{"id":11363,"date":"2011-02-28T07:00:00","date_gmt":"2011-02-28T07:00:00","guid":{"rendered":"https:\/\/blogs.msdn.microsoft.com\/oldnewthing\/2011\/02\/28\/if-you-want-to-use-guids-to-identify-your-files-then-nobodys-stopping-you\/"},"modified":"2011-02-28T07:00:00","modified_gmt":"2011-02-28T07:00:00","slug":"if-you-want-to-use-guids-to-identify-your-files-then-nobodys-stopping-you","status":"publish","type":"post","link":"https:\/\/devblogs.microsoft.com\/oldnewthing\/20110228-00\/?p=11363","title":{"rendered":"If you want to use GUIDs to identify your files, then nobody's stopping you"},"content":{"rendered":"

\nIgor Levicki proposes solving the problem of file extensions by\n\nusing a GUID instead of a file name to identify a file<\/a>.\n<\/p>\n

\nYou can do this already.\nEvery file on an NTFS volume has an object identifier<\/i>\nwhich is formally 16-byte buffer, but let’s just call it a GUID.\nBy default a file doesn’t have an object identifier,\nbut you can ask for one to be created with\nFSCTL_CREATE_OR_GET_OBJECT_ID<\/code>,\nwhich will retrieve the existing object identifier associated with\na file, or create one if there isn’t one already.\nIf you are a control freak, you can use\nFSCTL_SET_OBJECT_ID<\/code> to specify the GUID you want to use\nas the object identifier.\n(The call fails if the file already has an object identifier.)\nAnd of course there is\nFSCTL_GET_OBJECT_ID<\/code> to retrieve the object identifier, if any.\n<\/p>\n

