Annotation of researchv10dc/cmd/gcc/cpp.texinfo, revision 1.1

1.1     ! root        1: \input texinfo
        !             2: @setfilename cpp.info
        !             3: @settitle The C Preprocessor
        !             4: @ifinfo
        !             5: This file documents the GNU C Preprocessor.
        !             6: 
        !             7: Copyright (C) 1987 Richard M. Stallman.
        !             8: 
        !             9: Permission is granted to make and distribute verbatim copies of
        !            10: this manual provided the copyright notice and this permission notice
        !            11: are preserved on all copies.
        !            12: 
        !            13: @ignore
        !            14: Permission is granted to process this file through Tex and print the
        !            15: results, provided the printed document carries copying permission
        !            16: notice identical to this one except for the removal of this paragraph
        !            17: (this paragraph not being relevant to the printed manual).
        !            18: 
        !            19: @end ignore
        !            20: Permission is granted to copy and distribute modified versions of this
        !            21: manual under the conditions for verbatim copying, provided also that
        !            22: the entire resulting derived work is distributed under the terms of a
        !            23: permission notice identical to this one.
        !            24: 
        !            25: Permission is granted to copy and distribute translations of this manual
        !            26: into another language, under the above conditions for modified versions.
        !            27: @end ifinfo
        !            28: 
        !            29: @titlepage
        !            30: @sp 6
        !            31: @center @titlefont{The C Preprocessor}
        !            32: @sp 4
        !            33: @center First Edition
        !            34: @sp 1
        !            35: @center January 1987
        !            36: @sp 5
        !            37: @center Richard M. Stallman
        !            38: @page
        !            39: @vskip 0pt plus 1filll
        !            40: Copyright @copyright{} 1987 Richard M. Stallman.
        !            41: 
        !            42: Permission is granted to make and distribute verbatim copies of
        !            43: this manual provided the copyright notice and this permission notice
        !            44: are preserved on all copies.
        !            45: 
        !            46: Permission is granted to copy and distribute modified versions of this
        !            47: manual under the conditions for verbatim copying, provided also that
        !            48: the entire resulting derived work is distributed under the terms of a
        !            49: permission notice identical to this one.
        !            50: 
        !            51: Permission is granted to copy and distribute translations of this manual
        !            52: into another language, under the above conditions for modified versions.
        !            53: @end titlepage
        !            54: @page
        !            55: 
        !            56: @node Top, Global Actions,, (DIR)
        !            57: @chapter The C Preprocessor
        !            58: 
        !            59: The C preprocessor is a @dfn{macro processor} that is used automatically by
        !            60: the C compiler to transform your program before actual compilation.  It is
        !            61: called a macro processor because it allows you to define @dfn{macros},
        !            62: which are brief abbreviations for longer constructs.
        !            63: 
        !            64: The C preprocessor provides four separate facilities that you can use as
        !            65: you see fit:
        !            66: 
        !            67: @itemize @bullet
        !            68: @item
        !            69: Inclusion of header files.  These are files of declarations that can be
        !            70: substituted into your program.
        !            71: 
        !            72: @item
        !            73: Macro expansion.  You can define @dfn{macros}, which are abbreviations
        !            74: for arbitrary fragments of C code, and then the C preprocessor will
        !            75: replace the macros with their definitions throughout the program.
        !            76: 
        !            77: @item
        !            78: Conditional compilation.  Using special preprocessor commands, you
        !            79: can include or exclude parts of the program according to various
        !            80: conditions.
        !            81: 
        !            82: @item
        !            83: Line control.  If you use a program to combine or rearrange source files into
        !            84: an intermediate file which is then compiled, you can use line control
        !            85: to inform the compiler of where each source line originally came from.
        !            86: @end itemize
        !            87: 
        !            88: C preprocessors vary in some details.  This manual discusses the GNU C
        !            89: preprocessor, the C Compatible Compiler Preprocessor.  The GNU C
        !            90: preprocessor provides a superset of the features of ANSI Standard C.
        !            91: 
        !            92: ANSI Standard C requires the rejection of many harmless constructs commonly
        !            93: used by today's C programs.  Such incompatibility would be inconvenient for
        !            94: users, so the GNU C preprocessor is configured to accept these constructs
        !            95: by default.  Strictly speaking, to get ANSI Standard C, you must use the
        !            96: options @samp{-T}, @samp{-undef} and @samp{-pedantic}, but in practice the
        !            97: consequences of having strict ANSI Standard C make it undesirable to do
        !            98: this.  @xref{Invocation}.
        !            99: 
        !           100: @menu
        !           101: * Global Actions::    Actions made uniformly on all input files.
        !           102: * Commands::          General syntax of preprocessor commands.
        !           103: * Header Files::      How and why to use header files.
        !           104: * Macros::            How and why to use macros.
        !           105: * Conditionals::      How and why to use conditionals.
        !           106: * Combining Sources:: Use of line control when you combine source files.
        !           107: * Other Commands::    Miscellaneous preprocessor commands.
        !           108: * Output::            Format of output from the C preprocessor.
        !           109: * Invocation::        How to invoke the preprocessor; command options.
        !           110: * Concept Index::     Index of concepts and terms.
        !           111: * Index::             Index of commands, predefined macros and options.
        !           112: @end menu
        !           113: 
        !           114: @node Global Actions, Commands, Top, Top
        !           115: @section Transformations Made Globally
        !           116: 
        !           117: Most C preprocessor features are inactive unless you give specific commands
        !           118: to request their use.  (Preprocessor commands are lines starting with
        !           119: @samp{#}; @pxref{Commands}).  But there are three transformations that the
        !           120: preprocessor always makes on all the input it receives, even in the absence
        !           121: of commands.
        !           122: 
        !           123: @itemize @bullet
        !           124: @item
        !           125: All C comments are replaced with single spaces.
        !           126: 
        !           127: @item
        !           128: Backslash-Newline sequences are deleted, no matter where.  This
        !           129: feature allows you to break long lines for cosmetic purposes without
        !           130: changing their meaning.
        !           131: 
        !           132: @item
        !           133: Predefined macro names are replaced with their expansions
        !           134: (@pxref{Predefined}).
        !           135: @end itemize
        !           136: 
        !           137: The first two transformations are done @emph{before} nearly all other parsing
        !           138: and before preprocessor commands are recognized.  Thus, for example, you
        !           139: can split a line cosmetically with Backslash-Newline anywhere (except
        !           140: when trigraphs are in use; see below).
        !           141: 
        !           142: @example
        !           143: /*
        !           144: */ # /*
        !           145: */ defi\
        !           146: ne FO\
        !           147: O 10\
        !           148: 20
        !           149: @end example
        !           150: 
        !           151: @noindent
        !           152: is equivalent into @samp{#define FOO 1020}.  You can split even an escape
        !           153: sequence with Backslash-Newline.  For example, you can split @code{"foo\bar"}
        !           154: between the @samp{\} and the @samp{b} to get
        !           155: 
        !           156: @example
        !           157: "foo\\
        !           158: bar"
        !           159: @end example
        !           160: 
        !           161: @noindent
        !           162: This behavior is unclean: in all other contexts, a Backslash can be
        !           163: inserted in a string constant as an ordinary character by writing a double
        !           164: Backslash, and this creates an exception.  But the ANSI C standard requires
        !           165: it.  (Strict ANSI C does not allow Newlines in string constants, so they
        !           166: do not consider this a problem.)
        !           167: 
        !           168: But there are a few exceptions to all three transformations.
        !           169: 
        !           170: @itemize @bullet
        !           171: @item
        !           172: C comments and predefined macro names are not recognized inside a
        !           173: @samp{#include} command in which the file name is delimited with
        !           174: @samp{<} and @samp{>}.
        !           175: 
        !           176: @item
        !           177: C comments and predefined macro names are never recognized within a
        !           178: character or string constant.  (Strictly speaking, this is the rule,
        !           179: not an exception, but it is worth noting here anyway.)
        !           180: 
        !           181: @item
        !           182: Backslash-Newline may not safely be used within an ANSI ``trigraph''.
        !           183: Trigraphs are converted before Backslash-Newline is deleted.  If you
        !           184: write what looks like a trigraph with a Backslash-Newline inside, the
        !           185: Backslash-Newline is deleted as usual, but it is then too late to
        !           186: recognize the trigraph.
        !           187: 
        !           188: This exception is relevant only if you use the @samp{-T} option to
        !           189: enable trigraph processing.  @xref{Invocation}.
        !           190: @end itemize
        !           191: 
        !           192: @node Commands, Header Files, Global Actions, Top
        !           193: @section Preprocessor Commands
        !           194: 
        !           195: @cindex preprocessor commands
        !           196: @cindex commands
        !           197: Most preprocessor features are active only if you use preprocessor commands
        !           198: to request their use.
        !           199: 
        !           200: Preprocessor commands are lines in your program that start with @samp{#}.
        !           201: The @samp{#} is followed by an identifier that is the @dfn{command name}.
        !           202: For example, @samp{#define} is the command that defines a macro.
        !           203: Whitespace is also allowed before and after the @samp{#}.
        !           204: 
        !           205: The set of valid command names is fixed.  Programs cannot define new
        !           206: preprocessor commands.
        !           207: 
        !           208: Some command names require arguments; these make up the rest of the command
        !           209: line and must be separated from the command name by whitespace.  For example,
        !           210: @samp{#define} must be followed by a macro name and the intended expansion
        !           211: of the macro.
        !           212: 
        !           213: A preprocessor command cannot be more than one line in normal circumstances.
        !           214: It may be split cosmetically with Backslash-Newline, but that has no effect
        !           215: on its meaning.  Comments containing Newlines can also divide the command into
        !           216: multiple lines, but the comments are changed to Spaces before the command
        !           217: is interpreted.  The only way a significant Newline can occur in a preprocessor
        !           218: command is within a string constant or character constant.  Note that
        !           219: most C compilers that might be applied to the output from the preprocessor
        !           220: do not accept string or character constants containing Newlines.
        !           221: 
        !           222: The @samp{#} and the command name cannot come from a macro expansion.  For
        !           223: example, if @samp{foo} is defined as a macro expanding to @samp{define},
        !           224: that does not make @samp{#foo} a valid preprocessor command.
        !           225: 
        !           226: @node Header Files, Macros, Commands, Top
        !           227: @section Header Files
        !           228: 
        !           229: @cindex header file
        !           230: A header file is a file containing C declarations and macro definitions
        !           231: (@pxref{Macros}) to be shared between several source files.  You request
        !           232: the use of a header file in your program with the C preprocessor command
        !           233: @samp{#include}.
        !           234: 
        !           235: @node Header Uses, Include Syntax, Header Files, Header Files
        !           236: @subsection Uses of Header Files
        !           237: 
        !           238: Header files serve two kinds of purposes.
        !           239: 
        !           240: @itemize @bullet
        !           241: @item
        !           242: @findex system header files
        !           243: System header files declare the interfaces to parts of the operating
        !           244: system.  You include them in your program to supply the definitions
        !           245: you need to invoke system calls and libraries.
        !           246: 
        !           247: @item
        !           248: Your own header files contain declarations for interfaces between the
        !           249: source files of your program.  Each time you have a group of related
        !           250: declarations and macro definitions all or most of which are needed in
        !           251: several different source files, it is a good idea to create a header
        !           252: file for them.
        !           253: @end itemize
        !           254: 
        !           255: Including a header file produces the same results in C compilation as
        !           256: copying the header file into each source file that needs it.  But such
        !           257: copying would be time-consuming and error-prone.  With a header file, the
        !           258: related declarations appear in only one place.  If they need to be changed,
        !           259: they can be changed in one place, and programs that include the header file
        !           260: will automatically use the new version when next recompiled.  The header
        !           261: file eliminates the labor of finding and changing all the copies as well as
        !           262: the risk that a failure to find one copy will result in inconsistencies
        !           263: within a program.
        !           264: 
        !           265: The usual convention is to give header files names that end with @file{.h}.
