libdap++ Updated for version 3.12.1

libdap::UInt32 Class Reference

Holds a 32-bit unsigned integer. More...

#include <UInt32.h>

Inheritance diagram for libdap::UInt32:
Collaboration diagram for libdap::UInt32:

List of all members.

Public Types

typedef stack< BaseType * > btp_stack

Public Member Functions

virtual void add_var (BaseType *bt, Part part=nil)
 Add a variable.
virtual unsigned int buf2val (void **val)
 Reads the class data.
virtual bool check_semantics (string &msg, bool all=false)
 Compare an object's current state with the semantics of its type.
string dataset () const
 Returns the name of the dataset used to create this instance.
virtual bool deserialize (UnMarshaller &um, DDS *dds, bool reuse=false)
 Receive data from the net.
virtual void dump (ostream &strm) const
 dumps information about this object
virtual int element_count (bool leaves=false)
 Count the members of constructor types.
virtual AttrTableget_attr_table ()
virtual BaseTypeget_parent ()
virtual bool is_constructor_type ()
 Returns true if the instance is a constructor (i.e., Structure, Sequence or Grid) type variable.
bool is_dap4 ()
virtual bool is_in_selection ()
 Is this variable part of the current selection?
virtual bool is_simple_type ()
 Returns true if the instance is a numeric, string or URL type variable.
virtual bool is_vector_type ()
 Returns true if the instance is a vector (i.e., array) type variable.
string name () const
 Returns the name of the class instance.
UInt32operator= (const UInt32 &rhs)
virtual bool ops (BaseType *b, int op)
 Evaluate relational operators.
virtual void print_decl (FILE *out, string space=" ", bool print_semi=true, bool constraint_info=false, bool constrained=false)
 Print an ASCII representation of the variable structure.
virtual void print_decl (ostream &out, string space=" ", bool print_semi=true, bool constraint_info=false, bool constrained=false)
 Print an ASCII representation of the variable structure.
virtual void print_val (FILE *out, string space="", bool print_decl_p=true)
 Prints the value of the variable.
virtual void print_val (ostream &out, string space="", bool print_decl_p=true)
 Prints the value of the variable.
virtual void print_xml (ostream &out, string space=" ", bool constrained=false)
virtual void print_xml (FILE *out, string space=" ", bool constrained=false)
virtual void print_xml_writer (XMLWriter &xml, bool constrained=false)
virtual BaseTypeptr_duplicate ()
virtual bool read ()
 Read data into a local buffer.
virtual bool read_p ()
 Has this variable been read?
virtual bool send_p ()
 Should this variable be sent?
virtual bool serialize (ConstraintEvaluator &eval, DDS &dds, Marshaller &m, bool ce_eval=true)
 Move data to the net.
virtual void set_attr_table (const AttrTable &at)
virtual void set_in_selection (bool state)
void set_is_dap4 (const bool v)
virtual void set_name (const string &n)
 Sets the name of the class instance.
virtual void set_parent (BaseType *parent)
virtual void set_read_p (bool state)
 Sets the value of the read_p property.
virtual void set_send_p (bool state)
virtual void set_synthesized_p (bool state)
void set_type (const Type &t)
 Sets the type of the class instance.
virtual bool set_value (dods_uint32 val)
virtual bool synthesized_p ()
virtual string toString ()
virtual void transfer_attributes (AttrTable *at)
Type type () const
 Returns the type of the class instance.
string type_name () const
 Returns the type of the class instance as a string.
 UInt32 (const string &n)
 UInt32 (const string &n, const string &d)
 UInt32 (const UInt32 &copy_from)
virtual unsigned int val2buf (void *val, bool reuse=false)
 Loads class data.
virtual dods_uint32 value () const
virtual BaseTypevar (const string &name="", bool exact_match=true, btp_stack *s=0)
 Returns a pointer to a member of a constructor class.
virtual BaseTypevar (const string &name, btp_stack &s)
virtual unsigned int width (bool constrained=false)
virtual ~UInt32 ()
Abstract Methods
virtual void intern_data (ConstraintEvaluator &eval, DDS &dds)

Protected Member Functions

void m_duplicate (const BaseType &bt)
 Perform a deep copy.

Protected Attributes

dods_uint32 d_buf

Detailed Description

See also:
BaseType

Definition at line 59 of file UInt32.h.


Member Typedef Documentation

typedef stack<BaseType *> libdap::BaseType::btp_stack [inherited]

Definition at line 233 of file BaseType.h.


Constructor & Destructor Documentation

libdap::UInt32::UInt32 ( const string &  n)

The UInt32 constructor accepts the name of the variable to be created.

Parameters:
nA string containing the name of the variable to be created. variable is created

Definition at line 75 of file UInt32.cc.

Referenced by ptr_duplicate().

libdap::UInt32::UInt32 ( const string &  n,
const string &  d 
)

The UInt32 server-side constructor accepts the name of the variable and the dataset name from which this instance is created.

Parameters:
nA string containing the name of the variable to be created.
dA string containing the name of the dataset from which this variable is created

Definition at line 86 of file UInt32.cc.

virtual libdap::UInt32::~UInt32 ( ) [inline, virtual]

Definition at line 67 of file UInt32.h.

libdap::UInt32::UInt32 ( const UInt32 copy_from)

Definition at line 90 of file UInt32.cc.

References d_buf.


Member Function Documentation

void libdap::BaseType::add_var ( BaseType bt,
Part  part = nil 
) [virtual, inherited]

Adds a variable to an instance of a constructor class, such as Array, Structure et cetera. This function is only used by those classes. For constructors with more than one variable, the variables appear in the same order in which they were added (i.e., the order in which add_var() was called). Since this method is only for use by Vectors and Constructors, the BaseType implementation throws InternalErr.

