split include_file_infos into include_file_owner and include_file_group

.chloggen/add_include_file_infos.yaml

@@ -7,7 +7,7 @@ change_type: enhancement
 component: filelogreceiver
 
 # A brief description of the change.  Surround your text with quotes ("") if it needs to start with a backtick (`).
-note: If include_file_infos is true when reading a file on filelogreceiver, it will add the file owner as the attribute `log.file.owner` and the file group as the attribute `log.file.group`.
+note: When reading a file on filelogreceiver, if include_file_owner is true, it will add the file owner as the attribute `log.file.owner` and if include_file_group is true, it will add the file group as the attribute `log.file.group`.
 
 # Mandatory: One or more tracking issues related to the change. You can use the PR number here if no issue exists.
 issues: [30775]

pkg/stanza/docs/operators/file_input.md

@@ -18,6 +18,8 @@ The `file_input` operator reads logs from files. It will place the lines read in
 | `include_file_path`             | `false`          | Whether to add the file path as the attribute `log.file.path`. |
 | `include_file_name_resolved`    | `false`          | Whether to add the file name after symlinks resolution as the attribute `log.file.name_resolved`. |
 | `include_file_path_resolved`    | `false`          | Whether to add the file path after symlinks resolution as the attribute `log.file.path_resolved`. |
+| `include_file_owner`            | `false`          | Whether to add the file owner as the attribute `log.file.owner`  |
+| `include_file_group`            | `false`          | Whether to add the file group as the attribute `log.file.group`  |
 | `preserve_leading_whitespaces`  | `false`          | Whether to preserve leading whitespaces.                                                                                                                                                                                                                         |
 | `preserve_trailing_whitespaces` | `false`          | Whether to preserve trailing whitespaces.                                                                                                                                                                                                                            |
 | `start_at`                      | `end`            | At startup, where to start reading logs from the file. Options are `beginning` or `end`. This setting will be ignored if previously read file offsets are retrieved from a persistence mechanism. |

pkg/stanza/fileconsumer/attrs/attrs.go

@@ -26,7 +26,8 @@ type Resolver struct {
 	IncludeFilePath         bool `mapstructure:"include_file_path,omitempty"`
 	IncludeFileNameResolved bool `mapstructure:"include_file_name_resolved,omitempty"`
 	IncludeFilePathResolved bool `mapstructure:"include_file_path_resolved,omitempty"`
-	IncludeFileInfos        bool `mapstructure:"include_file_infos,omitempty"`
+	IncludeFileOwner        bool `mapstructure:"include_file_owner,omitempty"`
+	IncludeFileGroup        bool `mapstructure:"include_file_group,omitempty"`
 }
 
 func (r *Resolver) Resolve(path string) (attributes map[string]any, err error) {
@@ -38,19 +39,23 @@ func (r *Resolver) Resolve(path string) (attributes map[string]any, err error) {
 	if r.IncludeFilePath {
 		attributes[LogFilePath] = path
 	}
-	if r.IncludeFileInfos {
+	if r.IncludeFileOwner || r.IncludeFileGroup {
 		var file, fileErr = os.OpenFile(fmt.Sprint(path), os.O_RDONLY, 0000)
 		if fileErr == nil {
 			var fileInfo, errStat = file.Stat()
 			if errStat == nil {
 				var fileStat = fileInfo.Sys().(*syscall.Stat_t)
-				var fileOwner, errFileUser = user.LookupId(fmt.Sprint(fileStat.Uid))
-				if errFileUser == nil {
-					attributes[LogFileOwner] = fileOwner.Username
+				if r.IncludeFileOwner {
+					var fileOwner, errFileUser = user.LookupId(fmt.Sprint(fileStat.Uid))
+					if errFileUser == nil {
+						attributes[LogFileOwner] = fileOwner.Username
+					}
 				}
-				var fileGroup, errFileGroup = user.LookupGroupId(fmt.Sprint(fileStat.Gid))
-				if errFileGroup == nil {
-					attributes[LogFileGroup] = fileGroup.Name
+				if r.IncludeFileGroup {
+					var fileGroup, errFileGroup = user.LookupGroupId(fmt.Sprint(fileStat.Gid))
+					if errFileGroup == nil {
+						attributes[LogFileGroup] = fileGroup.Name
+					}
 				}
 			}
 			defer file.Close()

pkg/stanza/fileconsumer/attrs/attrs_test.go

@@ -16,18 +16,19 @@ import (
 func TestResolver(t *testing.T) {
 	t.Parallel()
 
-	for i := 0; i < 32; i++ {
+	for i := 0; i < 64; i++ {
 
 		// Create a 4 bit string where each bit represents the value of a config option
-		bitString := fmt.Sprintf("%05b", i)
+		bitString := fmt.Sprintf("%06b", i)
 