        !           266: 
        !           267: @node Include Syntax, Include Operation, Header Uses, Header Files
        !           268: @subsection The @samp{#include} Command
        !           269: 
        !           270: @findex #include
        !           271: Both user and system header files are included using the preprocessor
        !           272: command @samp{#include}.  It has three variants:
        !           273: 
        !           274: @table @code
        !           275: @item #include <@var{file}>
        !           276: This variant is used for system header files.  It searches for a file
        !           277: named @var{file} in a list of directories specified by you, then in a
        !           278: standard list of system directories.  You specify directories to
        !           279: search for header files with the command option @samp{-I}
        !           280: (@pxref{Invocation}).  The option @samp{-nostdinc} inhibits searching
        !           281: the standard system directories; in this case only the directories
        !           282: you specify are searched.
        !           283: 
        !           284: The parsing of this form of @samp{#include} is slightly special
        !           285: because comments are not recognized within the @samp{<@dots{}>}.
        !           286: Thus, in @samp{#include <x/*y>} the @samp{/*} does not start a comment
        !           287: and the command specifies inclusion of a system header file named
        !           288: @file{x/*y}.  Of course, a header file with such a name is unlikely to
        !           289: exist on Unix, where shell wildcard features would make it hard to
        !           290: manipulate.@refill
        !           291: 
        !           292: The argument @var{file} may not contain a @samp{>} character.  It may,
        !           293: however, contain a @samp{<} character.
        !           294: 
        !           295: @item #include "@var{file}"
        !           296: This variant is used for header files of your own program.  It
        !           297: searches for a file named @var{file} first in the current
        !           298: directory, then in the same directories used for system header
        !           299: files.  The current directory is tried first because it is
        !           300: presumed to be the location of the files of the program being
        !           301: compiled.  (If the @samp{-I-} option is used, the special treatment
        !           302: of the current directory is inhibited.)
        !           303: 
        !           304: The argument @var{file} may not contain @samp{"} characters.  If
        !           305: backslashes occur within @var{file}, they are considered ordinary text
        !           306: characters, not escape characters.  None of the character escape
        !           307: sequences appropriate to string constants in C are processed.  Thus,
        !           308: @samp{#include "x\n\\y"} specifies a filename containing three
        !           309: backslashes.  It is not clear why this behavior is ever useful, but
        !           310: the ANSI standard specifies it.
        !           311: 
        !           312: @item #include @var{anything else}
        !           313: This variant is called a @dfn{computed #include}.  Any @samp{#include}
        !           314: command whose argument does not fit the above two forms is a computed
        !           315: include.  The text @var{anything else} is checked for macro calls,
        !           316: which are expanded (@pxref{Macros}).  When this is done, the result
        !           317: must fit one of the above two variants.
        !           318: 
        !           319: This feature allows you to define a macro which controls the file name
        !           320: to be used at a later point in the program.  One application of this
        !           321: is to allow a site-configuration file for your program to specify the
        !           322: names of the system include files to be used.  This can help in
        !           323: porting the program to various operating systems in which the
        !           324: necessary system header files are found in different places.
        !           325: @end table
        !           326: 
        !           327: @node Include Operation,, Include Syntax, Header Files
        !           328: @subsection How @samp{#include} Works
        !           329: 
        !           330: The @samp{#include} command works by directing the C preprocessor to scan
        !           331: the specified file as input before continuing with the rest of the current
        !           332: file.  The output from the preprocessor contains the output already
        !           333: generated, followed by the output resulting from the included file,
        !           334: followed by the output that comes from the text after the @samp{#include}
        !           335: command.  For example, given two files as follows:
        !           336: 
        !           337: @example
        !           338: /* File program.c */
        !           339: int x;
        !           340: #include "header.h"
        !           341: 
        !           342: main ()
        !           343: @{
        !           344:   printf (test ());
        !           345: @}
        !           346: 
        !           347: 
        !           348: /* File header.h */
        !           349: char *test ();
        !           350: @end example
        !           351: 
        !           352: @noindent
        !           353: the output generated by the C preprocessor for @file{program.c} as input
        !           354: would be
        !           355: 
        !           356: @example
        !           357: int x;
        !           358: char *test ();
        !           359: 
        !           360: main ()
        !           361: @{
        !           362:   printf (test ());
        !           363: @}
        !           364: @end example
        !           365: 
        !           366: Included files are not limited to declarations and macro definitions; they
        !           367: are merely the typical use.  Any fragment of a C program can be included
        !           368: from another file.  The include file could even contain the beginning of a
        !           369: statement that is concluded in the containing file, or the end of a
        !           370: statement that was started in the including file.  However, a comment or a
        !           371: string or character constant may not start in the included file and finish
        !           372: in the including file.  An unterminated comment, string constant or
        !           373: character constant in an included file is considered to end (with an error
        !           374: message) at the end of the file.
        !           375: 
        !           376: The line following the @samp{#include} command is always treated as a
        !           377: separate line by the C preprocessor even if the included file lacks a final
        !           378: newline.
        !           379: 
        !           380: @node Macros, Conditionals, Header Files, Top
        !           381: @section Macros
        !           382: 
        !           383: A macro is a sort of abbreviation which you can define once and then
        !           384: use later.  There are many complicated features associated with macros
        !           385: in the C preprocessor.
        !           386: 
        !           387: @menu
        !           388: * Simple Macros::    Macros that always expand the same way.
        !           389: * Argument Macros::  Macros that accept arguments that are substituted
        !           390:                        into the macro expansion.
        !           391: * Predefined::       Predefined macros that are always available.
        !           392: * Stringification::  Macro arguments converted into string constants.
        !           393: * Concatenation::    Building tokens from parts taken from macro arguments.
        !           394: * Undefining::       Cancelling a macro's definition.
        !           395: * Redefining::       Changing a macro's definition.
        !           396: * Macro Pitfalls::   Macros can confuse the unwary.  Here we explain
        !           397:                        several common problems and strange features.
        !           398: @end menu
        !           399: 
        !           400: @node Simple Macros, Argument Macros, Macros, Macros
        !           401: @subsection Simple Macros
        !           402: 
        !           403: A @dfn{simple macro} is a kind of abbreviation.  It is a name which stands
        !           404: for a fragment of code.
        !           405: 
        !           406: Before you can use a macro, you must @dfn{define} it explicitly with the
        !           407: @samp{#define} command.  @samp{#define} is followed by the name of the
        !           408: macro and then the code it should be an abbreviation for.  For example,
        !           409: 
        !           410: @example
        !           411: #define BUFFER_SIZE 1020
        !           412: @end example
        !           413: 
        !           414: @noindent
        !           415: defines a macro named @samp{BUFFER_SIZE} as an abbreviation for the text
        !           416: @samp{1020}.  Therefore, if somewhere after this @samp{#define} command
        !           417: there comes a C statement of the form
        !           418: 
        !           419: @example
        !           420: foo = (char *) xmalloc (BUFFER_SIZE);
        !           421: @end example
        !           422: 
        !           423: @noindent
        !           424: then the C preprocessor will recognize and @dfn{expand} the macro
        !           425: @samp{BUFFER_SIZE}, resulting in
        !           426: 
        !           427: @example
        !           428: foo = (char *) xmalloc (1020);
        !           429: @end example
        !           430: 
        !           431: @noindent
        !           432: the definition must be a single line; however, it may not end in the
        !           433: middle of a multi-line string constant or character constant.
        !           434: 
        !           435: The use of all upper case for macro names is a standard convention.
        !           436: Programs are easier to read when it is possible to tell at a glance which
        !           437: names are macros.
        !           438: 
        !           439: Normally, a macro definition must be a single line, like all C preprocessor
        !           440: commands.  (You can split a long macro definition cosmetically with
        !           441: Backslash-Newline.)  There is one exception: Newlines can be included in
        !           442: the macro definition if within a string or character constant.  By the same
        !           443: token, it is not possible for a macro definition to contain an unbalanced
        !           444: quote character; the definition automatically extends to include the
        !           445: matching quote character that ends the string or character constant.
        !           446: Comments within a macro definition may contain Newlines, which make no
        !           447: difference since the comments are entirely replaced with Spaces regardless
        !           448: of their contents.
        !           449: 
        !           450: Aside from the above, there is no restriction on what can go in a macro
        !           451: body.  Parentheses need not balance.  The body need not resemble valid C
        !           452: code.  (Of course, you might get error messages from the C compiler when
        !           453: you use the macro.)
        !           454: 
        !           455: The C preprocessor scans your program sequentially, so macro definitions
        !           456: take effect at the place you write them.  Therefore, the following input to
        !           457: the C preprocessor
        !           458: 
        !           459: @example
        !           460: foo = X;
        !           461: #define X 4
        !           462: bar = X;
        !           463: @end example
        !           464: 
        !           465: @noindent
        !           466: produces as output
        !           467: 
        !           468: @example
        !           469: foo = X;
        !           470: 
        !           471: bar = 4;
        !           472: @end example
        !           473: 
        !           474: After the preprocessor expands a macro name, the macro's definition body is
        !           475: appended to the front of the remaining input, and the check for macro calls
        !           476: continues.  Therefore, the macro body can contain calls to other macros.
        !           477: For example, after
        !           478: 
        !           479: @example
        !           480: #define BUFSIZE 1020
        !           481: #define TABLESIZE BUFSIZE
        !           482: @end example
        !           483: 
        !           484: @noindent
        !           485: the name @samp{TABLESIZE} when used in the program would go through two
        !           486: stages of expansion, resulting ultimately in @samp{1020}.
        !           487: 
        !           488: This is not at all the same as defining @samp{TABLESIZE} to be @samp{1020}.
        !           489: The @samp{#define} for @samp{TABLESIZE} uses exactly the body you
        !           490: specify---in this case, @samp{BUFSIZE}---and does not check to see whether
        !           491: it too is the name of a macro.  It's only when you @emph{use} @samp{TABLESIZE}
        !           492: that the result of its expansion is checked for more macro names.
        !           493: @xref{Cascaded Macros}.
        !           494: 
        !           495: @node Argument Macros, Predefined, Simple Macros, Macros
        !           496: @subsection Macros with Arguments
        !           497: 
        !           498: A simple macro always stands for exactly the same text, each time it is
        !           499: used.  Macros can be more flexible when they accept @dfn{arguments}.
        !           500: Arguments are fragments of code that you supply each time the macro is
        !           501: used.  These fragments are included in the expansion of the macro according
        !           502: to the directions in the macro definition.
        !           503: 
        !           504: To define a macro that uses arguments, you write a @samp{#define} command
        !           505: with a list of @dfn{argument names} in parentheses after the name of the
        !           506: macro.  The argument names may be any valid C identifiers, separated by
        !           507: commas and optionally whitespace.  The open-parenthesis must follow the
        !           508: macro name immediately, with no space in between.
        !           509: 
        !           510: For example, here is a macro that computes the minimum of two numeric
        !           511: values, as it is defined in many C programs:
        !           512: 
        !           513: @example
        !           514: #define min(X, Y)  ((X) < (Y) ? (X) : (Y))
        !           515: @end example
        !           516: 
        !           517: @noindent
        !           518: (This is not the best way to define a ``minimum'' macro in GNU C.
        !           519: @xref{Side Effects}, for more information.)
        !           520: 
        !           521: To use a macro that expects arguments, you write the name of the macro
        !           522: followed by a list of @dfn{actual arguments} in parentheses. separated by
        !           523: commas.  The number of actual arguments you give must match the number of
        !           524: arguments the macro expects.   Examples of use of the macro @samp{min}
        !           525: include @samp{min (1, 2)} and @samp{min (x + 28, *p)}.
        !           526: 
        !           527: The expansion text of the macro depends on the arguments you use.
        !           528: Each of the argument names of the macro is replaced, throughout the
        !           529: macro definition, with the corresponding actual argument.  Using the
        !           530: same macro @samp{min} defined above, @samp{min (1, 2)} expands into
        !           531: 
        !           532: @example
        !           533: ((1) < (2) ? (1) : (2))
        !           534: @end example
        !           535: 
        !           536: @noindent
        !           537: where @samp{1} has been substituted for @samp{X} and @samp{2} for @samp{Y}.
