1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
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
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
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
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
use regex::Regex;
use lazy_static::lazy_static;
use select::node::{Descendants, Node};
use select::predicate::{Name, Predicate};
use std::borrow::Cow;
use std::collections::HashSet;
lazy_static! {
pub static ref RE_BAD_NODES_ATTR : Regex = Regex::new(r###"(?mi)^side$|combx|retweet|mediaarticlerelated|menucontainer|navbar|storytopbar-bucket|utility-bar|inline-share-tools|comment|PopularQuestions|contact|foot(er|note)?|cnn_strycaptiontxt|cnn_html_slideshow|cnn_strylftcntnt|links|meta$|shoutbox|sponsor|tags|socialnetworking|socialNetworking|cnnStryHghLght|cnn_stryspcvbx|^inset$|pagetools|post-attributes|welcome_form|contentTools2|the_answers|communitypromo|runaroundLeft|subscribe|vcard|articleheadings|date|^print$|popup|author-dropdown|tools|socialtools|byline|konafilter|breadcrumbs|^fn$|wp-caption-text|legende|ajoutVideo|timestamp|js_replies|[^-]facebook(-broadcasting)?|google|[^-]twitter|styln-briefing-block|read-more-link|js-body-read-more"###).unwrap();
}
pub const BAD_NODE_NAMES: &[&str; 5] = &["script", "style", "figcaption", "figure", "button"];
const ATTR_TO_CHECK: [&str; 3] = ["id", "class", "name"];
pub trait DocumentCleaner {
fn clean_node_text(&self, node: Node) -> String {
fn recur_text<T: DocumentCleaner + ?Sized>(
node: Node,
txt: &mut String,
cleaner: &T,
) -> bool {
if cleaner.is_bad_node_name(node) {
return false;
}
let mut txt_added = false;
if cleaner.is_good_node(node) {
let mut needs_ws = false;
for child in node.children() {
if needs_ws {
txt.push(' ');
needs_ws = false;
}
if let Some(txt_fragment) = child.as_text().map(str::trim) {
if !txt_fragment.is_empty() {
txt.push_str(txt_fragment);
txt_added = true
}
} else if Name("a").matches(&child) {
let mut a = String::new();
if recur_text(child, &mut a, cleaner) && txt_added {
txt.push(' ');
txt.push_str(&a);
needs_ws = true;
}
} else {
recur_text(child, txt, cleaner);
}
}
if (txt_added && is_para(node)) || needs_ws {
txt.push('\n');
}
}
txt_added
}
let mut txt = String::new();
recur_text(node, &mut txt, self);
txt
}
fn is_good_node(&self, node: Node) -> bool {
!has_bad_attr(node)
}
fn is_bad_node_name(&self, node: Node) -> bool {
is_bad_node(node)
}
fn iter_clean_nodes<'a>(&'a self, node: Node<'a>) -> CleanNodeIter<'a, Self>
where
Self: Sized,
{
CleanNodeIter {
cleaner: self,
inner: node.descendants(),
}
}
}
pub struct CleanNodeIter<'a, T: DocumentCleaner> {
cleaner: &'a T,
inner: Descendants<'a>,
}
impl<'a, T: DocumentCleaner> Iterator for CleanNodeIter<'a, T> {
type Item = Node<'a>;
fn next(&mut self) -> Option<Self::Item> {
let node = self.inner.next()?;
if self.cleaner.is_bad_node_name(node) || !self.cleaner.is_good_node(node) {
for ignore in node.descendants() {
let next = self.inner.next()?;
if ignore.index() != next.index() {
return Some(next);
}
}
}
Some(node)
}
}
pub struct DefaultDocumentCleaner;
impl DocumentCleaner for DefaultDocumentCleaner {}
#[derive(Debug)]
pub struct CommonCleaner<P>
where
P: Fn(Node) -> bool,
{
pub bad_node_names: HashSet<Cow<'static, str>>,
pub is_good_node: P,
}
impl<P> CommonCleaner<P>
where
P: Fn(Node) -> bool,
{
pub fn new(is_good_node: P) -> Self {
Self::with_names(is_good_node, BAD_NODE_NAMES.iter().cloned())
}
pub fn with_names<T, I>(is_good_node: P, bad_names: T) -> Self
where
T: IntoIterator<Item = I>,
I: Into<Cow<'static, str>>,
{
Self {
bad_node_names: bad_names.into_iter().map(I::into).collect(),
is_good_node,
}
}
}
impl Default for CommonCleaner<for<'s> fn(Node<'s>) -> bool> {
fn default() -> Self {
Self::new(|n| !is_bad_node(n))
}
}
impl<P> DocumentCleaner for CommonCleaner<P>
where
P: Fn(Node) -> bool,
{
fn is_good_node(&self, node: Node) -> bool {
(self.is_good_node)(node)
}
fn is_bad_node_name(&self, node: Node) -> bool {
if let Some(n) = node.name().map(Cow::Borrowed) {
self.bad_node_names.contains(&n)
} else {
true
}
}
}
pub fn is_bad_node(node: Node) -> bool {
if let Some(n) = node.name() {
BAD_NODE_NAMES.contains(&n)
} else {
false
}
}
fn is_para(node: Node) -> bool {
if let Some(n) = node.name() {
let names = &[
"blockquote",
"dl",
"div",
"img",
"ol",
"p",
"pre",
"table",
"ul",
];
names.contains(&n)
} else {
false
}
}
pub fn has_bad_attr(node: Node) -> bool {
for attr in ATTR_TO_CHECK.iter() {
if let Some(id) = node.attr(attr) {
if RE_BAD_NODES_ATTR.is_match(id) {
return true;
}
}
}
false
}