Mythli

No, Database data is by default 0 (that means dbc-data is used) Thanks for your help

WAD is a very old emulator supporting patch 1.x if you want to call it so (it is more about playing with the wow world) for more info read http://www.gotwow.ic.cz/wowd/ "Database" files in WAD look like this: #include scripts/extra/creatures_auctioneers.scp ... #include scripts/afg/afgcreatures.scp #include scripts/afg/afgqiraj.scp #include scripts/chronos/chronos_vendors.scp #include scripts/chronos/chronos_creatures.scp [creature 0] name=Spawn Points (Only GM can see it) faction=35 level=255 attack=2000 2200 bounding_radius=1.5 combat_reach=2.25 damage=0 2 maxhealth=56 model=262 size=0.5 type=1 maxmana=0 money=127500 [creature 1] name=Spawn Point (Only GM can see it) faction=35 level=255 attack=2000 2200 bounding_radius=1.5 combat_reach=2.25 damage=0 2 maxhealth=56 model=262 size=0.5 type=1 maxmana=0 money=127500

I have written this tool long time ago for personal use and learning reasons (first c# programm i have ever wrote) SQLUpdate Merging Tool This is my little tool for merging various sql-updates to one big update. i have written this in .net (c#) but dont look at the code it's horrible Features: 1. Support SD2, Mangos, UDB and many other SQL Update formats ^\\\\d+_{1}(\\\\d+)_{1}\\\\d+_{1}([A-Z]|[a-z]+)_{1}(.*)$", //match mangos update style (provide revision number, databasename, tablename) ^r{1}(\\\\d+)_{1}(\\\\w+).*$", //match SD2 Update style with some on official style (provide revision number& database name) ^r{1}(\\\\d+)_{1}(.*)$", //match SD2 update style (provide revision number& database name) 2. Grouping 3. Editing 4. Drag& Drop 5. Apply filter Screenshots& Explanation: Feel free to comment this tool/ post feedback. regards Mythli <3 Download: http://mythli.ath.cx/websvn/wsvn/Downloads/sqlupdatemerge_1.0/?op=dl&isdir=1& Changelog: http://mythli.ath.cx/websvn/wsvn/sqlMerge/trunk/?op=log&isdir=1& RSS: http://mythli.ath.cx/websvn/wsvn/sqlMerge/trunk/?op=rss&isdir=1& requires .net framework 2.0 or newest mono release.

In the hope someone find this snippet usefull using System; using System.Collections.Generic; using System.Linq; using System.Text; using System.IO; using System.Text.RegularExpressions; using System.Data; namespace scpparser.Classes { public class ScpParser { public String IncludeRoot; public String FileName; public List<String> ParsedFiles = new List<String>(); private Regex IncludeSearchPattern = new Regex("^(?!//)#include\\\\s*(.*)"); private Regex SectionStartSearchPattern = new Regex("^(?!//)\\\\[(\\\\w+)\\\\s*([^]]+)\\\\]$"); private Regex ValuePairSearchPattern = new Regex("^(?!//)(\\\\w+)\\\\s*=\\\\s*(.*)$"); private String currentSection; private String currentPrimaryKey; public Dictionary<String, Dictionary<String, List<String>>> ParseResult = new Dictionary<String, Dictionary<String, List<String>>>(); public Dictionary<String, String> ParsedColumns = new Dictionary<String, String>(); public ScpParser() { } public ScpParser(String IncludeRoot, String FileName) : this() { this.Open(IncludeRoot, FileName); } public void HandleInclude(String FileName) { FileName = Path.Combine(this.IncludeRoot, FileName); this.ParseFile(FileName); } public void HandleSection(String SectionName, String PrimaryKey) { this.currentSection = SectionName; this.currentPrimaryKey = PrimaryKey; this.ParseResult.Add(PrimaryKey, new Dictionary<String, List<String>>()); } public void HandleValuePair(String ValueName, String Value) { if (String.IsNullOrEmpty(this.currentSection) || String.IsNullOrEmpty(this.currentPrimaryKey)) return; if (!this.ParseResult[this.currentPrimaryKey].ContainsKey(ValueName)) { this.ParseResult[this.currentPrimaryKey].Add(ValueName, new List<String>()); } if (!this.ParsedColumns.ContainsKey(ValueName)) this.ParsedColumns.Add(ValueName, ""); this.ParseResult[this.currentPrimaryKey][ValueName].Add(Value); } public void ParseLine(String Line) { Match match = null; match = this.IncludeSearchPattern.Match(Line); if (match.Success) { this.HandleInclude(match.Groups[1].Value); return; } match = this.SectionStartSearchPattern.Match(Line); if (match.Success) { this.HandleSection(match.Groups[1].Value, match.Groups[2].Value); return; } match = this.ValuePairSearchPattern.Match(Line); if (match.Success) { this.HandleValuePair(match.Groups[1].Value, match.Groups[2].Value); return; } } public void ParseFile(String FileName) { using (StreamReader fStream = File.OpenText(FileName)) { String line; while ((line = fStream.ReadLine()) != null) { this.ParseLine(line); } } this.ParsedFiles.Add(FileName); } public void Open(String IncludeRoot, String FileName) { this.IncludeRoot = IncludeRoot; this.FileName = FileName; this.ParseFile(FileName); } } }

