|
public selectQuery ( string query ) : |
||
| query | string | |
| return | ||
private void parseUA(string _userAgent)
{
int client_id = 0;
int client_class_id = -1;
int os_id = 0;
int deviceclass_id = 0;
System.Text.RegularExpressions.Regex searchTerm;
PerlRegExpConverter regConv;
if (!string.IsNullOrEmpty(_userAgent))
{
userAgent.UaString = this.ua;
userAgent.UaClass = "Unrecognized";
userAgent.UaClassCode = "unrecognized";
if (dt.Connected)
{
DataTable crawlerList = dt.selectQuery(@"SELECT udger_crawler_list.id as botid,name,ver,ver_major,last_seen,respect_robotstxt,family,family_code,family_homepage,family_icon,vendor,vendor_code,vendor_homepage,crawler_classification,crawler_classification_code
FROM udger_crawler_list
LEFT JOIN udger_crawler_class ON udger_crawler_class.id = udger_crawler_list.class_id
WHERE ua_string = '" + _userAgent + "'");
// crawler
if (crawlerList.Rows.Count > 0)
{
client_class_id = 99;
this.prepareCrawler(crawlerList.Rows[0]);
}
else
{
//client
DataTable clientRegex = dt.selectQuery(@"SELECT class_id,client_id,regstring,name,name_code,homepage,icon,icon_big,engine,vendor,vendor_code,vendor_homepage,uptodate_current_version,client_classification,client_classification_code
FROM udger_client_regex
JOIN udger_client_list ON udger_client_list.id = udger_client_regex.client_id
JOIN udger_client_class ON udger_client_class.id = udger_client_list.class_id
ORDER BY sequence ASC");
if (clientRegex != null)
{
foreach (DataRow row in clientRegex.Rows)
{
regConv = new PerlRegExpConverter(row["regstring"].ToString(), "", Encoding.UTF8);
searchTerm = regConv.Regex;
if (searchTerm.IsMatch(_userAgent))
{
var match = searchTerm.Match(_userAgent);
if (match.Success && match.Groups.Count > 1)
{
this.prepareClientRegex(row, match, ref client_id, ref client_class_id);
break;
}
else if (match.Success && match.Groups.Count == 1)
{
this.prepareClientRegex(row, match, ref client_id, ref client_class_id);
break;
}
}
}
}
}
// OS
DataTable osRegex = dt.selectQuery(@"SELECT os_id,regstring,family,family_code,name,name_code,homepage,icon,icon_big,vendor,vendor_code,vendor_homepage
FROM udger_os_regex
JOIN udger_os_list ON udger_os_list.id = udger_os_regex.os_id
ORDER BY sequence ASC");
if (osRegex != null)
{
foreach (DataRow row in osRegex.Rows)
{
regConv = new PerlRegExpConverter(row["regstring"].ToString(), "", Encoding.UTF8);
searchTerm = regConv.Regex;
if (searchTerm.IsMatch(_userAgent))
{
this.prepareOs(row, ref os_id);
break;
}
}
}
// client_os_relation
if (os_id == 0 && client_id != 0)
{
DataTable clientOSRelations = dt.selectQuery(@"SELECT os_id,family,family_code,name,name_code,homepage,icon,icon_big,vendor,vendor_code,vendor_homepage
FROM udger_client_os_relation
JOIN udger_os_list ON udger_os_list.id = udger_client_os_relation.os_id
WHERE client_id = '" + client_id.ToString() + "'");
if (clientOSRelations != null && clientOSRelations.Rows.Count > 0)
{
this.prepareOs(clientOSRelations.Rows[0], ref os_id);
}
}
// device
DataTable device = dt.selectQuery(@"SELECT deviceclass_id,regstring,name,name_code,icon,icon_big
FROM udger_deviceclass_regex
JOIN udger_deviceclass_list ON udger_deviceclass_list.id=udger_deviceclass_regex.deviceclass_id
ORDER BY sequence ASC");
if (device != null)
{
foreach (DataRow row in device.Rows)
{
regConv = new PerlRegExpConverter(row["regstring"].ToString(), "", Encoding.UTF8);
searchTerm = regConv.Regex;
if (searchTerm.IsMatch(_userAgent))
{
this.prepareDevice(row, ref deviceclass_id);
break;
}
}
}
if (deviceclass_id == 0 && client_class_id != -1)
{
DataTable deviceList = dt.selectQuery(@"SELECT deviceclass_id,name,name_code,icon,icon_big
FROM udger_deviceclass_list
JOIN udger_client_class ON udger_client_class.deviceclass_id = udger_deviceclass_list.id
WHERE udger_client_class.id = '" + client_class_id.ToString() + "'");
if (deviceList != null && deviceList.Rows.Count > 0)
{
this.prepareDevice(deviceList.Rows[0], ref deviceclass_id);
}
}
}
}
}
/// <summary>
/// Check if useragent string and/or IP address is bot
/// </summary>
/// <param name="_useragent">user agent string</param>
/// <param name="_ip">IP address v4 or v6</param>
/// <returns>Dictionary</returns>
public Dictionary <string, object> isBot(string _useragent = "", string _ip = "")
{
this.WriteDebug("isBot: start");
Dictionary <string, object> ret = new Dictionary <string, object>();;
if (_useragent == "" && _ip == "")
{
this.WriteDebug("isBot: Missing mandatory parameter");
ret.Add("flag", 1);
ret.Add("errortext", "missing mandatory parameter");
return(ret);
}
if (_ip != "" && !this.IsValidIp(_ip))
{
this.WriteDebug("isBot: IP address is not valid");
ret.Add("flag", 2);
ret.Add("errortext", "ip address is not valid");
return(ret);
}
dt.connect(this);
if (!dt.Connected)
{
this.WriteDebug("Data file not found, download the data manually");
ret.Add("flag", 3);
ret.Add("errortext", "data file not found");
return(ret);
}
bool botInfo = false;
bool botInfoUA = false;
bool botInfoIP = false;
bool harmony = false;
string botName = "";
string family = "";
string botURL = "";
if (_useragent != "")
{
this.WriteDebug("isBot: test useragent");
DataTable table1 = dt.selectQuery("SELECT name,family FROM c_robots where md5='" + this.CreateMD5(_useragent) + "'");
if (table1.Rows.Count > 0)
{
DataRow rowusr = table1.Rows[0];
botInfo = true;
botInfoUA = true;
botName = rowusr["name"].ToString();
family = rowusr["family"].ToString();
botURL = "http://udger.com/resources/ua-list/bot-detail?bot=" + family;
}
}
if (_ip != "")
{
this.WriteDebug("isBot: test IP address");
DataTable table2 = dt.selectQuery("SELECT name,family from c_robots AS C JOIN bot_ip as B ON C.id=B.robot and B.md5='" + this.CreateMD5(_ip) + "' ");
if (table2.Rows.Count > 0)
{
DataRow row = table2.Rows[0];
botInfo = true;
botInfoIP = true;
if (family == row["family"].ToString())
{
harmony = true;
}
botName = row["name"].ToString();
botURL = "http://udger.com/resources/ua-list/bot-detail?bot=" + row["family"].ToString();
}
}
this.WriteDebug("isBot: completed");
ret.Add("flag", 0);
ret.Add("is_bot", botInfo);
ret.Add("bot_by_ua", botInfoUA);
ret.Add("bot_by_ip", botInfoIP);
ret.Add("harmony_ua_ip", harmony);
ret.Add("bot_name", botName);
ret.Add("bot_udger_url", botURL);
return(ret);
}
