Fix GH-12532: PharData created from zip has incorrect timestamp

The datetime stored in the DOS time fields, which is what zip standard
uses, is local time without a timezone. There's an extension to the zip
file format since '97 that allows storing a unix timestamp (in UTC) in
the header for both the central directory and the local entries.
This patch adds support for this.

Closes GH-12548.

Author: nielsdos

NEWS

@@ -24,6 +24,10 @@ PDO_PGSQL:
 PGSQL:
   . Added the possibility to have no conditions for pg_select. (OmarEmaraDev)
 
+Phar:
+  . Fixed bug GH-12532 (PharData created from zip has incorrect timestamp).
+    (nielsdos)
+
 SimpleXML:
   . Fixed bug GH-12192 (SimpleXML infinite loop when getName() is called
     within foreach). (nielsdos)

UPGRADING

@@ -79,6 +79,9 @@ PHP 8.4 UPGRADE NOTES
   . Added constant DOMNode::DOCUMENT_POSITION_CONTAINED_BY.
   . Added constant DOMNode::DOCUMENT_POSITION_IMPLEMENTATION_SPECIFIC.
 
+- Phar:
+  . Added support for the unix timestamp extension for zip archives.
+
 - SOAP:
   . Added support for clark notation for namespaces in class map.
     It is now possible to specify entries in a class map with clark notation

ext/phar/pharzip.h

@@ -146,6 +146,13 @@ typedef struct _phar_zip_unix3 {
 /* (var.)        variable    symbolic link filename */
 } phar_zip_unix3;
 
+/* See https://libzip.org/specifications/extrafld.txt */
+typedef struct _phar_zip_unix_time {
+    phar_zip_extra_field_header header;
+    char flags;                 /* flags                            1 byte  */
+    char time[4];               /* time in standard Unix format     4 bytes	*/
+} phar_zip_unix_time;
+
 typedef struct _phar_zip_central_dir_file {
 	char signature[4];            /* central file header signature   4 bytes  (0x02014b50) */
 	char madeby[2];               /* version made by                 2 bytes */

ext/phar/tests/gh12532.phpt

@@ -0,0 +1,22 @@
+--TEST--
+GH-12532 (PharData created from zip has incorrect timestamp)
+--EXTENSIONS--
+phar
+--FILE--
+<?php
+
+date_default_timezone_set('UTC');
+$phar = new PharData(__DIR__ . '/gh12532.zip');
+echo $phar->getMTime(), "\n";
+echo $phar->getFileInfo()->getMTime(), "\n";
+echo date('Y-m-d H:i:s', $phar->getMTime()), "\n";
+echo date('Y-m-d H:i:s', $phar->getCTime()), "\n";
+echo date('Y-m-d H:i:s', $phar->getATime()), "\n";
+
+?>
+--EXPECT--
+1680284661
+1680284661
+2023-03-31 17:44:21
+2023-03-31 17:44:21
+2023-03-31 17:44:21

ext/phar/tests/gh12532.zip

@@ -44,6 +44,7 @@ static int phar_zip_process_extra(php_stream *fp, phar_entry_info *entry, uint16
 	union {
 		phar_zip_extra_field_header header;
 		phar_zip_unix3 unix3;
+		phar_zip_unix_time time;
 	} h;
 	size_t read;
 
@@ -52,6 +53,35 @@ static int phar_zip_process_extra(php_stream *fp, phar_entry_info *entry, uint16
 			return FAILURE;
 		}
 
+		if (h.header.tag[0] == 'U' && h.header.tag[1] == 'T') {
+			/* Unix timestamp header found.
+			 * The flags field indicates which timestamp fields are present.
+			 * The size with a timestamp is at least 5 (== 9 - tag size) bytes, but may be larger.
+			 * We only store the modification time in the entry, so only read that.
+			 */
+			const size_t min_size = 5;
+			uint16_t header_size = PHAR_GET_16(h.header.size);
+			if (header_size >= min_size) {
+				read = php_stream_read(fp, &h.time.flags, min_size);
+				if (read != min_size) {
+					return FAILURE;
+				}
+				if (h.time.flags & (1 << 0)) {
+					/* Modification time set */
+					entry->timestamp = PHAR_GET_32(h.time.time);
+				}
+
+				len -= header_size + 4;
+
+				/* Consume remaining bytes */
+				if (header_size != read) {
+					php_stream_seek(fp, header_size - read, SEEK_CUR);
+				}
+				continue;
+			}
+			/* Fallthrough to next if to skip header */
+		}
+
 		if (h.header.tag[0] != 'n' || h.header.tag[1] != 'u') {
 			/* skip to next header */
 			php_stream_seek(fp, PHAR_GET_16(h.header.size), SEEK_CUR);