cloudos-cli 2.53.0__tar.gz → 2.56.0__tar.gz

{cloudos_cli-2.53.0 → cloudos_cli-2.56.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: cloudos_cli
-Version: 2.53.0
+Version: 2.56.0
 Summary: Python package for interacting with CloudOS
 Home-page: https://github.com/lifebit-ai/cloudos-cli
 Author: David Piñeyro
@@ -597,6 +597,34 @@ Executing results...
 results: s3://path/to/location/of/results/results/
 ```
 
+#### Query working directory of job
+
+To get the working directory of a job submitted to CloudOS:
+
+```shell
+cloudos job workdir \
+    --apikey $MY_API_KEY \
+    --cloudos-url $CLOUDOS \
+    --job-id 62c83a1191fe06013b7ef355
+```
+
+Or with a defined profile:
+
+```shell
+cloudos job workdir \
+    --profile profile-name \
+    --job-id 62c83a1191fe06013b7ef355
+```
+
+The output should be something similar to:
+
+```console
+CloudOS job functionality: run, check and abort jobs in CloudOS.
+
+Finding working directory path...
+Working directory for job 68747bac9e7fe38ec6e022ad: az://123456789000.blob.core.windows.net/cloudos-987652349087/projects/455654676/jobs/54678856765/work
+```
+
 #### Abort single or multiple jobs from CloudOS
 
 Aborts jobs in the CloudOS workspace that are either running or initialising. It can be used with one or more job IDs provided as a comma separated string using the `--job-ids` parameter.
@@ -615,20 +643,20 @@ Aborting jobs...
 ```
 
 
-#### Clone a job with optional parameter overrides
+#### Clone/resume a job with optional parameter overrides
 
-The `clone` command allows you to create a new job based on an existing job's configuration, with the ability to override specific parameters. This is useful for re-running jobs with slight modifications without having to specify all parameters from scratch.
+The `clone` and `resume` commands allows you to create a new job based on an existing job's configuration, with the ability to override specific parameters. This is useful for re-running jobs with slight modifications without having to specify all parameters from scratch.
 
 Basic usage:
 ```console
-cloudos job clone \
+cloudos job clone/resume \
     --profile MY_PROFILE
     --job-id "60a7b8c9d0e1f2g3h4i5j6k7"
 ```
 
-Clone with parameter overrides:
+Clone/resume with parameter overrides:
 ```console
-cloudos job clone \
+cloudos job clone/resume \
     --profile MY_PROFILE
     --job-id "60a7b8c9d0e1f2g3h4i5j6k7" \
     --job-queue "high-priority-queue" \
@@ -649,7 +677,7 @@ Available override options:
 - `--job-queue`: Specify a different job queue
 - `--cost-limit`: Set a new cost limit (use -1 for no limit)
 - `--instance-type`: Change the master instance type
-- `--job-name`: Assign a custom name to the cloned job
+- `--job-name`: Assign a custom name to the cloned/resumed job
 - `--nextflow-version`: Use a different Nextflow version
 - `--git-branch`: Switch to a different git branch
 - `--nextflow-profile`: Change the Nextflow profile

{cloudos_cli-2.53.0 → cloudos_cli-2.56.0}/README.md

@@ -562,6 +562,34 @@ Executing results...
 results: s3://path/to/location/of/results/results/
 ```
 
+#### Query working directory of job
+
+To get the working directory of a job submitted to CloudOS:
+
+```shell
+cloudos job workdir \
+    --apikey $MY_API_KEY \
+    --cloudos-url $CLOUDOS \
+    --job-id 62c83a1191fe06013b7ef355
+```
+
+Or with a defined profile:
+
+```shell
+cloudos job workdir \
+    --profile profile-name \
+    --job-id 62c83a1191fe06013b7ef355
+```
+
+The output should be something similar to:
+
+```console
+CloudOS job functionality: run, check and abort jobs in CloudOS.
+
+Finding working directory path...
+Working directory for job 68747bac9e7fe38ec6e022ad: az://123456789000.blob.core.windows.net/cloudos-987652349087/projects/455654676/jobs/54678856765/work
+```
+
 #### Abort single or multiple jobs from CloudOS
 
