Upstream status: submitted to bug-libextractor@gnu.org.

From 1fc6daaeaf829fb941a176831c011888a73c43b9 Mon Sep 17 00:00:00 2001
From: Maxim Cournoyer <maxim.cournoyer@gmail.com>
Date: Mon, 11 Mar 2024 09:36:26 -0400
Subject: [PATCH] html_extractor: Add support for modern tidy-html.

* configure.ac: Use PKG_PROG_PKG_CONFIG to initialize pkg-config detection.
<tidy>: Check for library via pkg-config.
* src/plugins/html_extractor.c: Standardize tidy include file names.
---
 configure.ac                 | 28 +++++++++-------------------
 src/plugins/html_extractor.c |  4 ++--
 2 files changed, 11 insertions(+), 21 deletions(-)

diff --git a/configure.ac b/configure.ac
index d17ff39..e89d70c 100644
--- a/configure.ac
+++ b/configure.ac
@@ -176,6 +176,8 @@ AS_CASE(["$target_os"],
 
 AM_ICONV
 
+PKG_PROG_PKG_CONFIG()
+
 # We define the paths here, because MinGW/GCC expands paths
 # passed through the command line ("-DLOCALEDIR=..."). This would
 # lead to hard-coded paths ("C:\mingw\mingw\bin...") that do
@@ -424,25 +426,13 @@ AC_CHECK_LIB(magic, magic_open,
    AM_CONDITIONAL(HAVE_MAGIC, false))],
   AM_CONDITIONAL(HAVE_MAGIC, false))
 
-AC_MSG_CHECKING(for tidyNodeGetValue -ltidy)
-AC_LANG_PUSH(C++)
-SAVED_LIBS=$LIBS
-LIBS="$LIBS -ltidy"
-AC_LINK_IFELSE(
-  [AC_LANG_PROGRAM([[#include <tidy/tidy.h>]],
-    [[ Bool b = tidyNodeGetValue (NULL, NULL, NULL); ]])],
-  [AC_MSG_RESULT(yes)
-   AM_CONDITIONAL(HAVE_TIDY, true)
-   AC_DEFINE(HAVE_TIDY,1,[Have tidyNodeGetValue in libtidy])],
-  [AC_MSG_RESULT(no)
-   AM_CONDITIONAL(HAVE_TIDY, false)])
-LIBS=$SAVED_LIBS
-AC_LANG_POP(C++)
-
-# restore LIBS
-LIBS=$LIBSOLD
-
-
+dnl tidyNodeGetValue was already available in 5.0.0, released in 2015.
+PKG_CHECK_MODULES([TIDY], [tidy >= 5.0.0],
+ [AC_DEFINE(HAVE_TIDY, 1, [Have tidy])
+  AM_CONDITIONAL(HAVE_TIDY, true)],
+ [AM_CONDITIONAL(HAVE_TIDY, false)])
+CFLAGS="$CFLAGS $TIDY_CFLAGS"
+LIBS="$LIBS $TIDY_LIBS"
 
 # should 'make check' run tests?
 AC_MSG_CHECKING(whether to run tests)
diff --git a/src/plugins/html_extractor.c b/src/plugins/html_extractor.c
index 5ebf97b..88100d3 100644
--- a/src/plugins/html_extractor.c
+++ b/src/plugins/html_extractor.c
@@ -26,8 +26,8 @@
 #include "platform.h"
 #include "extractor.h"
 #include <magic.h>
-#include <tidy/tidy.h>
-#include <tidy/tidybuffio.h>
+#include <tidy.h>
+#include <tidybuffio.h>
 
 /**
  * Mapping of HTML META names to LE types.

base-commit: a75f40b64b5868967c95ea214e8eaac4f7088b23
-- 
2.41.0