Note:
For the implementation of this method in Structure, Sequence, et c., first copy bt and then insert the copy. If bt is itself a constructor type you must either use the var() method to get a pointer to the actual instance added to *this or you must first add all of bt's children to it before adding it to *this. The implementations should use m_duplicate() to perform a deep copy of bt.
Todo:
We should get rid of the Part parameter and adopt the convention that the first variable is the Array and all subsequent ones are Maps (when dealing with a Grid, the only time Part matters). This would enable several methods to migrate from Structure, Sequence and Grid to Constructor.
Parameters:
btThe variable to be added to this instance. The caller of this method must free memory it allocates for v. This method will make a deep copy of the object pointed to by v.
partThe part of the constructor data to be modified. Only meaningful for Grid variables.
See also:
Part

Reimplemented in libdap::Array, libdap::Constructor, libdap::Grid, and libdap::Vector.

Definition at line 852 of file BaseType.cc.

Referenced by libdap::D4ParserSax2::ddx_end_element(), and libdap::DDXParser::ddx_sax2_end_element().

unsigned int libdap::UInt32::buf2val ( void **  val) [virtual]

Reads the class data into the memory referenced by val. The caller should either allocate enough storage to val to hold the class data or set *val to null. If *val is NULL, memory will be allocated by this function with new(). If the memory is allocated this way, the caller is responsible for deallocating that memory. Array and values for simple types are stored as C would store an array.

Deprecated:
Use value() in the leaf classes.
Parameters:
valA pointer to a pointer to the memory into which the class data will be copied. If the value pointed to is NULL, memory will be allocated to hold the data, and the pointer value modified accordingly. The calling program is responsible for deallocating the memory references by this pointer.
Returns:
The size (in bytes) of the information copied to val.

Implements libdap::BaseType.

Definition at line 167 of file UInt32.cc.

References d_buf, and width().

Here is the call graph for this function:

bool libdap::BaseType::check_semantics ( string &  msg,
bool  all = false 
) [virtual, inherited]

This function checks the class instance for internal consistency. This is important to check for complex constructor classes. For BaseType, an object is semantically correct if it has both a non-null name and type.

For example, an Int32 instance would return FALSE if it had no name or no type defined. A Grid instance might return FALSE for more complex reasons, such as having Map arrays of the wrong size or shape.

This function is used by the DDS class, and will rarely, if ever, be explicitly called by a DODS application program. A variable must pass this test before it is sent, but there may be many other stages in a retrieve operation where it would fail.

Returns:
Returns FALSE when the current state violates some aspect of the type semantics, TRUE otherwise.
Parameters:
msgA returned string, containing a message indicating the source of any problem.
allFor complex constructor types (Grid, Sequence, Structure), this flag indicates whether to check the semantics of the member variables, too.
See also:
DDS::check_semantics

Reimplemented in libdap::Array, libdap::Constructor, libdap::Grid, and libdap::Vector.

Definition at line 1153 of file BaseType.cc.

References libdap::dods_null_c.

Referenced by libdap::Grid::check_semantics(), and libdap::Constructor::check_semantics().

string libdap::BaseType::dataset ( ) const [inherited]

A dataset from which the data is to be read. The meaning of this string will vary among different types of data sources. It may be the name of a data file or an identifier used to read data from a relational database.

Definition at line 275 of file BaseType.cc.

Referenced by libdap::Sequence::read_row(), libdap::Vector::serialize(), libdap::UInt64::serialize(), serialize(), libdap::UInt16::serialize(), libdap::Str::serialize(), libdap::Int8::serialize(), libdap::Int64::serialize(), libdap::Int32::serialize(), libdap::Int16::serialize(), libdap::Grid::serialize(), libdap::Float64::serialize(), libdap::Float32::serialize(), libdap::Constructor::serialize(), and libdap::Byte::serialize().

bool libdap::UInt32::deserialize ( UnMarshaller um,
DDS dds,
bool  reuse = false 
) [virtual]

Receives data from the network connection identified by the source parameter. The data is put into the class data buffer according to the input dds.

This function is only used on the client side of the DODS client/server connection.

Parameters:
umAn UnMarshaller that knows how to deserialize data types
ddsThe Data Descriptor Structure object corresponding to this dataset. See The DODS User Manual for information about this structure. This would have been received from the server in an earlier transmission.
reuseA boolean value, indicating whether the class internal data storage can be reused or not. If this argument is TRUE, the class buffer is assumed to be large enough to hold the incoming data, and it is not reallocated. If FALSE, new storage is allocated. If the internal buffer has not been allocated at all, this argument has no effect.
Returns:
Always returns TRUE.
Exceptions:
Errorwhen a problem reading from the UnMarshaller is found.
See also:
DDS

Implements libdap::BaseType.

Definition at line 142 of file UInt32.cc.

References d_buf, and libdap::UnMarshaller::get_uint32().

Here is the call graph for this function:

void libdap::UInt32::dump ( ostream strm) const [virtual]

Displays the pointer value of this instance and information about this instance.

Parameters:
strmC++ i/o stream to dump the information to
Returns:
void

Reimplemented from libdap::BaseType.

Definition at line 274 of file UInt32.cc.

References d_buf, libdap::DapIndent::Indent(), libdap::DapIndent::LMarg(), and libdap::DapIndent::UnIndent().

Here is the call graph for this function:

int libdap::BaseType::element_count ( bool  leaves = false) [virtual, inherited]

Return a count of the total number of variables in this variable. This is used to count the number of variables held by a constructor variable - for simple type and vector variables it always returns 1.

