1 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
2 <html><head><title>A Few Coding Standards</title></head>
5 <table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0>
6 <tr><td> <font size=+5 color="#EEEEFF" face="Georgia,Palatino,Times,Roman"><b>A Few Coding Standards</b></font></td>
10 <li><a href="#introduction">Introduction</a>
11 <li><a href="#mechanicalissues">Mechanical Source Issues</a>
13 <li><a href="#sourceformating">Source Code Formatting</a>
15 <li><a href="#scf_commenting">Commenting</a>
16 <li><a href="#scf_commentformat">Comment Formatting</a>
17 <li><a href="#scf_includes">#include Style</a>
18 <li><a href="#scf_codewidth">Source Code Width</a>
19 <li><a href="#scf_spacestabs">Use Spaces Instead of Tabs</a>
20 <li><a href="#scf_indentation">Indent Code Consistently</a>
22 <li><a href="#compilerissues">Compiler Issues</a>
24 <li><a href="#ci_warningerrors">Treat Compiler Warnings Like Errors</a>
25 <li><a href="#ci_cpp_features">Which C++ features can I use?</a>
26 <li><a href="#ci_portable_code">Write Portable Code</a>
29 <li><a href="#styleissues">Style Issues</a>
31 <li><a href="#macro">The High Level Issues</a>
33 <li><a href="#hl_module">A Public Header File <b>is</b> a Module</a>
34 <li><a href="#hl_dontinclude">#include as Little as Possible</a>
35 <li><a href="#hl_privateheaders">Keep "internal" Headers Private</a>
37 <li><a href="#micro">The Low Level Issues</a>
39 <li><a href="#hl_assert">Assert Liberally</a>
40 <li><a href="#hl_preincrement">Prefer Preincrement</a>
41 <li><a href="#hl_avoidendl">Avoid endl</a>
42 <li><a href="#hl_exploitcpp">Exploit C++ to its Fullest</a>
44 <li><a href="#iterators">Writing Iterators</a>
46 <li><a href="#seealso">See Also</a>
50 <!-- *********************************************************************** -->
51 </ul><table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0><tr><td align=center><font color="#EEEEFF" size=+2 face="Georgia,Palatino"><b>
52 <a name="introduction">Introduction
53 </b></font></td></tr></table><ul>
54 <!-- *********************************************************************** -->
56 This document attempts to describe a few coding standards that are being used in the LLVM source tree. Although no coding standards should be regarded as absolute requirements to be followed in all instances, coding standards can be useful.<p>
58 This document intentionally does not prescribe fixed standards for religious issues such as brace placement and space usage. For issues like this, follow the golden rule:
61 <blockquote><b>If you are adding a significant body of source to a project, feel free to use whatever style you are most comfortable with. If you are extending, enhancing, or bug fixing already implemented code, use the style that is already being used so that the source is uniform and easy to follow.</b></blockquote>
63 The ultimate goal of these guidelines is the increase readability and maintainability of our common source base. If you have suggestions for topics to be included, please mail them to <a href="mailto:sabre@nondot.org">Chris</a>.<p>
66 <!-- *********************************************************************** -->
67 </ul><table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0><tr><td align=center><font color="#EEEEFF" size=+2 face="Georgia,Palatino"><b>
68 <a name="mechanicalissues">Mechanical Source Issues
69 </b></font></td></tr></table><ul>
70 <!-- *********************************************************************** -->
72 <!-- ======================================================================= -->
73 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0><tr><td> </td><td width="100%"> <font color="#EEEEFF" face="Georgia,Palatino"><b>
74 <a name="sourceformating">Source Code Formatting
75 </b></font></td></tr></table><ul>
78 <!-- _______________________________________________________________________ -->
79 </ul><a name="scf_commenting"><h4><hr size=0>Commenting</h4><ul>
81 Comments are one critical part of readability and maintainability. Everyone knows they should comment, so should you. :) Although we all should probably comment our code more than we do, there are a few very critical places that documentation is very useful:<p>
84 <h4><li>File Headers</h4>
85 Every source file should have a header on it that describes the basic purpose of the file. If a file does not have a header, it should not be checked into CVS. Most source trees will probably have a standard file header format. The standard format for the LLVM source tree looks like this:<p>
88 //===-- llvm/Instruction.h - Instruction class definition --------*- C++ -*--=//
90 // This file contains the declaration of the Instruction class, which is the
91 // base class for all of the VM instructions.
