Update libmagic to version 5.45 (TileDB-Inc#4673)

[SC-38521](https://app.shortcut.com/tiledb-inc/story/38521/update-libmagic-and-use-the-upstream-vcpkg-port)

Split from TileDB-Inc#4553.

This PR updates libmagic to version 5.45 and switches to using a vcpkg
port closer to the upstream one, which we can easily consume with
find_package(unofficial-libmagic) since microsoft/vcpkg#35274.

One complication is that the upstream port builds libmagic with its
official autotools-based build system, which is significantly slower on
Windows (on Linux it builds pretty fast). I tried to upstream the
CMake-based port I had added in TileDB-Inc#4119, but the PR was rejected.

Apart from binary caching, there is unfortunately nothing we can do
about the build performance regression. We could maintain the
CMake-based port for our own use, but it will split what we build and
what a future user of TileDB from vcpkg will build, and that port is not
without its problems (it failed for example when I tried cross-compiling
to arm64-windows, because it tried to execute the arm64 file.exe on my
x64 machine).

---
TYPE: BUILD
DESC: Update libmagic to version 5.45

.readthedocs.yml

@@ -12,6 +12,17 @@ build:
   os: "ubuntu-22.04"
   tools:
     python: "3.8"
+  apt_packages:
+    - autoconf
+    - autoconf-archive
+    - automake
+    - curl
+    - git
+    - libtool
+    - make
+    - pkg-config
+    - unzip
+    - zip
 
 python:
   install:

doc/dev/BUILD.md

@@ -12,6 +12,7 @@ title: Building TileDB from source
         * Apple Clang 14
 * Git (required by vcpkg)
 * curl (required by vcpkg on non-Windows)
+* autoconf (required by building libmagic on non-Windows)
 
 ## Downloading the source code
 

ports/README.md

@@ -30,7 +30,6 @@ After copying the port, add an entry to the table below. You should also contrib
 
 |Port|Reason|
 |----|------|
-|`libmagic`|Updating to the upstream port deferred due to failures.|
 |`openssl`|Pinning to OpenSSL 1.1 until we can move to 3.0 in January 2024.|
-|`pcre2`|To be removed alongside libmagic.|
 |`azure-storage-common-cpp`|Patching to disable default features on libxml2 (https://github.com/Azure/azure-sdk-for-cpp/pull/5221).|
+|`libmagic`|Patching to add features explicitly enabling compression support.|

ports/libmagic/0001-Use-pcre2.patch → ports/libmagic/0001-Use-libtre.patch

@@ -44,7 +44,7 @@ index c548e97..299ac0c 100644
  #include <errno.h>
  #include <fcntl.h>	/* For open and flags */
 -#include <regex.h>
-+#include <pcre2posix.h>
++#include <tre/regex.h>
  #include <time.h>
  #include <sys/types.h>
  #ifndef WIN32

