CMS 3D CMS Logo

Public Member Functions

DDLSpecPar Class Reference

DDLSpecPar processes SpecPar elements. More...

#include <DDLSpecPar.h>

Inheritance diagram for DDLSpecPar:
DDXMLElement

List of all members.

Public Member Functions

 DDLSpecPar (DDLElementRegistry *myreg)
 Constructor.
void processElement (const std::string &name, const std::string &nmspace, DDCompactView &cpv)
 Processing the element.
 ~DDLSpecPar (void)
 Destructor.

Detailed Description

DDLSpecPar processes SpecPar elements.

Author:
Michael Case

DDLSpecPar.h - description ------------------- begin: Tue Nov 21 2001 email: case@ucdhep.ucdavis.edu

This element is used to specify parameters for a part in the detector. PartSelector provides a way to associate Parameters with specific parts of the detector.

Definition at line 23 of file DDLSpecPar.h.


Constructor & Destructor Documentation

DDLSpecPar::DDLSpecPar ( DDLElementRegistry myreg)

Constructor.

Definition at line 26 of file DDLSpecPar.cc.

  : DDXMLElement( myreg )
{}
DDLSpecPar::~DDLSpecPar ( void  )

Destructor.

Definition at line 30 of file DDLSpecPar.cc.

{}

Member Function Documentation

void DDLSpecPar::processElement ( const std::string &  name,
const std::string &  nmspace,
DDCompactView cpv 
) [virtual]

Processing the element.

The processElement method completes any necessary work to process the XML element.

For example, this can be used to call the DDCore to make the geometry in memory. There is a default for this so that if not declared in the inheriting class, no processing is done.

08/13/03 doNotEval for Parameter is based on the value of the eval flag. For String it is always false and for Numeric it is always true. But for "legacy" Parameter, remember, we need to check eval. Default is NOT to evaluate.

1. Check eval flag of each level (SpecParSection, SpecPar and Parameter). 2. Default is the closest specified eval attribute with any value other than "false".

Reimplemented from DDXMLElement.

Definition at line 36 of file DDLSpecPar.cc.

References DDXMLElement::clear(), DCOUT_V, DDXMLElement::getAttributeSet(), DDXMLElement::getDDName(), DDLElementRegistry::getElement(), i, if(), DDI::Singleton< I >::instance(), DDXMLElement::myRegistry_, edm::second(), DDValue::setEvalState(), DDXMLElement::size(), and python::multivaluedict::sort().

{
  DCOUT_V('P',"DDLSpecPar::processElement started");

  // sends the call to the DDD Core OR does nothing if it is a sub-element

  // What I want to do here is the following:
  // 1.  output PartSelector information.
  // 2.  pass the Path and parameters to DDSpecifics
  // for each of the above, use the name of the SpecPar, since DDL does not
  // provide a name for a PartSelector.

  DDXMLElement* myParameter      = myRegistry_->getElement("Parameter");
  DDXMLElement* myNumeric        = myRegistry_->getElement("Numeric");
  DDXMLElement* myString         = myRegistry_->getElement("String");
  DDXMLElement* myPartSelector   = myRegistry_->getElement("PartSelector");
  DDXMLElement* mySpecParSection = myRegistry_->getElement("SpecParSection");

  // Because of namespace magic "!" means namespaces should be provided
  // in the names of the XML elements for the DDD.  So if this is
  // the state/case then we need to force the expression evaluator to 
  // use the namespace of the SpecPar element being processed.
  // --  Michael Case 2008-11-06
  std::string ns(nmspace);
  DDXMLAttribute spatts = getAttributeSet();
  std::string rn = spatts.find("name")->second;
  if ( ns == "!" ) {
    size_t foundColon= rn.find(':');
    if (foundColon != std::string::npos) {
      ns = rn.substr(0,foundColon);
      //       rn = rn.substr(foundColon+1);
    }
  }

  // DDPartSelector name comes from DDLSpecPar (this class, there is no analogue to 
  // DDLSpecPar in DDCore)
  std::vector <std::string> partsels;
  size_t i;

  //    if (getName("name") == "")
  //      {
  //        std::cout << "ERROR: no name for SpecPar" << std::endl;
  //        partsels = myPartSelector->getVectorAttribute("path");
  //        snames = myParameter->getVectorAttribute("name");
  //        std::cout << "\tParameter Names" << std::endl;
  //        size_t i;
  //        for (i = 0; i < snames.size(); ++i)
  //    {
  //      std::cout << "\t\t" << snames[i] << std::endl;
  //    }
  //        std::cout << "\tPart Selectors:" << std::endl;
  //        for (i = 0; i < partsels.size(); ++i)
  //    {
  //      std::cout << "\t\t" << partsels[i] << std::endl;
  //    }
  //      }
  //    else 
  //      {

  //should i keep this? partsels = myPartSelector->getVectorAttribute("path");
  //otherise I have to do this block...
  for (i = 0; i < myPartSelector->size(); ++i)
    partsels.push_back((myPartSelector->getAttributeSet(i).find("path"))->second);
  DDsvalues_type svt;

