Ensure that qualified-name segments start with a valid start character (#35530)

* Add spec comments to various methods

Signed-off-by: Simon Wülker <simon.wuelker@arcor.de>

* Ensure that qualified-name segments start with a valid start character

Signed-off-by: Simon Wülker <simon.wuelker@arcor.de>

* Update WPT expectations

Signed-off-by: Simon Wülker <simon.wuelker@arcor.de>

---------

Signed-off-by: Simon Wülker <simon.wuelker@arcor.de>
This commit is contained in:
Simon Wülker 2025-02-19 05:34:42 +01:00 committed by GitHub
parent b57eba2919
commit 29e0fad21e
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
6 changed files with 141 additions and 161 deletions

View file

@ -6,17 +6,91 @@
use html5ever::{namespace_url, ns, LocalName, Namespace, Prefix};
use crate::dom::bindings::error::{Error, ErrorResult, Fallible};
use crate::dom::bindings::error::{Error, Fallible};
use crate::dom::bindings::str::DOMString;
/// Check if an element name is valid. See <http://www.w3.org/TR/xml/#NT-Name>
/// for details.
fn is_valid_start(c: char) -> bool {
matches!(c, ':' |
'A'..='Z' |
'_' |
'a'..='z' |
'\u{C0}'..='\u{D6}' |
'\u{D8}'..='\u{F6}' |
'\u{F8}'..='\u{2FF}' |
'\u{370}'..='\u{37D}' |
'\u{37F}'..='\u{1FFF}' |
'\u{200C}'..='\u{200D}' |
'\u{2070}'..='\u{218F}' |
'\u{2C00}'..='\u{2FEF}' |
'\u{3001}'..='\u{D7FF}' |
'\u{F900}'..='\u{FDCF}' |
'\u{FDF0}'..='\u{FFFD}' |
'\u{10000}'..='\u{EFFFF}')
}
fn is_valid_continuation(c: char) -> bool {
is_valid_start(c) ||
matches!(c,
'-' |
'.' |
'0'..='9' |
'\u{B7}' |
'\u{300}'..='\u{36F}' |
'\u{203F}'..='\u{2040}')
}
/// Validate a qualified name. See <https://dom.spec.whatwg.org/#validate> for details.
pub(crate) fn validate_qualified_name(qualified_name: &str) -> ErrorResult {
// Step 2.
match xml_name_type(qualified_name) {
XMLName::Invalid => Err(Error::InvalidCharacter),
XMLName::Name => Err(Error::InvalidCharacter), // see whatwg/dom#671
XMLName::QName => Ok(()),
///
/// On success, this returns a tuple `(prefix, local name)`.
pub(crate) fn validate_and_extract_qualified_name(
qualified_name: &str,
) -> Fallible<(Option<&str>, &str)> {
if qualified_name.is_empty() {
// Qualified names must not be empty
return Err(Error::InvalidCharacter);
}
let mut colon_offset = None;
let mut at_start_of_name = true;
for (byte_position, c) in qualified_name.char_indices() {
if c == ':' {
if colon_offset.is_some() {
// Qualified names must not contain more than one colon
return Err(Error::InvalidCharacter);
}
colon_offset = Some(byte_position);
at_start_of_name = true;
continue;
}
if at_start_of_name {
if !is_valid_start(c) {
// Name segments must begin with a valid start character
return Err(Error::InvalidCharacter);
}
at_start_of_name = false;
} else if !is_valid_continuation(c) {
// Name segments must consist of valid characters
return Err(Error::InvalidCharacter);
}
}
let Some(colon_offset) = colon_offset else {
// Simple case: there is no prefix
return Ok((None, qualified_name));
};
let (prefix, local_name) = qualified_name.split_at(colon_offset);
let local_name = &local_name[1..]; // Remove the colon
if prefix.is_empty() || local_name.is_empty() {
// Neither prefix nor local name can be empty
return Err(Error::InvalidCharacter);
}
Ok((Some(prefix), local_name))
}
/// Validate a namespace and qualified name and extract their parts.
@ -25,139 +99,52 @@ pub(crate) fn validate_and_extract(
namespace: Option<DOMString>,
qualified_name: &str,
) -> Fallible<(Namespace, Option<Prefix>, LocalName)> {
// Step 1.
// Step 1. If namespace is the empty string, then set it to null.
