Merge pull request #436 from pangeachat/pos-morph

Pos-morph

lib/pangea/controllers/language_list_controller.dart

@@ -93,7 +93,6 @@ class PangeaLanguage {
   }
 
   static LanguageModel byLangCode(String langCode) {
-    final list = _langList;
     for (final element in _langList) {
       if (element.langCode == langCode) return element;
     }

lib/pangea/enum/activity_display_instructions_enum.dart

@@ -0,0 +1,13 @@
+enum ActivityDisplayInstructionsEnum { highlight, hide }
+
+extension ActivityDisplayInstructionsEnumExt
+    on ActivityDisplayInstructionsEnum {
+  String get string {
+    switch (this) {
+      case ActivityDisplayInstructionsEnum.highlight:
+        return 'highlight';
+      case ActivityDisplayInstructionsEnum.hide:
+        return 'hide';
+    }
+  }
+}

lib/pangea/matrix_event_wrappers/pangea_message_event.dart

@@ -687,7 +687,8 @@ class PangeaMessageEvent {
 
     for (final itStep in originalSent!.choreo!.itSteps) {
       for (final continuance in itStep.continuances) {
-        // this seems to always be false for continuances right now
+        final List<PangeaToken> tokensToSave =
+            continuance.tokens.where((t) => t.lemma.saveVocab).toList();
 
         if (originalSent!.choreo!.finalMessage.contains(continuance.text)) {
           continue;
@@ -695,18 +696,21 @@ class PangeaMessageEvent {
         if (continuance.wasClicked) {
           //PTODO - account for end of flow score
           if (continuance.level != ChoreoConstants.levelThresholdForGreen) {
-            uses.addAll(
+            for (final token in tokensToSave) {
-              _lemmasToVocabUses(
+              uses.add(
-                continuance.lemmas,
+                _lemmaToVocabUse(
+                  token.lemma,
                   ConstructUseTypeEnum.incIt,
                 ),
               );
             }
+          }
         } else {
           if (continuance.level != ChoreoConstants.levelThresholdForGreen) {
-            uses.addAll(
+            for (final token in tokensToSave) {
-              _lemmasToVocabUses(
+              uses.add(
-                continuance.lemmas,
+                _lemmaToVocabUse(
+                  token.lemma,
                   ConstructUseTypeEnum.ignIt,
                 ),
               );
@@ -714,6 +718,7 @@ class PangeaMessageEvent {
           }
         }
       }
+    }
     return uses;
   }
 
@@ -728,14 +733,16 @@ class PangeaMessageEvent {
     }
 
     // for each token, record whether selected in ga, ta, or wa
-    for (final token in originalSent!.tokens!) {
+    for (final token in originalSent!.tokens!
-      uses.addAll(_getVocabUseForToken(token));
+        .where((token) => token.lemma.saveVocab)
+        .toList()) {
+      uses.add(_getVocabUseForToken(token));
     }
 
     return uses;
   }
 
-  /// Returns a list of [OneConstructUse] objects for the given [token]
+  /// Returns a [OneConstructUse] for the given [token]
   /// If there is no [originalSent] or [originalSent.choreo], the [token] is
   /// considered to be a [ConstructUseTypeEnum.wa] as long as it matches the target language.
   /// Later on, we may want to consider putting it in some category of like 'pending'
@@ -744,11 +751,11 @@ class PangeaMessageEvent {
   /// If the [token] is in the [originalSent.choreo.acceptedOrIgnoredMatch.choices],
   /// it is considered to be a [ConstructUseTypeEnum.corIt].
   /// If the [token] is not included in any choreoStep, it is considered to be a [ConstructUseTypeEnum.wa].
-  List<OneConstructUse> _getVocabUseForToken(PangeaToken token) {
+  OneConstructUse _getVocabUseForToken(PangeaToken token) {
     if (originalSent?.choreo == null) {
       final bool inUserL2 = originalSent?.langCode == l2Code;
-      return _lemmasToVocabUses(
+      return _lemmaToVocabUse(
-        token.lemmas,
+        token.lemma,
         inUserL2 ? ConstructUseTypeEnum.wa : ConstructUseTypeEnum.unk,
       );
     }
@@ -763,31 +770,25 @@ class PangeaMessageEvent {
                     step.text.contains(r.value),
               ) ??
               false)) {
-        return _lemmasToVocabUses(token.lemmas, ConstructUseTypeEnum.ga);
+        return _lemmaToVocabUse(token.lemma, ConstructUseTypeEnum.ga);
       }
       if (step.itStep != null) {
         final bool pickedThroughIT =
             step.itStep!.chosenContinuance?.text.contains(token.text.content) ??
                 false;
         if (pickedThroughIT) {
-          return _lemmasToVocabUses(token.lemmas, ConstructUseTypeEnum.corIt);
+          return _lemmaToVocabUse(token.lemma, ConstructUseTypeEnum.corIt);
           //PTODO - check if added via custom input in IT flow
         }
       }
     }
-    return _lemmasToVocabUses(token.lemmas, ConstructUseTypeEnum.wa);
+    return _lemmaToVocabUse(token.lemma, ConstructUseTypeEnum.wa);
   }
 
