xhtml2pdf 为本地 html 文件生成空白 pdf

Question

I have an interactive html file generated using Bokeh. I am trying to convert the html file into pdf using xhtml2pdf but the images on the html file dont show up on the pdf with xhtml2pdf or pdfkit.

I tried PyQt to generate the pdfs and it worked fine but had issues with being converted to an executable and pdfcrowd worked too but it is a paid library. I have looked at all the other solutions online but none of those seem to work.

This is the html file generation:

    source = ColumnDataSource(data=dict(x=xx, y=yy, z=val))
    
    surface1 = Surface3d1(x="x", y="y", z="z", data_source=source, width=600, height=600)
    surface2 = Surface3d2(x="x", y="y", z="z", data_source=source, width=600, height=600)
    surface3 = Surface3d3(x="x", y="y", z="z", data_source=source, width=600, height=600)
    surface4 = Surface3d4(x="x", y="y", z="z", data_source=source, width=600, height=600)
    
    l1 = gridplot([[Div(text = 'Side View 1'), Div(text = 'Front View')], [surface1, surface2]])
    l2 = gridplot([[Div(text = 'Side View 2'), Div(text = 'Top View')], [surface3, surface4]])

    filename = filename_out.split("\\")
    
    layout = column(Div(text = str(filename[-1].replace('_', ' '))),Div(text = 'Absolute Warpage = ' + str(warpage)),l1,l2)
    output_file(filename_out[:-1]+'.html', title = 'CMM Scan', mode = 'inline')
    save(layout)

This is the code for converting to pdf

def convert_html_to_pdf(source_html, output_filename):
from xhtml2pdf import pisa 
# open output file for writing (truncated binary)
result_file = open(output_filename, "w+b")

# convert HTML to PDF
pisa_status = pisa.CreatePDF(
        source_html,                # the HTML to convert
        dest=result_file)           # file handle to recieve result

# close output file
result_file.close()                 # close output file

# return True on success and False on errors
return pisa_status.err

convert_html_to_pdf(str(filename_out+'.html'),str(filename_out+'.pdf'))

Answer 1

Standard Bokeh plots are JavaScript programs that render to an HTML raster canvas. I am a little surprised to hear that anything works to covert to PDF. Any conversion tool that only looks at the base HTML content, without actually executing the JS code, certainly cannot work. Most likely, PyQt works because it embeds an actual real browser engine, while the others tools do not.

You might have better luck by setting output_backend="svg" on your plots (but note there are also some known issues still being fixed with Bokeh's SVG support, so YMMV). However, I suspect this may also run in to the same problem above, where running JS code is what produces the SVG, and absent a JS engine, no plot will show up.

Ultimately, the only option may be to separately export the Bokeh plots (eg using export_svgs or export_png ) into static PNGs or SVGs that can be included in a PDF (or in an HTML document that then can be naively converted in to a PDF without running JS code).