nokogiri 1.16.3 → 1.18.1

data/patches/libxml2/0019-xpath-Use-separate-static-hash-table-for-standard-fu.patch

@@ -0,0 +1,244 @@
+From d3e3526111097560cf7c002613e2cb1d469b59e0 Mon Sep 17 00:00:00 2001
+From: Nick Wellnhofer <wellnhofer@aevum.de>
+Date: Sat, 21 Dec 2024 16:03:46 +0100
+Subject: [PATCH] xpath: Use separate static hash table for standard functions
+
+This avoids registering standard functions when creating an XPath
+context.
+
+Lookup of extension functions is a bit slower now, but ultimately, all
+function lookups should be moved to the compilation phase.
+
+(cherry picked from commit bf5fcf6e646bb51a0f6a3655a1d64bea97274867)
+---
+ xpath.c | 170 ++++++++++++++++++++++++++++++++------------------------
+ 1 file changed, 98 insertions(+), 72 deletions(-)
+
+diff --git a/xpath.c b/xpath.c
+index 485d7747..21711653 100644
+--- a/xpath.c
++++ b/xpath.c
+@@ -136,11 +136,48 @@
+ 
+ #if defined(LIBXML_XPATH_ENABLED) || defined(LIBXML_SCHEMAS_ENABLED)
+ 
+-/************************************************************************
+- *									*
+- *			Floating point stuff				*
+- *									*
+- ************************************************************************/
++static void
++xmlXPathNameFunction(xmlXPathParserContextPtr ctxt, int nargs);
++
++static const struct {
++    const char *name;
++    xmlXPathFunction func;
++} xmlXPathStandardFunctions[] = {
++    { "boolean", xmlXPathBooleanFunction },
++    { "ceiling", xmlXPathCeilingFunction },
++    { "count", xmlXPathCountFunction },
++    { "concat", xmlXPathConcatFunction },
++    { "contains", xmlXPathContainsFunction },
++    { "id", xmlXPathIdFunction },
++    { "false", xmlXPathFalseFunction },
++    { "floor", xmlXPathFloorFunction },
++    { "last", xmlXPathLastFunction },
++    { "lang", xmlXPathLangFunction },
++    { "local-name", xmlXPathLocalNameFunction },
++    { "not", xmlXPathNotFunction },
++    { "name", xmlXPathNameFunction },
++    { "namespace-uri", xmlXPathNamespaceURIFunction },
++    { "normalize-space", xmlXPathNormalizeFunction },
++    { "number", xmlXPathNumberFunction },
++    { "position", xmlXPathPositionFunction },
++    { "round", xmlXPathRoundFunction },
++    { "string", xmlXPathStringFunction },
++    { "string-length", xmlXPathStringLengthFunction },
++    { "starts-with", xmlXPathStartsWithFunction },
++    { "substring", xmlXPathSubstringFunction },
++    { "substring-before", xmlXPathSubstringBeforeFunction },
++    { "substring-after", xmlXPathSubstringAfterFunction },
++    { "sum", xmlXPathSumFunction },
++    { "true", xmlXPathTrueFunction },
++    { "translate", xmlXPathTranslateFunction }
++};
++
++#define NUM_STANDARD_FUNCTIONS \
++    (sizeof(xmlXPathStandardFunctions) / sizeof(xmlXPathStandardFunctions[0]))
++
++#define SF_HASH_SIZE 64
++
++static unsigned char xmlXPathSFHash[SF_HASH_SIZE];
+ 
+ double xmlXPathNAN = 0.0;
+ double xmlXPathPINF = 0.0;
+@@ -156,6 +193,18 @@ xmlXPathInit(void) {
+     xmlInitParser();
+ }
+ 
++ATTRIBUTE_NO_SANITIZE_INTEGER
++static unsigned
++xmlXPathSFComputeHash(const xmlChar *name) {
++    unsigned hashValue = 5381;
++    const xmlChar *ptr;
++
++    for (ptr = name; *ptr; ptr++)
++        hashValue = hashValue * 33 + *ptr;
++
++    return(hashValue);
++}
++
+ /**
+  * xmlInitXPathInternal:
+  *
+@@ -164,6 +213,8 @@ xmlXPathInit(void) {
+ ATTRIBUTE_NO_SANITIZE("float-divide-by-zero")
+ void
+ xmlInitXPathInternal(void) {
++    size_t i;
++
+ #if defined(NAN) && defined(INFINITY)
+     xmlXPathNAN = NAN;
