meta.anchors.content.literal_syntax

Boundary-looking text inside fenced or indented code blocks, inline code, or escaped text does not open or close an anchor. Raw HTML examples do not supply anchor definitions, section headings, or end markers. A standalone end comment acts as a boundary only outside these literal contexts and enclosing HTML blocks.

Read in specification context · design/meta/anchors.md

Implementation

Implemented crates/sage-meta/src/spec.rs:322–327

sage_meta::spec · lib:sage_meta

#![allow(unused)]
fn main() {
            // {anchor}`meta.anchors.content.literal_syntax`
            Event::Text(_) if !in_code_block && !in_html_block && html_tags.is_empty() => {
                let text = &source.source[range.clone()];
                for (relative, _) in text.match_indices("{anchor}") {
                    let start = range.start + relative;
                    if is_escaped(&source.source, start) {

}

See more source ↗

Tests and review

Indexed tests and the implementation-completeness and test-adequacy review; no test run is implied.

Implemented sage_meta::prose_rendering_hides_control_comments_but_preserves_literal_examples

crates/sage-meta/tests/browser.rs · test:browser

#![allow(unused)]
fn main() {
#[test]
#[sage::test(meta.anchors.content.explicit_end)]
#[sage::test(meta.anchors.content.literal_syntax)]
fn prose_rendering_hides_control_comments_but_preserves_literal_examples() {
    use sage_meta::spec::{SpecChapter, SpecIndex, SpecSource};
    let source = "{anchor}`demo.rule` A rule.\n\nMore of the rule.\n\n<!-- end anchor -->\n\nOutside commentary.\n\n```html\n<!-- end anchor -->\n```\n\nInline `<!-- end anchor -->`.\n\n<div>\n\n<!-- end anchor -->\n\n</div>\n";
    let mut spec = SpecIndex::from_sources(&[SpecSource {
        file: "design/demo.md".into(),
        source: source.into(),
    }]);
    assert!(spec.diagnostics.is_empty(), "{:?}", spec.diagnostics);
    spec.chapters.insert(
        "design/demo.md".into(),
        SpecChapter {
            file: "design/demo.md".into(),
            namespace: String::new(),
            url: "design/demo.html".into(),
            normative: true,
        },
    );
    let html =
        sage_meta::render_spec_markdown(&spec, "design/demo.md", source, "sage-explorer.html")
            .unwrap();
    assert!(html.contains("More of the rule."));
    assert!(html.contains("Outside commentary."));
    assert!(!html.contains("<!-- end anchor -->"));
    assert_eq!(html.matches("&lt;!-- end anchor --&gt;").count(), 3);
    assert!(html.contains("<code>&lt;!-- end anchor --&gt;</code>"));
    assert!(html.contains("<pre><code class=\"language-html\">&lt;!-- end anchor --&gt;"));
    assert!(html.contains("&lt;div&gt;"));
}
}

See more source ↗

Implemented sage_meta::boundary_examples_in_code_escapes_and_enclosing_html_are_literal

crates/sage-meta/tests/spec.rs · test:spec

#![allow(unused)]
fn main() {
#[test]
#[sage::test(meta.anchors.content.literal_syntax)]
fn boundary_examples_in_code_escapes_and_enclosing_html_are_literal() {
    let body = r#"Opening.

```markdown
{anchor}`example.fenced` Example.
# Heading
<!-- end anchor -->
```

    {anchor}`example.indented` Example.
    Heading
    -------
    <!-- end anchor -->

`` {anchor}`example.inline` <!-- end anchor --> # Heading ``

\{anchor}`example.escaped` Example.

\# Heading

\<!-- end anchor -->

<div><span title="a > b and </span>">

# HTML heading

{anchor}`example.html` Example.

<!-- end anchor -->

</span></div>

<div
    class="example">

HTML heading
============

{anchor}`example.multiline_html` Example.

<!-- end anchor -->

</div>

Inline <span>

# Inline HTML heading

{anchor}`example.inline_html` Example.