 Aborts jobs in the CloudOS workspace that are either running or initialising. It can be used with one or more job IDs provided as a comma separated string using the `--job-ids` parameter.
@@ -580,20 +608,20 @@ Aborting jobs...
 ```
 
 
-#### Clone a job with optional parameter overrides
+#### Clone/resume a job with optional parameter overrides
 
-The `clone` command allows you to create a new job based on an existing job's configuration, with the ability to override specific parameters. This is useful for re-running jobs with slight modifications without having to specify all parameters from scratch.
+The `clone` and `resume` commands allows you to create a new job based on an existing job's configuration, with the ability to override specific parameters. This is useful for re-running jobs with slight modifications without having to specify all parameters from scratch.
 
 Basic usage:
 ```console
-cloudos job clone \
+cloudos job clone/resume \
     --profile MY_PROFILE
     --job-id "60a7b8c9d0e1f2g3h4i5j6k7"
 ```
 
-Clone with parameter overrides:
+Clone/resume with parameter overrides:
 ```console
-cloudos job clone \
+cloudos job clone/resume \
     --profile MY_PROFILE
     --job-id "60a7b8c9d0e1f2g3h4i5j6k7" \
     --job-queue "high-priority-queue" \
@@ -614,7 +642,7 @@ Available override options:
 - `--job-queue`: Specify a different job queue
 - `--cost-limit`: Set a new cost limit (use -1 for no limit)
 - `--instance-type`: Change the master instance type
-- `--job-name`: Assign a custom name to the cloned job
+- `--job-name`: Assign a custom name to the cloned/resumed job
 - `--nextflow-version`: Use a different Nextflow version
 - `--git-branch`: Switch to a different git branch
 - `--nextflow-profile`: Change the Nextflow profile

{cloudos_cli-2.53.0 → cloudos_cli-2.56.0}/cloudos_cli/__main__.py

@@ -100,9 +100,11 @@ def run_cloudos_cli(ctx, debug):
                 'status': shared_config,
                 'list': shared_config,
                 'logs': shared_config,
+                'workdir': shared_config,
                 'results': shared_config,
                 'details': shared_config,
-                'clone': shared_config
+                'clone': shared_config,
+                'resume': shared_config
             },
             'workflow': {
                 'list': shared_config,
@@ -162,9 +164,11 @@ def run_cloudos_cli(ctx, debug):
                 'status': shared_config,
                 'list': shared_config,
                 'logs': shared_config,
+                'workdir': shared_config,
                 'results': shared_config,
                 'details': shared_config,
-                'clone': shared_config
+                'clone': shared_config,
+                'resume': shared_config
             },
             'workflow': {
                 'list': shared_config,
@@ -207,7 +211,7 @@ def run_cloudos_cli(ctx, debug):
 
 @run_cloudos_cli.group()
 def job():
-    """CloudOS job functionality: run, check and abort jobs in CloudOS."""
+    """CloudOS job functionality: run, clone, resume, check and abort jobs in CloudOS."""
     print(job.__doc__ + '\n')
 
 
@@ -803,6 +807,83 @@ def job_status(ctx,
           'or repeat the command you just used.')
 
