libxml-ruby 5.0.5 → 6.0.0

data/ext/libxml/ruby_xml_html_parser.c

@@ -1,91 +1,91 @@
-/* Please see the LICENSE file for copyright and distribution information */
-
-#include "ruby_libxml.h"
-
-#include <libxml/HTMLparser.h>
-
-/* Document-class: LibXML::XML::HTMLParser
- *
- * The HTML parser implements an HTML 4.0 non-verifying parser with an API
- * compatible with the XML::Parser.  In contrast with the XML::Parser,
- * it can parse "real world" HTML, even if it severely broken from a
- * specification point of view.
- *
- * The HTML parser creates an in-memory document object
- * that consist of any number of XML::Node instances.  This is simple
- * and powerful model, but has the major limitation that the size of
- * the document that can be processed is limited by the amount of
- * memory available.
- *
- * Using the html parser is simple:
- *
- *   parser = XML::HTMLParser.file('my_file')
- *   doc = parser.parse
- *
- * You can also parse documents (see XML::HTMLParser.document), 
- * strings (see XML::HTMLParser.string) and io objects (see
- * XML::HTMLParser.io).
- */
-
-VALUE cXMLHtmlParser;
-static ID CONTEXT_ATTR;
-
-
-/* call-seq:
- *    XML::HTMLParser.initialize -> parser
- *
- * Initializes a new parser instance with no pre-determined source.
- */
-static VALUE rxml_html_parser_initialize(int argc, VALUE *argv, VALUE self)
-{
-  VALUE context = Qnil;
-
-  rb_scan_args(argc, argv, "01", &context);
-
-  if (context == Qnil)
-  {
-    rb_raise(rb_eArgError, "An instance of a XML::Parser::Context must be passed to XML::HTMLParser.new");
-  }
-
-  rb_ivar_set(self, CONTEXT_ATTR, context);
-  return self;
-}
-
-/*
- * call-seq:
- *    parser.parse -> XML::Document
- *
- * Parse the input XML and create an XML::Document with
- * it's content. If an error occurs, XML::Parser::ParseError
- * is thrown.
- */
-static VALUE rxml_html_parser_parse(VALUE self)
-{
-  xmlParserCtxtPtr ctxt;
-  VALUE context = rb_ivar_get(self, CONTEXT_ATTR);
-  
-  Data_Get_Struct(context, xmlParserCtxt, ctxt);
-
-  if (htmlParseDocument(ctxt) == -1 && ! ctxt->recovery)
-  {
-    rxml_raise(&ctxt->lastError);
-  }
-
-  rb_funcall(context, rb_intern("close"), 0);
-
-  return rxml_document_wrap(ctxt->myDoc);
-}
-
-void rxml_init_html_parser(void)
-{
-  CONTEXT_ATTR = rb_intern("@context");
-
-  cXMLHtmlParser = rb_define_class_under(mXML, "HTMLParser", rb_cObject);
-
-  /* Atributes */
-  rb_define_attr(cXMLHtmlParser, "input", 1, 0);
-
-  /* Instance methods */
-  rb_define_method(cXMLHtmlParser, "initialize", rxml_html_parser_initialize, -1);
-  rb_define_method(cXMLHtmlParser, "parse", rxml_html_parser_parse, 0);
-}
+/* Please see the LICENSE file for copyright and distribution information */
+
+#include "ruby_libxml.h"
+
+#include <libxml/HTMLparser.h>
+
+/* Document-class: LibXML::XML::HTMLParser
+ *
+ * The HTML parser implements an HTML 4.0 non-verifying parser with an API
+ * compatible with the XML::Parser.  In contrast with the XML::Parser,
+ * it can parse "real world" HTML, even if it severely broken from a
+ * specification point of view.
+ *
+ * The HTML parser creates an in-memory document object
+ * that consist of any number of XML::Node instances.  This is simple
+ * and powerful model, but has the major limitation that the size of
+ * the document that can be processed is limited by the amount of
+ * memory available.
+ *
+ * Using the html parser is simple:
+ *
+ *   parser = XML::HTMLParser.file('my_file')
+ *   doc = parser.parse
+ *
+ * You can also parse documents (see XML::HTMLParser.document), 
+ * strings (see XML::HTMLParser.string) and io objects (see
+ * XML::HTMLParser.io).
+ */
+
+VALUE cXMLHtmlParser;
+static ID CONTEXT_ATTR;
+
+
+/* call-seq:
+ *    XML::HTMLParser.initialize -> parser
+ *
+ * Initializes a new parser instance with no pre-determined source.
+ */
+static VALUE rxml_html_parser_initialize(int argc, VALUE *argv, VALUE self)
+{
+  VALUE context = Qnil;
+
+  rb_scan_args(argc, argv, "01", &context);
+
+  if (context == Qnil)
+  {
+    rb_raise(rb_eArgError, "An instance of a XML::Parser::Context must be passed to XML::HTMLParser.new");
+  }
+
+  rb_ivar_set(self, CONTEXT_ATTR, context);
+  return self;
+}
+
+/*
+ * call-seq:
+ *    parser.parse -> XML::Document
+ *
+ * Parse the input XML and create an XML::Document with
+ * it's content. If an error occurs, XML::Parser::ParseError
+ * is thrown.
+ */
+static VALUE rxml_html_parser_parse(VALUE self)
+{
+  xmlParserCtxtPtr ctxt;
+  VALUE context = rb_ivar_get(self, CONTEXT_ATTR);
+  
+  TypedData_Get_Struct(context, xmlParserCtxt, &rxml_html_parser_context_type, ctxt);
+
+  if (htmlParseDocument(ctxt) == -1 && ! ctxt->recovery)
+  {
+    rxml_raise(&ctxt->lastError);
+  }
+
+  rb_funcall(context, rb_intern("close"), 0);
+
+  return rxml_document_wrap(ctxt->myDoc);
+}
+
+void rxml_init_html_parser(void)
+{
+  CONTEXT_ATTR = rb_intern("@context");
+
+  cXMLHtmlParser = rb_define_class_under(mXML, "HTMLParser", rb_cObject);
+
+  /* Atributes */
+  rb_define_attr(cXMLHtmlParser, "input", 1, 0);
+
+  /* Instance methods */
+  rb_define_method(cXMLHtmlParser, "initialize", rxml_html_parser_initialize, -1);
+  rb_define_method(cXMLHtmlParser, "parse", rxml_html_parser_parse, 0);
+}