+     xmlXPathPINF = INFINITY;
+@@ -175,8 +226,34 @@ xmlInitXPathInternal(void) {
+     xmlXPathPINF = 1.0 / zero;
+     xmlXPathNINF = -xmlXPathPINF;
+ #endif
++
++    /*
++     * Initialize hash table for standard functions
++     */
++
++    for (i = 0; i < SF_HASH_SIZE; i++)
++        xmlXPathSFHash[i] = UCHAR_MAX;
++
++    for (i = 0; i < NUM_STANDARD_FUNCTIONS; i++) {
++        const char *name = xmlXPathStandardFunctions[i].name;
++        int bucketIndex = xmlXPathSFComputeHash(BAD_CAST name) % SF_HASH_SIZE;
++
++        while (xmlXPathSFHash[bucketIndex] != UCHAR_MAX) {
++            bucketIndex += 1;
++            if (bucketIndex >= SF_HASH_SIZE)
++                bucketIndex = 0;
++        }
++
++        xmlXPathSFHash[bucketIndex] = i;
++    }
+ }
+ 
++/************************************************************************
++ *									*
++ *			Floating point stuff				*
++ *									*
++ ************************************************************************/
++
+ /**
+  * xmlXPathIsNaN:
+  * @val:  a double value
+@@ -3979,18 +4056,6 @@ xmlXPathRegisterFuncLookup (xmlXPathContextPtr ctxt,
+  */
+ xmlXPathFunction
+ xmlXPathFunctionLookup(xmlXPathContextPtr ctxt, const xmlChar *name) {
+-    if (ctxt == NULL)
+-	return (NULL);
+-
+-    if (ctxt->funcLookupFunc != NULL) {
+-	xmlXPathFunction ret;
+-	xmlXPathFuncLookupFunc f;
+-
+-	f = ctxt->funcLookupFunc;
+-	ret = f(ctxt->funcLookupData, name, NULL);
+-	if (ret != NULL)
+-	    return(ret);
+-    }
+     return(xmlXPathFunctionLookupNS(ctxt, name, NULL));
+ }
+ 
+@@ -4015,6 +4080,22 @@ xmlXPathFunctionLookupNS(xmlXPathContextPtr ctxt, const xmlChar *name,
+     if (name == NULL)
+ 	return(NULL);
+ 
++    if (ns_uri == NULL) {
++        int bucketIndex = xmlXPathSFComputeHash(name) % SF_HASH_SIZE;
++
++        while (xmlXPathSFHash[bucketIndex] != UCHAR_MAX) {
++            int funcIndex = xmlXPathSFHash[bucketIndex];
++
++            if (strcmp(xmlXPathStandardFunctions[funcIndex].name,
++                       (char *) name) == 0)
++                return(xmlXPathStandardFunctions[funcIndex].func);
++
++            bucketIndex += 1;
++            if (bucketIndex >= SF_HASH_SIZE)
++                bucketIndex = 0;
++        }
++    }
++
+     if (ctxt->funcLookupFunc != NULL) {
+ 	xmlXPathFuncLookupFunc f;
+ 
+@@ -13494,61 +13575,6 @@ xmlXPathEscapeUriFunction(xmlXPathParserContextPtr ctxt, int nargs) {
+ void
+ xmlXPathRegisterAllFunctions(xmlXPathContextPtr ctxt)
+ {
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"boolean",
+-                         xmlXPathBooleanFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"ceiling",
+-                         xmlXPathCeilingFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"count",
+-                         xmlXPathCountFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"concat",
+-                         xmlXPathConcatFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"contains",
+-                         xmlXPathContainsFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"id",
+-                         xmlXPathIdFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"false",
+-                         xmlXPathFalseFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"floor",
+-                         xmlXPathFloorFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"last",
+-                         xmlXPathLastFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"lang",
+-                         xmlXPathLangFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"local-name",
