Class: Nokogiri::XML::Reader

Inherits:
Object
  • Object
show all
Includes:
Enumerable
Defined in:
lib/nokogiri/xml/reader.rb,
ext/nokogiri/xml_reader.c

Class Method Summary collapse

Instance Method Summary collapse

Class Method Details

.from_memory(string, url = nil, encoding = nil, options = 0) ⇒ Object

Create a new reader that parses string



363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
 
# File 'ext/nokogiri/xml_reader.c', line 363

static VALUE from_memory(int argc, VALUE *argv, VALUE klass)
{
  VALUE rb_buffer, rb_url, encoding, rb_options;

  const char * c_url      = NULL;
  const char * c_encoding = NULL;
  int c_options           = 0; 

  rb_scan_args(argc, argv, "13", &rb_buffer, &rb_url, &encoding, &rb_options);

  rb_buffer = StringValue(rb_buffer) ;
  if (RTEST(rb_url)) c_url = StringValuePtr(rb_url);
  if (RTEST(encoding)) c_encoding = StringValuePtr(rb_url);
  if (RTEST(rb_options)) c_options = NUM2INT(rb_options);

  xmlTextReaderPtr reader = xmlReaderForMemory(
      StringValuePtr(rb_buffer),
      NUM2INT(rb_funcall(rb_buffer, rb_intern("length"), 0)),
      c_url,
      c_encoding,
      c_options
  );

  if(reader == NULL) {
    xmlFreeTextReader(reader);
    rb_raise(rb_eRuntimeError, "couldn't create a parser");
  }

  return Data_Wrap_Struct(klass, NULL, dealloc, reader);
}

.bool ⇒ Object

Set the global XML default for load external subsets.



129
130
131
132
133
 
# File 'ext/nokogiri/xml_document.c', line 129

static VALUE load_external_subsets_set(VALUE klass, VALUE value)
{
    xmlLoadExtDtdDefaultValue = NUM2INT(value);
    return Qnil ;
}

.new(*args) ⇒ Object 



101
102
103
104
105
106
107
108
109
 
# File 'ext/nokogiri/xml_document.c', line 101

static VALUE new(int argc, VALUE *argv, VALUE klass)
{
  VALUE version;
  if(rb_scan_args(argc, argv, "01", &version) == 0)
    version = rb_str_new2("1.0");

  xmlDocPtr doc = xmlNewDoc((xmlChar *)StringValuePtr(version));
  return Nokogiri_wrap_xml_document(klass, doc);
}

.parse_stylesheet_doc(document) ⇒ Object

Parse a stylesheet from document.



20
21
22
23
24
25
26
27
 
# File 'ext/nokogiri/xslt_stylesheet.c', line 20

static VALUE parse_stylesheet_doc(VALUE klass, VALUE xmldocobj)
{
    xmlDocPtr xml ;
    xsltStylesheetPtr ss ;
    Data_Get_Struct(xmldocobj, xmlDoc, xml);
    ss = xsltParseStylesheetDoc(xmlCopyDoc(xml, 1)); /* 1 => recursive */
    return Data_Wrap_Struct(klass, NULL, dealloc, ss);
}

.read_memory(string, url, encoding, options) ⇒ Object

Read the HTML document contained in string with given url, encoding, and options. See Nokogiri::HTML.parse



79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
 
# File 'ext/nokogiri/xml_document.c', line 79

static VALUE read_memory( VALUE klass,
                          VALUE string,
                          VALUE url,
                          VALUE encoding,
                          VALUE options )
{
  const char * c_buffer = StringValuePtr(string);
  const char * c_url    = (url == Qnil) ? NULL : StringValuePtr(url);
  const char * c_enc    = (encoding == Qnil) ? NULL : StringValuePtr(encoding);
  int len               = NUM2INT(rb_funcall(string, rb_intern("length"), 0));

  xmlInitParser();
  xmlDocPtr doc = xmlReadMemory(c_buffer, len, c_url, c_enc, NUM2INT(options));

  if(doc == NULL) {
    xmlFreeDoc(doc);
    rb_raise(rb_eRuntimeError, "Couldn't create a document");
  }

  return Nokogiri_wrap_xml_document(klass, doc);
}

.bool ⇒ Object

Set the global XML default for substitute entities.



