6 my $arg_string = undef;
8 # give Macs a chance to provide command line parameters
10 $arg_string = MacPerl::Ask(
11 'Please enter @ARGV (-h for help)',
12 defined $ARGV[0] ? "\"$ARGV[0]\"" : ""
16 Perl::Tidy::perltidy( argv => $arg_string );
22 perltidy - a perl script indenter and reformatter
26 perltidy [ options ] file1 file2 file3 ...
27 (output goes to file1.tdy, file2.tdy, file3.tdy, ...)
28 perltidy [ options ] file1 -o outfile
29 perltidy [ options ] file1 -st >outfile
30 perltidy [ options ] <infile >outfile
34 Perltidy reads a perl script and writes an indented, reformatted script.
36 Many users will find enough information in L<"EXAMPLES"> to get
37 started. New users may benefit from the short tutorial
39 http://perltidy.sourceforge.net/tutorial.html
41 A convenient aid to systematically defining a set of style parameters
43 http://perltidy.sourceforge.net/stylekey.html
45 Perltidy can produce output on either of two modes, depending on the
46 existence of an B<-html> flag. Without this flag, the output is passed
47 through a formatter. The default formatting tries to follow the
48 recommendations in perlstyle(1), but it can be controlled in detail with
49 numerous input parameters, which are described in L<"FORMATTING
52 When the B<-html> flag is given, the output is passed through an HTML
53 formatter which is described in L<"HTML OPTIONS">.
59 This will produce a file F<somefile.pl.tdy> containing the script reformatted
60 using the default options, which approximate the style suggested in
61 perlstyle(1). The source file F<somefile.pl> is unchanged.
65 Execute perltidy on all F<.pl> files in the current directory with the
66 default options. The output will be in files with an appended F<.tdy>
67 extension. For any file with an error, there will be a file with extension
70 perltidy -b file1.pl file2.pl
72 Modify F<file1.pl> and F<file2.pl> in place, and backup the originals to
73 F<file1.pl.bak> and F<file2.pl.bak>. If F<file1.pl.bak> and/or F<file2.pl.bak>
74 already exist, they will be overwritten.
76 perltidy -b -bext='/' file1.pl file2.pl
78 Same as the previous example except that the backup files F<file1.pl.bak> and F<file2.pl.bak> will be deleted if there are no errors.
80 perltidy -gnu somefile.pl
82 Execute perltidy on file F<somefile.pl> with a style which approximates the
83 GNU Coding Standards for C programs. The output will be F<somefile.pl.tdy>.
85 perltidy -i=3 somefile.pl
87 Execute perltidy on file F<somefile.pl>, with 3 columns for each level of
88 indentation (B<-i=3>) instead of the default 4 columns. There will not be any
89 tabs in the reformatted script, except for any which already exist in comments,
90 pod documents, quotes, and here documents. Output will be F<somefile.pl.tdy>.
92 perltidy -i=3 -et=8 somefile.pl
94 Same as the previous example, except that leading whitespace will
95 be entabbed with one tab character per 8 spaces.
97 perltidy -ce -l=72 somefile.pl
99 Execute perltidy on file F<somefile.pl> with all defaults except use "cuddled
100 elses" (B<-ce>) and a maximum line length of 72 columns (B<-l=72>) instead of
101 the default 80 columns.
103 perltidy -g somefile.pl
105 Execute perltidy on file F<somefile.pl> and save a log file F<somefile.pl.LOG>
106 which shows the nesting of braces, parentheses, and square brackets at
107 the start of every line.
109 perltidy -html somefile.pl
111 This will produce a file F<somefile.pl.html> containing the script with
112 html markup. The output file will contain an embedded style sheet in
113 the <HEAD> section which may be edited to change the appearance.
115 perltidy -html -css=mystyle.css somefile.pl
117 This will produce a file F<somefile.pl.html> containing the script with
118 html markup. This output file will contain a link to a separate style
119 sheet file F<mystyle.css>. If the file F<mystyle.css> does not exist,
120 it will be created. If it exists, it will not be overwritten.
122 perltidy -html -pre somefile.pl
124 Write an html snippet with only the PRE section to F<somefile.pl.html>.
125 This is useful when code snippets are being formatted for inclusion in a
126 larger web page. No style sheet will be written in this case.
128 perltidy -html -ss >mystyle.css
130 Write a style sheet to F<mystyle.css> and exit.
132 perltidy -html -frm mymodule.pm
134 Write html with a frame holding a table of contents and the source code. The
135 output files will be F<mymodule.pm.html> (the frame), F<mymodule.pm.toc.html>
136 (the table of contents), and F<mymodule.pm.src.html> (the source code).
138 =head1 OPTIONS - OVERVIEW
140 The entire command line is scanned for options, and they are processed
141 before any files are processed. As a result, it does not matter
142 whether flags are before or after any filenames. However, the relative
143 order of parameters is important, with later parameters overriding the
144 values of earlier parameters.
146 For each parameter, there is a long name and a short name. The short
147 names are convenient for keyboard input, while the long names are
148 self-documenting and therefore useful in scripts. It is customary to
149 use two leading dashes for long names, but one may be used.
151 Most parameters which serve as on/off flags can be negated with a
152 leading "n" (for the short name) or a leading "no" or "no-" (for the
153 long name). For example, the flag to outdent long quotes is B<-olq>
154 or B<--outdent-long-quotes>. The flag to skip this is B<-nolq>
155 or B<--nooutdent-long-quotes> or B<--no-outdent-long-quotes>.
157 Options may not be bundled together. In other words, options B<-q> and
158 B<-g> may NOT be entered as B<-qg>.
160 Option names may be terminated early as long as they are uniquely identified.
161 For example, instead of B<--dump-token-types>, it would be sufficient to enter
162 B<--dump-tok>, or even B<--dump-t>, to uniquely identify this command.
166 The following parameters concern the files which are read and written.
170 =item B<-h>, B<--help>
172 Show summary of usage and exit.
174 =item B<-o>=filename, B<--outfile>=filename
176 Name of the output file (only if a single input file is being
177 processed). If no output file is specified, and output is not
178 redirected to the standard output, the output will go to F<filename.tdy>.
180 =item B<-st>, B<--standard-output>
182 Perltidy must be able to operate on an arbitrarily large number of files
183 in a single run, with each output being directed to a different output
184 file. Obviously this would conflict with outputting to the single
185 standard output device, so a special flag, B<-st>, is required to
186 request outputting to the standard output. For example,
188 perltidy somefile.pl -st >somefile.new.pl
190 This option may only be used if there is just a single input file.
191 The default is B<-nst> or B<--nostandard-output>.
193 =item B<-se>, B<--standard-error-output>
195 If perltidy detects an error when processing file F<somefile.pl>, its
196 default behavior is to write error messages to file F<somefile.pl.ERR>.
197 Use B<-se> to cause all error messages to be sent to the standard error
198 output stream instead. This directive may be negated with B<-nse>.
199 Thus, you may place B<-se> in a F<.perltidyrc> and override it when
200 desired with B<-nse> on the command line.
202 =item B<-oext>=ext, B<--output-file-extension>=ext
204 Change the extension of the output file to be F<ext> instead of the
205 default F<tdy> (or F<html> in case the -B<-html> option is used).
206 See L<Specifying File Extensions>.
208 =item B<-opath>=path, B<--output-path>=path
210 When perltidy creates a filename for an output file, by default it merely
211 appends an extension to the path and basename of the input file. This
212 parameter causes the path to be changed to F<path> instead.
214 The path should end in a valid path separator character, but perltidy will try
215 to add one if it is missing.
219 perltidy somefile.pl -opath=/tmp/
221 will produce F</tmp/somefile.pl.tdy>. Otherwise, F<somefile.pl.tdy> will
222 appear in whatever directory contains F<somefile.pl>.
224 If the path contains spaces, it should be placed in quotes.
226 This parameter will be ignored if output is being directed to standard output,
227 or if it is being specified explicitly with the B<-o=s> parameter.
229 =item B<-b>, B<--backup-and-modify-in-place>
231 Modify the input file or files in-place and save the original with the
232 extension F<.bak>. Any existing F<.bak> file will be deleted. See next
233 item for changing the default backup extension, and for eliminating the
234 backup file altogether.
236 A B<-b> flag will be ignored if input is from standard input or goes to
237 standard output, or if the B<-html> flag is set.
239 In particular, if you want to use both the B<-b> flag and the B<-pbp>
240 (--perl-best-practices) flag, then you must put a B<-nst> flag after the
241 B<-pbp> flag because it contains a B<-st> flag as one of its components,
242 which means that output will go to the standard output stream.
244 =item B<-bext>=ext, B<--backup-file-extension>=ext
246 This parameter serves two purposes: (1) to change the extension of the backup
247 file to be something other than the default F<.bak>, and (2) to indicate
248 that no backup file should be saved.
250 To change the default extension to something other than F<.bak> see
251 L<Specifying File Extensions>.
253 A backup file of the source is always written, but you can request that it
254 be deleted at the end of processing if there were no errors. This is risky
255 unless the source code is being maintained with a source code control
258 To indicate that the backup should be deleted include one forward slash,
259 B</>, in the extension. If any text remains after the slash is removed
260 it will be used to define the backup file extension (which is always
261 created and only deleted if there were no errors).
263 Here are some examples:
265 Parameter Extension Backup File Treatment
266 <-bext=bak> F<.bak> Keep (same as the default behavior)
267 <-bext='/'> F<.bak> Delete if no errors
268 <-bext='/backup'> F<.backup> Delete if no errors
269 <-bext='original/'> F<.original> Delete if no errors
271 =item B<-w>, B<--warning-output>
273 Setting B<-w> causes any non-critical warning
274 messages to be reported as errors. These include messages
275 about possible pod problems, possibly bad starting indentation level,
276 and cautions about indirect object usage. The default, B<-nw> or
277 B<--nowarning-output>, is not to include these warnings.
279 =item B<-q>, B<--quiet>
281 Deactivate error messages and syntax checking (for running under
284 For example, if you use a vi-style editor, such as vim, you may execute
285 perltidy as a filter from within the editor using something like
289 where C<n1,n2> represents the selected text. Without the B<-q> flag,
290 any error message may mess up your screen, so be prepared to use your
293 =item B<-log>, B<--logfile>
295 Save the F<.LOG> file, which has many useful diagnostics. Perltidy always
296 creates a F<.LOG> file, but by default it is deleted unless a program bug is
297 suspected. Setting the B<-log> flag forces the log file to be saved.
299 =item B<-g=n>, B<--logfile-gap=n>
301 Set maximum interval between input code lines in the logfile. This purpose of
302 this flag is to assist in debugging nesting errors. The value of C<n> is
303 optional. If you set the flag B<-g> without the value of C<n>, it will be
304 taken to be 1, meaning that every line will be written to the log file. This
305 can be helpful if you are looking for a brace, paren, or bracket nesting error.
307 Setting B<-g> also causes the logfile to be saved, so it is not necessary to
308 also include B<-log>.
310 If no B<-g> flag is given, a value of 50 will be used, meaning that at least
311 every 50th line will be recorded in the logfile. This helps prevent
312 excessively long log files.
314 Setting a negative value of C<n> is the same as not setting B<-g> at all.
316 =item B<-npro> B<--noprofile>
318 Ignore any F<.perltidyrc> command file. Normally, perltidy looks first in
319 your current directory for a F<.perltidyrc> file of parameters. (The format
320 is described below). If it finds one, it applies those options to the
321 initial default values, and then it applies any that have been defined
322 on the command line. If no F<.perltidyrc> file is found, it looks for one
323 in your home directory.
325 If you set the B<-npro> flag, perltidy will not look for this file.
327 =item B<-pro=filename> or B<--profile=filename>
329 To simplify testing and switching .perltidyrc files, this command may be
330 used to specify a configuration file which will override the default
331 name of .perltidyrc. There must not be a space on either side of the
332 '=' sign. For example, the line
334 perltidy -pro=testcfg
336 would cause file F<testcfg> to be used instead of the
337 default F<.perltidyrc>.
