CSV Import Schema

Upload a CSV on Media Catalog → Import → CSV File to create many assets in one go. Each row becomes one asset. Rows are validated individually: invalid rows are skipped and reported back to you, and the valid rows are imported.

At a glance

• One asset per row. The header row is required and must contain at least asset_type, ingest_id, and title.
• Supported asset types in v1: MOVIE, EPISODE, SERIES, SEASON, PODCAST, SHORT, NEWS. CHANNEL, LIVE_EVENT, and EVENT are not supported yet.
• Unknown columns are ignored (reported as warnings on the summary screen).
• List columns (age_ratings, genres, content_groups) use | as the separator — e.g. Action|Drama.
• Upserts by ingest_id. Re-uploading a CSV with the same ingest_id updates the existing asset rather than creating a duplicate.

Columns

Required for every row

Column	Description
asset_type	One of: MOVIE, EPISODE, SERIES, SEASON, PODCAST, SHORT, NEWS.
ingest_id	Your stable external identifier for this asset. Used both for idempotent upserts and as the cross-row reference key for parents.
title	Display title.

Optional metadata

Column	Description
description	Long-form description.
language	ISO language code for title / description. Defaults to en.
production_year	Integer, e.g. 2024.
duration	Integer, seconds.
age_ratings	Pipe-separated list of age-rating codes, e.g. US_PG-13|DE_12.

Parent linkage (for SEASON and EPISODE)

Episodes and seasons reference their parent SERIES/PODCAST by ingest_id — either from another row in the same CSV, or from an asset that already exists in your catalog.

Column	Required for	Description
series_ingest_id	SEASON, EPISODE	ingest_id of the parent series or podcast.
season_number	SEASON, EPISODE	Integer. For an EPISODE, identifies the season within the series.
episode_number	EPISODE	Integer.

Tip

If an EPISODE’s series_ingest_id + season_number doesn’t match any row in this CSV and no matching SEASON exists in your catalog, the row is skipped with that reason. The remainder of the CSV still imports.

Video source

Column	Required for	Description
source_url	MOVIE, EPISODE, NEWS, SHORT	URL to the video. Two kinds are accepted: a file URL (.mp4, .mov, etc.) kicks off the usual transcoding pipeline; an external manifest URL (.m3u8 for HLS or .mpd for DASH) is passed through without transcoding — the asset will play directly from your CDN. Detection is automatic based on the URL extension; no separate column is required.

Taxonomy and curation

Both taxonomy columns accept pipe-separated names (not UUIDs), matched case-sensitively against whatever you’ve configured in the admin panel. Unknown names produce a warning against the row but don’t fail the import — you can fix them up later in the admin panel.

Column	Description
genres	Pipe-separated list of genre names.
content_groups	Pipe-separated list of static content group names. Dynamic content groups auto-populate via their own rules and are not assigned here.

Posters

One URL per column. Images are fetched asynchronously by the poster service after the import completes.

Column	Description
poster_swimlane_16x9_url	Landscape thumbnail used on swimlane rows.
poster_swimlane_2x3_url	Portrait thumbnail used on swimlane rows.
poster_details_16x9_url	Large landscape hero image on the asset details screen.

Processing order

The import runs in three passes so that cross-row references always resolve:

1. Containers — SERIES / PODCAST are created first, then SEASON rows are attached to their series.
2. Leaf assets — MOVIE / EPISODE / NEWS / SHORT rows are upserted. Each one also gets a transcoding (or pass-through for HLS/DASH) job enqueued.
3. Taxonomy + posters — genres, content groups, and poster URLs are attached to every asset created in passes 1 and 2.

You don’t need to pre-sort your CSV — the import re-orders the rows internally.

Examples

Series + seasons + episodes

asset_type,ingest_id,title,description,language,production_year,duration,age_ratings,series_ingest_id,season_number,episode_number,source_url,genres,content_groups,poster_swimlane_16x9_url,poster_swimlane_2x3_url,poster_details_16x9_url
SERIES,bd,Breaking Drama,A gripping drama.,en,2024,,US_TV-14,,,,,Drama,Featured,https://cdn.example.com/bd-swim-16x9.jpg,https://cdn.example.com/bd-swim-2x3.jpg,https://cdn.example.com/bd-hero.jpg
SEASON,bd-s1,Breaking Drama: Season 1,,en,,,,bd,1,,,,,,,
EPISODE,bd-s1e1,Pilot,The one where it starts.,en,2024,2700,US_TV-14,bd,1,1,https://cdn.example.com/bd/s1e1.mp4,Drama,,,,
EPISODE,bd-s1e2,Complications,,en,2024,2640,US_TV-14,bd,1,2,https://cdn.example.com/bd/s1e2.m3u8,Drama,,,,

Row bd-s1e2 uses an HLS manifest — Vidori will play that back directly from your CDN without transcoding. Row bd-s1e1 uses an MP4 file and will go through the normal transcode pipeline.

Flat movie catalog

asset_type,ingest_id,title,description,language,production_year,duration,age_ratings,series_ingest_id,season_number,episode_number,source_url,genres,content_groups,poster_swimlane_16x9_url,poster_swimlane_2x3_url,poster_details_16x9_url
MOVIE,mov-001,The Quiet Harbour,,en,2022,6420,,,,,https://cdn.example.com/quiet.mp4,Drama|Documentary,Featured,,,
MOVIE,mov-002,Midnight in Lagos,,en,2023,5400,US_R,,,,https://cdn.example.com/midnight.mpd,Drama,Featured|New,,,
MOVIE,mov-003,Hello Again,,en,2021,5760,US_PG,,,,https://cdn.example.com/hello.m3u8,Comedy,,,,

Row mov-002 uses DASH (.mpd) and row mov-003 uses HLS (.m3u8); both skip transcoding. Row mov-001 is a plain MP4 and runs the full transcode pipeline.

Error handling

The import always returns a per-row summary. Common skip reasons:

Reason	What it means
ingest_id is required / title is required	A required column is empty on that row.
unknown asset_type '…'	The value in asset_type doesn’t match any supported type.
asset_type '…' is not supported in v1 CSV import	You used CHANNEL, LIVE_EVENT, or EVENT — not supported yet.
MOVIE row requires source_url	A leaf row is missing its video source.
parent series with ingest_id '…' not found	The series an EPISODE/SEASON points to isn’t in the CSV or the catalog.
duplicate ingest_id '…'	Two rows in the same CSV share an ingest_id. Only the first is kept.

Warnings (import succeeds, but something was not applied):

Warning	What it means
unknown genre '…' — skipped	The genre name doesn’t match anything in your tenant. Create the genre first, then re-upload.
unknown content_group '…' — skipped	Same idea, for content groups.
Unknown columns ignored	The CSV has columns that aren’t part of this schema. Safe to ignore, useful to catch typos.