 
+@job.command('workdir')
+@click.option('-k',
+              '--apikey',
+              help='Your CloudOS API key',
+              required=True)
+@click.option('-c',
+              '--cloudos-url',
+              help=(f'The CloudOS url you are trying to access to. Default={CLOUDOS_URL}.'),
+              default=CLOUDOS_URL,
+              required=True)
+@click.option('--workspace-id',
+              help='The specific CloudOS workspace id.',
+              required=True)
+@click.option('--job-id',
+              help='The job id in CloudOS to search for.',
+              required=True)
+@click.option('--verbose',
+              help='Whether to print information messages or not.',
+              is_flag=True)
+@click.option('--disable-ssl-verification',
+              help=('Disable SSL certificate verification. Please, remember that this option is ' +
+                    'not generally recommended for security reasons.'),
+              is_flag=True)
+@click.option('--ssl-cert',
+              help='Path to your SSL certificate file.')
+@click.option('--profile', help='Profile to use from the config file', default=None)
+@click.pass_context
+def job_workdir(ctx,
+             apikey,
+             cloudos_url,
+             workspace_id,
+             job_id,
+             verbose,
+             disable_ssl_verification,
+             ssl_cert,
+             profile):
+    """Get the path to the working directory of a specified job."""
+    profile = profile or ctx.default_map['job']['workdir']['profile']
+    # Create a dictionary with required and non-required params
+    required_dict = {
+        'apikey': True,
+        'workspace_id': True,
+        'workflow_name': False,
+        'project_name': False,
+        'procurement_id': False
+    }
+    # determine if the user provided all required parameters
+    config_manager = ConfigurationProfile()
+    user_options = (
+        config_manager.load_profile_and_validate_data(
+            ctx,
+            INIT_PROFILE,
+            CLOUDOS_URL,
+            profile=profile,
+            required_dict=required_dict,
+            apikey=apikey,
+            cloudos_url=cloudos_url,
+            workspace_id=workspace_id
+        )
+    )
+    apikey = user_options['apikey']
+    cloudos_url = user_options['cloudos_url']
+    workspace_id = user_options['workspace_id']
+
+    print('Finding working directory path...')
+    verify_ssl = ssl_selector(disable_ssl_verification, ssl_cert)
+    if verbose:
+        print('\t...Preparing objects')
+    cl = Cloudos(cloudos_url, apikey, None)
+    if verbose:
+        print('\tThe following Cloudos object was created:')
+        print('\t' + str(cl) + '\n')
+        print(f'\tSearching for job id: {job_id}')
+    workdir = cl.get_job_workdir(job_id, workspace_id, verify_ssl)
+    print(f"Working directory for job {job_id}: {workdir}")
+
+
 @job.command('logs')
 @click.option('-k',
               '--apikey',
@@ -1455,8 +1536,7 @@ def abort_jobs(ctx,
             cl.abort_job(job, workspace_id, verify_ssl)
             print(f"\tJob '{job}' aborted successfully.")
 
-
-@job.command('clone')
+@click.command()
 @click.option('-k',
               '--apikey',
               help='Your CloudOS API key',
@@ -1526,7 +1606,7 @@ def abort_jobs(ctx,
               help='Profile to use from the config file',
               default=None)
 @click.pass_context
-def clone_job(ctx,
+def clone_resume(ctx,
               apikey,
               cloudos_url,
               workspace_id,
@@ -1547,8 +1627,13 @@ def clone_job(ctx,
               disable_ssl_verification,
               ssl_cert,
               profile):
-    """Clone an existing job with optional parameter overrides."""
-    profile = profile or ctx.default_map['job']['clone']['profile']
+    if ctx.info_name == "clone":
+        mode, action = "clone", "cloning"
+    elif ctx.info_name == "resume":
+        mode, action = "resume", "resuming"
+
+    f"""{mode.capitalize()} an existing job with optional parameter overrides."""
+    profile = profile or ctx.default_map['job'][mode]['profile']
 
     # Create a dictionary with required and non-required params
     required_dict = {
@@ -1580,7 +1665,7 @@ def clone_job(ctx,
 
     verify_ssl = ssl_selector(disable_ssl_verification, ssl_cert)
 
-    print('Cloning job...')
+    print(f'{action.capitalize()} job...')
     if verbose:
         print('\t...Preparing objects')
 
@@ -1591,11 +1676,11 @@ def clone_job(ctx,
     if verbose:
         print('\tThe following Job object was created:')
         print('\t' + str(job_obj) + '\n')
-        print(f'\tCloning job {job_id} in workspace: {workspace_id}')
+        print(f'\t{action.capitalize()} job {job_id} in workspace: {workspace_id}')
 
     try:
-        # Clone the job with provided overrides
-        cloned_job_id = job_obj.clone_job(
+        # Clone/resume the job with provided overrides
+        cloned_resumed_job_id = job_obj.clone_or_resume_job(
             source_job_id=job_id,
             queue_name=job_queue,
             cost_limit=cost_limit,
@@ -1610,21 +1695,25 @@ def clone_job(ctx,
             # only when explicitly setting --project-name will be overridden, else using the original project
             project_name=project_name if ctx.get_parameter_source("project_name") == click.core.ParameterSource.COMMANDLINE else None,
             parameters=list(parameter) if parameter else None,
-            verify=verify_ssl
+            verify=verify_ssl,
+            mode=mode
         )
         if verbose:
-            print(f'\tCloned job ID: {cloned_job_id}')
+            print(f'\t{mode.capitalize()}d job ID: {cloned_resumed_job_id}')
 
