script/dom/
characterdata.rs

1/* This Source Code Form is subject to the terms of the Mozilla Public
2 * License, v. 2.0. If a copy of the MPL was not distributed with this
3 * file, You can obtain one at https://mozilla.org/MPL/2.0/. */
4
5//! DOM bindings for `CharacterData`.
6use std::cell::LazyCell;
7
8use dom_struct::dom_struct;
9use script_bindings::codegen::InheritTypes::{CharacterDataTypeId, NodeTypeId, TextTypeId};
10
11use crate::dom::bindings::cell::{DomRefCell, Ref};
12use crate::dom::bindings::codegen::Bindings::CharacterDataBinding::CharacterDataMethods;
13use crate::dom::bindings::codegen::Bindings::NodeBinding::Node_Binding::NodeMethods;
14use crate::dom::bindings::codegen::Bindings::ProcessingInstructionBinding::ProcessingInstructionMethods;
15use crate::dom::bindings::codegen::UnionTypes::NodeOrString;
16use crate::dom::bindings::error::{Error, ErrorResult, Fallible};
17use crate::dom::bindings::inheritance::Castable;
18use crate::dom::bindings::root::{DomRoot, LayoutDom};
19use crate::dom::bindings::str::DOMString;
20use crate::dom::cdatasection::CDATASection;
21use crate::dom::comment::Comment;
22use crate::dom::document::Document;
23use crate::dom::element::Element;
24use crate::dom::mutationobserver::{Mutation, MutationObserver};
25use crate::dom::node::{ChildrenMutation, Node, NodeDamage};
26use crate::dom::processinginstruction::ProcessingInstruction;
27use crate::dom::text::Text;
28use crate::dom::virtualmethods::vtable_for;
29use crate::script_runtime::CanGc;
30
31// https://dom.spec.whatwg.org/#characterdata
32#[dom_struct]
33pub(crate) struct CharacterData {
34    node: Node,
35    data: DomRefCell<DOMString>,
36}
37
38impl CharacterData {
39    pub(crate) fn new_inherited(data: DOMString, document: &Document) -> CharacterData {
40        CharacterData {
41            node: Node::new_inherited(document),
42            data: DomRefCell::new(data),
43        }
44    }
45
46    pub(crate) fn clone_with_data(
47        &self,
48        data: DOMString,
49        document: &Document,
50        can_gc: CanGc,
51    ) -> DomRoot<Node> {
52        match self.upcast::<Node>().type_id() {
53            NodeTypeId::CharacterData(CharacterDataTypeId::Comment) => {
54                DomRoot::upcast(Comment::new(data, document, None, can_gc))
55            },
56            NodeTypeId::CharacterData(CharacterDataTypeId::ProcessingInstruction) => {
57                let pi = self.downcast::<ProcessingInstruction>().unwrap();
58                DomRoot::upcast(ProcessingInstruction::new(
59                    pi.Target(),
60                    data,
61                    document,
62                    can_gc,
63                ))
64            },
65            NodeTypeId::CharacterData(CharacterDataTypeId::Text(TextTypeId::CDATASection)) => {
66                DomRoot::upcast(CDATASection::new(data, document, can_gc))
67            },
68            NodeTypeId::CharacterData(CharacterDataTypeId::Text(TextTypeId::Text)) => {
69                DomRoot::upcast(Text::new(data, document, can_gc))
70            },
71            _ => unreachable!(),
72        }
73    }
74
75    #[inline]
76    pub(crate) fn data(&self) -> Ref<'_, DOMString> {
77        self.data.borrow()
78    }
79
80    #[inline]
81    pub(crate) fn append_data(&self, data: &str) {
82        self.queue_mutation_record();
83        self.data.borrow_mut().push_str(data);
84        self.content_changed();
85    }
86
87    fn content_changed(&self) {
88        let node = self.upcast::<Node>();
89        node.dirty(NodeDamage::Other);
90
91        // If this is a Text node, we might need to re-parse (say, if our parent
92        // is a <style> element.) We don't need to if this is a Comment or
93        // ProcessingInstruction.
94        if self.is::<Text>() {
95            if let Some(parent_node) = node.GetParentNode() {
96                let mutation = ChildrenMutation::ChangeText;
97                vtable_for(&parent_node).children_changed(&mutation);
98            }
99        }
100    }
101
102    // Queue a MutationObserver record before changing the content.