data/ext/libxml/ruby_xml_html_parser.h

@@ -1,10 +1,10 @@
-/* Please see the LICENSE file for copyright and distribution information */
-
-#ifndef __RXML_HTML_PARSER__
-#define __RXML_HTML_PARSER__
-
-extern VALUE cXMLHtmlParser;
-
-void rxml_init_html_parser(void);
-
-#endif
+/* Please see the LICENSE file for copyright and distribution information */
+
+#ifndef __RXML_HTML_PARSER__
+#define __RXML_HTML_PARSER__
+
+extern VALUE cXMLHtmlParser;
+
+void rxml_init_html_parser(void);
+
+#endif

data/ext/libxml/ruby_xml_html_parser_context.c

@@ -129,14 +129,21 @@ static htmlParserCtxtPtr htmlNewParserCtxt(void)
 }
 #endif
 
-static void rxml_html_parser_context_free(htmlParserCtxtPtr ctxt)
+static void rxml_html_parser_context_free(void* data)
 {
+  htmlParserCtxtPtr ctxt = (htmlParserCtxtPtr)data;
   htmlFreeParserCtxt(ctxt);
 }
 
+const rb_data_type_t rxml_html_parser_context_type = {
+  "LibXML::XML::HTMLParser::Context",
+  {NULL, rxml_html_parser_context_free, NULL},
+  &rxml_parser_context_type, NULL, 0
+};
+
 static VALUE rxml_html_parser_context_wrap(htmlParserCtxtPtr ctxt)
 {
-  return Data_Wrap_Struct(cXMLHtmlParserContext, NULL, rxml_html_parser_context_free, ctxt);
+  return TypedData_Wrap_Struct(cXMLHtmlParserContext, &rxml_html_parser_context_type, ctxt);
 }
 
 /* call-seq:
@@ -275,7 +282,7 @@ static VALUE rxml_html_parser_context_close(VALUE self)
 {
   htmlParserCtxtPtr ctxt;
   xmlParserInputPtr xinput;
-  Data_Get_Struct(self, htmlParserCtxt, ctxt);
+  TypedData_Get_Struct(self, htmlParserCtxt, &rxml_html_parser_context_type, ctxt);
 
   while ((xinput = inputPop(ctxt)) != NULL)
   {
@@ -293,7 +300,7 @@ static VALUE rxml_html_parser_context_close(VALUE self)
 static VALUE rxml_html_parser_context_disable_cdata_set(VALUE self, VALUE value)
 {
   htmlParserCtxtPtr ctxt;
-  Data_Get_Struct(self, htmlParserCtxt, ctxt);
+  TypedData_Get_Struct(self, htmlParserCtxt, &rxml_html_parser_context_type, ctxt);
 
   if (ctxt->sax == NULL)
     rb_raise(rb_eRuntimeError, "Sax handler is not yet set");
@@ -320,9 +327,8 @@ static VALUE rxml_html_parser_context_options_set(VALUE self, VALUE options)
 {
   int xml_options = NUM2INT(options);
   htmlParserCtxtPtr ctxt;
-  Check_Type(options, T_FIXNUM);
 
-  Data_Get_Struct(self, htmlParserCtxt, ctxt);
+  TypedData_Get_Struct(self, htmlParserCtxt, &rxml_html_parser_context_type, ctxt);
   htmlCtxtUseOptions(ctxt, xml_options);
 
 #if LIBXML_VERSION >= 20707
@@ -337,10 +343,17 @@ static VALUE rxml_html_parser_context_options_set(VALUE self, VALUE options)
   return self;
 }
 
+static VALUE rxml_html_parser_context_alloc(VALUE klass)
+{
+  xmlParserCtxtPtr ctxt = htmlNewParserCtxt();
+  return TypedData_Wrap_Struct(klass, &rxml_html_parser_context_type, ctxt);
+}
+
 void rxml_init_html_parser_context(void)
 {
   IO_ATTR = ID2SYM(rb_intern("@io"));
   cXMLHtmlParserContext = rb_define_class_under(cXMLHtmlParser, "Context", cXMLParserContext);
+  rb_define_alloc_func(cXMLHtmlParserContext, rxml_html_parser_context_alloc);
 
   rb_define_singleton_method(cXMLHtmlParserContext, "file", rxml_html_parser_context_file, -1);
   rb_define_singleton_method(cXMLHtmlParserContext, "io", rxml_html_parser_context_io, -1);

data/ext/libxml/ruby_xml_html_parser_context.h

@@ -1,10 +1,11 @@
-/* Please see the LICENSE file for copyright and distribution information */
-
-#ifndef __RXML_HTML_PARSER_CONTEXT__
-#define __RXML_HTML_PARSER_CONTEXT__
-
-extern VALUE cXMLHtmlParserContext;
-
-void rxml_init_html_parser_context(void);
-
-#endif
+/* Please see the LICENSE file for copyright and distribution information */
+
+#ifndef __RXML_HTML_PARSER_CONTEXT__
+#define __RXML_HTML_PARSER_CONTEXT__
+
+extern VALUE cXMLHtmlParserContext;
+extern const rb_data_type_t rxml_html_parser_context_type;
+
+void rxml_init_html_parser_context(void);
+
+#endif

