oratos 0.3.1

Website visibility intelligence for SEO, accessibility, structured metadata, and AI readiness.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153

use crate::html::{is_site_root_path, HtmlPage};

/// Generate a conservative `llms.txt` from discovered pages.
pub fn generate_llms_txt(pages: &[HtmlPage], site_title: Option<&str>) -> String {
    let mut out = String::new();

    let title = site_title
        .or_else(|| pages.first().and_then(|p| p.title.as_deref()))
        .unwrap_or("Website");

    out.push_str(&format!("# {title}\n\n"));
    out.push_str(
        "> Generated by Oratos. llms.txt is an emerging convention — supplement, do not replace, \
         good HTML, metadata, sitemaps, and structured data.\n\n",
    );

    let summary = pages
        .iter()
        .find_map(|p| p.meta_description.clone())
        .or_else(|| pages.first().map(|p| truncate(&p.main_text, 200)));

    if let Some(s) = summary {
        out.push_str("## Summary\n\n");
        out.push_str(&s);
        out.push_str("\n\n");
    }

    out.push_str("## Important pages\n\n");
    for page in important_pages(pages) {
        let desc = page_description(page);
        out.push_str(&format!(
            "- [{}]({})",
            page.display_name(),
            page.url_or_path
        ));
        if !desc.is_empty() {
            out.push_str(&format!(" — {desc}"));
        }
        out.push('\n');
    }

    out.push_str("\n## Page summaries\n\n");
    for page in pages {
        let summary = page_description(page);
        if summary.is_empty() {
            continue;
        }
        out.push_str(&format!("### {}\n\n", page.display_name()));
        out.push_str(&summary);
        out.push_str("\n\n");
    }

    out
}

fn important_pages(pages: &[HtmlPage]) -> Vec<&HtmlPage> {
    let mut important: Vec<&HtmlPage> = pages
        .iter()
        .filter(|p| is_site_root_path(&p.url_or_path) || p.headings.iter().any(|h| h.level == 1))
        .collect();
    if important.len() < 10 {
        for page in pages {
            if important.iter().any(|p| p.url_or_path == page.url_or_path) {
                continue;
            }
            important.push(page);
            if important.len() == 10 {
                break;
            }
        }
    }
    if important.is_empty() {
        pages.iter().take(10).collect()
    } else if important.len() > 10 {
        important.into_iter().take(10).collect()
    } else {
        important
    }
}

fn page_description(page: &HtmlPage) -> String {
    if let Some(desc) = &page.meta_description {
        return desc.clone();
    }
    if let Some(h1) = page.headings.iter().find(|h| h.level == 1) {
        return h1.text.clone();
    }
    truncate(&page.main_text, 160)
}

fn truncate(s: &str, max: usize) -> String {
    if s.len() <= max {
        s.to_string()
    } else {
        let end = char_boundary_before(s, max.saturating_sub(1));
        format!("{}…", &s[..end])
    }
}

fn char_boundary_before(s: &str, max: usize) -> usize {
    if max >= s.len() {
        return s.len();
    }
    let mut boundary = 0usize;
    for (idx, _) in s.char_indices() {
        if idx > max {
            break;
        }
        boundary = idx;
    }
    boundary
}

#[cfg(test)]
mod tests {
    use super::*;
    use crate::html::parse_html;

    #[test]
    fn generates_llms_txt_header() {
        let page = parse_html("/index.html", "<html><head><title>My Site</title></head><body><h1>Home</h1><p>Welcome to my site with enough text for extraction purposes here.</p></body></html>", true);
        let txt = generate_llms_txt(&[page], None);
        assert!(txt.contains("# My Site"));
        assert!(txt.contains("Important pages"));
    }

    #[test]
    fn truncate_handles_utf8_boundaries() {
        let s = "Good Example Site — Home with extra text";
        let truncated = truncate(s, 20);
        assert!(truncated.ends_with('…'));
    }

    #[test]
    fn important_pages_fills_to_ten_when_h1_subset_is_small() {
        let mut pages = Vec::new();
        pages.push(parse_html(
            "/one.html",
            "<html><head><title>One</title></head><body><h1>One</h1></body></html>",
            true,
        ));
        for i in 2..=12 {
            let path = format!("/{i}.html");
            pages.push(parse_html(
                &path,
                "<html><head><title>Page</title></head><body><p>body</p></body></html>",
                true,
            ));
        }
        let selected = important_pages(&pages);
        assert_eq!(selected.len(), 10);
    }
}