103    fn queue_mutation_record(&self) {
104        let mutation = LazyCell::new(|| Mutation::CharacterData {
105            old_value: self.data.borrow().clone(),
106        });
107        MutationObserver::queue_a_mutation_record(self.upcast::<Node>(), mutation);
108    }
109}
110
111impl CharacterDataMethods<crate::DomTypeHolder> for CharacterData {
112    // https://dom.spec.whatwg.org/#dom-characterdata-data
113    fn Data(&self) -> DOMString {
114        self.data.borrow().clone()
115    }
116
117    // https://dom.spec.whatwg.org/#dom-characterdata-data
118    fn SetData(&self, data: DOMString) {
119        self.queue_mutation_record();
120        let old_length = self.Length();
121        let new_length = data.encode_utf16().count() as u32;
122        *self.data.borrow_mut() = data;
123        self.content_changed();
124        let node = self.upcast::<Node>();
125        node.ranges()
126            .replace_code_units(node, 0, old_length, new_length);
127    }
128
129    // https://dom.spec.whatwg.org/#dom-characterdata-length
130    fn Length(&self) -> u32 {
131        self.data.borrow().encode_utf16().count() as u32
132    }
133
134    // https://dom.spec.whatwg.org/#dom-characterdata-substringdata
135    fn SubstringData(&self, offset: u32, count: u32) -> Fallible<DOMString> {
136        let data = self.data.borrow();
137        // Step 1.
138        let mut substring = String::new();
139        let remaining = match split_at_utf16_code_unit_offset(&data, offset) {
140            Ok((_, astral, s)) => {
141                // As if we had split the UTF-16 surrogate pair in half
142                // and then transcoded that to UTF-8 lossily,
143                // since our DOMString is currently strict UTF-8.
144                if astral.is_some() {
145                    substring += "\u{FFFD}";
146                }
147                s
148            },
149            // Step 2.
150            Err(()) => return Err(Error::IndexSize),
151        };
152        match split_at_utf16_code_unit_offset(remaining, count) {
153            // Steps 3.
154            Err(()) => substring += remaining,
155            // Steps 4.
156            Ok((s, astral, _)) => {
157                substring += s;
158                // As if we had split the UTF-16 surrogate pair in half
159                // and then transcoded that to UTF-8 lossily,
160                // since our DOMString is currently strict UTF-8.
161                if astral.is_some() {
162                    substring += "\u{FFFD}";
163                }
164            },
165        };
166        Ok(DOMString::from(substring))
167    }
168
169    // https://dom.spec.whatwg.org/#dom-characterdata-appenddatadata
170    fn AppendData(&self, data: DOMString) {
171        // FIXME(ajeffrey): Efficient append on DOMStrings?
172        self.append_data(&data);
173    }
174
175    // https://dom.spec.whatwg.org/#dom-characterdata-insertdataoffset-data
176    fn InsertData(&self, offset: u32, arg: DOMString) -> ErrorResult {
177        self.ReplaceData(offset, 0, arg)
178    }
179
180    // https://dom.spec.whatwg.org/#dom-characterdata-deletedataoffset-count
181    fn DeleteData(&self, offset: u32, count: u32) -> ErrorResult {
182        self.ReplaceData(offset, count, DOMString::new())
183    }
184
185    // https://dom.spec.whatwg.org/#dom-characterdata-replacedata
186    fn ReplaceData(&self, offset: u32, count: u32, arg: DOMString) -> ErrorResult {
187        let mut new_data;
188        {
189            let data = self.data.borrow();
190            let prefix;
191            let replacement_before;
192            let remaining;
193            match split_at_utf16_code_unit_offset(&data, offset) {
194                Ok((p, astral, r)) => {
195                    prefix = p;
196                    // As if we had split the UTF-16 surrogate pair in half
197                    // and then transcoded that to UTF-8 lossily,
198                    // since our DOMString is currently strict UTF-8.
199                    replacement_before = if astral.is_some() { "\u{FFFD}" } else { "" };
200                    remaining = r;
201                },
202                // Step 2.
203                Err(()) => return Err(Error::IndexSize),
204            };
205            let replacement_after;
206            let suffix;
207            match split_at_utf16_code_unit_offset(remaining, count) {
208                // Steps 3.
209                Err(()) => {
210                    replacement_after = "";
211                    suffix = "";
212                },
213                Ok((_, astral, s)) => {
214                    // As if we had split the UTF-16 surrogate pair in half
215                    // and then transcoded that to UTF-8 lossily,
216                    // since our DOMString is currently strict UTF-8.
217                    replacement_after = if astral.is_some() { "\u{FFFD}" } else { "" };
218                    suffix = s;
219                },
220            };
221            // Step 4: Mutation observers.
222            self.queue_mutation_record();
223
224            // Step 5 to 7.