+-                         xmlXPathLocalNameFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"not",
+-                         xmlXPathNotFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"name",
+-                         xmlXPathNameFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"namespace-uri",
+-                         xmlXPathNamespaceURIFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"normalize-space",
+-                         xmlXPathNormalizeFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"number",
+-                         xmlXPathNumberFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"position",
+-                         xmlXPathPositionFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"round",
+-                         xmlXPathRoundFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"string",
+-                         xmlXPathStringFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"string-length",
+-                         xmlXPathStringLengthFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"starts-with",
+-                         xmlXPathStartsWithFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"substring",
+-                         xmlXPathSubstringFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"substring-before",
+-                         xmlXPathSubstringBeforeFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"substring-after",
+-                         xmlXPathSubstringAfterFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"sum",
+-                         xmlXPathSumFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"true",
+-                         xmlXPathTrueFunction);
+-    xmlXPathRegisterFunc(ctxt, (const xmlChar *)"translate",
+-                         xmlXPathTranslateFunction);
+-
+     xmlXPathRegisterFuncNS(ctxt, (const xmlChar *)"escape-uri",
+ 	 (const xmlChar *)"http://www.w3.org/2002/08/xquery-functions",
+                          xmlXPathEscapeUriFunction);
+-- 
+2.47.1
+

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: nokogiri
 version: !ruby/object:Gem::Version
-  version: 1.16.3
+  version: 1.18.1
 platform: ruby
 authors:
 - Mike Dalessio
@@ -17,10 +17,9 @@ authors:
 - Sergio Arbeo
 - Timothy Elliott
 - Nobuyoshi Nakada
-autorequire:
 bindir: bin
 cert_chain: []
-date: 2024-03-15 00:00:00.000000000 Z
+date: 2024-12-29 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: mini_portile2
@@ -36,7 +35,6 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: 2.8.2
-  force_ruby_platform: false
 - !ruby/object:Gem::Dependency
   name: racc
   requirement: !ruby/object:Gem::Requirement
@@ -51,7 +49,6 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '1.4'
-  force_ruby_platform: false
 description: |
   Nokogiri (鋸) makes it easy and painless to work with XML and HTML from Ruby. It provides a
   sensible, easy-to-understand API for reading, writing, modifying, and querying documents. It is
@@ -66,9 +63,10 @@ extra_rdoc_files:
 - ext/nokogiri/html4_document.c
 - ext/nokogiri/html4_element_description.c
 - ext/nokogiri/html4_entity_lookup.c
+- ext/nokogiri/html4_sax_parser.c
 - ext/nokogiri/html4_sax_parser_context.c
 - ext/nokogiri/html4_sax_push_parser.c
-- ext/nokogiri/libxml2_backwards_compat.c
+- ext/nokogiri/libxml2_polyfill.c
 - ext/nokogiri/nokogiri.c
 - ext/nokogiri/test_global_handlers.c
 - ext/nokogiri/xml_attr.c
@@ -111,9 +109,10 @@ files:
 - ext/nokogiri/html4_document.c