        !           538: 
        !           539: Likewise, @samp{min (x + 28, *p)} expands into
        !           540: 
        !           541: @example
        !           542: ((x + 28) < (*p) ? (x + 28) : (*p))
        !           543: @end example
        !           544: 
        !           545: Parentheses in the actual arguments must balance; a comma within
        !           546: parentheses does not end an argument.  However, there is no requirement for
        !           547: brackets or braces to balance; thus, if you want to supply @samp{array[x =
        !           548: y, x + 1]} as an argument, you must write it as @samp{array[(x = y, x +
        !           549: 1)]}, which is equivalent C code.
        !           550: 
        !           551: After the actual arguments are substituted into the macro body, the entire
        !           552: result is appended to the front of the remaining input, and the check for
        !           553: macro calls continues.  Therefore, the actual arguments can contain calls
        !           554: to other macros, either with or without arguments, or even to the same
        !           555: macro.  The macro body can also contain calls to other macros.  For
        !           556: example, @samp{min (min (a, b), c)} expands into
        !           557: 
        !           558: @example
        !           559: ((((a) < (b) ? (a) : (b))) < (c)
        !           560:  ? (((a) < (b) ? (a) : (b)))
        !           561:  : (c))
        !           562: @end example
        !           563: 
        !           564: @noindent
        !           565: (Line breaks shown here for clarity would not actually be generated.)
        !           566: 
        !           567: If you use the macro name followed by something other than an
        !           568: open-parenthesis (after ignoring any spaces, tabs and comments that
        !           569: follow), it is not a call to the macro, and the preprocessor does not
        !           570: change what you have written.  Therefore, it is possible for the same name
        !           571: to be a variable or function in your program as well as a macro, and you
        !           572: can choose in each instance whether to refer to the macro (if an actual
        !           573: argument list follows) or the variable or function (if an argument list
        !           574: does not follow).
        !           575: 
        !           576: Such dual use of one name could be confusing and should be avoided
        !           577: except when the two meanings are effectively synonymous: that is, when the
        !           578: name is both a macro and a function and the two have similar effects.  You
        !           579: can think of the name simply as a function; use of the name for purposes
        !           580: other than calling it (such as, to take the address) will refer to the
        !           581: function, while calls will expand the macro and generate better but
        !           582: equivalent code.  For example, you can use a function named @samp{min} in
        !           583: the same source file that defines the macro.  If you write @samp{&min} with
        !           584: no argument list, you refer to the function.  If you write @samp{min (x,
        !           585: bb)}, with an argument list, the macro is expanded.  If you write
        !           586: @samp{(min) (a, bb)}, where the name @samp{min} is not followed by an
        !           587: open-parenthesis, the macro is not expanded, so you wind up with a call to
        !           588: the function @samp{min}.
        !           589: 
        !           590: It is not allowed to define the same name as both a simple macro and
        !           591: a macro with arguments.
        !           592: 
        !           593: In the definition of a macro with arguments, the list of argument names
        !           594: must follow the macro name immediately with no space in between.  If there
        !           595: is a space after the macro name, the macro is defined as taking no
        !           596: arguments, and all the rest of the name is taken to be the expansion.  The
        !           597: reason for this is that it is often useful to define a macro that takes no
        !           598: arguments and whose definition begins with an identifier in parentheses.
        !           599: This rule about spaces makes it possible for you to do either this:
        !           600: 
        !           601: @example
        !           602: #define FOO(x) - 1 / (x)
        !           603: @end example
        !           604: 
        !           605: @noindent
        !           606: (which defines @samp{FOO} to take an argument and expand into minus the
        !           607: reciprocal of that argument) or this:
        !           608: 
        !           609: @example
        !           610: #define BAR (x) - 1 / (x)
        !           611: @end example
        !           612: 
        !           613: @noindent
        !           614: (which defines @samp{BAR} to take no argument and always expand into
        !           615: @samp{(x) - 1 / (x)}).
        !           616: 
        !           617: Note that the @emph{uses} of a macro with arguments can have spaces before
        !           618: the left parenthesis; it's the @emph{definition} where it matters whether
        !           619: there is a space.
        !           620: 
        !           621: @node Predefined, Stringification, Argument Macros, Macros
        !           622: @subsection Predefined Macros
        !           623: 
        !           624: @cindex predefined macros
        !           625: Several simple macros are predefined.  You can use them without giving
        !           626: definitions for them.  They fall into two classes: standard macros and
        !           627: system-specific macros.
        !           628: 
        !           629: @menu
        !           630: * Standard Predefined::     Standard predefined macros.
        !           631: * Nonstandard Predefined::  Nonstandard predefined macros.
        !           632: @end menu
        !           633: 
        !           634: @node Standard Predefined, Nonstandard Predefined, Predefined, Predefined
        !           635: @subsubsection Standard Predefined Macros
        !           636: 
        !           637: The standard predefined macros are available with the same meanings
        !           638: regardless of the machine or operating system on which you are using GNU C.
        !           639: Their names all start and end with double underscores.  Those preceding
        !           640: @code{__GNUC__} in this table are standardized by ANSI C; the rest are
        !           641: GNU C extensions.
        !           642: 
        !           643: @table @code
        !           644: @item __FILE__
        !           645: @findex __FILE__
        !           646: This macro expands to the name of the current input file, in the form
        !           647: of a C string constant.
        !           648: 
        !           649: @item __LINE__
        !           650: @findex __LINE__
        !           651: This macro expands to the current input line number, in the form of a
        !           652: decimal integer constant.  While we call it a predefined macro, it's
        !           653: a pretty strange macro, since its ``definition'' changes with each
        !           654: new line of source code.
        !           655: 
        !           656: This and @samp{__FILE__} are useful in generating an error message to
        !           657: report an inconsistency detected by the program; the message can state
        !           658: the source line at which the inconsistency was detected.  For example,
        !           659: 
        !           660: @example
        !           661: fprintf (stderr, "Internal error: negative string length "
        !           662:                  "%d at %s, line %d.",
        !           663:          length, __FILE__, __LINE__);
        !           664: @end example
        !           665: 
        !           666: A @samp{#include} command changes the expansions of @samp{__FILE__}
        !           667: and @samp{__LINE__} to correspond to the included file.  At the end of
        !           668: that file, when processing resumes on the input file that contained
        !           669: the @samp{#include} command, the expansions of @samp{__FILE__} and
        !           670: @samp{__LINE__} revert to the values they had before the
        !           671: @samp{#include} (but @samp{__LINE__} is then incremented by one as
        !           672: processing moves to the line after the @samp{#include}).
        !           673: 
        !           674: The expansions of both @samp{__FILE__} and @samp{__LINE__} are altered
        !           675: if a @samp{#line} command is used.  @xref{Combining Sources}.
        !           676: 
        !           677: @item __DATE__
        !           678: @findex __DATE__
        !           679: This macro expands to a string constant that describes the date on
        !           680: which the preprocessor is being run.  The string constant contains
        !           681: eleven characters and looks like @samp{"Jan 29 1987"} or @w{@samp{"Apr
        !           682: 1 1905"}}.
        !           683: 
        !           684: @item __TIME__
        !           685: @findex __TIME__
        !           686: This macro expands to a string constant that describes the time at
        !           687: which the preprocessor is being run.  The string constant contains
        !           688: eight characters and looks like @samp{"23:59:01"}.
        !           689: 
        !           690: @item __STDC__
        !           691: @findex __STDC__
        !           692: This macro expands to the constant 1, to signify that this is ANSI
        !           693: Standard C.  (Whether that is actually true depends on what C compiler
        !           694: will operate on the output from the preprocessor.)
        !           695: 
        !           696: @item __GNUC__
        !           697: This macro is defined if and only if this is GNU C.  This macro is
        !           698: defined only when the entire GNU C compiler is in use; if you invoke
        !           699: the preprocessor directly, @samp{__GNUC__} is undefined.
        !           700: 
        !           701: @item __STRICT_ANSI__
        !           702: This macro is defined if and only if the @samp{-ansi} switch was
        !           703: specified when GNU C was invoked.  Its definition is the null string.
        !           704: This macro exists primarily to direct certain GNU header files not to
        !           705: define certain traditional Unix constructs which are incompatible with
        !           706: ANSI C.
        !           707: 
        !           708: @item __VERSION__
        !           709: This macro expands to a string which describes the version number of
        !           710: GNU C.  The string is normally a sequence of decimal numbers separated
        !           711: by periods, such as @samp{"1.18"}.  The only reasonable use of this
        !           712: macro is to incorporate it into a string constant.
        !           713: 
        !           714: @item __OPTIMIZE__
        !           715: This macro is defined in optimizing compilations.  It causes certain
        !           716: GNU header files to define alternative macro definitions for some
        !           717: system library functions.  It is unwise to refer to or test the
        !           718: definition of this macro unless you make very sure that programs will
        !           719: execute with the same effect regardless.
        !           720: 
        !           721: @item __CHAR_UNSIGNED__
        !           722: This macro is defined if and only if the data type @code{char} is
        !           723: unsigned on the target machine.  It exists to cause the standard
        !           724: header file @file{limit.h} to work correctly.  It is bad practice
        !           725: to refer to this macro yourself; instead, refer to the standard
        !           726: macros defined in @file{limit.h}.
        !           727: @end table
        !           728: 
        !           729: @node Nonstandard Predefined,, Standard Predefined, Predefined
        !           730: @subsubsection Nonstandard Predefined Macros
        !           731: 
        !           732: The C preprocessor normally has several predefined macros that vary between
        !           733: machines because their purpose is to indicate what type of system and
        !           734: machine is in use.  This manual, being for all systems and machines, cannot
        !           735: tell you exactly what their names are; instead, we offer a list of some
        !           736: typical ones.
        !           737: 
        !           738: Some nonstandard predefined macros describe the operating system in use,
        !           739: with more or less specificity.  For example,
        !           740: 
        !           741: @table @code
        !           742: @item unix
        !           743: @findex unix
        !           744: @samp{unix} is normally predefined on all Unix systems.
        !           745: 
        !           746: @item BSD
        !           747: @findex BSD
        !           748: @samp{BSD} is predefined on recent versions of Berkeley Unix
        !           749: (perhaps only in version 4.3).
        !           750: @end table
        !           751: 
        !           752: Other nonstandard predefined macros describe the kind of CPU, with more or
        !           753: less specificity.  For example,
        !           754: 
        !           755: @table @code
        !           756: @item vax
        !           757: @findex vax
        !           758: @samp{vax} is predefined on Vax computers.
        !           759: 
        !           760: @item mc68000
        !           761: @findex mc68000
        !           762: @samp{mc68000} is predefined on most computers whose CPU is a Motorola
        !           763: 68000, 68010 or 68020.
        !           764: 
        !           765: @item m68k
        !           766: @findex m68k
        !           767: @samp{m68k} is also predefined on most computers whose CPU is a 68000,
        !           768: 68010 or 68020; however, some makers use @samp{mc68000} and some use
        !           769: @samp{m68k}.  Some predefine both names.  What happens in GNU C
        !           770: depends on the system you are using it on.
        !           771: 
        !           772: @item M68020
        !           773: @findex M68020
        !           774: @samp{M68020} has been observed to be predefined on some systems that
        !           775: use 68020 CPUs---in addition to @samp{mc68000} and @samp{m68k} that
        !           776: are less specific.
        !           777: 
        !           778: @item ns32000
        !           779: @findex ns32000
        !           780: @samp{ns32000} is predefined on computers which use the National
        !           781: Semiconductor 32000 series CPU.
        !           782: @end table
        !           783: 
        !           784: Yet other nonstandard predefined macros describe the manufacturer of
        !           785: the system.  For example,
        !           786: 
        !           787: @table @code
        !           788: @item sun
        !           789: @findex sun
        !           790: @samp{sun} is predefined on all models of Sun computers.
        !           791: 
        !           792: @item pyr
        !           793: @findex pyr
        !           794: @samp{pyr} is predefined on all models of Pyramid computers.
        !           795: 
        !           796: @item sequent
        !           797: @findex sequent
        !           798: @samp{sequent} is predefined on all models of Sequent computers.
        !           799: @end table
        !           800: 
        !           801: These predefined symbols are not only nonstandard, they are contrary to the
        !           802: ANSI standard because their names do not start with underscores.  However,
        !           803: the GNU C preprocessor would be useless if it did not predefine the same
        !           804: names that are normally predefined on the system and machine you are using.
