nokogiri 1.10.3 → 1.12.5

data/ext/nokogiri/xslt_stylesheet.c

@@ -1,34 +1,29 @@
-#include <xslt_stylesheet.h>
+#include <nokogiri.h>
 
-#include <libxslt/xsltInternals.h>
-#include <libxslt/xsltutils.h>
-#include <libxslt/transform.h>
-#include <libexslt/exslt.h>
-
-VALUE xslt;
-
-int vasprintf (char **strp, const char *fmt, va_list ap);
-void vasprintf_free (void *p);
+VALUE cNokogiriXsltStylesheet ;
 
-static void mark(nokogiriXsltStylesheetTuple *wrapper)
+static void
+mark(nokogiriXsltStylesheetTuple *wrapper)
 {
   rb_gc_mark(wrapper->func_instances);
 }
 
-static void dealloc(nokogiriXsltStylesheetTuple *wrapper)
+static void
+dealloc(nokogiriXsltStylesheetTuple *wrapper)
 {
-    xsltStylesheetPtr doc = wrapper->ss;
+  xsltStylesheetPtr doc = wrapper->ss;
 
-    NOKOGIRI_DEBUG_START(doc);
-    xsltFreeStylesheet(doc); /* commented out for now. */
-    NOKOGIRI_DEBUG_END(doc);
+  NOKOGIRI_DEBUG_START(doc);
+  xsltFreeStylesheet(doc); /* commented out for now. */
+  NOKOGIRI_DEBUG_END(doc);
 
-    free(wrapper);
+  free(wrapper);
 }
 
-static void xslt_generic_error_handler(void * ctx, const char *msg, ...)
+static void
+xslt_generic_error_handler(void *ctx, const char *msg, ...)
 {
-  char * message;
+  char *message;
 
   va_list args;
   va_start(args, msg);
@@ -37,10 +32,11 @@ static void xslt_generic_error_handler(void * ctx, const char *msg, ...)
 
   rb_str_cat2((VALUE)ctx, message);
 
-  vasprintf_free(message);
+  free(message);
 }
 
-VALUE Nokogiri_wrap_xslt_stylesheet(xsltStylesheetPtr ss)
+VALUE
+Nokogiri_wrap_xslt_stylesheet(xsltStylesheetPtr ss)
 {
   VALUE self;
   nokogiriXsltStylesheetTuple *wrapper;
@@ -61,29 +57,29 @@ VALUE Nokogiri_wrap_xslt_stylesheet(xsltStylesheetPtr ss)
  *
  * Parse a stylesheet from +document+.
  */
-static VALUE parse_stylesheet_doc(VALUE klass, VALUE xmldocobj)
+static VALUE
+parse_stylesheet_doc(VALUE klass, VALUE xmldocobj)
 {
-    xmlDocPtr xml, xml_cpy;
-    VALUE errstr, exception;
-    xsltStylesheetPtr ss ;
-    Data_Get_Struct(xmldocobj, xmlDoc, xml);
-    exsltRegisterAll();
+  xmlDocPtr xml, xml_cpy;
+  VALUE errstr, exception;
+  xsltStylesheetPtr ss ;
+  Data_Get_Struct(xmldocobj, xmlDoc, xml);
 
-    errstr = rb_str_new(0, 0);
-    xsltSetGenericErrorFunc((void *)errstr, xslt_generic_error_handler);
+  errstr = rb_str_new(0, 0);
+  xsltSetGenericErrorFunc((void *)errstr, xslt_generic_error_handler);
 
-    xml_cpy = xmlCopyDoc(xml, 1); /* 1 => recursive */
-    ss = xsltParseStylesheetDoc(xml_cpy);
+  xml_cpy = xmlCopyDoc(xml, 1); /* 1 => recursive */
+  ss = xsltParseStylesheetDoc(xml_cpy);
 
-    xsltSetGenericErrorFunc(NULL, NULL);
+  xsltSetGenericErrorFunc(NULL, NULL);
 
