generated from Nodarx/template
feat: speed improvements
Signed-off-by: Sebastian Krupinski <root@LAPTOP-7DVOR6NC>
This commit is contained in:
Sebastian Krupinski
@@ -4,6 +4,7 @@ namespace Gricob\IMAP\Protocol\Response\Parser;
|
||||
|
||||
use DateTimeImmutable;
|
||||
use Doctrine\Common\Lexer\Token;
|
||||
use Gricob\IMAP\Mime\Part\Body;
|
||||
use Gricob\IMAP\Protocol\Response\Line\CommandContinuation;
|
||||
use Gricob\IMAP\Protocol\Response\Line\Data\CapabilityData;
|
||||
use Gricob\IMAP\Protocol\Response\Line\Data\ExistsData;
|
||||
@@ -11,6 +12,9 @@ use Gricob\IMAP\Protocol\Response\Line\Data\ExpungeData;
|
||||
use Gricob\IMAP\Protocol\Response\Line\Data\Fetch\Address;
|
||||
use Gricob\IMAP\Protocol\Response\Line\Data\Fetch\BodySection;
|
||||
use Gricob\IMAP\Protocol\Response\Line\Data\Fetch\BodyStructure;
|
||||
use Gricob\IMAP\Protocol\Response\Line\Data\Fetch\BodyStructure\MultiPart;
|
||||
use Gricob\IMAP\Protocol\Response\Line\Data\Fetch\BodyStructure\Part;
|
||||
use Gricob\IMAP\Protocol\Response\Line\Data\Fetch\BodyStructure\SinglePart;
|
||||
use Gricob\IMAP\Protocol\Response\Line\Data\Fetch\Envelope;
|
||||
use Gricob\IMAP\Protocol\Response\Line\Data\FetchData;
|
||||
use Gricob\IMAP\Protocol\Response\Line\Data\FlagsData;
|
||||
@@ -353,7 +357,7 @@ readonly class Parser
|
||||
$this->space();
|
||||
$text = $this->literal();
|
||||
|
||||
$bodySections[] = new BodySection($section, $text);
|
||||
$bodySections = $this->fetchBody($bodyStructure, $text);
|
||||
}
|
||||
break;
|
||||
case TokenType::ENVELOPE:
|
||||
@@ -384,6 +388,145 @@ readonly class Parser
|
||||
);
|
||||
}
|
||||
|
||||
/**
|
||||
* @return BodySection[]
|
||||
*/
|
||||
private function fetchBody(?BodyStructure $node, string $data): array {
|
||||
return $this->fetchBodyNode($node->part, $data);
|
||||
}
|
||||
|
||||
/**
|
||||
* @return BodySection[]
|
||||
*/
|
||||
private function fetchBodyNode(?Part $node, string $data, string $partId = ''): array {
|
||||
if ($node instanceof MultiPart) {
|
||||
return $this->fetchBodyMultipart($node, $data, $partId);
|
||||
}
|
||||
|
||||
if ($node instanceof SinglePart) {
|
||||
return [$this->fetchBodySinglePart($data, $partId)];
|
||||
}
|
||||
|
||||
return [];
|
||||
}
|
||||
|
||||
/**
|
||||
* @return BodySection
|
||||
*/
|
||||
private function fetchBodySinglePart(string $data, string $partId = ''): BodySection
|
||||
{
|
||||
$partId = empty($partId) ? '1' : $partId;
|
||||
return new BodySection($partId, $data);
|
||||
}
|
||||
|
||||
/**
|
||||
* @return BodySection[]
|
||||
*/
|
||||
private function fetchBodyMultipart(MultiPart $structure, string $data, string $partId = ''): array
|
||||
{
|
||||
$boundary = null;
|
||||
foreach ($structure->attributes as $key => $value) {
|
||||
if (strtolower($key) === 'boundary') {
|
||||
$boundary = $value;
|
||||
break;
|
||||
}
|
||||
}
|
||||
|
||||
if ($boundary === null) {
|
||||
throw new \RuntimeException('Multipart missing boundary attribute');
|
||||
}
|
||||
|
||||
$chunks = $this->splitOnBoundary($data, $boundary);
|
||||
|
||||
$parts = [];
|
||||
foreach ($structure->parts as $i => $childStructure) {
|
||||
$chunk = $chunks[$i] ?? '';
|
||||
$chunk = $this->stripPartHeaders($chunk);
|
||||
$id = empty($partId) ? (string)($i + 1) : $partId . '.' . ($i + 1);
|
||||
$parts = array_merge($parts, $this->fetchBodyNode($childStructure, $chunk, $id));
|
||||
}
|
||||
|
||||
return $parts;
|
||||
}
|
||||
|
||||
/**
|
||||
* Split $raw on MIME boundary delimiter lines, returning one string per
|
||||
* body part. The preamble (before the first delimiter) and epilogue
|
||||
* (after the close delimiter) are discarded.