For compound data types, there are two ways to count members. You can count the members, or you can count the simple members and add that to the count of the compound members. For example, if a Structure contains an Int32 and another Structure that itself contains two Int32 members, the element count of the top-level structure could be two (one Int32 and one Structure) or three (one Int32 by itself and two Int32's in the subsidiary Structure). Use the leaves parameter to control which kind of counting you desire.

Returns:
Returns 1 for simple types. For compound members, the count depends on the leaves argument.
Parameters:
leavesThis parameter is only relevant if the object contains other compound data types. If FALSE, the function counts only the data variables mentioned in the object's declaration. If TRUE, it counts the simple members, and adds that to the sum of the counts for the compound members. This parameter has no effect for simple type variables.

Reimplemented in libdap::Constructor, libdap::Grid, and libdap::Vector.

Definition at line 542 of file BaseType.cc.

Referenced by libdap::Vector::element_count().

AttrTable & libdap::BaseType::get_attr_table ( ) [virtual, inherited]

Get this variable's AttrTable. It's generally a bad idea to return a reference to a contained object, but in this case it seems that building an interface inside BaseType is overkill.

Use the AttrTable methods to manipulate the table.

Definition at line 666 of file BaseType.cc.

Referenced by libdap::D4ParserSax2::ddx_start_element(), libdap::Grid::print_xml_writer(), libdap::D4Group::print_xml_writer(), libdap::Constructor::print_xml_writer(), libdap::BaseType::print_xml_writer(), libdap::Array::print_xml_writer_core(), libdap::Grid::transfer_attributes(), and libdap::BaseType::transfer_attributes().

BaseType * libdap::BaseType::get_parent ( ) [virtual, inherited]

Return a pointer to the Constructor or Vector which holds (contains) this variable. If this variable is at the top level, this method returns null.

Returns:
A BaseType pointer to the variable's parent.

Definition at line 788 of file BaseType.cc.

Referenced by libdap::Sequence::intern_data_for_leaf(), libdap::Sequence::intern_data_parent_part_two(), libdap::Sequence::serialize_leaf(), and libdap::Sequence::serialize_parent_part_two().

void libdap::BaseType::intern_data ( ConstraintEvaluator eval,
DDS dds 
) [virtual, inherited]

Similar to using serialize() and deserialize() together in one object. Data are read as for serialize and those values are stored in the objects as deserialize() does but does not write and then read data to/from a stream.

This method is defined by the various data type classes. It calls the read() abstract method.

Parameters:
evalUse this as the constraint expression evaluator.
ddsThe Data Descriptor Structure object corresponding to this dataset. See The DODS User Manual for information about this structure.

Reimplemented in libdap::Constructor, libdap::Grid, libdap::Sequence, and libdap::Vector.

Definition at line 932 of file BaseType.cc.

References DBG2, libdap::BaseType::name(), libdap::BaseType::read(), libdap::BaseType::read_p(), libdap::DDS::timeout_off(), and libdap::DDS::timeout_on().

Referenced by libdap::Grid::intern_data().

Here is the call graph for this function:

bool libdap::BaseType::is_dap4 ( ) [inline, inherited]

Definition at line 252 of file BaseType.h.

bool libdap::BaseType::is_in_selection ( ) [virtual, inherited]

Does this variable appear in either the selection part or as a function argument in the current constrain expression. If this property is set (true) then implementations of the read() method should read this variable.

Note:
This method does not check, nor does it know about the semantics of, string arguments passed to functions. Those functions might include variable names in strings; they are responsible for reading those variables. See the grid (func_grid_select()) for an example.
See also:
BaseType::read()

Definition at line 740 of file BaseType.cc.

void libdap::BaseType::m_duplicate ( const BaseType bt) [protected, inherited]

Perform a deep copy. Copies the values of bt into *this. Pointers are dereferenced and their values are copied into a newly allocated instance.

Parameters:
btThe source object.

Definition at line 80 of file BaseType.cc.

References DBG2.

Referenced by libdap::BaseType::BaseType(), and libdap::BaseType::operator=().

string libdap::BaseType::name ( ) const [inherited]

Definition at line 254 of file BaseType.cc.

Referenced by libdap::Vector::add_var(), libdap::Vector::add_var_nocopy(), libdap::Grid::check_semantics(), libdap::Constructor::check_semantics(), libdap::Sequence::deserialize(), libdap::DDS::exact_match(), libdap::extract_double_array(), libdap::Vector::intern_data(), libdap::Sequence::intern_data(), libdap::Constructor::intern_data(), libdap::BaseType::intern_data(), libdap::Sequence::intern_data_for_leaf(), libdap::Sequence::intern_data_parent_part_one(), libdap::Sequence::intern_data_parent_part_two(), libdap::Sequence::intern_data_private(), libdap::DDS::leaf_match(), libdap::Constructor::m_exact_match(), libdap::Constructor::m_leaf_match(), libdap::Grid::print_decl(), libdap::Constructor::print_decl(), libdap::Grid::print_xml_writer(), libdap::D4Group::print_xml_writer(), libdap::Constructor::print_xml_writer(), libdap::Array::print_xml_writer_core(), libdap::BaseType::read(), libdap::Sequence::read_row(), libdap::DODSFilter::send_data_ddx(), libdap::Str::serialize(), libdap::Sequence::serialize(), libdap::Sequence::serialize_leaf(), libdap::Sequence::serialize_parent_part_one(), libdap::Sequence::serialize_parent_part_two(), libdap::Sequence::set_leaf_sequence(), libdap::BaseType::set_name(), libdap::BaseType::set_read_p(), libdap::BaseType::set_send_p(), libdap::Grid::transfer_attributes(), libdap::BaseType::transfer_attributes(), libdap::Vector::value(), libdap::rvalue::value_name(), libdap::Vector::var(), libdap::Grid::var(), libdap::Constructor::var(), and libdap::Sequence::var_value().