339 A pathname begins with three dots, e.g. ".../.perltidyrc", indicates that
340 the file should be searched for starting in the current directory and
341 working upwards. This makes it easier to have multiple projects each with
342 their own .perltidyrc in their root directories.
344 =item B<-opt>, B<--show-options>
346 Write a list of all options used to the F<.LOG> file.
347 Please see B<--dump-options> for a simpler way to do this.
349 =item B<-f>, B<--force-read-binary>
351 Force perltidy to process binary files. To avoid producing excessive
352 error messages, perltidy skips files identified by the system as non-text.
353 However, valid perl scripts containing binary data may sometimes be identified
354 as non-text, and this flag forces perltidy to process them.
358 =head1 FORMATTING OPTIONS
366 This flag disables all formatting and causes the input to be copied unchanged
367 to the output except for possible changes in line ending characters and any
368 pre- and post-filters. This can be useful in conjunction with a hierarchical
369 set of F<.perltidyrc> files to avoid unwanted code tidying. See also
370 L<Skipping Selected Sections of Code> for a way to avoid tidying specific
373 =item B<-i=n>, B<--indent-columns=n>
375 Use n columns per indentation level (default n=4).
377 =item B<-l=n>, B<--maximum-line-length=n>
379 The default maximum line length is n=80 characters. Perltidy will try
380 to find line break points to keep lines below this length. However, long
381 quotes and side comments may cause lines to exceed this length.
382 Setting B<-l=0> is equivalent to setting B<-l=(a large number)>.
384 =item B<-vmll>, B<--variable-maximum-line-length>
386 A problem arises using a fixed maximum line length with very deeply nested code
387 and data structures because eventually the amount of leading whitespace used
388 for indicating indentation takes up most or all of the available line width,
389 leaving little or no space for the actual code or data. One solution is to use
390 a vary long line length. Another solution is to use the B<-vmll> flag, which
391 basically tells perltidy to ignore leading whitespace when measuring the line
394 To be precise, when the B<-vmll> parameter is set, the maximum line length of a
395 line of code will be M+L*I, where
397 M is the value of --maximum-line-length=M (-l=M), default 80,
398 I is the value of --indent-columns=I (-i=I), default 4,
399 L is the indentation level of the line of code
401 When this flag is set, the choice of breakpoints for a block of code should be
402 essentially independent of its nesting depth. However, the absolute line
403 lengths, including leading whitespace, can still be arbitrarily large. This
404 problem can be avoided by including the next parameter.
406 The default is not to do this (B<-nvmll>).
408 =item B<-wc=n>, B<--whitespace-cycle=n>
410 This flag also addresses problems with very deeply nested code and data
411 structures. When the nesting depth exceeds the value B<n> the leading
412 whitespace will be reduced and start at a depth of 1 again. The result is that
413 blocks of code will shift back to the left rather than moving arbitrarily far
414 to the right. This occurs cyclically to any depth.
416 For example if one level of indentation equals 4 spaces (B<-i=4>, the default),
417 and one uses B<-wc=15>, then if the leading whitespace on a line exceeds about
418 4*15=60 spaces it will be reduced back to 4*1=4 spaces and continue increasing
419 from there. If the whitespace never exceeds this limit the formatting remains
422 The combination of B<-vmll> and B<-wc=n> provides a solution to the problem of
423 displaying arbitrarily deep data structures and code in a finite window,
424 although B<-wc=n> may of course be used without B<-vmll>.
426 The default is not to use this, which can also be indicated using B<-wc=0>.
430 Using tab characters will almost certainly lead to future portability
431 and maintenance problems, so the default and recommendation is not to
432 use them. For those who prefer tabs, however, there are two different
435 Except for possibly introducing tab indentation characters, as outlined
436 below, perltidy does not introduce any tab characters into your file,
437 and it removes any tabs from the code (unless requested not to do so
438 with B<-fws>). If you have any tabs in your comments, quotes, or
439 here-documents, they will remain.
443 =item B<-et=n>, B<--entab-leading-whitespace>
445 This flag causes each B<n> initial space characters to be replaced by
446 one tab character. Note that the integer B<n> is completely independent
447 of the integer specified for indentation parameter, B<-i=n>.
449 =item B<-t>, B<--tabs>
451 This flag causes one leading tab character to be inserted for each level
452 of indentation. Certain other features are incompatible with this
453 option, and if these options are also given, then a warning message will
454 be issued and this flag will be unset. One example is the B<-lp>
457 =item B<-dt=n>, B<--default-tabsize=n>
459 If the first line of code passed to perltidy contains leading tabs but no
460 tab scheme is specified for the output stream then perltidy must guess how many
461 spaces correspond to each leading tab. This number of spaces B<n>
462 corresponding to each leading tab of the input stream may be specified with
463 B<-dt=n>. The default is B<n=8>.
465 This flag has no effect if a tab scheme is specified for the output stream,
466 because then the input stream is assumed to use the same tab scheme and
467 indentation spaces as for the output stream (any other assumption would lead to
472 =item B<-syn>, B<--check-syntax>
474 This flag is now ignored for safety, but the following documentation
475 has been retained for reference.
477 This flag causes perltidy to run C<perl -c -T> to check syntax of input
478 and output. (To change the flags passed to perl, see the next
479 item, B<-pscf>). The results are written to the F<.LOG> file, which
480 will be saved if an error is detected in the output script. The output
481 script is not checked if the input script has a syntax error. Perltidy
482 does its own checking, but this option employs perl to get a "second
485 If perl reports errors in the input file, they will not be reported in
486 the error output unless the B<--warning-output> flag is given.
488 The default is B<NOT> to do this type of syntax checking (although
489 perltidy will still do as much self-checking as possible). The reason
490 is that it causes all code in BEGIN blocks to be executed, for all
491 modules being used, and this opens the door to security issues and
492 infinite loops when running perltidy.
494 =item B<-pscf=s>, B<-perl-syntax-check-flags=s>
496 When perl is invoked to check syntax, the normal flags are C<-c -T>. In
497 addition, if the B<-x> flag is given to perltidy, then perl will also be
498 passed a B<-x> flag. It should not normally be necessary to change
499 these flags, but it can be done with the B<-pscf=s> flag. For example,
500 if the taint flag, C<-T>, is not wanted, the flag could be set to be just
503 Perltidy will pass your string to perl with the exception that it will
504 add a B<-c> and B<-x> if appropriate. The F<.LOG> file will show
505 exactly what flags were passed to perl.
507 =item B<-xs>, B<--extended-syntax>
509 A problem with formatting Perl code is that some modules can introduce new
510 syntax. This flag allows perltidy to handle certain common extensions
511 to the standard syntax without complaint.
513 For example, without this flag a structure such as the following would generate
514 a syntax error and the braces would not be balanced:
516 method deposit( Num $amount) {
517 $self->balance( $self->balance + $amount );
520 This flag is enabled by default but it can be deactivated with B<-nxs>.
521 Probably the only reason to deactivate this flag is to generate more diagnostic
522 messages when debugging a script.
525 =item B<-io>, B<--indent-only>
527 This flag is used to deactivate all whitespace and line break changes
528 within non-blank lines of code.
529 When it is in effect, the only change to the script will be
530 to the indentation and to the number of blank lines.
531 And any flags controlling whitespace and newlines will be ignored. You
532 might want to use this if you are perfectly happy with your whitespace
533 and line breaks, and merely want perltidy to handle the indentation.
534 (This also speeds up perltidy by well over a factor of two, so it might be
535 useful when perltidy is merely being used to help find a brace error in
538 Setting this flag is equivalent to setting B<--freeze-newlines> and
539 B<--freeze-whitespace>.
541 If you also want to keep your existing blank lines exactly
542 as they are, you can add B<--freeze-blank-lines>.
544 With this option perltidy is still free to modify the indenting (and
545 outdenting) of code and comments as it normally would. If you also want to
546 prevent long comment lines from being outdented, you can add either B<-noll> or
549 Setting this flag will prevent perltidy from doing any special operations on
550 closing side comments. You may still delete all side comments however when
551 this flag is in effect.
554 =item B<-enc=s>, B<--character-encoding=s>
556 where B<s>=B<none> or B<utf8>. This flag tells perltidy the character encoding
557 of both the input and output character streams. The value B<utf8> causes the
558 stream to be read and written as UTF-8. The value B<none> causes the stream to
559 be processed without special encoding assumptions. At present there is no
560 automatic detection of character encoding (even if there is a C<'use utf8'>
561 statement in your code) so this flag must be set for streams encoded in UTF-8.
562 Incorrectly setting this parameter can cause data corruption, so please
563 carefully check the output.
565 The default is B<none>.
567 The abbreviations B<-utf8> or B<-UTF8> are equivalent to B<-enc=utf8>.
568 So to process a file named B<file.pl> which is encoded in UTF-8 you can use:
570 perltidy -utf8 file.pl
572 =item B<-ole=s>, B<--output-line-ending=s>
574 where s=C<win>, C<dos>, C<unix>, or C<mac>. This flag tells perltidy
575 to output line endings for a specific system. Normally,
576 perltidy writes files with the line separator character of the host
577 system. The C<win> and C<dos> flags have an identical result.
579 =item B<-ple>, B<--preserve-line-endings>
581 This flag tells perltidy to write its output files with the same line
582 endings as the input file, if possible. It should work for
583 B<dos>, B<unix>, and B<mac> line endings. It will only work if perltidy
584 input comes from a filename (rather than stdin, for example). If
585 perltidy has trouble determining the input file line ending, it will
586 revert to the default behavior of using the line ending of the host system.
588 =item B<-it=n>, B<--iterations=n>
590 This flag causes perltidy to do B<n> complete iterations. The reason for this
591 flag is that code beautification is an iterative process and in some
592 cases the output from perltidy can be different if it is applied a second time.
593 For most purposes the default of B<n=1> should be satisfactory. However B<n=2>
594 can be useful when a major style change is being made, or when code is being
595 beautified on check-in to a source code control system. It has been found to
596 be extremely rare for the output to change after 2 iterations. If a value
597 B<n> is greater than 2 is input then a convergence test will be used to stop
598 the iterations as soon as possible, almost always after 2 iterations. See
599 the next item for a simplified iteration control.
601 This flag has no effect when perltidy is used to generate html.
603 =item B<-conv>, B<--converge>
605 This flag is equivalent to B<-it=4> and is included to simplify iteration
606 control. For all practical purposes one either does or does not want to be
607 sure that the output is converged, and there is no penalty to using a large
608 iteration limit since perltidy will check for convergence and stop iterating as
609 soon as possible. The default is B<-nconv> (no convergence check). Using
610 B<-conv> will approximately double run time since normally one extra iteration
611 is required to verify convergence.
615 =head2 Code Indentation Control
619 =item B<-ci=n>, B<--continuation-indentation=n>
621 Continuation indentation is extra indentation spaces applied when
622 a long line is broken. The default is n=2, illustrated here:
625 ( $max_index_to_go >= 0 ) ? $levels_to_go[0] : $last_output_level;
627 The same example, with n=0, is a little harder to read:
630 ( $max_index_to_go >= 0 ) ? $levels_to_go[0] : $last_output_level;
632 The value given to B<-ci> is also used by some commands when a small
633 space is required. Examples are commands for outdenting labels,
634 B<-ola>, and control keywords, B<-okw>.
636 When default values are not used, it is suggested that the value B<n>
637 given with B<-ci=n> be no more than about one-half of the number of
638 spaces assigned to a full indentation level on the B<-i=n> command.
640 =item B<-sil=n> B<--starting-indentation-level=n>
642 By default, perltidy examines the input file and tries to determine the
643 starting indentation level. While it is often zero, it may not be
644 zero for a code snippet being sent from an editing session.