let namespace = namespace_from_domstring(namespace);
// Step 2.
validate_qualified_name(qualified_name)?;
// Step 2. Validate qualifiedName.
// Step 3. Let prefix be null.
// Step 4. Let localName be qualifiedName.
// Step 5. If qualifiedName contains a U+003A (:):
// NOTE: validate_and_extract_qualified_name does all of these things for us, because
// it's easier to do them together
let (prefix, local_name) = validate_and_extract_qualified_name(qualified_name)?;
debug_assert!(!local_name.contains(':'));
let colon = ':';
// Step 5.
let mut parts = qualified_name.splitn(2, colon);
let (maybe_prefix, local_name) = {
let maybe_prefix = parts.next();
let maybe_local_name = parts.next();
debug_assert!(parts.next().is_none());
if let Some(local_name) = maybe_local_name {
debug_assert!(!maybe_prefix.unwrap().is_empty());
(maybe_prefix, local_name)
} else {
(None, maybe_prefix.unwrap())
}
};
debug_assert!(!local_name.contains(colon));
match (namespace, maybe_prefix) {
match (namespace, prefix) {
(ns!(), Some(_)) => {
// Step 6.
// Step 6. If prefix is non-null and namespace is null, then throw a "NamespaceError" DOMException.
Err(Error::Namespace)
},
(ref ns, Some("xml")) if ns != &ns!(xml) => {
// Step 7.
// Step 7. If prefix is "xml" and namespace is not the XML namespace,
// then throw a "NamespaceError" DOMException.
Err(Error::Namespace)
},
(ref ns, p) if ns != &ns!(xmlns) && (qualified_name == "xmlns" || p == Some("xmlns")) => {
// Step 8.
// Step 8. If either qualifiedName or prefix is "xmlns" and namespace is not the XMLNS namespace,
// then throw a "NamespaceError" DOMException.
Err(Error::Namespace)
},
(ns!(xmlns), p) if qualified_name != "xmlns" && p != Some("xmlns") => {
// Step 9.
// Step 9. If namespace is the XMLNS namespace and neither qualifiedName nor prefix is "xmlns",
// then throw a "NamespaceError" DOMException.
Err(Error::Namespace)
},
(ns, p) => {
// Step 10.
// Step 10. Return namespace, prefix, and localName.
Ok((ns, p.map(Prefix::from), LocalName::from(local_name)))
},
}
}
/// Results of `xml_name_type`.
#[derive(PartialEq)]
#[allow(missing_docs)]
pub(crate) enum XMLName {
QName,
Name,
Invalid,
}
/// Check if an element name is valid. See <http://www.w3.org/TR/xml/#NT-Name>
/// for details.
pub(crate) fn xml_name_type(name: &str) -> XMLName {
fn is_valid_start(c: char) -> bool {
matches!(c, ':' |
'A'..='Z' |
'_' |
'a'..='z' |
'\u{C0}'..='\u{D6}' |
'\u{D8}'..='\u{F6}' |
'\u{F8}'..='\u{2FF}' |
'\u{370}'..='\u{37D}' |
'\u{37F}'..='\u{1FFF}' |
'\u{200C}'..='\u{200D}' |
'\u{2070}'..='\u{218F}' |
'\u{2C00}'..='\u{2FEF}' |
'\u{3001}'..='\u{D7FF}' |
'\u{F900}'..='\u{FDCF}' |
'\u{FDF0}'..='\u{FFFD}' |
'\u{10000}'..='\u{EFFFF}')
}
fn is_valid_continuation(c: char) -> bool {
is_valid_start(c) ||
matches!(c,
'-' |
'.' |
'0'..='9' |
'\u{B7}' |
'\u{300}'..='\u{36F}' |
'\u{203F}'..='\u{2040}')
}
pub(crate) fn matches_name_production(name: &str) -> bool {
let mut iter = name.chars();
let mut non_qname_colons = false;
let mut seen_colon = false;
let mut last = match iter.next() {
None => return XMLName::Invalid,
Some(c) => {
if !is_valid_start(c) {
return XMLName::Invalid;
}
if c == ':' {
non_qname_colons = true;
}
c
},
};
for c in iter {
if !is_valid_continuation(c) {
return XMLName::Invalid;
}
if c == ':' {
if seen_colon {
non_qname_colons = true;
} else {
seen_colon = true;
}
}
last = c
}
if last == ':' {
non_qname_colons = true
}
if non_qname_colons {
XMLName::Name
} else {
XMLName::QName
if iter.next().is_none_or(|c| !is_valid_start(c)) {
return false;
}
iter.all(is_valid_continuation)
}
/// Convert a possibly-null URL to a namespace.