93 //===----------------------------------------------------------------------===//
96 A few things to note about this particular format. The "<tt>-*- C++ -*-</tt>" string on the first line is there to tell Emacs that the source file is a C++ file, not a C file (Emacs assumes .h files are C files by default [Note that tag this is not necessary in .cpp files]). The name of the file is also on the first line, along with a very short description of the purpose of the file. This is important when printing out code and flipping though lots of pages.<p>
98 The main body of the description does not have to be very long in most cases. Here it's only two lines. If an algorithm is being implemented or something tricky is going on, a reference to the paper where it is published should be included, as well as any notes or "gotchas" in the code to watch out for.<p>
101 <h4><li>Class overviews</h4>
103 Classes are one fundemental part of a good object oriented design. As such, a class definition should have a comment block that explains what the class is used for... if it's not obvious. If it's so completely obvious your grandma could figure it out, it's probably safe to leave it out. Naming classes something sane goes a long ways towards avoiding writing documentation. :)<p>
106 <h4><li>Method information</h4>
108 Methods defined in a class (as well as any global functions) should also be documented properly. A quick note about what it does any a description of the borderline behaviour is all that is necessary here (unless something particularly tricky or insideous is going on). The hope is that people can figure out how to use your interfaces without reading the code itself... that is the goal metric.<p>
110 Good things to talk about here are what happens when something unexpected happens: does the method return null? Abort? Format your hard disk?<p>
114 <!-- _______________________________________________________________________ -->
115 </ul><a name="scf_commentformat"><h4><hr size=0>Comment Formatting</h4><ul>
117 In general, prefer C++ style (<tt>//</tt>) comments. They take less space, require less typing, don't have nesting problems, etc. There are a few cases when it is useful to use C style (<tt>/* */</tt>) comments however:<p>
120 <li>When writing a C code: Obviously if you are writing C code, use C style comments. :)
121 <li>When writing a header file that may be #included by a C source file.
122 <li>When writing a source file that is used by a tool that only accepts C style comments.
125 To comment out a large block of code, use <tt>#if 0</tt> and <tt>#endif</tt>. These nest properly and are better behaved in general than C style comments.<p>
127 <!-- _______________________________________________________________________ -->
128 </ul><a name="scf_includes"><h4><hr size=0>#include Style</h4><ul>
130 Immediately after the <a href="#scf_commenting">header file comment</a> (and
131 include guards if working on a header file), the <a
132 href="hl_dontinclude">minimal</a> list of #includes required by the file should
133 be listed. We prefer these #includes to be listed in this order:<p>
136 <li><a href="#mmheader">Main Module header</a>
137 <li><a href="#hl_privateheaders">Local/Private Headers</a>
149 ... and each catagory should be sorted by name.<p>
151 <a name="mmheader">The "Main Module Header" file applies to .cpp file which
152 implement an interface defined by a .h file. This #include should always be
153 included <b>first</b> regardless of where it lives on the file system. By
154 including a header file first in the .cpp files that implement the interfaces,
155 we ensure that the header does not have any hidden dependencies which are not
156 explicitly #included in the header, but should be. It is also a form of
157 documentation in the .cpp file to indicate where the interfaces it implements
161 <!-- _______________________________________________________________________ -->
162 </ul><a name="scf_codewidth"><h4><hr size=0>Source Code Width</h4><ul>
164 Write your code to fit within 80 columns of text. This helps those of us who like to print out code and look at your code in an xterm without resizing it.
