first experiments

sort_pages.py

@@ -0,0 +1,156 @@
+from pathlib import Path
+import shutil
+import re
+import json
+
+INPUT_DIR = Path("unique_pages")
+OUTPUT_DIR = Path("classified_pages")
+
+CATEGORY_RE = re.compile(r'"wgCategories":\[(.*?)\]')
+
+# ---------- CONFIG ----------
+
+FACTIONS = {
+    "cygnar",
+    "khador",
+    "protectorate of menoth",
+    "cryx",
+    "retribution of scyrah",
+    "convergence of cyriss",
+    "cephalyx",
+    "mercenary",
+    "crucible guard",
+    "infernals",
+    "trollblood",
+    "circle orboros",
+    "skorne",
+    "legion of everblight",
+    "grymkin",
+    "minion",
+}
+
+MODEL_TYPES = {
+    "warcaster unit": "warcaster unit",
+    "warlock unit": "warlock unit",
+    "warcaster": "warcaster",
+    "warlock": "warlocks",
+    "infernal master": "masters",
+    "warjack": "warjacks",
+    "warbeast": "warbeasts",
+    "horror": "horrors",
+    "monstrosity": "monstrosities",
+    "solo": "solos",
+    "unit": "units",
+    "battle engine": "battle engines",
+    "structure": "structure",
+}
+
+# ---------- helpers ----------
+
+
+def extract_categories(html: str):
+    m = CATEGORY_RE.search(html)
+    if not m:
+        return []
+
+    raw = "[" + m.group(1) + "]"
+    try:
+        return [c.lower() for c in json.loads(raw)]
+    except:
+        return []
+
+
+def detect_faction(categories):
+    for c in categories:
+        if c in FACTIONS:
+            return c
+    return None
+
+
+def detect_model_type(categories):
+    for c, folder in MODEL_TYPES.items():
+        if c in categories:
+            return folder
+    return None
+
+
+# ---------- classification ----------
+
+
+def classify(categories):
+
+    # SPELLS
+    if "spell" in categories:
+        return Path("common/warmahordes rules/spells")
+
+    if "animus" in categories:
+        return Path("common/warmahordes rules/animi")
+
+    if "model ability" in categories:
+        return Path("common/warmahordes rules/model advantages")
+
+    if "weapon ability" in categories:
+        return Path("common/warmahordes rules/weapon qualities")
+
+    # MODELS
+    if "model" in categories:
+        faction = detect_faction(categories)
+        model_type = detect_model_type(categories)
+
+        if faction and model_type:
+            system = (
+                "warmachine"
+                if faction
+                not in {
+                    "trollblood",
+                    "circle orboros",
+                    "skorne",
+                    "legion of everblight",
+                    "grymkin",
+                    "minion",
+                }
+                else "hordes"
+            )
+
+            return Path(f"{system}/{faction}/{model_type}")
+
+    # THEME FORCE
+    if "theme force" in categories:
+        faction = detect_faction(categories)
+        if faction:
+            system = (
+                "warmachine"
+                if faction
+                not in {
+                    "trollblood",
+                    "circle orboros",
+                    "skorne",
+                    "legion of everblight",
+                    "grymkin",
+                    "minion",
+                }
+                else "hordes"
+            )
+
+            return Path(f"{system}/{faction}/theme forces")
+
+    # fallback
+    return Path("common/others")
+
+
+# ---------- main ----------
+
+files = list(INPUT_DIR.glob("*.html"))
+print(f"{len(files)} pages à classifier")
+
+for file_path in files:
+
+    html = file_path.read_text(encoding="utf-8", errors="ignore")
+    categories = extract_categories(html)
+
+    dest = OUTPUT_DIR / classify(categories)
+    dest.mkdir(parents=True, exist_ok=True)
+
+    shutil.copy2(file_path, dest / file_path.name)
+
+print("✅ Classification terminée")