Minimize DTD subsystem

configure.php

@@ -611,73 +611,51 @@ function git_status()
 
 function dtd_conf_entities()
 {
-    function dtd_pe_body( string $filename = '' )
+    function dtd_pe_load( string $name , string $filename ) : string
     {
         if ( file_exists( $filename ) )
         {
             $filename = realpain( $filename );
-            return "SYSTEM '$filename'";
+            $ret  = "<!ENTITY % $name SYSTEM '{$filename}'>\n";
+            $ret .= "%{$name};\n\n";
+            return $ret;
         }
-        return "''";
+        return "";
     }
 
-    global $ac;
-    $lang = $ac["LANG"];
-
-    // When all is converted to XML Entities,
-    // all this can be reduced to:
-    // $ent1 = dtd_pe_body( __DIR__ . '/temp/text-entities.ent' );
-    // $ent2 = dtd_pe_body( __DIR__ . '/temp/file-entities.ent' );
-
-    $baseEnt1 = dtd_pe_body( __DIR__ . '/entities/global.ent' );
-    $baseEnt2 = dtd_pe_body( __DIR__ . '/temp/file-entities.ent' );
-    $baseEnt3 = dtd_pe_body( __DIR__ . '/temp/entities.ent' );
+    // After DTD entities are converted to XML, all this can be reduced to:
+    // $contents .= dtd_pe_load( "text-entities" , __DIR__ . '/temp/text-entities.dtd' );
+    // $contents .= dtd_pe_load( "file-entities" , __DIR__ . '/temp/file-entities.dtd' );
 
-    $base = $ac['LANG_BASE_DIR'];
+    $base = $GLOBALS['ac']['LANG_BASE_DIR'];
+    $lang = $GLOBALS['ac']["LANG"];
 
-    $langOne1 = dtd_pe_body( __DIR__ . "/../$base/language-defs.ent" );
-    $langOne2 = dtd_pe_body( __DIR__ . "/../$base/language-snippets.ent" );
-    $langOne3 = dtd_pe_body( __DIR__ . "/../$base/extensions.ent" );
+    $contents = "<!ENTITY LANG '{$lang}'>\n\n";
 
-    if ( is_single_language() )
+    if ( ! is_single_language() )
     {
-        $langTwo1 = dtd_pe_body();
-        $langTwo2 = dtd_pe_body();
-        $langTwo3 = dtd_pe_body();
-    }
-    else
-    {
-        $langTwo1 = dtd_pe_body( __DIR__ . "/../$lang/language-defs.ent" );
-        $langTwo2 = dtd_pe_body( __DIR__ . "/../$lang/language-snippets.ent" );
-        $langTwo3 = dtd_pe_body( __DIR__ . "/../$lang/extensions.ent" );
+        $contents .= dtd_pe_load( "translation-defs"       , __DIR__ . "/../$lang/language-defs.ent" );
+        $contents .= dtd_pe_load( "translation-snippets"   , __DIR__ . "/../$lang/language-snippets.ent" );
+        $contents .= dtd_pe_load( "translation-extensions" , __DIR__ . "/../$lang/extensions.ent" );
     }
+    $contents .= dtd_pe_load( "language-defs"       , __DIR__ . "/../$base/language-defs.ent" );
+    $contents .= dtd_pe_load( "language-snippets"   , __DIR__ . "/../$base/language-snippets.ent" );
+    $contents .= dtd_pe_load( "language-extensions" , __DIR__ . "/../$base/extensions.ent" );
 
-    if ( $ac['CHMENABLED'] == 'yes' )
-        $chmpath = dtd_pe_body( __DIR__ . "/chm/manual.chm.xml" );
-    else
-        $chmpath = dtd_pe_body();
+    $contents .= dtd_pe_load( "base-entities" , __DIR__ . '/entities/global.ent' );
+    $contents .= dtd_pe_load( "text-entities" , __DIR__ . '/temp/text-entities.dtd' );
+    $contents .= dtd_pe_load( "file-entities" , __DIR__ . '/temp/file-entities.dtd' );
 