        !           805: Even system header files check the predefined names and will generate
        !           806: incorrect declarations if they do not find the names that are expected.
        !           807: 
        !           808: The set of nonstandard predefined names in the GNU C preprocessor is
        !           809: controlled by the macro @samp{CPP_PREDEFINES}, which should be a string
        !           810: containing @samp{-D} options, separated by spaces.  For example, on the Sun,
        !           811: the definition
        !           812: 
        !           813: @example
        !           814: #define CPP_PREDEFINES "-Dmc68000 -Dsun -Dunix -Dm68k"
        !           815: @end example
        !           816: 
        !           817: @noindent
        !           818: is used.
        !           819: 
        !           820: The @samp{-ansi} option which requests complete support for ANSI C
        !           821: inhibits the definition of these predefined symbols.
        !           822: 
        !           823: @node Stringification, Concatenation, Predefined, Macros
        !           824: @subsection Stringification
        !           825: 
        !           826: @cindex stringification
        !           827: @dfn{Stringification} means turning a code fragment into a string constant
        !           828: whose contents are the text for the code fragment.  For example,
        !           829: stringifying @samp{foo (z)} results in @samp{"foo (z)"}.
        !           830: 
        !           831: In the C preprocessor, stringification is an option available when macro
        !           832: arguments are substituted into the macro definition.  In the body of the
        !           833: definition, when an argument name appears, the character @samp{#} before
        !           834: the name specifies stringification of the corresponding actual argument
        !           835: when it is substituted at that point in the definition.  The same argument
        !           836: may be substituted in other places in the definition without
        !           837: stringification if the argument name appears in those places with no
        !           838: @samp{#}.
        !           839: 
        !           840: Here is an example of a macro definition that uses stringification:
        !           841: 
        !           842: @example
        !           843: #define WARN_IF(EXP) \
        !           844: do @{ if (EXP) fprintf (stderr, "Warning: " #EXP "\n"); @} while (0)
        !           845: @end example
        !           846: 
        !           847: @noindent
        !           848: Here the actual argument for @samp{EXP} is substituted once as given,
        !           849: into the @samp{if} statement, and once as stringified, into the
        !           850: argument to @samp{fprintf}.  The @samp{do} and @samp{while (0)} are
        !           851: a kludge to make it possible to write @samp{WARN_IF (@var{arg});},
        !           852: which the resemblance of @samp{WARN_IF} to a function would make
        !           853: C programmers want to do; @pxref{Swallow Semicolon}).
        !           854: 
        !           855: The stringification feature is limited to transforming one macro argument
        !           856: into one string constant: there is no way to combine the argument with
        !           857: other text and then stringify it all together.  But the example above shows
        !           858: how an equivalent result can be obtained in ANSI Standard C using the
        !           859: feature that adjacent string constants are concatenated as one string
        !           860: constant.  The preprocessor stringifies @samp{EXP}'s actual argument
        !           861: into a separate string constant, resulting in text like
        !           862: 
        !           863: @example
        !           864: do @{ if (x == 0) fprintf (stderr, "Warning: " "x == 0" "\n"); @} while (0)
        !           865: @end example
        !           866: 
        !           867: @noindent
        !           868: but the C compiler then sees three consecutive string constants and
        !           869: concatenates them into one, producing effectively
        !           870: 
        !           871: @example
        !           872: do @{ if (x == 0) fprintf (stderr, "Warning: x == 0\n"); @} while (0)
        !           873: @end example
        !           874: 
        !           875: Stringification in C involves more than putting doublequote characters
        !           876: around the fragment; it is necessary to put backslashes in front of all
        !           877: doublequote characters, and all backslashes in string and character
        !           878: constants, in order to get a valid C string constant with the proper
        !           879: contents.  Thus, stringifying @samp{p = "foo\n";} results in @samp{"p =
        !           880: \"foo\\n\";"}.  However, backslashes that are not inside of string or
        !           881: character constants are not duplicated: @samp{\n} by itself stringifies to
        !           882: @samp{"\n"}.
        !           883: 
        !           884: Whitespace (including comments) in the text being stringified is handled
        !           885: according to precise rules.  All leading and trailing whitespace is ignored.
        !           886: Any sequence of whitespace in the middle of the text is converted to
        !           887: a single space in the stringified result.
        !           888: 
        !           889: @node Concatenation, Undefining, Stringification, Macros
        !           890: @subsection Concatenation
        !           891: 
        !           892: @cindex concatenation
        !           893: @dfn{Concatenation} means joining two strings into one.  In the context
        !           894: of macro expansion, concatenation refers to joining two lexical units
        !           895: into one longer one.  Specifically, an actual argument to the macro can be
        !           896: concatenated with another actual argument or with fixed text to produce
        !           897: a longer name.  The longer name might be the name of a function,
        !           898: variable or type, or a C keyword; it might even be the name of another
        !           899: macro, in which case it will be expanded.
        !           900: 
        !           901: When you define a macro, you request concatenation with the special
        !           902: operator @samp{##} in the macro body.  When the macro is called,
        !           903: after actual arguments are substituted, all @samp{##} operators are
        !           904: deleted, and so is any whitespace next to them (including whitespace
        !           905: that was part of an actual argument).  The result is to concatenate
        !           906: the syntactic tokens on either side of the @samp{##}.
        !           907: 
        !           908: Consider a C program that interprets named commands.  There probably needs
        !           909: to be a table of commands, perhaps an array of structures declared as
        !           910: follows:
        !           911: 
        !           912: @example
        !           913: struct command
        !           914: @{
        !           915:   char *name;
        !           916:   void (*function) ();
        !           917: @};
        !           918: 
        !           919: struct command commands[] =
        !           920: @{
        !           921:   @{ "quit", quit_command@},
        !           922:   @{ "help", help_command@},
        !           923:   @dots{}
        !           924: @};
        !           925: @end example
        !           926: 
        !           927: It would be cleaner not to have to give each command name twice, once in
        !           928: the string constant and once in the function name.  A macro which takes the
        !           929: name of a command as an argument can make this unnecessary.  The string
        !           930: constant can be created with stringification, and the function name by
        !           931: concatenating the argument with @samp{_command}.  Here is how it is done:
        !           932: 
        !           933: @example
        !           934: #define COMMAND(NAME)  @{ #NAME, NAME ## _command @}
        !           935: 
        !           936: struct command commands[] =
        !           937: @{
        !           938:   COMMAND (quit),
        !           939:   COMMAND (help),
        !           940:   @dots{}
        !           941: @};
        !           942: @end example
        !           943: 
        !           944: The usual case of concatenation is concatenating two names (or a name and a
        !           945: number) into a longer name.  But this isn't the only valid case.  It is
        !           946: also possible to concatenate two numbers (or a number and a name, such as
        !           947: @samp{1.5} and @samp{e3}) into a number.  Also, multi-character operators
        !           948: such as @samp{+=} can be formed by concatenation.  In some cases it is even
        !           949: possible to piece together a string constant.  However, two pieces of text
        !           950: that don't together form a valid lexical unit cannot be concatenated.  For
        !           951: example, concatenation with @samp{x} on one side and @samp{+} on the other
        !           952: is not meaningful because those two characters can't fit together in any
        !           953: lexical unit of C.  The ANSI standard says that such attempts at
        !           954: concatenation are undefined, but in the GNU C preprocessor it is well
        !           955: defined: it puts the @samp{x} and @samp{+} side by side with no particular
        !           956: special results.
        !           957: 
        !           958: Keep in mind that the C preprocessor converts comments to whitespace before
        !           959: macros are even considered.  Therefore, you cannot create a comment by
        !           960: concatenating @samp{/} and @samp{*}: the @samp{/*} sequence that starts a
        !           961: comment is not a lexical unit, but rather the beginning of a ``long'' space
        !           962: character.  Also, you can freely use comments next to a @samp{##} in a
        !           963: macro definition, or in actual arguments that will be concatenated, because
        !           964: the comments will be converted to spaces at first sight, and concatenation
        !           965: will later discard the spaces.
        !           966: 
        !           967: @node Undefining, Redefining, Concatenation, Macros
        !           968: @subsection Undefining Macros
        !           969: 
        !           970: @cindex undefining macros
        !           971: To @dfn{undefine} a macro means to cancel its definition.  This is done
        !           972: with the @samp{#undef} command.  @samp{#undef} is followed by the macro
        !           973: name to be undefined.
        !           974: 
        !           975: Like definition, undefinition occurs at a specific point in the source
        !           976: file, and it applies starting from that point.  The name ceases to be a
        !           977: macro name, and from that point on it is treated by the preprocessor as if
        !           978: it had never been a macro name.
        !           979: 
        !           980: For example,
        !           981: 
        !           982: @example
        !           983: #define FOO 4
        !           984: x = FOO;
        !           985: #undef FOO
        !           986: x = FOO;
        !           987: @end example
        !           988: 
        !           989: @noindent
        !           990: expands into
        !           991: 
        !           992: @example
        !           993: x = 4;
        !           994: 
        !           995: x = FOO;
        !           996: @end example
        !           997: 
        !           998: @noindent
        !           999: In this example, @samp{FOO} had better be a variable or function as well
        !          1000: as (temporarily) a macro, in order for the result of the expansion to be
        !          1001: valid C code.
        !          1002: 
        !          1003: The same form of @samp{#undef} command will cancel definitions with
        !          1004: arguments or definitions that don't expect arguments.  The @samp{#undef}
        !          1005: command has no effect when used on a name not currently defined as a macro.
        !          1006: 
        !          1007: @node Redefining, Macro Pitfalls, Undefining, Macros
        !          1008: @subsection Redefining Macros
        !          1009: 
        !          1010: @cindex redefining macros
        !          1011: @dfn{Redefining} a macro means defining (with @samp{#define}) a name that
        !          1012: is already defined as a macro.
        !          1013: 
        !          1014: A redefinition is trivial if the new definition is transparently identical
        !          1015: to the old one.  You probably wouldn't deliberately write a trivial
        !          1016: redefinition, but they can happen automatically when a header file is
        !          1017: included more than once (@pxref{Header Files}), so they are accepted
        !          1018: silently and without effect.
        !          1019: 
        !          1020: Nontrivial redefinition is considered likely to be an error, so
        !          1021: it provokes a warning message from the preprocessor.  However, sometimes it
        !          1022: is useful to change the definition of a macro in mid-compilation.  You can
        !          1023: inhibit the warning by undefining the macro with @samp{#undef} before the
        !          1024: second definition.
        !          1025: 
        !          1026: In order for a reefinition to be trivial, the new definition must exactly match
        !          1027: the one already in effect, with two possible exceptions:
        !          1028: 
        !          1029: @itemize @bullet
        !          1030: @item
        !          1031: Whitespace may be added or deleted at the beginning or the end.
        !          1032: 
        !          1033: @item
        !          1034: Whitespace may be changed in the middle (but not inside strings).
        !          1035: However, it may not be eliminated entirely, and it may not be added
        !          1036: where there was no whitespace at all.
        !          1037: @end itemize
        !          1038: 
        !          1039: Recall that a comment counts as whitespace.
        !          1040: 
        !          1041: @node Macro Pitfalls,, Redefining, Macros
        !          1042: @subsection Pitfalls and Subtleties of Macros
        !          1043: 
        !          1044: In this section we describe some special rules that apply to macros and
        !          1045: macro expansion, and point out certain cases in which the rules have 
        !          1046: counterintuitive consequences that you must watch out for.
        !          1047: 
        !          1048: @menu
        !          1049: * Misnesting::        Macros can contain unmatched parentheses.
        !          1050: * Macro Parentheses:: Why apparently superfluous parentheses
        !          1051:                          may be necessary to avoid incorrect grouping.
        !          1052: * Swallow Semicolon:: Macros that look like functions
        !          1053:                          but expand into compound statements.
        !          1054: * Side Effects::      Unsafe macros that cause trouble when
        !          1055:                          arguments contain side effects.
        !          1056: * Self-Reference::    Macros whose definitions use the macros' own names.
        !          1057: * Argument Prescan::  Actual arguments are checked for macro calls
        !          1058:                          before they are substituted.
        !          1059: * Cascaded Macros::   Macros whose definitions use other macros.