UInt32 & libdap::UInt32::operator= ( const UInt32 rhs)

Definition at line 102 of file UInt32.cc.

References d_buf.

bool libdap::UInt32::ops ( BaseType b,
int  op 
) [virtual]

This method contains the relational operators used by the constraint expression evaluator in the DDS class. Each class that wants to be able to evaluate relational expressions must overload this function. The implementation in BaseType throws an InternalErr exception. The DAP library classes Byte, ..., Url provide specializations of this method. It is not meaningful for classes such as Array because relational expressions using Array are not supported.

The op argument refers to a table generated by bison from the constraint expression parser. Use statements like the following to correctly interpret its value:

    switch (op) {
        case EQUAL: return i1 == i2;
        case NOT_EQUAL: return i1 != i2;
        case GREATER: return i1 > i2;
        case GREATER_EQL: return i1 >= i2;
        case LESS: return i1 < i2;
        case LESS_EQL: return i1 <= i2;
        case REGEXP: throw Error("Regular expressions are not supported for integer values");
        default: throw Error("Unknown operator");
    }
    

This function is used by the constraint expression evaluator.

Parameters:
bCompare the value of this instance with b.
opAn integer index indicating which relational operator is implied. Choose one from the following: EQUAL, NOT_EQUAL, GREATER, GREATER_EQL, LESS, LESS_EQL, and REGEXP.
Returns:
The boolean value of the comparison.

Reimplemented from libdap::BaseType.

Definition at line 217 of file UInt32.cc.

References d_buf, libdap::dods_byte_c, libdap::dods_float32_c, libdap::dods_float64_c, libdap::dods_int16_c, libdap::dods_int32_c, libdap::dods_int64_c, libdap::dods_int8_c, libdap::dods_uint16_c, libdap::dods_uint32_c, libdap::dods_uint64_c, libdap::BaseType::read(), libdap::BaseType::read_p(), libdap::BaseType::type(), and value().

Here is the call graph for this function:

void libdap::BaseType::print_decl ( FILE *  out,
string  space = "    ",
bool  print_semi = true,
bool  constraint_info = false,
bool  constrained = false 
) [virtual, inherited]

Write the variable's declaration in a C-style syntax. This function is used to create textual representation of the Data Descriptor Structure (DDS). See The DODS User Manual for information about this structure.

A simple array declaration might look like this:

    Float64 lat[lat = 180];
    

While a more complex declaration (for a Grid, in this case), would look like this:

    Grid {
    ARRAY:
    Int32 sst[time = 404][lat = 180][lon = 360];
    MAPS:
    Float64 time[time = 404];
    Float64 lat[lat = 180];
    Float64 lon[lon = 360];
    } sst;
    
Parameters:
outThe output stream on which to print the declaration.
spaceEach line of the declaration will begin with the characters in this string. Usually used for leading spaces.
print_semiA boolean value indicating whether to print a semicolon at the end of the declaration.
constraint_infoA boolean value indicating whether constraint information is to be printed with the declaration. If the value of this parameter is TRUE, print_decl() prints the value of the variable's send_p() flag after the declaration.
constrainedIf this boolean value is TRUE, the variable's declaration is only printed if is the send_p() flag is TRUE. If a constraint expression is in place, and this variable is not requested, the send_p() flag is FALSE.
See also:
DDS
DDS::CE

Reimplemented in libdap::Array, libdap::Constructor, and libdap::Grid.

Definition at line 985 of file BaseType.cc.

Referenced by libdap::DODSFilter::functional_constraint(), libdap::Grid::print_decl(), libdap::Array::print_decl(), libdap::UInt64::print_val(), print_val(), libdap::UInt16::print_val(), libdap::Str::print_val(), libdap::Int8::print_val(), libdap::Int64::print_val(), libdap::Int32::print_val(), libdap::Int16::print_val(), libdap::Float64::print_val(), libdap::Float32::print_val(), and libdap::Byte::print_val().

void libdap::BaseType::print_decl ( ostream out,
string  space = "    ",
bool  print_semi = true,
bool  constraint_info = false,
bool  constrained = false 
) [virtual, inherited]

Write the variable's declaration in a C-style syntax. This function is used to create textual representation of the Data Descriptor Structure (DDS). See The DODS User Manual for information about this structure.

A simple array declaration might look like this:

    Float64 lat[lat = 180];
    

While a more complex declaration (for a Grid, in this case), would look like this:

    Grid {
    ARRAY:
    Int32 sst[time = 404][lat = 180][lon = 360];
    MAPS:
    Float64 time[time = 404];
    Float64 lat[lat = 180];
    Float64 lon[lon = 360];
    } sst;
    
Parameters:
outThe output stream on which to print the declaration.
spaceEach line of the declaration will begin with the characters in this string. Usually used for leading spaces.
print_semiA boolean value indicating whether to print a semicolon at the end of the declaration.
constraint_infoA boolean value indicating whether constraint information is to be printed with the declaration. If the value of this parameter is TRUE, print_decl() prints the value of the variable's send_p() flag after the declaration.
constrainedIf this boolean value is TRUE, the variable's declaration is only printed if is the send_p() flag is TRUE. If a constraint expression is in place, and this variable is not requested, the send_p() flag is FALSE.
See also:
DDS
DDS::CE

Reimplemented in libdap::Array, libdap::Constructor, and libdap::Grid.

