Simon Willison @simonwillison.net · Feb 27

Built a fun prototype this morning of binary search using HTTP range requests, in this case to lookup characters in ~77MB of unicode data simonwillison.net/2026/Feb/27/...

58 likes 3 replies

?

Replies

Laurence Rowe · Feb 27

I think you could create an index of the offsets of the various contiguous ranges of code points to avoid the binary search.

Banda Bassotti · Feb 27

There's 77MB of Unicode data?!

Pawel Jozefiak · Mar 2

Binary search over HTTP range requests for unicode lookups is genuinely clever. The laziness-as-a-feature angle is underused in tooling design.