        !          1060: @end menu
        !          1061: 
        !          1062: @node Misnesting, Macro Parentheses, Macro Pitfalls, Macro Pitfalls
        !          1063: @subsubsection Improperly Nested Constructs
        !          1064: 
        !          1065: Recall that when a macro is called with arguments, the arguments are
        !          1066: substituted into the macro body and the result is checked, together with
        !          1067: the rest of the input file, for more macro calls.
        !          1068: 
        !          1069: It is possible to piece together a macro call coming partially from the
        !          1070: macro body and partially from the actual arguments.  For example,
        !          1071: 
        !          1072: @example
        !          1073: #define double(x) (2*(x))
        !          1074: #define call_with_1(x) x(1)
        !          1075: @end example
        !          1076: 
        !          1077: @noindent
        !          1078: would expand @samp{call_with_1 (double)} into @samp{(2*(1))}.
        !          1079: 
        !          1080: Macro definitions do not have to have balanced parentheses.  By writing an
        !          1081: unbalanced open parenthesis in a macro body, it is possible to create a
        !          1082: macro call that begins inside the macro body but ends outside of it.  For
        !          1083: example,
        !          1084: 
        !          1085: @example
        !          1086: #define strange(file) fprintf (file, "%s %d",
        !          1087: @dots{}
        !          1088: strange(stderr) p, 35)
        !          1089: @end example
        !          1090: 
        !          1091: @noindent
        !          1092: This bizarre example expands to @samp{fprintf (stderr, "%s %d", p, 35)}!
        !          1093: 
        !          1094: @node Macro Parentheses, Swallow Semicolon, Misnesting, Macro Pitfalls
        !          1095: @subsubsection Unintended Grouping of Arithmetic
        !          1096: 
        !          1097: You may have noticed that in most of the macro definition examples shown
        !          1098: above, each occurrence of a macro argument name had parentheses around it.
        !          1099: In addition, another pair of parentheses usually surround the entire macro
        !          1100: definition.  Here is why it is best to write macros that way.
        !          1101: 
        !          1102: Suppose you define a macro as follows
        !          1103: 
        !          1104: @example
        !          1105: #define ceil_div(x, y) (x + y - 1) / y
        !          1106: @end example
        !          1107: 
        !          1108: @noindent
        !          1109: whose purpose is to divide, rounding up.  (One use for this
        !          1110: operation is to compute how many @samp{int}'s are needed to hold
        !          1111: a certain number of @samp{char}'s.)  Then suppose it is used as follows:
        !          1112: 
        !          1113: @example
        !          1114: a = ceil_div (b & c, sizeof (int));
        !          1115: @end example
        !          1116: 
        !          1117: @noindent
        !          1118: This expands into
        !          1119: 
        !          1120: @example
        !          1121: a = (b & c + sizeof (int) - 1) / sizeof (int);
        !          1122: @end example
        !          1123: 
        !          1124: @noindent
        !          1125: which does not do what is intended.  The operator-precedence rules of
        !          1126: C make it equivalent to this:  
        !          1127: 
        !          1128: @example
        !          1129: a = (b & (c + sizeof (int) - 1)) / sizeof (int);
        !          1130: @end example
        !          1131: 
        !          1132: @noindent
        !          1133: But what we want is this:
        !          1134: 
        !          1135: @example
        !          1136: a = ((b & c) + sizeof (int) - 1)) / sizeof (int);
        !          1137: @end example
        !          1138: 
        !          1139: @noindent
        !          1140: Defining the macro as
        !          1141: 
        !          1142: @example
        !          1143: #define ceil_div(x, y) ((x) + (y) - 1) / (y)
        !          1144: @end example
        !          1145: 
        !          1146: @noindent
        !          1147: provides the desired result.
        !          1148: 
        !          1149: However, unintended grouping can result in another way.  Consider
        !          1150: @samp{sizeof ceil_div(1, 2)}.  That has the appearance of a C expression
        !          1151: that would compute the size of the type of @samp{ceil_div (1, 2)}, but in
        !          1152: fact it means something very different.  Here is what it expands to:
        !          1153: 
        !          1154: @example
        !          1155: sizeof ((1) + (2) - 1) / (2)
        !          1156: @end example
        !          1157: 
        !          1158: @noindent
        !          1159: This would take the size of an integer and divide it by two.  The precedence
        !          1160: rules have put the division outside the @samp{sizeof} when it was intended
        !          1161: to be inside.
        !          1162: 
        !          1163: Parentheses around the entire macro definition can prevent such problems.
        !          1164: Here, then, is the recommended way to define @samp{ceil_div}:
        !          1165: 
        !          1166: @example
        !          1167: #define ceil_div(x, y) (((x) + (y) - 1) / (y))
        !          1168: @end example
        !          1169: 
        !          1170: @node Swallow Semicolon, Side Effects, Macro Parentheses, Macro Pitfalls
        !          1171: @subsubsection Swallowing the Semicolon
        !          1172: 
        !          1173: @cindex semicolons (after macro calls)
        !          1174: Often it is desirable to define a macro that expands into a compound
        !          1175: statement.  Consider, for example, the following macro, that advances a
        !          1176: pointer (the argument @samp{p} says where to find it) across whitespace
        !          1177: characters:
        !          1178: 
        !          1179: @example
        !          1180: #define SKIP_SPACES (p, limit)  \
        !          1181: @{ register char *lim = (limit); \
        !          1182:   while (p != lim) @{            \
        !          1183:     if (*p++ != ' ') @{          \
        !          1184:       p--; break; @}@}@}
        !          1185: @end example
        !          1186: 
        !          1187: @noindent
        !          1188: Here Backslash-Newline is used to split the macro definition, which must
        !          1189: be a single line, so that it resembles the way such C code would be
        !          1190: layed out if not part of a macro definition.
        !          1191: 
        !          1192: A call to this macro might be @samp{SKIP_SPACES (p, lim)}.  Strictly
        !          1193: speaking, the call expands to a compound statement, which is a complete
        !          1194: statement with no need for a semicolon to end it.  But it looks like a
        !          1195: function call.  So it minimizes confusion if you can use it like a function
        !          1196: call, writing a semicolon afterward, as in @samp{SKIP_SPACES (p, lim);}
        !          1197: 
        !          1198: But this can cause trouble before @samp{else} statements, because the
        !          1199: semicolon is actually a null statement.  Suppose you write
        !          1200: 
        !          1201: @example
        !          1202: if (*p != 0)
        !          1203:   SKIP_SPACES (p, lim);
        !          1204: else @dots{}
        !          1205: @end example
        !          1206: 
        !          1207: @noindent
        !          1208: The presence of two statements---the compound statement and a null
        !          1209: statement---in between the @samp{if} condition and the @samp{else}
        !          1210: makes invalid C code.
        !          1211: 
        !          1212: The definition of the macro @samp{SKIP_SPACES} can be altered to solve
        !          1213: this problem, using a @samp{do @dots{} while} statement.  Here is how:
        !          1214:   
        !          1215: @example
        !          1216: #define SKIP_SPACES (p, limit)     \
        !          1217: do @{ register char *lim = (limit); \
        !          1218:      while (p != lim) @{            \
        !          1219:        if (*p++ != ' ') @{          \
        !          1220:          p--; break; @}@}@}           \
        !          1221: while (0)
        !          1222: @end example
        !          1223: 
        !          1224: Now @samp{SKIP_SPACES (p, lim);} expands into
        !          1225: 
        !          1226: @example
        !          1227: do @{@dots{}@} while (0);
        !          1228: @end example
        !          1229: 
        !          1230: @noindent
        !          1231: which is one statement.
        !          1232: 
        !          1233: @node Side Effects, Self-Reference, Swallow Semicolon, Macro Pitfalls
        !          1234: @subsubsection Duplication of Side Effects
        !          1235: 
        !          1236: @cindex side effects (in macro arguments)
        !          1237: @cindex unsafe macros
        !          1238: Many C programs define a macro @samp{min}, for ``minimum'', like this:
        !          1239: 
        !          1240: @example
        !          1241: #define min(X, Y)  ((X) < (Y) ? (X) : (Y))
        !          1242: @end example
        !          1243: 
        !          1244: When you use this macro with an argument containing a side effect,
        !          1245: as shown here,
        !          1246: 
        !          1247: @example
        !          1248: next = min (x + y, foo (z));
        !          1249: @end example
        !          1250: 
        !          1251: @noindent
        !          1252: it expands as follows:
        !          1253: 
        !          1254: @example
        !          1255: next = ((x + y) < (foo (z)) ? (x + y) : (foo (z)));
        !          1256: @end example
        !          1257: 
        !          1258: @noindent
        !          1259: where @samp{x + y} has been substituted for @samp{X} and @samp{foo (z)}
        !          1260: for @samp{Y}.
        !          1261: 
        !          1262: The function @samp{foo} is used only once in the statement as it appears
        !          1263: in the program, but the expression @samp{foo (z)} has been substituted
        !          1264: twice into the macro expansion.  As a result, @samp{foo} might be called
        !          1265: two times when the statement is executed.  If it has side effects or
        !          1266: if it takes a long time to compute, the results might not be what you
        !          1267: intended.  We say that @samp{min} is an @dfn{unsafe} macro.
        !          1268: 
        !          1269: The best solution to this problem is to define @samp{min} in a way that
        !          1270: computes the value of @samp{foo (z)} only once.  The C language offers no
        !          1271: way standard way to do this, but it can be done with GNU C extensions as
        !          1272: follows:
        !          1273: 
        !          1274: @example
        !          1275: #define min(X, Y)                     \
        !          1276: (@{ typeof (X) __x = (X), __y = (Y);   \
        !          1277:    (__x < __y) ? __x : __y; @})
        !          1278: @end example
        !          1279: 
        !          1280: If you do not wish to use GNU C extensions, the only solution is to be
        !          1281: careful when @emph{using} the macro @samp{min}.  For example, you can
        !          1282: calculate the value of @samp{foo (z)}, save it in a variable, and use that
        !          1283: variable in @samp{min}:
        !          1284: 
        !          1285: @example
        !          1286: #define min(X, Y)  ((X) < (Y) ? (X) : (Y))
        !          1287: @dots{}
        !          1288: @{
        !          1289:   int tem = foo (z);
        !          1290:   next = min (x + y, tem);
        !          1291: @}
        !          1292: @end example
        !          1293: 
        !          1294: @noindent
        !          1295: (where I assume that @samp{foo} returns type @samp{int}).
        !          1296: 
        !          1297: @node Self-Reference, Argument Prescan, Side Effects, Macro Pitfalls
        !          1298: @subsubsection Self-Referential Macros
        !          1299: 
        !          1300: @cindex self-reference
        !          1301: A @dfn{self-referential} macro is one whose name appears in its definition.
        !          1302: A special feature of ANSI Standard C is that the self-reference is not
        !          1303: considered a macro call.  It is passed into the preprocessor output
        !          1304: unchanged.
        !          1305: 
        !          1306: Let's consider an example:
        !          1307: 
        !          1308: @example
        !          1309: #define foo (4 + foo)
        !          1310: @end example
        !          1311: 
        !          1312: @noindent
        !          1313: where @samp{foo} is also a variable in your program.
        !          1314: 
        !          1315: Following the ordinary rules, each reference to @samp{foo} will expand into
        !          1316: @samp{(4 + foo)}; then this will be rescanned and will expand into @samp{(4
        !          1317: + (4 + foo))}; and so on until it causes a fatal error (memory full) in the
        !          1318: preprocessor.
        !          1319: 
        !          1320: However, the special rule about self-reference cuts this process short
        !          1321: after one step, at @samp{(4 + foo)}.  Therefore, this macro definition
        !          1322: has the possibly useful effect of causing the program to add 4 to
        !          1323: the value of @samp{foo} wherever @samp{foo} is referred to.
        !          1324: 
        !          1325: In most cases, it is a bad idea to take advantage of this feature.
        !          1326: A person reading the program who sees that @samp{foo} is a variable will
        !          1327: not expect that it is a macro as well.  The reader will come across a
        !          1328: the identifier @samp{foo} in the program and think its value should be
        !          1329: that of the variable @samp{foo}, whereas in fact the value is four
        !          1330: greater.