ports/libmagic/0002-Change-zlib-lib-name-to-match-CMake-output.patch

@@ -0,0 +1,39 @@
+From 2fa43ece9ec7564e1fbb9867bb5852b834643aa4 Mon Sep 17 00:00:00 2001
+From: Long Nguyen <[email protected]>
+Date: Sat, 8 May 2021 19:36:11 +0700
+Subject: [PATCH 02/14] Change zlib lib name to match CMake output
+
+---
+ configure.ac | 6 +++---
+ 1 file changed, 3 insertions(+), 3 deletions(-)
+
+diff --git a/configure.ac b/configure.ac
+index b05c334..dd4063c 100644
+--- a/configure.ac
++++ b/configure.ac
+@@ -173,7 +173,7 @@ AC_REPLACE_FUNCS(getopt_long asprintf vasprintf strlcpy strlcat getline ctime_r
+
+ dnl Checks for libraries
+ if test "$enable_zlib" != "no"; then
+-  AC_CHECK_LIB(z, gzopen)
++  AC_SEARCH_LIBS(gzopen, [z zlib zlibd], have_zlib = "yes", have_zlib = "no")
+ fi
+ if test "$enable_bzlib" != "no"; then
+   AC_CHECK_LIB(bz2, BZ2_bzCompressInit)
+@@ -193,11 +193,11 @@ AM_CONDITIONAL(IS_CROSS_COMPILE, test "$cross_compiling" = yes)
+
+ dnl Final sanity checks
+ if test "$enable_zlib" = "yes"; then
+-  if test "$ac_cv_header_zlib_h$ac_cv_lib_z_gzopen" != "yesyes"; then
++  if test "$ac_cv_header_zlib_h$have_zlib" != "yesyes"; then
+     AC_MSG_ERROR([zlib support requested but not found])
+   fi
+ fi
+-if  test "$ac_cv_header_zlib_h$ac_cv_lib_z_gzopen" = "yesyes"; then
++if  test "$ac_cv_header_zlib_h$have_zlib" = "yesyes"; then
+   AC_DEFINE([ZLIBSUPPORT], 1, [Enable zlib compression support])
+ fi
+ if test "$enable_bzlib" = "yes"; then
+-- 
+2.29.2.windows.2
+

ports/libmagic/0003-Fix-WIN32-macro-checks.patch

@@ -12,7 +12,7 @@ index 299ac0c..2c365a6 100644
 --- a/src/file.h
 +++ b/src/file.h
 @@ -82,7 +82,7 @@
- #include <pcre2posix.h>
+ #include <tre/regex.h>
  #include <time.h>
  #include <sys/types.h>
 -#ifndef WIN32
@@ -31,13 +31,13 @@ index 299ac0c..2c365a6 100644
  #define PATHSEP	':'
 @@ -103,7 +103,7 @@
 
- #define private static
+ #define file_private static
 
 -#if HAVE_VISIBILITY && !defined(WIN32)
 +#if HAVE_VISIBILITY && !defined(_WIN32)
- #define public  __attribute__ ((__visibility__("default")))
- #ifndef protected
- #define protected __attribute__ ((__visibility__("hidden")))
+ #define file_public  __attribute__ ((__visibility__("default")))
+ #ifndef file_protected
+ #define file_protected __attribute__ ((__visibility__("hidden")))
 -- 
 2.29.2.windows.2
 

ports/libmagic/0005-Include-dirent.h-for-S_ISREG-and-S_ISDIR.patch

@@ -11,14 +11,15 @@ diff --git a/src/file.h b/src/file.h
 index 0332506..4aa9f60 100644
 --- a/src/file.h
 +++ b/src/file.h
-@@ -88,6 +88,7 @@
+@@ -88,7 +88,8 @@
  /* Do this here and now, because struct stat gets re-defined on solaris */
  #include <sys/stat.h>
  #include <stdarg.h>
 +#include <dirent.h>
-
- #define ENABLE_CONDITIONALS
-
+ #include <locale.h>
+ #if defined(HAVE_XLOCALE_H)
+ #include <xlocale.h>
+ #endif
 -- 
 2.29.2.windows.2
 

ports/libmagic/0006-Remove-Wrap-POSIX-headers.patch

@@ -161,7 +161,7 @@ index 3ab52d1..fc48d84 100644
 --- a/src/pread.c
 +++ b/src/pread.c
 @@ -3,7 +3,9 @@
- FILE_RCSID("@(#)$File: pread.c,v 1.2 2013/04/02 16:23:07 christos Exp $")
+ FILE_RCSID("@(#)$File: pread.c,v 1.5 2022/09/24 20:30:13 christos Exp $")
  #endif  /* lint */
  #include <fcntl.h>
 +#ifdef HAVE_UNISTD_H

ports/libmagic/0010-Properly-check-for-the-presence-of-bitmasks.patch

@@ -17,7 +17,7 @@ index 5204f20..7244841 100644
  #undef HAVE_MAJOR
 -#ifdef	S_IFLNK
 +#if S_IFLNK != 0
- private int
+ file_private int
  bad_link(struct magic_set *ms, int err, char *buf)
  {
 @@ -108,7 +108,7 @@ file_fsmagic(struct magic_set *ms, const char *fn, struct stat *sb)

