httplite

A minimal and single-header HTTP/1.1 request line and header parser written in pure C.

Overview

httplite is an ultra lightweight, zero-allocation HTTP/1.1 parser designed for blazing-fast parsing using only pointer arithmetic. It parses HTTP request or response lines and headers directly from the input buffer without copying data.

LOC

The header file is exactly 50 lines in total...

Extended version

Check out the "extended-version" path in the repository to get more information.

Usage Example

#include "httplite.h"
#include <stdio.h>

void main() {
    const char *request = "GET /index.html HTTP/1.1\r\nHost: example.com\r\nConnection: close\r\n\r\n"; /* A typical http request */
    size_t len = 66; /* Length of request without the null terminator */

    http_message message; /* Http message structure */
    if (http_parse_message(request, len, &message)) {
        printf("Method: %.*s\n", (int)message.part1_length, message.part1);
        printf("Path: %.*s\n", (int)message.part2_length, message.part2);
        printf("Version: %.*s\n", (int)message.part3_length, message.part3);

        const char *name, *value;
        size_t name_len, value_len;

        printf("Headers:\n");
        while (http_next_header(&message, &name, &name_len, &value, &value_len)) {
            if (name == NULL || name_len == 0) break;
            printf("  %.*s: %.*s\n", (int)name_len, name, (int)value_len, value);
        }

        if (message.next_length != 0) {
            printf("Body:\n%.*s\n", (int)message.next_length, message.next);
        }
    }
    else {
        printf("Failed to parse HTTP message\n");
    }
}

How It Works

• Parses HTTP/1.x request or response line (method, path, version).
• Iterates over headers with zero-copy pointer arithmetic like a stream.
• No dependencies, dynamic memory allocation or even the standart library.

API

• int http_parse_message(const char *buffer, size_t buffer_size, http_message *out); Parses the start line and initializes the message for header parsing.

• int http_next_header(http_message *message, const char **name, size_t *name_length, const char **value, size_t *value_length); Iterates headers one by one; returns 1 while headers remain, 0 at end or error.

Performance Notes

• Zero-copy parsing ensures minimal CPU and memory overhead.
• Suitable for high-throughput or embedded environments.

Limitations

• Supports HTTP/1.x request line, response line and headers only.
• Expects CRLF (\r\n) line endings.
• Does not validate message bodies or headers, just does its primary job which is parsing.
• No support for multiline headers or chunked encoding.

License

MIT License

---

Minimal, blazing-fast HTTP/1.1 parsing using only pointer arithmetic for zero-copy efficiency.