        !          1331: 
        !          1332: The special rule for self-reference applies also to @dfn{indirect}
        !          1333: self-reference.  This is the case where a macro @var{x} expands to use a
        !          1334: macro @samp{y}, and @samp{y}'s expansion refers to the macro @samp{x}.  The
        !          1335: resulting reference to @samp{x} comes indirectly from the expansion of
        !          1336: @samp{x}, so it is a self-reference and is not further expanded.  Thus,
        !          1337: after
        !          1338: 
        !          1339: @example
        !          1340: #define x (4 + y)
        !          1341: #define y (2 * x)
        !          1342: @end example
        !          1343: 
        !          1344: @noindent
        !          1345: @samp{x} would expand into @samp{(4 + (2 * x))}.  Clear?
        !          1346: 
        !          1347: But suppose @samp{y} is used elsewhere, not from the definition of @samp{x}.
        !          1348: Then the use of @samp{x} in the expansion of @samp{y} is not a self-reference
        !          1349: because @samp{x} is not ``in progress''.  So it does expand.  However,
        !          1350: the expansion of @samp{x} contains a reference to @samp{y}, and that
        !          1351: is an indirect self-reference now because @samp{y} is ``in progress''.
        !          1352: The result is that @samp{y} expands to @samp{(2 * (4 + y))}.
        !          1353: 
        !          1354: It is not clear that this behavior would ever be useful, but it is specified
        !          1355: by the ANSI C standard, so you need to understand it.
        !          1356: 
        !          1357: @node Argument Prescan, Cascaded Macros, Self-Reference, Macro Pitfalls
        !          1358: @subsubsection Separate Expansion of Macro Arguments
        !          1359: 
        !          1360: We have explained that the expansion of a macro, including the substituted
        !          1361: actual arguments, is scanned over again for macro calls to be expanded.
        !          1362: 
        !          1363: What really happens is more subtle: first each actual argument text is scanned
        !          1364: separately for macro calls.  Then the results of this are substituted into
        !          1365: the macro body to produce the macro expansion, and the macro expansion
        !          1366: is scanned again for macros to expand.
        !          1367: 
        !          1368: The result is that the actual arguments are scanned @emph{twice} to expand
        !          1369: macro calls in them.
        !          1370: 
        !          1371: Most of the time, this has no effect.  If the actual argument contained
        !          1372: any macro calls, they are expanded during the first scan.  The result
        !          1373: therefore contains no macro calls, so the second scan does not change it.
        !          1374: If the actual argument were substituted as given, with no prescan,
        !          1375: the single remaining scan would find the same macro calls and produce
        !          1376: the same results.
        !          1377: 
        !          1378: You might expect the double scan to change the results when a
        !          1379: self-referential macro is used in an actual argument of another macro
        !          1380: (@pxref{Self-Reference}): the self-referential macro would be expanded once
        !          1381: in the first scan, and a second time in the second scan.  But this is not
        !          1382: what happens.  The self-references that do not expand in the first scan are
        !          1383: marked so that they will not expand in the second scan either.
        !          1384: 
        !          1385: The prescan is not done when an argument is stringified or concatenated.
        !          1386: (More precisely, stringification and concatenation use the argument as
        !          1387: written, in un-prescanned form.  The same actual argument would be used in
        !          1388: prescanned form if it is substituted elsewhere without stringification or
        !          1389: concatenation.)  Thus,
        !          1390: 
        !          1391: @example
        !          1392: #define str(s) #s
        !          1393: #define foo 4
        !          1394: str (foo)
        !          1395: @end example
        !          1396: 
        !          1397: @noindent
        !          1398: expands to @samp{"foo"}.  Once more, prescan has been prevented from
        !          1399: having any noticeable effect.
        !          1400: 
        !          1401: You might now ask, ``Why mention the prescan, if it makes no difference?
        !          1402: Why not skip it and make the preprocessor faster?''  The answer is that the
        !          1403: prescan does make a difference in two special cases:
        !          1404: 
        !          1405: @itemize @bullet
        !          1406: @item
        !          1407: Nested calls to a macro.
        !          1408: 
        !          1409: @item
        !          1410: Macros that call other macros that stringify or concatenate.
        !          1411: @end itemize
        !          1412: 
        !          1413: We say that @dfn{nested} calls to a macro occur when a macro's actual
        !          1414: argument contains a call to that very macro.  For example, if @samp{f}
        !          1415: is a macro that expects one argument, @samp{f (f (1))} is a nested
        !          1416: pair of calls to @samp{f}.  The desired expansion is made by
        !          1417: expanding @samp{f (1)} and substituting that into the definition of
        !          1418: @samp{f}.  The prescan causes the expected result to happen.
        !          1419: Without the prescan, @samp{f (1)} itself would be substituted as
        !          1420: an actual argument, and the inner use of @samp{f} would appear
        !          1421: during the main scan as an indirect self-reference and would not
        !          1422: be expanded.  Here, the prescan cancels an undesirable side effect
        !          1423: (in the medical, not computational, sense of the term) of the special
        !          1424: rule for self-referential macros.
        !          1425: 
        !          1426: There is also one case where prescan is useful.  It is possible
        !          1427: to use prescan to expand an argument and then stringify it---if you use
        !          1428: two levels of macros.  Let's add a new macro @samp{xstr} to the
        !          1429: example shown above:
        !          1430: 
        !          1431: @example
        !          1432: #define xstr(s) str(s)
        !          1433: #define str(s) #s
        !          1434: #define foo 4
        !          1435: xstr (foo)
        !          1436: @end example
        !          1437: 
        !          1438: This expands into @samp{"4"}, not @samp{"foo"}.  The reason for the
        !          1439: difference is that the argument of @samp{xstr} is expanded at prescan
        !          1440: (because @samp{xstr} does not specify stringification or concatenation of
        !          1441: the argument).  The result of prescan then forms the actual argument for
        !          1442: @samp{str}.  @samp{str} uses its argument without prescan because it
        !          1443: performs strigification; but it cannot prevent or undo the prescanning
        !          1444: already done by @samp{xstr}.
        !          1445: 
        !          1446: @node Cascaded Macros,, Argument Prescan, Macro Pitfalls
        !          1447: @subsubsection Cascaded Use of Macros
        !          1448: 
        !          1449: @cindex cascaded macros
        !          1450: @cindex macro body uses macro
        !          1451: A @dfn{cascade} of macros is when one macro's body contains a reference
        !          1452: to another macro.  This is very common practice.  For example,
        !          1453: 
        !          1454: @example
        !          1455: #define BUFSIZE 1020
        !          1456: #define TABLESIZE BUFSIZE
        !          1457: @end example
        !          1458: 
        !          1459: This is not at all the same as defining @samp{TABLESIZE} to be @samp{1020}.
        !          1460: The @samp{#define} for @samp{TABLESIZE} uses exactly the body you
        !          1461: specify---in this case, @samp{BUFSIZE}---and does not check to see whether
        !          1462: it too is the name of a macro.
        !          1463: 
        !          1464: It's only when you @emph{use} @samp{TABLESIZE} that the result of its expansion
        !          1465: is checked for more macro names.
        !          1466: 
        !          1467: This makes a difference if you change the definition of @samp{BUFSIZE}
        !          1468: at some point in the source file.  @samp{TABLESIZE}, defined as shown,
        !          1469: will always expand using the definition of @samp{BUFSIZE} that is
        !          1470: currently in effect:
        !          1471: 
        !          1472: @example
        !          1473: #define BUFSIZE 1020
        !          1474: #define TABLESIZE BUFSIZE
        !          1475: #undef BUFSIZE
        !          1476: #define BUFSIZE 37
        !          1477: @end example
        !          1478: 
        !          1479: @noindent
        !          1480: Now @samp{TABLESIZE} expands (in two stages) to @samp{37}.
        !          1481: 
        !          1482: @node Conditionals, Combining Sources, Macros, Top
        !          1483: @section Conditionals
        !          1484: 
        !          1485: @cindex conditionals
        !          1486: In a macro processor, a @dfn{conditional} is a command that allows a part
        !          1487: of the program to be ignored during compilation, on some conditions.
        !          1488: In the C preprocessor, a conditional can test either an arithmetic expression
        !          1489: or whether a name is defined as a macro.
        !          1490: 
        !          1491: A conditional in the C preprocessor resembles in some ways an @samp{if}
        !          1492: statement in C, but it is important to understand the difference between
        !          1493: them.  The condition in an @samp{if} statement is tested during the execution
        !          1494: of your program.  Its purpose is to allow your program to behave differently
        !          1495: from run to run, depending on the data it is operating on.  The condition
        !          1496: in a preprocessor conditional command is tested when your program is compiled.
        !          1497: Its purpose is to allow different code to be included in the program depending
        !          1498: on the situation at the time of compilation.
        !          1499: 
        !          1500: @menu
        !          1501: * Uses: Conditional Uses.       What conditionals are for.
        !          1502: * Syntax: Conditional Syntax.   How conditionals are written.
        !          1503: * Deletion: Deleted Code.       Making code into a comment.
        !          1504: * Macros: Conditionals-Macros.  Why conditionals are used with macros.
        !          1505: * Errors: #error Command.       Detecting inconsistent compilation parameters.
        !          1506: @end menu
        !          1507: 
        !          1508: @node Conditional Uses, Conditional Syntax, Conditionals, Conditionals
        !          1509: Generally there are three kinds of reason to use a conditional.
        !          1510: 
        !          1511: @itemize @bullet
        !          1512: @item
        !          1513: A program may need to use different code depending on the machine or
        !          1514: operating system it is to run on.  In some cases the code for one
        !          1515: operating system may be erroneous on another operating system; for
        !          1516: example, it might refer to library routines that do not exist on the
        !          1517: other system.  When this happens, it is not enough to avoid executing
        !          1518: the invalid code: merely having it in the program makes it impossible
        !          1519: to link the program and run it.  With a preprocessor conditional, the
        !          1520: offending code can be effectively excised from the program when it is
        !          1521: not valid.
        !          1522: 
        !          1523: @item
        !          1524: You may want to be able to compile the same source file into two
        !          1525: different programs.  Sometimes the difference between the programs is
        !          1526: that one makes frequent time-consuming consistency checks on its
        !          1527: intermediate data while the other does not.
        !          1528: 
        !          1529: @item
        !          1530: A conditional whose condition is always false is a good way to exclude
        !          1531: code from the program but keep it as a sort of comment for future
        !          1532: reference.
        !          1533: @end itemize
        !          1534: 
        !          1535: Most simple programs that are intended to run on only one machine will
        !          1536: not need to use preprocessor conditionals.
        !          1537: 
        !          1538: @node Conditional Syntax, Conditionals-Macros, Conditional Uses, Conditionals
        !          1539: @subsection Syntax of Conditionals
        !          1540: 
        !          1541: @findex #if
        !          1542: A conditional in the C preprocessor begins with a @dfn{conditional
        !          1543: command}: @samp{#if}, @samp{#ifdef} or @samp{#ifndef}.@xref{Conditionals},
        !          1544: for info on @samp{#ifdef} and @samp{#ifndef}; only @samp{#if} is explained
        !          1545: here.
        !          1546: 
        !          1547: @menu
        !          1548: * If: #if Command.     Basic conditionals using @samp{#if} and @samp{#endif}.
        !          1549: * Else: #else Command. Including some text if the condition fails.
        !          1550: * Elif: #elif Command. Testing several alternative possibilities.
        !          1551: @end menu
        !          1552: 
        !          1553: @node #if Command, #else Command, Conditional Syntax, Conditional Syntax
        !          1554: @subsubsection The @samp{#if} Command
        !          1555: 
        !          1556: The @samp{#if} command in its simplest form consists of
        !          1557: 
        !          1558: @example
        !          1559: #if @var{expression}
        !          1560: @var{controlled text}
        !          1561: #endif /* @var{expression} */
        !          1562: @end example
        !          1563: 
        !          1564: The comment following the @samp{#endif} is not required, but it is a good
        !          1565: practice because it helps people match the @samp{#endif} to the
        !          1566: corresponding @samp{#if}.  Such comments should always be used, except in
        !          1567: short conditionals that are not nested.  In fact, you can put anything at
        !          1568: all after the @samp{#endif} and it will be ignored by the GNU C preprocessor,
        !          1569: but only comments are acceptable in ANSI Standard C.