Definition at line 1036 of file BaseType.cc.

References libdap::id2www(), libdap::BaseType::send_p(), and libdap::BaseType::type_name().

Here is the call graph for this function:

void libdap::UInt32::print_val ( FILE *  out,
string  space = "",
bool  print_decl_p = true 
) [virtual]

Prints the value of the variable, with its declaration. This function is primarily intended for debugging DODS applications. However, it can be overloaded and used to do some useful things. Take a look at the asciival and writeval clients, both of which overload this to output the values of variables in different ways.

Parameters:
outThe output stream on which to print the value.
spaceThis value is passed to the print_decl() function, and controls the leading spaces of the output.
print_decl_pA boolean value controlling whether the variable declaration is printed as well as the value.

Implements libdap::BaseType.

Definition at line 198 of file UInt32.cc.

void libdap::UInt32::print_val ( ostream out,
string  space = "",
bool  print_decl_p = true 
) [virtual]

Prints the value of the variable, with its declaration. This function is primarily intended for debugging DODS applications. However, it can be overloaded and used to do some useful things. Take a look at the asciival and writeval clients, both of which overload this to output the values of variables in different ways.

Parameters:
outThe output ostream on which to print the value.
spaceThis value is passed to the print_decl() function, and controls the leading spaces of the output.
print_decl_pA boolean value controlling whether the variable declaration is printed as well as the value.

Implements libdap::BaseType.

Definition at line 206 of file UInt32.cc.

References d_buf, and libdap::BaseType::print_decl().

Here is the call graph for this function:

void libdap::BaseType::print_xml ( ostream out,
string  space = "    ",
bool  constrained = false 
) [virtual, inherited]

Write the XML representation of this variable. This method is used to build the DDX XML response.

Parameters:
outDestination output stream
spaceUse this to indent child declarations. Default is "".
constrainedIf true, only print this if it's part part of the current projection. Default is False.
Deprecated:

Reimplemented in libdap::Array, libdap::Constructor, and libdap::Grid.

Definition at line 1080 of file BaseType.cc.

References XMLWriter::get_doc(), and libdap::BaseType::print_xml_writer().

Here is the call graph for this function:

void libdap::BaseType::print_xml ( FILE *  out,
string  space = "    ",
bool  constrained = false 
) [virtual, inherited]

Write the XML representation of this variable. This method is used to build the DDX XML response.

Parameters:
outDestination.
spaceUse this to indent child declarations. Default is "".
constrainedIf true, only print this if it's part part of the current projection. Default is False.
Deprecated:

Reimplemented in libdap::Array, libdap::Constructor, and libdap::Grid.

Definition at line 1065 of file BaseType.cc.

References XMLWriter::get_doc(), XMLWriter::get_doc_size(), and libdap::BaseType::print_xml_writer().

Here is the call graph for this function:

void libdap::BaseType::print_xml_writer ( XMLWriter xml,
bool  constrained = false 
) [virtual, inherited]

Write the XML representation of this variable. This method is used to build the DDX XML response.

Parameters:
outDestination output stream
spaceUse this to indent child declarations. Default is "".
constrainedIf true, only print this if it's part part of the current projection. Default is False.

Reimplemented in libdap::Array, libdap::Constructor, libdap::D4Group, and libdap::Grid.

Definition at line 1094 of file BaseType.cc.

References libdap::BaseType::get_attr_table(), XMLWriter::get_writer(), libdap::AttrTable::print_xml_writer(), libdap::BaseType::send_p(), and libdap::BaseType::type_name().

Referenced by libdap::BaseType::print_xml(), and libdap::Array::print_xml_writer_core().

Here is the call graph for this function:

BaseType * libdap::UInt32::ptr_duplicate ( ) [virtual]

Clone this instance. Allocate a new instance and copy *this into it. This method must perform a deep copy.

Note:
This method should not copy data values, but must copy all other fields in the object.
Returns:
A newly allocated copy of this.

Implements libdap::BaseType.

Definition at line 96 of file UInt32.cc.

References UInt32().

Here is the call graph for this function:

bool libdap::BaseType::read ( ) [virtual, inherited]

This method should be implemented for each of the data type classes (Byte, ..., Grid) when using the DAP class library to build a server. This method is only for DAP servers. The library provides a default definition here which throws an InternalErr exception unless the read_p property has been set. In that case it returns false, indicating that all the data have been read. The latter case can happen when building a constant value that needs to be passed to a function. The variable/constant is loaded with a value when it is created.

When implementing a new DAP server, the Byte, ..., Grid data type classes are usually specialized. In each of those specializations read() should be defined to read values from the data source and store them in the object's local buffer. The read() method is called by other methods in this library. When writing read(), follow these rules:

  • read() should throw Error if it encounters an error. The message should be verbose enough to be understood by someone running a client on a different machine.
  • The value(s) should be read if and only if either send_p() or is_in_selection() return true. If neither of these return true, the value(s) should not be read. This is important when writing read() for a Constructor type such as Grid where a client may ask for only the map vectors (and thus reading the much larger Array part is not needed).
  • For each specialization of read(), the method should first test the value of the read_p property (using the read_p() method) and read values only if the value of read_p() is false. Once the read() method reads data and stores it in the instance, it must set the value of the read_p property to true using set_read_p(). If your read() methods fail to do this data may not serialize correctly.
  • The Array::read() and Grid::read() methods should take into account any restrictions on Array sizes.
  • If you are writing Sequence::read(), be sure to check the documentation for Sequence::read_row() and Sequence::serialize() so you understand how Sequence::read() is being called.
  • For Sequence::read(), your specialization must correctly manage the unsent_data property and row count in addition to the read_p property (handle the read_p property as describe above). For a Sequence to serialize correctly, once all data from the Sequence has been read, unsent_data property must be set to false (use Sequence::set_unsent_data()). Also, at that time the row number counter must be reset (use Sequence::reset_row_counter()). Typically the correct time to set unsent_data to false and reset the row counter is the time when Sequence::read() return false indicating that all the data for the Sequence have been read. Failure to handle these tasks will break serialization of nested Sequences. Note that when Sequence::read() returns with a result of true (indicating there is more data to send, the value of the unsent_data property should be true.