167 <!-- _______________________________________________________________________ -->
168 </ul><a name="scf_spacestabs"><h4><hr size=0>Use Spaces Instead of Tabs</h4><ul>
170 In all cases, prefer spaces to tabs in source files. People have different prefered indentation levels, and different styles of indentation that they like... this is fine. What isn't is that different editors/viewers expand tabs out to different tab stops. This can cause your code to look completely unreadable, and it is not worth dealing with.<p>
172 As always, follow the <a href="#goldenrule">Golden Rule</a> above: follow the style of existing code if your are modifying and extending it. If you like four spaces of indentation, <b>DO NOT</b> do that in the middle of a chunk of code with two spaces of indentation. Also, do not reindent a whole source file: it make for incredible diffs that are absolutely worthless.<p>
175 <!-- _______________________________________________________________________ -->
176 </ul><a name="scf_indentation"><h4><hr size=0>Indent Code Consistently</h4><ul>
178 Okay, your first year of programming you were told that indentation is important. If you didn't believe and internalize this then, now is the time. Just do it.<p>
183 <!-- ======================================================================= -->
184 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0><tr><td> </td><td width="100%"> <font color="#EEEEFF" face="Georgia,Palatino"><b>
185 <a name="compilerissues">Compiler Issues
186 </b></font></td></tr></table><ul>
189 <!-- _______________________________________________________________________ -->
190 </ul><a name="ci_warningerrors"><h4><hr size=0>Treat Compiler Warnings Like Errors</h4><ul>
192 If your code has compiler warnings in it, something is wrong: you aren't casting values correctly, your have "questionable" constructs in your code, or you are doing something legitimately wrong. Compiler warnings can cover up legitimate errors in output and make dealing with a translation unit difficult.<p>
194 It is not possible to prevent all warnings from all compilers, nor is it desirable. Instead, pick a standard compiler (like <tt>gcc</tt>) that provides a good thorough set of warnings, and stick to them. At least in the case of <tt>gcc</tt>, it is possible to work around any spurious errors by changing the syntax of the code slightly. For example, an warning that annoys me occurs when I write code like this:<p>
197 if (V = getValue()) {
202 <tt>gcc</tt> will warn me that I probably want to use the <tt>==</tt> operator, and that I probably mistyped it. In most cases, I haven't, and I really don't want the spurious errors. To fix this particular problem, I rewrite the code like this:<p>
205 if ((V = getValue())) {
210 ...which shuts <tt>gcc</tt> up. Any <tt>gcc</tt> warning that annoys you can be
211 fixed by massaging the code appropriately.<p>
213 These are the <tt>gcc</tt> warnings that I prefer to enable: <tt>-Wall -Winline
214 -W -Wwrite-strings -Wno-unused</tt><p>
217 <!-- _______________________________________________________________________ -->
218 </ul><a name="ci_cpp_features"><h4><hr size=0>Which C++ features can I use?</h4><ul>
220 Compilers are finally catching up to the C++ standard. Most compilers implement
221 most features, so you can use just about any features that you would like. In
222 the LLVM source tree, I have chosen to not use these features:<p>
225 <li>Exceptions: Exceptions are very useful for error reporting and handling
226 exceptional conditions. I do not use them in LLVM because they do have an
227 associated performance impact (by restricting restructuring of code), and parts
228 of LLVM are designed for performance critical purposes.<p>
230 Just like most of the rules in this document, this isn't a hard and fast
231 requirement. Exceptions are used in the Parser, because it simplifies error
232 reporting <b>significantly</b>, and the LLVM parser is not at all in the
235 <li>RTTI: RTTI has a large cost in terms of executable size, and compilers are
236 not yet very good at stomping out "dead" class information blocks. Because of
237 this, typeinfo and dynamic cast are not used.
240 Other features, such as templates (without partial specialization) can be used
241 freely. The general goal is to have clear, consise, performant code... if a
242 technique assists with that then use it.<p>
245 <!-- _______________________________________________________________________ -->
246 </ul><a name="ci_portable_code"><h4><hr size=0>Write Portable Code</h4><ul>
248 In almost all cases, it is possible and within reason to write completely
249 portable code. If there are cases where it isn't possible to write portable
250 code, isolate it behind a well defined (and well documented) interface.<p>
252 In practice, this means that you shouldn't assume much about the host compiler,
253 including its support for "high tech" features like partial specialization of
254 templates. In fact, Visual C++ 6 could be an important target for our work in
255 the future, and we don't want to have to rewrite all of our code to support
260 <!-- *********************************************************************** -->
261 </ul><table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0><tr><td align=center><font color="#EEEEFF" size=+2 face="Georgia,Palatino"><b>
262 <a name="styleissues">Style Issues
263 </b></font></td></tr></table><ul>
264 <!-- *********************************************************************** -->
267 <!-- ======================================================================= -->
268 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0><tr><td> </td><td width="100%"> <font color="#EEEEFF" face="Georgia,Palatino"><b>
269 <a name="macro">The High Level Issues
270 </b></font></td></tr></table><ul>
273 <!-- _______________________________________________________________________ -->
274 </ul><a name="hl_module"><h4><hr size=0>A Public Header File <b>is</b> a Module</h4><ul>
276 C++ doesn't do too well in the modularity department. There is no real
277 encapsulation or data hiding (unless you use expensive protocol classes), but it
278 is what we have to work with. When you write a public header file (in the LLVM
279 source tree, they live in the top level "include" directory), you are defining a
280 module of functionality.<p>
282 Ideally, modules should be completely independent of each other, and their
283 header files should only include the absolute minimum number of headers
284 possible. A module is not just a class, a function, or a namespace: <a
285 href="http://www.cuj.com/articles/2000/0002/0002c/0002c.htm">it's a collection
286 of these</a> that defines an interface. This interface may be several
287 functions, classes or data structures, but the important issue is how they work
