feat: support boolean and integer expansion by Zebradil · Pull Request #934 · helmfile/vals

Zebradil · 2026-01-09T22:44:37Z

This PR implements support for boolean and integer values in secret sources.
Previously, only string values were read from secret sources, and non-string
values were skipped, so this should not be a breaking change.

Fixes #190
Related to #492

I'm aware of the idea of having as=int and as=bool URL parameters brought by @mumoshu in #492.
It is not implemented in this PR, but it can be further discussed.

In my opinion, these parameters aren't strictly necessary for the initial goal of the PR to support non-string values in secret sources.
However, such parameters can be useful in case a user needs to convert types. This can also be implemented in a separate PR.

Furthermore, contrary to the approach taken in #492, this PR aims to preserve types instead of just supporting lookups of non-string secret values. This is because I, as a user, do not always control how expanded secret values are used, but I usually control how secret values are stored.

For example, I use vals to inject secret values into a Kubernetes Custom Resource manifest (YAML). Since the official vals supports only strings, I can't inject integer or boolean values. The injection happens on the server-side and the results are not stored but applied immediately, so I don't have access to the manifests with injected secrets. The Custom Resource Definition requires the values to be integers instead of strings. This makes the manifests invalid, so they can't be used.

P.S.: My formatter changed some unrelated code, I'll roll it back if needed.

Zebradil · 2026-01-09T23:08:19Z

pkg/expansion/expand_match.go

+			sb.WriteString(s)
+			// FIXME: this skips the rest of the string, is this intended?
+			return sb.String(), nil


It seems like the code should just skip to the end of the current match.
Something like this:

Suggested change

sb.WriteString(s)

// FIXME: this skips the rest of the string, is this intended?

return sb.String(), nil

sb.WriteString(s[:ixs[1]]) // Write as is from the beginning of the string to the end of the current match

s = s[ixs[1]:]

continue

Copilot

Pull request overview

This pull request adds support for boolean and integer value expansion from secret sources. Previously, only string values could be fetched from secret providers, and non-string values were skipped. The change enables type preservation so that boolean and integer values maintain their types in the output YAML/JSON instead of being converted to strings.

Key Changes:

Modified the Lookup function signature to return interface{} instead of string to support multiple value types
Added isTerminalValue helper to identify acceptable scalar types (bool, int, string)
Introduced InValue function that preserves types for full-match expansions while converting partial matches to strings
Added test coverage for boolean and integer type preservation

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 6 comments.

File	Description
vals.go	Changed Lookup signature from string to interface{}, added isTerminalValue helper, updated type handling in switch statement to include bool/int
pkg/expansion/expand_match.go	Added InValue function for type-preserving expansion, extracted shouldExpand helper, updated InMap to use InValue
pkg/expansion/expand_match_test.go	Updated test signatures to match new Lookup interface{} return type
vals_test.go	Added TestEvalNodesTypes to verify bool/int type preservation; minor formatting improvements (octal literals, var declarations)

Copilot · 2026-01-10T02:25:37Z

pkg/expansion/expand_match.go

+			sb.WriteString(s)
+			// FIXME: this skips the rest of the string, is this intended?
+			return sb.String(), nil


This FIXME comment highlights a potential bug in the existing logic. When a reference kind is not in the expansion list, the function returns immediately, skipping any subsequent references in the same string. This means if you have multiple references in a string and the first one should be skipped, the remaining ones won't be processed either. Consider whether this behavior is intended or if the function should continue processing the rest of the string after skipping non-matching references.

