//! Shared code-file classification for extraction and pre-ranking scope filters.
use std::{path::Path, sync::OnceLock};

pub const CODE_EXTENSIONS: &[&str] = &[
    "rs", "py", "pyi", "js", "jsx", "mjs", "cjs", "ts", "tsx", "mts", "cts", "c", "h", "cpp", "cc",
    "cxx", "hpp", "hxx", "hh", "go", "java", "kt", "kts", "cs", "swift", "m", "mm", "rb", "php",
    "phtml", "scala", "sc", "sql", "sh", "bash", "zsh", "fish", "ps1", "psm1", "psd1", "bat",
    "cmd", "lua", "r", "pl", "pm", "ex", "exs", "erl", "hrl", "hs", "lhs", "clj", "cljs", "cljc",
    "dart", "vue", "svelte", "graphql", "gql", "tf", "tfvars", "proto", "zig", "vb", "vbs", "f",
    "for", "f90", "f95", "f03", "pas", "d", "cmake",
];
const CODE_FILENAMES: &[&str] = &[
    "makefile",
    "gnumakefile",
    "dockerfile",
    "cmakelists.txt",
    "justfile",
];

pub fn is_code_path(path: &Path) -> bool {
    let extension = path
        .extension()
        .and_then(|value| value.to_str())
        .unwrap_or("")
        .to_ascii_lowercase();
    let filename = path
        .file_name()
        .and_then(|value| value.to_str())
        .unwrap_or("")
        .to_ascii_lowercase();
    CODE_EXTENSIONS.contains(&extension.as_str()) || CODE_FILENAMES.contains(&filename.as_str())
}

/// Static trusted SQL predicate for the shared internal `files f` alias. No
/// caller-controlled query, path or extension is interpolated into this SQL.
pub fn code_file_predicate() -> &'static str {
    static SQL: OnceLock<String> = OnceLock::new();
    SQL.get_or_init(|| {
        let literals = |values: &[&str]| {
            values
                .iter()
                .map(|value| format!("'{value}'"))
                .collect::<Vec<_>>()
                .join(",")
        };
        format!(
            "(f.extension IN ({}) OR lower(f.filename) IN ({}))",
            literals(CODE_EXTENSIONS),
            literals(CODE_FILENAMES)
        )
    })
}