-    $conf = [];
-
-    $conf[] = "<!ENTITY LANG '$lang'>";
-
-    $conf[] = "<!ENTITY manual.chmonly           $chmpath>";
-
-    $conf[] = "<!ENTITY % base-entities          $baseEnt1>";
-    $conf[] = "<!ENTITY % file-entities          $baseEnt2>";
-    $conf[] = "<!ENTITY % text-entities          $baseEnt3>";
-
-    $conf[] = "<!ENTITY % language-defs          $langOne1>";
-    $conf[] = "<!ENTITY % language-snippets      $langOne2>";
-    $conf[] = "<!ENTITY % language-extensions    $langOne3>";
+    if ( $GLOBALS['ac']['CHMENABLED'] == 'yes' )
+        $contents .= dtd_pe_load( "manual.chmonly" , __DIR__ . "/chm/manual.chm.xml" );
+    else
+        $contents .= "<!ENTITY manual.chmonly ''>\n";
 
-    $conf[] = "<!ENTITY % translation-defs       $langTwo1>";
-    $conf[] = "<!ENTITY % translation-snippets   $langTwo2>";
-    $conf[] = "<!ENTITY % translation-extensions $langTwo3>";
+    $langTempDir = __DIR__ . "/../$base/temp";
+    realpain( $langTempDir , mkdir: true );
 
-    $outdir = __DIR__ . "/../$base/temp";
-    realpain( $outdir , mkdir: true );
-    file_put_contents( "{$outdir}/conf.dtd" , implode( "\n" , $conf ) );
+    file_put_contents(  __DIR__ . "/temp/lang" , $lang );
+    file_put_contents( "{$langTempDir}/conf.dtd" , $contents );
 }
 
 function dtd_file_entities()
@@ -750,7 +728,11 @@ function dom_load( DOMDocument $dom , string $filename , bool $firstLoad ) : boo
 {
     $filename = realpath( $filename );
     $options = LIBXML_NOENT | LIBXML_COMPACT | LIBXML_BIGLINES | LIBXML_PARSEHUGE;
-    return $dom->load( $filename , $options );
+    $ret = $dom->load( $filename , $options );
+
+    $dom->documentElement->setAttribute( 'xml:lang' , $GLOBALS['ac']["LANG"] );
+
+    return $ret;
 }
 
 function dom_saveload( DOMDocument $dom , string $filename = "" ) : string
@@ -922,7 +904,7 @@ function xinclude_run_xpointer( DOMDocument $dom ) : int
 
 function xinclude_residual_fixup( DOMDocument $dom )
 {
-    // XInclude failures are soft errors on translations, so we replace
+    // XInclude failures are soft errors on translations, so we erase
     // residual XInclude tags on translations to keep them validating.
 
     $fixups = 0;

funcindex.xml

@@ -1,4 +1,16 @@
 <?xml version="1.0" encoding="UTF-8"?>
+
+<!--
+
+This file is now being ignored, after DTD subsystem minimization.
+Yet, it is mentioned by name in configure.php, file-entities.php
+and in PhD repo. All cases are related to PhD, but configure.php
+includes this file as is, before any PhD stuff has a chance to run.
+
+After it is removed by name from all these points, it can be deleted.
+
+-->
+
 <!-- DO NOT EDIT THIS FILE. The index info comes from PhD -->
 <appendix xmlns="http://docbook.org/ns/docbook" xml:id="indexes">
  <title>&IndexListing;</title>

manual.xml

@@ -1,4 +1,19 @@
 <?xml version="1.0" encoding="utf-8"?>
+
+
+
+This file is now being ignored, after third-party manual split.
+Yet, it is mentioned by name in configure.php, on PhD related
+code paths.
+
+After these code paths are updated or moved into PhD, this can
+be deleted.
+
+Keep these comments outside SGML comment tags, as to make this
+file invalid for XML loading.
+
+
+
 <!DOCTYPE PHPDOC [
 