646 To guess the starting indentation level perltidy simply assumes that
647 indentation scheme used to create the code snippet is the same as is being used
648 for the current perltidy process. This is the only sensible guess that can be
649 made. It should be correct if this is true, but otherwise it probably won't.
650 For example, if the input script was written with -i=2 and the current peltidy
651 flags have -i=4, the wrong initial indentation will be guessed for a code
652 snippet which has non-zero initial indentation. Likewise, if an entabbing
653 scheme is used in the input script and not in the current process then the
654 guessed indentation will be wrong.
656 If the default method does not work correctly, or you want to change the
657 starting level, use B<-sil=n>, to force the starting level to be n.
659 =item List indentation using B<-lp>, B<--line-up-parentheses>
661 By default, perltidy indents lists with 4 spaces, or whatever value
662 is specified with B<-i=n>. Here is a small list formatted in this way:
666 'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun',
667 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec'
670 Use the B<-lp> flag to add extra indentation to cause the data to begin
671 past the opening parentheses of a sub call or list, or opening square
672 bracket of an anonymous array, or opening curly brace of an anonymous
673 hash. With this option, the above list would become:
677 'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun',
678 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec'
681 If the available line length (see B<-l=n> ) does not permit this much
682 space, perltidy will use less. For alternate placement of the
683 closing paren, see the next section.
685 This option has no effect on code BLOCKS, such as if/then/else blocks,
686 which always use whatever is specified with B<-i=n>. Also, the
687 existence of line breaks and/or block comments between the opening and
688 closing parens may cause perltidy to temporarily revert to its default
691 Note: The B<-lp> option may not be used together with the B<-t> tabs option.
692 It may, however, be used with the B<-et=n> tab method.
694 In addition, any parameter which significantly restricts the ability of
695 perltidy to choose newlines will conflict with B<-lp> and will cause
696 B<-lp> to be deactivated. These include B<-io>, B<-fnl>, B<-nanl>, and
697 B<-ndnl>. The reason is that the B<-lp> indentation style can require
698 the careful coordination of an arbitrary number of break points in
699 hierarchical lists, and these flags may prevent that.
701 =item B<-cti=n>, B<--closing-token-indentation>
703 The B<-cti=n> flag controls the indentation of a line beginning with
704 a C<)>, C<]>, or a non-block C<}>. Such a line receives:
706 -cti = 0 no extra indentation (default)
707 -cti = 1 extra indentation such that the closing token
708 aligns with its opening token.
709 -cti = 2 one extra indentation level if the line looks like:
711 -cti = 3 one extra indentation level always
713 The flags B<-cti=1> and B<-cti=2> work well with the B<-lp> flag (previous
716 # perltidy -lp -cti=1
718 'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun',
719 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec'
722 # perltidy -lp -cti=2
724 'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun',
725 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec'
728 These flags are merely hints to the formatter and they may not always be
729 followed. In particular, if -lp is not being used, the indentation for
730 B<cti=1> is constrained to be no more than one indentation level.
732 If desired, this control can be applied independently to each of the
733 closing container token types. In fact, B<-cti=n> is merely an
734 abbreviation for B<-cpi=n -csbi=n -cbi=n>, where:
735 B<-cpi> or B<--closing-paren-indentation> controls B<)>'s,
736 B<-csbi> or B<--closing-square-bracket-indentation> controls B<]>'s,
737 B<-cbi> or B<--closing-brace-indentation> controls non-block B<}>'s.
739 =item B<-icp>, B<--indent-closing-paren>
741 The B<-icp> flag is equivalent to
742 B<-cti=2>, described in the previous section. The B<-nicp> flag is
743 equivalent B<-cti=0>. They are included for backwards compatibility.
745 =item B<-icb>, B<--indent-closing-brace>
747 The B<-icb> option gives one extra level of indentation to a brace which
748 terminates a code block . For example,
757 The default is not to do this, indicated by B<-nicb>.
759 =item B<-olq>, B<--outdent-long-quotes>
761 When B<-olq> is set, lines which is a quoted string longer than the
762 value B<maximum-line-length> will have their indentation removed to make
763 them more readable. This is the default. To prevent such out-denting,
764 use B<-nolq> or B<--nooutdent-long-lines>.
766 =item B<-oll>, B<--outdent-long-lines>
768 This command is equivalent to B<--outdent-long-quotes> and
769 B<--outdent-long-comments>, and it is included for compatibility with previous
770 versions of perltidy. The negation of this also works, B<-noll> or
771 B<--nooutdent-long-lines>, and is equivalent to setting B<-nolq> and B<-nolc>.
773 =item Outdenting Labels: B<-ola>, B<--outdent-labels>
775 This command will cause labels to be outdented by 2 spaces (or whatever B<-ci>
776 has been set to), if possible. This is the default. For example:
779 LOOP: while ( $i = <FOTOS> ) {
785 Use B<-nola> to not outdent labels.
787 =item Outdenting Keywords
791 =item B<-okw>, B<--outdent-keywords>
793 The command B<-okw> will cause certain leading control keywords to
794 be outdented by 2 spaces (or whatever B<-ci> has been set to), if
795 possible. By default, these keywords are C<redo>, C<next>, C<last>,
796 C<goto>, and C<return>. The intention is to make these control keywords
797 easier to see. To change this list of keywords being outdented, see
800 For example, using C<perltidy -okw> on the previous example gives:
803 LOOP: while ( $i = <FOTOS> ) {
809 The default is not to do this.
811 =item Specifying Outdented Keywords: B<-okwl=string>, B<--outdent-keyword-list=string>
813 This command can be used to change the keywords which are outdented with
814 the B<-okw> command. The parameter B<string> is a required list of perl
815 keywords, which should be placed in quotes if there are more than one.
816 By itself, it does not cause any outdenting to occur, so the B<-okw>
817 command is still required.
819 For example, the commands C<-okwl="next last redo goto" -okw> will cause
820 those four keywords to be outdented. It is probably simplest to place
821 any B<-okwl> command in a F<.perltidyrc> file.
827 =head2 Whitespace Control
829 Whitespace refers to the blank space between variables, operators,
830 and other code tokens.
834 =item B<-fws>, B<--freeze-whitespace>
836 This flag causes your original whitespace to remain unchanged, and
837 causes the rest of the whitespace commands in this section, the
838 Code Indentation section, and
839 the Comment Control section to be ignored.
841 =item Tightness of curly braces, parentheses, and square brackets.
843 Here the term "tightness" will mean the closeness with which
844 pairs of enclosing tokens, such as parentheses, contain the quantities
845 within. A numerical value of 0, 1, or 2 defines the tightness, with
846 0 being least tight and 2 being most tight. Spaces within containers
847 are always symmetric, so if there is a space after a C<(> then there
848 will be a space before the corresponding C<)>.
850 The B<-pt=n> or B<--paren-tightness=n> parameter controls the space within
851 parens. The example below shows the effect of the three possible
854 if ( ( my $len_tab = length( $tabstr ) ) > 0 ) { # -pt=0
855 if ( ( my $len_tab = length($tabstr) ) > 0 ) { # -pt=1 (default)
856 if ((my $len_tab = length($tabstr)) > 0) { # -pt=2
858 When n is 0, there is always a space to the right of a '(' and to the left
859 of a ')'. For n=2 there is never a space. For n=1, the default, there
860 is a space unless the quantity within the parens is a single token, such
861 as an identifier or quoted string.
863 Likewise, the parameter B<-sbt=n> or B<--square-bracket-tightness=n>
864 controls the space within square brackets, as illustrated below.
866 $width = $col[ $j + $k ] - $col[ $j ]; # -sbt=0
867 $width = $col[ $j + $k ] - $col[$j]; # -sbt=1 (default)
868 $width = $col[$j + $k] - $col[$j]; # -sbt=2
870 Curly braces which do not contain code blocks are controlled by
871 the parameter B<-bt=n> or B<--brace-tightness=n>.
873 $obj->{ $parsed_sql->{ 'table' }[0] }; # -bt=0
874 $obj->{ $parsed_sql->{'table'}[0] }; # -bt=1 (default)
875 $obj->{$parsed_sql->{'table'}[0]}; # -bt=2
877 And finally, curly braces which contain blocks of code are controlled by the
878 parameter B<-bbt=n> or B<--block-brace-tightness=n> as illustrated in the
881 %bf = map { $_ => -M $_ } grep { /\.deb$/ } dirents '.'; # -bbt=0 (default)
882 %bf = map { $_ => -M $_ } grep {/\.deb$/} dirents '.'; # -bbt=1
883 %bf = map {$_ => -M $_} grep {/\.deb$/} dirents '.'; # -bbt=2
885 To simplify input in the case that all of the tightness flags have the same
886 value <n>, the parameter <-act=n> or B<--all-containers-tightness=n> is an
887 abbreviation for the combination <-pt=n -sbt=n -bt=n -bbt=n>.
890 =item B<-tso>, B<--tight-secret-operators>
892 The flag B<-tso> causes certain perl token sequences (secret operators)
893 which might be considered to be a single operator to be formatted "tightly"
894 (without spaces). The operators currently modified by this flag are:
896 0+ +0 ()x!! ~~<> ,=> =( )=
898 For example the sequence B<0 +>, which converts a string to a number,
899 would be formatted without a space: B<0+> when the B<-tso> flag is set. This
900 flag is off by default.
902 =item B<-sts>, B<--space-terminal-semicolon>
904 Some programmers prefer a space before all terminal semicolons. The
905 default is for no such space, and is indicated with B<-nsts> or
906 B<--nospace-terminal-semicolon>.
909 $i = 1; # -nsts (default)
911 =item B<-sfs>, B<--space-for-semicolon>
913 Semicolons within B<for> loops may sometimes be hard to see,
914 particularly when commas are also present. This option places spaces on
915 both sides of these special semicolons, and is the default. Use
916 B<-nsfs> or B<--nospace-for-semicolon> to deactivate it.
918 for ( @a = @$ap, $u = shift @a ; @a ; $u = $v ) { # -sfs (default)
919 for ( @a = @$ap, $u = shift @a; @a; $u = $v ) { # -nsfs
921 =item B<-asc>, B<--add-semicolons>
923 Setting B<-asc> allows perltidy to add any missing optional semicolon at the end
924 of a line which is followed by a closing curly brace on the next line. This
925 is the default, and may be deactivated with B<-nasc> or B<--noadd-semicolons>.
927 =item B<-dsm>, B<--delete-semicolons>
929 Setting B<-dsm> allows perltidy to delete extra semicolons which are
930 simply empty statements. This is the default, and may be deactivated
931 with B<-ndsm> or B<--nodelete-semicolons>. (Such semicolons are not
932 deleted, however, if they would promote a side comment to a block
935 =item B<-aws>, B<--add-whitespace>
937 Setting this option allows perltidy to add certain whitespace improve
938 code readability. This is the default. If you do not want any
939 whitespace added, but are willing to have some whitespace deleted, use
940 B<-naws>. (Use B<-fws> to leave whitespace completely unchanged).
942 =item B<-dws>, B<--delete-old-whitespace>
944 Setting this option allows perltidy to remove some old whitespace
945 between characters, if necessary. This is the default. If you
946 do not want any old whitespace removed, use B<-ndws> or
947 B<--nodelete-old-whitespace>.
949 =item Detailed whitespace controls around tokens
951 For those who want more detailed control over the whitespace around
952 tokens, there are four parameters which can directly modify the default
953 whitespace rules built into perltidy for any token. They are:
955 B<-wls=s> or B<--want-left-space=s>,
957 B<-nwls=s> or B<--nowant-left-space=s>,
959 B<-wrs=s> or B<--want-right-space=s>,
961 B<-nwrs=s> or B<--nowant-right-space=s>.
963 These parameters are each followed by a quoted string, B<s>, containing a
964 list of token types. No more than one of each of these parameters
965 should be specified, because repeating a command-line parameter
966 always overwrites the previous one before perltidy ever sees it.
