config: replace file-path settings with inline structs

Today the pii, filter, and summary settings are specified as paths to
separate YAML files:

  pii:    /path/to/pii.yml
  filter: /path/to/filter.yml
  summary: /path/to/summary.yml

This means it is not possible to have a single, combined config file
for an OpenTelemetry collector that uses the trace2receiver. It also
duplicates YAML loading logic that is already built-in using the
${file:PATH} syntax.

Replace the path-string fields (PiiSettingsPath, FilterSettingsPath,
SummaryPath) with inline struct fields (Pii, Filter, Summary) that are
populated directly by mapstructure during config unmarshalling. Users
who prefer separate files can use the ${file:PATH} syntax.

As part of this, rename the previously private Config fields
(piiSettings, filterSettings, summary) to exported names so
mapstructure can populate them, and extract validate() methods on
FilterSettings and SummarySettings so they can be called from
Config.Validate() directly instead of at parse-from-file time.

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: Matthew John Cheetham <mjcheetham@outlook.com>

Author: mjcheetham

config.go

@@ -7,6 +7,11 @@ import (
 	"strings"
 )
 
+// Note: The `pii`, `filter`, and `summary` fields accept inline
+// YAML configuration.  If you prefer to keep the configuration
+// in a separate file, use the `${file:PATH}` syntax to reference
+// an external YAML file.
+
 // `Config` represents the complete configuration settings for
 // an individual receiver declaration from the `config.yaml`.
 //
@@ -45,17 +50,15 @@ type Config struct {
 	// data stream.
 	AllowCommandControlVerbs bool `mapstructure:"enable_commands"`
 
-	// Pathname to YML file containing PII settings.
-	PiiSettingsPath string `mapstructure:"pii"`
-	piiSettings     *PiiSettings
+	// PII settings control whether possibly GDPR-sensitive fields
+	// are included in the telemetry output.
+	Pii *PiiSettings `mapstructure:"pii"`
 
-	// Pathname to YML file containing our filter settings.
-	FilterSettingsPath string `mapstructure:"filter"`
-	filterSettings     *FilterSettings
+	// Filter settings control how the OTLP output is filtered.
+	Filter *FilterSettings `mapstructure:"filter"`
 
-	// Pathname to YML file containing summary settings.
-	SummaryPath string `mapstructure:"summary"`
-	summary     *SummarySettings
+	// Summary settings control aggregated metrics from trace2 events.
+	Summary *SummarySettings `mapstructure:"summary"`
 }
 
 // `Validate()` checks if the receiver configuration is valid.
@@ -101,23 +104,14 @@ func (cfg *Config) Validate() error {
 		cfg.UnixSocketPath = path
 	}
 
-	if len(cfg.PiiSettingsPath) > 0 {
-		cfg.piiSettings, err = parsePiiFile(cfg.PiiSettingsPath)
-		if err != nil {
-			return err
-		}
-	}
-
-	if len(cfg.FilterSettingsPath) > 0 {
-		cfg.filterSettings, err = parseFilterSettings(cfg.FilterSettingsPath)
-		if err != nil {
+	if cfg.Filter != nil {
+		if err = cfg.Filter.validate(); err != nil {
 			return err
 		}
 	}
 
-	if len(cfg.SummaryPath) > 0 {
-		cfg.summary, err = parseSummarySettings(cfg.SummaryPath)
-		if err != nil {
+	if cfg.Summary != nil {
+		if err = cfg.Summary.validate(); err != nil {
 			return err
 		}
 	}

factory.go

@@ -18,10 +18,6 @@ func createDefaultConfig() component.Config {
 		NamedPipePath:            "",
 		UnixSocketPath:           "",
 		AllowCommandControlVerbs: false,
-		PiiSettingsPath:          "",
-		piiSettings:              nil,
-		FilterSettingsPath:       "",
-		filterSettings:           nil,
 	}
 }
 

filter_settings.go

@@ -82,25 +82,32 @@ func parseFilterSettingsFromBuffer(data []byte, path string) (*FilterSettings, e
 		return nil, err
 	}
 
-	// After parsing the YML and populating the `mapstructure` fields, we need
-	// to validate them and/or build internal structures from them.
+	if err = fs.validate(); err != nil {
+		return nil, err
+	}
+
+	return fs, nil
+}
 