225            new_data = String::with_capacity(
226                prefix.len() +
227                    replacement_before.len() +
228                    arg.len() +
229                    replacement_after.len() +
230                    suffix.len(),
231            );
232            new_data.push_str(prefix);
233            new_data.push_str(replacement_before);
234            new_data.push_str(&arg);
235            new_data.push_str(replacement_after);
236            new_data.push_str(suffix);
237        }
238        *self.data.borrow_mut() = DOMString::from(new_data);
239        self.content_changed();
240        // Steps 8-11.
241        let node = self.upcast::<Node>();
242        node.ranges()
243            .replace_code_units(node, offset, count, arg.encode_utf16().count() as u32);
244        Ok(())
245    }
246
247    // https://dom.spec.whatwg.org/#dom-childnode-before
248    fn Before(&self, nodes: Vec<NodeOrString>, can_gc: CanGc) -> ErrorResult {
249        self.upcast::<Node>().before(nodes, can_gc)
250    }
251
252    // https://dom.spec.whatwg.org/#dom-childnode-after
253    fn After(&self, nodes: Vec<NodeOrString>, can_gc: CanGc) -> ErrorResult {
254        self.upcast::<Node>().after(nodes, can_gc)
255    }
256
257    // https://dom.spec.whatwg.org/#dom-childnode-replacewith
258    fn ReplaceWith(&self, nodes: Vec<NodeOrString>, can_gc: CanGc) -> ErrorResult {
259        self.upcast::<Node>().replace_with(nodes, can_gc)
260    }
261
262    // https://dom.spec.whatwg.org/#dom-childnode-remove
263    fn Remove(&self, can_gc: CanGc) {
264        let node = self.upcast::<Node>();
265        node.remove_self(can_gc);
266    }
267
268    // https://dom.spec.whatwg.org/#dom-nondocumenttypechildnode-previouselementsibling
269    fn GetPreviousElementSibling(&self) -> Option<DomRoot<Element>> {
270        self.upcast::<Node>()
271            .preceding_siblings()
272            .filter_map(DomRoot::downcast)
273            .next()
274    }
275
276    // https://dom.spec.whatwg.org/#dom-nondocumenttypechildnode-nextelementsibling
277    fn GetNextElementSibling(&self) -> Option<DomRoot<Element>> {
278        self.upcast::<Node>()
279            .following_siblings()
280            .filter_map(DomRoot::downcast)
281            .next()
282    }
283}
284
285pub(crate) trait LayoutCharacterDataHelpers<'dom> {
286    fn data_for_layout(self) -> &'dom str;
287}
288
289impl<'dom> LayoutCharacterDataHelpers<'dom> for LayoutDom<'dom, CharacterData> {
290    #[allow(unsafe_code)]
291    #[inline]
292    fn data_for_layout(self) -> &'dom str {
293        unsafe { self.unsafe_get().data.borrow_for_layout() }
294    }
295}
296
297/// Split the given string at the given position measured in UTF-16 code units from the start.
298///
299/// * `Err(())` indicates that `offset` if after the end of the string
300/// * `Ok((before, None, after))` indicates that `offset` is between Unicode code points.
301///   The two string slices are such that:
302///   `before == s.to_utf16()[..offset].to_utf8()` and
303///   `after == s.to_utf16()[offset..].to_utf8()`
304/// * `Ok((before, Some(ch), after))` indicates that `offset` is "in the middle"
305///   of a single Unicode code point that would be represented in UTF-16 by a surrogate pair
306///   of two 16-bit code units.
307///   `ch` is that code point.
308///   The two string slices are such that:
309///   `before == s.to_utf16()[..offset - 1].to_utf8()` and
310///   `after == s.to_utf16()[offset + 1..].to_utf8()`
311fn split_at_utf16_code_unit_offset(s: &str, offset: u32) -> Result<(&str, Option<char>, &str), ()> {
312    let mut code_units = 0;
313    for (i, c) in s.char_indices() {
314        if code_units == offset {
315            let (a, b) = s.split_at(i);
316            return Ok((a, None, b));
317        }
318        code_units += 1;
319        if c > '\u{FFFF}' {
320            if code_units == offset {
321                debug_assert_eq!(c.len_utf8(), 4);
322                warn!("Splitting a surrogate pair in CharacterData API.");
323                return Ok((&s[..i], Some(c), &s[i + c.len_utf8()..]));
324            }
325            code_units += 1;
326        }
327    }
328    if code_units == offset {
329        Ok((s, None, ""))
330    } else {
331        Err(())
332    }
333}
script/dom/characterdata.rs

script/dom/
characterdata.rs