\n#define UNICODE\n#define _UNICODE\n#include <windows.h>\n#include <stdio.h>\n#include <tchar.h>\n#include <ole2.h>\n#include <winioctl.h>\nint __cdecl _tmain(int argc, PTSTR *argv)\n{\n HANDLE h = CreateFile(argv[1], 0,\n                 FILE_SHARE_READ | FILE_SHARE_WRITE |\n                 FILE_SHARE_DELETE, NULL,\n                 OPEN_EXISTING, 0, NULL);\n if (h != INVALID_HANDLE_VALUE) {\n  FILE_OBJECTID_BUFFER buf;\n  DWORD cbOut;\n  if (DeviceIoControl(h, FSCTL_CREATE_OR_GET_OBJECT_ID,\n                 NULL, 0, &buf, sizeof(buf),\n                 &cbOut, NULL)) {\n    GUID guid;\n    CopyMemory(&guid, &buf.ObjectId, sizeof(GUID));\n    WCHAR szGuid[39];\n    StringFromGUID2(guid, szGuid, 39);\n    _tprintf(_T(\"GUID is %ws\\n\"), szGuid);\n  }\n  CloseHandle(h);\n }\n return 0;\n}\n<\/pre>\n
\nThis program takes a file or directory name\nas its sole parameter and prints the associated object identifier.\n<\/p>\n
\nBig deal, now we have a GUID associated with each file.\n<\/p>\n
\nThe other half is, of course, using this GUID to open the file:\n<\/p>\n
\n#define UNICODE\n#define _UNICODE\n#include <windows.h>\n#include <stdio.h>\n#include <tchar.h>\n#include <ole2.h>\nint __cdecl _tmain(int argc, PTSTR *argv)\n{\n HANDLE hRoot = CreateFile(_T(\"C:\\\\\"), 0,\n                 FILE_SHARE_READ | FILE_SHARE_WRITE |\n                 FILE_SHARE_DELETE, NULL,\n                 OPEN_EXISTING,\n                 FILE_FLAG_BACKUP_SEMANTICS, NULL);\n if (hRoot != INVALID_HANDLE_VALUE) {\n  FILE_ID_DESCRIPTOR desc;\n  desc.dwSize = sizeof(desc);\n  desc.Type = ObjectIdType;\n  if (SUCCEEDED(CLSIDFromString(argv[1], &desc.ObjectId))) {\n   HANDLE h = OpenFileById(hRoot, &desc, GENERIC_READ,\n                 FILE_SHARE_READ | FILE_SHARE_WRITE |\n                 FILE_SHARE_DELETE, NULL, 0);\n   if (h != INVALID_HANDLE_VALUE) {\n    BYTE b;\n    DWORD cb;\n    if (ReadFile(h, &b, 1, &cb, NULL)) {\n     _tprintf(_T(\"First byte of file is 0x%02x\\n\"), b);\n    }\n    CloseHandle(h);\n   }\n  }\n  CloseHandle(hRoot);\n }\n return 0;\n}\n<\/pre>\n
\nTo open a file by its GUID, you first need to open\nsomething—anything—on the volume the file resides on.\nDoesn’t matter what you open;\nthe only reason for having this handle is so that\nOpenFileById<\/code> knows which volume you’re talking about.\nIn our little test program, we use the C: drive,\nwhich means that the file search will take place on the\nC: drive.\n<\/p>\n
\nNext, you fill in the FILE_ID_DESCRIPTOR<\/code>,\nsaying that you want to open the file by its object identifier,\nand then it’s off to the races with OpenFileById<\/code>.\nJust as a proof of concept, we read and print the first byte\nof the file that was opened as a result.\n<\/p>\n
\nNotice that the file you open by its object identifier does not\nhave to be in the current directory.\nIt can be anywhere on the C: drive<\/i>.\nAs long as you have the GUID for a file, you can open it\nno matter where it is on the drive.\n<\/p>\n
\nYou can run these two programs just to enjoy the thrill of\nopening a file by its GUID.\nNotice that once you get the GUID for a file, you can move\nit anywhere on the drive, and OpenFileById<\/code>\nwill still open it.\n<\/p>\n
\n(And if you want to get rid of those pesky drive letters,\nyou can use the volume GUID instead.\nNow every file is identified by a pair of GUIDs:\nthe volume GUID and the object identifier.)\n<\/p>\n
\nSo Igor’s dream world where all files are referenced by GUID\nalready exists.\nWhy isn’t everybody switching over to this utopia of GUID-based\nfile identification?\n<\/p>\n
\nYou probably know the answer already:\nBecause people prefer to name things with something mnemonic\nrather than a GUID.\nImagine a file open dialog in this dream world.\n“Enter the GUID of the file you wish to open, or click Browse\nto see the GUIDs of all the files on this volume so you can pick\nfrom a list.”\nHow long would this dialog survive?\n<\/p>\n
\nFor today, you don’t have to call me Raymond.\nYou can call me\n\n{7ecf65a0-4b78-5f9b-e77c-8770091c0100}<\/a>,\nor “91c” for short.\n<\/p>\n
\n(And I’ve totally ignored the fact that\nusing GUIDs to identify files does nothing\nto solve the problem of trying to figure out what program should be\nused to open a particular file.)\n<\/p>\n
\nBonus chatter<\/b>:\nYou can also open files by their file identifer, which is a\nvolume-specific 64-bit value.\nBut I chose to use the GUID both for the extra challenge,\nand just to show that Igor’s dream world already exists.<\/p>\n","protected":false},"excerpt":{"rendered":"
Igor Levicki proposes solving the problem of file extensions by using a GUID instead of a file name to identify a file. You can do this already. Every file on an NTFS volume has an object identifier which is formally 16-byte buffer, but let’s just call it a GUID. By default a file doesn’t have […]<\/p>\n","protected":false},"author":1069,"featured_media":111744,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[1],"tags":[26],"class_list":["post-11363","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-oldnewthing","tag-other"],"acf":[],"blog_post_summary":"
Igor Levicki proposes solving the problem of file extensions by using a GUID instead of a file name to identify a file. You can do this already. Every file on an NTFS volume has an object identifier which is formally 16-byte buffer, but let’s just call it a GUID. By default a file doesn’t have […]<\/p>\n","_links":{"self":[{"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/posts\/11363","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/users\/1069"}],"replies":[{"embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/comments?post=11363"}],"version-history":[{"count":0,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/posts\/11363\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/media\/111744"}],"wp:attachment":[{"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/media?parent=11363"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/categories?post=11363"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/tags?post=11363"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}