data/ext/libxml/ruby_xml_html_parser_options.c

@@ -1,48 +1,48 @@
-/* Please see the LICENSE file for copyright and distribution information */
-
-#include "ruby_libxml.h"
-
-#include <libxml/HTMLparser.h>
-
-/* Document-class: LibXML::XML::HTMLParser::Options
- *
- * Options to control the operation of the HTMLParser.  The easiest
- * way to set a parser's options is via the methods
- * XML::HTMLParser.file, XML::HTMLParser.io or XML::HTMLParser.string.
- * For additional control, see XML::HTMLParser::Context#options=.
-*/
-
-VALUE mXMLHtmlParserOptions;
-
-void rxml_init_html_parser_options(void)
-{
-  mXMLHtmlParserOptions = rb_define_module_under(cXMLHtmlParser, "Options");
-
-
-#if LIBXML_VERSION >= 20621
-  /* 1: Relax parsing. */
-  rb_define_const(mXMLHtmlParserOptions, "RECOVER", INT2NUM(HTML_PARSE_RECOVER)); 
-#endif
-#if LIBXML_VERSION >= 20708
-  /* 2:  Do not default a doctype if not found */
-  rb_define_const(mXMLHtmlParserOptions, "NODEFDTD", INT2NUM(HTML_PARSE_NODEFDTD));
-#endif
-  /* 32: Suppress error reports. */
-  rb_define_const(mXMLHtmlParserOptions, "NOERROR", INT2NUM(HTML_PARSE_NOERROR)); 
-  /* 64: Suppress warning reports. */
-  rb_define_const(mXMLHtmlParserOptions, "NOWARNING", INT2NUM(HTML_PARSE_NOWARNING));
-  /* 128: Enable pedantic error reporting. */
-  rb_define_const(mXMLHtmlParserOptions, "PEDANTIC", INT2NUM(HTML_PARSE_PEDANTIC)); 
-  /* 256: Remove blank nodes. */
-  rb_define_const(mXMLHtmlParserOptions, "NOBLANKS", INT2NUM(HTML_PARSE_NOBLANKS)); 
-#if LIBXML_VERSION >= 20621
-  /* 2048: Forbid network access. */
-  rb_define_const(mXMLHtmlParserOptions, "NONET", INT2NUM(HTML_PARSE_NONET)); 
-  /* 65536: Compact small text nodes. */
-  rb_define_const(mXMLHtmlParserOptions, "COMPACT", INT2NUM(HTML_PARSE_COMPACT));
-#endif
-#if LIBXML_VERSION >= 20707
-  /* 8192:  Do not add implied html/body... elements */
-  rb_define_const(mXMLHtmlParserOptions, "NOIMPLIED", INT2NUM(HTML_PARSE_NOIMPLIED));
-#endif
-}
+/* Please see the LICENSE file for copyright and distribution information */
+
+#include "ruby_libxml.h"
+
+#include <libxml/HTMLparser.h>
+
+/* Document-class: LibXML::XML::HTMLParser::Options
+ *
+ * Options to control the operation of the HTMLParser.  The easiest
+ * way to set a parser's options is via the methods
+ * XML::HTMLParser.file, XML::HTMLParser.io or XML::HTMLParser.string.
+ * For additional control, see XML::HTMLParser::Context#options=.
+*/
+
+VALUE mXMLHtmlParserOptions;
+
+void rxml_init_html_parser_options(void)
+{
+  mXMLHtmlParserOptions = rb_define_module_under(cXMLHtmlParser, "Options");
+
+
+#if LIBXML_VERSION >= 20621
+  /* 1: Relax parsing. */
+  rb_define_const(mXMLHtmlParserOptions, "RECOVER", INT2NUM(HTML_PARSE_RECOVER)); 
+#endif
+#if LIBXML_VERSION >= 20708
+  /* 2:  Do not default a doctype if not found */
+  rb_define_const(mXMLHtmlParserOptions, "NODEFDTD", INT2NUM(HTML_PARSE_NODEFDTD));
+#endif
+  /* 32: Suppress error reports. */
+  rb_define_const(mXMLHtmlParserOptions, "NOERROR", INT2NUM(HTML_PARSE_NOERROR)); 
+  /* 64: Suppress warning reports. */
+  rb_define_const(mXMLHtmlParserOptions, "NOWARNING", INT2NUM(HTML_PARSE_NOWARNING));
+  /* 128: Enable pedantic error reporting. */
+  rb_define_const(mXMLHtmlParserOptions, "PEDANTIC", INT2NUM(HTML_PARSE_PEDANTIC)); 
+  /* 256: Remove blank nodes. */
+  rb_define_const(mXMLHtmlParserOptions, "NOBLANKS", INT2NUM(HTML_PARSE_NOBLANKS)); 
+#if LIBXML_VERSION >= 20621
+  /* 2048: Forbid network access. */
+  rb_define_const(mXMLHtmlParserOptions, "NONET", INT2NUM(HTML_PARSE_NONET)); 
+  /* 65536: Compact small text nodes. */
+  rb_define_const(mXMLHtmlParserOptions, "COMPACT", INT2NUM(HTML_PARSE_COMPACT));
+#endif
+#if LIBXML_VERSION >= 20707
+  /* 8192:  Do not add implied html/body... elements */
+  rb_define_const(mXMLHtmlParserOptions, "NOIMPLIED", INT2NUM(HTML_PARSE_NOIMPLIED));
+#endif
+}