ports/libmagic/0011-Remove-pipe-related-functions-in-funcs.c.patch

@@ -11,21 +11,19 @@ diff --git a/src/funcs.c b/src/funcs.c
 index b926625..b585486 100644
 --- a/src/funcs.c
 +++ b/src/funcs.c
-@@ -809,6 +809,7 @@ file_print_guid(char *str, size_t len, const uint64_t *guid)
- 	    g->data4[6], g->data4[7]);
+@@ -888,5 +888,6 @@
  }
 
 +#ifndef _WIN32
- protected int
+ file_protected int
  file_pipe_closexec(int *fds)
  {
-@@ -827,6 +828,7 @@ protected int
- file_clear_closexec(int fd) {
- 	return fcntl(fd, F_SETFD, 0);
+@@ -914,5 +915,6 @@
+ #endif
  }
 +#endif
 
- protected char *
+ file_protected char *
  file_strtrim(char *str)
 -- 
 2.29.2.windows.2

ports/libmagic/0012-Convert-MSYS2-paths-to-Windows-paths.patch

@@ -0,0 +1,24 @@
+From f0139ced57c1579450f3d09b6e3ae0159aae031b Mon Sep 17 00:00:00 2001
+From: Long Nguyen <[email protected]>
+Date: Mon, 10 May 2021 08:43:28 +0700
+Subject: [PATCH 12/14] Convert MSYS2 paths to Windows paths
+
+---
+ src/Makefile.am | 3 +++
+ 1 file changed, 3 insertions(+)
+
+diff --git a/src/Makefile.am b/src/Makefile.am
+index 34781b9..26f853f 100644
+--- a/src/Makefile.am
++++ b/src/Makefile.am
+@@ -1,4 +1,7 @@
+ MAGIC = $(pkgdatadir)/magic
++if MINGW
++MAGIC != cygpath -m -a -l '$(MAGIC)'
++endif
+ lib_LTLIBRARIES = libmagic.la
+ nodist_include_HEADERS = magic.h
+
+-- 
+2.29.2.windows.2
+

ports/libmagic/0013-Check-for-backslash-in-argv-0-on-Windows.patch

@@ -0,0 +1,42 @@
+From 3f10c7d31627b64b068b84ba72e706991f672560 Mon Sep 17 00:00:00 2001
+From: Long Nguyen <[email protected]>
+Date: Fri, 14 May 2021 08:14:05 +0700
+Subject: [PATCH 13/14] Check for backslash in argv[0] on Windows
+
+---
+ magic/Makefile.am | 2 +-
+ src/file.c        | 5 +++++
+ 2 files changed, 6 insertions(+), 1 deletion(-)
+
+diff --git a/magic/Makefile.am b/magic/Makefile.am
+index 0eb5865..170bbab 100644
+--- a/magic/Makefile.am
++++ b/magic/Makefile.am
+@@ -353,7 +353,7 @@ ${MAGIC}: $(EXTRA_DIST) $(FILE_COMPILE_DEP)
+ 	@(if expr "${FILE_COMPILE}" : '.*/.*' > /dev/null; then \
+ 	    echo "Using ${FILE_COMPILE} to generate ${MAGIC}" > /dev/null; \
+ 	  else \
+-	    v=$$(${FILE_COMPILE} --version | sed -e s/file-// -e q); \
++	    v=$$(${FILE_COMPILE} --version | sed -e s/file${EXEEXT}-// -e q); \
+ 	    if [ "$$v" != "${PACKAGE_VERSION}" ]; then \
+ 		echo "Cannot use the installed version of file ($$v) to"; \
+ 		echo "cross-compile file ${PACKAGE_VERSION}"; \
+diff --git a/src/file.c b/src/file.c
+index 2889f8a..12a604b 100644
+--- a/src/file.c
++++ b/src/file.c
+@@ -200,6 +200,11 @@ main(int argc, char *argv[])
+ 	_wildcard(&argc, &argv);
+ #endif
+
++#ifdef _WIN32
++	if ((progname = strrchr(argv[0], '\\')) != NULL)
++		progname++;
++	else
++#endif
+ 	if ((progname = strrchr(argv[0], '/')) != NULL)
+ 		progname++;
+ 	else
+-- 
+2.29.2.windows.2
+