-        print(f"Job successfully cloned. New job ID: {cloned_job_id}")
+        print(f"Job successfully {mode}d. New job ID: {cloned_resumed_job_id}")
 
     except BadRequestException as e:
         if verbose:
             print(f'\tError details: {e}')
-        raise ValueError(f"Failed to clone job: {e}")
+        raise ValueError(f"Failed to {mode} job: {e}")
     except Exception as e:
         if verbose:
             print(f'\tError details: {e}')
-        raise ValueError(f"An error occurred while cloning the job: {e}")
+        raise ValueError(f"An error occurred while {action} the job: {e}")
+# Register the same function under two names
+job.add_command(clone_resume, "clone")
+job.add_command(clone_resume, "resume")
 
 
 @workflow.command('list')
@@ -3035,7 +3124,7 @@ def run_bash_array_job(ctx,
 @click.option('--details',
               help=('When selected, it prints the details of the listed files. ' +
                     'Details contains "Type", "Owner", "Size", "Last Updated", ' +
-                    '"File Name", "Storage Path".'),
+                    '"Virtual Name", "Storage Path".'),
               is_flag=True)
 @click.pass_context
 def list_files(ctx,
@@ -3104,7 +3193,7 @@ def list_files(ctx,
             table.add_column("Owner", style="white")
             table.add_column("Size", style="magenta")
             table.add_column("Last Updated", style="green")
-            table.add_column("File Name", style="bold", overflow="fold")
+            table.add_column("Virtual Name", style="bold", overflow="fold")
             table.add_column("Storage Path", style="dim", no_wrap=False, overflow="fold", ratio=2)
 
             for item in contents:
@@ -3295,7 +3384,7 @@ def move_files(ctx, source_path, destination_path, apikey, cloudos_url, workspac
         if folder_type in ("VirtualFolder", "Folder"):
             target_kind = "Folder"
         elif folder_type=="S3Folder":
-            click.echo(f"[ERROR] Item '{source_item_name}' could not be moved to '{destination_path}' as the destination folder is not modifiable.",
+            click.echo(f"[ERROR] Unable to move item '{source_item_name}' to '{destination_path}'. The destination is an S3 folder, and only virtual folders can be selected as valid move destinations.",
                    err=True)
             sys.exit(1)
         elif isinstance(folder_type, bool) and folder_type:  # legacy dataset structure
@@ -3785,20 +3874,20 @@ def rm_item(ctx, target_path, apikey, cloudos_url,
         click.echo(f"[ERROR] Item '{item_name}' could not be removed as the parent folder is not modifiable.",
                    err=True)
         sys.exit(1)
-    click.echo(f"Deleting {kind} '{item_name}' from '{parent_path or '[root]'}'...")
+    click.echo(f"Removing {kind} '{item_name}' from '{parent_path or '[root]'}'...")
     try:
         response = client.delete_item(item_id=item_id, kind=kind)
         if response.ok:
             click.secho(
-                f"[SUCCESS] {kind} '{item_name}' was deleted from '{parent_path or '[root]'}'.",
+                f"[SUCCESS] {kind} '{item_name}' was removed from '{parent_path or '[root]'}'.",
                 fg="green", bold=True
             )
             click.secho("This item will still be available on your Cloud Provider.", fg="yellow")
         else:
-            click.echo(f"[ERROR] Deletion failed: {response.status_code} - {response.text}", err=True)
+            click.echo(f"[ERROR] Removal failed: {response.status_code} - {response.text}", err=True)
             sys.exit(1)
     except Exception as e:
-        click.echo(f"[ERROR] Delete operation failed: {str(e)}", err=True)
+        click.echo(f"[ERROR] Remove operation failed: {str(e)}", err=True)
         sys.exit(1)
 
