thesa 4.1.34

use std::path::PathBuf;

use clap::{Args, Parser, Subcommand};
use siteforge::ExportFormat as SiteforgeExportFormat;

use crate::types::{DatasetProvider, Mode, ModelProvider, SiteArchiveProfile};

#[derive(Parser, Clone)]
#[command(
    name = "thesa",
    version,
    about = "A compendium to archive GitHub repositories, ML models, datasets, and websites",
    subcommand_precedence_over_arg = true
)]
pub(crate) struct Cli {
    #[command(subcommand)]
    pub(crate) command: Option<Commands>,

    /// Target: GitHub owner/repo (repos), model target, dataset target, or website URL
    pub(crate) target: Option<String>,

    /// Output directory where archives are stored
    #[arg(short = 'o', long = "output", default_value = "./archives")]
    pub(crate) output: PathBuf,

    /// GitHub token for API requests (optional, falls back to GITHUB_TOKEN env)
    #[arg(short = 't', long = "token")]
    pub(crate) token: Option<String>,

    /// Hugging Face token for API requests (optional, falls back to HF_TOKEN env)
    #[arg(long = "hf-token")]
    pub(crate) hf_token: Option<String>,

    /// CivitAI token for API requests (optional, falls back to CIVITAI_TOKEN env)
    #[arg(long = "civitai-token")]
    pub(crate) civitai_token: Option<String>,

    /// Clone depth for each repository
    #[arg(long = "depth")]
    pub(crate) depth: Option<u32>,

    /// Number of items to process at once
    #[arg(short = 'c', long = "concurrency", default_value_t = 4)]
    pub(crate) concurrency: usize,

    /// Skip existing directories instead of failing or replacing
    #[arg(long = "skip-existing")]
    pub(crate) skip_existing: bool,

    /// Filter items by substring in name
    #[arg(long = "filter")]
    pub(crate) filter: Option<String>,

    /// Print plan only; do not execute
    #[arg(long = "dry-run")]
    pub(crate) dry_run: bool,

    /// Force the interactive Scrin interface
    #[arg(short = 'T', long = "tui")]
    pub(crate) tui: bool,

    /// Archive mode: repos (GitHub), models, datasets, or sites (Siteforge)
    #[arg(
        long = "mode",
        value_enum,
        default_value_t = Mode::Repos,
        help = "Archive mode: repos (GitHub), models, datasets, or sites (Siteforge)"
    )]
    pub(crate) mode: Mode,

    /// Model provider: hf (Hugging Face), ollama, civitai
    #[arg(long = "provider", value_enum, default_value_t = ModelProvider::Hf)]
    pub(crate) provider: ModelProvider,

    /// Dataset provider: hf (Hugging Face Datasets), zenodo
    #[arg(long = "dataset-provider", value_enum, default_value_t = DatasetProvider::Hf)]
    pub(crate) dataset_provider: DatasetProvider,

    /// Use Hugging Face mirror (hf-mirror.com) — no token required
    #[arg(long = "hf-mirror")]
    pub(crate) hf_mirror: bool,
}

#[derive(Debug, Clone, Subcommand)]
#[allow(clippy::large_enum_variant)]
pub(crate) enum Commands {
    /// Archive artifacts from a supported source provider
    Archive {
        /// Print plan only; do not execute
        #[arg(long = "dry-run")]
        dry_run: bool,

        #[command(subcommand)]
        source: ArchiveCommands,
    },

    /// Verify a thesa archive directory using .thesa/manifest.json and checksums
    Verify {
        /// Archive output directory containing a .thesa manifest sidecar
        archive: PathBuf,

        /// Emit machine-readable JSON report
        #[arg(long)]
        json: bool,
    },
}

#[derive(Debug, Clone, Subcommand)]
pub(crate) enum ArchiveCommands {
    /// Refresh existing GitHub repository archives with git pull --ff-only
    #[command(alias = "update", alias = "pull")]
    Refresh(GitRefreshArgs),

    /// Archive a website with siteforge
    Site(SiteArchiveArgs),

    /// Archive datasets with baseforge
    Dataset(DatasetArchiveArgs),

    /// Archive a generic URL/website with siteforge
    Url(SiteArchiveArgs),

    /// List Siteforge website archives
    List(SiteforgeRootArgs),

    /// Inspect one Siteforge archive
    Inspect(SiteforgeArchiveIdArgs),

    /// Search Siteforge archives
    Search(SiteforgeSearchArgs),

    /// Export a Siteforge archive
    Export(SiteforgeExportArgs),

    /// Create AI context packs for a Siteforge archive
    Pack(SiteforgePackArgs),

    /// Verify a Siteforge archive by archive id
    Verify(SiteforgeArchiveIdArgs),

    /// Resume an interrupted Siteforge archive by archive id
    Resume(SiteforgeArchiveIdArgs),

