fix output for query cli

[jperez999]

Description

cleans output of query sub command to only display source, page_number and text.

Checklist

• I am familiar with the Contributing Guidelines.
• New or existing tests cover these changes.
• The documentation is up to date with these changes.
• If adjusting docker-compose.yaml environment variables have you ensured those are mimicked in the Helm values.yaml file.

[greptile-apps]

Greptile Summary

This PR filters the output of the query CLI sub-command to expose only text, source, and page_number, stripping internal fields like _distance from the JSON printed to the user.

• sdk_workflow.py: a projection list-comprehension is added after retriever.query(), using .get() with defaults for each public field.
• test_root_cli_workflow.py: the fixture is split into raw_hits (full VDB response) and public_hits (projected output), but the final assertions were not updated and still reference the now-undefined hits variable, causing a NameError at test runtime.

Confidence Score: 4/5

Safe to merge once the stale hits reference in the test assertions is corrected — the source change itself is correct.

The production code change is sound, but the test was left with two assertions referencing a deleted variable (hits), so the test suite will error out every time it runs until this is fixed.

nemo_retriever/tests/test_root_cli_workflow.py — lines 369-370 reference the undefined hits variable.

Important Files Changed

Filename	Overview
nemo_retriever/src/nemo_retriever/adapters/cli/sdk_workflow.py	Adds a projection step after retriever.query() to expose only text/source/page_number; uses .get() safely for all fields.
nemo_retriever/tests/test_root_cli_workflow.py	Renames hits to raw_hits/public_hits but leaves two assertions referencing the deleted hits variable, causing a NameError at runtime.

Sequence Diagram

sequenceDiagram
    participant User
    participant CLI as CLI (main.py)
    participant QD as query_documents()
    participant R as Retriever

    User->>CLI: retriever query "..."
    CLI->>QD: "query_documents(query, **opts)"
    QD->>R: "Retriever(**retriever_kwargs)"
    QD->>R: retriever.query(query)
    R-->>QD: raw_hits [text, source, page_number, _distance, ...]
    Note over QD: Project to {text, source, page_number} only
    QD-->>CLI: filtered_hits
    CLI-->>User: JSON output (no scores/internal fields)

Loading

Comments Outside Diff (1)

1. nemo_retriever/tests/test_root_cli_workflow.py, line 369-370 (link)

   Stale variable reference breaks the test. The variable hits was renamed to raw_hits / public_hits in this PR, but these two assertions still reference the old name. Running the test suite as-is will raise a NameError: name 'hits' is not defined at both lines, so the test always fails.

   Prompt To Fix With AI

   This is a comment left during a code review.
   Path: nemo_retriever/tests/test_root_cli_workflow.py
   Line: 369-370
   
   Comment:
   Stale variable reference breaks the test. The variable `hits` was renamed to `raw_hits` / `public_hits` in this PR, but these two assertions still reference the old name. Running the test suite as-is will raise a `NameError: name 'hits' is not defined` at both lines, so the test always fails.
   
   
   
   How can I resolve this? If you propose a fix, please make it concise.

Prompt To Fix All With AI

Fix the following 1 code review issue. Work through them one at a time, proposing concise fixes.

---

### Issue 1 of 1
nemo_retriever/tests/test_root_cli_workflow.py:369-370
Stale variable reference breaks the test. The variable `hits` was renamed to `raw_hits` / `public_hits` in this PR, but these two assertions still reference the old name. Running the test suite as-is will raise a `NameError: name 'hits' is not defined` at both lines, so the test always fails.

```suggestion
    assert json.loads(result.output) == public_hits
    assert result.output == json.dumps(public_hits, indent=2, sort_keys=True, default=str) + "\n"
```

_{Reviews (4): Last reviewed commit: "Merge branch 'main' into clean-cli-skill..." | Re-trigger Greptile}

[greptile-apps]

The list comprehension on this line is ~130 characters, exceeding the project's enforced 120-character limit (Black + Flake8). The pre-commit hook will reject the file as-is, blocking CI. Black would reformat it to a multi-line style.

Suggested change

	hits = [{"text": hit.get("text", ""), "source": hit.get("source", ""), "page_number": hit.get("page_number")} for hit in hits]
	hits = [
	{
	"text": hit.get("text", ""),
	"source": hit.get("source", ""),
	"page_number": hit.get("page_number"),
	}
	for hit in hits
	]

Prompt To Fix With AI

This is a comment left during a code review.
Path: nemo_retriever/src/nemo_retriever/adapters/cli/sdk_workflow.py
Line: 85

Comment:
The list comprehension on this line is ~130 characters, exceeding the project's enforced 120-character limit (Black + Flake8). The pre-commit hook will reject the file as-is, blocking CI. Black would reformat it to a multi-line style.

```suggestion
    hits = [
        {
            "text": hit.get("text", ""),
            "source": hit.get("source", ""),
            "page_number": hit.get("page_number"),
        }
        for hit in hits
    ]
```

How can I resolve this? If you propose a fix, please make it concise.

[randerzander]

returns are much cleaner, thanks!