117
118
119
120
121
 
# File 'ext/nokogiri/xml_document.c', line 117

static VALUE substitute_entities_set(VALUE klass, VALUE value)
{
    xmlSubstituteEntitiesDefault(NUM2INT(value));
    return Qnil ;
}

Instance Method Details

#[](i) ⇒ Object

Get the node at index i



43
44
45
46
47
48
49
50
51
52
53
54
55
56
 
# File 'ext/nokogiri/xml_node_set.c', line 43

static VALUE index_at(VALUE self, VALUE number)
{
  int i = NUM2INT(number);
  xmlNodeSetPtr node_set;
  Data_Get_Struct(self, xmlNodeSet, node_set);

  if(i >= node_set->nodeNr || abs(i) > node_set->nodeNr)
    return Qnil;

  if(i < 0)
    i = i + node_set->nodeNr;

  return Nokogiri_wrap_xml_node(node_set->nodeTab[i]);
}

#apply_to(document, params) ⇒ Object

Apply an XSLT stylesheet to an XML::Document. params is an array of strings used as XSLT parameters.



60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
 
# File 'ext/nokogiri/xslt_stylesheet.c', line 60

static VALUE apply_to(int argc, VALUE* argv, VALUE self)
{
    VALUE xmldoc, paramobj ;
    xmlDocPtr xml ;
    xmlDocPtr result ;
    xsltStylesheetPtr ss ;
    const char** params ;
    int param_len, j ;
    VALUE resultobj ;

    rb_scan_args(argc, argv, "11", &xmldoc, &paramobj);
    if (paramobj == Qnil) { paramobj = rb_ary_new2(0) ; }

    Data_Get_Struct(xmldoc, xmlDoc, xml);
    Data_Get_Struct(self, xsltStylesheet, ss);

    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 = StringValuePtr(entry);
      params[j] = ptr;
    }
    params[param_len] = 0 ;

    result = xsltApplyStylesheet(ss, xml, params);
    free(params);
    resultobj = Nokogiri_wrap_xml_document(0, result) ;
    return rb_funcall(self, rb_intern("serialize"), 1, resultobj);
}

#attribute(name) ⇒ Object

Get the value of attribute named name



137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
 
# File 'ext/nokogiri/xml_reader.c', line 137

static VALUE reader_attribute(VALUE self, VALUE name)
{
  xmlTextReaderPtr reader;
  xmlChar *value ;
  Data_Get_Struct(self, xmlTextReader, reader);

  if(name == Qnil) return Qnil;
  name = StringValue(name) ;

  value = xmlTextReaderGetAttribute(reader, (xmlChar*)StringValuePtr(name));
  if(value == NULL) {
    /* this section is an attempt to workaround older versions of libxml that
       don't handle namespaces properly in all attribute-and-friends functions */
    xmlChar *prefix = NULL ;
    xmlChar *localname = xmlSplitQName2((xmlChar*)StringValuePtr(name), &prefix);
    if (localname != NULL) {
      value = xmlTextReaderLookupNamespace(reader, localname);
      free(localname) ;
    } else {
      value = xmlTextReaderLookupNamespace(reader, prefix);
    }
  }
  if(value == NULL) return Qnil;

  VALUE rb_value = rb_str_new2((const char *)value);
  xmlFree(value);
  return rb_value;
}

#attribute_at(index) ⇒ Object

Get the value of attribute at index



112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
 
# File 'ext/nokogiri/xml_reader.c', line 112

static VALUE attribute_at(VALUE self, VALUE index)
{
  xmlTextReaderPtr reader;
  Data_Get_Struct(self, xmlTextReader, reader);

  if(index == Qnil) return Qnil;
  index = rb_funcall(index, rb_intern("to_i"), 0);

  xmlChar * value = xmlTextReaderGetAttributeNo(
      reader,
      NUM2INT(index)
  );
  if(value == NULL) return Qnil;

