lol_html/
lib.rs

1//! ***LOL HTML*** is a **L**ow **O**utput **L**atency streaming **HTML** rewriter/parser with
2//! CSS-selector based API.
3//!
4//! It is designed to modify HTML on the fly with minimal buffering. It can quickly handle very large
5//! documents, and operate in environments with limited memory resources.
6//!
7//! The crate serves as a back-end for the HTML rewriting functionality of [Cloudflare Workers], but
8//! can be used as a standalone library with the convenient API for a wide variety of HTML
9//! rewriting/analysis tasks.
10//!
11//! The crate provides two main API entry points:
12//!
13//! * [`HtmlRewriter`] - a streaming HTML rewriter;
14//! * [`rewrite_str`] - one-off HTML string rewriting function.
15//!
16//! [Cloudflare Workers]: https://www.cloudflare.com/en-gb/products/cloudflare-workers/
17//! [`HtmlRewriter`]: struct.HtmlRewriter.html
18//! [`rewrite_str`]: fn.rewrite_str.html
19#![forbid(unsafe_code)]
20#![allow(clippy::default_trait_access)]
21#![allow(clippy::module_name_repetitions)]
22#![allow(clippy::redundant_pub_crate)]
23#![deny(rustdoc::broken_intra_doc_links)]
24#![cfg_attr(not(any(feature = "integration_test", test)), warn(missing_docs))]
25#![cfg_attr(any(feature = "integration_test", test), allow(unnameable_types))]
26
27#[macro_use]
28mod base;
29
30#[macro_use]
31mod html;
32
33#[macro_use]
34mod rewriter;
35
36mod memory;
37mod parser;
38mod rewritable_units;
39mod transform_stream;
40
41use cfg_if::cfg_if;
42
43pub use self::rewriter::{
44    AsciiCompatibleEncoding, CommentHandler, DoctypeHandler, DocumentContentHandlers,
45    ElementContentHandlers, ElementHandler, EndHandler, EndTagHandler, HandlerResult, HandlerTypes,
46    HtmlRewriter, LocalHandlerTypes, MemorySettings, RewriteStrSettings, Settings, TextHandler,
47    rewrite_str,
48};
49pub use self::selectors_vm::Selector;
50pub use self::transform_stream::OutputSink;
51
52/// This module contains type aliases that make the [`HtmlRewriter`] safe to move between threads (have the [`Send`] bound).
53///
54/// The bound requires content handlers to be thread-safe, which prevents them from mutating external state without synchronization.
55///
56/// Rewriting is sequential, so there's no benefit from using the `Send`-compatible rewriter.
57pub mod send {
58    pub use crate::rewriter::{
59        CommentHandlerSend as CommentHandler, DoctypeHandlerSend as DoctypeHandler,
60        ElementHandlerSend as ElementHandler, EndHandlerSend as EndHandler,
61        EndTagHandlerSend as EndTagHandler, TextHandlerSend as TextHandler,
62    };
63    pub use crate::rewriter::{IntoHandler, SendHandlerTypes};
64
65    /// An [`HtmlRewriter`](crate::HtmlRewriter) that implements [`Send`].
66    pub type HtmlRewriter<'handlers, O> = crate::HtmlRewriter<'handlers, O, SendHandlerTypes>;
67    /// [`Settings`](crate::Settings) for [`Send`]able [`HtmlRewriter`](crate::HtmlRewriter)s.
68    pub type Settings<'handlers, 'selectors> =
69        crate::Settings<'handlers, 'selectors, SendHandlerTypes>;
70    /// [`RewriteStrSettings`](crate::RewriteStrSettings) for [`Send`]able [`HtmlRewriter`](crate::HtmlRewriter)s.
71    pub type RewriteStrSettings<'handlers, 'selectors> =
72        crate::RewriteStrSettings<'handlers, 'selectors, SendHandlerTypes>;
73
74    /// [`ElementContentHandlers`](crate::ElementContentHandlers) for [`Send`]able [`HtmlRewriter`](crate::HtmlRewriter)s.
75    pub type ElementContentHandlers<'h> = crate::ElementContentHandlers<'h, SendHandlerTypes>;
76    /// [`DocumentContentHandlers`](crate::DocumentContentHandlers) for [`Send`]able [`HtmlRewriter`](crate::HtmlRewriter)s.
77    pub type DocumentContentHandlers<'h> = crate::DocumentContentHandlers<'h, SendHandlerTypes>;
78
79    /// [`Element`](crate::rewritable_units::Element) for [`Send`]able [`HtmlRewriter`](crate::HtmlRewriter)s.
