From 51d5e2ea82a7540d91cc4161e890bc19eab85bf3 Mon Sep 17 00:00:00 2001 From: Christoph Reiter Date: Wed, 19 Jan 2022 09:26:23 +0100 Subject: [PATCH] libxml2: backport fix for lxml Fixes #10576 Fixes #8864 --- .../0028-work-around-lxml-api-abuse.patch | 211 ++++++++++++++++++ mingw-w64-libxml2/PKGBUILD | 9 +- 2 files changed, 219 insertions(+), 1 deletion(-) create mode 100644 mingw-w64-libxml2/0028-work-around-lxml-api-abuse.patch diff --git a/mingw-w64-libxml2/0028-work-around-lxml-api-abuse.patch b/mingw-w64-libxml2/0028-work-around-lxml-api-abuse.patch new file mode 100644 index 0000000000..8a9de7fc2b --- /dev/null +++ b/mingw-w64-libxml2/0028-work-around-lxml-api-abuse.patch @@ -0,0 +1,211 @@ +From 7955b0d6fbbe49392ccc2e511edd00fbbfcb5a10 Mon Sep 17 00:00:00 2001 +From: Nick Wellnhofer +Date: Tue, 18 May 2021 20:08:28 +0200 +Subject: [PATCH] Work around lxml API abuse + +Make xmlNodeDumpOutput and htmlNodeDumpFormatOutput work with corrupted +parent pointers. This used to work with the old recursive code but the +non-recursive rewrite required parent pointers to be set correctly. + +Unfortunately, lxml relies on the old behavior and passes subtrees with +a corrupted structure. Fall back to a recursive function call if an +invalid parent pointer is detected. + +Fixes #255. +--- + HTMLtree.c | 46 ++++++++++++++++++++++++++++------------------ + xmlsave.c | 31 +++++++++++++++++++++---------- + 2 files changed, 49 insertions(+), 28 deletions(-) + +diff --git a/HTMLtree.c b/HTMLtree.c +index 24434d453..bdd639c7f 100644 +--- a/HTMLtree.c ++++ b/HTMLtree.c +@@ -744,7 +744,7 @@ void + htmlNodeDumpFormatOutput(xmlOutputBufferPtr buf, xmlDocPtr doc, + xmlNodePtr cur, const char *encoding ATTRIBUTE_UNUSED, + int format) { +- xmlNodePtr root; ++ xmlNodePtr root, parent; + xmlAttrPtr attr; + const htmlElemDesc * info; + +@@ -755,6 +755,7 @@ htmlNodeDumpFormatOutput(xmlOutputBufferPtr buf, xmlDocPtr doc, + } + + root = cur; ++ parent = cur->parent; + while (1) { + switch (cur->type) { + case XML_HTML_DOCUMENT_NODE: +@@ -762,13 +763,25 @@ htmlNodeDumpFormatOutput(xmlOutputBufferPtr buf, xmlDocPtr doc, + if (((xmlDocPtr) cur)->intSubset != NULL) { + htmlDtdDumpOutput(buf, (xmlDocPtr) cur, NULL); + } +- if (cur->children != NULL) { ++ /* Always validate cur->parent when descending. */ ++ if ((cur->parent == parent) && (cur->children != NULL)) { ++ parent = cur; + cur = cur->children; + continue; + } + break; + + case XML_ELEMENT_NODE: ++ /* ++ * Some users like lxml are known to pass nodes with a corrupted ++ * tree structure. Fall back to a recursive call to handle this ++ * case. ++ */ ++ if ((cur->parent != parent) && (cur->children != NULL)) { ++ htmlNodeDumpFormatOutput(buf, doc, cur, encoding, format); ++ break; ++ } ++ + /* + * Get specific HTML info for that node. + */ +@@ -817,6 +830,7 @@ htmlNodeDumpFormatOutput(xmlOutputBufferPtr buf, xmlDocPtr doc, + (cur->name != NULL) && + (cur->name[0] != 'p')) /* p, pre, param */ + xmlOutputBufferWriteString(buf, "\n"); ++ parent = cur; + cur = cur->children; + continue; + } +@@ -825,9 +839,9 @@ htmlNodeDumpFormatOutput(xmlOutputBufferPtr buf, xmlDocPtr doc, + (info != NULL) && (!info->isinline)) { + if ((cur->next->type != HTML_TEXT_NODE) && + (cur->next->type != HTML_ENTITY_REF_NODE) && +- (cur->parent != NULL) && +- (cur->parent->name != NULL) && +- (cur->parent->name[0] != 'p')) /* p, pre, param */ ++ (parent != NULL) && ++ (parent->name != NULL) && ++ (parent->name[0] != 'p')) /* p, pre, param */ + xmlOutputBufferWriteString(buf, "\n"); + } + +@@ -842,9 +856,9 @@ htmlNodeDumpFormatOutput(xmlOutputBufferPtr buf, xmlDocPtr doc, + break; + if (((cur->name == (const xmlChar *)xmlStringText) || + (cur->name != (const xmlChar *)xmlStringTextNoenc)) && +- ((cur->parent == NULL) || +- ((xmlStrcasecmp(cur->parent->name, BAD_CAST "script")) && +- (xmlStrcasecmp(cur->parent->name, BAD_CAST "style"))))) { ++ ((parent == NULL) || ++ ((xmlStrcasecmp(parent->name, BAD_CAST "script")) && ++ (xmlStrcasecmp(parent->name, BAD_CAST "style"))))) { + xmlChar *buffer; + + buffer = xmlEncodeEntitiesReentrant(doc, cur->content); +@@ -902,13 +916,9 @@ htmlNodeDumpFormatOutput(xmlOutputBufferPtr buf, xmlDocPtr doc, + break; + } + +- /* +- * The parent should never be NULL here but we want to handle +- * corrupted documents gracefully. +- */ +- if (cur->parent == NULL) +- return; +- cur = cur->parent; ++ cur = parent; ++ /* cur->parent was validated when descending. */ ++ parent = cur->parent; + + if ((cur->type == XML_HTML_DOCUMENT_NODE) || + (cur->type == XML_DOCUMENT_NODE)) { +@@ -939,9 +949,9 @@ htmlNodeDumpFormatOutput(xmlOutputBufferPtr buf, xmlDocPtr doc, + (cur->next != NULL)) { + if ((cur->next->type != HTML_TEXT_NODE) && + (cur->next->type != HTML_ENTITY_REF_NODE) && +- (cur->parent != NULL) && +- (cur->parent->name != NULL) && +- (cur->parent->name[0] != 'p')) /* p, pre, param */ ++ (parent != NULL) && ++ (parent->name != NULL) && ++ (parent->name[0] != 'p')) /* p, pre, param */ + xmlOutputBufferWriteString(buf, "\n"); + } + } +diff --git a/xmlsave.c b/xmlsave.c +index 61a40459b..aedbd5e70 100644 +--- a/xmlsave.c ++++ b/xmlsave.c +@@ -847,7 +847,7 @@ htmlNodeDumpOutputInternal(xmlSaveCtxtPtr ctxt, xmlNodePtr cur) { + static void + xmlNodeDumpOutputInternal(xmlSaveCtxtPtr ctxt, xmlNodePtr cur) { + int format = ctxt->format; +- xmlNodePtr tmp, root, unformattedNode = NULL; ++ xmlNodePtr tmp, root, unformattedNode = NULL, parent; + xmlAttrPtr attr; + xmlChar *start, *end; + xmlOutputBufferPtr buf; +@@ -856,6 +856,7 @@ xmlNodeDumpOutputInternal(xmlSaveCtxtPtr ctxt, xmlNodePtr cur) { + buf = ctxt->buf; + + root = cur; ++ parent = cur->parent; + while (1) { + switch (cur->type) { + case XML_DOCUMENT_NODE: +@@ -868,7 +869,9 @@ xmlNodeDumpOutputInternal(xmlSaveCtxtPtr ctxt, xmlNodePtr cur) { + break; + + case XML_DOCUMENT_FRAG_NODE: +- if (cur->children != NULL) { ++ /* Always validate cur->parent when descending. */ ++ if ((cur->parent == parent) && (cur->children != NULL)) { ++ parent = cur; + cur = cur->children; + continue; + } +@@ -887,7 +890,18 @@ xmlNodeDumpOutputInternal(xmlSaveCtxtPtr ctxt, xmlNodePtr cur) { + break; + + case XML_ELEMENT_NODE: +- if ((cur != root) && (ctxt->format == 1) && (xmlIndentTreeOutput)) ++ /* ++ * Some users like lxml are known to pass nodes with a corrupted ++ * tree structure. Fall back to a recursive call to handle this ++ * case. ++ */ ++ if ((cur->parent != parent) && (cur->children != NULL)) { ++ xmlNodeDumpOutputInternal(ctxt, cur); ++ break; ++ } ++ ++ if ((ctxt->level > 0) && (ctxt->format == 1) && ++ (xmlIndentTreeOutput)) + xmlOutputBufferWrite(buf, ctxt->indent_size * + (ctxt->level > ctxt->indent_nr ? + ctxt->indent_nr : ctxt->level), +@@ -942,6 +956,7 @@ xmlNodeDumpOutputInternal(xmlSaveCtxtPtr ctxt, xmlNodePtr cur) { + xmlOutputBufferWrite(buf, 1, ">"); + if (ctxt->format == 1) xmlOutputBufferWrite(buf, 1, "\n"); + if (ctxt->level >= 0) ctxt->level++; ++ parent = cur; + cur = cur->children; + continue; + } +@@ -1058,13 +1073,9 @@ xmlNodeDumpOutputInternal(xmlSaveCtxtPtr ctxt, xmlNodePtr cur) { + break; + } + +- /* +- * The parent should never be NULL here but we want to handle +- * corrupted documents gracefully. +- */ +- if (cur->parent == NULL) +- return; +- cur = cur->parent; ++ cur = parent; ++ /* cur->parent was validated when descending. */ ++ parent = cur->parent; + + if (cur->type == XML_ELEMENT_NODE) { + if (ctxt->level > 0) ctxt->level--; +-- +GitLab + diff --git a/mingw-w64-libxml2/PKGBUILD b/mingw-w64-libxml2/PKGBUILD index 131b562970..305f3e6213 100644 --- a/mingw-w64-libxml2/PKGBUILD +++ b/mingw-w64-libxml2/PKGBUILD @@ -6,7 +6,7 @@ _realname=libxml2 pkgbase=mingw-w64-${_realname} pkgname="${MINGW_PACKAGE_PREFIX}-${_realname}" pkgver=2.9.12 -pkgrel=3 +pkgrel=4 arch=('any') mingw_arch=('mingw32' 'mingw64' 'ucrt64' 'clang64' 'clang32' 'clangarm64') pkgdesc="XML parsing library, version 2 (mingw-w64)" @@ -33,6 +33,7 @@ source=("http://xmlsoft.org/sources/${_realname}-${pkgver}.tar.gz" 0026-mingw-relocate.patch libxml2-2.9.9-python.patch 0027-decoding-segfault.patch + 0028-work-around-lxml-api-abuse.patch libxml2-disable-version-script.patch) sha256sums=('c8d6681e38c56f172892c85ddc0852e1fd4b53b4209e7f4ebf17f7e2eae71d92' '9b61db9f5dbffa545f4b8d78422167083a8568c59bd1129f94138f936cf6fc1f' @@ -46,6 +47,7 @@ sha256sums=('c8d6681e38c56f172892c85ddc0852e1fd4b53b4209e7f4ebf17f7e2eae71d92' 'e98c2f5ae8cdbcf54354a833c550c26563c1ebdfbf92cc790d40eafadb72f985' '7b9bc59b174da93a547429b2f751047c758657953f6e2fb398e73b5ebf0f55e3' '0391a4b267ba7251ca74ff2e98bf4c0332a14b618e8147a9341ec5617e45d9d9' + '6af21506236a16c154d1e72f987e8a0fb1bc2ea01f0fd6911eda1191df4f544c' '96f5b4de10b56563da0cb3c8b9ee9a88f89e5385acd66cadf4acab9628dc9848') # Helper macros to help make tasks easier # @@ -82,6 +84,11 @@ prepare() { apply_patch_with_msg \ 0027-decoding-segfault.patch + # https://gitlab.gnome.org/nwellnhof/libxml2/-/commit/7955b0d6fbbe49392ccc + # https://github.com/msys2/MINGW-packages/issues/10576 + apply_patch_with_msg \ + 0028-work-around-lxml-api-abuse.patch + libtoolize --copy --force aclocal automake --add-missing