Behaviour and default implementation for extracting file metadata.

PhxMediaLibrary automatically extracts metadata from uploaded files and stores it in the metadata field of the Media schema. This includes image dimensions, video/audio duration, EXIF data, and more.

Extracted Metadata

The metadata map may contain the following keys depending on file type:

Images

• "width" — image width in pixels
• "height" — image height in pixels
• "format" — detected image format (e.g. "jpeg", "png", "webp")
• "exif" — EXIF data map (orientation, camera make/model, GPS, etc.)
• "has_alpha" — whether the image has an alpha channel

Video

• "width" — video width in pixels
• "height" — video height in pixels
• "duration" — duration in seconds (float)
• "format" — container format (e.g. "mp4", "webm")

Audio

• "duration" — duration in seconds (float)
• "format" — audio format (e.g. "mp3", "wav", "flac")

All files

• "extracted_at" — ISO 8601 timestamp of when metadata was extracted

Custom Implementation

You can provide your own extractor by implementing the behaviour and configuring it:

defmodule MyApp.MetadataExtractor do
  @behaviour PhxMediaLibrary.MetadataExtractor

  @impl true
  def extract(file_path, mime_type, _opts) do
    # Your custom extraction logic
    {:ok, %{"custom_field" => "value"}}
  end
end

config :phx_media_library,
  metadata_extractor: MyApp.MetadataExtractor

Disabling Extraction

To disable automatic metadata extraction:

config :phx_media_library,
  extract_metadata: false

Or per-call:

post
|> PhxMediaLibrary.add(upload)
|> PhxMediaLibrary.without_metadata()
|> PhxMediaLibrary.to_collection(:images)