- determine dependencies

[ option... ] filename

-Help

-VERSion

The program is used to traverse source files looking for include dependencies suitable for CW[collect]ion or CW#include-cooked-ing by cook.

The filename ``-'' is understood to mean the standard input. When you use this file name, caching is ignored.

Several input languages are supported, see the options list for details.

The following options are understood.

-C

The source file is a C source file. It is assumed that it will have the dependencies resolved by the cpp(1) command. The same include semantics as the cpp(1) command will be employed. This is the default. This is short-hand for ``--language=c''

--Language=name

This option may be used to specify the language of the source file. Know names include ``C'', ``M4'', ``optimistic'' and ``roff''.

The ``optimistic'' language will take on almost anything. It accepts an CW]include keyword in any case, including mixed, with leading white space, but at most one leading punctuation character. It assumes that the filename follows the include keyword and does not contain white space, and does not start or end with punctuation characters (it strips off any it may find). The rest of the line is ignored. The drawback is that it will sometimes recognise commands and other text as unintended include directives, hence the name. This is often used to recognise include directives in a wide variety of assembler input.

-Roff

The source file is a *roff source file. It is assumed that it will have the dependencies resolved by the roffpp(1) command. The same include semantics as the roffpp(1) command will be employed. This is short-hand for ``--language=roff''

-Verbose

Tell what is happening.

-Ipath

Specify include path, a la cc(1).

-I-

Any directories you specify with -I options before the -I- option are searched only for the case of #include "file"; they are not searched for #include <file>.

If additional directories are specified with -I options after the -I-, these directories are searched for all #include directives. (Ordinarily all -I directories are used this way.)

In addition, the -I- option inhibits the use of the current directory (where the current input file came from) as the first search directory for #include "file". There is no way to override this effect of -I-. With -I. you can specify searching the directory which was current when was invoked. That is not exactly the same as what the preprocessor does by default, but it is often satisfactory.

The -I- option does not inhibit the use of the standard system directories for header files. Thus, -I- and -No_System are independent.

-Absolute_Paths

This option may be used to allow absolute paths in the output. This is usually the default.

-No_Absolute_Paths

This option may be used to exclude absolute paths from the output.

-Absent_Local_Ignore

For files included using a #include ''filename.h'' directive, ignore the file if it cannot be found.

-Absent_Local_Mention

For files included using a #include ''filename.h'' directive, print the file name even if the file cannot be found. This is the default (it probably needs to be built).

-Absent_Local_Error

For files included using a #include ''filename.h'' directive, print a fatal error if the file cannot be found.

-Absent_System_Ignore

For files included with a #include <filename.h> directive, ignore the file if it cannot be found. This is the default (it was probably ifdef'ed out).

-Absent_System_Mention

For files included with a #include <filename.h> directive, print the file name even if the file cannot be found.

-Absent_System_Error

For files included with a #include <filename.h> directive, print a fatal error if the file cannot be found.

-Absent_Program_Ignore

If the file named on the command line cannot be found, behave as if the file were found, but was empty.

-Absent_Program_Error

If the file named on the command line cannot be found, print a fatal error message. This is the default.

-Escape_Newlines

This option may be used to request that newlines in the output are escaped with backslash (``\'') characters.

-Help

Give information on how to use .

-EXclude filename

This option may be used to nominate include file names which are not to be used.

-VERSion

Tell what version of is being run.

-Interior_Files filename...

This option may be used to tell c_incl about include files which don't exist yet. This is because they are interior to the dependency graph, but cook(1) hasn't finished walking it yet. Often used with Cook's CW[interior-files] function. (Note: the filename list has an arbitrary number of files; it ends at the next option or end-of-line, so you need to be careful where you put the input filename.)

-No_System

Do not search the /usr/include directory. By default this is searched last. This option implies the -No_Absolute_Paths option, unless explicitly contradicted.

-CAche

This option may be used to turn caching on. This is the default.

-No_Cache

This option may be used to turn caching off.

-PREfix string

This option may be used to print a string before any of the filenames are printed. It will not be printed if no file names are printed.

-Quote_FileNames

This option may be used to have c_incl quote filenames. This permits filenames to contain characters which are special to Cook, including spaces.

-SUFfix string

This option may be used to print a string after all of the filenames are printed. It will not be printed if no file names are printed.

-Output filename

This option may be used to specify the output file. Defaults to the standard output if not set.

-No_Source_Relative_Includes

This option will give a fatal error if a #include ''filename.h'' directive is used. This is necessary when you are using Cook's CWsearch_list functionality to stitch together a baseline and a private work area.

-RECursion

This option may be used to specify that nested include files are to be scanned, so that their includes may also be discovered. This is the default.

-No_RECursion

This option may be use to specify that nested include files are not to be scanned. This option is recommended for use with the Cook CWcascade-for recipes. This option implies -No_Cache, unless a -Cache option is specified.

-Remove_Leading_Path path

This option may be used to remove path prefixes from the included filenames. May be used more than once. This is necessary when you are using Cook's CWsearch_list functionality to stitch together a baseline and a private work area; usually as ``CW[prepost "-rlp=" "" [search_list]]''

-STripdot

This option may be used to specify that leading redundant dot directories are to be removed from paths before processing. This is the default.

-No_STripdot

This option may be used to specify that leading redundant dot directories need not be removed from paths before processing. (Some path flattening may still occur.)

-Substitute_Leading_Path from to

This option may be used to modify path prefixes from the included filenames. May be used more than once. This is necessary when you are performing heterogeneous builds in the same directory tree. By using an ``arch'' variable to hold the architecture, and placing each architecture's objects in a separate directory tree, this option may be used as ``CW-slp [arch] "'[arch]'"'' (The outer quotes protect from Cook, the inner quotes protect from the shell.) If you need more intricate editing, used sed(1).

Any other options will generate an error.

All options may be abbreviated; the abbreviation is documented as the upper case letters, all lower case letters and underscores (_) are optional. You must use consecutive sequences of optional letters.

All options are case insensitive, you may type them in upper case or lower case or a combination of both, case is not important.

For example: the arguments "-help", "-HEL" and "-h" are all interpreted to mean the -Help option. The argument "-hlp" will not be understood, because consecutive optional characters were not supplied.

Options and other command line arguments may be mixed arbitrarily on the command line.

The GNU long option names are understood. Since all option names for are long, this means ignoring the extra leading '-'. The "--option=value" convention is also understood.

The caching mechanism use by the program caches the results of searching files for include files (in a file called .rc in the current directory). The cache is only refreshed when a file changes.

The use of this cache has been shown to dramatically increase the performance of the program. Typically, only a small proportions files in a project change between builds, resulting in a very high cache hit rate.

When using caching, always use the same command line options, otherwise weird and wonderful things will happen.

The .rc file is a binary file. If you wish to rebuild the cache, simply delete this file with the rm(1) command. Being a binary file, the .rc file is not portable across machines or operating systems, so you will need to delete it when you move your sources. It is a binary file for performance.

Accesses to the .rc file use file locking, so recipies using need not use the CWsingle-thread clause.

The command will exit with a status of 1 on any error. The command will only exit with a status of 0 if there are no errors.

version

Copyright Peter Miller; All rights reserved.

The program comes with ABSOLUTELY NO WARRANTY; for details use the ' -VERSion License' command. This is free software and you are welcome to redistribute it under certain conditions; for details use the ' -VERSion License' command.

tab(;); l r l. Peter Miller;E-Mail:;millerp@canb.auug.org.au /\/\*;WWW:;http://www.canb.auug.org.au/~millerp/