    /// Print Siteforge diagnostics
    Diagnostics(SiteforgeDiagnosticsArgs),
}

#[derive(Debug, Clone, Args)]
pub(crate) struct GitRefreshArgs {
    /// Optional GitHub owner/user/org to refresh from the archive tree
    #[arg(value_name = "OWNER")]
    pub(crate) owner: Option<String>,

    /// Root to scan for GitHub repository archives
    #[arg(long = "archive-root", short = 'r', default_value = "./archives")]
    pub(crate) archive_root: PathBuf,

    /// GitHub owner/user/org to refresh from the archive tree
    #[arg(
        long = "owner",
        alias = "user",
        alias = "username",
        value_name = "OWNER"
    )]
    pub(crate) owner_filter: Option<String>,

    /// Number of repositories to refresh at once
    #[arg(short = 'c', long = "concurrency", default_value_t = 4)]
    pub(crate) concurrency: usize,

    /// Only refresh archives whose owner/repo slug or path contains this text
    #[arg(long = "filter")]
    pub(crate) filter: Option<String>,

    /// Print discovered archives only; do not run git pull
    #[arg(long = "dry-run")]
    pub(crate) dry_run: bool,
}

#[derive(Debug, Clone, Args)]
pub(crate) struct SiteforgeRootArgs {
    /// Siteforge archive root
    #[arg(long = "archive-root", default_value = "./archives/sites")]
    pub(crate) archive_root: PathBuf,
}

#[derive(Debug, Clone, Args)]
pub(crate) struct DatasetArchiveArgs {
    /// Dataset target: namespace, dataset id, `search:<query>`, top, latest, or provider URL
    pub(crate) target: String,

    /// Dataset provider
    #[arg(long = "provider", value_enum, default_value_t = DatasetProvider::Hf)]
    pub(crate) provider: DatasetProvider,

    /// Dataset archive root
    #[arg(
        long = "archive-root",
        short = 'r',
        default_value = "./archives/datasets"
    )]
    pub(crate) archive_root: PathBuf,

    /// Number of dataset files to download at once
    #[arg(short = 'c', long = "concurrency", default_value_t = 4)]
    pub(crate) concurrency: usize,

    /// Skip existing dataset directories
    #[arg(long = "skip-existing")]
    pub(crate) skip_existing: bool,

    /// Filter discovered datasets by substring
    #[arg(long = "filter")]
    pub(crate) filter: Option<String>,

    /// Hugging Face token for private/gated datasets (falls back to HF_TOKEN env)
    #[arg(long = "hf-token")]
    pub(crate) hf_token: Option<String>,

    /// Print plan only; do not download
    #[arg(long = "dry-run")]
    pub(crate) dry_run: bool,
}

#[derive(Debug, Clone, Args)]
pub(crate) struct SiteforgeArchiveIdArgs {
    /// Siteforge archive id
    pub(crate) archive_id: String,

    #[command(flatten)]
    pub(crate) root: SiteforgeRootArgs,
}

#[derive(Debug, Clone, Args)]
pub(crate) struct SiteforgeSearchArgs {
    /// Query text
    pub(crate) query: String,

    /// Restrict search to one archive id
    #[arg(long = "archive-id")]
    pub(crate) archive_id: Option<String>,

    /// Maximum matches to print
    #[arg(long = "limit", default_value_t = 20)]
    pub(crate) limit: usize,

    #[command(flatten)]
    pub(crate) root: SiteforgeRootArgs,
}

#[derive(Debug, Clone, Args)]
pub(crate) struct SiteforgeExportArgs {
    /// Siteforge archive id
    pub(crate) archive_id: String,

    /// Export format
    #[arg(long = "format", value_enum)]
    pub(crate) format: SiteforgeExportFormat,

    /// Export output path
    #[arg(long = "output")]
    pub(crate) output: Option<PathBuf>,

    #[command(flatten)]
    pub(crate) root: SiteforgeRootArgs,
}

#[derive(Debug, Clone, Args)]
pub(crate) struct SiteforgePackArgs {
    /// Siteforge archive id
    pub(crate) archive_id: String,

    /// Maximum tokens per generated pack
    #[arg(long = "max-tokens")]
    pub(crate) max_tokens: Option<usize>,

    #[command(flatten)]
    pub(crate) root: SiteforgeRootArgs,
}

#[derive(Debug, Clone, Args)]
pub(crate) struct SiteforgeDiagnosticsArgs {
    /// Optional Siteforge archive id
    pub(crate) archive_id: Option<String>,