|
||||
*
|
||||
* @return string[]
|
||||
*/
|
||||
private function splitOnBoundary(string $raw, string $boundary): array
|
||||
{
|
||||
$delimiter = '--' . $boundary;
|
||||
$closeDelimiter = '--' . $boundary . '--';
|
||||
|
||||
$parts = [];
|
||||
$current = null;
|
||||
|
||||
// Handle both CRLF and bare-LF line endings
|
||||
$lines = preg_split('/\r?\n/', $raw);
|
||||
|
||||
foreach ($lines as $line) {
|
||||
$trimmed = rtrim($line);
|
||||
|
||||
if ($trimmed === $closeDelimiter) {
|
||||
if ($current !== null) {
|
||||
$parts[] = rtrim($current, "\r\n");
|
||||
}
|
||||
break;
|
||||
}
|
||||
|
||||
if ($trimmed === $delimiter) {
|
||||
if ($current !== null) {
|
||||
$parts[] = rtrim($current, "\r\n");
|
||||
}
|
||||
$current = '';
|
||||
continue;
|
||||
}
|
||||
|
||||
if ($current !== null) {
|
||||
$current .= $line . "\r\n";
|
||||
}
|
||||
// Lines before the first delimiter are preamble — ignored
|
||||
}
|
||||
|
||||
// If the close delimiter was absent, flush whatever is buffered
|
||||
if ($current !== null && $current !== '') {
|
||||
$trimmed = rtrim($current, "\r\n");
|
||||
if (!in_array($trimmed, $parts, true)) {
|
||||
$parts[] = $trimmed;
|
||||
}
|
||||
}
|
||||
|
||||
return $parts;
|
||||
}
|
||||
|
||||
/**
|
||||
* Strip MIME part headers from a body chunk.
|
||||
*
|
||||
* Each part chunk begins with its own headers (Content-Type,
|
||||
* Content-Transfer-Encoding, etc.) followed by a blank line.
|
||||
* Since BODYSTRUCTURE already supplies all encoding/charset info,
|
||||
* we discard the part headers and return the raw body bytes only.
|
||||
*/
|
||||
private function stripPartHeaders(string $raw): string
|
||||
{
|
||||
// Try CRLF blank line first, then bare LF
|
||||
$crlfPos = strpos($raw, "\r\n\r\n");
|
||||
$lfPos = strpos($raw, "\n\n");
|
||||
|
||||
if ($crlfPos !== false && ($lfPos === false || $crlfPos <= $lfPos)) {
|
||||
return substr($raw, $crlfPos + 4);
|
||||
}
|
||||
|
||||
if ($lfPos !== false) {
|
||||
return substr($raw, $lfPos + 2);
|
||||
}
|
||||
|
||||
return $raw;
|
||||
}
|
||||
|
||||
/**
|
||||
* @throws ParseError
|
||||
*/
|
||||
@@ -1005,13 +1148,48 @@ readonly class Parser
|
||||
return $raw;
|
||||
}
|
||||
|
||||
for ($i = 0; $i < strlen($raw); $i++) {
|
||||
$character = $raw[$i];
|
||||
if (!mb_check_encoding($character, 'US-ASCII')) {
|
||||
$raw[$i] = ' ';
|
||||
$result = '';
|
||||
$pos = 0;
|
||||
$len = strlen($raw);
|
||||
|
||||
while ($pos < $len) {
|
||||
if (preg_match('/\{(\d+)\}\r\n/', $raw, $m, PREG_OFFSET_CAPTURE, $pos)) {
|
||||
$braceOff = (int) $m[0][1];
|
||||
$literalLen = (int) $m[1][0];
|
||||
$headerLen = strlen($m[0][0]);
|
||||
|
||||
// Sanitize structural text that precedes this literal
|
||||
$result .= $this->sanitizeChunk(substr($raw, $pos, $braceOff - $pos));
|
||||
|
||||
// Preserve the {N}\r\n marker verbatim
|
||||
$result .= $m[0][0];
|
||||
|
||||
// Preserve the literal body bytes verbatim (may be UTF-8 / 8-bit)
|
||||
$result .= substr($raw, $braceOff + $headerLen, $literalLen);
|
||||
|
||||
$pos = $braceOff + $headerLen + $literalLen;
|
||||
} else {
|
||||
// No more literals — sanitize the remainder
|
||||
$result .= $this->sanitizeChunk(substr($raw, $pos));
|
||||
break;
|
||||
}
|
||||
}
|
||||
|
||||
return $raw;
|
||||
return $result;
|
||||
}
|
||||
|
||||
private function sanitizeChunk(string $chunk): string
|
||||
{
|
||||
if (mb_check_encoding($chunk, 'US-ASCII')) {
|
||||
return $chunk;
|
||||
}
|
||||
|
||||
for ($i = 0, $len = strlen($chunk); $i < $len; $i++) {
|
||||
if (!mb_check_encoding($chunk[$i], 'US-ASCII')) {
|
||||
$chunk[$i] = ' ';
|
||||
}
|
||||
}
|
||||
|
||||
return $chunk;
|
||||
}
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user