sutro 0.0.0__py3-none-any.whl → 0.1.11__py3-none-any.whl

sutro/__init__.py

@@ -0,0 +1,14 @@
+from .sdk import Sutro
+
+# Create an instance of the class
+_instance = Sutro()
+
+# Import all methods from the instance into the package namespace
+from types import MethodType
+
+for attr in dir(_instance):
+    if callable(getattr(_instance, attr)) and not attr.startswith("__"):
+        globals()[attr] = MethodType(getattr(_instance, attr).__func__, _instance)
+
+# Clean up namespace
+del MethodType, attr

sutro/cli.py

@@ -0,0 +1,418 @@
+from datetime import datetime, timezone
+import click
+from colorama import Fore, Style
+import os
+import json
+from sutro.sdk import Sutro
+import polars as pl
+import warnings
+
+warnings.filterwarnings("ignore", category=pl.PolarsInefficientMapWarning)
+pl.Config.set_tbl_hide_dataframe_shape(True)
+
+CONFIG_DIR = os.path.expanduser("~/.sutro")
+CONFIG_FILE = os.path.join(CONFIG_DIR, "config.json")
+
+
+def load_config():
+    if os.path.exists(CONFIG_FILE):
+        with open(CONFIG_FILE, "r") as f:
+            return json.load(f)
+    return {}
+
+
+def save_config(config):
+    os.makedirs(CONFIG_DIR, exist_ok=True)
+    with open(CONFIG_FILE, "w") as f:
+        json.dump(config, f)
+
+
+def check_auth():
+    config = load_config()
+    return config.get("api_key") is not None
+
+
+def get_sdk():
+    config = load_config()
+    if config.get("base_url") != None:
+        return Sutro(
+            api_key=config.get("api_key"), base_url=config.get("base_url")
+        )
+    else:
+        return Sutro(api_key=config.get("api_key"))
+
+
+def set_config_base_url(base_url: str):
+    config = load_config()
+    config["base_url"] = base_url
+    save_config(config)
+
+
+def set_human_readable_dates(datetime_columns, df):
+    for col in datetime_columns:
+        if col in df.columns:
+            # Convert UTC string to local time string
+            df = df.with_columns(
+                pl.col(col)
+                .str.to_datetime()
+                .map_elements(
+                    lambda dt: dt.replace(tzinfo=timezone.utc)
+                    .astimezone()
+                    .strftime("%Y-%m-%d %H:%M:%S %Z")
+                    if dt
+                    else None,
+                    return_dtype=pl.Utf8,
+                )
+                .alias(col)
+            )
+    return df
+
+
+@click.group(invoke_without_command=True)
+@click.pass_context
+def cli(ctx):
+    # Allow login and set-base-url commands without authentication
+    if not check_auth() and ctx.invoked_subcommand not in ["login", "set-base-url"]:
+        click.echo("Please login using 'sutro login'.")
+        ctx.exit(1)
+
+    if ctx.invoked_subcommand is None:
+        message = """
+Welcome to the Sutro CLI! 
+
+To see a list of all available commands, use 'sutro --help'.
+    """
+        click.echo(Fore.GREEN + message + Style.RESET_ALL)
+
+        click.echo(ctx.get_help())
+
+
+@cli.command()
+def login():
+    """Set or update your API key for Sutro."""
+    config = load_config()
+    default_api_key = config.get("api_key", "")
+    default_base_url = config.get("base_url", "https://api.sutro.sh")
+    click.echo(
+        "Hint: An API key is already set. Press Enter to keep the existing key."
