Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 3 additions & 3 deletions src/domains.rs
Original file line number Diff line number Diff line change
Expand Up @@ -192,12 +192,12 @@ pub(crate) fn find_authority_end(
}
}

return (end, last_dot);
(end, last_dot)
} else {
return (None, None);
(None, None)
}
} else {
return (end, last_dot);
(end, last_dot)
}
}

Expand Down
16 changes: 9 additions & 7 deletions src/finder.rs
Original file line number Diff line number Diff line change
Expand Up @@ -102,12 +102,14 @@ pub struct LinkFinder {
url_can_be_iri: bool,
}

type TriggerFinder = dyn Fn(&[u8]) -> Option<usize>;

/// Iterator for finding links.
pub struct Links<'t> {
text: &'t str,
rewind: usize,

trigger_finder: Box<dyn Fn(&[u8]) -> Option<usize>>,
trigger_finder: Box<TriggerFinder>,
email_scanner: EmailScanner,
url_scanner: UrlScanner,
domain_scanner: DomainScanner,
Expand Down Expand Up @@ -232,7 +234,7 @@ impl<'t> Links<'t> {
};

// With optional schemes URLs don't have unique `:`, then search for `.` as well
let trigger_finder: Box<dyn Fn(&[u8]) -> Option<usize>> = match (url, email) {
let trigger_finder: Box<TriggerFinder> = match (url, email) {
(true, true) if url_must_have_scheme => Box::new(|s| memchr2(b':', b'@', s)),
(true, true) => Box::new(|s| memchr3(b':', b'@', b'.', s)),
(true, false) if url_must_have_scheme => Box::new(|s| memchr(b':', s)),
Expand Down Expand Up @@ -271,7 +273,7 @@ impl<'t> Iterator for Links<'t> {
let end = self.rewind + range.end;
self.rewind = end;
let link = Link {
text: &self.text,
text: self.text,
start,
end,
kind,
Expand Down Expand Up @@ -299,10 +301,10 @@ impl<'t> Iterator for Spans<'t> {

fn next(&mut self) -> Option<Span<'t>> {
match self.links.peek() {
Some(ref link) => {
Some(link) => {
if self.position < link.start {
let span = Span {
text: &self.text,
text: self.text,
start: self.position,
end: link.start,
kind: None,
Expand All @@ -314,7 +316,7 @@ impl<'t> Iterator for Spans<'t> {
None => {
if self.position < self.text.len() {
let span = Span {
text: &self.text,
text: self.text,
start: self.position,
end: self.text.len(),
kind: None,
Expand All @@ -327,7 +329,7 @@ impl<'t> Iterator for Spans<'t> {
self.links.next().map(|link| {
self.position = link.end;
Span {
text: &self.text,
text: self.text,
start: link.start,
end: link.end,
kind: Some(link.kind),
Expand Down
9 changes: 3 additions & 6 deletions src/url.rs
Original file line number Diff line number Diff line change
Expand Up @@ -9,7 +9,7 @@ use crate::scanner::Scanner;
/// The shortest valid URL (without a scheme) might be g.cn (Google China),
/// which consists of four characters.
/// We set this as a lower threshold for parsing URLs from plaintext
/// to avoid false-positives and as a slight performance optimization.
/// to avoid false positives and as a slight performance optimization.
/// This threshold might be adjusted in the future.
const MIN_URL_LENGTH: usize = 4;

Expand Down Expand Up @@ -146,10 +146,7 @@ fn find_scheme_start(s: &str) -> (Option<usize>, Option<char>) {
///
/// We could make this configurable, but let's keep it simple until someone asks (hi!).
fn scheme_requires_host(scheme: &str) -> bool {
match scheme {
"https" | "http" | "ftp" | "ssh" => true,
_ => false,
}
matches!(scheme, "https" | "http" | "ftp" | "ssh")
}

/// Find the start of a plain domain URL (no scheme), e.g. from `blog.`, start at `g` and end at `b`.
Expand All @@ -175,7 +172,7 @@ fn find_domain_start(s: &str, iri_parsing_enabled: bool) -> (Option<usize>, Opti
// If this was a valid domain, we'd have extracted it already from the previous "."
'.' => return (None, None),
'-' => {
if first == None {
if first.is_none() {
// Domain label can't end with `-`
return (None, None);
} else {
Expand Down