data/ext/libxml/ruby_xml_html_parser_options.h

@@ -1,10 +1,10 @@
-/* Please see the LICENSE file for copyright and distribution information */
-
-#ifndef __RXML_HTML_PARSER_OPTIONS__
-#define __RXML_HTML_PARSER_OPTIONS__
-
-extern VALUE mXMLHtmlParserOptions;
-
-void rxml_init_html_parser_options(void);
-
-#endif
+/* Please see the LICENSE file for copyright and distribution information */
+
+#ifndef __RXML_HTML_PARSER_OPTIONS__
+#define __RXML_HTML_PARSER_OPTIONS__
+
+extern VALUE mXMLHtmlParserOptions;
+
+void rxml_init_html_parser_options(void);
+
+#endif

data/ext/libxml/ruby_xml_input_cbg.c

@@ -8,8 +8,38 @@
  * Support for adding custom scheme handlers. */
 
 static ic_scheme *first_scheme = 0;
+static int input_callbacks_registered = 0;
 
-int ic_match(char const *filename)
+static int ic_match(char const *filename);
+static void* ic_open(char const *filename);
+static int ic_read(void *context, char *buffer, int len);
+static int ic_close(void *context);
+
+static void input_callbacks_register_once(void)
+{
+  if (input_callbacks_registered)
+    return;
+
+  xmlRegisterInputCallbacks(ic_match, ic_open, ic_read, ic_close);
+  input_callbacks_registered = 1;
+}
+
+static char *ic_strdup(const char *string)
+{
+  size_t length = strlen(string) + 1;
+  char *copy = ruby_xmalloc(length);
+  memcpy(copy, string, length);
+  return copy;
+}
+
+static void ic_scheme_free(ic_scheme *scheme)
+{
+  rb_gc_unregister_address(&scheme->class);
+  ruby_xfree(scheme->scheme_name);
+  ruby_xfree(scheme);
+}
+
+static int ic_match(char const *filename)
 {
   ic_scheme *scheme;
 
@@ -27,7 +57,7 @@ int ic_match(char const *filename)
   return 0;
 }
 
-void* ic_open(char const *filename)
+static void* ic_open(char const *filename)
 {
   ic_doc_context *ic_doc;
   ic_scheme *scheme;
@@ -38,12 +68,11 @@ void* ic_open(char const *filename)
   {
     if (!xmlStrncasecmp(BAD_CAST filename, BAD_CAST scheme->scheme_name, scheme->name_len))
     {
-      ic_doc = (ic_doc_context*) malloc(sizeof(ic_doc_context));
-
       res = rb_funcall(scheme->class, rb_intern("document_query"), 1,
           rb_str_new2(filename));
 
