Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

ClassCompiler: Validate field types -> Icinga type names correctly #9514

Merged
merged 1 commit into from
Sep 6, 2022

Conversation

yhabteab
Copy link
Member

@yhabteab yhabteab commented Sep 5, 2022

When the classcompiler is validating/transforming field types -> Icinga type names, it is currently returning
Icinga Number type for field type of bool, which is actually wrong. This PR ensures to always transform
into the correct Icinga type names.

When the classcompiler is validating/transforming field types -> Icinga type names, it is currently returning
Icinga `Number` type for field type of `bool`, which is actually wrong. This PR ensures to always transform
into the correct Icinga type names.
@yhabteab yhabteab added bug Something isn't working area/configuration DSL, parser, compiler, error handling labels Sep 5, 2022
@yhabteab yhabteab requested a review from julianbrost September 5, 2022 15:25
@yhabteab yhabteab self-assigned this Sep 5, 2022
@cla-bot cla-bot bot added the cla/signed label Sep 5, 2022
@julianbrost
Copy link
Contributor

There are two uses of FieldTypeToIcingaName:

  1. /* GetFieldInfo */
    m_Header << "\t" << "Field GetFieldInfo(int id) const override;" << std::endl;
    m_Impl << "Field TypeImpl<" << klass.Name << ">::GetFieldInfo(int id) const" << std::endl
    << "{" << std::endl;
    if (!klass.Parent.empty())
    m_Impl << "\t" << "int real_id = id - " << klass.Parent << "::TypeInstance->GetFieldCount();" << std::endl
    << "\t" << "if (real_id < 0) { return " << klass.Parent << "::TypeInstance->GetFieldInfo(id); }" << std::endl;
    if (!klass.Fields.empty()) {
    m_Impl << "\t" << "switch (";
    if (!klass.Parent.empty())
    m_Impl << "real_id";
    else
    m_Impl << "id";
    m_Impl << ") {" << std::endl;
    size_t num = 0;
    for (const Field& field : klass.Fields) {
    std::string ftype = FieldTypeToIcingaName(field, false);
    std::string nameref;
    if (field.Type.IsName)
    nameref = "\"" + field.Type.TypeName + "\"";
    else
    nameref = "nullptr";
    m_Impl << "\t\t" << "case " << num << ":" << std::endl
    << "\t\t\t" << "return {" << num << ", \"" << ftype << "\", \"" << field.Name << "\", \"" << (field.NavigationName.empty() ? field.Name : field.NavigationName) << "\", " << nameref << ", " << field.Attributes << ", " << field.Type.ArrayRank << "};" << std::endl;
    num++;
    }
    m_Impl << "\t\t" << "default:" << std::endl
    << "\t\t";
    }
    m_Impl << "\t" << "throw std::runtime_error(\"Invalid field ID.\");" << std::endl;
    if (!klass.Fields.empty())
    m_Impl << "\t" << "}" << std::endl;
    m_Impl << "}" << std::endl << std::endl;

    This affects the GetFieldInfo function in the generated code. This one has a few users, but the change only affects the value of the TypeName member, which has fewer users (git grep '\bTypeName\b', excluding matches in tools/mkclass/ which use a different TypeName than the one in use in code generated by mkclass):
    1. icinga2/lib/base/object.cpp

      Lines 140 to 151 in 3a8abdc

      try {
      SetField(fid, value);
      } catch (const boost::bad_lexical_cast&) {
      Field fieldInfo = type->GetFieldInfo(fid);
      Type::Ptr ftype = Type::GetByName(fieldInfo.TypeName);
      BOOST_THROW_EXCEPTION(ScriptError("Attribute '" + field + "' cannot be set to value of type '" + value.GetTypeName() + "', expected '" + ftype->GetName() + "'", debugInfo));
      } catch (const std::bad_cast&) {
      Field fieldInfo = type->GetFieldInfo(fid);
      Type::Ptr ftype = Type::GetByName(fieldInfo.TypeName);
      BOOST_THROW_EXCEPTION(ScriptError("Attribute '" + field + "' cannot be set to value of type '" + value.GetTypeName() + "', expected '" + ftype->GetName() + "'", debugInfo));
      }
      }

      Used in an error message, so probably even better to use the proper type name.
    2. if (strcmp(field.TypeName, "int") != 0 && strcmp(field.TypeName, "double") != 0
      && strcmp(field.TypeName, "bool") != 0 && strcmp(field.TypeName, "String") != 0)
      continue;

