Skip to content

Latest commit

 

History

History
56 lines (36 loc) · 1.29 KB

README.md

File metadata and controls

56 lines (36 loc) · 1.29 KB

Scrape opengles.gpuinfo.org

Steps

  1. Open Devtools
  2. Go to the network tab
  3. Go to `https://opengles.gpuinfo.org/listreports.php'
  4. In the network tab, right click the reports.php?... entry and pick "open in a new tab"
  5. In the new tab, open the devtools to the networ page.
  6. In the URL in the omnibar , change length=25 to length=10000 and press enter
  7. In the response tab on the request, right click and pick "Save As..."
  8. Save as record-entries.json.

in a terminal

npm ci
node scrape.js

This tries to download each report as HTML to records.html

When finished

node html-to-json.js

This parses each HTML file and writes the contents to records-json.

If it crashes (v8 runs out of memory), run it again. Continue until all files are done

node flatten-json.js`

This reads all the json files and writes both alldata.json (an array with the data flattened to key/values) and alldata.csv with column headings and the data normalized.

query.js can be modified as an example of querying yourself.

node query.js

License

All the data is from https://opengles.gpuinfo.org/

The license there is CC-BY-4.0 so the data in the json, csv, and html files is CC-BY-4.0

The code in this repo license MIT