+        if default_api_key
+        else ""
+    )
+    api_key = click.prompt(
+        "Enter your API key",
+        default=default_api_key,
+        hide_input=True,
+        show_default=False,
+    )
+
+    result = get_sdk().try_authentication(api_key)
+    if not result or "authenticated" not in result or result["authenticated"] != True:
+        raise click.ClickException(
+            Fore.RED + "Invalid API key. Try again." + Style.RESET_ALL
+        )
+    else:
+        ascii = """
+
+ 
+ ▄▄▄▄▄▄▄▄▄▄▄  ▄         ▄  ▄▄▄▄▄▄▄▄▄▄▄  ▄▄▄▄▄▄▄▄▄▄▄  ▄▄▄▄▄▄▄▄▄▄▄ 
+▐░░░░░░░░░░░▌▐░▌       ▐░▌▐░░░░░░░░░░░▌▐░░░░░░░░░░░▌▐░░░░░░░░░░░▌
+▐░█▀▀▀▀▀▀▀▀▀ ▐░▌       ▐░▌ ▀▀▀▀█░█▀▀▀▀ ▐░█▀▀▀▀▀▀▀█░▌▐░█▀▀▀▀▀▀▀█░▌
+▐░▌          ▐░▌       ▐░▌     ▐░▌     ▐░▌       ▐░▌▐░▌       ▐░▌
+▐░█▄▄▄▄▄▄▄▄▄ ▐░▌       ▐░▌     ▐░▌     ▐░█▄▄▄▄▄▄▄█░▌▐░▌       ▐░▌
+▐░░░░░░░░░░░▌▐░▌       ▐░▌     ▐░▌     ▐░░░░░░░░░░░▌▐░▌       ▐░▌
+ ▀▀▀▀▀▀▀▀▀█░▌▐░▌       ▐░▌     ▐░▌     ▐░█▀▀▀▀█░█▀▀ ▐░▌       ▐░▌
+          ▐░▌▐░▌       ▐░▌     ▐░▌     ▐░▌     ▐░▌  ▐░▌       ▐░▌
+ ▄▄▄▄▄▄▄▄▄█░▌▐░█▄▄▄▄▄▄▄█░▌     ▐░▌     ▐░▌      ▐░▌ ▐░█▄▄▄▄▄▄▄█░▌
+▐░░░░░░░░░░░▌▐░░░░░░░░░░░▌     ▐░▌     ▐░▌       ▐░▌▐░░░░░░░░░░░▌
+ ▀▀▀▀▀▀▀▀▀▀▀  ▀▀▀▀▀▀▀▀▀▀▀       ▀       ▀         ▀  ▀▀▀▀▀▀▀▀▀▀▀ 
+                                                                 
+
+"""
+        click.echo(Fore.BLUE + ascii + Style.RESET_ALL)
+        click.echo(
+            Fore.GREEN + "Successfully authenticated. Welcome back!" + Style.RESET_ALL
+        )
+
+    save_config({"api_key": api_key, "base_url": default_base_url})
+
+
+@cli.group()
+def jobs():
+    """Manage jobs."""
+    pass
+
+
+@jobs.command()
+@click.option(
+    "--all", is_flag=True, help="Include all jobs, including cancelled and failed ones."
+)
+def list(all=False):
+    """Lists historical and ongoing jobs. Will only list first 25 jobs by default. Use --all to see all jobs."""
+    sdk = get_sdk()
+    jobs = sdk.list_jobs()
+    if jobs is None or len(jobs) == 0:
+        click.echo(Fore.YELLOW + "No jobs found." + Style.RESET_ALL)
+        return
+
+    df = pl.DataFrame(jobs)
+    # TODO: this is a temporary fix to remove jobs where datetime_created is null. We should fix this on the backend.