<!-- end anchor -->

</span>

<!-- <div> does not open an enclosing element. -->

The actual final qualification."#;
    let text = format!(
        "{{anchor}}`meta.rule` {body}\n\n<!-- end anchor -->\n\nOutside.\n\n{{anchor}}`meta.fresh` Fresh."
    );
    let index = SpecIndex::from_sources(&[source("literals.md", &text)]);
    assert!(index.diagnostics.is_empty(), "{:?}", index.diagnostics);
    assert_eq!(index.anchors.len(), 2);
    assert_eq!(index.anchors["meta.rule"].content, body);
    assert_eq!(index.anchors["meta.fresh"].content, "Fresh.");
    let examples = SpecIndex::from_sources(&[source("examples.md", body)]);
    assert!(
        examples.diagnostics.is_empty(),
        "{:?}",
        examples.diagnostics
    );
    assert!(examples.anchors.is_empty());
}
}

See more source ↗

Implemented sage_meta::malformed_looking_comments_are_content_and_do_not_end_claims

crates/sage-meta/tests/spec.rs · test:spec

#![allow(unused)]
fn main() {
#[test]
#[sage::test(meta.anchors.content.explicit_end)]
#[sage::test(meta.anchors.content.literal_syntax)]
fn malformed_looking_comments_are_content_and_do_not_end_claims() {
    let body = "Opening.\n\n<!--end anchor-->\n\n<!-- end anchor --> trailing text\n\n<!-- end anchor --> <!-- another comment -->\n\nBefore <!-- end anchor -->\n\n<!-- end anchor\n-->\n\n<!-- outer comment\n<!-- end anchor -->\n\nAfter all examples.";
    let text = format!("{{anchor}}`meta.rule` {body}\n\n<!-- end anchor -->\nOutside.");
    let index = SpecIndex::from_sources(&[source("comments.md", &text)]);
    assert!(index.diagnostics.is_empty(), "{:?}", index.diagnostics);
    assert_eq!(index.anchors["meta.rule"].content, body);
}
}

See more source ↗

Implemented sage_meta::raw_html_text_does_not_hide_boundaries_after_its_closing_element

crates/sage-meta/tests/spec.rs · test:spec

#![allow(unused)]
fn main() {
#[test]
#[sage::test(meta.anchors.content.literal_syntax)]
fn raw_html_text_does_not_hide_boundaries_after_its_closing_element() {
    for html in [
        "<script>const example = \"<\";</script>",
        "<script>const example = \"<!--\";</script>",
        "<style>.example::before { content: \"<\"; }</style>",
        "<textarea>some < text</textarea>",
        "<pre>some < text</pre>",
        "<div>some < text</div>",
    ] {
        let body = format!("Claim.\n\n{html}");
        let text = format!(
            "{{anchor}}`meta.before` {body}\n\n<!-- end anchor -->\n\n{{anchor}}`meta.after` Following."
        );
        let index = SpecIndex::from_sources(&[source("html.md", &text)]);
        assert!(
            index.diagnostics.is_empty(),
            "{html}: {:?}",
            index.diagnostics
        );
        assert_eq!(index.anchors["meta.before"].content, body);
        assert_eq!(index.anchors["meta.after"].content, "Following.");
    }
}
}

See more source ↗

Implemented sage_meta::rendering_removes_only_parsed_end_markers

crates/sage-meta/tests/spec.rs · test:spec

#![allow(unused)]
fn main() {
#[test]
#[sage::test(meta.anchors.content.explicit_end)]
#[sage::test(meta.anchors.content.literal_syntax)]
fn rendering_removes_only_parsed_end_markers() {
    let text = "{anchor}`claim` A claim.\n\n```html\n<!-- end anchor -->\n```\n\n<div>\n\n<!-- end anchor -->\n\n</div>\n\n<!--end anchor-->\n\n  <!-- end anchor --> \t\n\nOutside prose.\n";
    let fixture = BookFixture::new("- [Rules](design/rules.md)\n");
    fixture.write("design/rules.md", text);
    let index = fixture.load();
    assert!(index.diagnostics.is_empty(), "{:?}", index.diagnostics);
    for rendered in [
        render_anchors(text).unwrap(),
        index.render_chapter("design/rules.md", text).unwrap(),
    ] {
        assert!(!rendered.contains("  <!-- end anchor --> \t"));
        assert_eq!(rendered.matches("<!-- end anchor -->").count(), 2);
        assert!(rendered.contains("```html\n<!-- end anchor -->\n```"));
        assert!(rendered.contains("<div>\n\n<!-- end anchor -->\n\n</div>"));
        assert!(rendered.contains("<!--end anchor-->"));
        assert!(rendered.ends_with("Outside prose.\n"));
    }
    let replacements = index.role_replacements("design/rules.md", text).unwrap();
    let removals = replacements
        .iter()
        .filter(|(_, value)| value.is_empty())
        .collect::<Vec<_>>();
    assert_eq!(removals.len(), 1);
    let (span, _) = removals[0];
    assert_eq!(&text[span.start..span.end], "  <!-- end anchor --> \t");
}
}