290 In general, a module should be implemented with one or more <tt>.cpp</tt> files.
291 Each of these <tt>.cpp</tt> files should include the header that defines their
292 interface first. This ensure that all of the dependences of the module header
293 have been properly added to the module header itself, and are not implicit.
294 System headers should be included after user headers for a translation unit.<p>
297 <!-- _______________________________________________________________________ -->
298 </ul><a name="hl_dontinclude"><h4><hr size=0>#include as Little as Possible</h4><ul>
300 <tt>#include</tt> hurts compile time performance. Don't do it unless you have
301 to, especially in header files.<p>
303 But wait, sometimes you need to have the definition of a class to use it, or to
304 inherit from it. In these cases go ahead and #include that header file. Be
305 aware however that there are many cases where you don't need to have the full
306 definition of a class. If you are using a pointer or reference to a class, you
307 don't need the header file. If you are simply returning a class instance from a
308 prototyped function or method, you don't need it. In fact, for most cases, you
309 simply don't need the definition of a class... and not <tt>#include</tt>'ing
310 speeds up compilation.<p>
312 It is easy to try to go too overboard on this recommendation, however. You
313 <b>must</b> include all of the header files that you are using, either directly
314 or indirectly (through another header file). To make sure that you don't
315 accidently forget to include a header file in your module header, make sure to
316 include your module header <b>first</b> in the implementation file (as mentioned
317 above). This way there won't be any hidden dependencies that you'll find out
321 <!-- _______________________________________________________________________ -->
322 </ul><a name="hl_privateheaders"><h4><hr size=0>Keep "internal" Headers Private</h4><ul>
324 Many modules have a complex implementation that causes them to use more than one
325 implementation (<tt>.cpp</tt>) file. It is often tempting to put the internal
326 communication interface (helper classes, extra functions, etc) in the public
327 module header file. Don't do this. :)<p>
329 If you really need to do something like this, put a private header file in the
330 same directory as the source files, and include it locally. This ensures that
331 your private interface remains private and undisturbed by outsiders.<p>
333 Note however, that it's okay to put extra implementation methods a public class
334 itself... just make them private (or protected), and all is well.<p>
337 <!-- ======================================================================= -->
338 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0><tr><td> </td><td width="100%"> <font color="#EEEEFF" face="Georgia,Palatino"><b>
339 <a name="micro">The Low Level Issues
340 </b></font></td></tr></table><ul>
343 <!-- _______________________________________________________________________ -->
344 </ul><a name="hl_assert"><h4><hr size=0>Assert Liberally</h4><ul>
346 Use the "<tt>assert</tt>" function to its fullest. Check all of your
347 preconditions and assumptions, you never know when a bug (not neccesarily even
348 yours) might be caught early by an assertion, which reduces debugging time
349 dramatically. The "<tt><cassert></tt>" header file is probably already
350 included by the header files you are using, so it doesn't cost anything to use
353 To further assist with debugging, make sure to put some kind of error message in
354 the assertion statement (which is printed if the assertion is tripped). This
355 helps the poor debugging make sense of why an assertion is being made and
356 enforced, and hopefully what to do about it. Here is one complete example:<p>
359 inline Value *getOperand(unsigned i) {
360 assert(i < Operands.size() && "getOperand() out of range!");
365 Here are some examples:
368 assert(Ty->isPointerType() && "Can't allocate a non pointer type!");
370 assert((Opcode == Shl || Opcode == Shr) && "ShiftInst Opcode invalid!");
372 assert(idx < getNumSuccessors() && "Successor # out of range!");
374 assert(V1.getType() == V2.getType() && "Constant types must be identical!");
376 assert(isa<PHINode>(Succ->front()) && "Only works on PHId BBs!");
379 You get the idea...<p>
382 <!-- _______________________________________________________________________ -->
383 </ul><a name="hl_preincrement"><h4><hr size=0>Prefer Preincrement</h4><ul>
385 Hard fast rule: Preincrement (++X) may be no slower than postincrement (X++) and
386 could very well be a lot faster than it. Use preincrementation whenever
389 The semantics of postincrement include making a copy of the value being
390 incremented, returning it, and then preincrementing the "work value". For
391 primitive types, this isn't a big deal... but for iterators, it can be a huge
392 issue (for example, some iterators contains stack and set objects in them...
