Implemented an initial transcript merge stage

This commit is contained in:
2026-04-26 13:57:13 -05:00
parent c32349a017
commit 18f1873776
8 changed files with 535 additions and 28 deletions

View File

@@ -9,7 +9,7 @@ func NewRegistry() *pipeline.Registry {
registry.RegisterInputReader(jsonFilesReader{})
registry.RegisterPreprocessor(noopPreprocessor{name: "validate-raw", requires: pipeline.StateRaw, produces: pipeline.StateRaw})
registry.RegisterPreprocessor(normalizeSpeakers{})
registry.RegisterPreprocessor(noopPreprocessor{name: "trim-text", requires: pipeline.StateCanonical, produces: pipeline.StateCanonical})
registry.RegisterPreprocessor(trimText{})
registry.RegisterPreprocessor(noopPreprocessor{name: "autocorrect", requires: pipeline.StateCanonical, produces: pipeline.StateCanonical})
registry.RegisterMerger(placeholderMerger{})
registry.RegisterPostprocessor(noopPostprocessor{name: "detect-overlaps"})