  VALUE rb_value = rb_str_new2((const char *)value);
  xmlFree(value);
  return rb_value;
}

#attribute_count ⇒ Object

Get the number of attributes for the current node



172
173
174
175
176
177
178
179
180
 
# File 'ext/nokogiri/xml_reader.c', line 172

static VALUE attribute_count(VALUE self)
{
  xmlTextReaderPtr reader;
  Data_Get_Struct(self, xmlTextReader, reader);
  int count = xmlTextReaderAttributeCount(reader);
  if(count == -1) return Qnil;

  return INT2NUM(count);
}

#attributes ⇒ Object

Get a Hash of attributes for this node



85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
 
# File 'ext/nokogiri/xml_reader.c', line 85

static VALUE attributes(VALUE self)
{
  xmlTextReaderPtr reader;
  VALUE attr ;

  Data_Get_Struct(self, xmlTextReader, reader);

  attr = rb_hash_new() ;

  if (! has_attributes(reader))
    return attr ;

  xmlNodePtr ptr = xmlTextReaderExpand(reader);
  if(ptr == NULL) return Qnil;

  Nokogiri_xml_node_namespaces(ptr, attr);
  Nokogiri_xml_node_properties(ptr, attr);

  return attr ;
}

#attributes? ⇒ Boolean

Does this node have attributes?

Returns:

  • (Boolean) 


68
69
70
71
72
73
74
75
76
77
 
# File 'ext/nokogiri/xml_reader.c', line 68

static VALUE attributes_eh(VALUE self)
{
  xmlTextReaderPtr reader;
  Data_Get_Struct(self, xmlTextReader, reader);
  int eh = has_attributes(reader);
  if(eh == 0) return Qfalse;
  if(eh == 1) return Qtrue;

  return Qnil;
}

#default? ⇒ Boolean

Was an attribute generated from the default value in the DTD or schema?

Returns:

  • (Boolean) 


34
35
36
37
38
39
40
41
42
43
 
# File 'ext/nokogiri/xml_reader.c', line 34

static VALUE default_eh(VALUE self)
{
  xmlTextReaderPtr reader;
  Data_Get_Struct(self, xmlTextReader, reader);
  int eh = xmlTextReaderIsDefault(reader);
  if(eh == 0) return Qfalse;
  if(eh == 1) return Qtrue;

  return Qnil;
}

#depth ⇒ Object

Get the depth of the node



188
189
190
191
192
193
194
195
196
 
# File 'ext/nokogiri/xml_reader.c', line 188

static VALUE depth(VALUE self)
{
  xmlTextReaderPtr reader;
  Data_Get_Struct(self, xmlTextReader, reader);
  int depth = xmlTextReaderDepth(reader);
  if(depth == -1) return Qnil;

  return INT2NUM(depth);
}

#each(&block) ⇒ Object 



6
7
8
9
10
 
# File 'lib/nokogiri/xml/reader.rb', line 6

def each(&block)
  while node = self.read
    block.call(node)
  end
end

#encoding ⇒ Object

Get the encoding for the document



204
205
206
207
208
209
210
211
212
 
# File 'ext/nokogiri/xml_reader.c', line 204

static VALUE encoding(VALUE self)
{
  xmlTextReaderPtr reader;
  Data_Get_Struct(self, xmlTextReader, reader);
  const char * encoding = (const char *)xmlTextReaderConstEncoding(reader);
  if(encoding == NULL) return Qnil;

  return rb_str_new2(encoding);
}

#lang ⇒ Object

Get the xml:lang scope within which the node resides.