393 copying an iterator could invoke the copy ctor's of these as well). In general,
394 get in the habit of always using preincrement, and you won't have a problem.<p>
397 <!-- _______________________________________________________________________ -->
398 </ul><a name="hl_avoidendl"><h4><hr size=0>Avoid endl</h4><ul>
400 The <tt>endl</tt> modifier, when used with iostreams outputs a newline to the
401 output stream specified. In addition to doing this, however, it also flushes
402 the output stream. In other words, these are equivalent:<p>
406 cout << "\n" << flush;
409 Most of the time, you probably have no reason to flush the output stream, so it's better to use a literal <tt>"\n"</tt>.<p>
412 <!-- _______________________________________________________________________ -->
413 </ul><a name="hl_exploitcpp"><h4><hr size=0>Exploit C++ to its Fullest</h4><ul>
415 C++ is a powerful language. With a firm grasp on its capabilities, you can make
416 write effective, consise, readable and maintainable code all at the same time.
417 By staying consistent, you reduce the amount of special cases that need to be
418 remembered. Reducing the total number of lines of code you write is a good way
419 to avoid documentation, and avoid giving bugs a place to hide.<p>
421 For these reasons, come to know and love the contents of your local
422 <algorithm> header file. Know about <functional> and what it can do
423 for you. C++ is just a tool that wants you to master it. :)<p>
427 <!-- ======================================================================= -->
428 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0><tr><td> </td><td width="100%"> <font color="#EEEEFF" face="Georgia,Palatino"><b>
429 <a name="iterators">Writing Iterators
430 </b></font></td></tr></table><ul>
432 Here's a pretty good summary of how to write your own data structure iterators in a way that is compatible with the STL, and with a lot of other code out there (slightly edited by Chris):<p>
435 From: Ross Smith <ross.s@ihug.co.nz>
436 Newsgroups: comp.lang.c++.moderated
437 Subject: Writing iterators (was: Re: Non-template functions that take iterators)
438 Date: 28 Jun 2001 12:07:10 -0400
441 > Any pointers handy on "writing STL-compatible iterators for
444 I'll give it a try...
446 The usual situation requiring user-defined iterators is that you have
447 a type that bears some resemblance to an STL container, and you want
448 to provide iterators so it can be used with STL algorithms. You need
449 to ask three questions:
451 First, is this simply a wrapper for an underlying collection of
452 objects that's held somewhere as a real STL container, or is it a
453 "virtual container" for which iteration is (under the hood) more
454 complicated than simply incrementing some underlying iterator (or
455 pointer or index or whatever)? In the former case you can frequently
456 get away with making your container's iterators simply typedefs for
457 those of the underlying container; your begin() function would call
458 member_container.begin(), and so on.
460 Second, do you only need read-only iterators, or do you need separate
461 read-only (const) and read-write (non-const) iterators?
463 Third, which kind of iterator (input, output, forward, bidirectional,
464 or random access) is appropriate? If you're familiar with the
465 properties of the iterator types (if not, visit
466 <a href="http://www.sgi.com/tech/stl/">http://www.sgi.com/tech/stl/</a>), the appropriate choice should be
467 obvious from the semantics of the container.
