6 my $arg_string = undef;
8 # give Macs a chance to provide command line parameters
11 MacPerl::Ask( 'Please enter @ARGV (-h for help)',
12 defined $ARGV[0] ? "\"$ARGV[0]\"" : "" );
15 Perl::Tidy::perltidy(argv => $arg_string);
21 perltidy - a perl script indenter and reformatter
25 perltidy [ options ] file1 file2 file3 ...
26 (output goes to file1.tdy, file2.tdy, file3.tdy, ...)
27 perltidy [ options ] file1 -o outfile
28 perltidy [ options ] file1 -st >outfile
29 perltidy [ options ] <infile >outfile
33 Perltidy reads a perl script and writes an indented, reformatted script.
35 Many users will find enough information in L<"EXAMPLES"> to get
36 started. New users may benefit from the short tutorial
38 http://perltidy.sourceforge.net/tutorial.html
40 A convenient aid to systematically defining a set of style parameters
42 http://perltidy.sourceforge.net/stylekey.html
44 Perltidy can produce output on either of two modes, depending on the
45 existence of an B<-html> flag. Without this flag, the output is passed
46 through a formatter. The default formatting tries to follow the
47 recommendations in perlstyle(1), but it can be controlled in detail with
48 numerous input parameters, which are described in L<"FORMATTING
51 When the B<-html> flag is given, the output is passed through an HTML
52 formatter which is described in L<"HTML OPTIONS">.
58 This will produce a file F<somefile.pl.tdy> containing the script reformatted
59 using the default options, which approximate the style suggested in
60 perlstyle(1). Perltidy never changes the input file.
64 Execute perltidy on all F<.pl> files in the current directory with the
65 default options. The output will be in files with an appended F<.tdy>
66 extension. For any file with an error, there will be a file with extension
69 perltidy -b file1.pl file2.pl
71 Modify F<file1.pl> and F<file2.pl> in place, and backup the originals to
72 F<file1.pl.bak> and F<file2.pl.bak>. If F<file1.pl.bak> and/or F<file2.pl.bak>
73 already exist, they will be overwritten.
75 perltidy -gnu somefile.pl
77 Execute perltidy on file F<somefile.pl> with a style which approximates the
78 GNU Coding Standards for C programs. The output will be F<somefile.pl.tdy>.
80 perltidy -i=3 somefile.pl
82 Execute perltidy on file F<somefile.pl>, with 3 columns for each level of
83 indentation (B<-i=3>) instead of the default 4 columns. There will not be any
84 tabs in the reformatted script, except for any which already exist in comments,
85 pod documents, quotes, and here documents. Output will be F<somefile.pl.tdy>.
87 perltidy -i=3 -et=8 somefile.pl
89 Same as the previous example, except that leading whitespace will
90 be entabbed with one tab character per 8 spaces.
92 perltidy -ce -l=72 somefile.pl
94 Execute perltidy on file F<somefile.pl> with all defaults except use "cuddled
95 elses" (B<-ce>) and a maximum line length of 72 columns (B<-l=72>) instead of
96 the default 80 columns.
98 perltidy -g somefile.pl
100 Execute perltidy on file F<somefile.pl> and save a log file F<somefile.pl.LOG>
101 which shows the nesting of braces, parentheses, and square brackets at
102 the start of every line.
104 perltidy -html somefile.pl
106 This will produce a file F<somefile.pl.html> containing the script with
107 html markup. The output file will contain an embedded style sheet in
108 the <HEAD> section which may be edited to change the appearance.
110 perltidy -html -css=mystyle.css somefile.pl
112 This will produce a file F<somefile.pl.html> containing the script with
113 html markup. This output file will contain a link to a separate style
114 sheet file F<mystyle.css>. If the file F<mystyle.css> does not exist,
115 it will be created. If it exists, it will not be overwritten.
117 perltidy -html -pre somefile.pl
119 Write an html snippet with only the PRE section to F<somefile.pl.html>.
120 This is useful when code snippets are being formatted for inclusion in a
121 larger web page. No style sheet will be written in this case.
123 perltidy -html -ss >mystyle.css
125 Write a style sheet to F<mystyle.css> and exit.
127 perltidy -html -frm mymodule.pm
129 Write html with a frame holding a table of contents and the source code. The
130 output files will be F<mymodule.pm.html> (the frame), F<mymodule.pm.toc.html>
131 (the table of contents), and F<mymodule.pm.src.html> (the source code).
133 =head1 OPTIONS - OVERVIEW
135 The entire command line is scanned for options, and they are processed
136 before any files are processed. As a result, it does not matter
137 whether flags are before or after any filenames. However, the relative
138 order of parameters is important, with later parameters overriding the
139 values of earlier parameters.
141 For each parameter, there is a long name and a short name. The short
142 names are convenient for keyboard input, while the long names are
143 self-documenting and therefore useful in scripts. It is customary to
144 use two leading dashes for long names, but one may be used.
146 Most parameters which serve as on/off flags can be negated with a
147 leading "n" (for the short name) or a leading "no" or "no-" (for the
148 long name). For example, the flag to outdent long quotes is is B<-olq>
149 or B<--outdent-long-quotes>. The flag to skip this is B<-nolq>
150 or B<--nooutdent-long-quotes> or B<--no-outdent-long-quotes>.
152 Options may not be bundled together. In other words, options B<-q> and
153 B<-g> may NOT be entered as B<-qg>.
155 Option names may be terminated early as long as they are uniquely identified.
156 For example, instead of B<--dump-token-types>, it would be sufficient to enter
157 B<--dump-tok>, or even B<--dump-t>, to uniquely identify this command.
161 The following parameters concern the files which are read and written.
165 =item B<-h>, B<--help>
167 Show summary of usage and exit.
169 =item B<-o>=filename, B<--outfile>=filename
171 Name of the output file (only if a single input file is being
172 processed). If no output file is specified, and output is not
173 redirected to the standard output (see B<-st>), the output will go to
174 F<filename.tdy>. [Note: - does not redirect to standard output. Use
177 =item B<-st>, B<--standard-output>
179 Perltidy must be able to operate on an arbitrarily large number of files
180 in a single run, with each output being directed to a different output
181 file. Obviously this would conflict with outputting to the single
182 standard output device, so a special flag, B<-st>, is required to
183 request outputting to the standard output. For example,
185 perltidy somefile.pl -st >somefile.new.pl
187 This option may only be used if there is just a single input file.
188 The default is B<-nst> or B<--nostandard-output>.
190 =item B<-se>, B<--standard-error-output>
192 If perltidy detects an error when processing file F<somefile.pl>, its
193 default behavior is to write error messages to file F<somefile.pl.ERR>.
194 Use B<-se> to cause all error messages to be sent to the standard error
195 output stream instead. This directive may be negated with B<-nse>.
196 Thus, you may place B<-se> in a F<.perltidyrc> and override it when
197 desired with B<-nse> on the command line.
199 =item B<-oext>=ext, B<--output-file-extension>=ext
201 Change the extension of the output file to be F<ext> instead of the
202 default F<tdy> (or F<html> in case the -B<-html> option is used).
203 See L<Specifying File Extensions>.
205 =item B<-opath>=path, B<--output-path>=path
207 When perltidy creates a filename for an output file, by default it merely
208 appends an extension to the path and basename of the input file. This
209 parameter causes the path to be changed to F<path> instead.
211 The path should end in a valid path separator character, but perltidy will try
212 to add one if it is missing.
216 perltidy somefile.pl -opath=/tmp/
218 will produce F</tmp/somefile.pl.tdy>. Otherwise, F<somefile.pl.tdy> will
219 appear in whatever directory contains F<somefile.pl>.
221 If the path contains spaces, it should be placed in quotes.
223 This parameter will be ignored if output is being directed to standard output,
224 or if it is being specified explicitly with the B<-o=s> parameter.
226 =item B<-b>, B<--backup-and-modify-in-place>
228 Modify the input file or files in-place and save the original with the
229 extension F<.bak>. Any existing F<.bak> file will be deleted. See next item
230 for changing the default backup extension.
232 A B<-b> flag will be ignored if input is from standard input, or
233 if the B<-html> flag is set.
235 =item B<-bext>=ext, B<--backup-file-extension>=ext
237 Change the extension of the backup file to be something other than the
238 default F<.bak>. See L<Specifying File Extensions>.
240 =item B<-w>, B<--warning-output>
242 Setting B<-w> causes any non-critical warning
243 messages to be reported as errors. These include messages
244 about possible pod problems, possibly bad starting indentation level,
245 and cautions about indirect object usage. The default, B<-nw> or
246 B<--nowarning-output>, is not to include these warnings.
248 =item B<-q>, B<--quiet>
250 Deactivate error messages and syntax checking (for running under
253 For example, if you use a vi-style editor, such as vim, you may execute
254 perltidy as a filter from within the editor using something like
258 where C<n1,n2> represents the selected text. Without the B<-q> flag,
259 any error message may mess up your screen, so be prepared to use your
262 =item B<-log>, B<--logfile>
264 Save the F<.LOG> file, which has many useful diagnostics. Perltidy always
265 creates a F<.LOG> file, but by default it is deleted unless a program bug is
266 suspected. Setting the B<-log> flag forces the log file to be saved.
268 =item B<-g=n>, B<--logfile-gap=n>
270 Set maximum interval between input code lines in the logfile. This purpose of
271 this flag is to assist in debugging nesting errors. The value of C<n> is
272 optional. If you set the flag B<-g> without the value of C<n>, it will be
273 taken to be 1, meaning that every line will be written to the log file. This
274 can be helpful if you are looking for a brace, paren, or bracket nesting error.
276 Setting B<-g> also causes the logfile to be saved, so it is not necessary to
277 also include B<-log>.
279 If no B<-g> flag is given, a value of 50 will be used, meaning that at least
280 every 50th line will be recorded in the logfile. This helps prevent
281 excessively long log files.
283 Setting a negative value of C<n> is the same as not setting B<-g> at all.
285 =item B<-npro> B<--noprofile>
287 Ignore any F<.perltidyrc> command file. Normally, perltidy looks first in
288 your current directory for a F<.perltidyrc> file of parameters. (The format
289 is described below). If it finds one, it applies those options to the
290 initial default values, and then it applies any that have been defined
291 on the command line. If no F<.perltidyrc> file is found, it looks for one
292 in your home directory.
294 If you set the B<-npro> flag, perltidy will not look for this file.
296 =item B<-pro=filename> or B<--profile=filename>
298 To simplify testing and switching .perltidyrc files, this command may be
299 used to specify a configuration file which will override the default
300 name of .perltidyrc. There must not be a space on either side of the
301 '=' sign. For example, the line
303 perltidy -pro=testcfg
305 would cause file F<testcfg> to be used instead of the
306 default F<.perltidyrc>.
308 A pathname begins with three dots, e.g. ".../.perltidyrc", indicates that
309 the file should be searched for starting in the current directory and
310 working upwards. This makes it easier to have multiple projects each with
311 their own .perltidyrc in their root directories.
313 =item B<-opt>, B<--show-options>
315 Write a list of all options used to the F<.LOG> file.
316 Please see B<--dump-options> for a simpler way to do this.
318 =item B<-f>, B<--force-read-binary>
320 Force perltidy to process binary files. To avoid producing excessive
321 error messages, perltidy skips files identified by the system as non-text.
322 However, valid perl scripts containing binary data may sometimes be identified
323 as non-text, and this flag forces perltidy to process them.
327 =head1 FORMATTING OPTIONS
335 This flag disables all formatting and causes the input to be copied unchanged
336 to the output except for possible changes in line ending characters and any
337 pre- and post-filters. This can be useful in conjunction with a hierarchical
338 set of F<.perltidyrc> files to avoid unwanted code tidying. See also
339 L<Skipping Selected Sections of Code> for a way to avoid tidying specific
342 =item B<-l=n>, B<--maximum-line-length=n>
344 The default maximum line length is n=80 characters. Perltidy will try
345 to find line break points to keep lines below this length. However, long
346 quotes and side comments may cause lines to exceed this length.