80    pub type Element<'rewriter, 'input_token> =
81        crate::rewritable_units::Element<'rewriter, 'input_token, SendHandlerTypes>;
82}
83
84/// The errors that can be produced by the crate's API.
85pub mod errors {
86    pub use super::memory::MemoryLimitExceededError;
87    pub use super::parser::ParsingAmbiguityError;
88    pub use super::rewritable_units::{
89        AttributeNameError, CommentTextError, TagNameError, Utf8Error,
90    };
91    pub use super::rewriter::RewritingError;
92    pub use super::selectors_vm::SelectorError;
93}
94
95/// HTML content descriptors that can be produced and modified by a rewriter.
96pub mod html_content {
97    pub use super::rewritable_units::{
98        Attribute, Comment, ContentType, Doctype, DocumentEnd, Element, EndTag, StartTag,
99        StreamingHandler, StreamingHandlerSink, TextChunk, UserData,
100    };
101
102    pub use super::base::SourceLocation;
103    pub use super::html::TextType;
104}
105
106#[cfg(any(test, feature = "integration_test"))]
107pub mod test_utils {
108    use encoding_rs::*;
109
110    pub static ASCII_COMPATIBLE_ENCODINGS: [&Encoding; 36] = [
111        BIG5,
112        EUC_JP,
113        EUC_KR,
114        GB18030,
115        GBK,
116        IBM866,
117        ISO_8859_2,
118        ISO_8859_3,
119        ISO_8859_4,
120        ISO_8859_5,
121        ISO_8859_6,
122        ISO_8859_7,
123        ISO_8859_8,
124        ISO_8859_8_I,
125        ISO_8859_10,
126        ISO_8859_13,
127        ISO_8859_14,
128        ISO_8859_15,
129        ISO_8859_16,
130        KOI8_R,
131        KOI8_U,
132        MACINTOSH,
133        SHIFT_JIS,
134        UTF_8,
135        WINDOWS_874,
136        WINDOWS_1250,
137        WINDOWS_1251,
138        WINDOWS_1252,
139        WINDOWS_1253,
140        WINDOWS_1254,
141        WINDOWS_1255,
142        WINDOWS_1256,
143        WINDOWS_1257,
144        WINDOWS_1258,
145        X_MAC_CYRILLIC,
146        X_USER_DEFINED,
147    ];
148
149    pub static NON_ASCII_COMPATIBLE_ENCODINGS: [&Encoding; 4] =
150        [UTF_16BE, UTF_16LE, ISO_2022_JP, REPLACEMENT];
151
152    pub struct Output {
153        bytes: Vec<u8>,
154        encoding: &'static Encoding,
155        finalizing_chunk_received: bool,
156    }
157
158    impl Output {
159        #[must_use]
160        #[inline]
161        pub fn new(encoding: &'static Encoding) -> Self {
162            Self {
163                bytes: Vec::default(),
164                encoding,
165                finalizing_chunk_received: false,
166            }
167        }
168
169        #[inline]
170        #[track_caller]
171        pub fn push(&mut self, chunk: &[u8]) {
172            if chunk.is_empty() {
173                self.finalizing_chunk_received = true;
174            } else {
175                assert!(
176                    !self.finalizing_chunk_received,
177                    "Chunk written to the output after the finalizing chunk."
178                );
179
180                self.bytes.extend_from_slice(chunk);
181            }
182        }
183    }
184
185    impl From<Output> for String {
186        #[inline]
187        #[track_caller]
188        fn from(output: Output) -> Self {
189            assert!(
190                output.finalizing_chunk_received,
191                "Finalizing chunk for the output hasn't been received."
192            );
193
194            output
195                .encoding
196                .decode_without_bom_handling(&output.bytes)
197                .0
198                .into_owned()
199        }
200    }
201}
202
203cfg_if! {
204    if #[cfg(feature = "integration_test")] {
205        pub mod selectors_vm;
206
207        pub use self::base::SharedEncoding;
208
209        pub use self::transform_stream::{
210            StartTagHandlingResult, TransformController, TransformStream,
211            TransformStreamSettings
212        };
213
214        pub use self::rewritable_units::{
215            EndTag, Serialize, StartTag, Token, TokenCaptureFlags,
216        };
217
218        pub use self::memory::SharedMemoryLimiter;
219        pub use self::html::{LocalName, LocalNameHash, Tag, Namespace};
220    } else {
221        mod selectors_vm;
222    }
223}
lol_html/lib.rs

lol_html/
lib.rs
