Initial pile-config

crates/pile-config/src/config.toml

@@ -0,0 +1,46 @@
+[dataset]
+# This dataset's name. Must be unique.
+name = "dataset"
+
+# root directory for fts indices, relative to the parent dir of this config file.
+# Files are written to {working_dir}/{dataset.name}/*.
+# Default is ".pile"
+#
+# working_dir = ".pile"
+
+# Data sources avaliable in this dataset
+source."music" = { type = "flac", path = ["music", "music-2"] }
+
+
+# This dataset's schema.
+# Defines normalized fields that are extracted from source entries on-demand.
+#
+# Format is as follows:
+#
+# "field-name" = {
+#   # The type of data this field contains.
+#   # only text is supportedin this verison.
+#   type = "text",
+#
+#   # An array of jsonpaths (rfc9535) used to extract this field from each source entry.
+#   # These are evaluated in order, the first non-null value is used.
+#   # A single string is equivalent to an array with one element.
+#   path = "$.json.path"
+# }
+[schema]
+album = { type = "text", path = "$.Album" }
+isrc = { type = "text", path = "$.Isrc" }
+artist = { type = "text", path = ["$.Artist", "$.TrackArtist"] }
+lyrics = { type = "text", path = "$.Lyrics" }
+genre = { type = "text", path = "$.Genre" }
+title = { type = "text", path = ["$.Title", "$.TrackTitle"] }
+
+# Fts configuration.
+# Determines which fields (defined in `schema`) are included in the fts index.
+[fts]
+field.album = { tokenize = true }
+field.isrc = { tokenize = true }
+field.artist = { tokenize = true }
+field.lyrics = { tokenize = true }
+field.genre = { tokenize = true }
+field.title = { tokenize = true }