    Also, if you server must handle nested sequences, be sure to read about subclassing set_read_p().

Returns:
The return value of this method for all types except Sequence should always be false. Sequences should return true to indicate more values remain in the Sequence, false to indicate no more values remain. (see Sequence::serialize() and Sequence::read_row()).
See also:
BaseType
Sequence

Reimplemented in libdap::Constructor.

Definition at line 923 of file BaseType.cc.

References libdap::BaseType::name().

Referenced by libdap::Vector::intern_data(), libdap::BaseType::intern_data(), libdap::UInt64::ops(), ops(), libdap::UInt16::ops(), libdap::Str::ops(), libdap::Int8::ops(), libdap::Int64::ops(), libdap::Int32::ops(), libdap::Int16::ops(), libdap::Float64::ops(), libdap::Float32::ops(), libdap::Byte::ops(), libdap::Vector::serialize(), libdap::UInt64::serialize(), serialize(), libdap::UInt16::serialize(), libdap::Str::serialize(), libdap::Int8::serialize(), libdap::Int64::serialize(), libdap::Int32::serialize(), libdap::Int16::serialize(), libdap::Float64::serialize(), libdap::Float32::serialize(), and libdap::Byte::serialize().

Here is the call graph for this function:

bool libdap::BaseType::send_p ( ) [virtual, inherited]

Returns the state of the send_p property. If true, this variable should be sent to the client, if false, it should not. If no constraint expression (CE) has been evaluated, this property is true for all variables in a data source (i.e., for all the variables listed in a DDS). If a CE has been evaluated, this property is true only for those variables listed in the projection part of the CE.

Returns:
True if the variable should be sent to the client, false otherwise.

Definition at line 638 of file BaseType.cc.

Referenced by libdap::Grid::components(), libdap::Grid::intern_data(), libdap::Grid::print_decl(), libdap::Constructor::print_decl(), libdap::BaseType::print_decl(), libdap::Array::print_decl(), libdap::Grid::print_val(), libdap::Grid::print_xml_writer(), libdap::D4Group::print_xml_writer(), libdap::Constructor::print_xml_writer(), libdap::BaseType::print_xml_writer(), libdap::Array::print_xml_writer_core(), libdap::Grid::projection_yields_grid(), libdap::Grid::serialize(), and libdap::Grid::width().

bool libdap::UInt32::serialize ( ConstraintEvaluator eval,
DDS dds,
Marshaller m,
bool  ce_eval = true 
) [virtual]

Sends the data from the indicated (local) dataset through the connection identified by the sink parameter. If the data is not already incorporated into the DDS object, read the data from the dataset.

This function is only used on the server side of the client/server connection, and is generally only called from the ResponseBuilder functions. It has no BaseType implementation; each datatype child class supplies its own implementation.

Parameters:
evalUse this as the constraint expression evaluator.
ddsThe Data Descriptor Structure object corresponding to this dataset. See The DODS User Manual for information about this structure.
mA marshaller used to serialize data types
ce_evalA boolean value indicating whether to evaluate the DODS constraint expression that may accompany this dataset. The constraint expression is stored in dds.
Returns:
This method always returns true. Older versions used the return value to signal success or failure.
Exceptions:
InternalErr.
Error.
See also:
DDS

Implements libdap::BaseType.

Definition at line 121 of file UInt32.cc.

References d_buf, libdap::BaseType::dataset(), libdap::ConstraintEvaluator::eval_selection(), libdap::Marshaller::put_uint32(), libdap::BaseType::read(), libdap::BaseType::read_p(), libdap::DDS::timeout_off(), and libdap::DDS::timeout_on().

Here is the call graph for this function:

void libdap::BaseType::set_attr_table ( const AttrTable at) [virtual, inherited]

Set this variable's attribute table.

Parameters:
atSource of the attributes.

Definition at line 674 of file BaseType.cc.

void libdap::BaseType::set_in_selection ( bool  state) [virtual, inherited]

Set the in_selection property to state. This property indicates that the variable is used as a parameter to a constraint expression function or that it appears as an argument in a selection sub-expression. If set (true), implementations of the BaseType::read() method should read this variable.

Parameters:
stateSet the in_selection property to this state.
See also:
BaseType::read()
BaseType::is_in_selection() for more information.

Reimplemented in libdap::Constructor, and libdap::Grid.

Definition at line 755 of file BaseType.cc.

Referenced by libdap::Grid::set_in_selection().

void libdap::BaseType::set_is_dap4 ( const bool  v) [inline, inherited]

Definition at line 253 of file BaseType.h.

void libdap::BaseType::set_name ( const string &  n) [virtual, inherited]

Definition at line 261 of file BaseType.cc.

References libdap::BaseType::name(), and libdap::www2id().

Referenced by libdap::Vector::add_var(), libdap::Vector::add_var_nocopy(), libdap::Array::print_xml_writer_core(), and libdap::Vector::set_name().