-  /// Convert a list of [lemmas] into a list of vocab uses
+  OneConstructUse _lemmaToVocabUse(
-  /// with the given [type]
+    Lemma lemma,
-  List<OneConstructUse> _lemmasToVocabUses(
-    List<Lemma> lemmas,
     ConstructUseTypeEnum type,
-  ) {
+  ) =>
-    final List<OneConstructUse> uses = [];
-    for (final lemma in lemmas) {
-      if (lemma.saveVocab) {
-        uses.add(
       OneConstructUse(
         useType: type,
         chatId: event.roomId!,
@@ -796,12 +797,7 @@ class PangeaMessageEvent {
         form: lemma.form,
         msgId: event.eventId,
         constructType: ConstructTypeEnum.vocab,
-          ),
       );
-      }
-    }
-    return uses;
-  }
 
   /// get construct uses of type grammar for the message
   List<OneConstructUse> get _grammarConstructUses {

lib/pangea/models/it_response_model.dart

@@ -2,11 +2,10 @@ import 'package:collection/collection.dart';
 import 'package:fluffychat/pangea/constants/choreo_constants.dart';
 import 'package:fluffychat/pangea/constants/model_keys.dart';
 import 'package:fluffychat/pangea/extensions/my_list_extension.dart';
+import 'package:fluffychat/pangea/models/pangea_token_model.dart';
 import 'package:flutter/material.dart';
 import 'package:flutter_gen/gen_l10n/l10n.dart';
 
-import 'lemma.dart';
-
 class ITResponseModel {
   String fullTextTranslation;
   List<Continuance> continuances;
@@ -79,7 +78,7 @@ class Continuance {
   double probability;
   int level;
   String text;
-  List<Lemma> lemmas;
+  List<PangeaToken> tokens;
 
   /// saving this in a full json form
   String description;
@@ -99,18 +98,17 @@ class Continuance {
     required this.inDictionary,
     required this.hasInfo,
     required this.gold,
-    required this.lemmas,
+    required this.tokens,
   });
 
   factory Continuance.fromJson(Map<String, dynamic> json) {
-    final List<Lemma> lemmaInternal =
+    final List<PangeaToken> tokensInternal = (json[ModelKey.tokens] != null)
-        (json[ModelKey.lemma] != null && json[ModelKey.lemma] is Iterable)
+        ? (json[ModelKey.tokens] as Iterable)
-            ? (json[ModelKey.lemma] as Iterable)
+            .map<PangeaToken>(
-                .map<Lemma>(
+              (e) => PangeaToken.fromJson(e as Map<String, dynamic>),
-                  (e) => Lemma.fromJson(e as Map<String, dynamic>),
             )
             .toList()
-                .cast<Lemma>()
+            .cast<PangeaToken>()
         : [];
     return Continuance(
       probability: json['probability'].toDouble(),
@@ -122,7 +120,7 @@ class Continuance {
       wasClicked: json['clkd'] ?? false,
       hasInfo: json['has_info'] ?? false,
       gold: json['gold'] ?? false,
-      lemmas: lemmaInternal,
+      tokens: tokensInternal,
     );
   }
 
@@ -132,7 +130,7 @@ class Continuance {
     data['level'] = level;
     data['text'] = text;
     data['clkd'] = wasClicked;
-    data[ModelKey.lemma] = lemmas.map((e) => e.toJson()).toList();
+    data[ModelKey.tokens] = tokens.map((e) => e.toJson()).toList();
 
     if (!condensed) {
       data['description'] = description;

lib/pangea/models/lemma.dart

@@ -8,22 +8,13 @@ class Lemma {
 
   /// [saveVocab] true - whether to save the lemma to the user's vocabulary
   /// vocab that are not saved: emails, urls, numbers, punctuation, etc.
+  /// server handles this determination
   final bool saveVocab;
 