968 To illustrate how these are used, suppose it is desired that there be no
969 space on either side of the token types B<= + - / *>. The following two
970 parameters would specify this desire:
972 -nwls="= + - / *" -nwrs="= + - / *"
974 (Note that the token types are in quotes, and that they are separated by
975 spaces). With these modified whitespace rules, the following line of math:
977 $root = -$b + sqrt( $b * $b - 4. * $a * $c ) / ( 2. * $a );
981 $root=-$b+sqrt( $b*$b-4.*$a*$c )/( 2.*$a );
983 These parameters should be considered to be hints to perltidy rather
984 than fixed rules, because perltidy must try to resolve conflicts that
985 arise between them and all of the other rules that it uses. One
986 conflict that can arise is if, between two tokens, the left token wants
987 a space and the right one doesn't. In this case, the token not wanting
988 a space takes priority.
990 It is necessary to have a list of all token types in order to create
991 this type of input. Such a list can be obtained by the command
992 B<--dump-token-types>. Also try the B<-D> flag on a short snippet of code
993 and look at the .DEBUG file to see the tokenization.
995 B<WARNING> Be sure to put these tokens in quotes to avoid having them
996 misinterpreted by your command shell.
998 =item Space between specific keywords and opening paren
1000 When an opening paren follows a Perl keyword, no space is introduced after the
1001 keyword, unless it is (by default) one of these:
1003 my local our and or eq ne if else elsif until unless
1004 while for foreach return switch case given when
1006 These defaults can be modified with two commands:
1008 B<-sak=s> or B<--space-after-keyword=s> adds keywords.
1010 B<-nsak=s> or B<--nospace-after-keyword=s> removes keywords.
1012 where B<s> is a list of keywords (in quotes if necessary). For example,
1014 my ( $a, $b, $c ) = @_; # default
1015 my( $a, $b, $c ) = @_; # -nsak="my local our"
1017 The abbreviation B<-nsak='*'> is equivalent to including all of the
1018 keywords in the above list.
1020 When both B<-nsak=s> and B<-sak=s> commands are included, the B<-nsak=s>
1021 command is executed first. For example, to have space after only the
1022 keywords (my, local, our) you could use B<-nsak="*" -sak="my local our">.
1024 To put a space after all keywords, see the next item.
1026 =item Space between all keywords and opening parens
1028 When an opening paren follows a function or keyword, no space is introduced
1029 after the keyword except for the keywords noted in the previous item. To
1030 always put a space between a function or keyword and its opening paren,
1033 B<-skp> or B<--space-keyword-paren>
1035 You will probably also want to use the flag B<-sfp> (next item) too.
1037 =item Space between all function names and opening parens
1039 When an opening paren follows a function the default is not to introduce
1040 a space. To cause a space to be introduced use:
1042 B<-sfp> or B<--space-function-paren>
1044 myfunc( $a, $b, $c ); # default
1045 myfunc ( $a, $b, $c ); # -sfp
1047 You will probably also want to use the flag B<-skp> (previous item) too.
1049 =item Trimming whitespace around C<qw> quotes
1051 B<-tqw> or B<--trim-qw> provide the default behavior of trimming
1052 spaces around multi-line C<qw> quotes and indenting them appropriately.
1054 B<-ntqw> or B<--notrim-qw> cause leading and trailing whitespace around
1055 multi-line C<qw> quotes to be left unchanged. This option will not
1056 normally be necessary, but was added for testing purposes, because in
1057 some versions of perl, trimming C<qw> quotes changes the syntax tree.
1059 =item B<-sbq=n> or B<--space-backslash-quote=n>
1066 can confuse syntax highlighters unless a space is included between the backslash and the single or double quotation mark.
1068 This can be controlled with the value of B<n> as follows:
1070 -sbq=0 means no space between the backslash and quote
1071 -sbq=1 means follow the example of the source code
1072 -sbq=2 means always put a space between the backslash and quote
1074 The default is B<-sbq=1>, meaning that a space will be used 0if there is one in the source code.
1076 =item Trimming trailing whitespace from lines of POD
1078 B<-trp> or B<--trim-pod> will remove trailing whitespace from lines of POD.
1079 The default is not to do this.
1083 =head2 Comment Controls
1085 Perltidy has a number of ways to control the appearance of both block comments
1086 and side comments. The term B<block comment> here refers to a full-line
1087 comment, whereas B<side comment> will refer to a comment which appears on a
1088 line to the right of some code.
1092 =item B<-ibc>, B<--indent-block-comments>
1094 Block comments normally look best when they are indented to the same
1095 level as the code which follows them. This is the default behavior, but
1096 you may use B<-nibc> to keep block comments left-justified. Here is an
1099 # this comment is indented (-ibc, default)
1100 if ($task) { yyy(); }
1102 The alternative is B<-nibc>:
1104 # this comment is not indented (-nibc)
1105 if ($task) { yyy(); }
1107 See also the next item, B<-isbc>, as well as B<-sbc>, for other ways to
1108 have some indented and some outdented block comments.
1110 =item B<-isbc>, B<--indent-spaced-block-comments>
1112 If there is no leading space on the line, then the comment will not be
1113 indented, and otherwise it may be.
1115 If both B<-ibc> and B<-isbc> are set, then B<-isbc> takes priority.
1117 =item B<-olc>, B<--outdent-long-comments>
1119 When B<-olc> is set, lines which are full-line (block) comments longer
1120 than the value B<maximum-line-length> will have their indentation
1121 removed. This is the default; use B<-nolc> to prevent outdenting.
1123 =item B<-msc=n>, B<--minimum-space-to-comment=n>
1125 Side comments look best when lined up several spaces to the right of
1126 code. Perltidy will try to keep comments at least n spaces to the
1127 right. The default is n=4 spaces.
1129 =item B<-fpsc=n>, B<--fixed-position-side-comment=n>
1131 This parameter tells perltidy to line up side comments in column number B<n>
1132 whenever possible. The default, n=0, will not do this.
1134 =item B<-iscl>, B<--ignore-side-comment-lengths>
1136 This parameter causes perltidy to ignore the length of side comments when
1137 setting line breaks. The default, B<-niscl>, is to include the length of
1138 side comments when breaking lines to stay within the length prescribed
1139 by the B<-l=n> maximum line length parameter. For example, the following
1140 long single line would remain intact with -l=80 and -iscl:
1142 perltidy -l=80 -iscl
1143 $vmsfile =~ s/;[\d\-]*$//; # Clip off version number; we can use a newer version as well
1145 whereas without the -iscl flag the line will be broken:
1148 $vmsfile =~ s/;[\d\-]*$//
1149 ; # Clip off version number; we can use a newer version as well
1152 =item B<-hsc>, B<--hanging-side-comments>
1154 By default, perltidy tries to identify and align "hanging side
1155 comments", which are something like this:
1157 my $IGNORE = 0; # This is a side comment
1158 # This is a hanging side comment
1161 A comment is considered to be a hanging side comment if (1) it immediately
1162 follows a line with a side comment, or another hanging side comment, and
1163 (2) there is some leading whitespace on the line.
1164 To deactivate this feature, use B<-nhsc> or B<--nohanging-side-comments>.
1165 If block comments are preceded by a blank line, or have no leading
1166 whitespace, they will not be mistaken as hanging side comments.
1168 =item Closing Side Comments
1170 A closing side comment is a special comment which perltidy can
1171 automatically create and place after the closing brace of a code block.
1172 They can be useful for code maintenance and debugging. The command
1173 B<-csc> (or B<--closing-side-comments>) adds or updates closing side
1174 comments. For example, here is a small code snippet
1177 if ( !defined( $_[0] ) ) {
1178 print("Hello, World\n");
1181 print( $_[0], "\n" );
1185 And here is the result of processing with C<perltidy -csc>:
1188 if ( !defined( $_[0] ) ) {
1189 print("Hello, World\n");
1192 print( $_[0], "\n" );
1194 } ## end sub message
1196 A closing side comment was added for C<sub message> in this case, but not
1197 for the C<if> and C<else> blocks, because they were below the 6 line
1198 cutoff limit for adding closing side comments. This limit may be
1199 changed with the B<-csci> command, described below.
1201 The command B<-dcsc> (or B<--delete-closing-side-comments>) reverses this
1202 process and removes these comments.
1204 Several commands are available to modify the behavior of these two basic
1205 commands, B<-csc> and B<-dcsc>:
1209 =item B<-csci=n>, or B<--closing-side-comment-interval=n>
1211 where C<n> is the minimum number of lines that a block must have in
1212 order for a closing side comment to be added. The default value is
1213 C<n=6>. To illustrate:
1215 # perltidy -csci=2 -csc
1217 if ( !defined( $_[0] ) ) {
1218 print("Hello, World\n");
1219 } ## end if ( !defined( $_[0] ))
1221 print( $_[0], "\n" );
1222 } ## end else [ if ( !defined( $_[0] ))
1223 } ## end sub message
1225 Now the C<if> and C<else> blocks are commented. However, now this has
1226 become very cluttered.
1228 =item B<-cscp=string>, or B<--closing-side-comment-prefix=string>
1230 where string is the prefix used before the name of the block type. The
1231 default prefix, shown above, is C<## end>. This string will be added to
1232 closing side comments, and it will also be used to recognize them in
1233 order to update, delete, and format them. Any comment identified as a
1234 closing side comment will be placed just a single space to the right of
1237 =item B<-cscl=string>, or B<--closing-side-comment-list>
1239 where C<string> is a list of block types to be tagged with closing side
1240 comments. By default, all code block types preceded by a keyword or
1241 label (such as C<if>, C<sub>, and so on) will be tagged. The B<-cscl>
1242 command changes the default list to be any selected block types; see
1243 L<Specifying Block Types>.
1244 For example, the following command
1245 requests that only C<sub>'s, labels, C<BEGIN>, and C<END> blocks be
1246 affected by any B<-csc> or B<-dcsc> operation:
1248 -cscl="sub : BEGIN END"
1250 =item B<-csct=n>, or B<--closing-side-comment-maximum-text=n>
1252 The text appended to certain block types, such as an C<if> block, is
1253 whatever lies between the keyword introducing the block, such as C<if>,
1254 and the opening brace. Since this might be too much text for a side
1255 comment, there needs to be a limit, and that is the purpose of this
1256 parameter. The default value is C<n=20>, meaning that no additional
1257 tokens will be appended to this text after its length reaches 20
1258 characters. Omitted text is indicated with C<...>. (Tokens, including
1259 sub names, are never truncated, however, so actual lengths may exceed
1260 this). To illustrate, in the above example, the appended text of the
1261 first block is C< ( !defined( $_[0] )...>. The existing limit of
1262 C<n=20> caused this text to be truncated, as indicated by the C<...>. See
1263 the next flag for additional control of the abbreviated text.
1265 =item B<-cscb>, or B<--closing-side-comments-balanced>
1267 As discussed in the previous item, when the
1268 closing-side-comment-maximum-text limit is exceeded the comment text must
1269 be truncated. Older versions of perltidy terminated with three dots, and this
1270 can still be achieved with -ncscb:
1272 perltidy -csc -ncscb
1273 } ## end foreach my $foo (sort { $b cmp $a ...
1275 However this causes a problem with editors which cannot recognize
1276 comments or are not configured to do so because they cannot "bounce" around in
1277 the text correctly. The B<-cscb> flag has been added to
1278 help them by appending appropriate balancing structure:
1281 } ## end foreach my $foo (sort { $b cmp $a ... })
1283 The default is B<-cscb>.
1285 =item B<-csce=n>, or B<--closing-side-comment-else-flag=n>
1287 The default, B<n=0>, places the text of the opening C<if> statement after any
1290 If B<n=2> is used, then each C<elsif> is also given the text of the opening
1291 C<if> statement. Also, an C<else> will include the text of a preceding
1292 C<elsif> statement. Note that this may result some long closing
1295 If B<n=1> is used, the results will be the same as B<n=2> whenever the
1296 resulting line length is less than the maximum allowed.