The Subject tells you everything you need to know Thanks in advance

look at my source, i don't make anything different.

that's true and... I have already said that this was just programmed for learning reasons so sscanf(..) is faster and i never said that regular expressions make the performance boost if we look at the code while(std::getline(this->fileStream, line)) { parselinehere(line); } would end in: * read line from file * parse line * waiting for parsing line to read next line until all lines are parsed with multiple threads we can read lines without waiting for parsing - on my hardware reading lines is very fast (ssd) and parsing lines is the time consuming thing. With multiple threads and a line buffer we can reduce the overhead to a minimum.

"additemset" is a method in the player class which is derived from the unit class. so do research on your own before posing...

i started programming with c++ a few weeks ago but i still think this is a good thing ("this->) "logical divide" member and local variables like m_ do. we are talking about 100-200ms or something around

I've programmed this for learning reasons and i need some feedback (coding style, correct pointers, ...) this codesnip will parse mangos-like configuration files using regular expressions and multiple threads (you need the boost c++ libary to compile) On a AMD Phenom x4 945 (+intel ssd) this code is up to 20 times faster in parsing than the dotconfpp mangos configuration class i don't think performance matter if we parse configuration files but this is still an interesting fact. here is my code: header-file: #ifndef CONFIG_H #define CONFIG_H #include "Common.h" #include <fstream> #include <boost/regex.hpp> class Config { public: Config(std::string ConfigFileName); void ReloadConfig(); std::string GetStringDefault(std::string Name, std::string DefaultValue = ""); bool GetBoolDefault(std::string Name, bool DefaultValue = false); int GetIntDefault(std::string Name, int DefaultValue = 0); float GetFloatDefault(std::string Name, float DefaultValue = 0); std::string GetFileName() { return this->configFileName; } ~Config(); private: std::string configFileName; std::ifstream fileStream; boost::regex searchPattern; stringMapType configMap; //--used for threading boost::mutex lineBufferMutex; boost::mutex configMapMutex; std::list<std::string> lineBuffer; bool isFileFinish; void lineParser(); //-- }; #endif code-file: #include "Config.h" Config::Config(std::string ConfigFileName) { this->configFileName = ConfigFileName; this->fileStream.open(this->configFileName); if (this->fileStream.is_open()) { this->searchPattern = boost::regex("^([ \\\\t]+)?([^#][^\\\\s]+)\\\\s*=\\\\s*(?\\\\d+)|(?:\\"([^\\"]*)\\")).*$"); this->ReloadConfig(); } } void Config::lineParser() { //--keep thread alive until file is fully pushed into lineBuffer and lineBuffer is empty while(this->isFileFinish == false || this->lineBuffer.size() != 0) { //--wait until lines for parsing are avaible if (this->lineBuffer.size() == 0) boost::this_thread::sleep(boost::Posix_time::millisec(1)); //-- else { boost::match_results<std::string::const_iterator> matches; //--lock lineBuffer object this->lineBufferMutex.lock(); //copy first line into temp var std::string line = *this->lineBuffer.begin(); //remove first line from lineBuffer this->lineBuffer.pop_front(); this->lineBufferMutex.unlock(); //-- if (boost::regex_match(line, matches, this->searchPattern)) { //--insert match into our configMap (we have to lock our configMap object first) this->configMapMutex.lock(); if (matches[3].matched) //insert number (without qoutes) this->configMap.insert(std::make_pair(matches[2], matches[3])); else //insert string (in quoutes) this->configMap.insert(std::make_pair(matches[2], matches[4])); this->configMapMutex.unlock(); //-- } } //-- } } void Config::ReloadConfig() { //--reset this->configMap.clear(); this->fileStream.seekg(0, ios::beg); this->isFileFinish = false; //-- //--define variables std::string line; boost::thread_group threadPool; //-- //--create parser-threads for each logical cpu for(uint32 i = 0; i < boost::thread::hardware_concurrency(); i++) { threadPool.create_thread(boost::bind(&Config::lineParser, this)); } //-- //--push lines from file in our lineBuffer while(std::getline(this->fileStream, line)) { this->lineBuffer.push_back(line); } //-- this->isFileFinish = true; //wait until all lines are parsed threadPool.join_all(); } std::string Config::GetStringDefault(std::string Name, std::string DefaultValue) { stringMapType::const_iterator configNode = this->configMap.find(Name); if(configNode != this->configMap.end()) return configNode->second; else return DefaultValue; } bool Config::GetBoolDefault(std::string Name, bool DefaultValue) { std::string baseValue = this->GetStringDefault(Name, ""); if (baseValue == "") return DefaultValue; else return boost::lexical_cast<bool>(baseValue); } int Config::GetIntDefault(std::string Name, int DefaultValue) { std::string baseValue = this->GetStringDefault(Name, ""); if (baseValue == "") return DefaultValue; else return boost::lexical_cast<int>(baseValue); } float Config::GetFloatDefault(std::string Name, float DefaultValue) { std::string baseValue = this->GetStringDefault(Name, ""); if (baseValue == "") return DefaultValue; else return boost::lexical_cast<float>(baseValue); } Config::~Config() { this->fileStream.close(); }

mangos use multiple threads if you mean that - but i think it only support 2 logical cores fully wasn't it?

public class DotConfParser : Dictionary<String, String> { public DotConfParser() :base() {} public DotConfParser(String FileName) { this.configFileName = FileName; this.Reload(); } public void Reload() { using (StreamReader streamReader = new StreamReader(File.OpenRead(this.configFileName))) { while (!streamReader.EndOfStream) { Match match = this.searchPattern.Match(streamReader.ReadLine()); if (match.Success) if (!base.ContainsKey(match.Groups[2].Value)) base.Add(match.Groups[2].Value, match.Groups[3].Value.Trim('"')); } } } private String configFileName; private Regex searchPattern = new Regex(@"^([ \\\\t]+)?([^#].+)\\\\s*=\\\\s*(?\\\\d)|(?:\\"([^\\"]*)\\")).*$", RegexOptions.Compiled | RegexOptions.Multiline); } A c# class to read mangos configuration files. Enjoy.

And the "new" already very shiny wiki would be more shiny with more activity inside take this as attention

If you look at DBCReader.cs you will see that there is no detection: for (int r = 0; r < header.RecordsCount; ++r) { uint key = reader.ReadUInt32(); reader.BaseStream.Position -= 4; T T_entry = reader.ReadStruct<T>(); dict.Add(key, T_entry); } He try to read a struct from DBCFile with predefined types, structure of Spell.dbc is defined in "Structure.cs" - this is exactly that what i don't want.

I have done some research on mangos code (DBCLoader) and figured out that you have string-reference fields which contain a Int32 value that point to a string with the following offset: StrReadOffset = StringStartOffset + FieldValue Problem now is that i have to detect if column is a string reference or not.