 - ext/nokogiri/html4_element_description.c
 - ext/nokogiri/html4_entity_lookup.c
+- ext/nokogiri/html4_sax_parser.c
 - ext/nokogiri/html4_sax_parser_context.c
 - ext/nokogiri/html4_sax_push_parser.c
-- ext/nokogiri/libxml2_backwards_compat.c
+- ext/nokogiri/libxml2_polyfill.c
 - ext/nokogiri/nokogiri.c
 - ext/nokogiri/nokogiri.h
 - ext/nokogiri/test_global_handlers.c
@@ -195,6 +194,7 @@ files:
 - lib/nokogiri/css/parser.rb
 - lib/nokogiri/css/parser.y
 - lib/nokogiri/css/parser_extras.rb
+- lib/nokogiri/css/selector_cache.rb
 - lib/nokogiri/css/syntax_error.rb
 - lib/nokogiri/css/tokenizer.rb
 - lib/nokogiri/css/tokenizer.rex
@@ -216,6 +216,7 @@ files:
 - lib/nokogiri/html4/sax/parser_context.rb
 - lib/nokogiri/html4/sax/push_parser.rb
 - lib/nokogiri/html5.rb
+- lib/nokogiri/html5/builder.rb
 - lib/nokogiri/html5/document.rb
 - lib/nokogiri/html5/document_fragment.rb
 - lib/nokogiri/html5/node.rb
@@ -267,13 +268,13 @@ files:
 - lib/xsd/xmlparser/nokogiri.rb
 - patches/libxml2/0001-Remove-script-macro-support.patch
 - patches/libxml2/0002-Update-entities-to-remove-handling-of-ssi.patch
-- patches/libxml2/0003-libxml2.la-is-in-top_builddir.patch
 - patches/libxml2/0009-allow-wildcard-namespaces.patch
 - patches/libxml2/0010-update-config.guess-and-config.sub-for-libxml2.patch
 - patches/libxml2/0011-rip-out-libxml2-s-libc_single_threaded-support.patch
+- patches/libxml2/0019-xpath-Use-separate-static-hash-table-for-standard-fu.patch
 - patches/libxslt/0001-update-config.guess-and-config.sub-for-libxslt.patch
-- ports/archives/libxml2-2.12.6.tar.xz
-- ports/archives/libxslt-1.1.39.tar.xz
+- ports/archives/libxml2-2.13.5.tar.xz
+- ports/archives/libxslt-1.1.42.tar.xz
 homepage: https://nokogiri.org
 licenses:
 - MIT
@@ -284,7 +285,6 @@ metadata:
   changelog_uri: https://nokogiri.org/CHANGELOG.html
   source_code_uri: https://github.com/sparklemotion/nokogiri
   rubygems_mfa_required: 'true'
-post_install_message:
 rdoc_options:
 - "--main"
 - README.md
@@ -294,15 +294,14 @@ required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
-      version: 3.0.0
+      version: 3.1.0
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.4.19
-signing_key:
+rubygems_version: 3.6.2
 specification_version: 4
 summary: Nokogiri (鋸) makes it easy and painless to work with XML and HTML from Ruby.
 test_files: []

data/ext/nokogiri/libxml2_backwards_compat.c

@@ -1,121 +0,0 @@
-#ifndef HAVE_XMLFIRSTELEMENTCHILD
-#include <nokogiri.h>
-/**
- * xmlFirstElementChild:
- * @parent: the parent node
- *
- * Finds the first child node of that element which is a Element node
- * Note the handling of entities references is different than in
- * the W3C DOM element traversal spec since we don't have back reference
- * from entities content to entities references.
- *
- * Returns the first element child or NULL if not available
- */
-xmlNodePtr
-xmlFirstElementChild(xmlNodePtr parent)
-{
-  xmlNodePtr cur = NULL;
-
-  if (parent == NULL) {
-    return (NULL);
-  }
-  switch (parent->type) {
-    case XML_ELEMENT_NODE:
-    case XML_ENTITY_NODE:
-    case XML_DOCUMENT_NODE:
-    case XML_HTML_DOCUMENT_NODE:
-      cur = parent->children;
-      break;
-    default:
-      return (NULL);
-  }
-  while (cur != NULL) {
-    if (cur->type == XML_ELEMENT_NODE) {
-      return (cur);
-    }
-    cur = cur->next;
-  }
-  return (NULL);
-}
-
-/**
- * xmlNextElementSibling:
- * @node: the current node
- *
- * Finds the first closest next sibling of the node which is an
- * element node.