-    if (!ss) {
-	xmlFreeDoc(xml_cpy);
-	exception = rb_exc_new3(rb_eRuntimeError, errstr);
-	rb_exc_raise(exception);
-    }
+  if (!ss) {
+    xmlFreeDoc(xml_cpy);
+    exception = rb_exc_new3(rb_eRuntimeError, errstr);
+    rb_exc_raise(exception);
+  }
 
-    return Nokogiri_wrap_xslt_stylesheet(ss);
+  return Nokogiri_wrap_xslt_stylesheet(ss);
 }
 
 
@@ -93,24 +89,21 @@ static VALUE parse_stylesheet_doc(VALUE klass, VALUE xmldocobj)
  *
  * Serialize +document+ to an xml string.
  */
-static VALUE serialize(VALUE self, VALUE xmlobj)
-{
-    xmlDocPtr xml ;
-    nokogiriXsltStylesheetTuple *wrapper;
-    xmlChar* doc_ptr ;
-    int doc_len ;
-    VALUE rval ;
-
-    Data_Get_Struct(xmlobj, xmlDoc, xml);
-    Data_Get_Struct(self, nokogiriXsltStylesheetTuple, wrapper);
-    xsltSaveResultToString(&doc_ptr, &doc_len, xml, wrapper->ss);
-    rval = NOKOGIRI_STR_NEW(doc_ptr, doc_len);
-    xmlFree(doc_ptr);
-    return rval ;
-}
-
-static void swallow_superfluous_xml_errors(void * userdata, xmlErrorPtr error, ...)
+static VALUE
+serialize(VALUE self, VALUE xmlobj)
 {
+  xmlDocPtr xml ;
+  nokogiriXsltStylesheetTuple *wrapper;
+  xmlChar *doc_ptr ;
+  int doc_len ;
+  VALUE rval ;
+
+  Data_Get_Struct(xmlobj, xmlDoc, xml);
+  Data_Get_Struct(self, nokogiriXsltStylesheetTuple, wrapper);
+  xsltSaveResultToString(&doc_ptr, &doc_len, xml, wrapper->ss);
+  rval = NOKOGIRI_STR_NEW(doc_ptr, doc_len);
+  xmlFree(doc_ptr);
+  return rval ;
 }
 
