Brotli forced by server, how to decode?

[gugatr0n1c]

Hi, pls, cannot find anything. Using HTTP.jl for scraping. When server forces brotli compression, how to deal with that? Gzip not working, getting 406 error. Only brotli is accepted, but is there in julia some way to decode? Thx

ie: https://www.cbssports.com/nba/transactions/

[gugatr0n1c]

workaround with python (PyCall, Python with brotli):

load_data = HTTP.request("GET", url, headers, decompress=false, readtimeout=180)
encoding = HTTP.header(load_data, "Content-Encoding", "")
if occursin("br", lowercase(encoding))
    brotli = pyimport("brotli")
    load_data = String(brotli.decompress(load_data.body))
else
...
end

[nhz2]

I made a package using the brotli C library
that you can use as an alternative to PyCall here.

https://github.com/JuliaIO/ChunkCodecs.jl/tree/main/LibBrotli

using ChunkCodecLibBrotli
String(decode(BrotliCodec(), load_data.body))

[quinnj]

Thanks @nhz2; that's awesome. Great solution to this.