Parsing web page requires playwright?

[gmgreg]

Is playwright required? Readme seems to suggest this is optional for CSS parsing only (e.g. "Note: Without Playwright, CSS selectors still work for text extraction, but no visual highlighting screenshots are generated.")

from attachments import Attachments
ctx = Attachments("https://en.wikipedia.org/wiki/Llama")
print(str(ctx))      # Pretty text view
print(len(ctx.images))  # Number of extracted images

output:

[Attachments] Running primary processor 'webpage_to_llm' for https://en.wikipedia.org/wiki/Llama
[Attachments]   Applying step 'load.url_to_bs4' to https://en.wikipedia.org/wiki/Llama
[Attachments]   Running AdditivePipeline(present.markdown + present.images + present.metadata)
[Attachments]     Applying additive step 'present.markdown' to https://en.wikipedia.org/wiki/Llama
[Attachments]     Applying additive step 'present.images' to https://en.wikipedia.org/wiki/Llama
Processor failed for https://en.wikipedia.org/wiki/Llama: Playwright not available. Install with: pip install playwright && playwright install chromium, falling back to universal pipeline
[Attachments] Applying step 'load.url_to_response' to https://en.wikipedia.org/wiki/Llama
[Attachments] Applying step 'modify.morph_to_detected_type' to https://en.wikipedia.org/wiki/Llama
[Attachments] Applying step 'load.url_to_bs4' to https://en.wikipedia.org/wiki/Llama
[Attachments] Applying step 'load.git_repo_to_structure' to https://en.wikipedia.org/wiki/Llama
[Attachments] Applying step 'load.directory_to_structure' to https://en.wikipedia.org/wiki/Llama
[Attachments] Applying step 'load.svg_to_svgdocument' to https://en.wikipedia.org/wiki/Llama
[Attachments] Applying step 'load.eps_to_epsdocument' to https://en.wikipedia.org/wiki/Llama
[Attachments] Applying step 'load.pdf_to_pdfplumber' to https://en.wikipedia.org/wiki/Llama
[Attachments] Applying step 'load.csv_to_pandas' to https://en.wikipedia.org/wiki/Llama
[Attachments] Applying step 'load.image_to_pil' to https://en.wikipedia.org/wiki/Llama
[Attachments] Applying step 'load.html_to_bs4' to https://en.wikipedia.org/wiki/Llama
[Attachments] Applying step 'load.text_to_string' to https://en.wikipedia.org/wiki/Llama
[Attachments] Applying step 'load.zip_to_images' to https://en.wikipedia.org/wiki/Llama
[Attachments] Applying step 'modify.pages' to https://en.wikipedia.org/wiki/Llama
[Attachments] Running AdditivePipeline(present.markdown + present.images + present.metadata)
[Attachments]   Applying additive step 'present.markdown' to https://en.wikipedia.org/wiki/Llama
[Attachments]   Applying additive step 'present.images' to https://en.wikipedia.org/wiki/Llama
⚠️ Could not process https://en.wikipedia.org/wiki/Llama: Playwright not available. Install with: pip install playwright && playwright install chromium

[MaximeRivest]

You would need [images: false] at the end of you string. Attachments goes all out, so tries to render an image of the site by default, with playwright.