Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision		Previous revision
projects:endian_template [2016/02/23] – [Running the unit tests] terenceprojects:endian_template [2016/02/26] (current) – [Q & A] terence
Line 7: Line 7:
   * **Author:** [[mailto:tjgrant@tatewake.com|Terence J. Grant]]   * **Author:** [[mailto:tjgrant@tatewake.com|Terence J. Grant]]
   * **License:** [[http://opensource.org/licenses/MIT|MIT License]]   * **License:** [[http://opensource.org/licenses/MIT|MIT License]]
-  * **Last Update:** 2016-02-23+  * **Last Update:** 2016-02-26
   * **Donate:** [[:donate|Your donations are appreciated!]]   * **Donate:** [[:donate|Your donations are appreciated!]]
  
Line 16: Line 16:
 After downloading, make sure to follow the [[#how_to_use|how to use]] instructions below; they're worth reading. After downloading, make sure to follow the [[#how_to_use|how to use]] instructions below; they're worth reading.
  
-  * Latest version: [[https://github.com/tatewake/endian-template|Endian Template(C++) version 2016-02-23 on GitHub]]+  * Latest version: [[https://github.com/tatewake/endian-template|Endian Template(C++) version 2016-02-26 on GitHub]]
  
 ===== About ===== ===== About =====
Line 33: Line 33:
 ==== Setup ==== ==== Setup ====
  
-  * First, copy the header into your project.+  * First, copy the header ''tEndian.h'' into your project.
   * Next, include the header somewhere in your project or [[wp>Precompiled_header|precompiled header]] if you use one.   * Next, include the header somewhere in your project or [[wp>Precompiled_header|precompiled header]] if you use one.
-  * You will also need to include ''assert.h'' and ''cstdint'' if they're not already included in your project 
      
 <code c++> <code c++>
-#include <assert.h> 
-#include <cstdint> 
 #include "tEndian.h" #include "tEndian.h"
 </code> </code>
Line 47: Line 44:
 You'll now have the following typedefs to use in your project: You'll now have the following typedefs to use in your project:
  
-  * ''leint16_t'', ''leint32_t'', ''leint64_t'' - Little endian signed types +  * ''leint16'', ''leint32'', ''leint64'' - Little endian signed types 
-  * ''leuint16_t'', ''leuint32_t'', ''leuint64_t'' - Little endian unsigned types +  * ''leuint16'', ''leuint32'', ''leuint64'' - Little endian unsigned types 
-  * ''lefloat32_t'', ''lefloat64_t'' - Little endian floating point types+  * ''lefloat32'', ''lefloat64'' - Little endian floating point types
  
-  * ''beint16_t'', ''beint32_t'', ''beint64_t'' - Big endian signed types +  * ''beint16'', ''beint32'', ''beint64'' - Big endian signed types 
-  * ''beuint16_t'', ''beuint32_t'', ''beuint64_t'' - Big endian unsigned types +  * ''beuint16'', ''beuint32'', ''beuint64'' - Big endian unsigned types 
-  * ''befloat32_t'', ''befloat64_t'' - Big endian floating point types+  * ''befloat32'', ''befloat64'' - Big endian floating point types
  
 The naming conventions are based on those in C++ standard library header [[http://en.cppreference.com/w/cpp/header/cstdint|cstdint]]; the only addition being a prefix of ''le'' or ''be'' to indicate little endian or big endian, respectively. The naming conventions are based on those in C++ standard library header [[http://en.cppreference.com/w/cpp/header/cstdint|cstdint]]; the only addition being a prefix of ''le'' or ''be'' to indicate little endian or big endian, respectively.
Line 69: Line 66:
  
 <code c++> <code c++>
-typedef tBigEndian<size_t>       besize_t;+typedef tBigEndian<size_t>       besize;
 typedef tBigEndian<short>        beshort; typedef tBigEndian<short>        beshort;
 typedef tLittleEndian<long long> lelonglong; typedef tLittleEndian<long long> lelonglong;
Line 78: Line 75:
 ==== Usage ==== ==== Usage ====
  
-When using the endian types, you can literally just drop them in to replace your current types (ensuring you remove any endian swapping functions you're calling), and you're done.+When using the endian types, you can literally just drop them in to replace your current types, as long as you also remove any endian swapping functions you're calling.
  