 /*
@@ -128,109 +121,114 @@ static void swallow_superfluous_xml_errors(void * userdata, xmlErrorPtr error, .
  *    puts xslt.transform(doc, ['key', 'value'])
  *
  */
-static VALUE transform(int argc, VALUE* argv, VALUE self)
+static VALUE
+transform(int argc, VALUE *argv, VALUE self)
 {
-    VALUE xmldoc, paramobj, errstr, exception ;
-    xmlDocPtr xml ;
-    xmlDocPtr result ;
-    nokogiriXsltStylesheetTuple *wrapper;
-    const char** params ;
-    long param_len, j ;
-    int parse_error_occurred ;
-
-    rb_scan_args(argc, argv, "11", &xmldoc, &paramobj);
-    if (NIL_P(paramobj)) { paramobj = rb_ary_new2(0L) ; }
-    if (!rb_obj_is_kind_of(xmldoc, cNokogiriXmlDocument))
-      rb_raise(rb_eArgError, "argument must be a Nokogiri::XML::Document");
-
-    /* handle hashes as arguments. */
-    if(T_HASH == TYPE(paramobj)) {
-      paramobj = rb_funcall(paramobj, rb_intern("to_a"), 0);
-      paramobj = rb_funcall(paramobj, rb_intern("flatten"), 0);
-    }
-
-    Check_Type(paramobj, T_ARRAY);
-
-    Data_Get_Struct(xmldoc, xmlDoc, xml);
-    Data_Get_Struct(self, nokogiriXsltStylesheetTuple, wrapper);
-
-    param_len = RARRAY_LEN(paramobj);
-    params = calloc((size_t)param_len+1, sizeof(char*));
-    for (j = 0 ; j < param_len ; j++) {
-      VALUE entry = rb_ary_entry(paramobj, j);
-      const char * ptr = StringValueCStr(entry);
-      params[j] = ptr;
-    }
-    params[param_len] = 0 ;
-
-    errstr = rb_str_new(0, 0);
-    xsltSetGenericErrorFunc((void *)errstr, xslt_generic_error_handler);
-    xmlSetGenericErrorFunc((void *)errstr, xslt_generic_error_handler);
-
-    result = xsltApplyStylesheet(wrapper->ss, xml, params);
-    free(params);
-
-    xsltSetGenericErrorFunc(NULL, NULL);
-    xmlSetGenericErrorFunc(NULL, NULL);
-
-    parse_error_occurred = (Qfalse == rb_funcall(errstr, rb_intern("empty?"), 0));
-
-    if (parse_error_occurred) {
-      exception = rb_exc_new3(rb_eRuntimeError, errstr);
-      rb_exc_raise(exception);
-    }
-
-    return Nokogiri_wrap_xml_document((VALUE)0, result) ;
+  VALUE xmldoc, paramobj, errstr, exception ;
+  xmlDocPtr xml ;
+  xmlDocPtr result ;
+  nokogiriXsltStylesheetTuple *wrapper;
+  const char **params ;
+  long param_len, j ;
+  int parse_error_occurred ;
+
+  rb_scan_args(argc, argv, "11", &xmldoc, &paramobj);
+  if (NIL_P(paramobj)) { paramobj = rb_ary_new2(0L) ; }
+  if (!rb_obj_is_kind_of(xmldoc, cNokogiriXmlDocument)) {
+    rb_raise(rb_eArgError, "argument must be a Nokogiri::XML::Document");
+  }
+
+  /* handle hashes as arguments. */
+  if (T_HASH == TYPE(paramobj)) {
+    paramobj = rb_funcall(paramobj, rb_intern("to_a"), 0);
+    paramobj = rb_funcall(paramobj, rb_intern("flatten"), 0);
+  }
+
+  Check_Type(paramobj, T_ARRAY);
+
+  Data_Get_Struct(xmldoc, xmlDoc, xml);
+  Data_Get_Struct(self, nokogiriXsltStylesheetTuple, wrapper);
+
+  param_len = RARRAY_LEN(paramobj);
+  params = calloc((size_t)param_len + 1, sizeof(char *));
+  for (j = 0 ; j < param_len ; j++) {
+    VALUE entry = rb_ary_entry(paramobj, j);
+    const char *ptr = StringValueCStr(entry);
+    params[j] = ptr;
+  }
+  params[param_len] = 0 ;
+
+  errstr = rb_str_new(0, 0);
+  xsltSetGenericErrorFunc((void *)errstr, xslt_generic_error_handler);
+  xmlSetGenericErrorFunc((void *)errstr, xslt_generic_error_handler);
+
+  result = xsltApplyStylesheet(wrapper->ss, xml, params);
+  free(params);
+
+  xsltSetGenericErrorFunc(NULL, NULL);
+  xmlSetGenericErrorFunc(NULL, NULL);
+
+  parse_error_occurred = (Qfalse == rb_funcall(errstr, rb_intern("empty?"), 0));
+
+  if (parse_error_occurred) {
+    exception = rb_exc_new3(rb_eRuntimeError, errstr);
+    rb_exc_raise(exception);
+  }
+
+  return noko_xml_document_wrap((VALUE)0, result) ;
 }
 
-static void method_caller(xmlXPathParserContextPtr ctxt, int nargs)
+static void
+method_caller(xmlXPathParserContextPtr ctxt, int nargs)
 {
-    VALUE handler;
-    const char *function_name;
-    xsltTransformContextPtr transform;
-    const xmlChar *functionURI;
+  VALUE handler;
+  const char *function_name;
+  xsltTransformContextPtr transform;
+  const xmlChar *functionURI;
 
