Index Strategy (`root_strategy: "index"`)

A single delimited index file (CSV / TSV / pipe) lists one row per sample, e.g. mapping audio paths to transcriptions and other metadata columns.

When the schema declares columns, the loader applies the column mappings (renaming, dtype conversion, file-path resolution). When columns is omitted, the raw index file is returned as-is.

Controlled by:

Field	Required	Description
`root_strategy`	✓	Must be `"index"`.
`index_file`	✓	Path to the index file, relative to the dataset root.
`columns`	✗	Mapping of logical column names to source columns and dtypes. When omitted, the raw index file is returned unchanged.
`base_audio_path`	✗	Directory prefix or list of directories used to resolve `file_path`/`file_content` dtype columns.
`format`	✗	Optional file format hint (`"csv"`, `"tsv"`, `"pipe"`). When omitted, the loader infers it from `index_file` where possible.
`separator`	✗	Explicit column separator (e.g. `"\\|"`).
`has_header`	✗	Whether the index file has a header row. When `false`, `source_column` must be a positional integer.
`encoding`	✗	File encoding (e.g. `"utf-8-sig"` for files with a BOM).
`quoting`	✗	`"minimal"` (CSV-style quoted fields) or `"none"` (`"` is literal). Defaults to `"none"` for tab-separated files and `"minimal"` otherwise; see Quoting in delimited files.
`strict`	✗	Disable archive heuristics: `index_file` must exist at its literal relative path (no recursive search), exact column-name matching only.

Examples

Basic index schema (ASR)

dataset_id: "cmj8u48g4005lnxzp98cpr7b2"
task: "ASR"
root_strategy: "index"
format: "tsv"

index_file: "ss-corpus-shi.tsv"
base_audio_path: "audios/"

columns:
  audio_path:
    source_column: "audio_file"
    dtype: "file_path"
  transcription:
    source_column: "transcription"
    dtype: "string"
  speaker_id:
    source_column: "client_id"
    dtype: "category"
    optional: true
  audio_id:
    source_column: "audio_id"
    dtype: "string"
    optional: true
  duration_ms:
    source_column: "duration_ms"
    dtype: "int"
    optional: true
  prompt_id:
    source_column: "prompt_id"
    dtype: "string"
    optional: true
  prompt:
    source_column: "prompt"
    dtype: "string"
    optional: true
  votes:
    source_column: "votes"
    dtype: "int"
    optional: true
  age:
    source_column: "age"
    dtype: "category"
    optional: true
  gender:
    source_column: "gender"
    dtype: "category"
    optional: true
  language:
    source_column: "language"
    dtype: "category"
    optional: true
  split:
    source_column: "split"
    dtype: "category"
    optional: true
  char_per_sec:
    source_column: "char_per_sec"
    dtype: "float"
    optional: true
  quality_tags:
    source_column: "quality_tags"
    dtype: "string"
    optional: true

Pipe-delimited, headerless metadata (TTS)

dataset_id: "aso-ckb-tts"
task: "TTS"
root_strategy: "index"
format: "pipe"
separator: "|"
has_header: false
index_file: "metadata.csv"
base_audio_path: "wavs/"
columns:
  audio_path:
    source_column: 0        # positional index (no header)
    dtype: "file_path"
  transcription:
    source_column: 1
    dtype: "string"

Search-based audio resolution

When the metadata stores an ID or partial filename instead of a directly joinable relative path, file_path columns can search within one or more audio roots:

dataset_id: "example-asr"
task: "ASR"
root_strategy: "index"
index_file: "data/metadata.csv"
base_audio_path:
  - "data/recipes/"
  - "data/giving_gift/"

columns:
  audio_path:
    source_column: "Sentence ID"
    dtype: "file_path"
    path_match_strategy: "exact"   # or "contains"
    file_extension: ".wav"
  transcription:
    source_column: "Sentences"
    dtype: "string"

path_match_strategy: "direct" remains the default and preserves the existing extract_dir / base_audio_path / value behavior. The loader also trims BOMs and surrounding header whitespace. The separator is never guessed from file contents — when the declared/inferred separator is wrong, loading fails with the raw (single-column) headers listed; declare separator explicitly.

If the true audio filename is composed from multiple metadata columns, use path_template instead of a fuzzy search:

dataset_id: "khmer-asr-cultural-dataset-4e33cd05"
task: "ASR"
root_strategy: "index"
index_file: "data/metadata.csv"
base_audio_path:
  - "data/recipes/"
  - "data/giving_gift/"

columns:
  audio_path:
    source_column: "Sentence ID"
    dtype: "file_path"
    file_extension: ".wav"
    path_template: "${Speaker ID}_khm_${Sentence ID}.wav"
  transcription:
    source_column: "Sentences"
    dtype: "string"

Template placeholders reference raw metadata column names exactly, and ${value} refers to the current source_column value. Relative paths are resolved from the dataset root inferred from the resolved index_file.

If the audio directory itself varies per row, base_audio_path can use the same placeholder syntax:

dataset_id: "khmer-asr-cultural-dataset-4e33cd05"
task: "ASR"
root_strategy: "index"
index_file: "data/metadata.csv"
base_audio_path: "data/${Split}/"

columns:
  audio_path:
    source_column: "Sentence ID"
    dtype: "file_path"
    file_extension: ".wav"
    path_template: "${Speaker ID}_khm_${value}"
  transcription:
    source_column: "Sentences"
    dtype: "string"

That resolves each row as dataset_root / data/<Split>/<Speaker ID>_khm_<Sentence ID>.wav.

File-content dtype

When the index file stores paths to transcription files instead of inline text, use dtype: "file_content" to read the file contents into the DataFrame:

dataset_id: "speech-data-nupe"
task: "ASR"
root_strategy: "index"
index_file: "Metadata.csv"
base_audio_path:
  - "Speaker_id_1"
  - "Speaker_id_2"

columns:
  audio_path:
    source_column: "Audio_File_Path"
    dtype: "file_path"
    file_extension: ".wav"
  transcription:
    source_column: "Transcript_File_Path"
    dtype: "file_content"
    file_extension: ".txt"
  speaker_id:
    source_column: "Speaker_ID"

The file_content dtype reuses the same path resolution as file_path (base_audio_path, file_extension, path_match_strategy, path_template) but returns the file's text content instead of the resolved path.

Raw index file (no column mappings)

A dataset described by a single delimited index file, returned as-is. Note that a schema without columns should only declare a task whose raw columns already satisfy the task contract (or a task without one, e.g. OTH) — otherwise every load emits a TaskValidationWarning.

dataset_id: "common-voice-text-langid"
task: "OTH"
root_strategy: "index"
format: "tsv"
index_file: "data.tsv"

columns:
  sentence:
    source_column: "sentence"
    dtype: "string"
  language:
    source_column: "lang"
    dtype: "category"

When columns is omitted, the index file is returned as a raw DataFrame with its original columns.

Index Strategy (root_strategy: "index")