CI: run xmllint against docs in the sanity workflow

.github/workflows/sanity.yml

@@ -8,8 +8,6 @@ on:
   push:
     branches: [ "*" ]
     paths-ignore:
-      - 'docs/**'
-      - '!docs/log-message-tags/**'
       - STATUS
       - CHANGES
       - README*
@@ -18,8 +16,6 @@ on:
   pull_request:
     branches: [ "trunk", "2.4.x" ]
     paths-ignore:
-      - 'docs/**'
-      - '!docs/log-message-tags/**'
       - STATUS
       - CHANGES
       - README*
@@ -54,3 +50,14 @@ jobs:
            sudo apt-get install -o Acquire::Retries=5 libapr1-dev libaprutil1-dev
     - name: Check for autoconf warnings
       run: ./buildconf --with-apr=/usr/bin/apr-1-config --with-errors
+
+  docs:
+    runs-on: ubuntu-latest
+    timeout-minutes: 5
+    steps:
+    - uses: actions/checkout@v6
+    - name: Install prerequisites
+      run: sudo apt-get -o Acquire::Retries=5 update &&
+           sudo apt-get install -o Acquire::Retries=5 libxml2-utils w3c-sgml-lib
+    - name: Check docs XML
+      run: ./test/gha-docs-check.sh

docs/manual/rewrite/avoid.xml.ko

@@ -1,4 +1,4 @@
-<?xml version='1.0' encoding='EUC-KR' ?>
+<?xml version='1.0' encoding='UTF-8' ?>
 <!DOCTYPE manualpage SYSTEM "../style/manualpage.dtd">
 <?xml-stylesheet type="text/xsl" href="../style/manual.ko.xsl"?>
 <!-- English Revision: 1933071:1938110 (outdated) -->

docs/manual/rewrite/flags.xml.ko

@@ -1,4 +1,4 @@
-<?xml version='1.0' encoding='EUC-KR' ?>
+<?xml version='1.0' encoding='UTF-8' ?>
 <!DOCTYPE manualpage SYSTEM "../style/manualpage.dtd">
 <?xml-stylesheet type="text/xsl" href="../style/manual.ko.xsl"?>
 <!-- English Revision: 1933062:1936255 (outdated) -->

docs/manual/rewrite/htaccess.xml.ko

@@ -1,4 +1,4 @@
-<?xml version='1.0' encoding='EUC-KR' ?>
+<?xml version='1.0' encoding='UTF-8' ?>
 <!DOCTYPE manualpage SYSTEM "../style/manualpage.dtd">
 <?xml-stylesheet type="text/xsl" href="../style/manual.ko.xsl"?>
 <!-- English Revision: 1933067:1935417 (outdated) -->

docs/manual/rewrite/index.xml.ko

@@ -1,4 +1,4 @@
-<?xml version='1.0' encoding='EUC-KR' ?>
+<?xml version='1.0' encoding='UTF-8' ?>
 <!DOCTYPE manualpage SYSTEM "../style/manualpage.dtd">
 <?xml-stylesheet type="text/xsl" href="../style/manual.ko.xsl"?>
 <!-- English Revision: 1933438:1934111 (outdated) -->

docs/manual/rewrite/intro.xml.ko

@@ -1,4 +1,4 @@
-<?xml version='1.0' encoding='EUC-KR' ?>
+<?xml version='1.0' encoding='UTF-8' ?>
 <!DOCTYPE manualpage SYSTEM "../style/manualpage.dtd">
 <?xml-stylesheet type="text/xsl" href="../style/manual.ko.xsl"?>
 <!-- English Revision: 1933423:1935426 (outdated) -->

docs/manual/rewrite/remapping.xml.ko

@@ -1,4 +1,4 @@
-<?xml version='1.0' encoding='EUC-KR' ?>
+<?xml version='1.0' encoding='UTF-8' ?>
 <!DOCTYPE manualpage SYSTEM "../style/manualpage.dtd">
 <?xml-stylesheet type="text/xsl" href="../style/manual.ko.xsl"?>
 <!-- English Revision: 1933068:1934878 (outdated) -->

