From fa8eee307578a563e2f63124bf774afc475f2f4a Mon Sep 17 00:00:00 2001
From: LVL1024 <70866179+LVL1024@users.noreply.github.com>
Date: Sat, 23 May 2026 22:14:12 +0000
Subject: [PATCH] feat: extract item investment stats into generic data

- Extracts `m_MapModCostBonuses` from selectable heroes
- Validates that investment scaling is identical across all heroes
- Maps engine slot types to standard categories (Weapon, Vitality, Spirit)
- Appends `ItemInvestments` to generic-data.json and adds it to structure validation
---
 src/parser/parser.py           | 50 +++++++++++++++++++++++++++++++++-
 src/parser/parsers/generics.py |  2 +-
 2 files changed, 50 insertions(+), 2 deletions(-)

diff --git a/src/parser/parser.py b/src/parser/parser.py
index c95ca888..6ab68a61 100644
--- a/src/parser/parser.py
+++ b/src/parser/parser.py
@@ -1,5 +1,6 @@
 import os
 import shutil
+import json
 
 from .parsers import (
     abilities,
@@ -160,10 +161,57 @@ def _parse_soul_unlocks(self):
     def _parse_generics(self):
         logger.trace('Parsing Generics...')
         generic_data_path = self.OUTPUT_DIR + '/json/generic-data.json'
-        parsed_generics = generics.GenericParser(generic_data_path, self.data['scripts']['generic_data']).run()
 
+        generic_data = dict(self.data['scripts']['generic_data'])
+        generic_data.update(self._extract_item_investment_stats())
+
+        parsed_generics = generics.GenericParser(generic_data_path, generic_data).run()
         json_utils.write(generic_data_path, json_utils.sort_dict(parsed_generics))
 
+    INVESTMENT_SLOT_MAP = {
+        'EItemSlotType_WeaponMod': 'Weapon',
+        'EItemSlotType_Armor': 'Vitality',
+        'EItemSlotType_Tech': 'Spirit',
+    }
+
+    def _extract_item_investment_stats(self) -> dict:
+        first = None
+        first_key = None
+        for hero_key, hero_data in self.data['scripts']['heroes'].items():
+            if not isinstance(hero_data, dict):
+                continue
+            if hero_key == 'hero_base':
+                continue
+            if not hero_data.get('m_bPlayerSelectable', False):
+                continue
+            if 'm_MapModCostBonuses' not in hero_data:
+                # This should not happen for a selectable hero, but we'll skip
+                logger.warning(f'Selectable hero {hero_key} missing m_MapModCostBonuses')
+                continue
+
+            remapped = {
+                self.INVESTMENT_SLOT_MAP[slot]: entries
+                for slot, entries in hero_data['m_MapModCostBonuses'].items()
+                if slot in self.INVESTMENT_SLOT_MAP
+            }
+
+            if first is None:
+                first = remapped
+                first_key = hero_key
+            else:
+                # deep compare dictionaries (order shouldn't matter because we use same remapping)
+                if json.dumps(first, sort_keys=True) != json.dumps(remapped, sort_keys=True):
+                    raise ValueError(
+                        f'Item investment stats differ between heroes {first_key} and {hero_key}.\n' f'{first_key}: {first}\n{hero_key}: {remapped}'
+                    )
+
+        if first is None:
+            logger.warning('Could not find m_MapModCostBonuses in any selectable hero')
+            return {}
+
+        logger.trace(f'Extracted item investment stats from {first_key} (validated across all selectable heroes)')
+        return {'ItemInvestments': first}
+
     def _parse_localizations(self):
         logger.trace('Parsing Localizations...')
         return localizations.LocalizationParser(self.localizations, self.OUTPUT_DIR).run()
diff --git a/src/parser/parsers/generics.py b/src/parser/parsers/generics.py
index 5a671827..7205f2d7 100644
--- a/src/parser/parsers/generics.py
+++ b/src/parser/parsers/generics.py
@@ -14,7 +14,7 @@ class GenericParser:
     """
 
     def __init__(self, output_dir, generic_data):
-        self.STRUCTURE_KEYS_TO_VALIDATE = ['ObjectiveParams', 'RejuvParams', 'ItemPricePerTier']
+        self.STRUCTURE_KEYS_TO_VALIDATE = ['ObjectiveParams', 'RejuvParams', 'ItemPricePerTier', 'ItemInvestments']
         self.POSSIBLE_PREFIXES = ['m_str', 'm_map', 'm_n', 'm_fl', 'm_', 'fl', 'E', 'n']
         self.generic_data_dir = output_dir
         self.generic_data = generic_data