+    df = df.filter(pl.col("datetime_created").is_not_null())
+    df = df.sort(by=["datetime_created"], descending=True)
+
+    # Format all datetime columns with a more readable format
+    datetime_columns = [
+        "datetime_created",
+        "datetime_added",
+        "datetime_started",
+        "datetime_completed",
+    ]
+    df = set_human_readable_dates(datetime_columns, df)
+
+    # TODO: get colors working
+    # df = df.with_columns([
+    #     pl.when(pl.col("status") == "SUCCEEDED")
+    #     .then(pl.concat_str([pl.lit(Fore.GREEN), pl.col("status"), pl.lit(Style.RESET_ALL)]))
+    #     .when(pl.col("status").is_in(["FAILED", "CANCELLED", "UNKNOWN"]))
+    #     .then(pl.concat_str([pl.lit(Fore.RED), pl.col("status"), pl.lit(Style.RESET_ALL)]))
+    #     .otherwise(pl.col("status"))
+    #     .alias("status")
+    # ])
+
+    # fill null input_tokens and output_tokens with 0
+    df = df.with_columns(
+        pl.col("input_tokens").fill_null(0).alias("input_tokens"),
+        pl.col("output_tokens").fill_null(0).alias("output_tokens"),
+    )
+
+    # fill null datetime_completed with empty string
+    df = df.with_columns(
+        pl.col("datetime_completed").fill_null("").alias("datetime_completed")
+    )
+
+    df = df.with_columns(
+        pl.col("job_cost")
+        .fill_null(0)
+        .map_elements(lambda x: f"${x:.5f}", return_dtype=pl.Utf8)
+        .alias("job_cost")
+    )
+
+    if all == False:
+        df = df.slice(0, 25)
+
+    with pl.Config(tbl_rows=-1, tbl_cols=-1, set_fmt_str_lengths=45):
+        print(df.select(pl.all()))
+
+
+@jobs.command()
+@click.argument("job_id")
+def status(job_id):
+    """Get the status of a job."""
+    sdk = get_sdk()
+    job_status = sdk.get_job_status(job_id)
+    if not job_status:
+        return
+
+    print(job_status)
+
+
+@jobs.command()
+@click.argument("job_id")
+@click.option(
+    "--include-inputs", is_flag=True, help="Include the inputs in the results."
+)
+@click.option(
+    "--include-cumulative-logprobs",
+    is_flag=True,
+    help="Include the cumulative logprobs in the results.",
+)
+@click.option(
+    "--save",
+    is_flag=True,
+    help="Download the results to the current working directory. The file name will be the job_id.",
+)
+@click.option(
+    "--save-format",
+    type=click.Choice(["parquet", "csv"]),
+    default="parquet",
+    help="The format of the output file. Options: parquet, csv",
+)
+def results(
+    job_id,
+    include_inputs,
+    include_cumulative_logprobs,
+    save=False,
+    save_format="parquet",
+):
+    """Get the results of a job."""
+    sdk = get_sdk()
+    job_results = sdk.get_job_results(
+        job_id, include_inputs, include_cumulative_logprobs
+    )
+    if not job_results:
+        return
+
+    df = pl.DataFrame(job_results)
+    if not save:
+        print(df)
+    elif save:
+        if save_format == "parquet":
+            df.write_parquet(f"{job_id}.parquet")
+        else:  # csv
+            df.write_csv(f"{job_id}.csv")
+        print(Fore.GREEN + f"Results saved to {job_id}.{save_format}" + Style.RESET_ALL)
+
+
+@jobs.command()
+@click.argument("job_id")
+def cancel(job_id):
+    """Cancel a running job."""
+    sdk = get_sdk()
+    result = sdk.cancel_job(job_id)
+    if not result:
+        return
+
+    click.echo(Fore.GREEN + "Job cancelled successfully." + Style.RESET_ALL)
+
+
+@cli.group()
+def stages():
+    """Manage stages."""
+    pass
+
+
+@stages.command()
+def create():
+    """Create a new stage."""
+    sdk = get_sdk()
+    stage_id = sdk.create_stage()
+    if not stage_id:
+        return
+    click.echo(
+        Fore.GREEN
+        + f"Stage created successfully. Stage ID: {stage_id}"
+        + Style.RESET_ALL
+    )
+
+
+@stages.command()
+def list():
+    """List all stages."""