-    transform = xsltXPathGetTransformContext(ctxt);
-    functionURI = ctxt->context->functionURI;
-    handler = (VALUE)xsltGetExtData(transform, functionURI);
-    function_name = (const char*)(ctxt->context->function);
+  transform = xsltXPathGetTransformContext(ctxt);
+  functionURI = ctxt->context->functionURI;
+  handler = (VALUE)xsltGetExtData(transform, functionURI);
+  function_name = (const char *)(ctxt->context->function);
 
-    Nokogiri_marshal_xpath_funcall_and_return_values(ctxt, nargs, handler, (const char*)function_name);
+  Nokogiri_marshal_xpath_funcall_and_return_values(ctxt, nargs, handler, (const char *)function_name);
 }
 
-static void * initFunc(xsltTransformContextPtr ctxt, const xmlChar *uri)
+static void *
+initFunc(xsltTransformContextPtr ctxt, const xmlChar *uri)
 {
-    VALUE modules = rb_iv_get(xslt, "@modules");
-    VALUE obj = rb_hash_aref(modules, rb_str_new2((const char *)uri));
-    VALUE args = { Qfalse };
-    VALUE methods = rb_funcall(obj, rb_intern("instance_methods"), 1, args);
-    VALUE inst;
-    nokogiriXsltStylesheetTuple *wrapper;
-    int i;
-
-    for(i = 0; i < RARRAY_LEN(methods); i++) {
-	VALUE method_name = rb_obj_as_string(rb_ary_entry(methods, i));
-	xsltRegisterExtFunction(ctxt,
-          (unsigned char *)StringValueCStr(method_name), uri, method_caller);
-    }
-
-    Data_Get_Struct((VALUE)ctxt->style->_private, nokogiriXsltStylesheetTuple,
-                    wrapper);
-    inst = rb_class_new_instance(0, NULL, obj);
-    rb_ary_push(wrapper->func_instances, inst);
-
-    return (void *)inst;
+  VALUE modules = rb_iv_get(mNokogiriXslt, "@modules");
+  VALUE obj = rb_hash_aref(modules, rb_str_new2((const char *)uri));
+  VALUE args = { Qfalse };
+  VALUE methods = rb_funcall(obj, rb_intern("instance_methods"), 1, args);
+  VALUE inst;
+  nokogiriXsltStylesheetTuple *wrapper;
+  int i;
+
+  for (i = 0; i < RARRAY_LEN(methods); i++) {
+    VALUE method_name = rb_obj_as_string(rb_ary_entry(methods, i));
+    xsltRegisterExtFunction(ctxt,
+                            (unsigned char *)StringValueCStr(method_name), uri, method_caller);
+  }
+
+  Data_Get_Struct((VALUE)ctxt->style->_private, nokogiriXsltStylesheetTuple,
+                  wrapper);
+  inst = rb_class_new_instance(0, NULL, obj);
+  rb_ary_push(wrapper->func_instances, inst);
+
+  return (void *)inst;
 }
 
-static void shutdownFunc(xsltTransformContextPtr ctxt,
-	const xmlChar *uri, void *data)
+static void
+shutdownFunc(xsltTransformContextPtr ctxt,
+             const xmlChar *uri, void *data)
 {
-    nokogiriXsltStylesheetTuple *wrapper;
+  nokogiriXsltStylesheetTuple *wrapper;
 
-    Data_Get_Struct((VALUE)ctxt->style->_private, nokogiriXsltStylesheetTuple,
-                    wrapper);
+  Data_Get_Struct((VALUE)ctxt->style->_private, nokogiriXsltStylesheetTuple,
+                  wrapper);
 
-    rb_ary_clear(wrapper->func_instances);
+  rb_ary_clear(wrapper->func_instances);
 }
 