Here is the call graph for this function:

void libdap::BaseType::set_parent ( BaseType parent) [virtual, inherited]

Set the parent property for this variable.

Note:
Added ability to set parent to null. 10/19/12 jhrg
Parameters:
parentPointer to the Constructor of Vector parent variable or null if the variable has no parent (if it is at the top-level of a DAP2/3 DDS).
Exceptions:
InternalErrthrown if called with anything other than a Constructor, Vector or Null.

Definition at line 770 of file BaseType.cc.

Referenced by libdap::Vector::_duplicate(), libdap::Grid::add_map(), libdap::Vector::add_var(), libdap::Grid::add_var(), libdap::Constructor::add_var(), libdap::Vector::add_var_nocopy(), libdap::Grid::add_var_nocopy(), libdap::Constructor::add_var_nocopy(), libdap::DDXParser::ddx_end_document(), libdap::Grid::m_duplicate(), libdap::Constructor::m_duplicate(), libdap::Grid::prepend_map(), libdap::Grid::set_array(), and libdap::Vector::Vector().

void libdap::BaseType::set_read_p ( bool  state) [virtual, inherited]

Sets the value of the read_p property. This indicates that the value(s) of this variable has/have been read. An implementation of the read() method would typically use this to set the read_p property to true.

Note:
For most of the types the default implementation of this method is fine. However, if you're building a server which must handle data represented using nested sequences, then you may need to provide a specialization of Sequence::set_read_p(). By default Sequence::set_read_() recursively sets the read_p property for all child variables to state. For servers where one Sequence reads an outer set of values and another reads an inner set, this is cumbersome. In such a case, it is easier to specialize Sequence::set_read_p() so that it does not recursively set the read_p property for the inner Sequence. Be sure to see the documentation for the read() method!
For synthesized variables, this method does nothing. Thus, if a synthesized variable is added to a Sequence, the Sequence can iteratively reset the read_p property without affecting the value of that property for the synthesized variable. That's important since a synthesized variable's value is calculated, not read.
Todo:
Look at making synthesized variables easier to implement and at making them more integrated into the overall CE evaluation process. Maybe the code that computes the synthesized var's value should be in the that variable's read() method? This might provide a way to get rid of the awkward 'projection functions' by replacing them with real children of BaseType. It would also provide a way to clean up the way the synthesized_p prop intrudes on the read_p prop.
See also:
BaseType::read()
Parameters:
stateSet the read_p property to this state.

Reimplemented in libdap::Constructor, libdap::Grid, and libdap::Vector.

Definition at line 618 of file BaseType.cc.

References DBG2, and libdap::BaseType::name().

Referenced by libdap::Vector::set_read_p(), libdap::Grid::set_read_p(), libdap::UInt64::set_value(), set_value(), libdap::UInt16::set_value(), libdap::Str::set_value(), libdap::Int8::set_value(), libdap::Int64::set_value(), libdap::Int32::set_value(), libdap::Int16::set_value(), libdap::Float64::set_value(), libdap::Float32::set_value(), and libdap::Byte::set_value().

Here is the call graph for this function:

void libdap::BaseType::set_send_p ( bool  state) [virtual, inherited]

Sets the value of the send_p flag. This function is meant to be called from within the constraint evaluator of other code which determines that this variable should be returned to the client. Data are ready to be sent when both the d_is_send and d_is_read flags are set to TRUE.

Parameters:
stateThe logical state to set the send_p flag.

Reimplemented in libdap::Constructor, libdap::Grid, and libdap::Vector.

Definition at line 652 of file BaseType.cc.

References DBG2, and libdap::BaseType::name().

Referenced by libdap::DDS::mark(), libdap::DODSFilter::send_data_ddx(), libdap::Vector::set_send_p(), and libdap::Grid::set_send_p().

Here is the call graph for this function:

void libdap::BaseType::set_synthesized_p ( bool  state) [virtual, inherited]

Set the synthesized flag. Before setting this flag be sure to set the read_p() state. Once this flag is set you cannot alter the state of the read_p flag!

See also:
synthesized_p()

Definition at line 562 of file BaseType.cc.

void libdap::BaseType::set_type ( const Type t) [inherited]

Definition at line 289 of file BaseType.cc.

Referenced by libdap::Url::Url().

bool libdap::UInt32::set_value ( dods_uint32  val) [virtual]

Definition at line 189 of file UInt32.cc.

References d_buf, and libdap::BaseType::set_read_p().

Here is the call graph for this function:

bool libdap::BaseType::synthesized_p ( ) [virtual, inherited]

Returns true if the variable is a synthesized variable. A synthesized variable is one that is added to the dataset by the server (usually with a `projection function'.

Definition at line 551 of file BaseType.cc.

string libdap::BaseType::toString ( ) [virtual, inherited]

Write out the object's internal fields in a string. To be used for debugging when regular inspection w/ddd or gdb isn't enough.

Returns:
A string which shows the object's internal stuff.

Reimplemented in libdap::Sequence.

Definition at line 205 of file BaseType.cc.

References libdap::BaseType::type_name().

Here is the call graph for this function:

void libdap::BaseType::transfer_attributes ( AttrTable at_container) [virtual, inherited]

Transfer attributes from a DAS object into this variable. Because of the rough history of the DAS object and the way that various server code built the DAS, this is necessarily a heuristic process. The intent is that this method will be overridden by handlers that need to look for certain patterns in the DAS (e.g., hdf4's odd variable_dim_n; where n = 0, 1, 2, ...) attribute containers.

