liboxen 0.50.1

//! Constants used throughout the codebase
//!

use crate::core::versions::MinOxenVersion;

/// Rust library version
pub const OXEN_VERSION: &str = env!("CARGO_PKG_VERSION");

/// # Filenames and dirs
/// .oxen is the name of the hidden directory where all our data lives
pub const OXEN_HIDDEN_DIR: &str = ".oxen";
/// Folder name for oxen home within `.cache`, `.config`., etc.
pub const OXEN: &str = "oxen";
/// ~/.config/oxen holds config files
pub const CONFIG_DIR: &str = ".config";
/// .oxenignore is the name of the file that contains the ignore patterns
pub const OXEN_IGNORE_FILE: &str = ".oxenignore";
/// Root path for repositories
pub const ROOT_PATH: &str = "/";
/// Config file for the repository
pub const REPO_CONFIG_FILENAME: &str = "config.toml";
/// HEAD file holds onto where the head commit is (commit_id or branch name)
pub const HEAD_FILE: &str = "HEAD";
/// refs/ is a key,val store of branch names to commit ids
pub const REFS_DIR: &str = "refs";
/// history/ dir is a list of directories named after commit ids
pub const HISTORY_DIR: &str = "history";
/// commits/ is a key-value database of commit ids to commit objects
pub const COMMITS_DIR: &str = "commits";
/// prefix for the cached stats dirs
pub const CACHE_DIR: &str = "cache";
/// prefix for the commit count dirs
pub const COMMIT_COUNT_DIR: &str = "cache/commit_count";
/// name of the schema db
pub const SCHEMAS_DIR: &str = "schemas";
/// schemas node in merkle tree
pub const SCHEMAS_TREE_PREFIX: &str = ".oxen";
// name of dir for locking branches during push
pub const BRANCH_LOCKS_DIR: &str = "locks";
/// prefix for the commit rows
pub const ROWS_DIR: &str = "rows";
/// prefix for the commit entry files
pub const FILES_DIR: &str = "files";
/// prefix for the cached dataframes
pub const DATA_FRAMES_DIR: &str = "data_frames";
/// prefix for the cached dataframes
pub const DATA_FRAME_STATUS_DIR: &str = "data_frame_status";
/// prefix for the commit entry dirs
pub const DIRS_DIR: &str = "dirs";
/// prefix for a commit dir => hash maping
pub const DIR_HASHES_DIR: &str = "dir_hashes";
/// prefix for the commit merkle tree db
pub const TREE_DIR: &str = "tree";
/// prefix for the commit merkle tree node dbs
pub const NODES_DIR: &str = "nodes";
/// prefix for cached compare dfs
pub const COMPARES_DIR: &str = "compares";
/// prefix for the left commit pointer in cached compares
pub const LEFT_COMPARE_COMMIT: &str = "LEFT";
/// prefix for the right commit pointer in cached compares
pub const RIGHT_COMPARE_COMMIT: &str = "RIGHT";
/// prefix for the stats dir
pub const STATS_DIR: &str = "stats";
/// prefix for the staged dirs
pub const STAGED_DIR: &str = "staged";
/// Name of the table in the duckdb db used for remote staging
pub const TABLE_NAME: &str = "df";
/// Oxen's internal row id column in duckdb remote staging tables
pub const OXEN_COLS: [&str; 4] = [OXEN_ID_COL, DIFF_STATUS_COL, OXEN_ROW_ID_COL, DIFF_HASH_COL];
/// Oxen's internal row id column to exclude from dfs
pub const EXCLUDE_OXEN_COLS: [&str; 7] = [
    OXEN_ID_COL,
    DIFF_STATUS_COL,
    OXEN_ROW_ID_COL,
    DIFF_HASH_COL,
    EVAL_STATUS_COL,
    EVAL_ERROR_COL,
    EVAL_DURATION_COL,
];
pub const OXEN_ROW_ID_COL: &str = "_oxen_row_id";
/// Oxen's internal id column in duckdb remote staging tables
pub const OXEN_ID_COL: &str = "_oxen_id";
/// Name of the folder of the cache dir in which dfs are indexed as duckdbs
pub const DUCKDB_CACHE_DIR: &str = "duckdb";
/// Default name for duckdb table used for indexing dataframes
pub const DUCKDB_DF_TABLE_NAME: &str = "df";
/// Max number of rows to query from a dataframe
pub const MAX_QUERYABLE_ROWS: usize = 5_000_000;
/// prefix for the sync status dirs to tell if commits are synced locally
pub const SYNC_STATUS_DIR: &str = "sync_status";
/// Flag for if the repository was cloned in a shallow fashion
pub const SHALLOW_FLAG: &str = "SHALLOW";
/// prefix for the commit indices
pub const INDICES_DIR: &str = "indices";
/// prefix for the schema fields that are indexed
pub const FIELDS_DIR: &str = "fields";
/// versions/ is where all the versions are stored so that we can use to quickly swap between versions of the file
pub const VERSIONS_DIR: &str = "versions";
/// chunks/ is where individual file chunks are stored
pub const CHUNKS_DIR: &str = "chunks";
/// Storage of file node representations in objects dir
pub const OBJECT_FILES_DIR: &str = "files";
/// File name for files stored in versions directory (>0.8.4). (Was commit id <= 0.8.4)
pub const VERSION_FILE_NAME: &str = "data";
/// File name for chunks of files stored in versions directory
pub const VERSION_CHUNK_FILE_NAME: &str = "chunk";
/// Chunks directory for version files
pub const VERSION_CHUNKS_DIR: &str = "chunks";
/// merge/ is where any merge conflicts are stored so that we can get rid of them
pub const MERGE_DIR: &str = "merge";
/// mods/ is where we can stage appends, modifications, deletions to files to be merged later
pub const MODS_DIR: &str = "mods";
/// workspaces/ is where we can make remote changes without having to clone locally
pub const WORKSPACES_DIR: &str = "workspaces";
/// workspace commit id
pub const WORKSPACE_CONFIG: &str = "WORKSPACE_CONFIG";
/// workspace_name_index/ is a RocksDB mapping workspace names to workspace IDs for O(1) lookup
pub const WORKSPACE_NAME_INDEX_DIR: &str = "workspace_name_index";