- * Note the handling of entities references is different than in
- * the W3C DOM element traversal spec since we don't have back reference
- * from entities content to entities references.
- *
- * Returns the next element sibling or NULL if not available
- */
-xmlNodePtr
-xmlNextElementSibling(xmlNodePtr node)
-{
-  if (node == NULL) {
-    return (NULL);
-  }
-  switch (node->type) {
-    case XML_ELEMENT_NODE:
-    case XML_TEXT_NODE:
-    case XML_CDATA_SECTION_NODE:
-    case XML_ENTITY_REF_NODE:
-    case XML_ENTITY_NODE:
-    case XML_PI_NODE:
-    case XML_COMMENT_NODE:
-    case XML_DTD_NODE:
-    case XML_XINCLUDE_START:
-    case XML_XINCLUDE_END:
-      node = node->next;
-      break;
-    default:
-      return (NULL);
-  }
-  while (node != NULL) {
-    if (node->type == XML_ELEMENT_NODE) {
-      return (node);
-    }
-    node = node->next;
-  }
-  return (NULL);
-}
-
-/**
- * xmlLastElementChild:
- * @parent: the parent node
- *
- * Finds the last child node of that element which is a Element node
- * Note the handling of entities references is different than in
- * the W3C DOM element traversal spec since we don't have back reference
- * from entities content to entities references.
- *
- * Returns the last element child or NULL if not available
- */
-xmlNodePtr
-xmlLastElementChild(xmlNodePtr parent)
-{
-  xmlNodePtr cur = NULL;
-
-  if (parent == NULL) {
-    return (NULL);
-  }
-  switch (parent->type) {
-    case XML_ELEMENT_NODE:
-    case XML_ENTITY_NODE:
-    case XML_DOCUMENT_NODE:
-    case XML_HTML_DOCUMENT_NODE:
-      cur = parent->last;
-      break;
-    default:
-      return (NULL);
-  }
-  while (cur != NULL) {
-    if (cur->type == XML_ELEMENT_NODE) {
-      return (cur);
-    }
-    cur = cur->prev;
-  }
-  return (NULL);
-}
-#endif

data/patches/libxml2/0003-libxml2.la-is-in-top_builddir.patch

@@ -1,25 +0,0 @@
-From 0b6ae484761fa01242fe8b67b54e3eb2d282d83d Mon Sep 17 00:00:00 2001
-From: Mike Dalessio <mike.dalessio@gmail.com>
-Date: Wed, 4 Dec 2019 08:43:51 -0500
-Subject: [PATCH] fix libxml2.la's path
-
----
- Makefile.in | 2 +-
- 1 file changed, 1 insertion(+), 1 deletion(-)
-
-diff --git a/Makefile.in b/Makefile.in
-index cf96d41..1372d8b 100644
---- a/Makefile.in
-+++ b/Makefile.in
-@@ -1057,7 +1057,7 @@ clean-noinstLTLIBRARIES:
- 	  rm -f $${locs}; \
- 	}
- 
--libxml2.la: $(libxml2_la_OBJECTS) $(libxml2_la_DEPENDENCIES) $(EXTRA_libxml2_la_DEPENDENCIES) 
-+$(top_builddir)/libxml2.la: $(libxml2_la_OBJECTS) $(libxml2_la_DEPENDENCIES) $(EXTRA_libxml2_la_DEPENDENCIES)
- 	$(AM_V_CCLD)$(libxml2_la_LINK) -rpath $(libdir) $(libxml2_la_OBJECTS) $(libxml2_la_LIBADD) $(LIBS)
- 
- testdso.la: $(testdso_la_OBJECTS) $(testdso_la_DEPENDENCIES) $(EXTRA_testdso_la_DEPENDENCIES) 
--- 
-2.17.1
-