347 Setting B<-l=0> is equivalent to setting B<-l=(a large number)>.
349 =item B<-i=n>, B<--indent-columns=n>
351 Use n columns per indentation level (default n=4).
355 Using tab characters will almost certainly lead to future portability
356 and maintenance problems, so the default and recommendation is not to
357 use them. For those who prefer tabs, however, there are two different
360 Except for possibly introducing tab indentation characters, as outlined
361 below, perltidy does not introduce any tab characters into your file,
362 and it removes any tabs from the code (unless requested not to do so
363 with B<-fws>). If you have any tabs in your comments, quotes, or
364 here-documents, they will remain.
368 =item B<-et=n>, B<--entab-leading-whitespace>
370 This flag causes each B<n> initial space characters to be replaced by
371 one tab character. Note that the integer B<n> is completely independent
372 of the integer specified for indentation parameter, B<-i=n>.
374 =item B<-t>, B<--tabs>
376 This flag causes one leading tab character to be inserted for each level
377 of indentation. Certain other features are incompatible with this
378 option, and if these options are also given, then a warning message will
379 be issued and this flag will be unset. One example is the B<-lp>
384 =item B<-syn>, B<--check-syntax>
386 This flag causes perltidy to run C<perl -c -T> to check syntax of input
387 and output. (To change the flags passed to perl, see the next
388 item, B<-pscf>). The results are written to the F<.LOG> file, which
389 will be saved if an error is detected in the output script. The output
390 script is not checked if the input script has a syntax error. Perltidy
391 does its own checking, but this option employs perl to get a "second
394 If perl reports errors in the input file, they will not be reported in
395 the error output unless the B<--warning-output> flag is given.
397 The default is B<not> to do this type of syntax checking (although
398 perltidy will still do as much self-checking as possible). The reason
399 is that it causes all code in BEGIN blocks to be executed, for all
400 modules being used, and this opens the door to security issues and
401 infinite loops when running perltidy.
403 =item B<-pscf=s>, B<-perl-syntax-check-flags=s>
405 When perl is invoked to check syntax, the normal flags are C<-c -T>. In
406 addition, if the B<-x> flag is given to perltidy, then perl will also be
407 passed a B<-x> flag. It should not normally be necessary to change
408 these flags, but it can be done with the B<-pscf=s> flag. For example,
409 if the taint flag, C<-T>, is not wanted, the flag could be set to be just
412 Perltidy will pass your string to perl with the exception that it will
413 add a B<-c> and B<-x> if appropriate. The F<.LOG> file will show
414 exactly what flags were passed to perl.
416 =item B<-io>, B<--indent-only>
418 This flag is used to deactivate all formatting and line break changes
419 within non-blank lines of code.
420 When it is in effect, the only change to the script will be
421 to the indentation and blank lines.
422 And any flags controlling whitespace and newlines will be ignored. You
423 might want to use this if you are perfectly happy with your whitespace
424 and line breaks, and merely want perltidy to handle the indentation.
425 (This also speeds up perltidy by well over a factor of two, so it might be
426 useful when perltidy is merely being used to help find a brace error in
429 Setting this flag is equivalent to setting B<--freeze-newlines> and
430 B<--freeze-whitespace>.
432 If you also want to keep your existing blank lines exactly
433 as they are, you can add B<--freeze-blank-lines>.
435 =item B<-ole=s>, B<--output-line-ending=s>
437 where s=C<win>, C<dos>, C<unix>, or C<mac>. This flag tells perltidy
438 to output line endings for a specific system. Normally,
439 perltidy writes files with the line separator character of the host
440 system. The C<win> and C<dos> flags have an identical result.
442 =item B<-ple>, B<--preserve-line-endings>
444 This flag tells perltidy to write its output files with the same line
445 endings as the input file, if possible. It should work for
446 B<dos>, B<unix>, and B<mac> line endings. It will only work if perltidy
447 input comes from a filename (rather than stdin, for example). If
448 perltidy has trouble determining the input file line ending, it will
449 revert to the default behavior of using the line ending of the host system.
451 =item B<-it=n>, B<--iterations=n>
453 This flag causes perltidy to do B<n> complete iterations. The reason for this
454 flag is that code beautification is a somewhat iterative process and in some
455 cases the output from perltidy can be different if it is applied a second time.
456 For most purposes the default of B<n=1> should be satisfactory. However B<n=2>
457 can be useful when a major style change is being made, or when code is being
458 beautified on check-in to a source code control system. The run time will be
459 approximately proportional to B<n>, and it should seldom be necessary to use a
460 value greater than B<n=2>. This flag has no effect when perltidy is used to generate html.
464 =head2 Code Indentation Control
468 =item B<-ci=n>, B<--continuation-indentation=n>
470 Continuation indentation is extra indentation spaces applied when
471 a long line is broken. The default is n=2, illustrated here:
474 ( $max_index_to_go >= 0 ) ? $levels_to_go[0] : $last_output_level;
476 The same example, with n=0, is a little harder to read:
479 ( $max_index_to_go >= 0 ) ? $levels_to_go[0] : $last_output_level;
481 The value given to B<-ci> is also used by some commands when a small
482 space is required. Examples are commands for outdenting labels,
483 B<-ola>, and control keywords, B<-okw>.
485 When default values are not used, it is suggested that the value B<n>
486 given with B<-ci=n> be no more than about one-half of the number of
487 spaces assigned to a full indentation level on the B<-i=n> command.
489 =item B<-sil=n> B<--starting-indentation-level=n>
491 By default, perltidy examines the input file and tries to determine the
492 starting indentation level. While it is often zero, it may not be
493 zero for a code snippet being sent from an editing session. If the
494 default method does not work correctly, or you want to change the
495 starting level, use B<-sil=n>, to force the starting level to be n.
497 =item List indentation using B<-lp>, B<--line-up-parentheses>
499 By default, perltidy indents lists with 4 spaces, or whatever value
500 is specified with B<-i=n>. Here is a small list formatted in this way:
504 'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun',
505 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec'
508 Use the B<-lp> flag to add extra indentation to cause the data to begin
509 past the opening parentheses of a sub call or list, or opening square
510 bracket of an anonymous array, or opening curly brace of an anonymous
511 hash. With this option, the above list would become:
515 'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun',
516 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec'
519 If the available line length (see B<-l=n> ) does not permit this much
520 space, perltidy will use less. For alternate placement of the
521 closing paren, see the next section.
523 This option has no effect on code BLOCKS, such as if/then/else blocks,
524 which always use whatever is specified with B<-i=n>. Also, the
525 existence of line breaks and/or block comments between the opening and
526 closing parens may cause perltidy to temporarily revert to its default
529 Note: The B<-lp> option may not be used together with the B<-t> tabs option.
530 It may, however, be used with the B<-et=n> tab method.
532 In addition, any parameter which significantly restricts the ability of
533 perltidy to choose newlines will conflict with B<-lp> and will cause
534 B<-lp> to be deactivated. These include B<-io>, B<-fnl>, B<-nanl>, and
535 B<-ndnl>. The reason is that the B<-lp> indentation style can require
536 the careful coordination of an arbitrary number of break points in
537 hierarchical lists, and these flags may prevent that.
539 =item B<-cti=n>, B<--closing-token-indentation>
541 The B<-cti=n> flag controls the indentation of a line beginning with
542 a C<)>, C<]>, or a non-block C<}>. Such a line receives:
544 -cti = 0 no extra indentation (default)
545 -cti = 1 extra indentation such that the closing token
546 aligns with its opening token.
547 -cti = 2 one extra indentation level if the line looks like:
549 -cti = 3 one extra indentation level always
551 The flags B<-cti=1> and B<-cti=2> work well with the B<-lp> flag (previous
554 # perltidy -lp -cti=1
556 'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun',
557 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec'
560 # perltidy -lp -cti=2
562 'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun',
563 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec'
566 These flags are merely hints to the formatter and they may not always be
567 followed. In particular, if -lp is not being used, the indentation for
568 B<cti=1> is constrained to be no more than one indentation level.
570 If desired, this control can be applied independently to each of the
571 closing container token types. In fact, B<-cti=n> is merely an
572 abbreviation for B<-cpi=n -csbi=n -cbi=n>, where:
573 B<-cpi> or B<--closing-paren-indentation> controls B<)>'s,
574 B<-csbi> or B<--closing-square-bracket-indentation> controls B<]>'s,
575 B<-cbi> or B<--closing-brace-indentation> controls non-block B<}>'s.
577 =item B<-icp>, B<--indent-closing-paren>
579 The B<-icp> flag is equivalent to
580 B<-cti=2>, described in the previous section. The B<-nicp> flag is
581 equivalent B<-cti=0>. They are included for backwards compatability.
583 =item B<-icb>, B<--indent-closing-brace>
585 The B<-icb> option gives one extra level of indentation to a brace which
586 terminates a code block . For example,
595 The default is not to do this, indicated by B<-nicb>.
597 =item B<-olq>, B<--outdent-long-quotes>
599 When B<-olq> is set, lines which is a quoted string longer than the
600 value B<maximum-line-length> will have their indentation removed to make
601 them more readable. This is the default. To prevent such out-denting,
602 use B<-nolq> or B<--nooutdent-long-lines>.
604 =item B<-oll>, B<--outdent-long-lines>
606 This command is equivalent to B<--outdent-long-quotes> and
607 B<--outdent-long-comments>, and it is included for compatibility with previous
608 versions of perltidy. The negation of this also works, B<-noll> or
609 B<--nooutdent-long-lines>, and is equivalent to setting B<-nolq> and B<-nolc>.
611 =item Outdenting Labels: B<-ola>, B<--outdent-labels>
613 This command will cause labels to be outdented by 2 spaces (or whatever B<-ci>
614 has been set to), if possible. This is the default. For example:
617 LOOP: while ( $i = <FOTOS> ) {
623 Use B<-nola> to not outdent labels.
625 =item Outdenting Keywords
629 =item B<-okw>, B<--outdent-keywords>
631 The command B<-okw> will will cause certain leading control keywords to
632 be outdented by 2 spaces (or whatever B<-ci> has been set to), if
633 possible. By default, these keywords are C<redo>, C<next>, C<last>,
634 C<goto>, and C<return>. The intention is to make these control keywords
635 easier to see. To change this list of keywords being outdented, see
638 For example, using C<perltidy -okw> on the previous example gives:
641 LOOP: while ( $i = <FOTOS> ) {
647 The default is not to do this.
649 =item Specifying Outdented Keywords: B<-okwl=string>, B<--outdent-keyword-list=string>
651 This command can be used to change the keywords which are outdented with
652 the B<-okw> command. The parameter B<string> is a required list of perl
653 keywords, which should be placed in quotes if there are more than one.
654 By itself, it does not cause any outdenting to occur, so the B<-okw>
655 command is still required.
657 For example, the commands C<-okwl="next last redo goto" -okw> will cause
658 those four keywords to be outdented. It is probably simplest to place
659 any B<-okwl> command in a F<.perltidyrc> file.
665 =head2 Whitespace Control
667 Whitespace refers to the blank space between variables, operators,
668 and other code tokens.
672 =item B<-fws>, B<--freeze-whitespace>
674 This flag causes your original whitespace to remain unchanged, and
675 causes the rest of the whitespace commands in this section, the
676 Code Indentation section, and
677 the Comment Control section to be ignored.
679 =item Tightness of curly braces, parentheses, and square brackets.
681 Here the term "tightness" will mean the closeness with which
682 pairs of enclosing tokens, such as parentheses, contain the quantities
683 within. A numerical value of 0, 1, or 2 defines the tightness, with
684 0 being least tight and 2 being most tight. Spaces within containers
685 are always symmetric, so if there is a space after a C<(> then there
686 will be a space before the corresponding C<)>.
688 The B<-pt=n> or B<--paren-tightness=n> parameter controls the space within
689 parens. The example below shows the effect of the three possible
692 if ( ( my $len_tab = length( $tabstr ) ) > 0 ) { # -pt=0
693 if ( ( my $len_tab = length($tabstr) ) > 0 ) { # -pt=1 (default)
694 if ((my $len_tab = length($tabstr)) > 0) { # -pt=2
696 When n is 0, there is always a space to the right of a '(' and to the left
697 of a ')'. For n=2 there is never a space. For n=1, the default, there
698 is a space unless the quantity within the parens is a single token, such
699 as an identifier or quoted string.