        !          1570: 
        !          1571: @var{expression} is a C expression of integer type, subject to stringent
        !          1572: restrictions.  It may contain
        !          1573: 
        !          1574: @itemize @bullet
        !          1575: @item
        !          1576: Integer constants, which are all regarded as @code{long} or
        !          1577: @code{unsigned long}.
        !          1578: 
        !          1579: @item
        !          1580: Character constants, which are interpreted according to the character
        !          1581: set and conventions of the machine and operating system on which the
        !          1582: preprocessor is running.  The GNU C preprocessor uses the C data type
        !          1583: @samp{char} for these character constants; therefore, whether some
        !          1584: character codes are negative is determined by the C compiler used to
        !          1585: compile the preprocessor.  If it treats @samp{char} as signed, then
        !          1586: character codes large enough to set the sign bit will be considered
        !          1587: negative; otherwise, no character code is considered negative.
        !          1588: 
        !          1589: @item
        !          1590: Arithmetic operators for addition, subtraction, multiplication,
        !          1591: division, bitwise operations, shifts, comparisons, and @samp{&&} and
        !          1592: @samp{||}.
        !          1593: 
        !          1594: @item
        !          1595: Identifiers that are not macros, which are all treated as zero(!).
        !          1596: 
        !          1597: @item
        !          1598: Macro calls.  All macro calls in the expression are expanded before
        !          1599: actual computation of the expression's value begins.
        !          1600: @end itemize
        !          1601: 
        !          1602: Note that @samp{sizeof} operators and @code{enum}-type values are not allowed.
        !          1603: @code{enum}-type values, like all other identifiers that are not taken
        !          1604: as macro calls and expanded, are treated as zero.
        !          1605: 
        !          1606: The text inside of a conditional can include preprocessor commands.  Then
        !          1607: the commands inside the conditional are obeyed only if that branch of the
        !          1608: conditional succeeds.  The text can also contain other conditional groups.
        !          1609: However, the @samp{#if}'s and @samp{#endif}'s must balance.
        !          1610: 
        !          1611: @node #else Command, #elif Command, #if Command, Conditional Syntax
        !          1612: @subsubsection The @samp{#else} Command
        !          1613: 
        !          1614: @findex #else
        !          1615: The @samp{#else} command can be added a conditional to provide alternative
        !          1616: text to be used if the condition is false.  This looks like
        !          1617: 
        !          1618: @example
        !          1619: #if @var{expression}
        !          1620: @var{text-if-true}
        !          1621: #else /* Not @var{expression} */
        !          1622: @var{text-if-false}
        !          1623: #endif /* Not @var{expression} */
        !          1624: @end example
        !          1625: 
        !          1626: If @var{expression} is nonzero, and the @var{text-if-true} is considered
        !          1627: included, then @samp{#else} acts like a failing conditional and the
        !          1628: @var{text-if-false} is ignored.  Contrariwise, if the @samp{#if}
        !          1629: conditional fails, the @var{text-if-false} is considered included.
        !          1630: 
        !          1631: @node #elif Command,, #else Command, Conditional Syntax
        !          1632: @subsubsection The @samp{#elif} Command
        !          1633: 
        !          1634: @findex #elif
        !          1635: One common case of nested conditionals is used to check for more than two
        !          1636: possible alternatives.  For example, you might have
        !          1637: 
        !          1638: @example
        !          1639: #if X == 1
        !          1640: @dots{}
        !          1641: #else /* X != 1 */
        !          1642: #if X == 2
        !          1643: @dots{}
        !          1644: #else /* X != 2 */
        !          1645: @dots{}
        !          1646: #endif /* X != 2 */
        !          1647: #endif /* X != 1 */
        !          1648: @end example
        !          1649: 
        !          1650: Another conditional command, @samp{#elif}, allows this to be abbreviated
        !          1651: as follows:
        !          1652: 
        !          1653: @example
        !          1654: #if X == 1
        !          1655: @dots{}
        !          1656: #elif X == 2
        !          1657: @dots{}
        !          1658: #else /* X != 2 and X != 1*/
        !          1659: @dots{}
        !          1660: #endif /* X != 2 and X != 1*/
        !          1661: @end example
        !          1662: 
        !          1663: @samp{#elif} stands for ``else if''.  Like @samp{#else}, it goes in the
        !          1664: middle of a @samp{#if}-@samp{#endif} pair and subdivides it; it does not
        !          1665: require a matching @samp{#endif} of its own.  Like @samp{#if}, the
        !          1666: @samp{#elif} command includes an expression to be tested.
        !          1667: 
        !          1668: The text following the @samp{#elif} is processed only if the original
        !          1669: @samp{#if}-condition failed and the @samp{#elif} condition succeeeds.  More
        !          1670: than one @samp{#elif} can go in the same @samp{#if}-@samp{#endif} group.
        !          1671: Then the text after each @samp{#elif} is processed only if the @samp{#elif}
        !          1672: condition succeeds after the original @samp{#if} and any previous
        !          1673: @samp{#elif}'s within it have failed.  @samp{#else} is equivalent to
        !          1674: @samp{#elif 1}, and @samp{#else} is allowed after any number of
        !          1675: @samp{#elif}'s, but @samp{#elif} may not follow a @samp{#else}.
        !          1676: 
        !          1677: @node Deleted Code, Conditionals-Macros, Conditional Syntax, Conditionals
        !          1678: @subsection Keeping Deleted Code for Future Reference
        !          1679: 
        !          1680: If you replace or delete a part of the program but want to keep the old
        !          1681: code around as a comment for future reference, the easy way to do this is
        !          1682: to put @samp{#if 0} before it and @samp{#endif} after it.
        !          1683: 
        !          1684: This works even if the code being turned off contains conditionals, but
        !          1685: they must be entire conditionals (balanced @samp{#if} and @samp{#endif}).
        !          1686: 
        !          1687: @node Conditionals-Macros, #error Command, Deleted Code, Conditionals
        !          1688: @subsection Conditionals and Macros
        !          1689: 
        !          1690: Conditionals are rarely useful except in connection with macros.  A
        !          1691: @samp{#if} command whose expression uses no macros is equivalent to
        !          1692: @samp{#if 1} or @samp{#if 0}; you might as well determine which one, by
        !          1693: computing the value of the expression yourself, and then simplify the
        !          1694: program.  But when the expression uses macros, its value can vary from
        !          1695: compilation to compilation.
        !          1696: 
        !          1697: For example, here is a conditional that tests the expression
        !          1698: @samp{BUFSIZE == 1020}, where @samp{BUFSIZE} must be a macro.
        !          1699: 
        !          1700: @example
        !          1701: #if BUFSIZE == 1020
        !          1702:   printf ("Large buffers!\n");
        !          1703: #endif /* BUFSIZE is large */
        !          1704: @end example
        !          1705: 
        !          1706: @findex defined
        !          1707: The special operator @samp{defined} may be used in @samp{#if} expressions
        !          1708: to test whether a certain name is defined as a macro.  Either @samp{defined
        !          1709: @var{name}} or @samp{defined (@var{name})} is an expression whose value is
        !          1710: 1 if @var{name} is defined as macro at the current point in the program,
        !          1711: and 0 otherwise.  For the @samp{defined} operator it makes no difference
        !          1712: what the definition of the macro is; all that matters is whether there is a
        !          1713: definition.  Thus, for example,@refill
        !          1714: 
        !          1715: @example
        !          1716: #if defined (vax) || defined (ns16000)
        !          1717: @end example
        !          1718: 
        !          1719: @noindent
        !          1720: would include the following code if either of the names @samp{vax} and
        !          1721: @samp{ns16000} is defined as a macro.
        !          1722: 
        !          1723: If a macro is defined and later undefined with @samp{#undef},
        !          1724: subsequent use of the @samp{defined} operator will return 0, because
        !          1725: the name is no longer defined.  If the macro is defined again with
        !          1726: another @samp{#define}, @samp{defined} will recommence returning 1.
        !          1727: 
        !          1728: @findex #ifdef
        !          1729: @findex #ifndef
        !          1730: Conditionals that test just the definedness of one name are very common, so
        !          1731: there are two special short conditional commands for this case.  They are
        !          1732: 
        !          1733: @table @code
        !          1734: @item #ifdef @var{name}
        !          1735: is equivalent to @samp{#if defined (@var{name})}.
        !          1736: 
        !          1737: @item #ifndef @var{name}
        !          1738: is equivalent to @samp{#if ! defined (@var{name})}.
        !          1739: @end table
        !          1740: 
        !          1741: Macro definitions can vary between compilations for several reasons.
        !          1742: 
        !          1743: @itemize @bullet
        !          1744: @item
        !          1745: Some macros are predefined on each kind of machine.  For example, on a
        !          1746: Vax, the name @samp{vax} is a predefined macro.  On other machines, it
        !          1747: would not be defined.
        !          1748: 
        !          1749: @item
        !          1750: Many more macros are defined by system header files.  Different
        !          1751: systems and machines define different macros, or give them different
        !          1752: values.  It is useful to test these macros with conditionals to avoid
        !          1753: using a system feature on a machine where it is not implemented.
        !          1754: 
        !          1755: @item
        !          1756: Macros are a common way of allowing users to customize a program for
        !          1757: different machines or applications.  For example, the macro
        !          1758: @samp{BUFSIZE} might be defined in a configuration file for your
        !          1759: program that is included as a header file in each source file.  You
        !          1760: would use @samp{BUFSIZE} in a preprocessor conditional in order to
        !          1761: generate different code depending on the chosen configuration.
        !          1762: 
        !          1763: @item
        !          1764: Macros can be defined or undefined with @samp{-D} and @samp{-U}
        !          1765: command options when you compile the program.  You can arrange to
        !          1766: compile the same source file into two different programs by choosing
        !          1767: a macro name to specify which program you want, writing conditionals
        !          1768: to test whether or how this macro is defined, and then controlling
        !          1769: the state of the macro with compiler command options.
        !          1770: @xref{Invocation}.
        !          1771: @end itemize
        !          1772: 
        !          1773: @node #error Command,, Conditionals-Macros, Conditionals
        !          1774: @subsection The @samp{#error} Command
        !          1775: 
        !          1776: @findex #error
        !          1777: The command @samp{#error} causes the preprocessor to report a fatal
        !          1778: error.  The rest of the line that follows @samp{#error} is used as the
        !          1779: error message.
        !          1780: 
        !          1781: You would use @samp{#error} inside of a conditional that detects a
        !          1782: combination of parameters which you know the program does not properly
        !          1783: support.  For example, if you know that the program will not run
        !          1784: properly on a Vax, you might write
        !          1785: 
        !          1786: @example
        !          1787: #ifdef vax
        !          1788: #error Won't work on Vaxen.  See comments at get_last_object.
        !          1789: #endif
        !          1790: @end example
        !          1791: 
        !          1792: @noindent
        !          1793: @xref{Nonstandard Predefined}, for why this works.
        !          1794: 
        !          1795: If you have several configuration parameters that must be set up by
        !          1796: the installation in a consistent way, you can use conditionals to detect
        !          1797: an inconsistency and report it with @samp{#error}.  For example,
        !          1798: 
        !          1799: @example
        !          1800: #if HASH_TABLE_SIZE % 2 == 0 || HASH_TABLE_SIZE % 3 == 0 \
        !          1801:     || HASH_TABLE_SIZE % 5 == 0
        !          1802: #error HASH_TABLE_SIZE should not be divisible by a small prime
        !          1803: #endif
        !          1804: @end example
        !          1805: 
        !          1806: @node Combining Sources, Other Commands, Conditionals, Top
        !          1807: @section Combining Source Files
        !          1808: 
        !          1809: @cindex line control
        !          1810: @findex #line
        !          1811: One of the jobs of the C preprocessor is to inform the C compiler of where
        !          1812: each line of C code came from: which source file and which line number.