docs/manual/rewrite/rewritemap.xml.ko

@@ -1,4 +1,4 @@
-<?xml version='1.0' encoding='EUC-KR' ?>
+<?xml version='1.0' encoding='UTF-8' ?>
 <!DOCTYPE manualpage SYSTEM "../style/manualpage.dtd">
 <?xml-stylesheet type="text/xsl" href="../style/manual.ko.xsl"?>
 <!-- English Revision: 1932819:1934327 (outdated) -->

docs/manual/rewrite/tech.xml.ko

@@ -1,4 +1,4 @@
-<?xml version='1.0' encoding='EUC-KR' ?>
+<?xml version='1.0' encoding='UTF-8' ?>
 <!DOCTYPE manualpage SYSTEM "../style/manualpage.dtd">
 <?xml-stylesheet type="text/xsl" href="../style/manual.ko.xsl"?>
 <!-- English Revision: 1933438:1935430 (outdated) -->

docs/manual/rewrite/vhosts.xml.ko

@@ -1,4 +1,4 @@
-<?xml version='1.0' encoding='EUC-KR' ?>
+<?xml version='1.0' encoding='UTF-8' ?>
 <!DOCTYPE manualpage SYSTEM "../style/manualpage.dtd">
 <?xml-stylesheet type="text/xsl" href="../style/manual.ko.xsl"?>
 <!-- English Revision: 1933071:1934122 (outdated) -->

docs/manual/style/lang/ko.xml

@@ -96,7 +96,7 @@
         <message id="experimental" letter="X">Experimental</message>
         <message id="external" letter="T">External</message>
 
-        <message id="deprecated" letter="D">사용 중단됨</message>
+        <message id="deprecated" letter="D">»ç¿ë ÁߴܵÊ</message>
 
         <!-- Used in directive context lists -->
         <message id="serverconfig" letter="s">ÁÖ¼­¹ö¼³Á¤</message>

test/gha-docs-check.sh

@@ -0,0 +1,46 @@
+#!/bin/sh
+#
+# Check the XML sources of the manual: every file must be well-formed,
+# and every file with a DOCTYPE must be valid against its DTD.  Pass
+# file names to check only those, otherwise all of docs/ is checked.
+#
+# The DTDs pull in the XHTML entity sets by public identifier, so these
+# must be resolvable from an XML catalog (on Debian/Ubuntu, install
+# w3c-sgml-lib; on Fedora, xhtml1-dtds, or point XML_CATALOG_FILES at
+# a catalog); nothing is fetched from the network.
+
+cd "`dirname "$0"`/.." || exit 1
+
+if ! command -v xmllint >/dev/null; then
+    echo "FAIL: xmllint not found"
+    exit 1
+fi
+
+if [ $# -eq 0 ]; then
+    set -- `find docs -name '*.xml' -o -name '*.xml.*' | sort`
+fi
+
+rv=0
+count=0
+
+for f in "$@"; do
+    count=$((count + 1))
+    # The language files use lang.dtd only to define entities, and
+    # declare no elements, so can only be checked for well-formedness.
+    if grep -q '<!DOCTYPE' "$f" && ! grep -q '<!DOCTYPE language ' "$f"; then
+        mode=--valid
+    else
+        mode=--loaddtd
+    fi
+    # An unresolvable entity set is only a warning; treat it as fatal.
+    if ! out=`xmllint --noout --nonet $mode "$f" 2>&1` \
+       || echo "$out" | grep -q 'failed to load external entity'; then
+        echo "$out"
+        echo "FAIL: $f"
+        rv=1
+    fi
+done
+
+[ $rv -eq 0 ] && echo "PASS: $count files checked"
+
+exit $rv