701 Likewise, the parameter B<-sbt=n> or B<--square-bracket-tightness=n>
702 controls the space within square brackets, as illustrated below.
704 $width = $col[ $j + $k ] - $col[ $j ]; # -sbt=0
705 $width = $col[ $j + $k ] - $col[$j]; # -sbt=1 (default)
706 $width = $col[$j + $k] - $col[$j]; # -sbt=2
708 Curly braces which do not contain code blocks are controlled by
709 the parameter B<-bt=n> or B<--brace-tightness=n>.
711 $obj->{ $parsed_sql->{ 'table' }[0] }; # -bt=0
712 $obj->{ $parsed_sql->{'table'}[0] }; # -bt=1 (default)
713 $obj->{$parsed_sql->{'table'}[0]}; # -bt=2
715 And finally, curly braces which contain blocks of code are controlled by the
716 parameter B<-bbt=n> or B<--block-brace-tightness=n> as illustrated in the
719 %bf = map { $_ => -M $_ } grep { /\.deb$/ } dirents '.'; # -bbt=0 (default)
720 %bf = map { $_ => -M $_ } grep {/\.deb$/} dirents '.'; # -bbt=1
721 %bf = map {$_ => -M $_} grep {/\.deb$/} dirents '.'; # -bbt=2
723 =item B<-sts>, B<--space-terminal-semicolon>
725 Some programmers prefer a space before all terminal semicolons. The
726 default is for no such space, and is indicated with B<-nsts> or
727 B<--nospace-terminal-semicolon>.
730 $i = 1; # -nsts (default)
732 =item B<-sfs>, B<--space-for-semicolon>
734 Semicolons within B<for> loops may sometimes be hard to see,
735 particularly when commas are also present. This option places spaces on
736 both sides of these special semicolons, and is the default. Use
737 B<-nsfs> or B<--nospace-for-semicolon> to deactivate it.
739 for ( @a = @$ap, $u = shift @a ; @a ; $u = $v ) { # -sfs (default)
740 for ( @a = @$ap, $u = shift @a; @a; $u = $v ) { # -nsfs
742 =item B<-asc>, B<--add-semicolons>
744 Setting B<-asc> allows perltidy to add any missing optional semicolon at the end
745 of a line which is followed by a closing curly brace on the next line. This
746 is the default, and may be deactivated with B<-nasc> or B<--noadd-semicolons>.
748 =item B<-dsm>, B<--delete-semicolons>
750 Setting B<-dsm> allows perltidy to delete extra semicolons which are
751 simply empty statements. This is the default, and may be deactivated
752 with B<-ndsm> or B<--nodelete-semicolons>. (Such semicolons are not
753 deleted, however, if they would promote a side comment to a block
756 =item B<-aws>, B<--add-whitespace>
758 Setting this option allows perltidy to add certain whitespace improve
759 code readability. This is the default. If you do not want any
760 whitespace added, but are willing to have some whitespace deleted, use
761 B<-naws>. (Use B<-fws> to leave whitespace completely unchanged).
763 =item B<-dws>, B<--delete-old-whitespace>
765 Setting this option allows perltidy to remove some old whitespace
766 between characters, if necessary. This is the default. If you
767 do not want any old whitespace removed, use B<-ndws> or
768 B<--nodelete-old-whitespace>.
770 =item Detailed whitespace controls around tokens
772 For those who want more detailed control over the whitespace around
773 tokens, there are four parameters which can directly modify the default
774 whitespace rules built into perltidy for any token. They are:
776 B<-wls=s> or B<--want-left-space=s>,
778 B<-nwls=s> or B<--nowant-left-space=s>,
780 B<-wrs=s> or B<--want-right-space=s>,
782 B<-nwrs=s> or B<--nowant-right-space=s>.
784 These parameters are each followed by a quoted string, B<s>, containing a
785 list of token types. No more than one of each of these parameters
786 should be specified, because repeating a command-line parameter
787 always overwrites the previous one before perltidy ever sees it.
789 To illustrate how these are used, suppose it is desired that there be no
790 space on either side of the token types B<= + - / *>. The following two
791 parameters would specify this desire:
793 -nwls="= + - / *" -nwrs="= + - / *"
795 (Note that the token types are in quotes, and that they are separated by
796 spaces). With these modified whitespace rules, the following line of math:
798 $root = -$b + sqrt( $b * $b - 4. * $a * $c ) / ( 2. * $a );
802 $root=-$b+sqrt( $b*$b-4.*$a*$c )/( 2.*$a );
804 These parameters should be considered to be hints to perltidy rather
805 than fixed rules, because perltidy must try to resolve conflicts that
806 arise between them and all of the other rules that it uses. One
807 conflict that can arise is if, between two tokens, the left token wants
808 a space and the right one doesn't. In this case, the token not wanting
809 a space takes priority.
811 It is necessary to have a list of all token types in order to create
812 this type of input. Such a list can be obtained by the command
813 B<--dump-token-types>. Also try the B<-D> flag on a short snippet of code
814 and look at the .DEBUG file to see the tokenization.
816 B<WARNING> Be sure to put these tokens in quotes to avoid having them
817 misinterpreted by your command shell.
819 =item Space between specific keywords and opening paren
821 When an opening paren follows a Perl keyword, no space is introduced after the
822 keyword, unless it is (by default) one of these:
824 my local our and or eq ne if else elsif until unless
825 while for foreach return switch case given when
827 These defaults can be modified with two commands:
829 B<-sak=s> or B<--space-after-keyword=s> adds keywords.
831 B<-nsak=s> or B<--nospace-after-keyword=s> removes keywords.
833 where B<s> is a list of keywords (in quotes if necessary). For example,
835 my ( $a, $b, $c ) = @_; # default
836 my( $a, $b, $c ) = @_; # -nsak="my local our"
838 To put a space after all keywords, see the next item.
840 =item Space between all keywords and opening parens
842 When an opening paren follows a function or keyword, no space is introduced
843 after the keyword except for the keywords noted in the previous item. To
844 always put a space between a function or keyword and its opening paren,
847 B<-skp> or B<--space-keyword-paren>
849 You will probably also want to use the flag B<-sfp> (next item) too.
851 =item Space between all function names and opening parens
853 When an opening paren follows a function the default is not to introduce
854 a space. To cause a space to be introduced use:
856 B<-sfp> or B<--space-function-paren>
858 myfunc( $a, $b, $c ); # default
859 myfunc ( $a, $b, $c ); # -sfp
861 You will probably also want to use the flag B<-skp> (previous item) too.
863 =item Trimming whitespace around C<qw> quotes
865 B<-tqw> or B<--trim-qw> provide the default behavior of trimming
866 spaces around multi-line C<qw> quotes and indenting them appropriately.
868 B<-ntqw> or B<--notrim-qw> cause leading and trailing whitespace around
869 multi-line C<qw> quotes to be left unchanged. This option will not
870 normally be necessary, but was added for testing purposes, because in
871 some versions of perl, trimming C<qw> quotes changes the syntax tree.
875 =head2 Comment Controls
877 Perltidy has a number of ways to control the appearance of both block comments
878 and side comments. The term B<block comment> here refers to a full-line
879 comment, whereas B<side comment> will refer to a comment which appears on a
880 line to the right of some code.
884 =item B<-ibc>, B<--indent-block-comments>
886 Block comments normally look best when they are indented to the same
887 level as the code which follows them. This is the default behavior, but
888 you may use B<-nibc> to keep block comments left-justified. Here is an
891 # this comment is indented (-ibc, default)
892 if ($task) { yyy(); }
894 The alternative is B<-nibc>:
896 # this comment is not indented (-nibc)
897 if ($task) { yyy(); }
899 See also the next item, B<-isbc>, as well as B<-sbc>, for other ways to
900 have some indented and some outdented block comments.
902 =item B<-isbc>, B<--indent-spaced-block-comments>
904 If there is no leading space on the line, then the comment will not be
905 indented, and otherwise it may be.
907 If both B<-ibc> and B<-isbc> are set, then B<-isbc> takes priority.
909 =item B<-olc>, B<--outdent-long-comments>
911 When B<-olc> is set, lines which are full-line (block) comments longer
912 than the value B<maximum-line-length> will have their indentation
913 removed. This is the default; use B<-nolc> to prevent outdenting.
915 =item B<-msc=n>, B<--minimum-space-to-comment=n>
917 Side comments look best when lined up several spaces to the right of
918 code. Perltidy will try to keep comments at least n spaces to the
919 right. The default is n=4 spaces.
921 =item B<-fpsc=n>, B<--fixed-position-side-comment=n>
923 This parameter tells perltidy to line up side comments in column number B<n>
924 whenever possible. The default, n=0, is not do do this.
926 =item B<-hsc>, B<--hanging-side-comments>
928 By default, perltidy tries to identify and align "hanging side
929 comments", which are something like this:
931 my $IGNORE = 0; # This is a side comment
932 # This is a hanging side comment
935 A comment is considered to be a hanging side comment if (1) it immediately
936 follows a line with a side comment, or another hanging side comment, and
937 (2) there is some leading whitespace on the line.
938 To deactivate this feature, use B<-nhsc> or B<--nohanging-side-comments>.
939 If block comments are preceded by a blank line, or have no leading
940 whitespace, they will not be mistaken as hanging side comments.
942 =item Closing Side Comments
944 A closing side comment is a special comment which perltidy can
945 automatically create and place after the closing brace of a code block.
946 They can be useful for code maintenance and debugging. The command
947 B<-csc> (or B<--closing-side-comments>) adds or updates closing side
948 comments. For example, here is a small code snippet
951 if ( !defined( $_[0] ) ) {
952 print("Hello, World\n");
955 print( $_[0], "\n" );
959 And here is the result of processing with C<perltidy -csc>:
962 if ( !defined( $_[0] ) ) {
963 print("Hello, World\n");
966 print( $_[0], "\n" );
970 A closing side comment was added for C<sub message> in this case, but not
971 for the C<if> and C<else> blocks, because they were below the 6 line
972 cutoff limit for adding closing side comments. This limit may be
973 changed with the B<-csci> command, described below.
975 The command B<-dcsc> (or B<--delete-closing-side-comments>) reverses this
976 process and removes these comments.
978 Several commands are available to modify the behavior of these two basic
979 commands, B<-csc> and B<-dcsc>:
983 =item B<-csci=n>, or B<--closing-side-comment-interval=n>
985 where C<n> is the minimum number of lines that a block must have in
986 order for a closing side comment to be added. The default value is
987 C<n=6>. To illustrate:
989 # perltidy -csci=2 -csc
991 if ( !defined( $_[0] ) ) {
992 print("Hello, World\n");
993 } ## end if ( !defined( $_[0] ))
995 print( $_[0], "\n" );
996 } ## end else [ if ( !defined( $_[0] ))
999 Now the C<if> and C<else> blocks are commented. However, now this has
1000 become very cluttered.
1002 =item B<-cscp=string>, or B<--closing-side-comment-prefix=string>
1004 where string is the prefix used before the name of the block type. The
1005 default prefix, shown above, is C<## end>. This string will be added to
1006 closing side comments, and it will also be used to recognize them in
1007 order to update, delete, and format them. Any comment identified as a
1008 closing side comment will be placed just a single space to the right of
1011 =item B<-cscl=string>, or B<--closing-side-comment-list-string>
1013 where C<string> is a list of block types to be tagged with closing side
1014 comments. By default, all code block types preceded by a keyword or
1015 label (such as C<if>, C<sub>, and so on) will be tagged. The B<-cscl>
1016 command changes the default list to be any selected block types; see
1017 L<Specifying Block Types>.
1018 For example, the following command
1019 requests that only C<sub>'s, labels, C<BEGIN>, and C<END> blocks be
1020 affected by any B<-csc> or B<-dcsc> operation:
1022 -cscl="sub : BEGIN END"
1024 =item B<-csct=n>, or B<--closing-side-comment-maximum-text=n>
1026 The text appended to certain block types, such as an C<if> block, is
1027 whatever lies between the keyword introducing the block, such as C<if>,
1028 and the opening brace. Since this might be too much text for a side
1029 comment, there needs to be a limit, and that is the purpose of this
1030 parameter. The default value is C<n=20>, meaning that no additional
1031 tokens will be appended to this text after its length reaches 20
1032 characters. Omitted text is indicated with C<...>. (Tokens, including
1033 sub names, are never truncated, however, so actual lengths may exceed
1034 this). To illustrate, in the above example, the appended text of the
1035 first block is C< ( !defined( $_[0] )...>. The existing limit of
1036 C<n=20> caused this text to be truncated, as indicated by the C<...>. See
1037 the next flag for additional control of the abbreviated text.