/// if we have merge conflicts we write to MERGE_HEAD and ORIG_HEAD to keep track of the parents
pub const MERGE_HEAD_FILE: &str = "MERGE_HEAD";
/// if we have merge conflicts we write to MERGE_HEAD and ORIG_HEAD to keep track of the parents
pub const ORIG_HEAD_FILE: &str = "ORIG_HEAD";
/// Written at the start of a client-side merge and removed after HEAD advances.
/// Contents are the target commit id. Presence means a merge was interrupted
/// mid-mutation and a retry targeting the same commit should force-restore.
pub const MERGE_IN_PROGRESS_FILE: &str = "MERGE_IN_PROGRESS";

/// Key for if something is synced
pub const IS_SYNCED: &str = "IS_SYNCED";

/// Default branch name: main
pub const DEFAULT_BRANCH_NAME: &str = "main";
/// Default remote name: origin
pub const DEFAULT_REMOTE_NAME: &str = "origin";
/// Default remote host: hub.oxen.ai
pub const DEFAULT_HOST: &str = "hub.oxen.ai";
/// Default remote scheme: https
pub const DEFAULT_SCHEME: &str = "https";

/// Default Namespace: ox
pub const DEFAULT_NAMESPACE: &str = "ox";

/// Initial Commit Message
pub const INITIAL_COMMIT_MSG: &str = "Initialized Repo 🐂";

/// Internal Name When Performing Computation
pub const ROW_NUM_COL_NAME: &str = "_row_num";
/// Internal Name When Performing Computation
pub const ROW_HASH_COL_NAME: &str = "_row_hash";
/// Internal Name When Performing Computation
pub const FILE_ROW_NUM_COL_NAME: &str = "_file_row_num";
// Internal Name When Performing Computation
pub const TARGETS_HASH_COL: &str = "_targets_hash";
// Internal Name When Performing Computation
pub const KEYS_HASH_COL: &str = "_keys_hash";
// Internal Name When Performing Computation
pub const DIFF_STATUS_COL: &str = "_oxen_diff_status";
// Internal Name When Performing Computation
pub const DIFF_HASH_COL: &str = "_oxen_diff_hash";

