phaneron/Kelsidavis-WoWee
1
0
Fork 0
mirror of https://github.com/Kelsidavis/WoWee.git synced 2026-03-26 00:40:15 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 5

Add clean.sh reset utility, exclude CSVs from repo, prefer binary DBCs

Browse source 
- Add clean.sh with granular options (--build, --assets, --cache, --csvs, --user, --all, --dry-run)
- Remove 119 CSV files from git tracking and add to .gitignore
- Flip DBC loading order: try binary from extracted MPQs first, CSV as fallback
- Remove forceBinaryDbc workaround and garbled-CSV validation (no longer needed)
This commit is contained in:
Kelsi 2026-02-28 09:04:43 -08:00
parent 0e1241ca60
commit 543535a892
122 changed files with 157 additions and 716135 deletions
Download patch file Download diff file Expand all files Collapse all files

105
src/pipeline/asset_manager.cpp
View file

@ -251,56 +251,14 @@ std::shared_ptr<DBCFile> AssetManager::loadDBC(const std::string& name) {
std::vector<uint8_t> dbcData;
// Some DBC CSV exports are known to be malformed in community datasets
// (string columns shifted, missing numeric ID field). Force binary MPQ data for
// these tables to keep mappings correct.
const bool forceBinaryDbc =
(name == "CreatureDisplayInfo.dbc" ||
name == "CreatureDisplayInfoExtra.dbc" ||
name == "ItemDisplayInfo.dbc" ||
name == "CreatureModelData.dbc" ||
name == "GroundEffectTexture.dbc" ||
name == "GroundEffectDoodad.dbc" ||
name == "SkillLine.dbc");
// Try expansion-specific CSV first (e.g. Data/expansions/wotlk/db/Spell.csv)
bool loadedFromCSV = false;
if (!forceBinaryDbc && !expansionDataPath_.empty()) {
// Derive CSV name from DBC name: "Spell.dbc" -> "Spell.csv"
std::string baseName = name;
auto dot = baseName.rfind('.');
if (dot != std::string::npos) {
baseName = baseName.substr(0, dot);
}
std::string csvPath = expansionDataPath_ + "/db/" + baseName + ".csv";
if (std::filesystem::exists(csvPath)) {
std::ifstream f(csvPath, std::ios::binary | std::ios::ate);
if (f) {
auto size = f.tellg();
if (size > 0) {
f.seekg(0);
dbcData.resize(static_cast<size_t>(size));
f.read(reinterpret_cast<char*>(dbcData.data()), size);
LOG_DEBUG("Found CSV DBC: ", csvPath);
loadedFromCSV = true;
}
}
}
}
if (forceBinaryDbc && !expansionDataPath_.empty()) {
LOG_INFO("Skipping CSV override for visual DBC, using binary: ", name);
}
// Fall back to manifest (binary DBC from extracted MPQs)
if (dbcData.empty()) {
// Try binary DBC from extracted MPQs first (preferred source)
{
std::string dbcPath = "DBFilesClient\\" + name;
dbcData = readFile(dbcPath);
}
// If binary DBC not found and we skipped CSV earlier (forceBinaryDbc),
// try CSV as a last resort — better than no data at all (e.g. Classic expansion
// where binary DBCs come from MPQ extraction the user may not have done).
if (dbcData.empty() && forceBinaryDbc && !expansionDataPath_.empty()) {
// Fall back to expansion-specific CSV (e.g. Data/expansions/wotlk/db/Spell.csv)
if (dbcData.empty() && !expansionDataPath_.empty()) {
std::string baseName = name;
auto dot = baseName.rfind('.');
if (dot != std::string::npos) {
@ -316,7 +274,6 @@ std::shared_ptr<DBCFile> AssetManager::loadDBC(const std::string& name) {
dbcData.resize(static_cast<size_t>(size));
f.read(reinterpret_cast<char*>(dbcData.data()), size);
LOG_INFO("Binary DBC not found, using CSV fallback: ", csvPath);
loadedFromCSV = true;
}
}
}
@ -329,58 +286,8 @@ std::shared_ptr<DBCFile> AssetManager::loadDBC(const std::string& name) {
auto dbc = std::make_shared<DBCFile>();
if (!dbc->load(dbcData)) {
// If CSV failed to parse, try binary DBC fallback
if (loadedFromCSV) {
LOG_WARNING("CSV DBC failed to parse: ", name, " — trying binary DBC fallback");
dbcData.clear();
std::string dbcPath = "DBFilesClient\\" + name;
dbcData = readFile(dbcPath);
if (!dbcData.empty()) {
dbc = std::make_shared<DBCFile>();
if (dbc->load(dbcData)) {
loadedFromCSV = false;
LOG_INFO("Binary DBC fallback succeeded: ", name);
} else {
LOG_ERROR("Failed to load DBC: ", name);
return nullptr;
}
} else {
LOG_ERROR("Failed to load DBC: ", name);
return nullptr;
}
} else {
LOG_ERROR("Failed to load DBC: ", name);
return nullptr;
}
}
// Validate CSV-loaded DBCs: if >50 records but all have ID 0 in field 0,
// the CSV data is garbled (e.g. string data in the ID column).
if (loadedFromCSV && dbc->getRecordCount() > 50) {
uint32_t zeroIds = 0;
const uint32_t sampleSize = std::min(dbc->getRecordCount(), 100u);
for (uint32_t i = 0; i < sampleSize; ++i) {
if (dbc->getUInt32(i, 0) == 0) ++zeroIds;
}
// If >80% of sampled records have ID 0, the CSV is garbled
if (zeroIds > sampleSize * 4 / 5) {
LOG_WARNING("CSV DBC '", name, "' has garbled field 0 (",
zeroIds, "/", sampleSize, " records with ID=0) — falling back to binary DBC");
dbcData.clear();
std::string dbcPath = "DBFilesClient\\" + name;
dbcData = readFile(dbcPath);
if (!dbcData.empty()) {
dbc = std::make_shared<DBCFile>();
if (!dbc->load(dbcData)) {
LOG_ERROR("Binary DBC fallback also failed: ", name);
return nullptr;
}
LOG_INFO("Binary DBC fallback succeeded: ", name);
} else {
LOG_ERROR("No binary DBC fallback available for: ", name, " — discarding garbled CSV");
return nullptr;
}
}
LOG_ERROR("Failed to load DBC: ", name);
return nullptr;
}
dbcCache[name] = dbc;