469 I'll start with forward iterators, as the simplest case that's likely
470 to come up in normal code. Input and output iterators have some odd
471 properties and rarely need to be implemented in user code; I'll leave
472 them out of discussion. Bidirectional and random access iterators are
475 The exact behaviour of a forward iterator is spelled out in the
476 Standard in terms of a set of expressions with specified behaviour,
477 rather than a set of member functions, which leaves some leeway in how
478 you actually implement it. Typically it looks something like this
479 (I'll start with the const-iterator-only situation):
481 #include <iterator>
485 typedef something_or_other value_type;
486 class const_iterator:
487 public std::iterator<std::forward_iterator_tag, value_type> {
488 friend class container;
490 const value_type& operator*() const;
491 const value_type* operator->() const;
492 const_iterator& operator++();
493 const_iterator operator++(int);
494 friend bool operator==(const_iterator lhs,
496 friend bool operator!=(const_iterator lhs,
504 An iterator should always be derived from an instantiation of the
505 std::iterator template. The iterator's life cycle functions
506 (constructors, destructor, and assignment operator) aren't declared
507 here; in most cases the compiler-generated ones are sufficient. The
508 container needs to be a friend of the iterator so that the container's
509 begin() and end() functions can fill in the iterator's private members
510 with the appropriate values.
512 <i>[Chris's Note: I prefer to not make my iterators friends. Instead, two
513 ctor's are provided for the iterator class: one to start at the end of the
514 container, and one at the beginning. Typically this is done by providing
515 two constructors with different signatures.]</i>
517 There are normally only three member functions that need nontrivial
518 implementations; the rest are just boilerplate.
520 const container::value_type&
521 container::const_iterator::operator*() const {
522 // find the element and return a reference to it
525 const container::value_type*
526 container::const_iterator::operator->() const {
530 If there's an underlying real container, operator*() can just return a
531 reference to the appropriate element. If there's no actual container
532 and the elements need to be generated on the fly -- what I think of as
533 a "virtual container" -- things get a bit more complicated; you'll
534 probably need to give the iterator a value_type member object, and
535 fill it in when you need to. This might be done as part of the
536 increment operator (below), or if the operation is nontrivial, you
537 might choose the "lazy" approach and only generate the actual value
538 when one of the dereferencing operators is called.
540 The operator->() function is just boilerplate around a call to
543 container::const_iterator&
544 container::const_iterator::operator++() {
545 // the incrementing logic goes here
549 container::const_iterator
550 container::const_iterator::operator++(int) {
551 const_iterator old(*this);
556 Again, the incrementing logic will usually be trivial if there's a
557 real container involved, more complicated if you're working with a
558 virtual container. In particular, watch out for what happens when you
559 increment past the last valid item -- this needs to produce an
560 iterator that will compare equal to container.end(), and making this
561 work is often nontrivial for virtual containers.
563 The post-increment function is just boilerplate again (and
564 incidentally makes it obvious why all the experts recommend using
565 pre-increment wherever possible).
567 bool operator==(container::const_iterator lhs,
568 container::const_iterator rhs) {
569 // equality comparison goes here
572 bool operator!=(container::const_iterator lhs,
573 container::const_iterator rhs) {
574 return !(lhs == rhs);
577 For a real container, the equality comparison will usually just
578 compare the underlying iterators (or pointers or indices or whatever).
579 The semantics of comparisons for virtual container iterators are often
580 tricky. Remember that iterator comparison only needs to be defined for
581 iterators into the same container, so you can often simplify things by
582 taking for granted that lhs and rhs both point into the same container
583 object. Again, the second function is just boilerplate.
585 It's a matter of taste whether iterator arguments are passed by value
586 or reference; I've shown tham passed by value to reduce clutter, but
587 if the iterator contains several data members, passing by reference
590 That convers the const-iterator-only situation. When we need separate
591 const and mutable iterators, one small complication is added beyond
592 the simple addition of a second class.
596 typedef something_or_other value_type;
597 class const_iterator;
599 public std::iterator<std::forward_iterator_tag, value_type> {
600 friend class container;
601 friend class container::const_iterator;
603 value_type& operator*() const;
604 value_type* operator->() const;
605 iterator& operator++();
606 iterator operator++(int);
607 friend bool operator==(iterator lhs, iterator rhs);
608 friend bool operator!=(iterator lhs, iterator rhs);
612 class const_iterator:
613 public std::iterator<std::forward_iterator_tag, value_type> {
614 friend class container;
617 const_iterator(const iterator& i);
618 const value_type& operator*() const;
619 const value_type* operator->() const;
620 const_iterator& operator++();
621 const_iterator operator++(int);
622 friend bool operator==(const_iterator lhs,
624 friend bool operator!=(const_iterator lhs,
632 There needs to be a conversion from iterator to const_iterator (so
633 that mixed-type operations, such as comparison between an iterator and
634 a const_iterator, will work). This is done here by giving
635 const_iterator a conversion constructor from iterator (equivalently,
636 we could have given iterator an operator const_iterator()), which
637 requires const_iterator to be a friend of iterator, so it can copy its
638 data members. (It also requires the addition of an explicit default
639 constructor to const_iterator, since the existence of another
640 user-defined constructor inhibits the compiler-defined one.)