1039 =item B<-cscb>, or B<--closing-side-comments-balanced>
1041 As discussed in the previous item, when the
1042 closing-side-comment-maximum-text limit is exceeded the comment text must
1043 be truncated. Older versions of perltidy terminated with three dots, and this
1044 can still be achieved with -ncscb:
1046 perltidy -csc -ncscb
1047 } ## end foreach my $foo (sort { $b cmp $a ...
1049 However this causes a problem with editors editors which cannot recognize
1050 comments or are not configured to do so because they cannot "bounce" around in
1051 the text correctly. The B<-cscb> flag has been added to
1052 help them by appending appropriate balancing structure:
1055 } ## end foreach my $foo (sort { $b cmp $a ... })
1057 The default is B<-cscb>.
1059 =item B<-csce=n>, or B<--closing-side-comment-else-flag=n>
1061 The default, B<n=0>, places the text of the opening C<if> statement after any
1064 If B<n=2> is used, then each C<elsif> is also given the text of the opening
1065 C<if> statement. Also, an C<else> will include the text of a preceding
1066 C<elsif> statement. Note that this may result some long closing
1069 If B<n=1> is used, the results will be the same as B<n=2> whenever the
1070 resulting line length is less than the maximum allowed.
1071 =item B<-cscb>, or B<--closing-side-comments-balanced>
1073 When using closing-side-comments, and the closing-side-comment-maximum-text
1074 limit is exceeded, then the comment text must be abbreviated.
1075 It is terminated with three dots if the B<-cscb> flag is negated:
1077 perltidy -csc -ncscb
1078 } ## end foreach my $foo (sort { $b cmp $a ...
1080 This causes a problem with older editors which do not recognize comments
1081 because they cannot "bounce" around in the text correctly. The B<-cscb>
1082 flag tries to help them by appending appropriate terminal balancing structures:
1085 } ## end foreach my $foo (sort { $b cmp $a ... })
1087 The default is B<-cscb>.
1090 =item B<-cscw>, or B<--closing-side-comment-warnings>
1092 This parameter is intended to help make the initial transition to the use of
1093 closing side comments.
1095 things to happen if a closing side comment replaces an existing, different
1096 closing side comment: first, an error message will be issued, and second, the
1097 original side comment will be placed alone on a new specially marked comment
1098 line for later attention.
1100 The intent is to avoid clobbering existing hand-written side comments
1101 which happen to match the pattern of closing side comments. This flag
1102 should only be needed on the first run with B<-csc>.
1106 B<Important Notes on Closing Side Comments:>
1112 Closing side comments are only placed on lines terminated with a closing
1113 brace. Certain closing styles, such as the use of cuddled elses
1114 (B<-ce>), preclude the generation of some closing side comments.
1118 Please note that adding or deleting of closing side comments takes
1119 place only through the commands B<-csc> or B<-dcsc>. The other commands,
1120 if used, merely modify the behavior of these two commands.
1124 It is recommended that the B<-cscw> flag be used along with B<-csc> on
1125 the first use of perltidy on a given file. This will prevent loss of
1126 any existing side comment data which happens to have the csc prefix.
1130 Once you use B<-csc>, you should continue to use it so that any
1131 closing side comments remain correct as code changes. Otherwise, these
1132 comments will become incorrect as the code is updated.
1136 If you edit the closing side comments generated by perltidy, you must also
1137 change the prefix to be different from the closing side comment prefix.
1138 Otherwise, your edits will be lost when you rerun perltidy with B<-csc>. For
1139 example, you could simply change C<## end> to be C<## End>, since the test is
1140 case sensitive. You may also want to use the B<-ssc> flag to keep these
1141 modified closing side comments spaced the same as actual closing side comments.
1145 Temporarily generating closing side comments is a useful technique for
1146 exploring and/or debugging a perl script, especially one written by someone
1147 else. You can always remove them with B<-dcsc>.
1151 =item Static Block Comments
1153 Static block comments are block comments with a special leading pattern,
1154 C<##> by default, which will be treated slightly differently from other
1155 block comments. They effectively behave as if they had glue along their
1156 left and top edges, because they stick to the left edge and previous line
1157 when there is no blank spaces in those places. This option is
1158 particularly useful for controlling how commented code is displayed.
1162 =item B<-sbc>, B<--static-block-comments>
1164 When B<-sbc> is used, a block comment with a special leading pattern, C<##> by
1165 default, will be treated specially.
1167 Comments so identified are treated as follows:
1173 If there is no leading space on the line, then the comment will not
1174 be indented, and otherwise it may be,
1178 no new blank line will be
1179 inserted before such a comment, and
1183 such a comment will never become
1184 a hanging side comment.
1188 For example, assuming C<@month_of_year> is
1191 @month_of_year = ( # -sbc (default)
1192 'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct',
1196 Without this convention, the above code would become
1198 @month_of_year = ( # -nsbc
1199 'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct',
1205 which is not as clear.
1206 The default is to use B<-sbc>. This may be deactivated with B<-nsbc>.
1208 =item B<-sbcp=string>, B<--static-block-comment-prefix=string>
1210 This parameter defines the prefix used to identify static block comments
1211 when the B<-sbc> parameter is set. The default prefix is C<##>,
1212 corresponding to C<-sbcp=##>. The prefix is actually part of a perl
1213 pattern used to match lines and it must either begin with C<#> or C<^#>.
1214 In the first case a prefix ^\s* will be added to match any leading
1215 whitespace, while in the second case the pattern will match only
1216 comments with no leading whitespace. For example, to
1217 identify all comments as static block comments, one would use C<-sbcp=#>.
1218 To identify all left-adjusted comments as static block comments, use C<-sbcp='^#'>.
1220 Please note that B<-sbcp> merely defines the pattern used to identify static
1221 block comments; it will not be used unless the switch B<-sbc> is set. Also,
1222 please be aware that since this string is used in a perl regular expression
1223 which identifies these comments, it must enable a valid regular expression to
1226 A pattern which can be useful is:
1230 This pattern requires a static block comment to have at least one character
1231 which is neither a # nor a space. It allows a line containing only '#'
1232 characters to be rejected as a static block comment. Such lines are often used
1233 at the start and end of header information in subroutines and should not be
1234 separated from the intervening comments, which typically begin with just a
1237 =item B<-osbc>, B<--outdent-static-block-comments>
1239 The command B<-osbc> will will cause static block comments to be outdented by 2
1240 spaces (or whatever B<-ci=n> has been set to), if possible.
1244 =item Static Side Comments
1246 Static side comments are side comments with a special leading pattern.
1247 This option can be useful for controlling how commented code is displayed
1248 when it is a side comment.
1252 =item B<-ssc>, B<--static-side-comments>
1254 When B<-ssc> is used, a side comment with a static leading pattern, which is
1255 C<##> by default, will be be spaced only a single space from previous
1256 character, and it will not be vertically aligned with other side comments.
1258 The default is B<-nssc>.
1260 =item B<-sscp=string>, B<--static-side-comment-prefix=string>
1262 This parameter defines the prefix used to identify static side comments
1263 when the B<-ssc> parameter is set. The default prefix is C<##>,
1264 corresponding to C<-sscp=##>.
1266 Please note that B<-sscp> merely defines the pattern used to identify
1267 static side comments; it will not be used unless the switch B<-ssc> is
1268 set. Also, note that this string is used in a perl regular expression
1269 which identifies these comments, so it must enable a valid regular
1270 expression to be formed.
1277 =head2 Skipping Selected Sections of Code
1279 Selected lines of code may be passed verbatim to the output without any
1280 formatting. This feature is enabled by default but can be disabled with
1281 the B<--noformat-skipping> or B<-nfs> flag. It should be used sparingly to
1282 avoid littering code with markers, but it might be helpful for working
1283 around occasional problems. For example it might be useful for keeping
1284 the indentation of old commented code unchanged, keeping indentation of
1285 long blocks of aligned comments unchanged, keeping certain list
1286 formatting unchanged, or working around a glitch in perltidy.
1290 =item B<-fs>, B<--format-skipping>
1292 This flag, which is enabled by default, causes any code between
1293 special beginning and ending comment markers to be passed to the
1294 output without formatting. The default beginning marker is #<<<
1295 and the default ending marker is #>>> but they
1296 may be changed (see next items below). Additional text may appear on
1297 these special comment lines provided that it is separated from the
1298 marker by at least one space. For example
1300 #<<< do not let perltidy touch this
1308 The comment markers may be placed at any location that a block comment may
1309 appear. If they do not appear to be working, use the -log flag and examine the
1310 F<.LOG> file. Use B<-nfs> to disable this feature.
1312 =item B<-fsb=string>, B<--format-skipping-begin=string>
1314 The B<-fsb=string> parameter may be used to change the beginning marker for
1315 format skipping. The default is equivalent to -fsb='#<<<'. The string that
1316 you enter must begin with a # and should be in quotes as necessary to get past
1317 the command shell of your system. It is actually the leading text of a pattern
1318 that is constructed by appending a '\s', so you must also include backslashes
1319 for characters to be taken literally rather than as patterns.
1321 Some examples show how example strings become patterns:
1323 -fsb='#\{\{\{' becomes /^#\{\{\{\s/ which matches #{{{ but not #{{{{
1324 -fsb='#\*\*' becomes /^#\*\*\s/ which matches #** but not #***
1325 -fsb='#\*{2,}' becomes /^#\*{2,}\s/ which matches #** and #*****
1327 =item B<-fse=string>, B<--format-skipping-end=string>
1329 The B<-fsb=string> is the corresponding parameter used to change the
1330 ending marker for format skipping. The default is equivalent to
1335 =head2 Line Break Control
1337 The parameters in this section control breaks after
1338 non-blank lines of code. Blank lines are controlled
1339 separately by parameters in the section L<Blank Line
1344 =item B<-fnl>, B<--freeze-newlines>
1346 If you do not want any changes to the line breaks within
1347 lines of code in your script, set
1348 B<-fnl>, and they will remain fixed, and the rest of the commands in
1349 this section and sections
1350 L<Controlling List Formatting>,
1351 L<Retaining or Ignoring Existing Line Breaks>.
1352 You may want to use B<-noll> with this.
1354 Note: If you also want to keep your blank lines exactly
1355 as they are, you can use the B<-fbl> flag which is described
1356 in the section L<Blank Line Control>.
1358 =item B<-ce>, B<--cuddled-else>
1360 Enable the "cuddled else" style, in which C<else> and C<elsif> are
1361 follow immediately after the curly brace closing the previous block.
1362 The default is not to use cuddled elses, and is indicated with the flag
1363 B<-nce> or B<--nocuddled-else>. Here is a comparison of the
1375 else { # -nce (default)
1379 =item B<-bl>, B<--opening-brace-on-new-line>
1381 Use the flag B<-bl> to place the opening brace on a new line:
1383 if ( $input_file eq '-' ) # -bl
1385 important_function();
1388 This flag applies to all structural blocks, including named sub's (unless
1389 the B<-sbl> flag is set -- see next item).
1391 The default style, B<-nbl>, places an opening brace on the same line as
1392 the keyword introducing it. For example,
1394 if ( $input_file eq '-' ) { # -nbl (default)
1396 =item B<-sbl>, B<--opening-sub-brace-on-new-line>
1398 The flag B<-sbl> can be used to override the value of B<-bl> for
1399 the opening braces of named sub's. For example,
1403 produces this result:
1407 if (!defined($_[0])) {
1408 print("Hello, World\n");
1415 This flag is negated with B<-nsbl>. If B<-sbl> is not specified,
1416 the value of B<-bl> is used.