Suggested change

sb.WriteString(s)

// FIXME: this skips the rest of the string, is this intended?

return sb.String(), nil

// Keep this reference as-is and continue processing the rest of the string.

sb.WriteString(s[:ixs[1]])

s = s[ixs[1]:]

continue

Copilot · 2026-01-10T02:25:37Z

pkg/expansion/expand_match.go

+// InValue expands matches in the given string value.
+// If the entire string matches the regex, it expands and preserves the type.
+// If only part of the string matches, it expands as a string.
+func (e *ExpandRegexMatch) InValue(s string) (interface{}, error) {
+	ixs := e.Target.FindStringSubmatchIndex(s)
+	switch {
+	// No match, return as is
+	case ixs == nil:
+		return s, nil
+	// Full match, expand preserving type
+	case ixs[0] == 0 && ixs[1] == len(s):
+		kind := s[ixs[2]:ixs[3]]
+		ref := s[ixs[6]:ixs[7]]
+		if !e.shouldExpand(kind) {
+			return s, nil
+		}
+		val, err := e.Lookup(ref)
+		if err != nil {
+			return nil, fmt.Errorf("expand %s: %v", ref, err)
+		}
+		return val, nil
+	// Partial match, expand as string
+	default:
+		return e.InString(s)
+	}
+}


The new InValue function that enables type preservation for boolean and integer values lacks test coverage. Consider adding tests that verify type preservation for full matches and string conversion for partial matches, including edge cases like booleans, integers, and mixed scenarios.

vals.go

Copilot · 2026-01-10T02:25:38Z

vals.go

 					newobj := map[string]interface{}{}
 					switch t := obj[k].(type) {
-					case string:
+					case bool, int, string:


The type check only handles int, but JSON unmarshaling (used by providers like Scaleway, and potentially returned by other providers) produces float64 for all numeric values. This means integer values from JSON-based secret sources will not be recognized as terminal values and will cause errors. Consider adding support for float64, int64, int32, and other numeric types that might be returned by different providers and unmarshaling libraries.

Suggested change

case bool, int, string:

case bool, int, int32, int64, float32, float64, string:

This suggestion seems reasonable. Should we support more scalar types?

Copilot · 2026-01-10T02:25:38Z

vals.go


+func isTerminalValue(v any) bool {
+	switch v.(type) {
+	case bool, int, string:


The type check only handles int, but JSON unmarshaling (used by providers like Scaleway, and potentially returned by other providers) produces float64 for all numeric values. This means integer values from JSON-based secret sources will not be recognized as terminal values. Consider adding support for float64, int64, int32, and other numeric types that might be returned by different providers and unmarshaling libraries.

Suggested change

case bool, int, string:

case bool,

int, int8, int16, int32, int64,

uint, uint8, uint16, uint32, uint64,

float32, float64,

string:

Copilot · 2026-01-10T02:25:38Z

vals_test.go

+func TestEvalNodesTypes(t *testing.T) {
+	tmpDir := t.TempDir()
+
+	createTmpFile := func(t *testing.T, dir, name, content string) string {
+		tmpFilePath := filepath.Join(dir, name)
+		err := os.WriteFile(tmpFilePath, []byte(content), 0o600)
+		require.NoError(t, err)
+		return tmpFilePath
+	}
+
+	secretYaml := `
+bool: true
+int: 42
+string: "It's a string"
+`
+	secretsFile := createTmpFile(t, tmpDir, "secrets.yaml", secretYaml)
+
+	replacer := strings.NewReplacer("{file-ref}", "ref+file://"+secretsFile)
+	inputYaml := replacer.Replace(`
+bool_value: {file-ref}#/bool
+int_value: {file-ref}#/int
+string_value: {file-ref}#/string
+`)
+	inputFile := createTmpFile(t, tmpDir, "input.yaml", inputYaml)
+
+	expected := `bool_value: true
+int_value: 42
+string_value: It's a string
+`
+
+	input, err := Inputs(inputFile)
+	require.NoError(t, err)
+
+	nodes, err := EvalNodes(input, Options{})
+	require.NoError(t, err)
+	buf := new(strings.Builder)
+
+	err = Output(buf, "", nodes)
+	require.NoError(t, err)
+
+	require.Equal(t, expected, buf.String())
+}


The test only covers YAML-based providers (file provider with .yaml file) which correctly produce int types. However, it doesn't test JSON-based providers (like Scaleway) that produce float64 for numeric values. Consider adding test coverage for JSON-based secret sources to ensure numeric type handling works correctly across all provider types.

