Skip to content

Offline, privacy-first grammar checker. Fast, open-source, Rust-powered

feat(html): condense spaces like most HTML parsers

Elijah Potterb1a76acb9a43
diff --git a/harper-html/src/lib.rs b/harper-html/src/lib.rs
index f8c7bff88..f4fb6d89a 100644
--- a/harper-html/src/lib.rs
+++ b/harper-html/src/lib.rs
@@ -1,9 +1,10 @@
use harper_core::parsers::{self, Parser, PlainEnglish};
-use harper_core::Token;
+use harper_core::{Token, TokenKind};
use harper_tree_sitter::TreeSitterMasker;
use tree_sitter::Node;
pub struct HtmlParser {
+ /// Used to grab the text nodes.
inner: parsers::Mask<TreeSitterMasker, PlainEnglish>,
}
@@ -26,6 +27,14 @@ impl Default for HtmlParser {
impl Parser for HtmlParser {
fn parse(&self, source: &[char]) -> Vec<Token> {
- self.inner.parse(source)
+ let mut tokens = self.inner.parse(source);
+
+ for token in &mut tokens {
+ if let TokenKind::Space(v) = &mut token.kind {
+ *v = (*v).clamp(0, 1);
+ }
+ }
+
+ tokens
}
}
diff --git a/harper-html/tests/run_tests.rs b/harper-html/tests/run_tests.rs
index dcea70978..f1c23b3db 100644
--- a/harper-html/tests/run_tests.rs
+++ b/harper-html/tests/run_tests.rs
@@ -38,3 +38,4 @@ macro_rules! create_test {
create_test!(run_on.html, 0);
create_test!(issue_156.html, 0);
+create_test!(issue_541.html, 0);
diff --git a/harper-html/tests/test_sources/issue_541.html b/harper-html/tests/test_sources/issue_541.html
new file mode 100644
index 000000000..d83fdbbfd
--- /dev/null
+++ b/harper-html/tests/test_sources/issue_541.html
@@ -0,0 +1,15 @@
+<!DOCTYPE html>
+<html lang="en">
+
+<head>
+ <meta charset="utf-8">
+</head>
+
+<body>
+ <p>
+ This block contains multiple lines of HTML. If Harper is throwing a "too many spaces" lint, it's
+ because <code>harper-html</code> isn't properly parsing the indent.
+ </p>
+</body>
+
+</html>