 /*
@@ -239,32 +237,28 @@ static void shutdownFunc(xsltTransformContextPtr ctxt,
  *
  *  Register a class that implements custom XSLT transformation functions.
  */
-static VALUE registr(VALUE self, VALUE uri, VALUE obj)
+static VALUE
+registr(VALUE self, VALUE uri, VALUE obj)
 {
-    VALUE modules = rb_iv_get(self, "@modules");
-    if(NIL_P(modules)) rb_raise(rb_eRuntimeError, "wtf! @modules isn't set");
+  VALUE modules = rb_iv_get(self, "@modules");
+  if (NIL_P(modules)) { rb_raise(rb_eRuntimeError, "wtf! @modules isn't set"); }
 
-    rb_hash_aset(modules, uri, obj);
-    xsltRegisterExtModule((unsigned char *)StringValueCStr(uri), initFunc, shutdownFunc);
-    return self;
+  rb_hash_aset(modules, uri, obj);
+  xsltRegisterExtModule((unsigned char *)StringValueCStr(uri), initFunc, shutdownFunc);
+  return self;
 }
 
-VALUE cNokogiriXsltStylesheet ;
-void init_xslt_stylesheet()
+void
+noko_init_xslt_stylesheet()
 {
-  VALUE nokogiri;
-  VALUE klass;
-
-  nokogiri = rb_define_module("Nokogiri");
-  xslt = rb_define_module_under(nokogiri, "XSLT");
-  klass = rb_define_class_under(xslt, "Stylesheet", rb_cObject);
+  rb_define_singleton_method(mNokogiriXslt, "register", registr, 2);
+  rb_iv_set(mNokogiriXslt, "@modules", rb_hash_new());
 
-  rb_iv_set(xslt, "@modules", rb_hash_new());
+  cNokogiriXsltStylesheet = rb_define_class_under(mNokogiriXslt, "Stylesheet", rb_cObject);
 
-  cNokogiriXsltStylesheet = klass;
+  rb_undef_alloc_func(cNokogiriXsltStylesheet);
 
-  rb_define_singleton_method(klass, "parse_stylesheet_doc", parse_stylesheet_doc, 1);
-  rb_define_singleton_method(xslt, "register", registr, 2);
-  rb_define_method(klass, "serialize", serialize, 1);
-  rb_define_method(klass, "transform", transform, -1);
+  rb_define_singleton_method(cNokogiriXsltStylesheet, "parse_stylesheet_doc", parse_stylesheet_doc, 1);
+  rb_define_method(cNokogiriXsltStylesheet, "serialize", serialize, 1);
+  rb_define_method(cNokogiriXsltStylesheet, "transform", transform, -1);
 }

