nokogumbo 1.4.9 → 1.4.10
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/ext/nokogumboc/nokogumbo.c +45 -8
- data/test-nokogumbo.rb +7 -0
- metadata +11 -11
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA1:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: c97bf9e43a0be85f20cf07ec15901052a1f9dbe1
|
4
|
+
data.tar.gz: 4c195ae6b3fc161648b7fc0c7eea06d14e8bd451
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: f49c427bdb96bce746e541c7864c755b0ece284f0f7fff8d0f3913403b8c9e9941b1d9c06ddec25142be4ee7ead2c97e96b9eac26eeee119cbb6cfc1879d5f7b
|
7
|
+
data.tar.gz: bf2b358e4f496f5f8e1c3b308e8c7befc308b9dda1af7399d4baa73cc59a395f630d315a5791565a088d7dfdf9c9a0eaaea5638798da6a2dfc6cea559df1fa25
|
data/ext/nokogumboc/nokogumbo.c
CHANGED
@@ -19,10 +19,13 @@
|
|
19
19
|
//
|
20
20
|
|
21
21
|
#include <ruby.h>
|
22
|
-
#include
|
22
|
+
#include "gumbo.h"
|
23
|
+
#include "error.h"
|
24
|
+
#include "parser.h"
|
23
25
|
|
24
26
|
// class constants
|
25
27
|
static VALUE Document;
|
28
|
+
static VALUE XMLSyntaxError;
|
26
29
|
|
27
30
|
#ifdef NGLIB
|
28
31
|
#include <nokogiri.h>
|
@@ -182,10 +185,10 @@ static xmlNodePtr walk_tree(xmlDocPtr document, GumboNode *node) {
|
|
182
185
|
|
183
186
|
// Parse a string using gumbo_parse into a Nokogiri document
|
184
187
|
static VALUE parse(VALUE self, VALUE string) {
|
185
|
-
|
186
|
-
|
187
|
-
|
188
|
-
);
|
188
|
+
const GumboOptions *options = &kGumboDefaultOptions;
|
189
|
+
const char *input = RSTRING_PTR(string);
|
190
|
+
size_t input_len = RSTRING_LEN(string);
|
191
|
+
GumboOutput *output = gumbo_parse_with_options(options, input, input_len);
|
189
192
|
xmlDocPtr doc = xmlNewDoc(CONST_CAST "1.0");
|
190
193
|
#ifdef NGLIB
|
191
194
|
doc->type = XML_HTML_DOCUMENT_NODE;
|
@@ -210,9 +213,42 @@ static VALUE parse(VALUE self, VALUE string) {
|
|
210
213
|
xmlAddChild((xmlNodePtr)doc, node);
|
211
214
|
}
|
212
215
|
}
|
213
|
-
gumbo_destroy_output(&kGumboDefaultOptions, output);
|
214
216
|
|
215
|
-
|
217
|
+
VALUE rdoc = Nokogiri_wrap_xml_document(Document, doc);
|
218
|
+
|
219
|
+
// Add parse errors to rdoc.
|
220
|
+
if (output->errors.length) {
|
221
|
+
GumboVector *errors = &output->errors;
|
222
|
+
GumboParser parser = { ._options = options };
|
223
|
+
GumboStringBuffer msg;
|
224
|
+
VALUE rerrors = rb_ary_new2(errors->length);
|
225
|
+
|
226
|
+
gumbo_string_buffer_init(&parser, &msg);
|
227
|
+
for (int i=0; i < errors->length; i++) {
|
228
|
+
GumboError *err = errors->data[i];
|
229
|
+
gumbo_string_buffer_clear(&parser, &msg);
|
230
|
+
gumbo_caret_diagnostic_to_string(&parser, err, input, &msg);
|
231
|
+
VALUE err_str = rb_str_new(msg.data, msg.length);
|
232
|
+
VALUE syntax_error = rb_class_new_instance(1, &err_str, XMLSyntaxError);
|
233
|
+
rb_iv_set(syntax_error, "@domain", INT2NUM(1)); // XML_FROM_PARSER
|
234
|
+
rb_iv_set(syntax_error, "@code", INT2NUM(1)); // XML_ERR_INTERNAL_ERROR
|
235
|
+
rb_iv_set(syntax_error, "@level", INT2NUM(2)); // XML_ERR_ERROR
|
236
|
+
rb_iv_set(syntax_error, "@file", Qnil);
|
237
|
+
rb_iv_set(syntax_error, "@line", INT2NUM(err->position.line));
|
238
|
+
rb_iv_set(syntax_error, "@str1", Qnil);
|
239
|
+
rb_iv_set(syntax_error, "@str2", Qnil);
|
240
|
+
rb_iv_set(syntax_error, "@str3", Qnil);
|
241
|
+
rb_iv_set(syntax_error, "@int1", INT2NUM(err->type));
|
242
|
+
rb_iv_set(syntax_error, "@column", INT2NUM(err->position.column));
|
243
|
+
rb_ary_push(rerrors, syntax_error);
|
244
|
+
}
|
245
|
+
rb_iv_set(rdoc, "@errors", rerrors);
|
246
|
+
gumbo_string_buffer_destroy(&parser, &msg);
|
247
|
+
}
|
248
|
+
|
249
|
+
gumbo_destroy_output(options, output);
|
250
|
+
|
251
|
+
return rdoc;
|
216
252
|
}
|
217
253
|
|
218
254
|