1298 =item B<-cscb>, or B<--closing-side-comments-balanced>
1300 When using closing-side-comments, and the closing-side-comment-maximum-text
1301 limit is exceeded, then the comment text must be abbreviated.
1302 It is terminated with three dots if the B<-cscb> flag is negated:
1304 perltidy -csc -ncscb
1305 } ## end foreach my $foo (sort { $b cmp $a ...
1307 This causes a problem with older editors which do not recognize comments
1308 because they cannot "bounce" around in the text correctly. The B<-cscb>
1309 flag tries to help them by appending appropriate terminal balancing structures:
1312 } ## end foreach my $foo (sort { $b cmp $a ... })
1314 The default is B<-cscb>.
1317 =item B<-cscw>, or B<--closing-side-comment-warnings>
1319 This parameter is intended to help make the initial transition to the use of
1320 closing side comments.
1322 things to happen if a closing side comment replaces an existing, different
1323 closing side comment: first, an error message will be issued, and second, the
1324 original side comment will be placed alone on a new specially marked comment
1325 line for later attention.
1327 The intent is to avoid clobbering existing hand-written side comments
1328 which happen to match the pattern of closing side comments. This flag
1329 should only be needed on the first run with B<-csc>.
1333 B<Important Notes on Closing Side Comments:>
1339 Closing side comments are only placed on lines terminated with a closing
1340 brace. Certain closing styles, such as the use of cuddled elses
1341 (B<-ce>), preclude the generation of some closing side comments.
1345 Please note that adding or deleting of closing side comments takes
1346 place only through the commands B<-csc> or B<-dcsc>. The other commands,
1347 if used, merely modify the behavior of these two commands.
1351 It is recommended that the B<-cscw> flag be used along with B<-csc> on
1352 the first use of perltidy on a given file. This will prevent loss of
1353 any existing side comment data which happens to have the csc prefix.
1357 Once you use B<-csc>, you should continue to use it so that any
1358 closing side comments remain correct as code changes. Otherwise, these
1359 comments will become incorrect as the code is updated.
1363 If you edit the closing side comments generated by perltidy, you must also
1364 change the prefix to be different from the closing side comment prefix.
1365 Otherwise, your edits will be lost when you rerun perltidy with B<-csc>. For
1366 example, you could simply change C<## end> to be C<## End>, since the test is
1367 case sensitive. You may also want to use the B<-ssc> flag to keep these
1368 modified closing side comments spaced the same as actual closing side comments.
1372 Temporarily generating closing side comments is a useful technique for
1373 exploring and/or debugging a perl script, especially one written by someone
1374 else. You can always remove them with B<-dcsc>.
1378 =item Static Block Comments
1380 Static block comments are block comments with a special leading pattern,
1381 C<##> by default, which will be treated slightly differently from other
1382 block comments. They effectively behave as if they had glue along their
1383 left and top edges, because they stick to the left edge and previous line
1384 when there is no blank spaces in those places. This option is
1385 particularly useful for controlling how commented code is displayed.
1389 =item B<-sbc>, B<--static-block-comments>
1391 When B<-sbc> is used, a block comment with a special leading pattern, C<##> by
1392 default, will be treated specially.
1394 Comments so identified are treated as follows:
1400 If there is no leading space on the line, then the comment will not
1401 be indented, and otherwise it may be,
1405 no new blank line will be
1406 inserted before such a comment, and
1410 such a comment will never become
1411 a hanging side comment.
1415 For example, assuming C<@month_of_year> is
1418 @month_of_year = ( # -sbc (default)
1419 'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct',
1423 Without this convention, the above code would become
1425 @month_of_year = ( # -nsbc
1426 'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct',
1432 which is not as clear.
1433 The default is to use B<-sbc>. This may be deactivated with B<-nsbc>.
1435 =item B<-sbcp=string>, B<--static-block-comment-prefix=string>
1437 This parameter defines the prefix used to identify static block comments
1438 when the B<-sbc> parameter is set. The default prefix is C<##>,
1439 corresponding to C<-sbcp=##>. The prefix is actually part of a perl
1440 pattern used to match lines and it must either begin with C<#> or C<^#>.
1441 In the first case a prefix ^\s* will be added to match any leading
1442 whitespace, while in the second case the pattern will match only
1443 comments with no leading whitespace. For example, to
1444 identify all comments as static block comments, one would use C<-sbcp=#>.
1445 To identify all left-adjusted comments as static block comments, use C<-sbcp='^#'>.
1447 Please note that B<-sbcp> merely defines the pattern used to identify static
1448 block comments; it will not be used unless the switch B<-sbc> is set. Also,
1449 please be aware that since this string is used in a perl regular expression
1450 which identifies these comments, it must enable a valid regular expression to
1453 A pattern which can be useful is:
1457 This pattern requires a static block comment to have at least one character
1458 which is neither a # nor a space. It allows a line containing only '#'
1459 characters to be rejected as a static block comment. Such lines are often used
1460 at the start and end of header information in subroutines and should not be
1461 separated from the intervening comments, which typically begin with just a
1464 =item B<-osbc>, B<--outdent-static-block-comments>
1466 The command B<-osbc> will cause static block comments to be outdented by 2
1467 spaces (or whatever B<-ci=n> has been set to), if possible.
1471 =item Static Side Comments
1473 Static side comments are side comments with a special leading pattern.
1474 This option can be useful for controlling how commented code is displayed
1475 when it is a side comment.
1479 =item B<-ssc>, B<--static-side-comments>
1481 When B<-ssc> is used, a side comment with a static leading pattern, which is
1482 C<##> by default, will be spaced only a single space from previous
1483 character, and it will not be vertically aligned with other side comments.
1485 The default is B<-nssc>.
1487 =item B<-sscp=string>, B<--static-side-comment-prefix=string>
1489 This parameter defines the prefix used to identify static side comments
1490 when the B<-ssc> parameter is set. The default prefix is C<##>,
1491 corresponding to C<-sscp=##>.
1493 Please note that B<-sscp> merely defines the pattern used to identify
1494 static side comments; it will not be used unless the switch B<-ssc> is
1495 set. Also, note that this string is used in a perl regular expression
1496 which identifies these comments, so it must enable a valid regular
1497 expression to be formed.
1504 =head2 Skipping Selected Sections of Code
1506 Selected lines of code may be passed verbatim to the output without any
1507 formatting. This feature is enabled by default but can be disabled with
1508 the B<--noformat-skipping> or B<-nfs> flag. It should be used sparingly to
1509 avoid littering code with markers, but it might be helpful for working
1510 around occasional problems. For example it might be useful for keeping
1511 the indentation of old commented code unchanged, keeping indentation of
1512 long blocks of aligned comments unchanged, keeping certain list
1513 formatting unchanged, or working around a glitch in perltidy.
1517 =item B<-fs>, B<--format-skipping>
1519 This flag, which is enabled by default, causes any code between
1520 special beginning and ending comment markers to be passed to the
1521 output without formatting. The default beginning marker is #<<<
1522 and the default ending marker is #>>> but they
1523 may be changed (see next items below). Additional text may appear on
1524 these special comment lines provided that it is separated from the
1525 marker by at least one space. For example
1527 #<<< do not let perltidy touch this
1535 The comment markers may be placed at any location that a block comment may
1536 appear. If they do not appear to be working, use the -log flag and examine the
1537 F<.LOG> file. Use B<-nfs> to disable this feature.
1539 =item B<-fsb=string>, B<--format-skipping-begin=string>
1541 The B<-fsb=string> parameter may be used to change the beginning marker for
1542 format skipping. The default is equivalent to -fsb='#<<<'. The string that
1543 you enter must begin with a # and should be in quotes as necessary to get past
1544 the command shell of your system. It is actually the leading text of a pattern
1545 that is constructed by appending a '\s', so you must also include backslashes
1546 for characters to be taken literally rather than as patterns.
1548 Some examples show how example strings become patterns:
1550 -fsb='#\{\{\{' becomes /^#\{\{\{\s/ which matches #{{{ but not #{{{{
1551 -fsb='#\*\*' becomes /^#\*\*\s/ which matches #** but not #***
1552 -fsb='#\*{2,}' becomes /^#\*{2,}\s/ which matches #** and #*****
1554 =item B<-fse=string>, B<--format-skipping-end=string>
1556 The B<-fsb=string> is the corresponding parameter used to change the
1557 ending marker for format skipping. The default is equivalent to
1562 =head2 Line Break Control
1564 The parameters in this section control breaks after
1565 non-blank lines of code. Blank lines are controlled
1566 separately by parameters in the section L<Blank Line
1571 =item B<-fnl>, B<--freeze-newlines>
1573 If you do not want any changes to the line breaks within
1574 lines of code in your script, set
1575 B<-fnl>, and they will remain fixed, and the rest of the commands in
1576 this section and sections
1577 L<Controlling List Formatting>,
1578 L<Retaining or Ignoring Existing Line Breaks>.
1579 You may want to use B<-noll> with this.
1581 Note: If you also want to keep your blank lines exactly
1582 as they are, you can use the B<-fbl> flag which is described
1583 in the section L<Blank Line Control>.
1585 =item B<-ce>, B<--cuddled-else>
1587 Enable the "cuddled else" style, in which C<else> and C<elsif> are
1588 follow immediately after the curly brace closing the previous block.
1589 The default is not to use cuddled elses, and is indicated with the flag
1590 B<-nce> or B<--nocuddled-else>. Here is a comparison of the
1602 else { # -nce (default)
1606 =item B<-cb>, B<--cuddled-blocks>
1608 This flag enables the "cuddled else" format style on a chain of specified block
1609 types. The default is to apply it to a chain consisting of try-catch-finally
1610 blocks, but it can apply to any desired chain of blocks by specifying their
1611 names on a separate parameter B<-cbl>, described in the next section.
1615 throw Error::Simple( "ok 2\n", 2 );
1616 } catch Error::Simple with {
1623 Cuddling between a pair of code blocks requires that the closing brace of the
1624 first block start a new line. If this block is entirely on one line in the
1625 input file, it is necessary to decide if it should be broken to allow cuddling.
1626 This decision is controlled by the flag B<-cbo=n> discussed below. The default
1627 and recommended value of B<-cbo=1> bases this decision on the first block in
1628 the chain. If it spans multiple lines then cuddling is made and continues
1629 along the chain, regardless of the sizes of subsequent blocks. Otherwise, short
1630 lines remain intact.
1632 So for example, the B<-cb> flag would not have any effect if the above snippet
1635 try { throw Error::Simple( "ok 2\n", 2 ); }
1636 catch Error::Simple with { my $err = shift; print "$err"; }
1637 finally { print "ok 3\n"; };
1639 If the first block spans multiple lines, then cuddling can be done and will
1640 continue for the subsequent blocks in the chain, as illustrated in the previous
1643 If there are blank lines between cuddled blocks they will be eliminated. If
1644 there are comments after the closing brace where cuddling would occur then
1645 cuddling will be prevented. If this occurs, cuddling will restart later in the
1648 The default for this parameter is B<--nocuddled-blocks>
1650 =item B<-cbl>, B<--cuddled-block-list>
1652 The block types to which the B<-cuddled-blocks> style applies is defined by
1653 this parameter. This parameter is a character string, giving a list of
1654 block types separated by dashes.
1656 The default value for this string is
1658 -cbl="try-catch-finally"
1660 This string will cause cuddled formatting to be applied to every block in a chain
1661 starting with a "try" and followed by any number of "catch" and "finally"
1664 In general, a string describing a chain of blocks has the form
1666 -cbl="word1-word2-word3-...-wordn"
1668 In this case, a chain begins when an opening block brace preceded by word1 in
1669 the list is encountered. The chain continues if the closing block brace is
1670 followed immediately by any of word2, word3, etc.
1672 If the leading word, word1, might be repeated later in a chain then it should
1673 also be included amoung the secondary words.
1675 Multiple chain types may be specified by separating the strings with commas or
1676 spaces. So for example if we have two chains of code blocks, f1-f2-f3 and g1-g2-g3-g4,
1677 they could be specified as
1679 -cbl="f1-f2-f3 g1-g2-g3-g4"
1681 -cbl=f1-f2-f3,g1-g2-g3-g4
1683 Spaces are easier to read but commas may avoid quotation difficulties when
1684 entering data in a command shell.
1686 To define secondary words that apply to all block types, other than those explicitly specified,
1687 the leading word can be omitted. For example, the built-in cuddled-else format specified by
1688 the B<-ce> flag can be approximately specified by
1690 -cbl="if-else-elsif unless-else-elsif -continue"
1692 -cbl=if-else-elsif,unless-else-elsif,-continue
1694 The final string -continue allows cuddling the optional continue block
1695 which may follow may other block types.
1697 As a diagnostic check, the flag B<--dump-cuddled-block-list> or B<-dcbl> can be
1698 used to view the hash of values this flag creates.
1700 Finally, note that the B<-cbl> flag by itself merely specifies which blocks are formatted
1701 with the cuddled format. It has no effect unless this formatting style is activated with
1705 =item B<-cbo=n>, B<--cuddled-break-option=n>
1707 Cuddled formatting is only possible between a pair of code blocks if the
1708 closing brace of the first block starts a new line. If a block is encountered
1709 which is entirely on a single line, and cuddled formatting is selected, it is
1710 necessary to make a decision as to whether or not to "break" the block, meaning
1711 to cause it to span multiple lines. This parameter controls that decision. The
1714 cbo=0 Never force a short block to break.
1715 cbo=1 If the first of a pair of blocks is broken in the input file,
1716 then break the second.
1717 cbo=2 Break open all blocks for maximal cuddled formatting.
1719 The default and recommended value is B<cbo=1>. With this value, if the starting
1720 block of a chain spans multiple lines, then a cascade of breaks will occur for
1721 remaining blocks causing the entire chain to be cuddled.
1723 The option B<cbo=0> can produce erratic cuddling if there are numerous one-line
1726 The option B<cbo=2> produces maximal cuddling but will not allow any short blocks.
1728 Note: at present, this option currently only applies to blocks controlled by
1729 the B<-cb> flag. Cuddling under the B<-ce> flag corresponds approximately to
1730 B<-cbo=1> but cannot currently be changed.
1733 =item B<-bl>, B<--opening-brace-on-new-line>
1735 Use the flag B<-bl> to place the opening brace on a new line:
1737 if ( $input_file eq '-' ) # -bl
1739 important_function();
1742 This flag applies to all structural blocks, including named sub's (unless
1743 the B<-sbl> flag is set -- see next item).
1745 The default style, B<-nbl>, places an opening brace on the same line as
1746 the keyword introducing it. For example,
1748 if ( $input_file eq '-' ) { # -nbl (default)
1750 =item B<-sbl>, B<--opening-sub-brace-on-new-line>
1752 The flag B<-sbl> can be used to override the value of B<-bl> for
1753 the opening braces of named sub's. For example,
1757 produces this result:
1761 if (!defined($_[0])) {
1762 print("Hello, World\n");
1769 This flag is negated with B<-nsbl>. If B<-sbl> is not specified,
1770 the value of B<-bl> is used.
1772 =item B<-asbl>, B<--opening-anonymous-sub-brace-on-new-line>
1774 The flag B<-asbl> is like the B<-sbl> flag except that it applies
1775 to anonymous sub's instead of named subs. For example
1779 produces this result:
1783 if ( !defined( $_[0] ) ) {
1784 print("Hello, World\n");
1787 print( $_[0], "\n" );
1791 This flag is negated with B<-nasbl>, and the default is B<-nasbl>.
1793 =item B<-bli>, B<--brace-left-and-indent>
1795 The flag B<-bli> is the same as B<-bl> but in addition it causes one
1796 unit of continuation indentation ( see B<-ci> ) to be placed before
1797 an opening and closing block braces.
1801 if ( $input_file eq '-' ) # -bli
1803 important_function();
1806 By default, this extra indentation occurs for blocks of type:
1807 B<if>, B<elsif>, B<else>, B<unless>, B<for>, B<foreach>, B<sub>,
1808 B<while>, B<until>, and also with a preceding label. The next item
1809 shows how to change this.
1811 =item B<-blil=s>, B<--brace-left-and-indent-list=s>
1813 Use this parameter to change the types of block braces for which the
1814 B<-bli> flag applies; see L<Specifying Block Types>. For example,
1815 B<-blil='if elsif else'> would apply it to only C<if/elsif/else> blocks.
1817 =item B<-bar>, B<--opening-brace-always-on-right>
1819 The default style, B<-nbl> places the opening code block brace on a new
1820 line if it does not fit on the same line as the opening keyword, like
1823 if ( $bigwasteofspace1 && $bigwasteofspace2
1824 || $bigwasteofspace3 && $bigwasteofspace4 )
1826 big_waste_of_time();
1829 To force the opening brace to always be on the right, use the B<-bar>
1830 flag. In this case, the above example becomes
1832 if ( $bigwasteofspace1 && $bigwasteofspace2
1833 || $bigwasteofspace3 && $bigwasteofspace4 ) {
1834 big_waste_of_time();
1837 A conflict occurs if both B<-bl> and B<-bar> are specified.
1839 =item B<-otr>, B<--opening-token-right> and related flags
1841 The B<-otr> flag is a hint that perltidy should not place a break between a
1842 comma and an opening token. For example:
1844 # default formatting
1845 push @{ $self->{$module}{$key} },
1847 accno => $ref->{accno},
1848 description => $ref->{description}
1852 push @{ $self->{$module}{$key} }, {
1853 accno => $ref->{accno},
1854 description => $ref->{description}
1857 The flag B<-otr> is actually an abbreviation for three other flags
1858 which can be used to control parens, hash braces, and square brackets
1859 separately if desired:
1861 -opr or --opening-paren-right
1862 -ohbr or --opening-hash-brace-right
1863 -osbr or --opening-square-bracket-right
1865 =item B<-wn>, B<--weld-nested-containers>
1867 The B<-wn> flag causes closely nested pairs of opening and closing container
1868 symbols (curly braces, brackets, or parens) to be "welded" together, meaning
1869 that they are treated as if combined into a single unit, with the indentation
1870 of the innermost code reduced to be as if there were just a single container
1875 # default formatting
1885 } } until $x++ > $z;
1887 When this flag is set perltidy makes a preliminary pass through the file and
1888 identifies all nested pairs of containers. To qualify as a nested pair, the
1889 closing container symbols must be immediately adjacent. The opening symbols
1890 must either be adjacent, or, if the outer opening symbol is an opening
1891 paren, they may be separated by any single non-container symbol or something
1892 that looks like a function evaluation.
1894 Any container symbol may serve as both the inner container of one pair and as
1895 the outer container of an adjacent pair. Consequently, any number of adjacent
1896 opening or closing symbols may join together in weld. For example, here are
1897 three levels of wrapped function calls:
1899 # default formatting
1900 my (@date_time) = Localtime(
1903 $year, $month, $day, $hour, $minute, $second,
1904 '0', $offset, '0', '0'
1910 my (@date_time) = Localtime( Date_to_Time( Add_Delta_DHMS(
1911 $year, $month, $day, $hour, $minute, $second,
1912 '0', $offset, '0', '0'
1915 Notice how the indentation of the inner lines are reduced by two levels in this
1916 case. This example also shows the typical result of this formatting, namely it
1917 is a sandwich consisting of an initial opening layer, a central section of any
1918 complexity forming the "meat" of the sandwich, and a final closing layer. This
1919 predictable structure helps keep the compacted structure readable.
1921 The inner sandwich layer is required to be at least one line thick. If this
1922 cannot be achieved, welding does not occur. This constraint can cause
1923 formatting to take a couple of iterations to stabilize when it is first applied
1924 to a script. The B<-conv> flag can be used to insure that the final format is
1925 achieved in a single run.
1927 Here is an example illustrating a welded container within a welded containers:
1929 # default formatting
1934 $sx * int( $xr->numify() ) & $sy * int( $yr->numify() )
1944 $sx * int( $xr->numify() ) & $sy * int( $yr->numify() )
1949 This format option is quite general but there are some limitations.
1951 One limitiation is that any line length limit still applies and can cause long
1952 welded sections to be broken into multiple lines.
1954 Another limitation is that an opening symbol which delimits quoted text cannot
1955 be included in a welded pair. This is because quote delimiters are treated
1956 specially in perltidy.
1958 Finally, the stacking of containers defined by this flag have priority over
1959 any other container stacking flags. This is because any welding is done first.
1962 =item B<Vertical tightness> of non-block curly braces, parentheses, and square brackets.
1964 These parameters control what shall be called vertical tightness. Here are the
1971 Opening tokens (except for block braces) are controlled by B<-vt=n>, or
1972 B<--vertical-tightness=n>, where
1974 -vt=0 always break a line after opening token (default).
1975 -vt=1 do not break unless this would produce more than one
1976 step in indentation in a line.
1977 -vt=2 never break a line after opening token
1981 You must also use the B<-lp> flag when you use the B<-vt> flag; the
1982 reason is explained below.
1986 Closing tokens (except for block braces) are controlled by B<-vtc=n>, or
1987 B<--vertical-tightness-closing=n>, where
1989 -vtc=0 always break a line before a closing token (default),
1990 -vtc=1 do not break before a closing token which is followed
1991 by a semicolon or another closing token, and is not in
1993 -vtc=2 never break before a closing token.
1995 The rules for B<-vtc=1> are designed to maintain a reasonable balance
1996 between tightness and readability in complex lists.
2000 Different controls may be applied to different token types,
2001 and it is also possible to control block braces; see below.
2005 Finally, please note that these vertical tightness flags are merely
2006 hints to the formatter, and it cannot always follow them. Things which
2007 make it difficult or impossible include comments, blank lines, blocks of
2008 code within a list, and possibly the lack of the B<-lp> parameter.
2009 Also, these flags may be ignored for very small lists (2 or 3 lines in
2014 Here are some examples:
2016 # perltidy -lp -vt=0 -vtc=0
2024 # perltidy -lp -vt=1 -vtc=0
2025 %romanNumerals = ( one => 'I',
2031 # perltidy -lp -vt=1 -vtc=1
2032 %romanNumerals = ( one => 'I',
2037 The difference between B<-vt=1> and B<-vt=2> is shown here:
2039 # perltidy -lp -vt=1
2041 mysprintf( "(void)find_threadsv(%s);",
2042 cstring( $threadsv_names[ $op->targ ] )
2046 # perltidy -lp -vt=2
2047 $init->add( mysprintf( "(void)find_threadsv(%s);",
2048 cstring( $threadsv_names[ $op->targ ] )
2052 With B<-vt=1>, the line ending in C<add(> does not combine with the next
2053 line because the next line is not balanced. This can help with
2054 readability, but B<-vt=2> can be used to ignore this rule.
2056 The tightest, and least readable, code is produced with both C<-vt=2> and
2059 # perltidy -lp -vt=2 -vtc=2
2060 $init->add( mysprintf( "(void)find_threadsv(%s);",
2061 cstring( $threadsv_names[ $op->targ ] ) ) );
2063 Notice how the code in all of these examples collapses vertically as
2064 B<-vt> increases, but the indentation remains unchanged. This is
2065 because perltidy implements the B<-vt> parameter by first formatting as
2066 if B<-vt=0>, and then simply overwriting one output line on top of the
2067 next, if possible, to achieve the desired vertical tightness. The
2068 B<-lp> indentation style has been designed to allow this vertical
2069 collapse to occur, which is why it is required for the B<-vt> parameter.
2071 The B<-vt=n> and B<-vtc=n> parameters apply to each type of container
2072 token. If desired, vertical tightness controls can be applied
2073 independently to each of the closing container token types.