 
@@ -3872,7 +3961,79 @@ def link(ctx, path, apikey, cloudos_url, project_name, workspace_id, session_id,
         project_name=project_name,
         verify=verify_ssl
     )
-    link_p.link_folder(path, session_id)
+
+    # Minimal folder validation and improved error messages
+    is_s3 = path.startswith("s3://")
+    is_folder = True
+    
+    if is_s3:
+        # S3 path validation - use heuristics to determine if it's likely a folder
+        try:
+            # If path ends with '/', it's likely a folder
+            if path.endswith('/'):
+                is_folder = True
+            else:
+                # Check the last part of the path
+                path_parts = path.rstrip("/").split("/")
+                if path_parts:
+                    last_part = path_parts[-1]
+                    # If the last part has no dot, it's likely a folder
+                    if '.' not in last_part:
+                        is_folder = True
+                    else:
+                        # If it has a dot, it might be a file - set to None for warning
+                        is_folder = None
+                else:
+                    # Empty path parts, set to None for uncertainty
+                    is_folder = None
+        except Exception:
+            # If we can't parse the S3 path, set to None for uncertainty
+            is_folder = None
+    else:
+        # File Explorer path validation (existing logic)
+        try:
+            datasets = Datasets(
+                cloudos_url=cloudos_url,
+                apikey=apikey,
+                workspace_id=workspace_id,
+                project_name=project_name,
+                verify=verify_ssl,
+                cromwell_token=None
+            )
+            parts = path.strip("/").split("/")
+            parent_path = "/".join(parts[:-1]) if len(parts) > 1 else ""
+            item_name = parts[-1]
+            contents = datasets.list_folder_content(parent_path)
+            found = None
+            for item in contents.get("folders", []):
+                if item.get("name") == item_name:
+                    found = item
+                    break
+            if not found:
+                for item in contents.get("files", []):
+                    if item.get("name") == item_name:
+                        found = item
+                        break
+            if found and ("folderType" not in found):
+                is_folder = False
+        except Exception:
+            is_folder = None
+
+    if is_folder is False:
+        if is_s3:
+            click.echo("[ERROR] The S3 path appears to point to a file, not a folder. You can only link folders. Please link the parent folder instead.", err=True)
+        else:
+            click.echo("[ERROR] Linking is only supported for folders, not individual files. Please link the parent folder instead.", err=True)
+        return
+    elif is_folder is None and is_s3:
+        click.echo("[WARNING] Unable to verify whether the S3 path is a folder. Proceeding with linking; however, if the operation fails, please confirm that you are linking a folder rather than a file.", err=True)
+
+    try:
+        link_p.link_folder(path, session_id)
+    except Exception as e:
+        click.echo(f"[ERROR] Could not link folder: {e}", err=True)
+        if is_s3:
+            click.echo("If you are linking an S3 path, please ensure it is a folder.", err=True)
 
 
 @images.command(name="ls")

cloudos_cli-2.56.0/cloudos_cli/_version.py

@@ -0,0 +1 @@
+__version__ = '2.56.0'

{cloudos_cli-2.53.0 → cloudos_cli-2.56.0}/cloudos_cli/clos.py

@@ -2,12 +2,13 @@
 This is the main class of the package.
 """
 
+from numpy import r_
 import requests
 import time
 import json
 from dataclasses import dataclass
 from cloudos_cli.utils.cloud import find_cloud
-from cloudos_cli.utils.errors import BadRequestException, JoBNotCompletedException, NotAuthorisedException
+from cloudos_cli.utils.errors import BadRequestException, JoBNotCompletedException, NotAuthorisedException, JobAccessDeniedException
 from cloudos_cli.utils.requests import retry_requests_get, retry_requests_post, retry_requests_put
 import pandas as pd
 from cloudos_cli.utils.last_wf import youngest_workflow_id_by_name
@@ -192,6 +193,124 @@ class Cloudos:
             raise BadRequestException(contents_req)
         return contents_req.json()["contents"]
 