+    sdk = get_sdk()
+    stages = sdk.list_stages()
+    if stages is None or len(stages) == 0:
+        click.echo(Fore.YELLOW + "No stages found." + Style.RESET_ALL)
+        return
+    df = pl.DataFrame(stages)
+
+    df = df.with_columns(
+        pl.col("schema")
+        .map_elements(lambda x: str(x), return_dtype=pl.Utf8)
+        .alias("schema")
+    )
+
+    # Format all datetime columns with a more readable format
+    datetime_columns = ["datetime_added", "updated_at"]
+    df = set_human_readable_dates(datetime_columns, df)
+
+    df = df.sort(by=["datetime_added"], descending=True)
+    with pl.Config(tbl_rows=-1, tbl_cols=-1, set_fmt_str_lengths=45):
+        print(df.select(pl.all()))
+
+
+@stages.command()
+@click.argument("stage_id")
+def files(stage_id):
+    """List all files in a stage."""
+    sdk = get_sdk()
+    files = sdk.list_stage_files(stage_id)
+    if not files:
+        return
+
+    print(Fore.YELLOW + "Files in stage " + stage_id + ":" + Style.RESET_ALL)
+    for file in files:
+        print(f"\t{file}")
+
+
+@stages.command()
+@click.argument("stage_id", required=False)
+@click.argument("file_path")
+def upload(file_path, stage_id):
+    """Upload files to a stage. You can provide a single file path or a directory path to upload all files in the directory."""
+    sdk = get_sdk()
+    sdk.upload_to_stage(file_path, stage_id)
+
+
+@stages.command()
+@click.argument("stage_id")
+@click.argument("file_name", required=False)
+@click.argument("output_path", required=False)
+def download(stage_id, file_name=None, output_path=None):
+    """Download a file/files from a stage. If no files are provided, all files in the stage will be downloaded. If no output path is provided, the file will be saved to the current working directory."""
+    sdk = get_sdk()
+    files = sdk.download_from_stage(stage_id, [file_name], output_path)
+    if not files:
+        return
+    for file in files:
+        if output_path is None:
+            with open(file_name, "wb") as f:
+                f.write(file)
+        else:
+            with open(output_path + "/" + file_name, "wb") as f:
+                f.write(file)
+
+
+@cli.command()
+def docs():
+    """Open the Sutro API docs."""
+    click.launch("https://docs.sutro.sh")
+
+
+@cli.command()
+@click.argument("base_url")
+def set_base_url(base_url):
+    """Set the base URL for the Sutro API."""
+    set_config_base_url(base_url)
+    click.echo(Fore.GREEN + f"Base URL set to {base_url}." + Style.RESET_ALL)
+
+
+@cli.command()
+def quotas():
+    """Get API quotas."""
+    sdk = get_sdk()
+    quotas = sdk.get_quotas()
+    if not quotas:
+        return
+    print(Fore.YELLOW + "Your current quotas are: \n" + Style.RESET_ALL)
+    for priority in range(len(quotas)):
+        quota = quotas[priority]
+        print(f"Job Priority: {priority}")
+        print(f"\tRow Quota (Maximum): {quota['row_quota']}")
+        print(f"\tToken Quota (Maximum): {quota['token_quota']}")
+        print("\n")
+    print(
+        Fore.YELLOW
+        + "To increase your quotas, contact us at team@sutro.sh."
+        + Style.RESET_ALL
+    )
+
+@jobs.command()
+@click.argument("job_id", required=False)
+@click.option("--latest", is_flag=True, help="Attach to the latest job.")
+def attach(job_id, latest):
+    """Attach to a running job and stream its progress."""
+    sdk = get_sdk()
+    if latest:
+        jobs = sdk.list_jobs()
+        if not jobs:
+            click.echo(Fore.YELLOW + "No jobs found." + Style.RESET_ALL)
+            return
+        job_id = jobs[0]["job_id"]
+        print(f"Attaching to latest job: {job_id}")
+    elif not job_id:
+        click.echo(Fore.YELLOW + "No job ID provided." + Style.RESET_ALL)
+        return
+    sdk.attach(job_id)
+
+
+if __name__ == "__main__":
+    cli()