// readLine reads a full line into buffer and returns it.
// In case of partial lines, readLine does return and error and en empty string
// This could potentialy be improved / replaced by https://github.com/elastic/beats/libbeat/tree/master/common/streambuf
func readLine(reader reader.Reader) (time.Time, string, int, common.MapStr, error) {
l, err := reader.Next()
// Full line read to be returned
if l.Bytes != 0 && err == nil {
return l.Ts, string(l.Content), l.Bytes, l.Fields, err
}
return time.Time{}, "", 0, nil, err
}
// Harvest reads files line by line and sends events to the defined output
func (h *Harvester) Harvest(r reader.Reader) {
harvesterStarted.Add(1)
harvesterRunning.Add(1)
defer harvesterRunning.Add(-1)
// Makes sure file is properly closed when the harvester is stopped
defer h.close()
// Channel to stop internal harvester routines
harvestDone := make(chan struct{})
defer close(harvestDone)
// Closes reader after timeout or when done channel is closed
// This routine is also responsible to properly stop the reader
go func() {
var closeTimeout <-chan time.Time
if h.config.CloseTimeout > 0 {
closeTimeout = time.After(h.config.CloseTimeout)
}
select {
// Applies when timeout is reached
case <-closeTimeout:
logp.Info("Closing harvester because close_timeout was reached: %s", h.state.Source)
// Required for shutdown when hanging inside reader
case <-h.done:
// Required when reader loop returns and reader finished
case <-harvestDone:
}
h.fileReader.Close()
}()
logp.Info("Harvester started for file: %s", h.state.Source)
for {
select {
case <-h.done:
return
default:
}
message, err := r.Next()
if err != nil {
switch err {
case ErrFileTruncate:
logp.Info("File was truncated. Begin reading file from offset 0: %s", h.state.Source)
h.state.Offset = 0
filesTruncated.Add(1)
case ErrRemoved:
logp.Info("File was removed: %s. Closing because close_removed is enabled.", h.state.Source)
case ErrRenamed:
logp.Info("File was renamed: %s. Closing because close_renamed is enabled.", h.state.Source)
case ErrClosed:
logp.Info("Reader was closed: %s. Closing.", h.state.Source)
case io.EOF:
logp.Info("End of file reached: %s. Closing because close_eof is enabled.", h.state.Source)
case ErrInactive:
logp.Info("File is inactive: %s. Closing because close_inactive of %v reached.", h.state.Source, h.config.CloseInactive)
default:
logp.Err("Read line error: %s; File: ", err, h.state.Source)
}
return
}
// Strip UTF-8 BOM if beginning of file
// As all BOMS are converted to UTF-8 it is enough to only remove this one
if h.state.Offset == 0 {
message.Content = bytes.Trim(message.Content, "\xef\xbb\xbf")
}
// Update offset
h.state.Offset += int64(message.Bytes)
// Create state event
event := input.NewEvent(h.getState())
text := string(message.Content)
// Check if data should be added to event. Only export non empty events.
if !message.IsEmpty() && h.shouldExportLine(text) {
event.ReadTime = message.Ts
event.Bytes = message.Bytes
event.Text = &text
event.JSONFields = message.Fields
event.EventMetadata = h.config.EventMetadata
event.InputType = h.config.InputType
event.DocumentType = h.config.DocumentType
event.JSONConfig = h.config.JSON
}
// Always send event to update state, also if lines was skipped
// Stop harvester in case of an error
if !h.sendEvent(event) {
return
}
}
}
// Harvest reads files line by line and sends events to the defined output
func (h *Harvester) Harvest(r reader.Reader) {
harvesterStarted.Add(1)
harvesterRunning.Add(1)
defer harvesterRunning.Add(-1)
// Makes sure file is properly closed when the harvester is stopped
defer h.close()
logp.Info("Harvester started for file: %s", h.state.Source)
for {
select {
case <-h.done:
return
default:
}
message, err := r.Next()
if err != nil {
switch err {
case ErrFileTruncate:
logp.Info("File was truncated. Begin reading file from offset 0: %s", h.state.Source)
h.state.Offset = 0
filesTruncated.Add(1)
case ErrRemoved:
logp.Info("File was removed: %s. Closing because close_removed is enabled.", h.state.Source)
case ErrRenamed:
logp.Info("File was renamed: %s. Closing because close_renamed is enabled.", h.state.Source)
case ErrClosed:
logp.Info("Reader was closed: %s. Closing.", h.state.Source)
case io.EOF:
logp.Info("End of file reached: %s. Closing because close_eof is enabled.", h.state.Source)
case ErrInactive:
logp.Info("File is inactive: %s. Closing because close_inactive of %v reached.", h.state.Source, h.config.CloseInactive)
default:
logp.Err("Read line error: %s; File: ", err, h.state.Source)
}
return
}
// Strip UTF-8 BOM if beginning of file
// As all BOMS are converted to UTF-8 it is enough to only remove this one
if h.state.Offset == 0 {
message.Content = bytes.Trim(message.Content, "\xef\xbb\xbf")
}
// Update offset
h.state.Offset += int64(message.Bytes)
// Create state event
event := input.NewEvent(h.getState())
text := string(message.Content)
// Check if data should be added to event. Only export non empty events.
if !message.IsEmpty() && h.shouldExportLine(text) {
event.ReadTime = message.Ts
event.Bytes = message.Bytes
event.Text = &text
event.JSONFields = message.Fields
event.EventMetadata = h.config.EventMetadata
event.InputType = h.config.InputType
event.DocumentType = h.config.DocumentType
event.JSONConfig = h.config.JSON
}
// Always send event to update state, also if lines was skipped
// Stop harvester in case of an error
if !h.sendEvent(event) {
return
}
}
}
