Expand description
Parsec: an HTML parser that follows the WHATWG HTML Standard, with CSS selectors and a tree you can change and write back out, and the other standards a page needs beside it.
The parser is at the top of this crate, and each standard is a module behind a cargo feature:
| Module | Feature | Default |
|---|---|---|
encoding | encoding | on |
url | url | on |
xpath | xpath | on |
stylesheet | css | off |
css is the selector engine, which is always here; stylesheet is
CSS Syntax and the cascade, which a scraper rarely needs.
use parsec::css::SelectorList;
let doc = parsec::parse("<ul><li class=hit>one<li>two</ul>", false);
let hits = SelectorList::parse("li.hit").unwrap().select(&doc, doc.root());
assert_eq!(doc.text(hits[0]), "one");Modules§
- css
- CSS selectors, as Selectors Level 4 defines them and browsers match them in a static page: parsed once, matched right to left.
- encoding
- Bytes into Parsec.
parse_bytesfinds a page’s encoding the way the HTML Standard says a browser does, decodes it and parses it.ByteParserdoes the same with the bytes as they arrive. - stylesheet
- CSS Syntax Level 3: the tokenizer, the parser and its entry points, the
An+B and
<urange>microsyntaxes, and serialization back to CSS, as the W3C’s current editor’s draft defines them, nesting included. - tokenizer
- The tokenizer, as section 13.2.5 of the HTML Standard describes it.
- url
- The WHATWG URL Standard: URLs with every getter and setter the
URLinterface has,URLSearchParams, and the host, IDNA and Punycode algorithms under them. - xpath
- XPath 1.0 over a parsec-html
Document, written from the W3C recommendation: every axis, node test and function of the core library, with the number, string and boolean rules of sections 3 and 4.
Structs§
- Document
- A parsed HTML document, or a fragment.
- Options
- How to parse.
- Parser
- Parses a document as it arrives, a piece at a time, into the tree
parse_withwould build from the whole: the same for any way of cutting the input up, with the same spans and errors when asked for. - Serialize
Options - How
Document::serialize_withwrites a node. The default is the HTML Standard’s serialization, byte for byte whatDocument::outer_htmlgives. Most fields are html5lib’s serializer options, under html5lib’s names. - Shadow
Root - A shadow root’s settings, as
attachShadowtakes them and a declarative<template shadowrootmode>sets them. - Syntax
Error - A parse error the HTML Standard defines, from
Document::errors. The parser recovered from it, as the spec says to; it’s for a linter or an editor to report.
Enums§
- AttrNs
- A namespaced attribute’s namespace, from the spec’s “adjust foreign
attributes” table;
Nonefor every ordinary attribute. - Edit
Error - Why an edit was refused.
- Kind
- What a node is, as
Document::kindreports it. - Ns
- Parse
Error - Why a parse returned no document.
- Quirks
- Shadow
Root Mode - Whether a shadow root’s host lets scripts reach it, from
shadowrootmodeorattachShadow’smode. - Slot
Assignment - How a shadow root’s slots get their nodes: by the
slotattribute of the host’s children, or only asDocument::assignsays.
Constants§
- MAX_
INPUT - The most input a document can hold. Spans keep offsets in 30 bits. The parse buffer, which holds the strings the parser had to change, can outgrow the input (each NUL it replaces is three bytes of U+FFFD), so a string that would take it past 30 bits goes in a string of its own instead.
Functions§
- is_
html_ element - Whether
name(in any case) names an element of HTML, current or obsolete, or the root of SVG or MathML:div,marquee,svg; notrss. - named_
reference - The character(s) HTML’s named reference
namestands for, without the&and with or without the;as the table has it:nbsp;gives U+00A0. - parse
- Parses a whole document. A leading byte order mark (U+FEFF) is dropped, as decoding the page would have.
- parse_
fragment - Parses
htmlas the contents of an element namedcontextin namespacens, asinnerHTMLdoes. - parse_
fragment_ with - Parses
htmlasparse_fragmentdoes, within a size limit and an optional budget, asparse_withdoes for a whole document. - parse_
with - Parses a whole document, within a size limit and an optional budget.
- tag_
number - The number
Document::known_taggives elements namedname(lowercase), if the parser knows the tag.
Type Aliases§
- NodeId
- A node’s index in its document; the document itself is node 0.