├── js.png ├── recognized-code.png ├── README.md └── index.html /js.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/bahmutov/image-as-script/HEAD/js.png -------------------------------------------------------------------------------- /recognized-code.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/bahmutov/image-as-script/HEAD/recognized-code.png -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # image-as-script 2 | 3 | People argue why the websites are so slow - is it the large images or lots of 4 | JavaScript? The next natural step is to send large images *with JavaScript*. 5 | Then one can run [character recognition library](https://github.com/antimatter15/ocrad.js) 6 | to get the JavaScript out of the image and execute it. Seems reasonable. 7 | 8 |  9 | 10 | This repo: [https://github.com/bahmutov/image-as-script](https://github.com/bahmutov/image-as-script) 11 | -------------------------------------------------------------------------------- /index.html: -------------------------------------------------------------------------------- 1 | 2 | 3 |
4 |Below is screenshot of some JavaScript
10 |
11 | Now this browser will run optical character recognition and 12 | then will execute the recognized code
13 |Uses ocrad.js library.
14 | 15 |This repo at github.com/bahmutov/image-as-script. Font "Times Roman" was most reliably recognized.
16 | Demo requires a server to run, will not work from file:// protocol.
Recognized code 18 |
19 | 20 | 42 | 43 | 44 | --------------------------------------------------------------------------------