+    def get_job_workdir(self, j_id, workspace_id, verify=True):
+        """
+        Get the working directory for the specified job
+        """
+        cloudos_url = self.cloudos_url
+        apikey = self.apikey
+        headers = {
+            "Content-type": "application/json",
+            "apikey": apikey
+        }
+        r = retry_requests_get(f"{cloudos_url}/api/v1/jobs/{j_id}", headers=headers, verify=verify)
+        if r.status_code == 401:
+            raise NotAuthorisedException
+        elif r.status_code == 403:
+            # Handle 403 with more informative error message
+            self._handle_job_access_denied(j_id, workspace_id, verify)
+        elif r.status_code >= 400:
+            raise BadRequestException(r)
+        r_json = r.json()
+        job_workspace = r_json["team"]
+        if job_workspace != workspace_id:
+            raise ValueError("Workspace provided or configured is different from workspace where the job was executed")
+        
+        # Check if logs field exists, if not fall back to original folder-based approach
+        if "logs" in r_json:
+            # Get workdir information from logs object using the same pattern as get_job_logs
+            logs_obj = r_json["logs"]
+            cloud_name, cloud_meta, cloud_storage = find_cloud(self.cloudos_url, self.apikey, workspace_id, logs_obj)
+            container_name = cloud_storage["container"]
+            prefix_name = cloud_storage["prefix"]
+            logs_bucket = logs_obj[container_name]
+            logs_path = logs_obj[prefix_name]
+            
+            # Construct workdir path by replacing '/logs' with '/work' in the logs path
+            workdir_path_suffix = logs_path.replace('/logs', '/work')
+            
+            if cloud_name == "aws":
+                workdir_path = f"s3://{logs_bucket}/{workdir_path_suffix}"
+            elif cloud_name == "azure":
+                storage_account_prefix = ''
+                cloude_scheme = cloud_storage["scheme"]
+                if cloude_scheme == 'az':
+                    storage_account_prefix = f"az://{cloud_meta['storage']['storageAccount']}.blob.core.windows.net"
+                workdir_path = f"{storage_account_prefix}/{logs_bucket}/{workdir_path_suffix}"
+            else:
+                raise ValueError("Unsupported cloud provider")
+            return workdir_path
+        else:
+            # Fallback to original folder-based approach for backward compatibility
+            workdir_id = r_json["resumeWorkDir"]
+
+            # This will fail, as the API endpoint is not open. This works when adding
+            # the authorisation bearer token manually to the headers
+            workdir_bucket_r = retry_requests_get(f"{cloudos_url}/api/v1/folders",
+                                                    params=dict(id=workdir_id, teamId=workspace_id), 
+                                                    headers=headers, verify=verify)
+            if workdir_bucket_r.status_code == 401:
+                raise NotAuthorisedException
+            elif workdir_bucket_r.status_code >= 400:
+                raise BadRequestException(workdir_bucket_r)
+
+            workdir_bucket_o = workdir_bucket_r.json()
+            if len(workdir_bucket_o) > 1:
+                raise ValueError(f"Request returned more than one result for folder id {workdir_id}")
+            workdir_bucket_info = workdir_bucket_o[0]
+            if workdir_bucket_info["folderType"] == "S3Folder":
+                cloud_name = "aws"
+            elif workdir_bucket_info["folderType"] == "AzureBlobFolder":
+                cloud_name = "azure"
+            else:
+                raise ValueError("Unsupported cloud provider")
+            if cloud_name == "aws":
+                bucket_name = workdir_bucket_info["s3BucketName"]
+                bucket_path = workdir_bucket_info["s3Prefix"]
+                workdir_path = f"s3://{bucket_name}/{bucket_path}"
+            elif cloud_name == "azure":
+                storage_account = f"az://{workspace_id}.blob.core.windows.net"
+                container_name = workdir_bucket_info["blobContainerName"]
+                blob_prefix = workdir_bucket_info["blobPrefix"]
+                workdir_path = f"{storage_account}/{container_name}/{blob_prefix}"
+            else:
+                raise ValueError("Unsupported cloud provider")
+            return workdir_path
+
+    def _handle_job_access_denied(self, job_id, workspace_id, verify=True):
+        """
+        Handle 403 errors with more informative messages by checking job ownership
+        """
+        try:
+            # Try to get current user info
+            current_user = self.get_user_info(verify)
+            current_user_name = f"{current_user.get('name', '')} {current_user.get('surname', '')}".strip()
+            if not current_user_name:
+                current_user_name = current_user.get('email', 'Unknown')
+        except Exception:
+            current_user_name = None
+
+        try:
+            # Try to get job info from job list to see the owner
+            jobs = self.get_job_list(workspace_id, last_n_jobs='all', verify=verify)
+            job_owner_name = None
+            
+            for job in jobs:
+                if job.get('_id') == job_id:
+                    user_info = job.get('user', {})
+                    job_owner_name = f"{user_info.get('name', '')} {user_info.get('surname', '')}".strip()
+                    if not job_owner_name:
+                        job_owner_name = user_info.get('email', 'Unknown')
+                    break
+            
+            raise JobAccessDeniedException(job_id, job_owner_name, current_user_name)
+        except JobAccessDeniedException:
+            # Re-raise the specific exception
+            raise
+        except Exception:
+            # If we can't get detailed info, fall back to generic message
+            raise JobAccessDeniedException(job_id)
+
     def get_job_logs(self, j_id, workspace_id, verify=True):
         """
         Get the location of the logs for the specified job