236
237
238
239
240
241
242
243
244
 
# File 'ext/nokogiri/xml_reader.c', line 236

static VALUE lang(VALUE self)
{
  xmlTextReaderPtr reader;
  Data_Get_Struct(self, xmlTextReader, reader);
  const char * lang = (const char *)xmlTextReaderConstXmlLang(reader);
  if(lang == NULL) return Qnil;

  return rb_str_new2(lang);
}

#length ⇒ Object

Get the length of the node set



9
10
11
12
13
14
15
16
17
18
 
# File 'ext/nokogiri/xml_node_set.c', line 9

static VALUE length(VALUE self)
{
  xmlNodeSetPtr node_set;
  Data_Get_Struct(self, xmlNodeSet, node_set);

  if(node_set)
    return INT2NUM(node_set->nodeNr);

  return INT2NUM(0);
}

#local_name ⇒ Object

Get the local name of the node



300
301
302
303
304
305
306
307
308
 
# File 'ext/nokogiri/xml_reader.c', line 300

static VALUE local_name(VALUE self)
{
  xmlTextReaderPtr reader;
  Data_Get_Struct(self, xmlTextReader, reader);
  const char * name = (const char *)xmlTextReaderConstLocalName(reader);
  if(name == NULL) return Qnil;

  return rb_str_new2(name);
}

#name ⇒ Object

Get the name of the node



316
317
318
319
320
321
322
323
324
 
# File 'ext/nokogiri/xml_reader.c', line 316

static VALUE name(VALUE self)
{
  xmlTextReaderPtr reader;
  Data_Get_Struct(self, xmlTextReader, reader);
  const char * name = (const char *)xmlTextReaderConstName(reader);
  if(name == NULL) return Qnil;

  return rb_str_new2(name);
}

#namespace_uri ⇒ Object

Get the URI defining the namespace associated with the node



284
285
286
287
288
289
290
291
292
 
# File 'ext/nokogiri/xml_reader.c', line 284

static VALUE namespace_uri(VALUE self)
{
  xmlTextReaderPtr reader;
  Data_Get_Struct(self, xmlTextReader, reader);
  const char * uri = (const char *)xmlTextReaderConstNamespaceUri(reader);
  if(uri == NULL) return Qnil;

  return rb_str_new2(uri);
}

#parse_memory(data) ⇒ Object

Parse the document stored in data



11
12
13
14
15
16
17
18
19
20
21
 
# File 'ext/nokogiri/xml_sax_parser.c', line 11

static VALUE parse_memory(VALUE self, VALUE data)
{
  xmlSAXHandlerPtr handler;
  Data_Get_Struct(self, xmlSAXHandler, handler);
  xmlSAXUserParseMemory(  handler,
                          (void *)self,
                          StringValuePtr(data),
                          NUM2INT(rb_funcall(data, rb_intern("length"), 0))
  );
  return data;
}

#prefix ⇒ Object

Get the shorthand reference to the namespace associated with the node.



268
269
270
271
272
273
274
275
276
 
# File 'ext/nokogiri/xml_reader.c', line 268

static VALUE prefix(VALUE self)
{
  xmlTextReaderPtr reader;
  Data_Get_Struct(self, xmlTextReader, reader);
  const char * prefix = (const char *)xmlTextReaderConstPrefix(reader);
  if(prefix == NULL) return Qnil;

  return rb_str_new2(prefix);
}

#push(node) ⇒ Object

Append node to the NodeSet.



26
27
28
29
30
31
32
33
34
35
 
# File 'ext/nokogiri/xml_node_set.c', line 26

static VALUE push(VALUE self, VALUE rb_node)
{
  xmlNodeSetPtr node_set;
  xmlNodePtr node;

  Data_Get_Struct(self, xmlNodeSet, node_set);
  Data_Get_Struct(rb_node, xmlNode, node);
  xmlXPathNodeSetAdd(node_set, node);
  return self;
}

#read ⇒ Object

Move the Reader forward through the XML document.