See more source ↗

Review: Accepted · Codex /root/anchor_094.

Fresh final followup review of literal_syntax. Saved/read the prior complete report and inputs, started fresh, recaptured all three dependencies and compared every fragment. The sole difference is an unrelated implementation_references association comment in spec.rs. The claim, all five tests, source/BookFixture helpers and browser rendering dependency are unchanged. No literal-context or rendering behavior changed.

The full claim requires definition/heading/end-looking text to remain inert inside fenced/indented/inline code, escapes and raw/enclosing HTML, with standalone end comments active only outside those contexts. The ordinary parser association processes definition Text events only outside code/HTML states and enclosing tags; inline code is not Text, and odd preceding backslashes suppress escaped definitions. Heading closure requires an empty HTML stack. End-comment recognition requires a complete standalone HTML block outside enclosing tags. finish_anchor retains the original contiguous source, so treating examples literally preserves rather than deletes them. update_anchor_html_tags tracks enclosing elements across blank lines, handles quoted delimiters and multiline tags, skips comment/CDATA/processing-instruction bodies and invalid less-than text, and treats script/style/textarea/pre as raw text until their actual matching close. Stack updates resume real boundaries after closure. The new association comment does not affect this mechanism.

All five ordinary tests were read. boundary_examples_in_code_escapes_and_enclosing_html_are_literal asserts exact complete active content and exactly two real anchors for fenced/indented/inline/escaped examples, nested and multiline HTML, quoted false closing tags and inline-opened containers; parsing examples alone requires no definitions or diagnostics. malformed_looking_comments_are_content_and_do_not_end_claims preserves exact nonstandalone, nonexact, multiline and nested-looking comments until the real end. raw_html_text_does_not_hide_boundaries_after_its_closing_element covers script strings containing less-than/comment openers, style, textarea, pre and div text, requiring preserved body and recognition of the later genuine end/new definition. rendering_removes_only_parsed_end_markers checks direct and chapter rendering, exact removal span/count, preserved code/HTML/nonexact markers and outside prose. prose_rendering_hides_control_comments_but_preserves_literal_examples checks real control removal while retaining exactly three escaped literal examples, inline/fenced code markup and escaped HTML container.

Captured browser rendering uses shared role replacements and removes only recognized empty replacements before displaying remaining raw HTML literally, so it does not introduce an independent marker interpretation. The captured source and BookFixture helpers invoke production parsing/loading. No TODO/N/A entries or deferred clause exist. Every named context and resumption after actual HTML closure remain accounted for by current source and substantive assertions. This is a full claim assessment of the actual marker-only delta, with no other relied-on changes, rather than blind refresh. No source changes or test/build/browser execution occurred; fresh review commands and source inspection only. No requested edits.

Browse meta anchors

116 anchors

Anchor Implementation Tests

meta (116) 102 Implemented24 TODO 103 Implemented36 TODO

meta.anchors (22) 22 Implemented 21 Implemented12 TODO

anchor_pages Implemented ImplementedTODO

Every specification anchor has a stable page identifying the rule and showing its definition, context, implementation entries, and tests/evidence. The page includes Rust excerpts and access to surrounding source, and exposes each entry’s placeholder status and explanation. Multiple sites remain individually inspectable rather than being reduced to a single coverage badge.

Symposiun AGentic Environment (SAGE)

meta.anchors.content.literal_syntax

Implementation

Tests and review