/// Internal Name For Evaluations Status
pub const EVAL_STATUS_COL: &str = "_oxen_eval_status";
// Internal Name For Evaluations Error
pub const EVAL_ERROR_COL: &str = "_oxen_eval_error";
// Internal Name For Evaluations Duration
pub const EVAL_DURATION_COL: &str = "_oxen_eval_duration";

// Data transfer
// Average chunk size of ~10mb
/// Average chunk size of ~10mb when chunking and sending data
pub const AVG_CHUNK_SIZE: u64 = 1024 * 1024 * 10;
// Allow up to N concurrent upload tasks
/// Allow up to N concurrent upload tasks
pub const MAX_CONCURRENT_UPLOADS: usize = 30;
/// Cap on parallel `version_exists` probes used when bulk operations need to
/// check many hashes at once. Sized for network HEAD operations rather than
/// CPU work, so it's higher than `DEFAULT_NUM_WORKERS`.
pub const MAX_CONCURRENT_VERSION_PROBES: usize = 32;
// Limit zip file downloads to batches of size N
/// Limit zip file downloads to batches of size N
pub const MAX_ZIP_DOWNLOAD_SIZE: u64 = 1024 * 1024 * 1024; // 1 GB
// Retry and back off of upload tasks N times
/// Retry and back off of upload tasks N times
#[cfg(test)]
pub const NUM_HTTP_RETRIES: u64 = 1;
#[cfg(not(test))]
pub const NUM_HTTP_RETRIES: u64 = 5;
/// Number of workers
pub const DEFAULT_NUM_WORKERS: usize = 8;
/// Default timeout for HTTP requests
pub const DEFAULT_TIMEOUT_SECS: u64 = 600;
/// Default vnode size
pub const DEFAULT_VNODE_SIZE: u64 = 10_000;

/// Pagination page size of 10
pub const DEFAULT_PAGE_SIZE: usize = 100;
/// Pagination page number of 1
pub const DEFAULT_PAGE_NUM: usize = 1;

/// Minimum allowable oxen version to push or pull data
pub const MIN_OXEN_VERSION: MinOxenVersion = MinOxenVersion::LATEST;

/// Filepath used to track repo and server-level migration status
pub const LAST_MIGRATION_FILE: &str = "last_migration.txt";

/// Constraints for diff and compare size
pub const MAX_DISPLAY_DIRS: usize = 10;

// Oxen stack size
pub const OXEN_STACK_SIZE: usize = 16_777_216;

// Parse the maximum number of retries allowed on upload from environment variable
pub fn max_retries() -> usize {
    if let Ok(max_retries) = std::env::var("OXEN_NUM_RETRIES") {
        // If the environment variable is set, use that
        if let Ok(max_retries) = max_retries.parse::<usize>() {
            max_retries
        } else {
            // If parsing failed, fall back to default
            NUM_HTTP_RETRIES.try_into().unwrap()
        }
    } else {
        // Environment variable not set, use default
        NUM_HTTP_RETRIES.try_into().unwrap()
    }
}

// Parse the timeout for http requests from environment variable
pub fn timeout() -> u64 {
    if let Ok(timeout) = std::env::var("OXEN_TIMEOUT_SECS") {
        // If the environment variable is set, use that
        if let Ok(timeout) = timeout.parse::<u64>() {
            timeout
        } else {
            // If parsing failed, fall back to default
            DEFAULT_TIMEOUT_SECS
        }
    } else {
        // Environment variable not set, use default
        DEFAULT_TIMEOUT_SECS
    }
}

// Parse the timeout for http requests from environment variable
pub fn chunk_size() -> u64 {
    if let Ok(chunk_size) = std::env::var("OXEN_AVG_CHUNK_SIZE") {
        // If the environment variable is set, use that
        if let Ok(chunk_size) = chunk_size.parse::<u64>() {
            chunk_size
        } else {
            // If parsing failed, fall back to default
            AVG_CHUNK_SIZE
        }
    } else {
        // Environment variable not set, use default
        AVG_CHUNK_SIZE
    }
}