  // boolean flag to indicate whether the std::vector<DDValuePair> has been evaluated 
  // using the Evaluator
  typedef std::map<std::string, std::pair<bool,std::vector<DDValuePair> > > vvvpType;

  vvvpType vvvp;

  bool doNotEval = true;
  bool doRegex = true;
  {
    // check parent level  
    const DDXMLAttribute & atts = mySpecParSection->getAttributeSet();
    
    if (atts.find("eval") != atts.end() && atts.find("eval")->second == "true")
      doNotEval = false;
    
    if (atts.find("regex") != atts.end() && atts.find("regex")->second == "false")
      doRegex = false;
  }
  {
    // check this level
    const DDXMLAttribute & atts = getAttributeSet();
    
    if (atts.find("eval") != atts.end() && atts.find("eval")->second == "true")
      doNotEval = false;
    else if (atts.find("eval") != atts.end())
      doNotEval = true;
    
    if (atts.find("regex") != atts.end() && atts.find("regex")->second == "false")
      doRegex = false;
    else if (atts.find("regex") != atts.end())
      doRegex = true;
  }
  for (i = 0; i < myParameter->size(); ++i)
  {
    const DDXMLAttribute & atts = myParameter->getAttributeSet(i);
    std::vector <DDValuePair> vvp;
    vvvpType::iterator itv = vvvp.find((atts.find("name")->second));
    if (itv != vvvp.end())
      vvp = itv->second.second;
    double tval = 0.0;
    bool isEvaluated = false;

    // bool notThis =  doNotEval  myParameter->get(std::string("eval"), i) != "true";

    // since eval is an optional attribute, we need to check if it exists for
    // the debug statement.  Unfortunately, I see no way to do this internal
    // to the DCOUT_V statement... ts is a "wasted" variable.
    std::string ts = "** no eval attribute **";
    if (atts.find("eval") != atts.end()) 
      ts = atts.find("eval")->second;
    DCOUT_V('P', std::string("about to process ") << atts.find("value")->second << std::string(" eval = ") << ts);
    if ((atts.find("eval") != atts.end() && atts.find("eval")->second !="false")
        || (atts.find("eval") == atts.end() && !doNotEval))
    { 
      tval = ExprEvalSingleton::instance().eval(ns, atts.find("value")->second);
      isEvaluated=true;
      DCOUT_V('P', std::string("EVALUATED"));
    }
    else
    {
      DCOUT_V('P', std::string("NOT Evaluated"));
    }
      
    DDValuePair vp(atts.find("value")->second, tval);
    vvp.push_back(vp);
    vvvp[atts.find("name")->second] = make_pair(isEvaluated,vvp);
  }

  // Process the String names and values.
  for (i = 0; i < myString->size(); ++i)
  {
    const DDXMLAttribute & atts = myString->getAttributeSet(i);
    std::vector <DDValuePair> vvp;
    vvvpType::iterator itv = vvvp.find(atts.find("name")->second);
    if (itv != vvvp.end())
      vvp = itv->second.second;
    DCOUT_V('P', std::string("about to process String ") << (atts.find("name")->second) << " = " << (atts.find("value")->second));
    DDValuePair vp(atts.find("value")->second, 0.0);
    vvp.push_back(vp);
    vvvp[atts.find("name")->second] = make_pair(false,vvp);
  }
  
  // Process the Numeric names and values.
  for (i = 0; i < myNumeric->size(); ++i)
  {
    const DDXMLAttribute & atts = myNumeric->getAttributeSet(i);
    std::vector <DDValuePair> vvp;
    vvvpType::iterator itv = vvvp.find(atts.find("name")->second);
    if (itv != vvvp.end())
      vvp = itv->second.second;
    DCOUT_V('P', std::string("about to process String ") << (atts.find("name")->second) << " = " << (atts.find("value")->second));
    double tval = ExprEvalSingleton::instance().eval(ns, atts.find("value")->second);
    DCOUT_V('P', std::string("EVALUATED"));
    DDValuePair vp(atts.find("value")->second, tval);
    vvp.push_back(vp);
    vvvp[atts.find("name")->second] = make_pair(true,vvp);
  }
  
  svt.reserve(vvvp.size());
  for (vvvpType::const_iterator it = vvvp.begin(); it != vvvp.end(); ++it)
  {
    DDValue val(it->first, it->second.second);
    bool isEvaluated = it->second.first;
    val.setEvalState(isEvaluated);
    svt.push_back(DDsvalues_Content_type(val,val));      
  }
  std::sort(svt.begin(),svt.end());


  DCOUT_V('p', "DDLSpecPar::processElement\n\tname " << getDDName(nmspace) << "\n\tpartsels.size() = " << myPartSelector->size() << "\n\tsvt " << svt);

  DDSpecifics ds(getDDName(nmspace), 
                 partsels,
                 svt,
                 doRegex);

  myParameter->clear();
  myPartSelector->clear();
  
  // after a SpecPar is done, we can clear
  clear();
  
  DCOUT_V('P',"DDLSpecPar::processElement(...)");
}