-      ic_doc->buffer = strdup(StringValuePtr(res));
+      ic_doc = ALLOC(ic_doc_context);
+      ic_doc->buffer = ic_strdup(StringValueCStr(res));
 
       ic_doc->bpos = ic_doc->buffer;
       ic_doc->remaining = (int)strlen(ic_doc->buffer);
@@ -54,7 +83,7 @@ void* ic_open(char const *filename)
   return 0;
 }
 
-int ic_read(void *context, char *buffer, int len)
+static int ic_read(void *context, char *buffer, int len)
 {
   ic_doc_context *ic_doc;
   int ret_len;
@@ -75,7 +104,7 @@ int ic_read(void *context, char *buffer, int len)
   return ret_len;
 }
 
-int ic_close(void *context)
+static int ic_close(void *context)
 {
   ruby_xfree(((ic_doc_context*) context)->buffer);
   ruby_xfree(context);
@@ -90,7 +119,7 @@ int ic_close(void *context)
  */
 static VALUE input_callbacks_register_input_callbacks(VALUE self)
 {
-  xmlRegisterInputCallbacks(ic_match, ic_open, ic_read, ic_close);
+  input_callbacks_register_once();
   return (Qtrue);
 }
 
@@ -107,11 +136,14 @@ static VALUE input_callbacks_add_scheme(VALUE self, VALUE scheme_name,
 
   Check_Type(scheme_name, T_STRING);
 
-  scheme = (ic_scheme*) malloc(sizeof(ic_scheme));
+  input_callbacks_register_once();
+
+  scheme = ALLOC(ic_scheme);
   scheme->next_scheme = 0;
-  scheme->scheme_name = strdup(StringValuePtr(scheme_name)); /* TODO alloc, dealloc */
+  scheme->scheme_name = ic_strdup(StringValueCStr(scheme_name));
   scheme->name_len = (int)strlen(scheme->scheme_name);
-  scheme->class = class; /* TODO alloc, dealloc */
+  scheme->class = class;
+  rb_gc_register_address(&scheme->class);
 
   //fprintf( stderr, "registered: %s, %d, %s\n", scheme->scheme_name, scheme->name_len, scheme->class );
 
@@ -150,9 +182,7 @@ static VALUE input_callbacks_remove_scheme(VALUE self, VALUE scheme_name)
   {
     save_scheme = first_scheme->next_scheme;
 
-    ruby_xfree(first_scheme->scheme_name);
-    ruby_xfree(first_scheme);
-
+    ic_scheme_free(first_scheme);
     first_scheme = save_scheme;
     return Qtrue;
   }
@@ -165,9 +195,7 @@ static VALUE input_callbacks_remove_scheme(VALUE self, VALUE scheme_name)
     {
       save_scheme = scheme->next_scheme->next_scheme;
 
-      ruby_xfree(scheme->next_scheme->scheme_name);
-      ruby_xfree(scheme->next_scheme);
-
+      ic_scheme_free(scheme->next_scheme);
       scheme->next_scheme = save_scheme;
       return Qtrue;
     }

data/ext/libxml/ruby_xml_input_cbg.h

@@ -1,20 +1,20 @@
-#ifndef _INPUT_CBG_
-#define _INPUT_CBG_
-
-void rxml_init_input_callbacks(void);
-
-typedef struct ic_doc_context {
-    char *buffer;
-    char *bpos;
-    int remaining;
-} ic_doc_context;
-
-typedef struct ic_scheme {
-    char *scheme_name;
-    VALUE class;
-    int name_len;
-
-    struct ic_scheme *next_scheme;
-} ic_scheme;
-
-#endif
+#ifndef _INPUT_CBG_
+#define _INPUT_CBG_
+
+void rxml_init_input_callbacks(void);
+
+typedef struct ic_doc_context {
+    char *buffer;
+    char *bpos;
+    int remaining;
+} ic_doc_context;
+
+typedef struct ic_scheme {
+    char *scheme_name;
+    VALUE class;
+    int name_len;
+
+    struct ic_scheme *next_scheme;
+} ic_scheme;
+
+#endif