-	// For each custom ruleset [<name> -> <path>] in the table (the map[string]string),
-	// create a peer entry in the internal [<name> -> <rsdef>] table and preload
-	// the various `ruleset.yml` files.
+// validate checks the parsed filter settings and builds internal
+// structures.  For each custom ruleset [<name> -> <path>] in the
+// table, create a peer entry in the internal [<name> -> <rsdef>]
+// table and preload the various `ruleset.yml` files.
+func (fs *FilterSettings) validate() error {
 	fs.rulesetDefs = make(map[string]*RulesetDefinition)
 	for k_rs_name, v_rs_path := range fs.Rulesets {
 		if !strings.HasPrefix(k_rs_name, "rs:") || len(k_rs_name) < 4 || len(v_rs_path) == 0 {
-			return nil, fmt.Errorf("ruleset has invalid name or pathname'%s':'%s'", k_rs_name, v_rs_path)
+			return fmt.Errorf("ruleset has invalid name or pathname'%s':'%s'", k_rs_name, v_rs_path)
 		}
 
+		var err error
 		fs.rulesetDefs[k_rs_name], err = parseRulesetFile(v_rs_path)
 		if err != nil {
-			return nil, err
+			return err
 		}
 	}
 
-	return fs, nil
+	return nil
 }
 
 // Add a ruleset to the filter settings.  This is primarily for writing test code.

summary.go

@@ -108,7 +108,7 @@ func apply__summary_message(tr2 *trace2Dataset, message string) {
 		return
 	}
 
-	css := tr2.rcvr_base.RcvrConfig.summary
+	css := tr2.rcvr_base.RcvrConfig.Summary
 	if css == nil {
 		return
 	}
@@ -134,7 +134,7 @@ func apply__summary_region(tr2 *trace2Dataset, region *TrRegion) {
 		return
 	}
 
-	css := tr2.rcvr_base.RcvrConfig.summary
+	css := tr2.rcvr_base.RcvrConfig.Summary
 	if css == nil {
 		return
 	}

summary_settings.go

@@ -60,49 +60,59 @@ func parseSummarySettingsFromBuffer(data []byte, path string) (*SummarySettings,
 		return nil, err
 	}
 
+	if err = css.validate(); err != nil {
+		return nil, err
+	}
+
+	return css, nil
+}
+
+// validate checks the parsed summary settings for errors such as
+// empty required fields and duplicate field names.
+func (css *SummarySettings) validate() error {
 	// Track all field names to detect duplicates
 	fieldNames := make(map[string]bool)
 
 	// Validate message pattern rules
 	for i, rule := range css.MessagePatterns {
 		if len(rule.Prefix) == 0 {
-			return nil, fmt.Errorf("message_patterns[%d]: prefix cannot be empty", i)
+			return fmt.Errorf("message_patterns[%d]: prefix cannot be empty", i)
 		}
 		if len(rule.FieldName) == 0 {
-			return nil, fmt.Errorf("message_patterns[%d]: field_name cannot be empty", i)
+			return fmt.Errorf("message_patterns[%d]: field_name cannot be empty", i)
 		}
 		if fieldNames[rule.FieldName] {
-			return nil, fmt.Errorf("message_patterns[%d]: duplicate field_name '%s'", i, rule.FieldName)
+			return fmt.Errorf("message_patterns[%d]: duplicate field_name '%s'", i, rule.FieldName)
 		}
 		fieldNames[rule.FieldName] = true
 	}
 
 	// Validate region timer rules
 	for i, rule := range css.RegionTimers {
 		if len(rule.Category) == 0 {
-			return nil, fmt.Errorf("region_timers[%d]: category cannot be empty", i)
+			return fmt.Errorf("region_timers[%d]: category cannot be empty", i)
 		}
 		if len(rule.Label) == 0 {
-			return nil, fmt.Errorf("region_timers[%d]: label cannot be empty", i)
+			return fmt.Errorf("region_timers[%d]: label cannot be empty", i)
 		}
 		if len(rule.CountField) == 0 && len(rule.TimeField) == 0 {
-			return nil, fmt.Errorf("region_timers[%d]: at least one of count_field or time_field must be specified", i)
+			return fmt.Errorf("region_timers[%d]: at least one of count_field or time_field must be specified", i)
 		}
 
 		if len(rule.CountField) > 0 {
 			if fieldNames[rule.CountField] {
-				return nil, fmt.Errorf("region_timers[%d]: duplicate field_name '%s'", i, rule.CountField)
+				return fmt.Errorf("region_timers[%d]: duplicate field_name '%s'", i, rule.CountField)
 			}
 			fieldNames[rule.CountField] = true
 		}
 
 		if len(rule.TimeField) > 0 {
 			if fieldNames[rule.TimeField] {
-				return nil, fmt.Errorf("region_timers[%d]: duplicate field_name '%s'", i, rule.TimeField)
+				return fmt.Errorf("region_timers[%d]: duplicate field_name '%s'", i, rule.TimeField)
 			}
 			fieldNames[rule.TimeField] = true
 		}
 	}
 
-	return css, nil
+	return nil
 }