2075 The parameters for controlling parentheses are B<-pvt=n> or
2076 B<--paren-vertical-tightness=n>, and B<-pcvt=n> or
2077 B<--paren-vertical-tightness-closing=n>.
2079 Likewise, the parameters for square brackets are B<-sbvt=n> or
2080 B<--square-bracket-vertical-tightness=n>, and B<-sbcvt=n> or
2081 B<--square-bracket-vertical-tightness-closing=n>.
2083 Finally, the parameters for controlling non-code block braces are
2084 B<-bvt=n> or B<--brace-vertical-tightness=n>, and B<-bcvt=n> or
2085 B<--brace-vertical-tightness-closing=n>.
2087 In fact, the parameter B<-vt=n> is actually just an abbreviation for
2088 B<-pvt=n -bvt=n sbvt=n>, and likewise B<-vtc=n> is an abbreviation
2089 for B<-pvtc=n -bvtc=n sbvtc=n>.
2091 =item B<-bbvt=n> or B<--block-brace-vertical-tightness=n>
2093 The B<-bbvt=n> flag is just like the B<-vt=n> flag but applies
2094 to opening code block braces.
2096 -bbvt=0 break after opening block brace (default).
2097 -bbvt=1 do not break unless this would produce more than one
2098 step in indentation in a line.
2099 -bbvt=2 do not break after opening block brace.
2101 It is necessary to also use either B<-bl> or B<-bli> for this to work,
2102 because, as with other vertical tightness controls, it is implemented by
2103 simply overwriting a line ending with an opening block brace with the
2104 subsequent line. For example:
2106 # perltidy -bli -bbvt=0
2107 if ( open( FILE, "< $File" ) )
2109 while ( $File = <FILE> )
2117 # perltidy -bli -bbvt=1
2118 if ( open( FILE, "< $File" ) )
2119 { while ( $File = <FILE> )
2126 By default this applies to blocks associated with keywords B<if>,
2127 B<elsif>, B<else>, B<unless>, B<for>, B<foreach>, B<sub>, B<while>,
2128 B<until>, and also with a preceding label. This can be changed with
2129 the parameter B<-bbvtl=string>, or
2130 B<--block-brace-vertical-tightness-list=string>, where B<string> is a
2131 space-separated list of block types. For more information on the
2132 possible values of this string, see L<Specifying Block Types>
2134 For example, if we want to just apply this style to C<if>,
2135 C<elsif>, and C<else> blocks, we could use
2136 C<perltidy -bli -bbvt=1 -bbvtl='if elsif else'>.
2138 There is no vertical tightness control for closing block braces; with
2139 one exception they will be placed on separate lines.
2140 The exception is that a cascade of closing block braces may
2141 be stacked on a single line. See B<-scbb>.
2143 =item B<-sot>, B<--stack-opening-tokens> and related flags
2145 The B<-sot> flag tells perltidy to "stack" opening tokens
2146 when possible to avoid lines with isolated opening tokens.
2151 $opt_c = Text::CSV_XS->new(
2160 $opt_c = Text::CSV_XS->new( {
2167 For detailed control of individual closing tokens the following
2168 controls can be used:
2170 -sop or --stack-opening-paren
2171 -sohb or --stack-opening-hash-brace
2172 -sosb or --stack-opening-square-bracket
2173 -sobb or --stack-opening-block-brace
2175 The flag B<-sot> is an abbreviation for B<-sop -sohb -sosb>.
2177 The flag B<-sobb> is a abbreviation for B<-bbvt=2 -bbvtl='*'>. This
2178 will case a cascade of opening block braces to appear on a single line,
2179 although this an uncommon occurrence except in test scripts.
2181 =item B<-sct>, B<--stack-closing-tokens> and related flags
2183 The B<-sct> flag tells perltidy to "stack" closing tokens
2184 when possible to avoid lines with isolated closing tokens.
2189 $opt_c = Text::CSV_XS->new(
2198 $opt_c = Text::CSV_XS->new(
2205 The B<-sct> flag is somewhat similar to the B<-vtc> flags, and in some
2206 cases it can give a similar result. The difference is that the B<-vtc>
2207 flags try to avoid lines with leading opening tokens by "hiding" them at
2208 the end of a previous line, whereas the B<-sct> flag merely tries to
2209 reduce the number of lines with isolated closing tokens by stacking them
2210 but does not try to hide them. For example:
2213 $opt_c = Text::CSV_XS->new(
2217 always_quote => 1, } );
2219 For detailed control of the stacking of individual closing tokens the
2220 following controls can be used:
2222 -scp or --stack-closing-paren
2223 -schb or --stack-closing-hash-brace
2224 -scsb or --stack-closing-square-bracket
2225 -scbb or --stack-closing-block-brace
2227 The flag B<-sct> is an abbreviation for stacking the non-block closing
2228 tokens, B<-scp -schb -scsb>.
2230 Stacking of closing block braces, B<-scbb>, causes a cascade of isolated
2231 closing block braces to be combined into a single line as in the following
2239 push( @lines, "$w1 $w2 $w3 $w4\n" );
2242 To simplify input even further for the case in which both opening and closing
2243 non-block containers are stacked, the flag B<-sac> or B<--stack-all-containers>
2244 is an abbreviation for B<-sot -sot>.
2246 =item B<-dnl>, B<--delete-old-newlines>
2248 By default, perltidy first deletes all old line break locations, and then it
2249 looks for good break points to match the desired line length. Use B<-ndnl>
2250 or B<--nodelete-old-newlines> to force perltidy to retain all old line break
2253 =item B<-anl>, B<--add-newlines>
2255 By default, perltidy will add line breaks when necessary to create
2256 continuations of long lines and to improve the script appearance. Use
2257 B<-nanl> or B<--noadd-newlines> to prevent any new line breaks.
2259 This flag does not prevent perltidy from eliminating existing line
2260 breaks; see B<--freeze-newlines> to completely prevent changes to line
2263 =item Controlling whether perltidy breaks before or after operators
2265 Four command line parameters provide some control over whether
2266 a line break should be before or after specific token types.
2267 Two parameters give detailed control:
2269 B<-wba=s> or B<--want-break-after=s>, and
2271 B<-wbb=s> or B<--want-break-before=s>.
2273 These parameters are each followed by a quoted string, B<s>, containing
2274 a list of token types (separated only by spaces). No more than one of each
2275 of these parameters should be specified, because repeating a
2276 command-line parameter always overwrites the previous one before
2277 perltidy ever sees it.
2279 By default, perltidy breaks B<after> these token types:
2280 % + - * / x != == >= <= =~ !~ < > | &
2281 = **= += *= &= <<= &&= -= /= |= >>= ||= //= .= %= ^= x=
2283 And perltidy breaks B<before> these token types by default:
2286 To illustrate, to cause a break after a concatenation operator, C<'.'>,
2287 rather than before it, the command line would be
2291 As another example, the following command would cause a break before
2292 math operators C<'+'>, C<'-'>, C<'/'>, and C<'*'>:
2296 These commands should work well for most of the token types that perltidy uses
2297 (use B<--dump-token-types> for a list). Also try the B<-D> flag on a short
2298 snippet of code and look at the .DEBUG file to see the tokenization. However,
2299 for a few token types there may be conflicts with hardwired logic which cause
2300 unexpected results. One example is curly braces, which should be controlled
2301 with the parameter B<bl> provided for that purpose.
2303 B<WARNING> Be sure to put these tokens in quotes to avoid having them
2304 misinterpreted by your command shell.
2306 Two additional parameters are available which, though they provide no further
2307 capability, can simplify input are:
2309 B<-baao> or B<--break-after-all-operators>,
2311 B<-bbao> or B<--break-before-all-operators>.
2313 The -baao sets the default to be to break after all of the following operators:
2315 % + - * / x != == >= <= =~ !~ < > | &
2316 = **= += *= &= <<= &&= -= /= |= >>= ||= //= .= %= ^= x=
2317 . : ? && || and or err xor
2319 and the B<-bbao> flag sets the default to break before all of these operators.
2320 These can be used to define an initial break preference which can be fine-tuned
2321 with the B<-wba> and B<-wbb> flags. For example, to break before all operators
2322 except an B<=> one could use --bbao -wba='=' rather than listing every
2323 single perl operator except B<=> on a -wbb flag.
2327 =head2 Controlling List Formatting
2329 Perltidy attempts to place comma-separated arrays of values in tables
2330 which look good. Its default algorithms usually work well, and they
2331 have been improving with each release, but several parameters are
2332 available to control list formatting.
2336 =item B<-boc>, B<--break-at-old-comma-breakpoints>
2338 This flag tells perltidy to try to break at all old commas. This is not
2339 the default. Normally, perltidy makes a best guess at list formatting,
2340 and seldom uses old comma breakpoints. Usually this works well,
2349 The default formatting will flatten this down to one line:
2351 # perltidy (default)
2352 my @list = ( 1, 1, 1, 1, 2, 1, 1, 3, 3, 1, 1, 4, 6, 4, 1, );
2354 which hides the structure. Using B<-boc>, plus additional flags
2355 to retain the original style, yields
2357 # perltidy -boc -lp -pt=2 -vt=1 -vtc=1
2364 A disadvantage of this flag is that all tables in the file
2365 must already be nicely formatted. For another possibility see
2366 the -fs flag in L<Skipping Selected Sections of Code>.
2368 =item B<-mft=n>, B<--maximum-fields-per-table=n>
2370 If the computed number of fields for any table exceeds B<n>, then it
2371 will be reduced to B<n>. The default value for B<n> is a large number,
2372 40. While this value should probably be left unchanged as a general
2373 rule, it might be used on a small section of code to force a list to
2374 have a particular number of fields per line, and then either the B<-boc>
2375 flag could be used to retain this formatting, or a single comment could
2376 be introduced somewhere to freeze the formatting in future applications
2389 =item B<-cab=n>, B<--comma-arrow-breakpoints=n>
2391 A comma which follows a comma arrow, '=>', is given special
2392 consideration. In a long list, it is common to break at all such
2393 commas. This parameter can be used to control how perltidy breaks at
2394 these commas. (However, it will have no effect if old comma breaks are
2395 being forced because B<-boc> is used). The possible values of B<n> are:
2397 n=0 break at all commas after =>
2398 n=1 stable: break at all commas after => if container is open,
2399 EXCEPT FOR one-line containers
2400 n=2 break at all commas after =>, BUT try to form the maximum
2401 maximum one-line container lengths
2402 n=3 do not treat commas after => specially at all
2403 n=4 break everything: like n=0 but ALSO break a short container with
2404 a => not followed by a comma when -vt=0 is used
2405 n=5 stable: like n=1 but ALSO break at open one-line containers when
2406 -vt=0 is used (default)
2408 For example, given the following single line, perltidy by default will
2409 not add any line breaks because it would break the existing one-line
2412 bless { B => $B, Root => $Root } => $package;
2414 Using B<-cab=0> will force a break after each comma-arrow item:
2422 If perltidy is subsequently run with this container broken, then by
2423 default it will break after each '=>' because the container is now
2424 broken. To reform a one-line container, the parameter B<-cab=2> could
2427 The flag B<-cab=3> can be used to prevent these commas from being
2428 treated specially. In this case, an item such as "01" => 31 is
2429 treated as a single item in a table. The number of fields in this table
2430 will be determined by the same rules that are used for any other table.
2435 "01" => 31, "02" => 29, "03" => 31, "04" => 30,
2436 "05" => 31, "06" => 30, "07" => 31, "08" => 31,
2437 "09" => 30, "10" => 31, "11" => 30, "12" => 31
2442 =head2 Retaining or Ignoring Existing Line Breaks
2444 Several additional parameters are available for controlling the extent
2445 to which line breaks in the input script influence the output script.
2446 In most cases, the default parameter values are set so that, if a choice
2447 is possible, the output style follows the input style. For example, if
2448 a short logical container is broken in the input script, then the
2449 default behavior is for it to remain broken in the output script.
2451 Most of the parameters in this section would only be required for a
2452 one-time conversion of a script from short container lengths to longer
2453 container lengths. The opposite effect, of converting long container
2454 lengths to shorter lengths, can be obtained by temporarily using a short
2455 maximum line length.