-  /// [pos] ex "v" - part of speech of the lemma
-  /// https://universaldependencies.org/u/pos/
-  final String pos;
-
-  /// [morph] ex {} - morphological features of the lemma
-  /// https://universaldependencies.org/u/feat/
-  final Map<String, dynamic> morph;
-
   Lemma({
     required this.text,
     required this.saveVocab,
     required this.form,
-    this.pos = '',
-    this.morph = const {},
   });
 
   factory Lemma.fromJson(Map<String, dynamic> json) {
@@ -31,8 +22,6 @@ class Lemma {
       text: json['text'],
       saveVocab: json['save_vocab'] ?? json['saveVocab'] ?? false,
       form: json["form"] ?? json['text'],
-      pos: json['pos'] ?? '',
-      morph: json['morph'] ?? {},
     );
   }
 
@@ -41,8 +30,6 @@ class Lemma {
       'text': text,
       'save_vocab': saveVocab,
       'form': form,
-      'pos': pos,
-      'morph': morph,
     };
   }
 

lib/pangea/models/pangea_token_model.dart

@@ -1,55 +1,60 @@
 import 'dart:developer';
 
 import 'package:flutter/foundation.dart';
-import 'package:sentry_flutter/sentry_flutter.dart';
 
 import '../constants/model_keys.dart';
-import '../utils/error_handler.dart';
 import 'lemma.dart';
 
 class PangeaToken {
   PangeaTokenText text;
-  List<Lemma> lemmas;
+  Lemma lemma;
+
+  /// [pos] ex "VERB" - part of speech of the token
+  /// https://universaldependencies.org/u/pos/
+  final String pos;
+
+  /// [morph] ex {} - morphological features of the token
+  /// https://universaldependencies.org/u/feat/
+  final Map<String, dynamic> morph;
 
   PangeaToken({
     required this.text,
-    required this.lemmas,
+    required this.lemma,
+    required this.pos,
+    required this.morph,
   });
 
-  static getLemmas(String text, Iterable? json) {
+  static Lemma _getLemmas(String text, dynamic json) {
     if (json != null) {
+      // July 24, 2024 - we're changing from a list to a single lemma and this is for backwards compatibility
+      // previously sent tokens have lists of lemmas
+      if (json is Iterable) {
         return json
                 .map<Lemma>(
                   (e) => Lemma.fromJson(e as Map<String, dynamic>),
                 )
                 .toList()
-          .cast<Lemma>();
+                .cast<Lemma>()
+                .firstOrNull ??
+            Lemma(text: text, saveVocab: false, form: text);
+      } else {
+        return Lemma.fromJson(json);
+      }
     } else {
-      return [Lemma(text: text, saveVocab: false, form: text)];
+      // earlier still, we didn't have lemmas so this is for really old tokens
+      return Lemma(text: text, saveVocab: false, form: text);
     }
   }
 
   factory PangeaToken.fromJson(Map<String, dynamic> json) {
-    try {
     final PangeaTokenText text =
         PangeaTokenText.fromJson(json[_textKey] as Map<String, dynamic>);
     return PangeaToken(
       text: text,
-        lemmas: getLemmas(text.content, json[_lemmaKey]),
+      lemma: _getLemmas(text.content, json[_lemmaKey]),
-      );
+      pos: json['pos'] ?? '',
-    } catch (err, s) {
+      morph: json['morph'] ?? {},
-      debugger(when: kDebugMode);
-      Sentry.addBreadcrumb(
-        Breadcrumb(
-          message: "PangeaToken.fromJson error",
-          data: {
-            "json": json,
-          },
-        ),
     );
-      ErrorHandler.logError(e: err, s: s);
-      rethrow;
-    }
   }
 
   static const String _textKey = "text";
@@ -57,7 +62,9 @@ class PangeaToken {
 
   Map<String, dynamic> toJson() => {
         _textKey: text.toJson(),
-        _lemmaKey: lemmas.map((e) => e.toJson()).toList(),
+        _lemmaKey: [lemma.toJson()],
+        'pos': pos,
+        'morph': morph,
       };
 
   int get end => text.offset + text.length;

lib/pangea/models/practice_activities.dart/multiple_choice_activity_model.dart

@@ -1,15 +1,18 @@
 import 'package:fluffychat/config/app_config.dart';
+import 'package:fluffychat/pangea/models/practice_activities.dart/practice_activity_model.dart';
 import 'package:flutter/material.dart';
 
 class MultipleChoice {
   final String question;
   final List<String> choices;
   final String answer;
+  final RelevantSpanDisplayDetails? spanDisplayDetails;
 
