manticoresearch/misc/http-log-to-test

#!/usr/bin/env manticore-executor
<?php declare(strict_types=1);

// Copyright (c) 2017-2026, Manticore Software LTD (https://manticoresearch.com)
// Copyright (c) 2001-2016, Andrew Aksyonoff
// Copyright (c) 2008-2016, Sphinx Technologies Inc
// All rights reserved
//
// This program is free software; you can redistribute it and/or modify
// it under the terms of the GNU General Public License. You should have
// received a copy of the GPL license along with this program; if you
// did not, you can find it at http://www.gnu.org/

function parseLogFile(string $filename): array {
	$result = [];
	$currentEntry = null;
	$isRequest = false;
	$isResponse = false;

	// Open file for reading
	$handle = fopen($filename, 'r');
	if (!$handle) {
		throw new Exception("Unable to open file: $filename");
	}

	while (($line = fgets($handle)) !== false) {
		// Remove trailing whitespace
		$line = rtrim($line);

		// Check for new entry
		if (preg_match('/^\[.*\] \[Request-ID: (\d+)\] - (Incoming HTTP Request|Outgoing HTTP Response):/', $line, $matches)) {
			// Start new entry if it's a request
			if (strpos($line, 'Incoming HTTP Request') !== false) {
				if ($currentEntry && shouldIncludeEntry($currentEntry)) {
					$result[] = $currentEntry;
				}
				$currentEntry = [
					'request_id' => $matches[1],
					'timestamp' => extractTimestamp($line),
					'request' => [],
					'response' => []
				];
				$isRequest = true;
				$isResponse = false;
			} elseif (strpos($line, 'Outgoing HTTP Response') !== false) {
				$isRequest = false;
				$isResponse = true;
			}

			continue;
		}

		// Skip delimiter lines
		if (preg_match('/^[<>]{10,}$/', $line)) {
			continue;
		}

		// Add content to current entry
		if ($currentEntry && !empty($line)) {
			if ($isRequest) {
				$row = parseHttpRequestLine($line);
				$currentEntry['request'] = mergeEntry($currentEntry['request'], $row);
			} elseif ($isResponse) {
				$row = parseHttpResponseLine($line);
				$currentEntry['response'] = mergeEntry($currentEntry['response'], $row);
			}
		}
	}

	// Process last entry from the list
	if ($currentEntry && shouldIncludeEntry($currentEntry)) {
		$result[] = $currentEntry;
	}

	fclose($handle);
	return $result;
}

/**
 * Parse a single HTTP request line into a struct
 * @param string $line
 * @return array
 */
function parseHttpRequestLine(string $line): array {
	static $isBody = false;
	$line = trim($line);

	// Return empty if line is empty
	if (empty($line)) {
		$isBody = true;
		return [];
	}

	// Check if it's the request line (first line)
	if (preg_match('/^(GET|POST|PUT|DELETE|HEAD|OPTIONS|PATCH)\s+(\S+)\s+HTTP\/([\d.]+)$/i', $line, $matches)) {
		return [
			'http:method' => $matches[1],
			'http:path' => $matches[2],
			'http:version' => $matches[3],
		];
	}

	// Header line check
	if (!$isBody && strpos($line, ':') !== false && preg_match('/^[\w\-]+:\s*.+$/', $line)) {
		list($key, $value) = array_map('trim', explode(':', $line, 2));
		return ["header:$key" => $value];
	}

	// If it's not a header and not the first line, treat it as body
	return [
		'body' => $line
	];
}

/**
 * Parse a single HTTP response line into a struct
 * @param string $line
 * @return array<string,string>
 */
function parseHttpResponseLine(string $line): array {
	static $isBody = false;
	$line = trim($line);
	// Empty line check
	if (empty($line)) {
		$isBody = true;
		return [];
	}

	// First line check (HTTP status line)
	if (preg_match('/^HTTP\/(\d\.\d)\s+(\d{3})\s+(.*)$/', $line, $matches)) {
		$isBody = false;
		return [
			'http:version' => $matches[1],
			'http:code' => (int)$matches[2],
			'http:message' => $matches[3],
		];
	}

	// Header line check
	if (!$isBody && strpos($line, ':') !== false && preg_match('/^[\w\-]+:\s*.+$/', $line)) {
		list($key, $value) = array_map('trim', explode(':', $line, 2));
		return ["header:$key" => $value];
	}

	// If none of the above, return the line as is
	return ['body' => $line];
}

function mergeEntry(array $entry, array $newEntry): array {
	foreach ($newEntry as $key => $value) {
		if ($key === 'body') {
			$entry[$key] = ($entry[$key] ?? '') . $value;
		} else {
			$entry[$key] = $value;
		}
	}
	return $entry;
}

/**
 * Check if the current entry should be included in the output
 * We are excluding all requests from Buddy that are still logged into the file
 * @param array $entry
 * @return bool
 */
function shouldIncludeEntry(array $entry): bool {
	return !str_starts_with($entry['request']['header:User-Agent'], 'Manticore Buddy');
}

function extractTimestamp($line) {
	if (preg_match('/^\[(.*?)\]/', $line, $matches)) {
		return $matches[1];
	}
	return null;
}

function printCLTTest(array $logEntries) {
	foreach ($logEntries as $entry) {
		$url = escapeshellarg($entry['request']['header:Host'] . $entry['request']['http:path']);
		unset($entry['request']['header:Host']);
		$args = [];
		foreach ($entry['request'] as $key => $value) {
			if (!str_starts_with($key, 'header:')) {
				continue;
			}
			$args[] = '-H ' . escapeshellarg(substr($key, 7) . ': ' . $value);
		}
		$method = $entry['request']['http:method'];
		if ($method === 'POST') {
			$args[] = '-d ' . escapeshellarg($entry['request']['body']);
		}
		$argLine = implode(' ', $args);
		echo "––– input –––\n";
		echo "curl -X {$entry['request']['http:method']} {$argLine} {$url}\n";
		echo "––– output –––\n";
		echo "{$entry['response']['body']}\n";
	}
}

try {
	if (!isset($argv[1])) {
		die("Usage: http-log-to-test <log-file>\n");
	}

	$logFile = $argv[1];
	if (!file_exists($logFile)) {
		die("Error: Log file '$logFile' not found.\n");
	}

	$logEntries = parseLogFile($logFile);
	printCLTTest($logEntries);
} catch (Exception $e) {
	fwrite(STDERR, "Error: " . $e->getMessage() . "\n");
	exit(1);
}