Extraction fails for frontend-heavy Angular SPA HTML with large inline style/font blocks

[shivangi2990]

I tested Defuddle against a large Angular SPA page source and observed extraction failure when the HTML primarily contained framework boilerplate and massive inline CSS/font definitions.

Observed Behavior

Defuddle returned empty/invalid markdown output and extraction failed because the HTML contained very little semantic readable content compared to DOM noise.

The HTML included:

large inline <style> blocks
thousands of @font-face declarations
bootstrap/material CSS
Angular app shell markup
tracking scripts and metadata
Expected Behavior

Defuddle should ideally:

ignore noisy/non-semantic nodes during preprocessing
or provide a preprocessing option for frontend-heavy SPA HTML
Suggested Improvement

A preprocessing step before readability extraction could help significantly, for example removing:

script
style
noscript
svg
stylesheet-related nodes

before running extraction.

Additional Context

The issue was reproduced consistently using fixture-based testing with a saved HTML payload from an Angular application page source.

[kepano]

Please provide a URL and the expected output

[shivangi2990]

raw-text.txt

[shivangi2990]

it is not able to extract content in li tags of the html