 <!-- Parameter entities used to configure manual.xml -->

scripts/file-entities.php

@@ -29,8 +29,9 @@
 - doc-base/temp/file-entites.ent
 - doc-base/temp/file-entites/dir.dir.ent
 
-The file entity for directories (file listings) are keep as individual
-files, to avoid these libxml errors, in some OS/versions:
+The file entity for directory listings is kept as an separate files,
+instead of a monolithic one, to avoid these libxml errors, in some
+OS and/or libxml versions:
 
 - Detected an entity reference loop [1]
 - Maximum entity amplification factor exceeded [2]
@@ -317,23 +318,23 @@ function writeEntities( array $entities )
 
     ksort( $entities );
 
-    $outFile = realpain(  __DIR__ . "/../temp/file-entities.ent" , touch: true );
-    $lstFile = realpain(  __DIR__ . "/../temp/file-entities.txt" , touch: true );
+    $dtdFile = realpain(  __DIR__ . "/../temp/file-entities.dtd" , touch: true );
+    $txtFile = realpain(  __DIR__ . "/../temp/file-entities.txt" , touch: true );
     $sepPath = realpain(  __DIR__ . "/../temp/file-entities" , mkdir: true );
 
-    $singleFile = fopen( $outFile , "w" );
+    $singleFile = fopen( $dtdFile , "w" );
     if ( ! $singleFile )
     {
-        print "Failed to open $outFile\n.";
+        print "Failed to open $dtdFile\n.";
         exit( 1 );
     }
-    fputs( $singleFile , "<!-- DON'T TOUCH - AUTOGENERATED BY file-entities.php -->\n\n" );
+    fputs( $singleFile , "<!-- AUTOGENERATED BY file-entities.php -->\n\n" );
 
     // Life could be simpler, but the building of PHP Manual is already
     // triping some hardcoded limits of bundled libxml2.
 
     // Off loading DTD entities that expand to more DTD entities,
-    // as external files, somehow avoid these limits.
+    // as external files, to somehow avoid these limits.
 
     if ( LIBXML_LIMITS_HACK )
     {
@@ -363,7 +364,7 @@ function writeEntities( array $entities )
     // entities.
 
     $contents = implode( "\n" , array_keys( $entities ) );
-    file_put_contents( $lstFile , $contents );
+    file_put_contents( $txtFile , $contents );
 }
 
 function writeEntityIndirectSlow( $singleFile , string $extraFile , string $name , string $text )

scripts/text-entities.php

@@ -12,7 +12,7 @@
 +----------------------------------------------------------------------+
 | Authors:     André L F S Bacci <ae php.net>                          |
 +----------------------------------------------------------------------+
-| Description: Collect individual entities into an temp/entities.ent.  |
+| Description: Collect individual entities into an text-entities.dtd.  |
 +----------------------------------------------------------------------+
 
 # Mental model for DTD <!ENTITY>,
@@ -39,7 +39,7 @@
 
 This script collects grouped and individual XML Entity files
 (detailed below), at some expected relative paths, and generates an
-doc-base/temp/entities.ent file with their respective DTD Entities.
+doc-base/temp/text-entities.dtd file with their respective DTD Entities.
 
 The output file has no duplications, so collection order is important
 to create some operational semantics. Here, latter loaded entities
@@ -175,7 +175,7 @@ public function __construct(
 
 class Entities
 {
-    private static string $filename = __DIR__ . "/../temp/entities.ent";
+    private static string $filename = __DIR__ . "/../temp/text-entities.dtd";
 
     private static array $merged = [];          // All EntityData, merged by name, no duplications
     private static array $unique = [];          // Any entity marked unique
@@ -425,7 +425,6 @@ function loadEntitySingle( string $path )
 function outputFiles( string $filename , array $entities )
 {
     $file = fopen( $filename , "w" );
-    fputs( $file , "\n<!-- DO NOT COPY / DO NOT TRANSLATE -->" );
     fputs( $file , "\n<!-- Autogenerated by text-entities.php -->\n\n" );
 
     $sepFileDir = __DIR__ . "/../temp/text-entities/";