Fix: Allow duplicate participant names across different workspaces

[FloWuenne]

This PR resolves issue where the same participant/team couldn't be added to multiple workspaces within a single YAML configuration file. Previously this would fail due to duplicate key handling - now properly supports adding the same user to different workspace contexts.

Thanks to Chris from Barnwell Bio for spotting this!

Copilot summary:

This pull request enhances the uniqueness validation logic in the parse_yaml_block function to correctly handle resources that may have duplicate names across different contexts (such as workspaces or organizations). It also introduces a new helper function to extract context values from command arguments.

Improvements to uniqueness validation:

• Updated parse_yaml_block to track unique combinations of resource name and context (workspace/organization) for context-sensitive resources (teams, members, participants), ensuring that duplicate names are only flagged within the same context.

Helper function addition:

• Added find_context_value to extract context information (e.g., workspace or organization) from command arguments, supporting the enhanced uniqueness logic.

[adamrtalbot]

I'm not massively following how this works so I think some clarity is required.

• Explain why this is needed
• How this works
• What problem this solves

[adamrtalbot]

Maybe it's better to treat all resources as workspace specific? I think only dataset is left as one that must be globally unique within your organisation.

[adamrtalbot]

I'm not following the strategy. I think it's making a unique value by joining name and workspace into a string?

[adamrtalbot]

I also think you're missing a value here:

Suggested change

	if block_name in context_sensitive_resources:
	# For context-sensitive resources, create a unique identifier
	# that includes both name and workspace/organization context
	workspace = find_context_value(cmd_args, ["--workspace", "--organization"])
	unique_id = f"{name}@{workspace}" if workspace else name
	else:
	# For other resources, use just the name as before
	unique_id = name
	if unique_id in unique_identifiers:
	context_info = f" in context '{workspace}'" if block_name in context_sensitive_resources and workspace else ""
	if block_name in context_sensitive_resources:
	# For context-sensitive resources, create a unique identifier
	# that includes both name and workspace/organization context
	workspace = find_context_value(cmd_args, ["--workspace", "--organization"])
	unique_id = f"{name}@{workspace}" if workspace else name
	else:
	# For other resources, use just the name as before
	unique_id = name
	if unique_id in unique_identifiers:
	context_info = f"{name} in context '{workspace}'" if block_name in context_sensitive_resources and workspace else ""

[adamrtalbot]

Rather than concatenating a string, it might be helpful to make a unique ID object although this might be a lot more work:

@dataclass
class SeqeraResourceIdentifier:
    name: str
    workspace: str | None = None

Then match that using a custom method:

    def __eq__(self, other):
        if self.workspace:
            return self.name == other.name and self.workspace == other.workspace
        else:
            return self.name == other.name

This encapsulates the caching logic into a reusable component without making flaky strings. Still, it's 100% premature optimisation.