    #[command(flatten)]
    pub(crate) root: SiteforgeRootArgs,
}

#[derive(Debug, Clone, Args)]
pub(crate) struct SiteArchiveArgs {
    /// Seed URL to archive. Optional when --input provides seeds.
    #[arg(value_name = "URL")]
    pub(crate) url: Option<String>,

    /// File containing seed URLs, one per line. Blank lines and # comments are ignored.
    #[arg(long = "input", value_name = "FILE")]
    pub(crate) input: Option<PathBuf>,

    /// Output root for website archives
    #[arg(short = 'o', long = "output", default_value = "./archives/sites")]
    pub(crate) output: PathBuf,

    /// Archive all in-scope pages under the site instead of one page
    #[arg(long = "full-site")]
    pub(crate) full_site: bool,

    /// Allow following links outside the seed domain
    #[arg(long = "allow-cross-domain")]
    pub(crate) allow_cross_domain: bool,

    /// Maximum crawl depth
    #[arg(long = "max-depth")]
    pub(crate) max_depth: Option<usize>,

    /// Maximum pages to archive
    #[arg(long = "max-pages")]
    pub(crate) max_pages: Option<usize>,

    /// Number of concurrent fetch workers
    #[arg(short = 'c', short_alias = 'j', long = "concurrency", alias = "jobs")]
    pub(crate) concurrency: Option<usize>,

    /// Politeness delay between requests per worker, in milliseconds
    #[arg(long = "delay-ms")]
    pub(crate) delay_ms: Option<u64>,

    /// Stable archive id for the siteforge archive
    #[arg(long = "archive-id")]
    pub(crate) archive_id: Option<String>,

    /// Include URL glob pattern. May be repeated.
    #[arg(long = "include", value_name = "GLOB")]
    pub(crate) include_url_patterns: Vec<String>,

    /// Exclude URL glob pattern. May be repeated.
    #[arg(long = "exclude", value_name = "GLOB")]
    pub(crate) exclude_url_patterns: Vec<String>,

    /// Siteforge archive output profile: full, agent, or lean
    #[arg(long = "archive-profile", value_enum)]
    pub(crate) archive_profile: Option<SiteArchiveProfile>,

    /// Maximum bytes to fetch per HTML page
    #[arg(long = "max-page-size-bytes")]
    pub(crate) max_page_size_bytes: Option<u64>,

    /// Maximum bytes to fetch per asset
    #[arg(long = "max-asset-size-bytes")]
    pub(crate) max_asset_size_bytes: Option<u64>,

    /// Maximum total archive size in bytes
    #[arg(long = "max-total-archive-size-bytes")]
    pub(crate) max_total_archive_size_bytes: Option<u64>,

    /// Render JavaScript pages through siteforge's browser renderer
    #[arg(long = "render-js")]
    pub(crate) render_js: bool,

    /// Render only pages that look like thin JavaScript app shells
    #[arg(long = "render-js-auto", conflicts_with = "render_js")]
    pub(crate) render_js_auto: bool,

    /// Browser command template for Siteforge JavaScript rendering
    #[arg(long = "browser-command", value_name = "COMMAND")]
    pub(crate) browser_command: Option<String>,

    /// Browser profile directory for Siteforge JavaScript rendering
    #[arg(long = "browser-profile", value_name = "DIR")]
    pub(crate) browser_profile: Option<PathBuf>,

    /// Milliseconds to wait after browser render before extracting DOM
    #[arg(long = "render-wait-ms")]
    pub(crate) render_wait_ms: Option<u64>,

    /// Extra request header. Repeatable. Format: "Name: Value"
    #[arg(long = "header", value_name = "NAME: VALUE")]
    pub(crate) headers: Vec<String>,

    /// Cookie header value to send with requests
    #[arg(long = "cookie", value_name = "COOKIE_HEADER")]
    pub(crate) cookie: Option<String>,

    /// User-Agent header used by Siteforge fetchers/renderers
    #[arg(long = "user-agent", value_name = "VALUE")]
    pub(crate) user_agent: Option<String>,

    /// HTTP timeout per request, in seconds
    #[arg(long = "timeout-secs")]
    pub(crate) timeout_secs: Option<u64>,

    /// Number of Siteforge retry attempts for failed frontier items
    #[arg(long = "retry-count")]
    pub(crate) retry_count: Option<usize>,

    /// Enable Siteforge OCR for code-like downloaded assets
    #[arg(long = "ocr")]
    pub(crate) ocr_enabled: bool,

    /// Print plan only; do not execute
    #[arg(long = "dry-run")]
    pub(crate) dry_run: bool,
}