 There are three primary reasons I see someone using the Endian Template: There are three primary reasons I see someone using the Endian Template:
Line 102: Line 99:
 If you read this data into an ''uint32_t'', you're fine as long as you're on a little endian machine (like an x86-based PC.) Not so much if you end up running your code on a PowerPC CPU, which is big endian. If you read this data into an ''uint32_t'', you're fine as long as you're on a little endian machine (like an x86-based PC.) Not so much if you end up running your code on a PowerPC CPU, which is big endian.
  
-So take the guesswork out of the equation-- read it into an ''leuint32_t'' instead and it'll always work, regardless of the architecture.+So take the guesswork out of the equation-- read it into an ''leuint32'' instead and it'll always work, regardless of the architecture.
  
 === Sending data over a network === === Sending data over a network ===
Line 108: Line 105:
 If you're familiar with [[http://www.tutorialspoint.com/unix_sockets/network_byte_orders.htm|Network Byte Order]], you'll know that the functions ''htons'', ''htonl'', ''ntohl'', ''ntohs'' are generally suggested to use before sending data over a network. If you're familiar with [[http://www.tutorialspoint.com/unix_sockets/network_byte_orders.htm|Network Byte Order]], you'll know that the functions ''htons'', ''htonl'', ''ntohl'', ''ntohs'' are generally suggested to use before sending data over a network.
  
-Since "Network Byte Order" is just "big endian"; instead you can now just define your data as ''beint32_t'', ''beint16_t'', or ''tBigEndian<unsigned short>'' and discard the conversion functions completely.+Since "Network Byte Order" is just "big endian"; instead you can now just define your data as ''beint32'', ''beint16'', or ''tBigEndian<unsigned short>'' and discard the conversion functions completely.
  
 And regardless of the endian of computer's CPU, the code will always work. And regardless of the endian of computer's CPU, the code will always work.
Line 132: Line 129:
 //If you have a question not answered here, please feel free to [[mailto:tjgrant@tatewake.com|contact me]] and ask.// //If you have a question not answered here, please feel free to [[mailto:tjgrant@tatewake.com|contact me]] and ask.//
  
-//No questions so far!//+=== vs Boost? ===
  
 +  * **Q:** What makes this better or different than [[http://www.boost.org/libs/endian/|Boost.Endian]]?
 +  * **A:** I can't say for certain if it's better. The Endian Template is small, and only provides a minimum of what's needed to handle automatic endian conversion.
 +
 +As I understand it Boost.Endian provides three different mechanisms for handling endian (one providing swapping functions), whereas my primary philosophy here is "don't use swapping functions."
 +
 +Just for trivia: I originally created this class in 2001 as a proof-of-concept, hoping I could use it in some emulation projects (never got around to those unfortunately.)
 +This predates when Boost.Endian came out, which was around 2002 or 2003.
 +
 +That said, I've used boost (in general) before on a day-to-day basis for a client project. I'm not a boost "hater", but I'm not a regular user of it either.
 ===== History ===== ===== History =====
 +**February 26, 2016**
 +
 +  * Renamed typedefs without ''_t'' postfix notation as this appears to be "reserved" by POSIX (thanks to reddit user "doodle77")
 +  * Added required includes directly into ''tEndian.h'' (thanks to reddit user "louiswins")
 +  * Changed endian testing method to use reinterpret_cast, as the previous was using "type punning", which in theory can have undefined behavior (thanks to reddit user "NasenSpray")
 +  * Removed ::PassThru method as a small optimization
  
 **February 23, 2016** **February 23, 2016**
Line 141: Line 153:
   * First public release   * First public release
   * Created project page   * Created project page
- 
Print/export
QR Code
QR Code Endian Template (C++) (generated for current page)