From 2ba67bc24cf6116349ad16d3bfa1a121c95a3173 Mon Sep 17 00:00:00 2001
From: SteveR-PMP <122415851+SteveR-PMP@users.noreply.github.com>
Date: Thu, 8 Feb 2024 10:39:50 -0800
Subject: [PATCH] feat: default text zero bias (#1330)

A positive value, in milliseconds. It is the threshold used to determine
if we should assume that the text stream actually starts at time zero.
If the first sample comes before default_text_zero_bias_ms, then the
start will be padded as the stream is assumed to start at zero. If the
first sample comes after default_text_zero_bias_ms then the start of the
stream will not be padded as we cannot assume the start time of the
stream.
---
 include/packager/packager.h                  |  4 ++++
 packager/app/muxer_flags.cc                  | 11 +++++++++++
 packager/app/muxer_flags.h                   |  1 +
 packager/app/packager_main.cc                |  2 ++
 packager/media/formats/webvtt/text_padder.cc |  4 +++-
 packager/packager.cc                         |  6 ++----
 6 files changed, 23 insertions(+), 5 deletions(-)

diff --git a/include/packager/packager.h b/include/packager/packager.h
index 6c7ad7a260b..96fd0b07209 100644
--- a/include/packager/packager.h
+++ b/include/packager/packager.h
@@ -46,6 +46,10 @@ struct PackagingParams {
   /// audio) timestamps to compensate for possible negative timestamps in the
   /// input.
   int32_t transport_stream_timestamp_offset_ms = 0;
+  // the threshold used to determine if we should assume that the text stream
+  // actually starts at time zero
+  int32_t default_text_zero_bias_ms = 0;
+
   /// Chunking (segmentation) related parameters.
   ChunkingParams chunking_params;
 
diff --git a/packager/app/muxer_flags.cc b/packager/app/muxer_flags.cc
index bbee0ccf040..2a68b1ee91b 100644
--- a/packager/app/muxer_flags.cc
+++ b/packager/app/muxer_flags.cc
@@ -62,3 +62,14 @@ ABSL_FLAG(int32_t,
           "input. For example, timestamps from ISO-BMFF after adjusted by "
           "EditList could be negative. In transport streams, timestamps are "
           "not allowed to be less than zero.");
+ABSL_FLAG(
+    int32_t,
+    default_text_zero_bias_ms,
+    0,
+    "A positive value, in milliseconds. It is the threshold used to "
+    "determine if we should assume that the text stream actually starts "
+    "at time zero. If the first sample comes before default_text_zero_bias_ms, "
+    "then the start will be padded as the stream is assumed to start at zero. "
+    "If the first sample comes after default_text_zero_bias_ms then the start "
+    "of the stream will not be padded as we cannot assume the start time of "
+    "the stream.");
diff --git a/packager/app/muxer_flags.h b/packager/app/muxer_flags.h
index 20ee2f72b72..a27aff8638a 100644
--- a/packager/app/muxer_flags.h
+++ b/packager/app/muxer_flags.h
@@ -21,5 +21,6 @@ ABSL_DECLARE_FLAG(bool, generate_sidx_in_media_segments);
 ABSL_DECLARE_FLAG(std::string, temp_dir);
 ABSL_DECLARE_FLAG(bool, mp4_include_pssh_in_stream);
 ABSL_DECLARE_FLAG(int32_t, transport_stream_timestamp_offset_ms);
+ABSL_DECLARE_FLAG(int32_t, default_text_zero_bias_ms);
 
 #endif  // APP_MUXER_FLAGS_H_
diff --git a/packager/app/packager_main.cc b/packager/app/packager_main.cc
index 7651a0cd01c..9e12798d60d 100644
--- a/packager/app/packager_main.cc
+++ b/packager/app/packager_main.cc
@@ -461,6 +461,8 @@ std::optional<PackagingParams> GetPackagingParams() {
 
   packaging_params.transport_stream_timestamp_offset_ms =
       absl::GetFlag(FLAGS_transport_stream_timestamp_offset_ms);
+  packaging_params.default_text_zero_bias_ms =
+      absl::GetFlag(FLAGS_default_text_zero_bias_ms);
 
   packaging_params.output_media_info = absl::GetFlag(FLAGS_output_media_info);
 
diff --git a/packager/media/formats/webvtt/text_padder.cc b/packager/media/formats/webvtt/text_padder.cc
index 8bb9cf4c6cb..f93b60200ba 100644
--- a/packager/media/formats/webvtt/text_padder.cc
+++ b/packager/media/formats/webvtt/text_padder.cc
@@ -40,7 +40,9 @@ Status TextPadder::OnTextSample(std::unique_ptr<StreamData> data) {
   // start at time zero.
   if (max_end_time_ms_ < 0) {
     max_end_time_ms_ =
-        sample.start_time() > zero_start_bias_ms_ ? sample.start_time() : 0;
+        zero_start_bias_ms_ && sample.start_time() > zero_start_bias_ms_
+            ? sample.start_time()
+            : 0;
   }
 
   // Check if there will be a gap between samples if we just dispatch this
diff --git a/packager/packager.cc b/packager/packager.cc
index 8f0196322d0..05993f978d8 100644
--- a/packager/packager.cc
+++ b/packager/packager.cc
@@ -59,8 +59,6 @@ namespace {
 
 const char kMediaInfoSuffix[] = ".media_info";
 
-const int64_t kDefaultTextZeroBiasMs = 10 * 60 * 1000;  // 10 minutes
-
 MuxerListenerFactory::StreamData ToMuxerListenerData(
     const StreamDescriptor& stream) {
   MuxerListenerFactory::StreamData data;
@@ -662,8 +660,8 @@ Status CreateAudioVideoJobs(
 
       std::vector<std::shared_ptr<MediaHandler>> handlers;
       if (is_text) {
-        handlers.emplace_back(
-            std::make_shared<TextPadder>(kDefaultTextZeroBiasMs));
+        handlers.emplace_back(std::make_shared<TextPadder>(
+            packaging_params.default_text_zero_bias_ms));
       }
       if (sync_points) {
         handlers.emplace_back(cue_aligner);