insert_htmlbox error: int too large to convert to float

[Mominadar]

Description of the bug

Hi,

I am using the PyMuPDF library to translate pdfs keeping the overall structure intact. Now, this has been working successfully but while trying to translate an RTL langue. I got this error in processing the pdf on insert_htmlbox function.

int too large to convert to float

The bbox for which this occured is:

[393.83990478515625, 245.69000244140625, 393.83990478515625, 256.7300109863281]

I tried to round off the values to 2 decimal places even but still got the same issue. This bbox value is gotten using this code snippet:

 pages_blocks = []
        for page_num in range(len(doc)):
            page = doc.load_page(page_num)
            blocks = page.get_text("dict")["blocks"]
            page_blocks = []
            for block in blocks:
                if "lines" in block:
                    for line in block["lines"]:
                        for span in line["spans"]:
                            if span["text"].strip():
                                bbox = span["bbox"]
                                if isinstance(bbox, (list, tuple)) and len(bbox) == 4:
                                    bbox = [float(coord) for coord in bbox]
                                    if span["text"]:
                                        page_blocks.append({
                                            "text": span["text"],
                                            "bbox": bbox,
                                        })
            pages_blocks.append(page_blocks)
        return pages_blocks   

Note that is value of bbox works:

[393.83990478515625, 245.69000244140625, 400.83990478515625, 256.7300109863281]

Only difference is the third value's whole part so I am unsure what exactly are the constraints here.

How to reproduce the bug

Code to reproduce output:

import fitz

class TranslatePdfClient: 
    def extract_text_blocks(self, doc):
        pages_blocks = []
        for page_num in range(len(doc)):
            page = doc.load_page(page_num)
            blocks = page.get_text("dict")["blocks"]
            page_blocks = []
            for block in blocks:
                if "lines" in block:
                    for line in block["lines"]:
                        for span in line["spans"]:
                            font_properties = {"size": span["size"], "color": "#%06x" % span["color"], "font-family": '%s' % span["font"]}
                            if span["text"].strip():
                                bbox = span["bbox"]
                                if isinstance(bbox, (list, tuple)) and len(bbox) == 4:
                                    bbox = [float(coord) for coord in bbox]
                                    if span["text"]:
                                        page_blocks.append({
                                            "text": span["text"],
                                            "bbox": bbox,
                                            "font_properties": font_properties
                                        })
            pages_blocks.append(page_blocks)
        return pages_blocks   
    
    def translate(self, source_language, target_language):
        scale = 0.5
        try:
            doc = fitz.open('test.pdf')
            pages_blocks = self.extract_text_blocks(doc)
            
            print("doc", doc)
            for page_num, page_blocks in enumerate(pages_blocks):
                page = doc.load_page(page_num)
                for block in page_blocks:
                    text = block["text"]
                    bbox = block["bbox"]
                    font_properties = block["font_properties"]
                    if isinstance(bbox, (list, tuple)) and len(bbox) == 4:
                        try:
                            translated_text = text #hardcoding for bug reproduction only
                            if any(c.isalpha() for c in text):
                                #bbox = [388.3199157714844, 250.489990234375, 400.7757873535156, 261.5299987792969] #works
                                #bbox = [393.83990478515625, 245.69000244140625, 393.83990478515625, 256.7300109863281] #crashes
                                #bbox = [393.83990478515625, 245.69000244140625, 400.83990478515625, 256.7300109863281] works
                                rect = fitz.Rect(bbox)
                                page.add_redact_annot(rect, text="")
                                page.apply_redactions(images=0, graphics=0, text=0)
                                adjusted_size = font_properties['size'] - 2 
                                adjusted_size = "%g" % adjusted_size
                                html = f'''
                                        <div style="font-size:{adjusted_size}px; font-family:{font_properties["font-family"]}; color:{font_properties["color"]}; overflow:visible;">
                                            {translated_text}
                                        </div>
                                        ''' 
                                page.insert_htmlbox(rect, html, scale_low=scale)
                            else:
                                print(f'No need to translate non-alpha text: {text}')
                        except Exception as e:
                            print(f"Error processing block: '{text}' with bbox: {bbox}")
                            print(e)
                            raise
                    else:
                        print(f"Invalid bbox: {bbox}")
            
            doc.save('output.pdf', garbage=4, clean=True, deflate=True, deflate_images=True, deflate_fonts=True)
            doc.close()
        except Exception as e:
            print(fitz.TOOLS.mupdf_warnings())
            print(f"Error processing block: {str(e)}")
            print(e)


if __name__ == "__main__":
    client = TranslatePdfClient()
    client.translate('arabic', 'english')

test file:
Getting error when using file directly as:
Error processing block: 'ر' with bbox: [393.83990478515625, 245.69000244140625, 393.83990478515625, 256.7300109863281]
'NoneType' object has no attribute 'y1'

Error processing block: 'NoneType' object has no attribute 'y1'
'NoneType' object has no attribute 'y1'

but when I use a file stream which I use on prod, I get:
Error processing block: 'ر' with bbox: [393.83990478515625, 245.69000244140625, 393.83990478515625, 256.7300109863281]
int too large to convert to float

file: test.pdf

PyMuPDF version

1.24.11

Operating system

Linux

Python version

3.10

[julian-smith-artifex-com]

Thanks for the detailed report, i was able to reproduce it easily.

It's a bug in Page.insert_htmlbox() where it fails to notice if the html content doesn't fit even at the minimum scaling scale_low (0.5 in your code).

I have a fix in my tree.

[Mominadar]

Hi Juilan,

Thanks so much for the quick response and fix. Really appreciate it. Will this be released soon?

[julian-smith-artifex-com]

I'm hoping to make a new release this week, but it's quite likely to be delayed until sometime next week.

[julian-smith-artifex-com]

Fixed in 1.24.12.