      The whole functions seems to be unused but doesn't check for "Number"/"Boolean" anyway.
    3. if (strcmp(fieldInfo.TypeName, "Timestamp") == 0)

      Only checks for "Timestamp", so also unaffected by the change.
    4. fieldInfo->Set("type", field.TypeName);

      Changes the returned type information in /v1/types, probably also better to include the correct type name.
  2. std::string ftype = FieldTypeToIcingaName(field, true);
    if (ftype == "Value") {
    m_Impl << "\t" << "if (" << valName << ".IsObjectType<Function>()) {" << std::endl
    << "\t\t" << "Function::Ptr func = " << valName << ";" << std::endl
    << "\t\t" << "if (func->IsDeprecated())" << std::endl
    << "\t\t\t" << "Log(LogWarning, \"" << klass.Name << "\") << \"Attribute '" << field.Name << R"(' for object '" << dynamic_cast<ConfigObject *>(this)->GetName() << "' of type '" << dynamic_cast<ConfigObject *>(this)->GetReflectionType()->GetName() << "' is set to a deprecated function: " << func->GetName();)" << std::endl
    << "\t" << "}" << std::endl << std::endl;
    }
    if (field.Type.IsName) {
    if (field.Type.ArrayRank > 0) {
    m_Impl << "\t\t\t" << "if (!" << valName << ".IsEmpty() && ";
    m_Impl << "!" << valName << ".IsString())" << std::endl
    << "\t\t\t\t" << "BOOST_THROW_EXCEPTION(ValidationError(dynamic_cast<ConfigObject *>(this), { \""
    << field.Name << R"(" }, "It is not allowed to specify '" + )" << valName << R"( + "' of type '" + )"
    << valName << ".GetTypeName()" << R"( + "' as ')" << field.Type.TypeName
    << "' name. Expected type of '" << field.Type.TypeName << "' name: 'String'.\"));" << std::endl;
    }
    m_Impl << "\t\t\t" << "if (";
    if (field.Type.ArrayRank > 0)
    m_Impl << valName << ".IsEmpty() || ";
    else
    m_Impl << "!" << valName << ".IsEmpty() && ";
    m_Impl << "!utils.ValidateName(\"" << field.Type.TypeName << "\", " << valName << "))" << std::endl
    << "\t\t\t\t" << "BOOST_THROW_EXCEPTION(ValidationError(dynamic_cast<ConfigObject *>(this), { \"" << field.Name << R"(" }, "Object '" + )" << valName << R"( + "' of type ')" << field.Type.TypeName
    << "' does not exist.\"));" << std::endl;
    } else if (field.Type.ArrayRank > 0 && (ftype == "Number" || ftype == "Boolean")) {
    m_Impl << "\t" << "try {" << std::endl
    << "\t\t" << "Convert::ToDouble(" << valName << ");" << std::endl
    << "\t" << "} catch (const std::invalid_argument&) {" << std::endl
    << "\t\t" << "BOOST_THROW_EXCEPTION(ValidationError(dynamic_cast<ConfigObject *>(this), { \"" << field.Name << R"(", "Array element '" + " << valName << " + "' of type '" + " << valName << ".GetReflectionType()->GetName() + "' is not valid here; expected type ')" << ftype << "'.\"));" << std::endl
    << "\t" << "}" << std::endl;
    }
    if (field.Type.ArrayRank > 0) {
    m_Impl << "\t\t" << "}" << std::endl
    << "\t" << "}" << std::endl;
    }
    m_Impl << "}" << std::endl << std::endl;

    The following is the only case where returning "Boolean" or "Number" would make a difference:
    } else if (field.Type.ArrayRank > 0 && (ftype == "Number" || ftype == "Boolean")) {

    But this is dead code at the moment anyways, i.e. no *.ti file defines an array of Number/Boolean anyways, so you can even put abort() into that branch and things still compile. Thus, no change in behavior here.

Copy link
Contributor

@julianbrost julianbrost left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Observable changes by this PR (highlighted in bold in my previous comment) are totally fine (even an improvement), so let's do this.

@julianbrost julianbrost merged commit 86b63a5 into master Sep 6, 2022
@icinga-probot icinga-probot bot deleted the bugfix/validate-array-types-correctly branch September 6, 2022 16:15
@Al2Klimov Al2Klimov added this to the 2.14.0 milestone Sep 7, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
area/configuration DSL, parser, compiler, error handling bug Something isn't working cla/signed
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants