def safe_font_attribute(run, attr):
try:
return getattr(run.font, attr)
except Exception:
return None
def safe_color(run):
try:
return str(run.font.color.rgb) if run.font.color and run.font.color.rgb else None
except Exception:
return None
def extract_paragraph_data(paragraph):
if not paragraph.runs:
return None
run = paragraph.runs[0]
return {
'text': paragraph.text,
'align': paragraph.alignment,
'font': {
'name': safe_font_attribute(run, 'name'),
'bold': safe_font_attribute(run, 'bold'),
'italic': safe_font_attribute(run, 'italic'),
'underline': safe_font_attribute(run, 'underline'),
'color': safe_color(run),
'language_id': safe_font_attribute(run, 'language_id'),
}
}
def transfer_textbox_content_in_group(group_shape):
group_shape_item = {}
for l, shape in enumerate(group_shape.shapes):
shape_item = {}
if shape.has_text_frame:
shape_item['type'] = "text"
shape_item['location'] = (shape.left, shape.top)
text_frame = shape.text_frame
for r, paragraph in enumerate(text_frame.paragraphs):
data = extract_paragraph_data(paragraph)
if data:
shape_item[f'paragraph_{r}'] = data
group_shape_item[f"shape_{l}"] = shape_item
return group_shape_item
def transfer_to_structure(pptx_file, images_dir_path):
item = {}
prs = Presentation(pptx_file)
image_path_list = []
for i, slide in enumerate(prs.slides):
slide_item = {}
for j, shape in enumerate(slide.shapes):
shape_item = {}
# Case 1: Normal text box
if shape.has_text_frame:
shape_item['type'] = "text"
text_frame = shape.text_frame
for r, paragraph in enumerate(text_frame.paragraphs):
data = extract_paragraph_data(paragraph)
if data:
shape_item[f'paragraph_{r}'] = data
# Case 2: Grouped shapes
elif isinstance(shape, GroupShape):
shape_item['type'] = "group"
shape_item['group_content'] = transfer_textbox_content_in_group(shape)
# Case 3: Picture
elif isinstance(shape, Picture):
shape_item['type'] = "picture"
image_path = os.path.join(images_dir_path, f"picture_{j}.png")
image_path_list.append(image_path)
shape_item['image_path'] = image_path
shape_item['size'] = shape.image.size
shape_item['dpi'] = shape.image.dpi
shape_item['location'] = (shape.left, shape.top)
shape_item['location_inches'] = (Inches(shape.left).inches, Inches(shape.top).inches)
try:
image_stream = io.BytesIO(shape.image.blob)
shape_image = Image.open(image_stream)
shape_image.save(image_path)
except Exception:
pass # Corrupt or unsupported image
slide_item[f"shape_{j}"] = shape_item
item[f"slide_{i}"] = slide_item
return print_json(item), image_path_list