There should be a one-to-one mapping between variables and attribute containers. However, in some cases one variable has attributes spread across several top level containers and in some cases one container is used by several variables

Note:
This method is technically unnecessary because a server (or client) can easily add attributes directly using the DDS::get_attr_table or BaseType::get_attr_table methods and then poke values in using any of the methods AttrTable provides. This method exists to ease the transition to DDS objects which contain attribute information for the existing servers (Since they all make DAS objects separately from the DDS). They could be modified to use the same AttrTable methods but operate on the AttrTable instances in a DDS/BaseType instead of those in a DAS.
Parameters:
at_containerTransfer attributes from this container.
Returns:
void

Reimplemented in libdap::Grid.

Definition at line 705 of file BaseType.cc.

References libdap::AttrTable::append_attr(), libdap::AttrTable::append_container(), libdap::AttrTable::attr_begin(), libdap::Attr_container, libdap::AttrTable::attr_end(), DBG, libdap::BaseType::get_attr_table(), libdap::AttrTable::get_attr_table(), libdap::AttrTable::get_attr_type(), libdap::AttrTable::get_attr_vector(), libdap::AttrTable::get_name(), libdap::AttrTable::get_type(), libdap::BaseType::name(), and libdap::AttrTable::set_is_global_attribute().

Referenced by libdap::Grid::transfer_attributes().

Here is the call graph for this function:

Type libdap::BaseType::type ( ) const [inherited]
unsigned int libdap::UInt32::val2buf ( void *  val,
bool  reuse = false 
) [virtual]

Store the value pointed to by val in the object's internal buffer. This function does not perform any checks, so users must be sure that the thing pointed to can actually be stored in the object's buffer.

Only simple objects (Int, Float, Byte, and so on) and arrays of these simple objects may be stored using this function. To put data into more complex constructor types, use the functions provided by that class.

Deprecated:
Use set_value() in the leaf classes.
Parameters:
valA pointer to the data to be inserted into the class data buffer.
reuseA boolean value, indicating whether the class internal data storage can be reused or not. If this argument is TRUE, the class buffer is assumed to be large enough to hold the incoming data, and it is not reallocated. If FALSE, new storage is allocated. If the internal buffer has not been allocated at all, this argument has no effect. This is currently used only in the Vector class.
Returns:
The size (in bytes) of the information copied from val.
See also:
Grid
Vector::val2buf

Implements libdap::BaseType.

Definition at line 150 of file UInt32.cc.

References d_buf, and width().

Here is the call graph for this function:

dods_uint32 libdap::UInt32::value ( ) const [virtual]

Definition at line 183 of file UInt32.cc.

References d_buf.

Referenced by ops().

BaseType * libdap::BaseType::var ( const string &  name = "",
bool  exact_match = true,
btp_stack s = 0 
) [virtual, inherited]

Returns a pointer to the contained variable in a composite class. The composite classes are those made up of aggregated simple data types. Array, Grid, and Structure are composite types, while Int and Float are simple types. This function is only used by composite classes. The BaseType implementation always returns null.

Several of the subclasses provide alternate access methods that make sense for that particular data type. For example, the Array class defines a *var(int i) method that returns the ith entry in the Array data, and the Structure provides a *var(Vars_iter) function using a pseudo-index to access the different members of the structure.

Parameters:
nameThe name of the class member. Defaults to ""
exact_matchTrue if only interested in variables whose full names match n exactly. If false, returns the first variable whose name matches name. For example, if name is x and point.x is a variable, then var("x", false) would return a BaseType pointer to point.x. If exact_match was true then name would need to be "point.x" for var to return that pointer. This feature simplifies constraint expressions for datasets which have complex, nested, constructor variables. Defaults to true.
sRecord the path to name. Defaults to null, in which case it is not used.
Returns:
A pointer to the member named in the n argument. If no name is given, the function returns the first (only) variable. For example, an Array has only one variable, while a Structure can have many.

Reimplemented in libdap::Constructor, libdap::Grid, and libdap::Vector.

Definition at line 795 of file BaseType.cc.

Referenced by libdap::DDS::exact_match(), libdap::DDS::leaf_match(), libdap::Constructor::m_exact_match(), libdap::Constructor::m_leaf_match(), and libdap::Vector::var().

BaseType * libdap::BaseType::var ( const string &  name,
btp_stack s 
) [virtual, inherited]

This version of var(...) searches for name and returns a pointer to the BaseType object if found. It uses the same search algorithm as BaseType::var(const string &, bool, btp_stack *) when exact_match is false. In addition to returning a pointer to the variable, it pushes onto s a BaseType pointer to each constructor type that ultimately contains name.

Note:
The BaseType implementation always returns null. There are no default values for the parameters. If var() is called w/o any params, the three parameter version will be used.
Deprecated:
This method is deprecated because it tries first to use exact_match and, if that fails, then tries leaf_match. It's better to use the alternate form of var(...) and specify exactly what you'd like to do.
Returns:
A pointer to the named variable.

Reimplemented in libdap::Constructor, libdap::Grid, and libdap::Vector.

Definition at line 817 of file BaseType.cc.

unsigned int libdap::UInt32::width ( bool  constrained = false) [virtual]

This version of width simply returns the same thing as width() for simple types and Arrays. For Constructors, it needs to be specialized. This is partly due to an inconsistency in the way Vector::width() is implemented. That method uses the constrained size of the array (while the Constructor versions do not take the constraint into account).

Parameters:
constrainedIf true, return the size after applying a constraint.
Returns:
The number of bytes used by the variable.

Reimplemented from libdap::BaseType.

Definition at line 115 of file UInt32.cc.

Referenced by buf2val(), and val2buf().


Member Data Documentation


The documentation for this class was generated from the following files: