nigig-org/nimanyatta/copy_code.rs

use std::fs::{self, File};
use std::io::{self, Read, Write};
use std::path::{Path, PathBuf};
use std::process::Command;

/// Directories to always skip, regardless of .gitignore
const IGNORED_DIRS: &[&str] = &[
    "target",
    "node_modules",
    ".git",
    "dist",
    "build",
    ".next",
    ".nuxt",
    ".cache",
    ".turbo",
    "out",
    "pkg",
    "vendor",
    "coverage",
    "__pycache__",
];

/// Files to always skip
const IGNORED_FILES: &[&str] = &[
    "package-lock.json",
    "yarn.lock",
    "pnpm-lock.yaml",
    "Cargo.lock",
    "concatenated_code.txt",
];

fn is_ignored_dir(path: &Path) -> bool {
    path.file_name()
        .and_then(|name| name.to_str())
        .map(|name| IGNORED_DIRS.contains(&name))
        .unwrap_or(false)
}

fn is_ignored_file(path: &Path) -> bool {
    path.file_name()
        .and_then(|name| name.to_str())
        .map(|name| IGNORED_FILES.contains(&name))
        .unwrap_or(false)
}

fn is_hidden(path: &Path) -> bool {
    path.file_name()
        .and_then(|name| name.to_str())
        .map(|name| name.starts_with('.'))
        .unwrap_or(false)
}

fn is_git_ignored(root: &Path, path: &Path) -> bool {
    let relative_path = path.strip_prefix(root).unwrap_or(path);
    let relative_path_str = match relative_path.to_str() {
        Some(s) => s,
        None => return false,
    };

    let output = Command::new("git")
        .args(["check-ignore", "-q", relative_path_str])
        .current_dir(root)
        .output();

    match output {
        Ok(output) => output.status.success(),
        Err(_) => false, // git not available — don't skip
    }
}

fn should_skip(root: &Path, path: &Path) -> bool {
    if is_hidden(path) {
        println!("  skip (hidden):       {}", path.display());
        return true;
    }

    if path.is_dir() && is_ignored_dir(path) {
        println!("  skip (ignored dir):  {}", path.display());
        return true;
    }

    if path.is_file() && is_ignored_file(path) {
        println!("  skip (ignored file): {}", path.display());
        return true;
    }

    if is_git_ignored(root, path) {
        println!("  skip (gitignored):   {}", path.display());
        return true;
    }

    false
}

const VALID_EXTENSIONS: &[&str] = &[
    "rs", "toml", "mjs", "md", "txt", "js", "ts", "tsx", "html", "css", "json", "yaml", "yml",
    "env", "sh", "sql", "graphql", "proto", "dockerfile",
];

fn has_valid_extension(path: &Path) -> bool {
    // Also allow files with no extension but known names
    let known_extensionless = &["Makefile", "Dockerfile", "Justfile", ".env", ".env.example"];

    if let Some(name) = path.file_name().and_then(|n| n.to_str()) {
        if known_extensionless.contains(&name) {
            return true;
        }
    }

    path.extension()
        .and_then(|ext| ext.to_str())
        .map(|ext| VALID_EXTENSIONS.contains(&ext.to_lowercase().as_str()))
        .unwrap_or(false)
}

fn copy_files_to_single_file(
    source_dir: &Path,
    output_file: &mut File,
    root: &Path,
    output_file_abs_path: &Path,
) -> Result<(), io::Error> {
    // Read and sort entries for deterministic output
    let mut entries: Vec<_> = fs::read_dir(source_dir)?
        .filter_map(|e| e.ok())
        .collect();
    entries.sort_by_key(|e| e.path());

    for entry in entries {
        let path = entry.path();

        if should_skip(root, &path) {
            continue;
        }

        if path.is_dir() {
            copy_files_to_single_file(&path, output_file, root, output_file_abs_path)?;
        } else if path.is_file() {
            // Skip the output file itself
            let file_abs_path = fs::canonicalize(&path)?;
            if file_abs_path == output_file_abs_path {
                println!("  skip (output file):  {}", path.display());
                continue;
            }

            if !has_valid_extension(&path) {
                println!("  skip (extension):    {}", path.display());
                continue;
            }

            let relative_path = path.strip_prefix(root).unwrap_or(&path);
            writeln!(output_file, "// File: {}\n", relative_path.display())?;

            let mut file = File::open(&path)?;
            let mut contents = String::new();
            file.read_to_string(&mut contents)?;
            writeln!(output_file, "{}\n", contents)?;

            println!("  ✓ included:          {}", relative_path.display());
        }
    }
    Ok(())
}

fn main() -> Result<(), io::Error> {
   // let project_root = PathBuf::from("./src/gateway/io_backend");
    // let project_root = PathBuf::from("./src/gateway/optimized");
    // let project_root = PathBuf::from("./src/gateway/inbox");
    // let project_root = PathBuf::from("./src/gateway/inbox");
    // let project_root = PathBuf::from("./crates/kra-etims-sdk");
    // let project_root = PathBuf::from("./src/observability");
    // let project_root = PathBuf::from("./src/monitoring");
    let project_root = PathBuf::from("./src/routes");
    // let project_root = PathBuf::from("./src/proxy");
    // let project_root = PathBuf::from("./src");
    let output_file_path = project_root.join("concatenated_code.txt");
    // let output_file_abs_path = fs::canonicalize(&output_file_path).unwrap_or_else(|_| {
    //     std::env::current_dir()
    //         .unwrap()
    //         .join("concatenated_code.txt")
    // });/Users/aok/Projects/rustdev/CratesCode/nimanyatta/src/gateway/optimized

    // Ensure the output file exists so canonicalize works
    if !output_file_path.exists() {
        File::create(&output_file_path)?;
    }
    let output_file_abs_path = fs::canonicalize(&output_file_path)?;

    let mut output_file = File::create(&output_file_path)?;

    println!("╔══════════════════════════════════════════╗");
    println!("║  Code Concatenation Tool                 ║");
    println!("╚══════════════════════════════════════════╝");
    println!();
    println!("Root:   {}", project_root.display());
    println!("Output: {}", output_file_abs_path.display());
    println!();
    println!("Ignored dirs:  {}", IGNORED_DIRS.join(", "));
    println!("Ignored files: {}", IGNORED_FILES.join(", "));
    println!("Valid exts:    {}", VALID_EXTENSIONS.join(", "));
    println!();

    copy_files_to_single_file(
        &project_root,
        &mut output_file,
        &project_root,
        &output_file_abs_path,
    )?;

    // Print summary
    let output_size = fs::metadata(&output_file_path)?.len();
    println!();
    println!("════════════════════════════════════════════");
    println!(
        "Done. Output: {} ({} bytes)",
        output_file_path.display(),
        output_size
    );

    Ok(())
}