Parse facebook message from archived data

fb-message-parse.php

<?php
# composer require sunra/php-simple-html-dom-parser
require __DIR__ . '/vendor/autoload.php';

// Change file size because fb message archive is very big data.
define( 'MAX_FILE_SIZE', 1000000000 );

$dom = Sunra\PhpSimple\HtmlDomParser::file_get_html( 'messages.htm' );

$messages = [];

foreach( $dom->find('div.message') as $message ) {
	$date = $message->find('.meta', 0)->innertext;
    // You can filter message here.
	if ( preg_match( '#2017年8月25日 1[34]#', $date ) ) {
		$messages[] = [
			'user'    => $message->find( '.user', 0 )->innertext,
			'message' => $message->nextsibling()->innertext, // Message content is p tag ;(
		];
	}
}

// Do stuff.
ob_start();
$former = '';
for ( $i = count( $messages ); $i > 0; $i-- ) {
	$message = $messages[$i-1];
	$first_letter = ucfirst( substr( $message['user'], 0, 1 ) );
	$class_name = '';
	$prefix = "<strong>{$first_letter}:</strong> ";
	if ( 'F' == $first_letter  ) {
		$prefix = '——';
		$class_name = ' class="interviewer"';
	}
	if ( $former == $first_letter  ) {
		$prefix = '';
	}
	$former = $first_letter;
	printf( '<p%s>%s%s</p>', $class_name, $prefix, $message['message'] );

	echo PHP_EOL;
}
file_put_contents( './message-log.html', ob_get_contents() );
ob_end_clean();