ConceptDBISAI-written · not yet proofread

HTTP Protocol

Also HTTP · Media Type · MIME Type

HTTP (Hypertext Transfer Protocol) is the text-based request–response protocol between browser (client) and web server, in which a request consists of a request line, headers and an optional body, and a response of a status line, headers (e.g. Content-Length, Content-Type), an empty line and the content.

Know these firstNetwork Socket

Formula#

Request format#

METHOD /path?param=value HTTP/1.1
Host: hostname:port
<more headers>

<body, only for e.g. POST>

/path?param=value is the part of the URL after http://<host>:<port>.

GET request example:

GET /search?query=bla HTTP/1.1
Host: localhost:8000

POST request example:

POST /submit HTTP/1.1
Host: localhost:8000
Content-Type: application/x-www-form-urlencoded
Content-Length: 9

query=bla

Response format#

HTTP/1.1 <status-code> <reason>
Content-Type: <media type>
Content-Length: <number of content bytes>

<content>

Example:

HTTP/1.1 200 OK
Content-Type: application/json
Content-Length: 18

{"result": "data"}

Status codes#

Code	Meaning
200	OK
301	Moved Permanently
400	Bad Request
403	Forbidden: the resource exists but may not be accessed by this client
404	Not Found: the requested resource does not exist
500	Internal Server Error

Media Types#

A media type (aka content type, formerly MIME type, MIME = Multipurpose Internet Mail Extensions) is a standard name for a type of content sent across the internet. The server declares it in the Content-Type header, optionally with the charset (text/html; charset=utf-8).

Media type	Used for
`text/html`	HTML pages
`text/css`	CSS stylesheets
`text/plain`	plain text
`application/javascript`	JavaScript
`application/json`	JSON data
`image/jpeg`	JPG image
`image/x-icon`	ICO image (notably `favicon.ico`)
`application/x-www-form-urlencoded`	form POST data

Properties#

HTTP/1 demands that line breaks are \r\n; the recommendation is to accept a single \n, too.
Content-Length counts bytes of the content, not characters (Character Encoding).
HTTP is stateless: each request is independent.
GET parameters are in the URL (visible, bookmarkable, limited length) and must be URL-encoded. POST sends arbitrarily large data in the body along with the URL. There are many more request types and headers.
Connection reuse: browsers send Connection: keep-alive to reuse the connection; a server answering Connection: close forces a new connection per request (HTTP Server > Concurrency).
Cross-origin headers: Origin (request) and Access-Control-Allow-Origin (response) implement CORS (Same-Origin Policy).
APIs use the same format: the LLM and embedding APIs of Sheet 12 are POST requests with Content-Type: application/json, an Authorization: Bearer <key> header and a JSON body.
Debugging: browser development console (F12), tab Network, shows all requests and responses; curl -v prints them on the command line.

Example#

ExampleWS19/20, T3.3

The page’s JavaScript sends the input to http://backend.org:8888/api?query=… after each keystroke. After the user typed bla, the last request and a successful response start with

GET /api?query=bla HTTP/1.1

HTTP/1.1 200 OK
Content-Type: application/json

The host and port are not part of the request line; they go into the Host header.

Sources#

Slides 09, p. 11 (request, response, \r\n), p. 12 (POST, 403, 404), p. 13 (media types), p. 14 (console), p. 21 (Connection)
DBIS_VL09 > HTTP, DBIS_VL09 > Request types and status codes, DBIS_VL09 > Media types, DBIS_VL09 > Reusing connections, DBIS_VL10 > CORS, DBIS_VL12 > Using an LLM via API (ES12)
DBIS Exam WS1920 (not published) T3.3, DBIS Exam WS2324 (not published) T4.3, DBIS Exam WS1516 (not published) T3.2, DBIS Exam WS1819 (not published) T4.2

Connections

Prerequisites know first1

Network Socket

Mentioned in inline links6

AJAX and Fetch API
AJAX (Asynchronous JavaScript and XML) is the technique of sending HTTP requests from JavaScript code in a page to a server without reloading the page, and writing the result into the page via the DOM. The modern standard is fetch(url): it… §
Character Encoding
Length in bytes ≠ length in characters. “knödel” has 6 characters but 7 UTF-8 bytes. Anything that counts bytes, e.g. the Content-Length header of an HTTP response, must be computed on the encoded bytes (HTTP Server). §
HTTP Server
Response headers. At least the status line, Content-Length (computed on the encoded bytes) and Content-Type with the media type matching the file ending: text/html, text/css, text/plain, application/javascript, application/json, image/x-ic… §
Network Socket
Read in rounds. recv(num_bytes) returns at most num_bytes bytes, possibly fewer, possibly an empty chunk. There is no universal “end of message” token, so the receiver loops until the application protocol says the message is complete. For… §
VL09 – Web Applications, Part 1
HTTP (Hypertext Transfer Protocol) is the protocol the browser uses to communicate with a (web) server. A typical request, where /search.html is the part of the URL after http://<host>:<port>: §
HTTP Protocol — request line, status line, headers, 200/403/404, GET vs. POST, Connection header §
Web Applications
HTTP Protocol — request line, status line, headers (Content-Length, Content-Type media types), status codes §

Search in these mentions →