1418 =item B<-asbl>, B<--opening-anonymous-sub-brace-on-new-line>
1420 The flag B<-asbl> is like the B<-sbl> flag except that it applies
1421 to anonymous sub's instead of named subs. For example
1425 produces this result:
1429 if ( !defined( $_[0] ) ) {
1430 print("Hello, World\n");
1433 print( $_[0], "\n" );
1437 This flag is negated with B<-nasbl>, and the default is B<-nasbl>.
1439 =item B<-bli>, B<--brace-left-and-indent>
1441 The flag B<-bli> is the same as B<-bl> but in addition it causes one
1442 unit of continuation indentation ( see B<-ci> ) to be placed before
1443 an opening and closing block braces.
1447 if ( $input_file eq '-' ) # -bli
1449 important_function();
1452 By default, this extra indentation occurs for blocks of type:
1453 B<if>, B<elsif>, B<else>, B<unless>, B<for>, B<foreach>, B<sub>,
1454 B<while>, B<until>, and also with a preceding label. The next item
1455 shows how to change this.
1457 =item B<-blil=s>, B<--brace-left-and-indent-list=s>
1459 Use this parameter to change the types of block braces for which the
1460 B<-bli> flag applies; see L<Specifying Block Types>. For example,
1461 B<-blil='if elsif else'> would apply it to only C<if/elsif/else> blocks.
1463 =item B<-bar>, B<--opening-brace-always-on-right>
1465 The default style, B<-nbl> places the opening code block brace on a new
1466 line if it does not fit on the same line as the opening keyword, like
1469 if ( $bigwasteofspace1 && $bigwasteofspace2
1470 || $bigwasteofspace3 && $bigwasteofspace4 )
1472 big_waste_of_time();
1475 To force the opening brace to always be on the right, use the B<-bar>
1476 flag. In this case, the above example becomes
1478 if ( $bigwasteofspace1 && $bigwasteofspace2
1479 || $bigwasteofspace3 && $bigwasteofspace4 ) {
1480 big_waste_of_time();
1483 A conflict occurs if both B<-bl> and B<-bar> are specified.
1485 =item B<-otr>, B<--opening-token-right> and related flags
1487 The B<-otr> flag is a hint that perltidy should not place a break between a
1488 comma and an opening token. For example:
1490 # default formatting
1491 push @{ $self->{$module}{$key} },
1493 accno => $ref->{accno},
1494 description => $ref->{description}
1498 push @{ $self->{$module}{$key} }, {
1499 accno => $ref->{accno},
1500 description => $ref->{description}
1503 The flag B<-otr> is actually a synonym for three other flags
1504 which can be used to control parens, hash braces, and square brackets
1505 separately if desired:
1507 -opr or --opening-paren-right
1508 -ohbr or --opening-hash-brace-right
1509 -osbr or --opening-square-bracket-right
1511 =item Vertical tightness of non-block curly braces, parentheses, and square brackets.
1513 These parameters control what shall be called vertical tightness. Here are the
1520 Opening tokens (except for block braces) are controlled by B<-vt=n>, or
1521 B<--vertical-tightness=n>, where
1523 -vt=0 always break a line after opening token (default).
1524 -vt=1 do not break unless this would produce more than one
1525 step in indentation in a line.
1526 -vt=2 never break a line after opening token
1530 You must also use the B<-lp> flag when you use the B<-vt> flag; the
1531 reason is explained below.
1535 Closing tokens (except for block braces) are controlled by B<-vtc=n>, or
1536 B<--vertical-tightness-closing=n>, where
1538 -vtc=0 always break a line before a closing token (default),
1539 -vtc=1 do not break before a closing token which is followed
1540 by a semicolon or another closing token, and is not in
1542 -vtc=2 never break before a closing token.
1544 The rules for B<-vtc=1> are designed to maintain a reasonable balance
1545 between tightness and readability in complex lists.
1549 Different controls may be applied to to different token types,
1550 and it is also possible to control block braces; see below.
1554 Finally, please note that these vertical tightness flags are merely
1555 hints to the formatter, and it cannot always follow them. Things which
1556 make it difficult or impossible include comments, blank lines, blocks of
1557 code within a list, and possibly the lack of the B<-lp> parameter.
1558 Also, these flags may be ignored for very small lists (2 or 3 lines in
1563 Here are some examples:
1565 # perltidy -lp -vt=0 -vtc=0
1573 # perltidy -lp -vt=1 -vtc=0
1574 %romanNumerals = ( one => 'I',
1580 # perltidy -lp -vt=1 -vtc=1
1581 %romanNumerals = ( one => 'I',
1586 The difference between B<-vt=1> and B<-vt=2> is shown here:
1588 # perltidy -lp -vt=1
1590 mysprintf( "(void)find_threadsv(%s);",
1591 cstring( $threadsv_names[ $op->targ ] )
1595 # perltidy -lp -vt=2
1596 $init->add( mysprintf( "(void)find_threadsv(%s);",
1597 cstring( $threadsv_names[ $op->targ ] )
1601 With B<-vt=1>, the line ending in C<add(> does not combine with the next
1602 line because the next line is not balanced. This can help with
1603 readability, but B<-vt=2> can be used to ignore this rule.
1605 The tightest, and least readable, code is produced with both C<-vt=2> and
1608 # perltidy -lp -vt=2 -vtc=2
1609 $init->add( mysprintf( "(void)find_threadsv(%s);",
1610 cstring( $threadsv_names[ $op->targ ] ) ) );
1612 Notice how the code in all of these examples collapses vertically as
1613 B<-vt> increases, but the indentation remains unchanged. This is
1614 because perltidy implements the B<-vt> parameter by first formatting as
1615 if B<-vt=0>, and then simply overwriting one output line on top of the
1616 next, if possible, to achieve the desired vertical tightness. The
1617 B<-lp> indentation style has been designed to allow this vertical
1618 collapse to occur, which is why it is required for the B<-vt> parameter.
1620 The B<-vt=n> and B<-vtc=n> parameters apply to each type of container
1621 token. If desired, vertical tightness controls can be applied
1622 independently to each of the closing container token types.
1624 The parameters for controlling parentheses are B<-pvt=n> or
1625 B<--paren-vertical-tightness=n>, and B<-pcvt=n> or
1626 B<--paren-vertical-tightness-closing=n>.
1628 Likewise, the parameters for square brackets are B<-sbvt=n> or
1629 B<--square-bracket-vertical-tightness=n>, and B<-sbcvt=n> or
1630 B<--square-bracket-vertical-tightness-closing=n>.
1632 Finally, the parameters for controlling non-code block braces are
1633 B<-bvt=n> or B<--brace-vertical-tightness=n>, and B<-bcvt=n> or
1634 B<--brace-vertical-tightness-closing=n>.
1636 In fact, the parameter B<-vt=n> is actually just an abbreviation for
1637 B<-pvt=n -bvt=n sbvt=n>, and likewise B<-vtc=n> is an abbreviation
1638 for B<-pvtc=n -bvtc=n sbvtc=n>.
1640 =item B<-bbvt=n> or B<--block-brace-vertical-tightness=n>
1642 The B<-bbvt=n> flag is just like the B<-vt=n> flag but applies
1643 to opening code block braces.
1645 -bbvt=0 break after opening block brace (default).
1646 -bbvt=1 do not break unless this would produce more than one
1647 step in indentation in a line.
1648 -bbvt=2 do not break after opening block brace.
1650 It is necessary to also use either B<-bl> or B<-bli> for this to work,
1651 because, as with other vertical tightness controls, it is implemented by
1652 simply overwriting a line ending with an opening block brace with the
1653 subsequent line. For example:
1655 # perltidy -bli -bbvt=0
1656 if ( open( FILE, "< $File" ) )
1658 while ( $File = <FILE> )
1666 # perltidy -bli -bbvt=1
1667 if ( open( FILE, "< $File" ) )
1668 { while ( $File = <FILE> )
1675 By default this applies to blocks associated with keywords B<if>,
1676 B<elsif>, B<else>, B<unless>, B<for>, B<foreach>, B<sub>, B<while>,
1677 B<until>, and also with a preceding label. This can be changed with
1678 the parameter B<-bbvtl=string>, or
1679 B<--block-brace-vertical-tightness-list=string>, where B<string> is a
1680 space-separated list of block types. For more information on the
1681 possible values of this string, see L<Specifying Block Types>
1683 For example, if we want to just apply this style to C<if>,
1684 C<elsif>, and C<else> blocks, we could use
1685 C<perltidy -bli -bbvt=1 -bbvtl='if elsif else'>.
1687 There is no vertical tightness control for closing block braces; with
1688 the exception of one-line blocks, they will normally remain on a
1691 =item B<-sot>, B<--stack-opening-tokens> and related flags
1693 The B<-sot> flag tells perltidy to "stack" opening tokens
1694 when possible to avoid lines with isolated opening tokens.
1699 $opt_c = Text::CSV_XS->new(
1708 $opt_c = Text::CSV_XS->new( {
1715 For detailed control of individual closing tokens the following
1716 controls can be used:
1718 -sop or --stack-opening-paren
1719 -sohb or --stack-opening-hash-brace
1720 -sosb or --stack-opening-square-bracket
1722 The flag B<-sot> is a synonym for B<-sop -sohb -sosb>.
1724 =item B<-sct>, B<--stack-closing-tokens> and related flags
1726 The B<-sct> flag tells perltidy to "stack" closing tokens
1727 when possible to avoid lines with isolated closing tokens.
1732 $opt_c = Text::CSV_XS->new(
1741 $opt_c = Text::CSV_XS->new(
1748 The B<-sct> flag is somewhat similar to the B<-vtc> flags, and in some
1749 cases it can give a similar result. The difference is that the B<-vtc>
1750 flags try to avoid lines with leading opening tokens by "hiding" them at
1751 the end of a previous line, whereas the B<-sct> flag merely tries to
1752 reduce the number of lines with isolated closing tokens by stacking them
1753 but does not try to hide them. For example:
1756 $opt_c = Text::CSV_XS->new(
1760 always_quote => 1, } );
1762 For detailed control of the stacking of individual closing tokens the
1763 following controls can be used:
1765 -scp or --stack-closing-paren
1766 -schb or --stack-closing-hash-brace
1767 -scsb or --stack-closing-square-bracket
1769 The flag B<-sct> is a synonym for B<-scp -schb -scsb>.
1771 =item B<-dnl>, B<--delete-old-newlines>
1773 By default, perltidy first deletes all old line break locations, and then it
1774 looks for good break points to match the desired line length. Use B<-ndnl>
1775 or B<--nodelete-old-newlines> to force perltidy to retain all old line break
1778 =item B<-anl>, B<--add-newlines>
1780 By default, perltidy will add line breaks when necessary to create
1781 continuations of long lines and to improve the script appearance. Use
1782 B<-nanl> or B<--noadd-newlines> to prevent any new line breaks.
1784 This flag does not prevent perltidy from eliminating existing line
1785 breaks; see B<--freeze-newlines> to completely prevent changes to line
1788 =item Controlling whether perltidy breaks before or after operators
1790 Four command line parameters provide some control over whether
1791 a line break should be before or after specific token types.
1792 Two parameters give detailed control:
1794 B<-wba=s> or B<--want-break-after=s>, and
1796 B<-wbb=s> or B<--want-break-before=s>.
1798 These parameters are each followed by a quoted string, B<s>, containing
1799 a list of token types (separated only by spaces). No more than one of each
1800 of these parameters should be specified, because repeating a
1801 command-line parameter always overwrites the previous one before
1802 perltidy ever sees it.