yxxhero · 2026-01-27T04:28:54Z

@Zebradil please fix conficts

This PR implements support for boolean and integer values in secret sources. Previously, only string values were read from secret sources, and non-string values were skipped, so this should not be a breaking change. Fixes helmfile#190 Related to helmfile#492 Signed-off-by: German Lashevich <german.lashevich@gmail.com>

Signed-off-by: German Lashevich <german.lashevich@gmail.com>

Zebradil · 2026-01-27T21:14:09Z

@yxxhero done

Zebradil requested review from itscaro, mumoshu, xiaomudk and yxxhero as code owners January 9, 2026 22:44

Zebradil force-pushed the main branch from 38cb43c to f31a02b Compare January 9, 2026 22:45

Zebradil commented Jan 9, 2026

View reviewed changes

Zebradil force-pushed the main branch from f31a02b to 0be7e88 Compare January 9, 2026 23:13

yxxhero requested a review from Copilot January 10, 2026 02:19

Copilot started reviewing on behalf of yxxhero January 10, 2026 02:20 View session

Copilot AI reviewed Jan 10, 2026

View reviewed changes

Zebradil force-pushed the main branch from 4224928 to b9b9c98 Compare January 10, 2026 12:40

Zebradil added 3 commits January 27, 2026 22:09

chore: ignore dist directory created by GoReleaser

405ba4c

Signed-off-by: German Lashevich <german.lashevich@gmail.com>

fix: error out on usupported type

063e5fd

Signed-off-by: German Lashevich <german.lashevich@gmail.com>

Zebradil force-pushed the main branch from b9b9c98 to 6232498 Compare January 27, 2026 21:11

refactor: reuse isTerminalValue function and simlplify code

1afe61c

Signed-off-by: German Lashevich <german.lashevich@gmail.com>

Zebradil force-pushed the main branch from 6232498 to 1afe61c Compare January 27, 2026 21:12

yxxhero merged commit c4c8e9f into helmfile:main Jan 28, 2026
5 checks passed

BrewTestBot mentioned this pull request Feb 9, 2026

vals 0.43.2 Homebrew/homebrew-core#266576

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat: support boolean and integer expansion#934

feat: support boolean and integer expansion#934
yxxhero merged 4 commits intohelmfile:mainfrom
Zebradil:main

Zebradil commented Jan 9, 2026 •

edited

Loading

Uh oh!

Zebradil Jan 9, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Jan 10, 2026

Uh oh!

Copilot AI Jan 10, 2026

Uh oh!

Uh oh!

Copilot AI Jan 10, 2026

Uh oh!

Zebradil Jan 10, 2026

Uh oh!

Copilot AI Jan 10, 2026

Uh oh!

Copilot AI Jan 10, 2026

Uh oh!

yxxhero commented Jan 27, 2026

Uh oh!

Zebradil commented Jan 27, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Comments

-			sb.WriteString(s)
-			// FIXME: this skips the rest of the string, is this intended?
-			return sb.String(), nil
+			// Keep this reference as-is and continue processing the rest of the string.
+			sb.WriteString(s[:ixs[1]])
+			s = s[ixs[1]:]
+			continue

	case bool, int, string:
	case bool, int, int32, int64, float32, float64, string:

-	case bool, int, string:
+	case bool,
+		int, int8, int16, int32, int64,
+		uint, uint8, uint16, uint32, uint64,
+		float32, float64,
+		string:

Conversation

Zebradil commented Jan 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Zebradil Jan 9, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Jan 10, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 10, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Copilot AI Jan 10, 2026

Choose a reason for hiding this comment

Uh oh!

Zebradil Jan 10, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 10, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 10, 2026

Choose a reason for hiding this comment

Uh oh!

yxxhero commented Jan 27, 2026

Uh oh!

Zebradil commented Jan 27, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Comments

Zebradil commented Jan 9, 2026 •

edited

Loading