Custom Log Sinks

turbo::LogSink is an extension point for processing logged messages (e.g., by writing them to disk files). A message can be directed to it using the .ToSinkOnly() or .ToSinkAlso() mutator methods, or a sink can be registered to observe all logged messages (except those using .ToSinkOnly()) with turbo::add_log_sink() and unregistered with turbo::remove_log_sink. For example:

Basic Interface

class LinePrinterLogSink : public turbo::LogSink {
 public:
  LinePrinterLogSink() : fp_(fopen("/dev/lp0", "a")) {
    PKCHECK(fp_ != nullptr) << "Failed to open /dev/lp0";
  }
  ~LinePrinterLogSink() {
    fputc('\f', fp_);
    PKCHECK(fclose(fp_) == 0) << "Failed to close /dev/lp0";
  }
  void Send(const turbo::LogEntry& entry) override {
    for (std::string_view line :
         turbo::sr_split(entry.text_message_with_prefix(), turbo::ByChar('\n'))) {
      // Overprint severe entries for emphasis:
      for (int i = static_cast<int>(turbo::LogSeverity::kInfo);
           i <= static_cast<int>(entry.log_severity()); i++) {
        turbo::FPrintF(fp_, "%s\r", line);
      }
      fputc('\n', fp_);
    }
  }

 private:
  FILE* const fp_;
};

A LogSink receives two copies of each FATAL message: one without a stacktrace, followed by one with it. This quirk ensures that some diagnostic data can be observed even if stack trace collection fails or takes too long. The process will terminate immediately after the LogSink returns—meaning the sink does not need to call abort().

Any logging that occurs within a registered LogSink (or functions called by a registered LogSink) is sent only to stderr, not to any registered LogSinks, to avoid infinite recursion.