Fix export on fresh clone (make import), add elision and duplicate-sentence guards, replace 14 duplicate sentences (0.5.3)

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>

tests/data/test_elision.gd

@@ -0,0 +1,110 @@
+extends GutTest
+## Garde-fou orthographique sur tous les textes du jeu : « le arbre » doit
+## s'ecrire « l'arbre ». Parcourt tous les JSON de data/ (chaines et listes de
+## mots consecutifs) a la recherche d'une elision manquante. Les mots a « h »
+## aspire et quelques exceptions legitimes sont listes ci-dessous.
+
+const DATA_DIR := "res://data"
+## Mots a « h » aspire ou vocalique sans elision : « le hibou », « le onze ».
+const NO_ELISION_WORDS: Array[String] = [
+    "hibou",
+    "haie",
+    "haies",
+    "héros",
+    "hérisson",
+    "hamac",
+    "hanche",
+    "hangar",
+    "haricot",
+    "haricots",
+    "hasard",
+    "hache",
+    "hauteur",
+    "haut",
+    "haute",
+    "hautes",
+    "hauts",
+    "hall",
+    "hameau",
+    "harpe",
+    "hâte",
+    "hérissons",
+    "hurlement",
+    "huit",
+    "onze",
+    "oui",
+    "ouistiti",
+    "yaourt",
+    "yeux",
+    "ukulélé",
+]
+
+## (*UCP) : \\w couvre les lettres accentuees. Motif : mot d'elision, espaces,
+## puis mot a voyelle ou « h » initial ; l'assertion arriere ecarte « écris-la en ».
+const PATTERN := (
+    "(*UCP)(?<![-\\w'’])(?i)(le|la|de|que|ne|se|me|te|je|ce)\\s+"
+    + "([aeiouâàäéèêëîïôöûùüœh][\\wàâäéèêëîïôöûùüçœ]*)"
+)
+
+var _regex := RegEx.new()
+var _problems: Array[String] = []
+
+
+func before_all() -> void:
+    _regex.compile(PATTERN)
+
+
+func test_no_missing_elision_in_game_texts() -> void:
+    _scan_dir(DATA_DIR)
+    assert_eq(_problems.size(), 0, "Elisions manquantes : %s" % "\n".join(_problems))
+
+
+func test_detector_catches_a_missing_elision() -> void:
+    _problems.clear()
+    _check_text("le arbre est grand", "exemple")
+    assert_eq(_problems.size(), 1)
+    _problems.clear()
+    _check_text("Le hibou dort. Écris-la en entier. Le onze.", "exemple")
+    assert_eq(_problems.size(), 0)
+
+
+func _scan_dir(path: String) -> void:
+    var dir := DirAccess.open(path)
+    if dir == null:
+        return
+    for sub in dir.get_directories():
+        if sub != "schemas":
+            _scan_dir(path.path_join(sub))
+    for file_name in dir.get_files():
+        if file_name.ends_with(".json"):
+            _scan_file(path.path_join(file_name))
+
+
+func _scan_file(path: String) -> void:
+    var file := FileAccess.open(path, FileAccess.READ)
+    var parsed: Variant = JSON.parse_string(file.get_as_text())
+    _scan_value(parsed, path)
+
+
+func _scan_value(value: Variant, where: String) -> void:
+    if typeof(value) == TYPE_STRING:
+        _check_text(value, where)
+    elif typeof(value) == TYPE_DICTIONARY:
+        for key: Variant in value:
+            _scan_value(value[key], where)
+    elif typeof(value) == TYPE_ARRAY:
+        var items: Array = value
+        # Mots consecutifs d'une phrase decoupee en jetons.
+        for i in range(items.size() - 1):
+            if typeof(items[i]) == TYPE_STRING and typeof(items[i + 1]) == TYPE_STRING:
+                _check_text("%s %s" % [items[i], items[i + 1]], where)
+        for item: Variant in items:
+            _scan_value(item, where)
+
+
+func _check_text(text: String, where: String) -> void:
+    for match in _regex.search_all(text):
+        var next_word := match.get_string(2).to_lower()
+        if NO_ELISION_WORDS.has(next_word):
+            continue
+        _problems.append("%s : « %s »" % [where, match.get_string(0)])

tests/data/test_elision.gd.uid

@@ -0,0 +1 @@
+uid://beh3tst4how7j

tests/data/test_no_duplicate_sentences.gd

@@ -0,0 +1,77 @@
+extends GutTest
+## Aucune phrase ne doit apparaitre deux fois dans le jeu (exercices et
+## dictees confondus) : un meme enonce revu dans deux missions est une
+## redite, pas un nouvel exercice. La phrase est reconstituee en version
+## correcte (blanc rempli, mot fautif corrige) puis comparee sans casse ni
+## ponctuation.
+
+const EXERCISES_DIR := "res://data/exercises"
+const DICTATIONS_DIR := "res://data/dictations"
+
+
+func test_no_sentence_is_used_twice() -> void:
+    var seen: Dictionary = {}
+    var duplicates: Array[String] = []
+    for entry in _collect_sentences():
+        var key := _normalize(entry["sentence"])
+        if seen.has(key):
+            duplicates.append("« %s » : %s et %s" % [entry["sentence"], seen[key], entry["id"]])
+        else:
+            seen[key] = entry["id"]
+    assert_eq(duplicates.size(), 0, "Phrases en double :\n%s" % "\n".join(duplicates))
+    assert_gt(seen.size(), 100, "le test doit avoir vu les phrases du jeu")
+
+
+func _collect_sentences() -> Array[Dictionary]:
+    var sentences: Array[Dictionary] = []
+    for path in _json_files(EXERCISES_DIR):
+        var data := _read(path)
+        var sentence := ""
+        if data.get("type", "") == "multiple_choice":
+            var answers: Array = data["answers"]
+            sentence = str(data["question"]).replace(
+                "...", str(answers[int(data["correct_answer"])])
+            )
+        elif data.get("type", "") == "sentence_correction":
+            var words: Array[String] = []
+            for token: Dictionary in data["tokens"]:
+                words.append(token["text"] if token["correct"] else token["expected"])
+            sentence = " ".join(words)
+        if not sentence.is_empty():
+            sentences.append({"id": data["id"], "sentence": sentence})
+    for path in _json_files(DICTATIONS_DIR):
+        var data := _read(path)
+        sentences.append({"id": data["id"], "sentence": data["target_text"]})
+    return sentences
+
+
+func _normalize(sentence: String) -> String:
+    var cleaned := ""
+    for character in sentence.to_lower():
+        if character.is_valid_int() or character == "'" or character == " ":
+            cleaned += character
+        elif character.to_upper() != character or "éèêàâîïôûùçœ".contains(character):
+            cleaned += character
+        else:
+            cleaned += " "
+    var parts := cleaned.split(" ", false)
+    return " ".join(parts)
+
+
+func _json_files(path: String) -> Array[String]:
+    var found: Array[String] = []
+    var dir := DirAccess.open(path)
+    if dir == null:
+        return found
+    for sub in dir.get_directories():
+        found.append_array(_json_files(path.path_join(sub)))
+    for file_name in dir.get_files():
+        if file_name.ends_with(".json"):
+            found.append(path.path_join(file_name))
+    return found
+
+
+func _read(path: String) -> Dictionary:
+    var file := FileAccess.open(path, FileAccess.READ)
+    var parsed: Variant = JSON.parse_string(file.get_as_text())
+    return parsed if typeof(parsed) == TYPE_DICTIONARY else {}

tests/data/test_no_duplicate_sentences.gd.uid

@@ -0,0 +1 @@
+uid://cekmrk80y81fk