1804 By default, perltidy breaks B<after> these token types:
1805 % + - * / x != == >= <= =~ !~ < > | &
1806 = **= += *= &= <<= &&= -= /= |= >>= ||= //= .= %= ^= x=
1808 And perltidy breaks B<before> these token types by default:
1811 To illustrate, to cause a break after a concatenation operator, C<'.'>,
1812 rather than before it, the command line would be
1816 As another example, the following command would cause a break before
1817 math operators C<'+'>, C<'-'>, C<'/'>, and C<'*'>:
1821 These commands should work well for most of the token types that perltidy uses
1822 (use B<--dump-token-types> for a list). Also try the B<-D> flag on a short
1823 snippet of code and look at the .DEBUG file to see the tokenization. However,
1824 for a few token types there may be conflicts with hardwired logic which cause
1825 unexpected results. One example is curly braces, which should be controlled
1826 with the parameter B<bl> provided for that purpose.
1828 B<WARNING> Be sure to put these tokens in quotes to avoid having them
1829 misinterpreted by your command shell.
1831 Two additional parameters are available which, though they provide no further
1832 capability, can simplify input are:
1834 B<-baao> or B<--break-after-all-operators>,
1836 B<-bbao> or B<--break-before-all-operators>.
1838 The -baao sets the default to be to break after all of the following operators:
1840 % + - * / x != == >= <= =~ !~ < > | &
1841 = **= += *= &= <<= &&= -= /= |= >>= ||= //= .= %= ^= x=
1842 . : ? && || and or err xor
1844 and the B<-bbao> flag sets the default to break before all of these operators.
1845 These can be used to define an initial break preference which can be fine-tuned
1846 with the B<-wba> and B<-wbb> flags. For example, to break before all operators
1847 except an B<=> one could use --bbao -wba='=' rather than listing every
1848 single perl operator except B<=> on a -wbb flag.
1852 =head2 Controlling List Formatting
1854 Perltidy attempts to place comma-separated arrays of values in tables
1855 which look good. Its default algorithms usually work well, and they
1856 have been improving with each release, but several parameters are
1857 available to control list formatting.
1861 =item B<-boc>, B<--break-at-old-comma-breakpoints>
1863 This flag tells perltidy to try to break at all old commas. This is not
1864 the default. Normally, perltidy makes a best guess at list formatting,
1865 and seldom uses old comma breakpoints. Usually this works well,
1874 The default formatting will flatten this down to one line:
1876 # perltidy (default)
1877 my @list = ( 1, 1, 1, 1, 2, 1, 1, 3, 3, 1, 1, 4, 6, 4, 1, );
1879 which hides the structure. Using B<-boc>, plus additional flags
1880 to retain the original style, yields
1882 # perltidy -boc -lp -pt=2 -vt=1 -vtc=1
1889 A disadvantage of this flag is that all tables in the file
1890 must already be nicely formatted. For another possibility see
1891 the -fs flag in L<Skipping Selected Sections of Code>.
1893 =item B<-mft=n>, B<--maximum-fields-per-table=n>
1895 If the computed number of fields for any table exceeds B<n>, then it
1896 will be reduced to B<n>. The default value for B<n> is a large number,
1897 40. While this value should probably be left unchanged as a general
1898 rule, it might be used on a small section of code to force a list to
1899 have a particular number of fields per line, and then either the B<-boc>
1900 flag could be used to retain this formatting, or a single comment could
1901 be introduced somewhere to freeze the formatting in future applications
1914 =item B<-cab=n>, B<--comma-arrow-breakpoints=n>
1916 A comma which follows a comma arrow, '=>', requires special
1917 consideration. In a long list, it is common to break at all such
1918 commas. This parameter can be used to control how perltidy breaks at
1919 these commas. (However, it will have no effect if old comma breaks are
1920 being forced because B<-boc> is used). The possible values of B<n> are:
1922 n=0 break at all commas after =>
1923 n=1 stable: break at all commas after => unless this would break
1924 an existing one-line container (default)
1925 n=2 break at all commas after =>, but try to form the maximum
1926 maximum one-line container lengths
1927 n=3 do not treat commas after => specially at all
1929 For example, given the following single line, perltidy by default will
1930 not add any line breaks because it would break the existing one-line
1933 bless { B => $B, Root => $Root } => $package;
1935 Using B<-cab=0> will force a break after each comma-arrow item:
1943 If perltidy is subsequently run with this container broken, then by
1944 default it will break after each '=>' because the container is now
1945 broken. To reform a one-line container, the parameter B<-cab=2> would
1948 The flag B<-cab=3> can be used to prevent these commas from being
1949 treated specially. In this case, an item such as "01" => 31 is
1950 treated as a single item in a table. The number of fields in this table
1951 will be determined by the same rules that are used for any other table.
1956 "01" => 31, "02" => 29, "03" => 31, "04" => 30,
1957 "05" => 31, "06" => 30, "07" => 31, "08" => 31,
1958 "09" => 30, "10" => 31, "11" => 30, "12" => 31
1963 =head2 Retaining or Ignoring Existing Line Breaks
1965 Several additional parameters are available for controlling the extent
1966 to which line breaks in the input script influence the output script.
1967 In most cases, the default parameter values are set so that, if a choice
1968 is possible, the output style follows the input style. For example, if
1969 a short logical container is broken in the input script, then the
1970 default behavior is for it to remain broken in the output script.
1972 Most of the parameters in this section would only be required for a
1973 one-time conversion of a script from short container lengths to longer
1974 container lengths. The opposite effect, of converting long container
1975 lengths to shorter lengths, can be obtained by temporarily using a short
1976 maximum line length.
1980 =item B<-bol>, B<--break-at-old-logical-breakpoints>
1982 By default, if a logical expression is broken at a C<&&>, C<||>, C<and>,
1983 or C<or>, then the container will remain broken. Also, breaks
1984 at internal keywords C<if> and C<unless> will normally be retained.
1985 To prevent this, and thus form longer lines, use B<-nbol>.
1987 =item B<-bok>, B<--break-at-old-keyword-breakpoints>
1989 By default, perltidy will retain a breakpoint before keywords which may
1990 return lists, such as C<sort> and <map>. This allows chains of these
1991 operators to be displayed one per line. Use B<-nbok> to prevent
1992 retaining these breakpoints.
1994 =item B<-bot>, B<--break-at-old-ternary-breakpoints>
1996 By default, if a conditional (ternary) operator is broken at a C<:>,
1997 then it will remain broken. To prevent this, and thereby
1998 form longer lines, use B<-nbot>.
2000 =item B<-iob>, B<--ignore-old-breakpoints>
2002 Use this flag to tell perltidy to ignore existing line breaks to the
2003 maximum extent possible. This will tend to produce the longest possible
2004 containers, regardless of type, which do not exceed the line length
2007 =item B<-kis>, B<--keep-interior-semicolons>
2009 Use the B<-kis> flag to prevent breaking at a semicolon if
2010 there was no break there in the input file. Normally
2011 perltidy places a newline after each semicolon which
2012 terminates a statement unless several statements are
2013 contained within a one-line brace block. To illustrate,
2014 consider the following input lines:
2016 dbmclose(%verb_delim); undef %verb_delim;
2017 dbmclose(%expanded); undef %expanded;
2019 The default is to break after each statement, giving
2021 dbmclose(%verb_delim);
2023 dbmclose(%expanded);
2026 With B<perltidy -kis> the multiple statements are retained:
2028 dbmclose(%verb_delim); undef %verb_delim;
2029 dbmclose(%expanded); undef %expanded;
2031 The statements are still subject to the specified value
2032 of B<maximum-line-length> and will be broken if this
2037 =head2 Blank Line Control
2039 Blank lines can improve the readability of a script if they are carefully
2040 placed. Perltidy has several commands for controlling the insertion,
2041 retention, and removal of blank lines.
2045 =item B<-fbl>, B<--freeze-blank-lines>
2047 Set B<-fbl> if you want to the blank lines in your script to
2048 remain exactly as they are. The rest of the parameters in
2049 this section may then be ignored. (Note: setting the B<-fbl> flag
2050 is equivalent to setting B<-mbl=0> and B<-kbl=2>).
2052 =item B<-bbc>, B<--blanks-before-comments>
2054 A blank line will be introduced before a full-line comment. This is the
2055 default. Use B<-nbbc> or B<--noblanks-before-comments> to prevent
2056 such blank lines from being introduced.
2058 =item B<-bbs>, B<--blanks-before-subs>
2060 A blank line will be introduced before a B<sub> definition, unless it is a
2061 one-liner or preceded by a comment. A blank line will also be introduced
2062 before a B<package> statement and a B<BEGIN> and B<END> block. This is the
2063 default. The intention is to help display the structure of a program by
2064 setting off certain key sections of code. This is negated with B<-nbbs> or
2065 B<--noblanks-before-subs>.
2067 =item B<-bbb>, B<--blanks-before-blocks>
2069 A blank line will be introduced before blocks of coding delimited by
2070 B<for>, B<foreach>, B<while>, B<until>, and B<if>, B<unless>, in the following
2077 The block is not preceded by a comment.
2081 The block is not a one-line block.
2085 The number of consecutive non-blank lines at the current indentation depth is at least B<-lbl>
2090 This is the default. The intention of this option is to introduce
2091 some space within dense coding.
2092 This is negated with B<-nbbb> or B<--noblanks-before-blocks>.
2094 =item B<-lbl=n> B<--long-block-line-count=n>
2096 This controls how often perltidy is allowed to add blank lines before
2097 certain block types (see previous section). The default is 8. Entering
2098 a value of B<0> is equivalent to entering a very large number.
2100 =item B<-mbl=n> B<--maximum-consecutive-blank-lines=n>
2102 This parameter specifies the maximum number of consecutive
2103 blank lines which will be output within code sections of a
2104 script. The default is n=1. If the input file has more
2105 than n consecutive blank lines, the number will be reduced
2106 to n. If B<n=0> then no blank lines will be output (unless
2107 all old blank lines are retained with the B<-kbl=2> flag of
2110 This flag obviously does not apply to pod sections,
2111 here-documents, and quotes.
2113 =item B<-kbl=n>, B<--keep-old-blank-lines=n>
2115 The B<-kbl=n> flag gives you control over how your existing blank lines are
2118 The possible values of B<n> are:
2120 n=0 ignore all old blank lines
2121 n=1 stable: keep old blanks, but limited by the value of the B<-mbl=n> flag
2122 n=2 keep all old blank lines, regardless of the value of the B<-mbl=n> flag
2124 The default is B<n=1>.
2126 =item B<-sob>, B<--swallow-optional-blank-lines>
2128 This is equivalent to B<kbl=0> and is included for compatability with
2131 =item B<-nsob>, B<--noswallow-optional-blank-lines>
2133 This is equivalent to B<kbl=1> and is included for compatability with
2140 A style refers to a convenient collection of existing parameters.
2144 =item B<-gnu>, B<--gnu-style>
2146 B<-gnu> gives an approximation to the GNU Coding Standards (which do
2147 not apply to perl) as they are sometimes implemented. At present, this
2148 style overrides the default style with the following parameters:
2150 -lp -bl -noll -pt=2 -bt=2 -sbt=2 -icp
2152 =item B<-pbp>, B<--perl-best-practices>
2154 B<-pbp> is an abbreviation for the parameters in the book B<Perl Best Practices>
2157 -l=78 -i=4 -ci=4 -st -se -vt=2 -cti=0 -pt=1 -bt=1 -sbt=1 -bbt=1 -nsfs -nolq
2158 -wbb="% + - * / x != == >= <= =~ !~ < > | & =
2159 **= += *= &= <<= &&= -= /= |= >>= ||= //= .= %= ^= x="
2161 Note that the -st and -se flags make perltidy act as a filter on one file only.
2162 These can be overridden with -nst and -nse if necessary.
2166 =head2 Other Controls
2170 =item Deleting selected text
2172 Perltidy can selectively delete comments and/or pod documentation. The
2173 command B<-dac> or B<--delete-all-comments> will delete all comments
2174 B<and> all pod documentation, leaving just code and any leading system
2177 The command B<-dp> or B<--delete-pod> will remove all pod documentation
2180 Two commands which remove comments (but not pod) are: B<-dbc> or
2181 B<--delete-block-comments> and B<-dsc> or B<--delete-side-comments>.
2182 (Hanging side comments will be deleted with block comments here.)
2184 The negatives of these commands also work, and are the defaults. When
2185 block comments are deleted, any leading 'hash-bang' will be retained.