345
346
347
348
349
350
351
352
353
354
355
 
# File 'ext/nokogiri/xml_reader.c', line 345

static VALUE read_more(VALUE self)
{
  xmlTextReaderPtr reader;
  Data_Get_Struct(self, xmlTextReader, reader);

  int ret = xmlTextReaderRead(reader);
  if(ret == 1) return self;
  if(ret == 0) return Qnil;

  rb_raise(rb_eRuntimeError, "Error pulling: %d", ret);
}

#root ⇒ Object

Get the root node for this document.



68
69
70
71
72
73
74
75
76
77
 
# File 'ext/nokogiri/xml_document.c', line 68

static VALUE root(VALUE self)
{
  xmlDocPtr doc;
  Data_Get_Struct(self, xmlDoc, doc);

  xmlNodePtr root = xmlDocGetRootElement(doc);

  if(!root) return Qnil;
  return Nokogiri_wrap_xml_node(root) ;
}

#root= ⇒ Object

Set the root element on this document



49
50
51
52
53
54
55
56
57
58
59
60
 
# File 'ext/nokogiri/xml_document.c', line 49

static VALUE set_root(VALUE self, VALUE root)
{
  xmlDocPtr doc;
  xmlNodePtr new_root;

  Data_Get_Struct(self, xmlDoc, doc);
  Data_Get_Struct(root, xmlNode, new_root);

  xmlDocSetRootElement(doc, new_root);
  Nokogiri_xml_node_owned_set(new_root);
  return root;
}

#serialize(document) ⇒ Object

Serialize document to an xml string.



30
31
32
33
34
35
36
37
38
39
40
41
 
# File 'ext/nokogiri/xml_document.c', line 30

static VALUE serialize(VALUE self)
{
  xmlDocPtr doc;
  xmlChar *buf;
  int size;
  Data_Get_Struct(self, xmlDoc, doc);

  xmlDocDumpMemory(doc, &buf, &size);
  VALUE rb_str = rb_str_new((char *)buf, (long)size);
  free(buf);
  return rb_str;
}

#state ⇒ Object

Get the state of the reader



332
333
334
335
336
337
 
# File 'ext/nokogiri/xml_reader.c', line 332

static VALUE state(VALUE self)
{
  xmlTextReaderPtr reader;
  Data_Get_Struct(self, xmlTextReader, reader);
  return INT2NUM(xmlTextReaderReadState(reader));
}

#type ⇒ Object

The type for this document



54
55
56
57
58
59
 
# File 'ext/nokogiri/html_document.c', line 54

static VALUE type(VALUE self)
{
  htmlDocPtr doc;
  Data_Get_Struct(self, xmlDoc, doc);
  return INT2NUM((int)doc->type);
}

#value ⇒ Object

Get the text value of the node if present



252
253
254
255
256
257
258
259
260
 
# File 'ext/nokogiri/xml_reader.c', line 252

static VALUE value(VALUE self)
{
  xmlTextReaderPtr reader;
  Data_Get_Struct(self, xmlTextReader, reader);
  const char * value = (const char *)xmlTextReaderConstValue(reader);
  if(value == NULL) return Qnil;

  return rb_str_new2(value);
}

#value? ⇒ Boolean

Does this node have a text value?

Returns:

  • (Boolean) 


51
52
53
54
55
56
57
58
59
60
 
# File 'ext/nokogiri/xml_reader.c', line 51

static VALUE value_eh(VALUE self)
{
  xmlTextReaderPtr reader;
  Data_Get_Struct(self, xmlTextReader, reader);
  int eh = xmlTextReaderHasValue(reader);
  if(eh == 0) return Qfalse;
  if(eh == 1) return Qtrue;

  return Qnil;
}

#xml_version ⇒ Object

Get the XML version of the document being read



220
221
222
223
224
225
226
227
228
 
# File 'ext/nokogiri/xml_reader.c', line 220

static VALUE xml_version(VALUE self)
{
  xmlTextReaderPtr reader;
  Data_Get_Struct(self, xmlTextReader, reader);
  const char * version = (const char *)xmlTextReaderConstXmlVersion(reader);
  if(version == NULL) return Qnil;

  return rb_str_new2(version);
}