diff options
Diffstat (limited to 'ace/CDR_Base.h')
| -rw-r--r-- | ace/CDR_Base.h | 140 |
1 files changed, 74 insertions, 66 deletions
diff --git a/ace/CDR_Base.h b/ace/CDR_Base.h index 159334dd9fb..9a8c16cccfe 100644 --- a/ace/CDR_Base.h +++ b/ace/CDR_Base.h @@ -1,31 +1,29 @@ // -*- C++ -*- -// -// $Id$ - -// ==================================================================== -// -// = LIBRARY -// ACE -// -// = FILENAME -// CDR_Base.h -// -// = DESCRIPTION -// ACE Common Data Representation (CDR) basic types. -// -// The current implementation assumes that the host has 1-byte, -// 2-byte and 4-byte integral types, and that it has single -// precision and double precision IEEE floats. -// Those assumptions are pretty good these days, with Crays being -// the only known exception. -// -// = AUTHORS -// Aniruddha Gokhale <gokhale@cs.wustl.edu> and Carlos O'Ryan -// <coryan@cs.wustl.edu> for the original implementation in TAO. -// ACE version by Jeff Parsons <parsons@cs.wustl.edu> -// and Istvan Buki <istvan.buki@euronet.be>. -// -// ===================================================================== + +//============================================================================= +/** + * @file CDR_Base.h + * + * $Id$ + * + * ACE Common Data Representation (CDR) basic types. + * + * The current implementation assumes that the host has 1-byte, + * 2-byte and 4-byte integral types, and that it has single + * precision and double precision IEEE floats. + * Those assumptions are pretty good these days, with Crays being + * the only known exception. + * + * + * @author TAO version by + * @author Aniruddha Gokhale <gokhale@cs.wustl.edu> + * @author Carlos O'Ryan<coryan@cs.wustl.edu> + * @author ACE version by + * @author Jeff Parsons <parsons@cs.wustl.edu> + * @author Istvan Buki <istvan.buki@euronet.be> + */ +//============================================================================= + #ifndef ACE_CDR_BASE_H #define ACE_CDR_BASE_H @@ -41,12 +39,14 @@ #include "ace/Basic_Types.h" #include "ace/Message_Block.h" +/** + * @class ACE_CDR + * + * @brief Keep constants and some routines common to both Output and + * Input CDR streams. + */ class ACE_Export ACE_CDR { - // = TITLE - // Keep constants and some routines common to both Output and - // Input CDR streams. - // public: // = Constants defined by the CDR protocol. // By defining as many of these constants as possible as enums we @@ -69,30 +69,38 @@ public: SHORT_ALIGN = 2, LONG_ALIGN = 4, LONGLONG_ALIGN = 8, + /// @note the CORBA LongDouble alignment requirements do not + /// match its size... LONGDOUBLE_ALIGN = 8, - // Note how the CORBA LongDouble alignment requirements do not - // match its size... + /// Maximal CDR 1.1 alignment: "quad precision" FP (i.e. "CDR::Long + /// double", size as above). MAX_ALIGNMENT = 8, - // Maximal CDR 1.1 alignment: "quad precision" FP (i.e. "CDR::Long - // double", size as above). + /// The default buffer size. + /** + * @todo We want to add options to control this + * default value, so this constant should be read as the default + * default value ;-) + */ DEFAULT_BUFSIZE = ACE_DEFAULT_CDR_BUFSIZE, - // The default buffer size. - // @@ TODO We want to add options to control this - // default value, so this constant should be read as the default - // default value ;-) + /// The buffer size grows exponentially until it reaches this size; + /// afterwards it grows linearly using the next constant EXP_GROWTH_MAX = ACE_DEFAULT_CDR_EXP_GROWTH_MAX, - // The buffer size grows exponentially until it reaches this size; - // afterwards it grows linearly using the next constant + /// Once exponential growth is ruled out the buffer size increases + /// in chunks of this size, note that this constants have the same + /// value right now, but it does not need to be so. LINEAR_GROWTH_CHUNK = ACE_DEFAULT_CDR_LINEAR_GROWTH_CHUNK - // Once exponential growth is ruled out the buffer size increases - // in chunks of this size, note that this constants have the same - // value right now, but it does not need to be so. }; + /** + * Do byte swapping for each basic IDL type size. There exist only + * routines to put byte, halfword (2 bytes), word (4 bytes), + * doubleword (8 bytes) and quadword (16 byte); because those are + * the IDL basic type sizes. + */ static void swap_2 (const char *orig, char *target); static void swap_4 (const char *orig, char *target); static void swap_8 (const char *orig, char *target); @@ -109,39 +117,39 @@ public: static void swap_16_array (const char *orig, char *target, size_t length); - // Do byte swapping for each basic IDL type size. There exist only - // routines to put byte, halfword (2 bytes), word (4 bytes), - // doubleword (8 bytes) and quadword (16 byte); because those are - // the IDL basic type sizes. + /// Align the message block to ACE_CDR::MAX_ALIGNMENT, + /// set by the CORBA spec at 8 bytes. static void mb_align (ACE_Message_Block *mb); - // Align the message block to ACE_CDR::MAX_ALIGNMENT, - // set by the CORBA spec at 8 bytes. + /** + * Compute the size of the smallest buffer that can contain at least + * <minsize> bytes. + * To understand how a "best fit" is computed look at the + * algorithm in the code. + * Basically the buffers grow exponentially, up to a certain point, + * then the buffer size grows linearly. + * The advantage of this algorithm is that is rapidly grows to a + * large value, but does not explode at the end. + */ static size_t first_size (size_t minsize); - // Compute the size of the smallest buffer that can contain at least - // <minsize> bytes. - // To understand how a "best fit" is computed look at the - // algorithm in the code. - // Basically the buffers grow exponentially, up to a certain point, - // then the buffer size grows linearly. - // The advantage of this algorithm is that is rapidly grows to a - // large value, but does not explode at the end. + /// Compute not the smallest, but the second smallest buffer that + /// will fir <minsize> bytes. static size_t next_size (size_t minsize); - // Compute not the smallest, but the second smallest buffer that - // will fir <minsize> bytes. + /** + * Increase the capacity of mb to contain at least <minsize> bytes. + * If <minsize> is zero the size is increased by an amount at least + * large enough to contain any of the basic IDL types. Return -1 on + * failure, 0 on success. + */ static int grow (ACE_Message_Block *mb, size_t minsize); - // Increase the capacity of mb to contain at least <minsize> bytes. - // If <minsize> is zero the size is increased by an amount at least - // large enough to contain any of the basic IDL types. Return -1 on - // failure, 0 on success. + /// Copy a message block chain into a single message block, + /// preserving the alignment of the original stream. static void consolidate (ACE_Message_Block *dst, const ACE_Message_Block *src); - // Copy a message block chain into a single message block, - // preserving the alignment of the original stream. static size_t total_length (const ACE_Message_Block *begin, const ACE_Message_Block *end); |