data/gumbo-parser/CHANGES.md

@@ -0,0 +1,63 @@
+## Gumbo 0.10.1 (2015-04-30)
+
+Same as 0.10.0, but with the version number bumped because the last version-number commit to v0.9.4 makes GitHub think that v0.9.4 is the latest version and so it's not highlighted on the webpage.
+
+## Gumbo 0.10.0 (2015-04-30)
+
+* Full support for `<template>` tag (kevinhendricks, nostrademons).
+* Some fixes for `<rtc>`/`<rt>` handling (kevinhendricks, vmg).
+* All html5lib-trunk tests pass now! (kevinhendricks, vmg, nostrademons)
+* Support for fragment parsing (vmg)
+* A couple additional example programs (kevinhendricks)
+* Performance improvements totaling an estimated 30-40% total improvement (vmg, nostrademons).
+
+## Gumbo 0.9.4 (2015-04-30)
+
+* Additional Visual Studio fixes (lowjoel, nostrademons)
+* Fixed some unused variable warnings.
+* Fix for glibtoolize vs. libtoolize build errors on Mac.
+* Fixed `CDATA` end tag handling.
+
+## Gumbo 0.9.3 (2015-02-17)
+
+* Bugfix for `&AElig;` entities (rgrove)
+* Fix `CDATA` handling; `CDATA` sections now generate a `GUMBO_NODE_CDATA` node rather
+than plain text.
+* Fix `get_title example` to handle whitespace nodes (gsnedders)
+* Visual Studio compilation fixes (fishioon)
+* Take the namespace into account when determining whether a node matches a
+certain tag (aroben)
+* Replace the varargs tag functions with a tagset bytevector, for a 20-30%
+speedup in overall parse time (kevinhendricks, vmg)
+* Add MacOS X support to Travis CI, and fix the deployment/DLL issues this
+uncovered (nostrademons, kevinhendricks, vmg)
+
+## Gumbo 0.9.2 (2014-09-21)
+
+* Performance improvements: Ragel-based char ref decoder and DFA-based UTF8
+decoder, totaling speedups of up to 300%.
+* Added benchmarking program and some sample data.
+* Fixed a compiler error under Visual Studio.
+* Fix an error in the ctypes bindings that could lead to memory corruption in
+the Python bindings.
+* Fix duplicate attributes when parsing `<isindex>` tags.
+* Don't leave semicolons behind when consuming entity references (rgrove)
+* Internally rename some functions in preparation for an amalgamation file
+(jdeng)
+* Add proper cflags for gyp builds (skabbes)
+
+## Gumbo 0.9.1 (2014-08-07)
+
+* First version listed on PyPi.
+* Autotools files excluded from GitHub and generated via autogen.sh. (endgame)
+* Numerous compiler warnings fixed. (bnoordhuis, craigbarnes)
+* Google security audit passed.
+* Gyp support (tfarina)
+* Naming convention for structs changed to avoid C reserved words.
+* Fix several integer and buffer overflows (Maxime2)
+* Some Visual Studio compiler support (bugparty)
+* Python3 compatibility for the ctypes bindings.
+
+## Gumbo 0.9.0 (2013-08-13)
+
+* Initial release open-sourced by Google.

data/gumbo-parser/Makefile

@@ -0,0 +1,101 @@
+.PHONY: all clean check coverage
+
+gumbo_objs := $(patsubst %.c,build/%.o,$(wildcard src/*.c))
+test_objs := $(patsubst %.cc,build/%.o,$(wildcard test/*.cc))
+gtest_lib := googletest/make/gtest_main.a
+
+# make SANITIZEFLAGS='-fsanitize=undefined -fsanitize=address'
+SANITIZEFLAGS :=
+CPPFLAGS := -Isrc
+CFLAGS := -std=c99 -Os -Wall
+CXXFLAGS := -isystem googletest/include -std=c++11 -Os -Wall
+LDFLAGS := -pthread
+
+all: check
+
+src/%.c: src/%.rl
+	ragel -F1 -o $@ $<
+
+build/src:
+	mkdir -p $@
+
+build/test:
+	mkdir -p $@
+
+build/src/%.o: src/%.c build/src/flags | build/src
+	$(CC) -MMD $(CPPFLAGS) $(CFLAGS) $(SANITIZEFLAGS) -c -o $@ $<
+
+build/test/%.o: test/%.cc build/test/flags | build/test
+	$(CXX) -MMD $(CPPFLAGS) $(CXXFLAGS) $(SANITIZEFLAGS) -c -o $@ $<
+
+build/run_tests: $(gumbo_objs) $(test_objs) $(gtest_lib)
+	$(CXX) -o $@ $+ $(LDFLAGS) $(SANITIZEFLAGS)
+
+check: build/run_tests
+	./build/run_tests
+
+coverage:
+	$(RM) build/{src,test}/*.gcda
+	$(RM) build/*.info
+	$(MAKE) CPPFLAGS='-Isrc -DNDEBUG=1' \
+		CFLAGS='-std=c99 --coverage -g -O0' \
+		CXXFLAGS='-isystem googletest/include -std=c++11 --coverage -g -O0' \
+		LDFLAGS='--coverage' \
+		build/run_tests
+	lcov	--no-external \
+		--initial \
+		--capture \
+		--base-directory . \
+		--directory build \
+		--output-file build/coverage-pre.info
+	awk	-F '[:,]' \
+		'/^SF:/ { delete defs } /^FN:/ { defs[$$2]=1 } /^DA:/ { if ($$3 == 0 && $$2 in defs) next } { print }' \
+		build/coverage-pre.info > build/coverage-initial.info
+	./build/run_tests
+	lcov	--no-external \
+		--capture \
+		--base-directory . \
+		--directory build \
+		--rc lcov_branch_coverage=1 \
+		--output-file build/coverage-test.info
+	lcov	--add-tracefile build/coverage-initial.info \
+		--add-tracefile build/coverage-test.info \
+		--rc lcov_branch_coverage=1 \
+		--output-file build/coverage.info
+	lcov	--remove build/coverage.info '$(CURDIR)/googletest/*' \
+		--rc lcov_branch_coverage=1 \
+		--output-file build/coverage.info
+	genhtml --branch-coverage \
+		--output-directory build/coverage \
+		build/coverage.info
+
+clean:
+	$(RM) -r build
+
+build/src/flags: | build/src
+	@echo 'old_CC := $(CC)' > $@
+	@echo 'old_CPPFLAGS := $(CPPFLAGS)' >> $@
+	@echo 'old_CFLAGS := $(CFLAGS)' >>$@
+	@echo 'old_SANITIZEFLAGS := $(SANITIZEFLAGS)' >> $@
+	@echo 'old_LDFLAGS := $(LDFLAGS)' >> $@
+
+build/test/flags: | build/test
+	@echo 'old_CXX := $(CXX)' > $@
+	@echo 'old_CPPFLAGS := $(CPPFLAGS)' >> $@
+	@echo 'old_CXXFLAGS := $(CXXFLAGS)' >> $@
+	@echo 'old_SANITIZEFLAGS := $(SANITIZEFLAGS)' >> $@
+	@echo 'old_LDFLAGS := $(LDFLAGS)' >> $@
+
+ifeq (,$(filter clean coverage,$(MAKECMDGOALS)))
+# Ensure that the flags are up to date.
+-include build/src/flags build/test/flags
+ifneq ($(old_CC) | $(old_CPPFLAGS) | $(old_CFLAGS) | $(old_SANITIZEFLAGS) | $(old_LDFLAGS),$(CC) | $(CPPFLAGS) | $(CFLAGS) | $(SANITIZEFLAGS) | $(LDFLAGS))
+.PHONY: build/src/flags
+endif
+ifneq ($(old_CXX) | $(old_CPPFLAGS) | $(old_CXXFLAGS) | $(old_SANITIZEFLAGS) | $(old_LDFLAGS),$(CXX) | $(CPPFLAGS) | $(CXXFLAGS) | $(SANITIZEFLAGS) | $(LDFLAGS))
+.PHONY: build/test/flags
+endif
+
+# Include dependencies.
+-include $(test_objs:.o=.d) $(gumbo_objs:.o=.d)
+endif

data/gumbo-parser/THANKS

@@ -0,0 +1,27 @@
+Gumbo HTML parser THANKS file
+
+Gumbo was originally written by Jonathan Tang, but many people helped out through suggestions, question-answering, code reviews, bugfixes, and organizational support.  Here is a list of these people.  Help me keep it complete and exempt of errors.
+
+Adam Barth
+Adam Roben
+Ben Noordhuis
+Bowen Han
+Constantinos Michael
+Craig Barnes
+Geoffrey Sneddon
+Ian Hickson
+Jack Deng
+Joel Low
+Jonathan Shneier
+Kevin Hendricks
+Mason Tang
+Maxim Zakharov
+Michal Zalewski
+Neal Norwitz
+Othar Hansson
+Ryan Grove
+Stefan Haustein
+Steffen Meschkat
+Steven Kabbes
+Thiago Farina
+Vicent Marti

data/gumbo-parser/src/Makefile

@@ -0,0 +1,34 @@
+# this Makefile is used by ext/nokogiri/extconf.rb
+# to enable a mini_portile2 recipe to build the gumbo parser
+.PHONY: clean
+
+CFLAGS += -std=c99 -Wall
+
+# allow the ENV var to override this
+RANLIB ?= ranlib
+
+gumbo_objs := \
+	ascii.o \
+	attribute.o \
+	char_ref.o \
+	error.o \
+	foreign_attrs.o \
+	parser.o \
+	string_buffer.o \
+	string_piece.o \
+	svg_attrs.o \
+	svg_tags.o \
+	tag.o \
+	tag_lookup.o \
+	token_buffer.o \
+	tokenizer.o \
+	utf8.o \
+	util.o \
+	vector.o
+
+libgumbo.a: $(gumbo_objs)
+	$(AR) $(ARFLAGS) $@ $(gumbo_objs)
+	- ($(RANLIB) $@ || true) >/dev/null 2>&1
+
+clean:
+	rm -f $(gumbo_objs) libgumbo.a

data/gumbo-parser/src/README.md

@@ -0,0 +1,41 @@
+libgumbo
+========
+
+This is an internal fork of the [libgumbo] library, which was copied and
+later modified under the terms of the Apache 2.0 [license]. See `lua-gumbo`
+commit [`0a04728`] for details of the original import.
+
+Since importing the code, the following notable fixes and improvements
+have been made:
+
+* `91cef89`: Re-implement `adjust_foreign_attributes()` with a gperf hash
+* `b11abe7`: Pass `TagSet` arrays into functions by reference instead of value
+* `b73dc03`: Simplify `maybe_replace_codepoint()` function
+* `d5d0bb3`: Remove special handling of `<menuitem>` tag
+* `7bd5162`: Remove special handling of `<isindex>` tag
+* `a5c1b0e`: Use `realloc(3)` instead of `malloc(3)` in `enlarge_vector_if_full()`
+* `dcbebd7`: Use `realloc(3)` instead of `malloc(3)` in `maybe_resize_string_buffer()`
+* `df15262`: Make `destroy_node()` function non-recursive
+* `2df37f5`: Fix signedness of some format specifiers
+* `176553e`: Add maximum element nesting limit
+* `bed0f4a`: Annotate `gumbo_debug()` with `PRINTF` macro and fix warnings
+* `7ffc218`: Annotate `print_message()` with `PRINTF` macro and fix warnings
+* `1bd8ab5`, `9136507`, `53a1f9a`: Deduplicate some identical `TagSet` arrays
+* `a7a9065`: Add some GCC/Clang function attributes
+* `8d3d4e4`: Remove custom allocator support
+* `8d3b006`: Fix recording of source positions for `</form>` end tags
+* `1a8d763`: Replace linear search in `maybe_replace_codepoint()` with a lookup table
+* `6dca79e`: Replace `strcasecmp()` and `strncasecmp()` with ascii-only equivalents
+* `17ab1d2`: Fix `TAGSET_INCLUDES` macro to work properly with multiple bit flags
+* `7e56d45`: Re-implement `gumbo_normalize_svg_tagname()` with a gperf hash
+* `a518d35`: Replace linear array search in `adjust_svg_attributes()` with a gperf hash
+* `a4a7433`: Fix duplicate `TagSet` initializer being ignored in `is_special_node()`
+* `8137fcd`: Add support for `<dialog>` tag
+* `4b35471`: Add missing `static` qualifiers to hide symbols that shouldn't be extern
+* `df57c59`, `03101f3`, `ea62330`: Replace use of locale-dependant `ctype.h` functions
+  with custom, ASCII-only equivalents
+
+
+[libgumbo]: https://github.com/google/gumbo-parser/tree/aa91b27b02c0c80c482e24348a457ed7c3c088e0/src
+[license]: https://github.com/google/gumbo-parser/blob/aa91b27b02c0c80c482e24348a457ed7c3c088e0/COPYING
+[`0a04728`]: https://gitlab.com/craigbarnes/lua-gumbo/commit/0a047282815af86f3367a7d95fefcfe5723ece48