        !          1813: 
        !          1814: C code can come from multiple source files if you use @samp{#include};
        !          1815: both @samp{#include} and the use of conditionals and macros can cause
        !          1816: the line number of a line in the preprocessor output to be different
        !          1817: from the line's number in the original source file.  You will appreciate
        !          1818: the value of making both the C compiler (in error messages) and symbolic
        !          1819: debuggers such as GDB use the line numbers in your source file.
        !          1820: 
        !          1821: The C preprocessor builds on this feature by offering a command by which
        !          1822: you can control the feature explicitly.  This is useful when a file for
        !          1823: input to the C preprocessor is the output from another program such as the
        !          1824: @code{bison} parser generator, which operates on another file that is the
        !          1825: true source file.  Parts of the output from @code{bison} are generated from
        !          1826: scratch, other parts come from a standard parser file.  The rest are copied
        !          1827: nearly verbatim from the source file, but their line numbers in the
        !          1828: @code{bison} output are not the same as their original line numbers.
        !          1829: Naturally you would like compiler error messages and symbolic debuggers to
        !          1830: know the original source file and line number of each line in the
        !          1831: @code{bison} output.
        !          1832: 
        !          1833: @code{bison} arranges this by writing @samp{#line} commands into the output
        !          1834: file.  @samp{#line} is a command that specifies the original line number
        !          1835: and source file name for subsequent input in the current preprocessor input
        !          1836: file.  @samp{#line} has three variants:
        !          1837: 
        !          1838: @table @code
        !          1839: @item #line @var{linenum}
        !          1840: Here @var{linenum} is a decimal integer constant.  This specifies that
        !          1841: the line number of the following line of input, in its original source file,
        !          1842: was @var{linenum}.
        !          1843: 
        !          1844: @item #line @var{linenum} @var{filename}
        !          1845: Here @var{linenum} is a decimal integer constant and @var{filename}
        !          1846: is a string constant.  This specifies that the following line of input
        !          1847: came originally from source file @var{filename} and its line number there
        !          1848: was @var{linenum}.  Keep in mind that @var{filename} is not just a
        !          1849: file name; it is surrounded by doublequote characters so that it looks
        !          1850: like a string constant.
        !          1851: 
        !          1852: @item #line @var{anything else}
        !          1853: @var{anything else} is checked for macro calls, which are expanded.
        !          1854: The result should be a decimal integer constant followed optionally
        !          1855: by a string constant, as described above.
        !          1856: @end table
        !          1857: 
        !          1858: @samp{#line} commands alter the results of the @samp{__FILE__} and
        !          1859: @samp{__LINE__} predefined macros from that point on.  @xref{Standard
        !          1860: Predefined}.
        !          1861: 
        !          1862: @node Other Commands, Output, Combining Sources, Top
        !          1863: @section Miscellaneous Preprocessor Commands
        !          1864: 
        !          1865: @findex #pragma
        !          1866: @cindex null command
        !          1867: This section describes two additional preprocesor commands.  They are
        !          1868: not very useful, but are mentioned for completeness.
        !          1869: 
        !          1870: The @dfn{null command} consists of a @samp{#} followed by a Newline, with
        !          1871: only whitespace (including comments) in between.  A null command is
        !          1872: understood as a preprocessor command but has no effect on the preprocessor
        !          1873: output.  The primary significance of the existence of the null command is
        !          1874: that an input line consisting of just a @samp{#} will produce no output,
        !          1875: rather than a line of output containing just a @samp{#}.  Supposedly
        !          1876: some old C programs contain such lines.
        !          1877: 
        !          1878: The @samp{#pragma} command is specified in the ANSI standard to have an
        !          1879: arbitrary implementation-defined effect.  In the GNU C preprocessor,
        !          1880: @samp{#pragma} first attempts to run the game @code{rogue}; if that fails,
        !          1881: it tries to run the game @code{hack}; if that fails, it tries to run
        !          1882: GNU Emacs displaying the Tower of Hanoi; if that fails, it reports a
        !          1883: fatal error.  In any case, preprocessing does not continue.
        !          1884: 
        !          1885: @node Output, Invocation, Other Commands, Top
        !          1886: @section C Preprocessor Output
        !          1887: 
        !          1888: @cindex output format
        !          1889: The output from the C preprocessor looks much like the input, except
        !          1890: that all preprocessor command lines have been replaced with blank lines
        !          1891: and all comments with spaces.  Whitespace within a line is not altered;
        !          1892: however, a space is inserted after the expansions of most macro calls.
        !          1893: 
        !          1894: Source file name and line number information is conveyed by lines of
        !          1895: the form
        !          1896: 
        !          1897: @example
        !          1898: # @var{linenum} @var{filename}
        !          1899: @end example
        !          1900: 
        !          1901: @noindent
        !          1902: which are inserted as needed into the middle of the input (but never within
        !          1903: a string or character constant).  Such a line means that the following line
        !          1904: originated in file @var{filename} at line @var{linenum}.
        !          1905: 
        !          1906: @node Invocation,, Output, Top
        !          1907: @section Invoking the C Preprocessor
        !          1908: 
        !          1909: Most often when you use the C preprocessor you will not have to invoke it
        !          1910: explicitly: the C compiler will do so automatically.  However, the
        !          1911: preprocessor is sometimes useful individually.
        !          1912: 
        !          1913: The C preprocessor expects two file names as arguments, @var{infile} and
        !          1914: @var{outfile}.  The preprocessor reads @var{infile} together with any other
        !          1915: files it specifies with @samp{#include}.  All the output generated by the
        !          1916: combined input files is written in @var{outfile}.
        !          1917: 
        !          1918: Either @var{infile} or @var{outfile} may be @samp{-}, which as @var{infile}
        !          1919: means to read from standard input and as @var{outfile} means to write to
        !          1920: standard output.  Also, if @var{outfile} or both file names are omitted,
        !          1921: the standard output and standard input are used for the omitted file names.
        !          1922: 
        !          1923: @cindex options
        !          1924: Here is a table of command options accepted by the C preprocessor.  Most
        !          1925: of them can also be given when compiling a C program; they are passed along
        !          1926: automatically to the preprocessor when it is invoked by the compiler.
        !          1927: 
        !          1928: @table @samp
        !          1929: @item -P
        !          1930: @findex -P
        !          1931: Inhibit generation of @samp{#}-lines with line-number information in
        !          1932: the output from the preprocessor (@pxref{Output}).  This might be
        !          1933: useful when running the preprocessor on something that is not C code
        !          1934: and will be sent to a program which might be confused by the
        !          1935: @samp{#}-lines
        !          1936: 
        !          1937: @item -C
        !          1938: @findex -C
        !          1939: Do not discard comments: pass them through to the output file.
        !          1940: Comments appearing in arguments of a macro call will be copied to the
        !          1941: output before the expansion of the macro call.
        !          1942: 
        !          1943: @item -T
        !          1944: @findex -T
        !          1945: Process ANSI standard trigraph sequences.  These are three-character
        !          1946: sequences, all starting with @samp{??}, that are defined by ANSI C to
        !          1947: stand for single characters.  For example, @samp{??/} stands for
        !          1948: @samp{\}, so @samp{'??/n'} is a character constant for Newline.
        !          1949: Strictly speaking, the GNU C preprocessor does not support all
        !          1950: programs in ANSI Standard C unless @samp{-T} is used, but if you
        !          1951: ever notice the difference it will be with relief.
        !          1952: 
        !          1953: You don't want to know any more about trigraphs.
        !          1954: 
        !          1955: @item -pedantic
        !          1956: @findex -pedantic
        !          1957: Issue warnings required by the ANSI C standard in certain cases such
        !          1958: as when text other than a comment follows @samp{#else} or @samp{#endif}.
        !          1959: 
        !          1960: @item -I @var{directory}
        !          1961: @findex -I
        !          1962: Add the directory @var{directory} to the end of the list of
        !          1963: directories to be searched for header files (@pxref{Include Syntax}).
        !          1964: This can be used to override a system header file, substituting your
        !          1965: own version, since these directories are searched before the system
        !          1966: header file directories.  If you use more than one @samp{-I} option,
        !          1967: the directories are scanned in left-to-right order; the standard
        !          1968: system directories come after.
        !          1969: 
        !          1970: @item -I-
        !          1971: Any directories specified with @samp{-I} options before the @samp{-I-}
        !          1972: option are searched only for the case of @samp{#include "@var{file}"};
        !          1973: they are not searched for @samp{#include <@var{file}>}.
        !          1974: 
        !          1975: If additional directories are specified with @samp{-I} options after
        !          1976: the @samp{-I-}, these directories are searched for all @samp{#include}
        !          1977: directives.
        !          1978: 
        !          1979: In addition, the @samp{-I-} option inhibits the use of the current
        !          1980: directory as the first search directory for @samp{#include "@var{file}"}.
        !          1981: Therefore, the current directory is searched only if it is requested
        !          1982: explicitly with @samp{-I.}.  Specifying both @samp{-I-} and @samp{-I.}
        !          1983: allows you to control precisely which directories are searched before
        !          1984: the current one and which are searched after.
        !          1985: 
        !          1986: @item -nostdinc
        !          1987: Do not search the standard system directories for header files.
        !          1988: Only the directories you have specified with @samp{-I} options
        !          1989: (and the current directory, if appropriate) are searched.
        !          1990: 
        !          1991: @item -D @var{name}
        !          1992: @findex -D
        !          1993: Predefine @var{name} as a macro, with definition @samp{1}.
        !          1994: 
        !          1995: @item -D @var{name}=@var{definition}
        !          1996: Predefine @var{name} as a macro, with definition @var{definition}.
        !          1997: There are no restrictions on the contents of @var{definition}, but if
        !          1998: you are invoking the preprocessor from a shell or shell-like program
        !          1999: you may need to use the shell's quoting syntax to protect characters
        !          2000: such as spaces that have a meaning in the shell syntax.
        !          2001: 
        !          2002: @item -U @var{name}
        !          2003: @findex -U
        !          2004: Do not predefine @var{name}.  If both @samp{-U} and @samp{-D} are
        !          2005: specified for one name, the @samp{-U} beats the @samp{-D} and the name
        !          2006: is not predefined.
        !          2007: 
        !          2008: @item -undef
        !          2009: @findex -undef
        !          2010: Do not predefine any nonstandard macros.
        !          2011: 
        !          2012: @item -d
        !          2013: @findex -d
        !          2014: Instead of outputting the result of preprocessing, output a list of
        !          2015: @samp{#define} commands for all the macros defined during the
        !          2016: execution of the preprocessor.
        !          2017: 
        !          2018: @item -M
        !          2019: @findex -M
        !          2020: Instead of outputting the result of preprocessing, output a rule
        !          2021: suitable for @code{make} describing the dependencies of the main
        !          2022: source file.  The preprocessor outputs one @code{make} rule containing
        !          2023: the object file name for that source file, a colon, and the names of
        !          2024: all the included files.  If there are many included files then the
        !          2025: rule is split into several lines using @samp{\}-newline.
        !          2026: 
        !          2027: This feature is used in automatic updating of makefiles.
        !          2028: 
        !          2029: @item -MM
        !          2030: @findex -MM
        !          2031: Like @samp{-M} but mention only the files included with @samp{#include
        !          2032: "@var{file}"}.  System header files included with @samp{#include
        !          2033: <@var{file}>} are omitted.
        !          2034: 
        !          2035: @item -i @var{file}
        !          2036: @findex -i
        !          2037: Process @var{file} as input, discarding the resulting output, before
        !          2038: processing the regular input file.  Because the output generated from
        !          2039: @var{file} is discarded, the only effect of @samp{-i @var{file}} is to
        !          2040: make the macros defined in @var{file} available for use in the main
        !          2041: input.
        !          2042: @end table
        !          2043: 
        !          2044: @node Concept Index, Index, Invocation, Top
        !          2045: @unnumbered Concept Index
        !          2046: @printindex cp
        !          2047: 
        !          2048: @node Index,, Concept Index, Top
        !          2049: @unnumbered Index of Commands, Macros and Options
        !          2050: @printindex fn
        !          2051: 
        !          2052: @contents
        !          2053: @bye

unix.superglobalmegacorp.com

This archive runs on limited infrastructure. Preserving old code on modern bandwidth. Automated agents are requested to crawl responsibly.