 		// Create a resolver with a config that matches the bit pattern of i
 		r := Resolver{
 			IncludeFileName:         bitString[0] == '1',
 			IncludeFilePath:         bitString[1] == '1',
 			IncludeFileNameResolved: bitString[2] == '1',
 			IncludeFilePathResolved: bitString[3] == '1',
-			IncludeFileInfos:        bitString[4] == '1',
+			IncludeFileOwner:        bitString[4] == '1',
+			IncludeFileGroup:        bitString[5] == '1',
 		}
 
 		t.Run(bitString, func(t *testing.T) {
@@ -68,16 +69,21 @@ func TestResolver(t *testing.T) {
 			} else {
 				assert.Empty(t, attributes[LogFilePathResolved])
 			}
-			if r.IncludeFileInfos {
+			if r.IncludeFileOwner {
 				expectLen++
 				assert.NotNil(t, attributes[LogFileOwner])
 				assert.IsType(t, "", attributes[LogFileOwner])
+			} else {
+				assert.Empty(t, attributes[LogFileOwner])
+				assert.Empty(t, attributes[LogFileOwner])
+			}
+			if r.IncludeFileGroup {
 				expectLen++
 				assert.NotNil(t, attributes[LogFileGroup])
 				assert.IsType(t, "", attributes[LogFileGroup])
 			} else {
-				assert.Empty(t, attributes[LogFileOwner])
-				assert.Empty(t, attributes[LogFileOwner])
+				assert.Empty(t, attributes[LogFileGroup])
+				assert.Empty(t, attributes[LogFileGroup])
 			}
 			assert.Equal(t, expectLen, len(attributes))
 		})

pkg/stanza/fileconsumer/config_test.go

@@ -38,7 +38,8 @@ func TestNewConfig(t *testing.T) {
 	assert.False(t, cfg.IncludeFilePath)
 	assert.False(t, cfg.IncludeFileNameResolved)
 	assert.False(t, cfg.IncludeFilePathResolved)
-	assert.False(t, cfg.IncludeFileInfos)
+	assert.False(t, cfg.IncludeFileOwner)
+	assert.False(t, cfg.IncludeFileGroup)
 }
 
 func TestUnmarshal(t *testing.T) {

pkg/stanza/operator/input/file/file_test.go

@@ -31,7 +31,8 @@ func TestAddFileResolvedFields(t *testing.T) {
 		cfg.IncludeFilePath = true
 		cfg.IncludeFileNameResolved = true
 		cfg.IncludeFilePathResolved = true
-		cfg.IncludeFileInfos = true
+		cfg.IncludeFileOwner = true
+		cfg.IncludeFileGroup = true
 	})
 
 	// Create temp dir with log file

receiver/filelogreceiver/README.md

@@ -33,7 +33,8 @@ Tails and parses logs from files.
 | `include_file_path`                 | `false`                              | Whether to add the file path as the attribute `log.file.path`.                                                                                                                                                                                                  |
 | `include_file_name_resolved`        | `false`                              | Whether to add the file name after symlinks resolution as the attribute `log.file.name_resolved`.                                                                                                                                                               |
 | `include_file_path_resolved`        | `false`                              | Whether to add the file path after symlinks resolution as the attribute `log.file.path_resolved`.                                                                                                                                                               |
-| `include_file_infos`            | `false`          | Whether to add the file owner as the attribute `log.file.owner` and the file group as the attribute `log.file.group`. |
+| `include_file_owner`                | `false`                              | Whether to add the file owner as the attribute `log.file.owner`  |
+| `include_file_group`                | `false`                              | Whether to add the file group as the attribute `log.file.group`  |
 | `poll_interval`                     | 200ms                                | The [duration](#time-parameters) between filesystem polls.                                                                                                                                                                                                      |
 | `fingerprint_size`                  | `1kb`                                | The number of bytes with which to identify a file. The first bytes in the file are used as the fingerprint. Decreasing this value at any point will cause existing fingerprints to forgotten, meaning that all files will be read from the beginning (one time) |
 | `max_log_size`                      | `1MiB`                               | The maximum size of a log entry to read. A log entry will be truncated if it is larger than `max_log_size`. Protects against reading large amounts of data into memory.                                                                                         |

receiver/otlpjsonfilereceiver/testdata/config.yaml

@@ -1,8 +1,8 @@
 otlpjsonfile:
   include:
-  - "/var/log/*.log"
+    - "/var/log/*.log"
   exclude:
-  - "/var/log/example.log"
+    - "/var/log/example.log"
 otlpjsonfile/all:
   include_file_name: true
   include_file_path: true
@@ -17,7 +17,7 @@ otlpjsonfile/all:
     line_start_pattern: "<"
     line_end_pattern: ">"
   include:
-  - "/var/log/*.log"
-  - "/tmp/*.log"
+    - "/var/log/*.log"
+    - "/tmp/*.log"
   exclude:
-  - "/var/log/example.log"
+    - "/var/log/example.log"