2186 Also, if the B<-x> flag is used, any system commands before a leading
2187 hash-bang will be retained (even if they are in the form of comments).
2189 =item Writing selected text to a file
2191 When perltidy writes a formatted text file, it has the ability to also
2192 send selected text to a file with a F<.TEE> extension. This text can
2193 include comments and pod documentation.
2195 The command B<-tac> or B<--tee-all-comments> will write all comments
2196 B<and> all pod documentation.
2198 The command B<-tp> or B<--tee-pod> will write all pod documentation (but
2201 The commands which write comments (but not pod) are: B<-tbc> or
2202 B<--tee-block-comments> and B<-tsc> or B<--tee-side-comments>.
2203 (Hanging side comments will be written with block comments here.)
2205 The negatives of these commands also work, and are the defaults.
2207 =item Using a F<.perltidyrc> command file
2209 If you use perltidy frequently, you probably won't be happy until you
2210 create a F<.perltidyrc> file to avoid typing commonly-used parameters.
2211 Perltidy will first look in your current directory for a command file
2212 named F<.perltidyrc>. If it does not find one, it will continue looking
2213 for one in other standard locations.
2215 These other locations are system-dependent, and may be displayed with
2216 the command C<perltidy -dpro>. Under Unix systems, it will first look
2217 for an environment variable B<PERLTIDY>. Then it will look for a
2218 F<.perltidyrc> file in the home directory, and then for a system-wide
2219 file F</usr/local/etc/perltidyrc>, and then it will look for
2220 F</etc/perltidyrc>. Note that these last two system-wide files do not
2221 have a leading dot. Further system-dependent information will be found
2222 in the INSTALL file distributed with perltidy.
2224 Under Windows, perltidy will also search for a configuration file named perltidy.ini since Windows does not allow files with a leading period (.).
2225 Use C<perltidy -dpro> to see the possbile locations for your system.
2226 An example might be F<C:\Documents and Settings\All Users\perltidy.ini>.
2228 Another option is the use of the PERLTIDY environment variable.
2229 The method for setting environment variables depends upon the version of
2230 Windows that you are using. Instructions for Windows 95 and later versions can
2233 http://www.netmanage.com/000/20021101_005_tcm21-6336.pdf
2235 Under Windows NT / 2000 / XP the PERLTIDY environment variable can be placed in
2236 either the user section or the system section. The later makes the
2237 configuration file common to all users on the machine. Be sure to enter the
2238 full path of the configuration file in the value of the environment variable.
2239 Ex. PERLTIDY=C:\Documents and Settings\perltidy.ini
2241 The configuation file is free format, and simply a list of parameters, just as
2242 they would be entered on a command line. Any number of lines may be used, with
2243 any number of parameters per line, although it may be easiest to read with one
2244 parameter per line. Blank lines are ignored, and text after a '#' is ignored
2245 to the end of a line.
2247 Here is an example of a F<.perltidyrc> file:
2249 # This is a simple of a .perltidyrc configuration file
2250 # This implements a highly spaced style
2251 -se # errors to standard error output
2252 -w # show all warnings
2253 -bl # braces on new lines
2254 -pt=0 # parens not tight at all
2255 -bt=0 # braces not tight
2256 -sbt=0 # square brackets not tight
2258 The parameters in the F<.perltidyrc> file are installed first, so any
2259 parameters given on the command line will have priority over them.
2261 To avoid confusion, perltidy ignores any command in the .perltidyrc
2262 file which would cause some kind of dump and an exit. These are:
2264 -h -v -ddf -dln -dop -dsn -dtt -dwls -dwrs -ss
2266 There are several options may be helpful in debugging a F<.perltidyrc>
2273 A very helpful command is B<--dump-profile> or B<-dpro>. It writes a
2274 list of all configuration filenames tested to standard output, and
2275 if a file is found, it dumps the content to standard output before
2276 exiting. So, to find out where perltidy looks for its configuration
2277 files, and which one if any it selects, just enter
2283 It may be simplest to develop and test configuration files with
2284 alternative names, and invoke them with B<-pro=filename> on the command
2285 line. Then rename the desired file to F<.perltidyrc> when finished.
2289 The parameters in the F<.perltidyrc> file can be switched off with
2290 the B<-npro> option.
2294 The commands B<--dump-options>, B<--dump-defaults>, B<--dump-long-names>,
2295 and B<--dump-short-names>, all described below, may all be helpful.
2299 =item Creating a new abbreviation
2301 A special notation is available for use in a F<.perltidyrc> file
2302 for creating an abbreviation for a group
2303 of options. This can be used to create a
2304 shorthand for one or more styles which are frequently, but not always,
2305 used. The notation is to group the options within curly braces which
2306 are preceded by the name of the alias (without leading dashes), like this:
2313 where B<newword> is the abbreviation, and B<opt1>, etc, are existing parameters
2314 I<or other abbreviations>. The main syntax requirement is that
2315 the new abbreviation must begin on a new line.
2316 Space before and after the curly braces is optional.
2318 specific example, the following line
2320 airy {-bl -pt=0 -bt=0 -sbt=0}
2322 could be placed in a F<.perltidyrc> file, and then invoked at will with
2324 perltidy -airy somefile.pl
2326 (Either C<-airy> or C<--airy> may be used).
2328 =item Skipping leading non-perl commands with B<-x> or B<--look-for-hash-bang>
2330 If your script has leading lines of system commands or other text which
2331 are not valid perl code, and which are separated from the start of the
2332 perl code by a "hash-bang" line, ( a line of the form C<#!...perl> ),
2333 you must use the B<-x> flag to tell perltidy not to parse and format any
2334 lines before the "hash-bang" line. This option also invokes perl with a
2335 -x flag when checking the syntax. This option was originally added to
2336 allow perltidy to parse interactive VMS scripts, but it should be used
2337 for any script which is normally invoked with C<perl -x>.
2339 =item Making a file unreadable
2341 The goal of perltidy is to improve the readability of files, but there
2342 are two commands which have the opposite effect, B<--mangle> and
2343 B<--extrude>. They are actually
2344 merely aliases for combinations of other parameters. Both of these
2345 strip all possible whitespace, but leave comments and pod documents,
2346 so that they are essentially reversible. The
2347 difference between these is that B<--mangle> puts the fewest possible
2348 line breaks in a script while B<--extrude> puts the maximum possible.
2349 Note that these options do not provided any meaningful obfuscation, because
2350 perltidy can be used to reformat the files. They were originally
2351 developed to help test the tokenization logic of perltidy, but they
2353 One use for B<--mangle> is the following:
2355 perltidy --mangle myfile.pl -st | perltidy -o myfile.pl.new
2357 This will form the maximum possible number of one-line blocks (see next
2358 section), and can sometimes help clean up a badly formatted script.
2360 A similar technique can be used with B<--extrude> instead of B<--mangle>
2361 to make the minimum number of one-line blocks.
2363 Another use for B<--mangle> is to combine it with B<-dac> to reduce
2364 the file size of a perl script.
2366 =item One-line blocks
2368 There are a few points to note regarding one-line blocks. A one-line
2369 block is something like this,
2371 if ($x > 0) { $y = 1 / $x }
2373 where the contents within the curly braces is short enough to fit
2376 With few exceptions, perltidy retains existing one-line blocks, if it
2377 is possible within the line-length constraint, but it does not attempt
2378 to form new ones. In other words, perltidy will try to follow the
2379 one-line block style of the input file.
2381 If an existing one-line block is longer than the maximum line length,
2382 however, it will be broken into multiple lines. When this happens, perltidy
2383 checks for and adds any optional terminating semicolon (unless the B<-nasc>
2384 option is used) if the block is a code block.
2386 The main exception is that perltidy will attempt to form new one-line
2387 blocks following the keywords C<map>, C<eval>, and C<sort>, because
2388 these code blocks are often small and most clearly displayed in a single
2391 One-line block rules can conflict with the cuddled-else option. When
2392 the cuddled-else option is used, perltidy retains existing one-line
2393 blocks, even if they do not obey cuddled-else formatting.
2395 Occasionally, when one-line blocks get broken because they exceed the
2396 available line length, the formatting will violate the requested brace style.
2397 If this happens, reformatting the script a second time should correct
2402 The following flags are available for debugging:
2404 B<--dump-defaults> or B<-ddf> will write the default option set to standard output and quit
2406 B<--dump-profile> or B<-dpro> will write the name of the current
2407 configuration file and its contents to standard output and quit.
2409 B<--dump-options> or B<-dop> will write current option set to standard
2412 B<--dump-long-names> or B<-dln> will write all command line long names (passed
2413 to Get_options) to standard output and quit.
2415 B<--dump-short-names> or B<-dsn> will write all command line short names
2416 to standard output and quit.
2418 B<--dump-token-types> or B<-dtt> will write a list of all token types
2419 to standard output and quit.
2421 B<--dump-want-left-space> or B<-dwls> will write the hash %want_left_space
2422 to standard output and quit. See the section on controlling whitespace
2425 B<--dump-want-right-space> or B<-dwrs> will write the hash %want_right_space
2426 to standard output and quit. See the section on controlling whitespace
2429 B<-DEBUG> will write a file with extension F<.DEBUG> for each input file
2430 showing the tokenization of all lines of code.
2432 =item Working with MakeMaker, AutoLoader and SelfLoader
2434 The first $VERSION line of a file which might be eval'd by MakeMaker
2435 is passed through unchanged except for indentation.
2436 Use B<--nopass-version-line>, or B<-npvl>, to deactivate this feature.
2438 If the AutoLoader module is used, perltidy will continue formatting
2439 code after seeing an __END__ line.
2440 Use B<--nolook-for-autoloader>, or B<-nlal>, to deactivate this feature.
2442 Likewise, if the SelfLoader module is used, perltidy will continue formatting
2443 code after seeing a __DATA__ line.
2444 Use B<--nolook-for-selfloader>, or B<-nlsl>, to deactivate this feature.
2446 =item Working around problems with older version of Perl
2448 Perltidy contains a number of rules which help avoid known subtleties
2449 and problems with older versions of perl, and these rules always
2450 take priority over whatever formatting flags have been set. For example,
2451 perltidy will usually avoid starting a new line with a bareword, because
2452 this might cause problems if C<use strict> is active.
2454 There is no way to override these rules.
2462 =item The B<-html> master switch
2464 The flag B<-html> causes perltidy to write an html file with extension
2465 F<.html>. So, for example, the following command
2467 perltidy -html somefile.pl
2469 will produce a syntax-colored html file named F<somefile.pl.html>
2470 which may be viewed with a browser.
2472 B<Please Note>: In this case, perltidy does not do any formatting to the
2473 input file, and it does not write a formatted file with extension
2474 F<.tdy>. This means that two perltidy runs are required to create a
2475 fully reformatted, html copy of a script.
2477 =item The B<-pre> flag for code snippets
2479 When the B<-pre> flag is given, only the pre-formatted section, within
2480 the <PRE> and </PRE> tags, will be output. This simplifies inclusion
2481 of the output in other files. The default is to output a complete
2484 =item The B<-nnn> flag for line numbering
2486 When the B<-nnn> flag is given, the output lines will be numbered.
2488 =item The B<-toc>, or B<--html-table-of-contents> flag
2490 By default, a table of contents to packages and subroutines will be
2491 written at the start of html output. Use B<-ntoc> to prevent this.
2492 This might be useful, for example, for a pod document which contains a
2493 number of unrelated code snippets. This flag only influences the code
2494 table of contents; it has no effect on any table of contents produced by
2495 pod2html (see next item).
2497 =item The B<-pod>, or B<--pod2html> flag
2499 There are two options for formatting pod documentation. The default is
2500 to pass the pod through the Pod::Html module (which forms the basis of
2501 the pod2html utility). Any code sections are formatted by perltidy, and
2502 the results then merged. Note: perltidy creates a temporary file when
2503 Pod::Html is used; see L<"FILES">. Also, Pod::Html creates temporary
2504 files for its cache.
2506 NOTE: Perltidy counts the number of C<=cut> lines, and either moves the
2507 pod text to the top of the html file if there is one C<=cut>, or leaves
2508 the pod text in its original order (interleaved with code) otherwise.
2510 Most of the flags accepted by pod2html may be included in the perltidy
2511 command line, and they will be passed to pod2html. In some cases,
2512 the flags have a prefix C<pod> to emphasize that they are for the
2513 pod2html, and this prefix will be removed before they are passed to
2514 pod2html. The flags which have the additional C<pod> prefix are:
2516 --[no]podheader --[no]podindex --[no]podrecurse --[no]podquiet
2517 --[no]podverbose --podflush
2519 The flags which are unchanged from their use in pod2html are:
2521 --backlink=s --cachedir=s --htmlroot=s --libpods=s --title=s
2522 --podpath=s --podroot=s
2524 where 's' is an appropriate character string. Not all of these flags are
2525 available in older versions of Pod::Html. See your Pod::Html documentation for
2528 The alternative, indicated with B<-npod>, is not to use Pod::Html, but
2529 rather to format pod text in italics (or whatever the stylesheet
2530 indicates), without special html markup. This is useful, for example,
2531 if pod is being used as an alternative way to write comments.
2533 =item The B<-frm>, or B<--frames> flag
2535 By default, a single html output file is produced. This can be changed
2536 with the B<-frm> option, which creates a frame holding a table of
2537 contents in the left panel and the source code in the right side. This
2538 simplifies code browsing. Assume, for example, that the input file is
2539 F<MyModule.pm>. Then, for default file extension choices, these three
2540 files will be created:
2542 MyModule.pm.html - the frame
2543 MyModule.pm.toc.html - the table of contents
2544 MyModule.pm.src.html - the formatted source code
2546 Obviously this file naming scheme requires that output be directed to a real
2547 file (as opposed to, say, standard output). If this is not the
2548 case, or if the file extension is unknown, the B<-frm> option will be
2551 =item The B<-text=s>, or B<--html-toc-extension> flag
2553 Use this flag to specify the extra file extension of the table of contents file
2554 when html frames are used. The default is "toc".
2555 See L<Specifying File Extensions>.
2557 =item The B<-sext=s>, or B<--html-src-extension> flag
2559 Use this flag to specify the extra file extension of the content file when html
2560 frames are used. The default is "src".
2561 See L<Specifying File Extensions>.
2563 =item The B<-hent>, or B<--html-entities> flag
2565 This flag controls the use of Html::Entities for html formatting. By
2566 default, the module Html::Entities is used to encode special symbols.
2567 This may not be the right thing for some browser/language
2568 combinations. Use --nohtml-entities or -nhent to prevent this.
2572 Style sheets make it very convenient to control and adjust the
2573 appearance of html pages. The default behavior is to write a page of
2574 html with an embedded style sheet.
2576 An alternative to an embedded style sheet is to create a page with a
2577 link to an external style sheet. This is indicated with the
2578 B<-css=filename>, where the external style sheet is F<filename>. The
2579 external style sheet F<filename> will be created if and only if it does
2580 not exist. This option is useful for controlling multiple pages from a
2583 To cause perltidy to write a style sheet to standard output and exit,
2584 use the B<-ss>, or B<--stylesheet>, flag. This is useful if the style
2585 sheet could not be written for some reason, such as if the B<-pre> flag
2586 was used. Thus, for example,
2588 perltidy -html -ss >mystyle.css
2590 will write a style sheet with the default properties to file
2593 The use of style sheets is encouraged, but a web page without a style
2594 sheets can be created with the flag B<-nss>. Use this option if you
2595 must to be sure that older browsers (roughly speaking, versions prior to
2596 4.0 of Netscape Navigator and Internet Explorer) can display the
2597 syntax-coloring of the html files.
2599 =item Controlling HTML properties
2601 Note: It is usually more convenient to accept the default properties
2602 and then edit the stylesheet which is produced. However, this section
2603 shows how to control the properties with flags to perltidy.
2605 Syntax colors may be changed from their default values by flags of the either
2606 the long form, B<-html-color-xxxxxx=n>, or more conveniently the short form,
2607 B<-hcx=n>, where B<xxxxxx> is one of the following words, and B<x> is the
2608 corresponding abbreviation:
2611 ---------- -------- --
2614 identifier identifier i
2615 bareword, function bareword w
2617 quite, pattern quote q
2618 here doc text here-doc-text h
2619 here doc target here-doc-target hh
2620 punctuation punctuation pu
2622 structural braces structure s
2623 semicolon semicolon sc
2627 sub definition name subroutine m
2628 pod text pod-text pd
2630 A default set of colors has been defined, but they may be changed by providing
2631 values to any of the following parameters, where B<n> is either a 6 digit
2632 hex RGB color value or an ascii name for a color, such as 'red'.
2634 To illustrate, the following command will produce an html
2635 file F<somefile.pl.html> with "aqua" keywords:
2637 perltidy -html -hck=00ffff somefile.pl
2639 and this should be equivalent for most browsers:
2641 perltidy -html -hck=aqua somefile.pl
2643 Perltidy merely writes any non-hex names that it sees in the html file.
2644 The following 16 color names are defined in the HTML 3.2 standard:
2663 Many more names are supported in specific browsers, but it is safest
2664 to use the hex codes for other colors. Helpful color tables can be
2665 located with an internet search for "HTML color tables".
2667 Besides color, two other character attributes may be set: bold, and italics.
2668 To set a token type to use bold, use the flag
2669 B<--html-bold-xxxxxx> or B<-hbx>, where B<xxxxxx> or B<x> are the long
2670 or short names from the above table. Conversely, to set a token type to
2671 NOT use bold, use B<--nohtml-bold-xxxxxx> or B<-nhbx>.
2673 Likewise, to set a token type to use an italic font, use the flag
2674 B<--html-italic-xxxxxx> or B<-hix>, where again B<xxxxxx> or B<x> are the
2675 long or short names from the above table. And to set a token type to
2676 NOT use italics, use B<--nohtml-italic-xxxxxx> or B<-nhix>.
2678 For example, to use bold braces and lime color, non-bold, italics keywords the
2679 following command would be used:
2681 perltidy -html -hbs -hck=00FF00 -nhbk -hik somefile.pl
2683 The background color can be specified with B<--html-color-background=n>,
2684 or B<-hcbg=n> for short, where n is a 6 character hex RGB value. The
2685 default color of text is the value given to B<punctuation>, which is
2688 Here are some notes and hints:
2690 1. If you find a preferred set of these parameters, you may want
2691 to create a F<.perltidyrc> file containing them. See the perltidy man
2692 page for an explanation.
2694 2. Rather than specifying values for these parameters, it is probably
2695 easier to accept the defaults and then edit a style sheet. The style
2696 sheet contains comments which should make this easy.
2698 3. The syntax-colored html files can be very large, so it may be best to
2699 split large files into smaller pieces to improve download times.
2703 =head1 SOME COMMON INPUT CONVENTIONS
2705 =head2 Specifying Block Types
2707 Several parameters which refer to code block types may be customized by also
2708 specifying an associated list of block types. The type of a block is the name
2709 of the keyword which introduces that block, such as B<if>, B<else>, or B<sub>.
2710 An exception is a labeled block, which has no keyword, and should be specified
2713 For example, the following parameter specifies C<sub>, labels, C<BEGIN>, and
2716 -cscl="sub : BEGIN END"
2718 (the meaning of the -cscl parameter is described above.) Note that
2719 quotes are required around the list of block types because of the
2722 =head2 Specifying File Extensions
2724 Several parameters allow default file extensions to be overridden. For
2725 example, a backup file extension may be specified with B<-bext=ext>,
2726 where B<ext> is some new extension. In order to provides the user some
2727 flexibility, the following convention is used in all cases to decide if
2728 a leading '.' should be used. If the extension C<ext> begins with
2729 C<A-Z>, C<a-z>, or C<0-9>, then it will be appended to the filename with
2730 an intermediate '.' (or perhaps an '_' on VMS systems). Otherwise, it
2731 will be appended directly.
2733 For example, suppose the file is F<somefile.pl>. For C<-bext=old>, a '.' is
2734 added to give F<somefile.pl.old>. For C<-bext=.old>, no additional '.' is
2735 added, so again the backup file is F<somefile.pl.old>. For C<-bext=~>, then no
2736 dot is added, and the backup file will be F<somefile.pl~> .
2738 =head1 SWITCHES WHICH MAY BE NEGATED
2740 The following list shows all short parameter names which allow a prefix
2741 'n' to produce the negated form:
2743 D anl asc aws b bbb bbc bbs bl bli boc bok bol bot ce
2744 csc dac dbc dcsc ddf dln dnl dop dp dpro dsc dsm dsn dtt dwls
2745 dwrs dws f fll frm fs hsc html ibc icb icp iob isbc lal log
2746 lp lsl ohbr okw ola oll opr opt osbr otr ple ple pod pvl q
2747 sbc sbl schb scp scsb sct se sfp sfs skp sob sohb sop sosb sot
2748 ssc st sts syn t tac tbc toc tp tqw tsc w x bar kis
2750 Equivalently, the prefix 'no' or 'no-' on the corresponding long names may be
2757 =item Parsing Limitations
2759 Perltidy should work properly on most perl scripts. It does a lot of
2760 self-checking, but still, it is possible that an error could be
2761 introduced and go undetected. Therefore, it is essential to make
2762 careful backups and to test reformatted scripts.
2764 The main current limitation is that perltidy does not scan modules
2765 included with 'use' statements. This makes it necessary to guess the
2766 context of any bare words introduced by such modules. Perltidy has good
2767 guessing algorithms, but they are not infallible. When it must guess,
2768 it leaves a message in the log file.
2770 If you encounter a bug, please report it.
2772 =item What perltidy does not parse and format
2774 Perltidy indents but does not reformat comments and C<qw> quotes.
2775 Perltidy does not in any way modify the contents of here documents or
2776 quoted text, even if they contain source code. (You could, however,
2777 reformat them separately). Perltidy does not format 'format' sections
2778 in any way. And, of course, it does not modify pod documents.
2786 =item Temporary files
2788 Under the -html option with the default --pod2html flag, a temporary file is
2789 required to pass text to Pod::Html. Unix systems will try to use the POSIX
2790 tmpnam() function. Otherwise the file F<perltidy.TMP> will be temporarily
2791 created in the current working directory.
2793 =item Special files when standard input is used
2795 When standard input is used, the log file, if saved, is F<perltidy.LOG>,
2796 and any errors are written to F<perltidy.ERR> unless the B<-se> flag is
2797 set. These are saved in the current working directory.
2799 =item Files overwritten
2801 The following file extensions are used by perltidy, and files with these
2802 extensions may be overwritten or deleted: F<.ERR>, F<.LOG>, F<.TEE>,
2803 and/or F<.tdy>, F<.html>, and F<.bak>, depending on the run type and
2806 =item Files extensions limitations
2808 Perltidy does not operate on files for which the run could produce a file with
2809 a duplicated file extension. These extensions include F<.LOG>, F<.ERR>,
2810 F<.TEE>, and perhaps F<.tdy> and F<.bak>, depending on the run type. The
2811 purpose of this rule is to prevent generating confusing filenames such as
2812 F<somefile.tdy.tdy.tdy>.
2818 perlstyle(1), Perl::Tidy(3)
2822 This man page documents perltidy version 20101217.
2826 Michael Cartmell supplied code for adaptation to VMS and helped with
2829 Yves Orton supplied code for adaptation to the various versions
2832 Axel Rose supplied a patch for MacPerl.
2834 Hugh S. Myers designed and implemented the initial Perl::Tidy module interface.
2836 Many others have supplied key ideas, suggestions, and bug reports;
2837 see the CHANGES file.
2842 email: perltidy at users.sourceforge.net
2843 http://perltidy.sourceforge.net
2847 Copyright (c) 2000-2010 by Steve Hancock
2851 This package is free software; you can redistribute it and/or modify it
2852 under the terms of the "GNU General Public License".
2854 Please refer to the file "COPYING" for details.
2858 This package is distributed in the hope that it will be useful,
2859 but WITHOUT ANY WARRANTY; without even the implied warranty of
2860 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
2862 See the "GNU General Public License" for more details.