642 Bidirectional iterators add just two member functions to forward
646 public std::iterator<std::bidirectional_iterator_tag, value_type> {
649 iterator& operator--();
650 iterator operator--(int);
654 I won't detail the implementations, they're obvious variations on
657 Random access iterators add several more member and friend functions:
660 public std::iterator<std::random_access_iterator_tag, value_type> {
663 iterator& operator+=(difference_type rhs);
664 iterator& operator-=(difference_type rhs);
665 friend iterator operator+(iterator lhs, difference_type rhs);
666 friend iterator operator+(difference_type lhs, iterator rhs);
667 friend iterator operator-(iterator lhs, difference_type rhs);
668 friend difference_type operator-(iterator lhs, iterator rhs);
669 friend bool operator<(iterator lhs, iterator rhs);
670 friend bool operator>(iterator lhs, iterator rhs);
671 friend bool operator<=(iterator lhs, iterator rhs);
672 friend bool operator>=(iterator lhs, iterator rhs);
676 container::iterator&
677 container::iterator::operator+=(container::difference_type rhs) {
678 // add rhs to iterator position
682 container::iterator&
683 container::iterator::operator-=(container::difference_type rhs) {
684 // subtract rhs from iterator position
688 container::iterator operator+(container::iterator lhs,
689 container::difference_type rhs) {
690 return iterator(lhs) += rhs;
693 container::iterator operator+(container::difference_type lhs,
694 container::iterator rhs) {
695 return iterator(rhs) += lhs;
698 container::iterator operator-(container::iterator lhs,
699 container::difference_type rhs) {
700 return iterator(lhs) -= rhs;
703 container::difference_type operator-(container::iterator lhs,
704 container::iterator rhs) {
705 // calculate distance between iterators
708 bool operator<(container::iterator lhs, container::iterator rhs) {
709 // perform less-than comparison
712 bool operator>(container::iterator lhs, container::iterator rhs) {
716 bool operator<=(container::iterator lhs, container::iterator rhs) {
717 return !(rhs < lhs);
720 bool operator>=(container::iterator lhs, container::iterator rhs) {
721 return !(lhs < rhs);
724 Four of the functions (operator+=(), operator-=(), the second
725 operator-(), and operator<()) are nontrivial; the rest are
728 One feature of the above code that some experts may disapprove of is
729 the declaration of all the free functions as friends, when in fact
730 only a few of them need direct access to the iterator's private data.
731 I originally got into the habit of doing this simply to keep the
732 declarations together; declaring some functions inside the class and
733 some outside seemed awkward. Since then, though, I've been told that
734 there's a subtle difference in the way name lookup works for functions
735 declared inside a class (as friends) and outside, so keeping them
736 together in the class is probably a good idea for practical as well as
739 I hope all this is some help to anyone who needs to write their own
740 STL-like containers and iterators.
743 Ross Smith <ross.s@ihug.co.nz> The Internet Group, Auckland, New Zealand
747 <!-- *********************************************************************** -->
748 </ul><table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0><tr><td align=center><font color="#EEEEFF" size=+2 face="Georgia,Palatino"><b>
749 <a name="seealso">See Also
750 </b></font></td></tr></table><ul>
751 <!-- *********************************************************************** -->
753 A lot of these comments and recommendations have been culled for other sources. Two particularly important books for our work are:<p>
756 <li><a href="http://www.aw.com/product/0,2627,0201924889,00.html">Effective C++</a> by Scott Meyers. There is an online version of the book (only some chapters though) <a href="http://www.awlonline.com/cseng/meyerscddemo/">available as well</a>.
757 <li><a href="http://cseng.aw.com/book/0,3828,0201633620,00.html">Large-Scale C++ Software Design</a> by John Lakos
760 If you get some free time, and you haven't read them: do so, you might learn something. :)
763 <!-- *********************************************************************** -->
765 <!-- *********************************************************************** -->
769 <address><a href="mailto:sabre@nondot.org">Chris Lattner</a></address>
770 <!-- Created: Tue Jan 23 15:19:28 CST 2001 -->
772 Last modified: Thu Aug 7 16:44:33 CDT 2003