{cloudos_cli-2.53.0 → cloudos_cli-2.56.0}/cloudos_cli/jobs/job.py

@@ -984,7 +984,34 @@ class Job(Cloudos):
                 return True
         return False
 
-    def clone_job(self, 
+    def get_resume_work_dir(self, job_id, verify=True):
+        """Get the resume work directory id for a job.
+
+        Parameters
+        ----------
+        job_id : str
+            The CloudOS job ID to get the resume work directory for.
+        verify : [bool|string]
+            Whether to use SSL verification or not. Alternatively, if
+            a string is passed, it will be interpreted as the path to
+            the SSL certificate file.
+
+        Returns
+        -------
+        str
+            The resume work directory id.
+        """
+        headers = {
+            "Content-type": "application/json",
+            "apikey": self.apikey
+        }
+        url = f"{self.cloudos_url}/api/v1/jobs/{job_id}?teamId={self.workspace_id}"
+        r = retry_requests_get(url, headers=headers, verify=verify)
+        if r.status_code >= 400:
+            raise BadRequestException(r)
+        return json.loads(r.content)["resumeWorkDir"]
+
+    def clone_or_resume_job(self, 
                   source_job_id,
                   queue_name=None,
                   cost_limit=None,
@@ -998,13 +1025,14 @@ class Job(Cloudos):
                   resumable=None,
                   project_name=None,
                   parameters=None,
-                  verify=True):
-        """Clone an existing job with optional parameter overrides.
+                  verify=True,
+                  mode=None):
+        """Clone or resume an existing job with optional parameter overrides.
         
         Parameters
         ----------
         source_job_id : str
-            The CloudOS job ID to clone from.
+            The CloudOS job ID to clone/resume from.
         queue_name : str, optional
             Name of the job queue to use.
         cost_limit : float, optional
@@ -1033,11 +1061,13 @@ class Job(Cloudos):
             Whether to use SSL verification or not. Alternatively, if
             a string is passed, it will be interpreted as the path to
             the SSL certificate file.
-            
+        mode : str, optional
+            The mode to use for the job (e.g. "clone", "resume").
+
         Returns
         -------
         str
-            The CloudOS job ID of the cloned job.
+            The CloudOS job ID of the cloned/resumed job.
         """
         # Get the original job payload
         original_payload = self.get_job_request_payload(source_job_id, verify=verify)
@@ -1048,6 +1078,11 @@ class Job(Cloudos):
         # remove unwanted fields
         del cloned_payload['_id']
         del cloned_payload['resourceId']
+        if mode == "resume":
+            try:
+                cloned_payload['resumeWorkDir'] = self.get_resume_work_dir(source_job_id, verify=verify)
+            except Exception as e:
+                print(f"Failed to get resume work directory: {e}, the job was not set as resumable when originally run\n")
 
         # Override job name if provided
         if job_name:
@@ -1097,8 +1132,10 @@ class Job(Cloudos):
             print("[Message]: Azure workspace does not use fusion filesystem, option '--accelerate-file-staging' is ignored.\n")
 
         # Override resumable if provided
-        if resumable:
+        if resumable and mode == "clone":
             cloned_payload['resumable'] = resumable
+        elif resumable and mode == "resume":
+            print("[Message]: 'resumable' option is only applicable when resuming a job, ignoring '--resumable' flag.\n")
 
         # Handle job queue override
         if queue_name:
@@ -1151,7 +1188,6 @@ class Job(Cloudos):
             "Content-type": "application/json",
             "apikey": self.apikey
         }
-
         r = retry_requests_post(f"{self.cloudos_url}/api/v2/jobs?teamId={self.workspace_id}",
                                 data=json.dumps(cloned_payload), 
                                 headers=headers, 
@@ -1161,6 +1197,6 @@ class Job(Cloudos):
             raise BadRequestException(r)
 
         j_id = json.loads(r.content)["jobId"]
-        print('\tJob successfully cloned and launched to CloudOS, please check the ' +
+        print(f'\tJob successfully {mode}d and launched to CloudOS, please check the ' +
               f"following link: {self.cloudos_url}/app/advanced-analytics/analyses/{j_id}\n")
         return j_id

