I have a very strange issue with HTTPSOCKET:
Try the following:
dim http as new HTTPSocket
http.Yield=true
msgbox http.Get("http://api.discogs.com/masters/3354", 15)
You will notice the msgbox will display garbage.
Now try : http://api.discogs.com/masters/3354 in any browser and you will notice it is readable JSON text.
http://api.discogs.com/masters/3354 is not Json, but a HTML page, where the contains a
element containing a string which can be read as JSON by human beings.JSON or HTML … the issue remains. I get garbage when using the HTTPSOCKET.
No, you’re right. Should have checked the InternetHeaders. They come through correctly (content type: JSON). But the encoding is missing (something like charset: utf-8). Can you add the encoding to the server’s response header?
The string you get in your example is gzipped!
Are you sure? Why does a browser display it correctly (readable)?
Did a quick test … its not a gzip (or any other compressed file)
Oeps … sorry … it is indeed a zipped code. Strange.
What could be the fastest methode to make it readable? (without writing it first as a zip file, unzip it, and read the file again)
The headers says so:
HTTP/1.1 200 OK
Reproxy-Status: yes
Access-Control-Allow-Origin: *
Content-Type: application/json
Content-Encoding: gzip
Server: lighttpd
Content-Length: 1055
Date: Sat, 15 Feb 2014 16:37:01 GMT
X-Varnish: 2046486739 2027595096
Age: 65473
Via: 1.1 varnish
Connection: close
Indeed. See my previous answer.
Strange it is loaded as a gzip file while in a browser it displays correctly (plain text).
Now I need a way to unzip it into memory without writing it first to a file.
As I wrote in my very first answer, it is shown as a HTML page by the browser (Safari in my case). Same as with XML and other data formats. They are interpreted by the browser and shown within a HTML page created on the fly.
I don’t know how to gunzip in Xojo. I would probably try to call gunzip in a Shell.
Ok, works when I use GZipFileMBS