better item assets matching

www/assets_finding.ipynb

@@ -2,7 +2,7 @@
  "cells": [
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 8,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -18,7 +18,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": 9,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -29,7 +29,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 15,
+   "execution_count": 10,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -41,7 +41,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 23,
+   "execution_count": 44,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -49,46 +49,123 @@
     "\n",
     "def filter_candidates(candidates):\n",
     "    max_match_len = 0\n",
-    "    max_len_matches = []\n",
+    "    filtered_matches = []\n",
+    "\n",
+    "    # go for longest match\n",
     "    for can in candidates:\n",
     "        name, match, mapping = can\n",
     "        match_len = len(match)\n",
     "        if match_len > max_match_len:\n",
-    "            max_match_len = 0\n",
-    "            max_len_matches = [(name, match, mapping)]\n",
+    "            max_match_len = match_len\n",
+    "            filtered_matches = [(name, match, mapping)]\n",
     "        elif match_len == max_match_len:\n",
-    "            max_len_matches.append((name, match, mapping))\n",
-    "    if len(max_len_matches) > 1:\n",
-    "        for can in max_len_matches:\n",
-    "            name, match = can\n",
-    "            if f\"{match}_White\" in name:\n",
+    "            filtered_matches.append((name, match, mapping))\n",
+    "\n",
+    "    # choose better matches\n",
+    "    if len(filtered_matches) > 1:\n",
+    "        better_matches = []\n",
+    "        for can in filtered_matches:\n",
+    "            name, match, mapping = can\n",
+    "            if mapping.startswith(\"Item\") and mapping in name:\n",
+    "                better_matches.append((name, match, mapping))\n",
+    "            elif mapping.startswith(\"Structure\") and mapping in name:\n",
+    "                better_matches.append((name, match, mapping))\n",
+    "        if len(better_matches) > 0:\n",
+    "            filtered_matches = better_matches\n",
+    "\n",
+    "    #exclude build states if we have non build states\n",
+    "    if len(filtered_matches) > 1:\n",
+    "        non_build_state = []\n",
+    "        for can in filtered_matches:\n",
+    "            name, match, mapping = can\n",
+    "            if \"BuildState\" not in name:\n",
+    "                non_build_state.append((name, match, mapping))\n",
+    "        if len(non_build_state) > 0:\n",
+    "            filtered_matches = non_build_state\n",
+    "\n",
+    "    #prefer matches without extra tags\n",
+    "    if len(filtered_matches) > 1:\n",
+    "        direct = []\n",
+    "        for can in filtered_matches:\n",
+    "            name, match, mapping = can\n",
+    "            if f\"{match}-\" in name:\n",
+    "                direct.append((name, match, mapping))\n",
+    "        if len(direct) > 0:\n",
+    "            filtered_matches = direct\n",
+    "    \n",
+    "    #filter to unique filenames\n",
+    "    if len(filtered_matches) > 1:\n",
+    "        unique_names = []\n",
+    "        unique_matches = []\n",
+    "        for can in filtered_matches:\n",
+    "            name, match, mapping = can\n",
+    "            if name not in unique_names:\n",
+    "                unique_names.append(name)\n",
+    "                unique_matches.append((name, match, mapping))\n",
+    "        filtered_matches = unique_matches\n",
+    "\n",
+    "    #prefer not worse matches\n",
+    "    if len(filtered_matches) > 1:\n",
+    "        not_worse = []\n",
+    "        for can in filtered_matches:\n",
+    "            name, match, mapping = can\n",
+    "            if name.startswith(\"Item\") and not mapping.startswith(\"Item\"):\n",
+    "                continue\n",
+    "            elif name.startswith(\"Structure\") and not mapping.startswith(\"Structure\"):\n",
+    "                continue\n",
+    "            elif name.startswith(\"Kit\") and not mapping.startswith(\"Kit\"):\n",
+    "                continue\n",
+    "            elif not name.startswith(match):\n",