{cloudos_cli-2.53.0 → cloudos_cli-2.56.0}/cloudos_cli/utils/errors.py

@@ -78,3 +78,21 @@ class NoCloudForWorkspaceException(Exception):
         msg = f"Workspace ID {workspace_id} is not associated with supported cloud providers. Check the workspace ID"
         super(NoCloudForWorkspaceException, self).__init__(msg)
         self.workspace_id = workspace_id
+
+
+class JobAccessDeniedException(Exception):
+    def __init__(self, job_id, job_owner_name=None, current_user_name=None):
+        if job_owner_name and current_user_name:
+            msg = (f"Access denied to job {job_id}. This job belongs to {job_owner_name}, "
+                   f"but you are authenticated as {current_user_name}. "
+                   f"You can only access jobs that belong to your account.")
+        elif job_owner_name:
+            msg = (f"Access denied to job {job_id}. This job belongs to {job_owner_name}. "
+                   f"You can only access jobs that belong to your account.")
+        else:
+            msg = (f"Access denied to job {job_id}. You can only access jobs that belong to your account. "
+                   f"This job may belong to another user or you may not have the required permissions.")
+        super(JobAccessDeniedException, self).__init__(msg)
+        self.job_id = job_id
+        self.job_owner_name = job_owner_name
+        self.current_user_name = current_user_name

{cloudos_cli-2.53.0 → cloudos_cli-2.56.0}/cloudos_cli.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: cloudos_cli
-Version: 2.53.0
+Version: 2.56.0
 Summary: Python package for interacting with CloudOS
 Home-page: https://github.com/lifebit-ai/cloudos-cli
 Author: David Piñeyro
@@ -597,6 +597,34 @@ Executing results...
 results: s3://path/to/location/of/results/results/
 ```
 
+#### Query working directory of job
+
+To get the working directory of a job submitted to CloudOS:
+
+```shell
+cloudos job workdir \
+    --apikey $MY_API_KEY \
+    --cloudos-url $CLOUDOS \
+    --job-id 62c83a1191fe06013b7ef355
+```
+
+Or with a defined profile:
+
+```shell
+cloudos job workdir \
+    --profile profile-name \
+    --job-id 62c83a1191fe06013b7ef355
+```
+
+The output should be something similar to:
+
+```console
+CloudOS job functionality: run, check and abort jobs in CloudOS.
+
+Finding working directory path...
+Working directory for job 68747bac9e7fe38ec6e022ad: az://123456789000.blob.core.windows.net/cloudos-987652349087/projects/455654676/jobs/54678856765/work
+```
+
 #### Abort single or multiple jobs from CloudOS
 
 Aborts jobs in the CloudOS workspace that are either running or initialising. It can be used with one or more job IDs provided as a comma separated string using the `--job-ids` parameter.
@@ -615,20 +643,20 @@ Aborting jobs...
 ```
 
 
-#### Clone a job with optional parameter overrides
+#### Clone/resume a job with optional parameter overrides
 
-The `clone` command allows you to create a new job based on an existing job's configuration, with the ability to override specific parameters. This is useful for re-running jobs with slight modifications without having to specify all parameters from scratch.
+The `clone` and `resume` commands allows you to create a new job based on an existing job's configuration, with the ability to override specific parameters. This is useful for re-running jobs with slight modifications without having to specify all parameters from scratch.
 
 Basic usage:
 ```console
-cloudos job clone \
+cloudos job clone/resume \
     --profile MY_PROFILE
     --job-id "60a7b8c9d0e1f2g3h4i5j6k7"
 ```
 
-Clone with parameter overrides:
+Clone/resume with parameter overrides:
 ```console
-cloudos job clone \
+cloudos job clone/resume \
     --profile MY_PROFILE
     --job-id "60a7b8c9d0e1f2g3h4i5j6k7" \
     --job-queue "high-priority-queue" \
@@ -649,7 +677,7 @@ Available override options:
 - `--job-queue`: Specify a different job queue
 - `--cost-limit`: Set a new cost limit (use -1 for no limit)
 - `--instance-type`: Change the master instance type
-- `--job-name`: Assign a custom name to the cloned job
+- `--job-name`: Assign a custom name to the cloned/resumed job
 - `--nextflow-version`: Use a different Nextflow version
 - `--git-branch`: Switch to a different git branch
 - `--nextflow-profile`: Change the Nextflow profile

cloudos_cli-2.53.0/cloudos_cli/_version.py

@@ -1 +0,0 @@
-__version__ = '2.53.0'