// Initialize the Nokogumbo class and fetch constants we will use later
|
@@ -224,10 +260,11 @@ void Init_nokogumboc() {
|
|
224
260
|
VALUE Nokogiri = rb_const_get(rb_cObject, rb_intern("Nokogiri"));
|
225
261
|
VALUE HTML = rb_const_get(Nokogiri, rb_intern("HTML"));
|
226
262
|
Document = rb_const_get(HTML, rb_intern("Document"));
|
263
|
+
VALUE XML = rb_const_get(Nokogiri, rb_intern("XML"));
|
264
|
+
XMLSyntaxError = rb_const_get(XML, rb_intern("SyntaxError"));
|
227
265
|
|
228
266
|
#ifndef NGLIB
|
229
267
|
// more class constants
|
230
|
-
VALUE XML = rb_const_get(Nokogiri, rb_intern("XML"));
|
231
268
|
Element = rb_const_get(XML, rb_intern("Element"));
|
232
269
|
Text = rb_const_get(XML, rb_intern("Text"));
|
233
270
|
CDATA = rb_const_get(XML, rb_intern("CDATA"));
|
data/test-nokogumbo.rb
CHANGED
@@ -125,6 +125,13 @@ class TestNokogumbo < Minitest::Test
|
|
125
125
|
assert_equal ["html", "comment", "html", "comment"], doc.children.map(&:name)
|
126
126
|
end
|
127
127
|
|
128
|
+
def test_parse_errors
|
129
|
+
doc = Nokogiri::HTML5("<!DOCTYPE html><html><!-- -- --></a>")
|
130
|
+
assert_equal doc.errors.length, 2
|
131
|
+
doc = Nokogiri::HTML5("<!DOCTYPE html><html>")
|
132
|
+
assert_empty doc.errors
|
133
|
+
end
|
134
|
+
|
128
135
|
private
|
129
136
|
|
130
137
|
def buffer
|
metadata
CHANGED
@@ -1,27 +1,27 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: nokogumbo
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 1.4.
|
4
|
+
version: 1.4.10
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Sam Ruby
|
8
8
|
autorequire:
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
|
-
date: 2016-
|
11
|
+
date: 2016-11-01 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: nokogiri
|
15
15
|
requirement: !ruby/object:Gem::Requirement
|
16
16
|
requirements:
|
17
|
-
- -
|
17
|
+
- - ">="
|
18
18
|
- !ruby/object:Gem::Version
|
19
19
|
version: '0'
|
20
20
|
type: :runtime
|
21
21
|
prerelease: false
|
22
22
|
version_requirements: !ruby/object:Gem::Requirement
|
23
23
|
requirements:
|
24
|
-
- -
|
24
|
+
- - ">="
|
25
25
|
- !ruby/object:Gem::Version
|
26
26
|
version: '0'
|
27
27
|
description: Nokogumbo allows a Ruby program to invoke the Gumbo HTML5 parser and
|
@@ -32,11 +32,10 @@ extensions:
|
|
32
32
|
- ext/nokogumboc/extconf.rb
|
33
33
|
extra_rdoc_files: []
|
34
34
|
files:
|
35
|
-
- ext/nokogumboc/extconf.rb
|
36
|
-
- ext/nokogumboc/nokogumbo.c
|
37
|
-
- lib/nokogumbo.rb
|
38
35
|
- LICENSE.txt
|
39
36
|
- README.md
|
37
|
+
- ext/nokogumboc/extconf.rb
|
38
|
+
- ext/nokogumboc/nokogumbo.c
|
40
39
|
- gumbo-parser/src/attribute.c
|
41
40
|
- gumbo-parser/src/attribute.h
|
42
41
|
- gumbo-parser/src/char_ref.c
|
@@ -69,10 +68,11 @@ files:
|
|
69
68
|
- gumbo-parser/src/vector.c
|
70
69
|
- gumbo-parser/src/vector.h
|
71
70
|
- gumbo-parser/visualc/include/strings.h
|
71
|
+
- lib/nokogumbo.rb
|
72
72
|
- test-nokogumbo.rb
|
73
73
|
homepage: https://github.com/rubys/nokogumbo/#readme
|
74
74
|
licenses:
|
75
|
-
- Apache
|
75
|
+
- Apache-2.0
|
76
76
|
metadata: {}
|
77
77
|
post_install_message:
|
78
78
|
rdoc_options: []
|
@@ -80,17 +80,17 @@ require_paths:
|
|
80
80
|
- lib
|
81
81
|
required_ruby_version: !ruby/object:Gem::Requirement
|
82
82
|
requirements:
|
83
|
-
- -
|
83
|
+
- - ">="
|
84
84
|
- !ruby/object:Gem::Version
|
85
85
|
version: '0'
|
86
86
|
required_rubygems_version: !ruby/object:Gem::Requirement
|
87
87
|
requirements:
|
88
|
-
- -
|
88
|
+
- - ">="
|
89
89
|
- !ruby/object:Gem::Version
|
90
90
|
version: '0'
|
91
91
|
requirements: []
|
92
92
|
rubyforge_project:
|
93
|
-
rubygems_version: 2.
|
93
|
+
rubygems_version: 2.5.1
|
94
94
|
signing_key:
|
95
95
|
specification_version: 4
|
96
96
|
summary: Nokogiri interface to the Gumbo HTML5 parser
|