+    "                continue\n",
+    "            not_worse.append((name, match, mapping))\n",
+    "        if len(not_worse) > 0:\n",
+    "            filtered_matches = not_worse\n",
+    "\n",
+    "    #if we have colored variants take White\n",
+    "    if len(filtered_matches) > 1:\n",
+    "        for can in filtered_matches:\n",
+    "            name, match, mapping = can\n",
+    "            if f\"_White\" in name:\n",
     "                return [name]\n",
-    "    return [(name, mapping) for name, _, mapping in max_len_matches]\n",
+    "\n",
+    "    return [name for name, _, _ in filtered_matches]\n",
     "\n",
     "for entry in db.values():\n",
     "    candidates = []\n",
     "    for name in names:\n",
     "        if entry[\"name\"] in name:\n",
-    "            candidates.append((name, entry[\"name\"],  entry[\"name\"]))\n",
+    "            candidates.append((name, entry[\"name\"], entry[\"name\"]))\n",
     "        if entry[\"name\"].removeprefix(\"Item\") in name:\n",
-    "            candidates.append((name, entry[\"name\"].removeprefix(\"Item\"),  entry[\"name\"]))\n",
+    "            candidates.append((name, entry[\"name\"].removeprefix(\"Item\"), entry[\"name\"]))\n",
     "        if entry[\"name\"].removeprefix(\"Structure\") in name:\n",
-    "            candidates.append((name, entry[\"name\"].removeprefix(\"Structure\"),  entry[\"name\"]))\n",
+    "            candidates.append((name, entry[\"name\"].removeprefix(\"Structure\"), entry[\"name\"]))\n",
     "    image_candidates[entry[\"name\"]] = filter_candidates(candidates)\n",
     "\n"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 32,
+   "execution_count": 46,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# rematch items to super structure?\n",
+    "for name in image_candidates.keys():\n",
+    "    for other in image_candidates.keys():\n",
+    "        if name != other and name in other:\n",
+    "            if len(image_candidates[name]) > 0 and len(image_candidates[other]) == 0:\n",
+    "                image_candidates[other] = image_candidates[name]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 47,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "ItemAuthoringToolRocketNetwork []\n",
-      "ImGuiCircuitboardAirlockControl []\n",
       "ItemBiomass []\n",
       "StructureBlocker []\n",
       "CartridgePlantAnalyser []\n",
@@ -99,17 +176,13 @@
       "Flag_ODA_6m []\n",
       "Flag_ODA_8m []\n",
       "ItemHorticultureBelt []\n",
-      "StructureHydroponicsTrayData []\n",
       "ItemKitLiquidRegulator []\n",
       "ItemKitPortablesConnector []\n",
       "Landingpad_GasConnectorInwardPiece []\n",
       "Landingpad_LiquidConnectorInwardPiece []\n",
+      "ItemMushroom ['ItemMushroom-resources.assets-3022.png', 'ItemMushroom-resources.assets-9304.png']\n",
       "StructurePlinth []\n",
-      "DynamicGasTankAdvancedOxygen []\n",
-      "Rover_MkI_build_states []\n",
-      "ItemPlantThermogenic_Creative []\n",
-      "StructureWaterBottleFillerPoweredBottom []\n",
-      "StructureWaterBottleFillerPowered []\n"
+      "ItemPlantThermogenic_Creative []\n"
      ]
     }
    ],
@@ -118,20 +191,24 @@
     "for name, candidates in image_candidates.items():\n",
     "    if len(candidates) != 1:\n",
     "        print(name, candidates)\n",
+    "        if len(candidates) > 1:\n",
+    "            #take first as fallback\n",
+    "            to_copy.append((name, candidates[0]))\n",
     "    else:\n",
-    "        to_copy.append((name, candidates[0][0]))\n"
+    "        # print(name, candidates)\n",
+    "        to_copy.append((name, candidates[0]))\n"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 34,
+   "execution_count": 48,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "|████████████████████████████████████████| 1216/1216 [100%] in 0.6s (1885.28/s) \n"
+      "|████████████████████████████████████████| 1223/1223 [100%] in 0.8s (1494.22/s) \n"
      ]
     }
    ],
@@ -145,7 +222,7 @@
     "    for name, file in to_copy:\n",
     "        source = datapath / file\n",
     "        dest = destpath / f\"{name}.png\"\n",
-    "        shutil.copyfile(source, dest)\n",
+    "        shutil.copy(source, dest)\n",
     "        bar()\n"
    ]
   }