1 files changed, 85 insertions, 0 deletions
diff --git a/sorsyl/lib/ipa_table.ml b/sorsyl/lib/ipa_table.ml
new file mode 100644
index 0000000..bee027a
--- /dev/null
+++ b/sorsyl/lib/ipa_table.ml
@@ -0,0 +1,85 @@
+(** Type representing a segment as a set of feature specifications *)
+(* an association list I guess . Use List.assoc to handle*)
+
+(** Decision tree for computing sonority values *)
+type bool_tree =
+  | Leaf of int  (** Terminal node with sonority value *)
+  | Node of {
+      test : Feature.segment -> bool;  (** Test function *)
+      t_branch : bool_tree;  (** Branch to follow if test is true *)
+      f_branch : bool_tree;  (** Branch to follow if test is false *)
+    }
+
+type ipa_entry = { ipa : string; features : Feature.segment }
+(** Type representing a row from the IPA CSV file *)
+
+(** Storage for loaded IPA data *)
+let ipa_table : (string, Feature.segment) Hashtbl.t = Hashtbl.create 1000
+
+(** Parse a single row from the CSV file *)
+let parse_row (row : string list) : ipa_entry option =
+  match row with
+  | [] -> None
+  | ipa :: features -> (
+      let feature_names =
+        [
+          "syl";
+          "son";
+          "cons";
+          "cont";
+          "delrel";
+          "lat";
+          "nas";
+          "strid";
+          "voi";
+          "sg";
+          "cg";
+          "ant";
+          "cor";
+          "distr";
+          "lab";
+          "hi";
+          "lo";
+          "back";
+          "round";
+          "velaric";
+          "tense";
+          "long";
+          "hitone";
+          "hireg";
+        ]
+      in
+      (* Skip the header row *)
+      if ipa = "ipa" then None
+      else
+        let rec build_features names values acc =
+          match (names, values) with
+          | [], [] -> Some (List.rev acc)
+          | name :: ns, value :: vs ->
+              let fval = Feature.value_of_string value in
+              let fname = Feature.string_of_feature name in
+              build_features ns vs ((fname, fval) :: acc)
+          | _ -> None (* Mismatched lengths *)
+        in
+        match build_features feature_names features [] with
+        | Some feature_list -> Some { ipa; features = feature_list }
+        | None -> None)
+
+(** Load IPA data from CSV file *)
+let load_csv filename =
+  let ic = open_in filename in
+  let csv = Csv.of_channel ic in
+  try
+    Csv.iter
+      ~f:(fun row ->
+        match parse_row row with
+        | Some entry -> Hashtbl.add ipa_table entry.ipa entry.features
+        | None -> ())
+      csv;
+    close_in ic
+  with e ->
+    close_in ic;
+    raise e
+
+(** Look up features for an IPA segment *)
+let lookup_segment ipa = Hashtbl.find_opt ipa_table ipa