tracking issue: pdf ecommerce mode

[cdxker]

Description

• feature: file worker needs to add extra metadata on page, and pdf url from pdf2md. #2929

• feature: render pdf preview for page a chunk appears in the new pdf mode. PDF mode should derive from ecommerce mode with a style similar to this image. (Less results per row).
  

• feature: add pdf mode as an option in public page mode

Target(s)

<replace w/ name of the service(s) which are associated with this issue>

Community channels

Matrix is preferred. Reach out on discord or Matrix for further assistance.

• discord
• matrix

[vtempest]

https://github.com/vtempest/ai-research-agent/tree/master/src/extractor/pdf-to-html
https://qwksearch.com/docs/docs/functions/src/extractor/pdf-to-html/
Check out this much more accurate pdf-to-html lib with 5k lines of code handling much v better formatting things like Footnote linking which ocr does not.

Trieves pdf2md is good for a few types of image pdf but its overkill too costly for use cases like 10k pdfs which this does for free.

https://blog.pgvecto.rs/why-hnsw-is-not-the-answer

You can easily run a 90% cheaper scaling stack with pgvector or cloud flare vector AI
It seems unclear who this is for, if it's for devs they can just use qdrant or type sense or HF transformers.js instead of a huge stack that's why it should be separated into modules. As a dev shop you need to find cost effective solutions or your clients will find them later.
Khami tried to justify this saying it's faster to use Hnsw qdrant but Hnsw is slower and older than ngt based algos that have dominated the benchmark
https://ann-benchmarks.com/index.html

We pledge to act and interact in ways that contribute to an open, welcoming, diverse, inclusive, and healthy community.
Demonstrating empathy and kindness toward other people
Being respectful of differing opinions, viewpoints, and experiences

It's ironic nick khami looks for bs excuses to ban great Ai researchers like me who found these ideas years ago and only after banning me he integrated. Khami was calling yc founders elitist douchey when I was encouraging him to apply... Then he became one and this will get worse next 20y investors have been messaged and there's a bunch of Ai accounts saying this is hostile and exclusionary. Hopefully you can learn to mature more and be forgiving whatever got you into a passive aggressive grudge in the spirit of the holidays 🎁✌️