   MultipleChoice({
     required this.question,
     required this.choices,
     required this.answer,
+    this.spanDisplayDetails,
   });
 
   bool isCorrect(int index) => index == correctAnswerIndex;
@@ -28,6 +31,9 @@ class MultipleChoice {
       question: json['question'] as String,
       choices: (json['choices'] as List).map((e) => e as String).toList(),
       answer: json['answer'] ?? json['correct_answer'] as String,
+      spanDisplayDetails: json['span_display_details'] != null
+          ? RelevantSpanDisplayDetails.fromJson(json['span_display_details'])
+          : null,
     );
   }
 
@@ -36,6 +42,7 @@ class MultipleChoice {
       'question': question,
       'choices': choices,
       'answer': answer,
+      'span_display_details': spanDisplayDetails,
     };
   }
 }

lib/pangea/models/practice_activities.dart/practice_activity_model.dart

@@ -1,5 +1,7 @@
 import 'dart:developer';
 
+import 'package:collection/collection.dart';
+import 'package:fluffychat/pangea/enum/activity_display_instructions_enum.dart';
 import 'package:fluffychat/pangea/enum/activity_type_enum.dart';
 import 'package:fluffychat/pangea/enum/construct_type_enum.dart';
 import 'package:fluffychat/pangea/models/practice_activities.dart/multiple_choice_activity_model.dart';
@@ -279,4 +281,58 @@ class PracticeActivityModel {
       'free_response': freeResponse?.toJson(),
     };
   }
+
+  RelevantSpanDisplayDetails? getRelevantSpanDisplayDetails() {
+    switch (activityType) {
+      case ActivityTypeEnum.multipleChoice:
+        return multipleChoice?.spanDisplayDetails;
+      case ActivityTypeEnum.listening:
+        return null;
+      case ActivityTypeEnum.speaking:
+        return null;
+      case ActivityTypeEnum.freeResponse:
+        return null;
+      default:
+        debugger(when: kDebugMode);
+        return null;
+    }
+  }
+}
+
+/// For those activities with a relevant span, this class will hold the details
+/// of the span and how it should be displayed
+/// e.g. hide the span for conjugation activities
+class RelevantSpanDisplayDetails {
+  final int offset;
+  final int length;
+  final ActivityDisplayInstructionsEnum displayInstructions;
+
+  RelevantSpanDisplayDetails({
+    required this.offset,
+    required this.length,
+    required this.displayInstructions,
+  });
+
+  factory RelevantSpanDisplayDetails.fromJson(Map<String, dynamic> json) {
+    final ActivityDisplayInstructionsEnum? display =
+        ActivityDisplayInstructionsEnum.values.firstWhereOrNull(
+      (e) => e.string == json['display_instructions'],
+    );
+    if (display == null) {
+      debugger(when: kDebugMode);
+    }
+    return RelevantSpanDisplayDetails(
+      offset: json['offset'] as int,
+      length: json['length'] as int,
+      displayInstructions: display ?? ActivityDisplayInstructionsEnum.hide,
+    );
+  }
+
+  Map<String, dynamic> toJson() {
+    return {
+      'offset': offset,
+      'length': length,
+      'display_instructions': displayInstructions,
+    };
+  }
 }

lib/pangea/repo/igc_repo.dart

@@ -47,23 +47,33 @@ class IgcRepo {
       tokens: [
         PangeaToken(
           text: PangeaTokenText(content: "This", offset: 0, length: 4),
-          lemmas: [Lemma(form: "This", text: "this", saveVocab: true)],
+          lemma: Lemma(form: "This", text: "this", saveVocab: true),
+          pos: "DET",
+          morph: {},
         ),
         PangeaToken(
           text: PangeaTokenText(content: "be", offset: 5, length: 2),
-          lemmas: [Lemma(form: "be", text: "be", saveVocab: true)],
+          lemma: Lemma(form: "be", text: "be", saveVocab: true),
+          pos: "VERB",
+          morph: {},
         ),
         PangeaToken(
           text: PangeaTokenText(content: "a", offset: 8, length: 1),
-          lemmas: [],
+          lemma: Lemma(form: "a", text: "a", saveVocab: true),
+          pos: "DET",
+          morph: {},
         ),
         PangeaToken(
           text: PangeaTokenText(content: "sample", offset: 10, length: 6),
-          lemmas: [],
+          lemma: Lemma(form: "sample", text: "sample", saveVocab: true),
+          pos: "NOUN",
+          morph: {},
         ),
         PangeaToken(
           text: PangeaTokenText(content: "text", offset: 17, length: 4),
-          lemmas: [],
+          lemma: Lemma(form: "text", text: "text", saveVocab: true),
+          pos: "NOUN",
+          morph: {},
         ),
       ],
       matches: [