ports/libmagic/0015-MSYS2-Remove-ioctl-call.patch

@@ -7,20 +7,20 @@ diff --git a/src/compress.c b/src/compress.c
 index 33ce2bc..f172eda 100644
 --- a/src/compress.c
 +++ b/src/compress.c
-@@ -378,7 +378,7 @@
+@@ -407,7 +407,7 @@
  sread(int fd, void *buf, size_t n, int canbepipe __attribute__((__unused__)))
  {
  	ssize_t rv;
--#ifdef FIONREAD
+-#if defined(FIONREAD) && !defined(__MINGW32__)
 +#if defined(FIONREAD) && !defined(__MINGW32__) && !defined(WIN32)
  	int t = 0;
  #endif
  	size_t rn = n;
-@@ -386,7 +386,7 @@
+@@ -418,7 +418,7 @@
  	if (fd == STDIN_FILENO)
  		goto nocheck;
 
--#ifdef FIONREAD
+-#if defined(FIONREAD) && !defined(__MINGW32__)
 +#if defined(FIONREAD) && !defined(__MINGW32__) && !defined(WIN32)
  	if (canbepipe && (ioctl(fd, FIONREAD, &t) == -1 || t == 0)) {
  #ifdef FD_ZERO

ports/libmagic/0016-Fix-file_famagic-function.patch

@@ -0,0 +1,40 @@
+diff --git a/src/fsmagic.c b/src/fsmagic.c
+index 7244841..2c553c1 100644
+--- a/src/fsmagic.c
++++ b/src/fsmagic.c
+@@ -66,7 +66,7 @@ # define major(dev)  (((dev) >> 8) & 0xff)
+ # define minor(dev)  ((dev) & 0xff)
+ #endif
+ #undef HAVE_MAJOR
+-#if S_IFLNK != 0
++#if S_IFLNK != 0 && ! defined(_WIN32)
+ file_private int
+ bad_link(struct magic_set *ms, int err, char *buf)
+ {
+@@ -108,7 +108,7 @@ file_fsmagic(struct magic_set *ms, const char *fn, struct stat *sb)
+ 	int ret, did = 0;
+ 	int mime = ms->flags & MAGIC_MIME;
+ 	int silent = ms->flags & (MAGIC_APPLE|MAGIC_EXTENSION);
+-#if S_IFLNK != 0
++#if S_IFLNK != 0 && ! defined(_WIN32)
+ 	char buf[BUFSIZ+4];
+ 	ssize_t nch;
+ 	struct stat tstatbuf;
+@@ -122,7 +122,7 @@ file_fsmagic(struct magic_set *ms, const char *fn, struct stat *sb)
+ 	 * Fstat is cheaper but fails for files you don't have read perms on.
+ 	 * On 4.2BSD and similar systems, use lstat() to identify symlinks.
+ 	 */
+-#if S_IFLNK != 0
++#if S_IFLNK != 0 && ! defined(_WIN32)
+ 	if ((ms->flags & MAGIC_SYMLINK) == 0)
+ 		ret = lstat(fn, sb);
+ 	else
+@@ -290,7 +290,7 @@ file_fsmagic(struct magic_set *ms, const char *fn, struct stat *sb)
+ 			return -1;
+ 		break;
+ #endif
+-#if S_IFLNK != 0
++#if S_IFLNK != 0 && ! defined(_WIN32)
+ 	case S_IFLNK:
+ 		if ((nch = readlink(fn, buf, BUFSIZ-1)) <= 0) {
+ 			if (ms->flags & MAGIC_ERROR) {