2459 =item B<-bol>, B<--break-at-old-logical-breakpoints>
2461 By default, if a logical expression is broken at a C<&&>, C<||>, C<and>,
2462 or C<or>, then the container will remain broken. Also, breaks
2463 at internal keywords C<if> and C<unless> will normally be retained.
2464 To prevent this, and thus form longer lines, use B<-nbol>.
2466 =item B<-bok>, B<--break-at-old-keyword-breakpoints>
2468 By default, perltidy will retain a breakpoint before keywords which may
2469 return lists, such as C<sort> and <map>. This allows chains of these
2470 operators to be displayed one per line. Use B<-nbok> to prevent
2471 retaining these breakpoints.
2473 =item B<-bot>, B<--break-at-old-ternary-breakpoints>
2475 By default, if a conditional (ternary) operator is broken at a C<:>,
2476 then it will remain broken. To prevent this, and thereby
2477 form longer lines, use B<-nbot>.
2479 =item B<-boa>, B<--break-at-old-attribute-breakpoints>
2481 By default, if an attribute list is broken at a C<:> in the source file, then
2482 it will remain broken. For example, given the following code, the line breaks
2483 at the ':'s will be retained:
2488 : Get('Name' => 'foo') : Set('Name');
2490 If the attributes are on a single line in the source code then they will remain
2491 on a single line if possible.
2493 To prevent this, and thereby always form longer lines, use B<-nboa>.
2495 =item B<-iob>, B<--ignore-old-breakpoints>
2497 Use this flag to tell perltidy to ignore existing line breaks to the
2498 maximum extent possible. This will tend to produce the longest possible
2499 containers, regardless of type, which do not exceed the line length
2502 =item B<-kis>, B<--keep-interior-semicolons>
2504 Use the B<-kis> flag to prevent breaking at a semicolon if
2505 there was no break there in the input file. Normally
2506 perltidy places a newline after each semicolon which
2507 terminates a statement unless several statements are
2508 contained within a one-line brace block. To illustrate,
2509 consider the following input lines:
2511 dbmclose(%verb_delim); undef %verb_delim;
2512 dbmclose(%expanded); undef %expanded;
2514 The default is to break after each statement, giving
2516 dbmclose(%verb_delim);
2518 dbmclose(%expanded);
2521 With B<perltidy -kis> the multiple statements are retained:
2523 dbmclose(%verb_delim); undef %verb_delim;
2524 dbmclose(%expanded); undef %expanded;
2526 The statements are still subject to the specified value
2527 of B<maximum-line-length> and will be broken if this
2528 maximum is exceeded.
2532 =head2 Blank Line Control
2534 Blank lines can improve the readability of a script if they are carefully
2535 placed. Perltidy has several commands for controlling the insertion,
2536 retention, and removal of blank lines.
2540 =item B<-fbl>, B<--freeze-blank-lines>
2542 Set B<-fbl> if you want to the blank lines in your script to
2543 remain exactly as they are. The rest of the parameters in
2544 this section may then be ignored. (Note: setting the B<-fbl> flag
2545 is equivalent to setting B<-mbl=0> and B<-kbl=2>).
2547 =item B<-bbc>, B<--blanks-before-comments>
2549 A blank line will be introduced before a full-line comment. This is the
2550 default. Use B<-nbbc> or B<--noblanks-before-comments> to prevent
2551 such blank lines from being introduced.
2553 =item B<-blbs=n>, B<--blank-lines-before-subs=n>
2555 The parameter B<-blbs=n> requests that least B<n> blank lines precede a sub
2556 definition which does not follow a comment and which is more than one-line
2557 long. The default is <-blbs=1>. B<BEGIN> and B<END> blocks are included.
2559 The requested number of blanks statement will be inserted regardless of the
2560 value of B<--maximum-consecutive-blank-lines=n> (B<-mbl=n>) with the exception
2561 that if B<-mbl=0> then no blanks will be output.
2563 This parameter interacts with the value B<k> of the parameter B<--maximum-consecutive-blank-lines=k> (B<-mbl=k>) as follows:
2565 1. If B<-mbl=0> then no blanks will be output. This allows all blanks to be suppressed with a single parameter. Otherwise,
2567 2. If the number of old blank lines in the script is less than B<n> then
2568 additional blanks will be inserted to make the total B<n> regardless of the
2571 3. If the number of old blank lines in the script equals or exceeds B<n> then
2572 this parameter has no effect, however the total will not exceed
2573 value specified on the B<-mbl=k> flag.
2576 =item B<-blbp=n>, B<--blank-lines-before-packages=n>
2578 The parameter B<-blbp=n> requests that least B<n> blank lines precede a package
2579 which does not follow a comment. The default is B<-blbp=1>.
2581 This parameter interacts with the value B<k> of the parameter
2582 B<--maximum-consecutive-blank-lines=k> (B<-mbl=k>) in the same way as described
2583 for the previous item B<-blbs=n>.
2586 =item B<-bbs>, B<--blanks-before-subs>
2588 For compatibility with previous versions, B<-bbs> or B<--blanks-before-subs>
2589 is equivalent to F<-blbp=1> and F<-blbs=1>.
2591 Likewise, B<-nbbs> or B<--noblanks-before-subs>
2592 is equivalent to F<-blbp=0> and F<-blbs=0>.
2594 =item B<-bbb>, B<--blanks-before-blocks>
2596 A blank line will be introduced before blocks of coding delimited by
2597 B<for>, B<foreach>, B<while>, B<until>, and B<if>, B<unless>, in the following
2604 The block is not preceded by a comment.
2608 The block is not a one-line block.
2612 The number of consecutive non-blank lines at the current indentation depth is at least B<-lbl>
2617 This is the default. The intention of this option is to introduce
2618 some space within dense coding.
2619 This is negated with B<-nbbb> or B<--noblanks-before-blocks>.
2621 =item B<-lbl=n> B<--long-block-line-count=n>
2623 This controls how often perltidy is allowed to add blank lines before
2624 certain block types (see previous section). The default is 8. Entering
2625 a value of B<0> is equivalent to entering a very large number.
2627 =item B<-blao=i> or B<--blank-lines-after-opening-block=i>
2629 This control places a minimum of B<i> blank lines B<after> a line which B<ends>
2630 with an opening block brace of a specified type. By default, this only applies
2631 to the block of a named B<sub>, but this can be changed (see B<-blaol> below).
2632 The default is not to do this (B<i=0>).
2634 Please see the note below on using the B<-blao> and B<-blbc> options.
2636 =item B<-blbc=i> or B<--blank-lines-before-closing-block=i>
2638 This control places a minimum of B<i> blank lines B<before> a line which
2639 B<begins> with a closing block brace of a specified type. By default, this
2640 only applies to the block of a named B<sub>, but this can be changed (see
2641 B<-blbcl> below). The default is not to do this (B<i=0>).
2643 =item B<-blaol=s> or B<--blank-lines-after-opening-block-list=s>
2645 The parameter B<s> is a list of block type keywords to which the flag B<-blao>
2646 should apply. The section L<"Specifying Block Types"> explains how to list
2649 =item B<-blbcl=s> or B<--blank-lines-before-closing-block-list=s>
2651 This parameter is a list of block type keywords to which the flag B<-blbc>
2652 should apply. The section L<"Specifying Block Types"> explains how to list
2655 =item Note on using the B<-blao> and B<-blbc> options.
2657 These blank line controls introduce a certain minimum number of blank lines in
2658 the text, but the final number of blank lines may be greater, depending on
2659 values of the other blank line controls and the number of old blank lines. A
2660 consequence is that introducing blank lines with these and other controls
2661 cannot be exactly undone, so some experimentation with these controls is
2662 recommended before using them.
2664 For example, suppose that for some reason we decide to introduce one blank
2665 space at the beginning and ending of all blocks. We could do
2668 perltidy -blao=2 -blbc=2 -blaol='*' -blbcl='*' filename
2670 Now suppose the script continues to be developed, but at some later date we
2671 decide we don't want these spaces after all. we might expect that running with
2672 the flags B<-blao=0> and B<-blbc=0> will undo them. However, by default
2673 perltidy retains single blank lines, so the blank lines remain.
2675 We can easily fix this by telling perltidy to ignore old blank lines by
2676 including the added parameter B<-kbl=0> and rerunning. Then the unwanted blank
2677 lines will be gone. However, this will cause all old blank lines to be
2678 ignored, perhaps even some that were added by hand to improve formatting. So
2679 please be cautious when using these parameters.
2681 =item B<-mbl=n> B<--maximum-consecutive-blank-lines=n>
2683 This parameter specifies the maximum number of consecutive blank lines which
2684 will be output within code sections of a script. The default is n=1. If the
2685 input file has more than n consecutive blank lines, the number will be reduced
2686 to n except as noted above for the B<-blbp> and B<-blbs> parameters. If B<n=0>
2687 then no blank lines will be output (unless all old blank lines are retained
2688 with the B<-kbl=2> flag of the next section).
2690 This flag obviously does not apply to pod sections,
2691 here-documents, and quotes.
2693 =item B<-kbl=n>, B<--keep-old-blank-lines=n>
2695 The B<-kbl=n> flag gives you control over how your existing blank lines are
2698 The possible values of B<n> are:
2700 n=0 ignore all old blank lines
2701 n=1 stable: keep old blanks, but limited by the value of the B<-mbl=n> flag
2702 n=2 keep all old blank lines, regardless of the value of the B<-mbl=n> flag
2704 The default is B<n=1>.
2706 =item B<-sob>, B<--swallow-optional-blank-lines>
2708 This is equivalent to B<kbl=0> and is included for compatibility with
2711 =item B<-nsob>, B<--noswallow-optional-blank-lines>
2713 This is equivalent to B<kbl=1> and is included for compatibility with
2720 A style refers to a convenient collection of existing parameters.
2724 =item B<-gnu>, B<--gnu-style>
2726 B<-gnu> gives an approximation to the GNU Coding Standards (which do
2727 not apply to perl) as they are sometimes implemented. At present, this
2728 style overrides the default style with the following parameters:
2730 -lp -bl -noll -pt=2 -bt=2 -sbt=2 -icp
2732 =item B<-pbp>, B<--perl-best-practices>
2734 B<-pbp> is an abbreviation for the parameters in the book B<Perl Best Practices>
2737 -l=78 -i=4 -ci=4 -st -se -vt=2 -cti=0 -pt=1 -bt=1 -sbt=1 -bbt=1 -nsfs -nolq
2738 -wbb="% + - * / x != == >= <= =~ !~ < > | & =
2739 **= += *= &= <<= &&= -= /= |= >>= ||= //= .= %= ^= x="
2741 Please note that this parameter set includes -st and -se flags, which make
2742 perltidy act as a filter on one file only. These can be overridden by placing
2743 B<-nst> and/or B<-nse> after the -pbp parameter.
2745 Also note that the value of continuation indentation, -ci=4, is equal to the
2746 value of the full indentation, -i=4. In some complex statements perltidy will
2747 produce nicer results with -ci=2. This can be implemented by including -ci=2
2748 after the -pbp parameter. For example,
2753 : $type eq 'item' ? "the $section entry"
2754 : "the section on $section"
2758 ? ( $section ? ' in ' : '' ) . "the $page$page_ext manpage"
2759 : ' elsewhere in this document'
2762 # perltidy -pbp -ci=2
2765 : $type eq 'item' ? "the $section entry"
2766 : "the section on $section"
2770 ? ( $section ? ' in ' : '' ) . "the $page$page_ext manpage"
2771 : ' elsewhere in this document'
2776 =head2 Controlling Vertical Alignment
2778 Vertical alignment refers to lining up certain symbols in list of consecutive
2779 similar lines to improve readability. For example, the "fat commas" are
2780 aligned in the following statement:
2783 PeerAddr => join( ".", @port[ 0 .. 3 ] ),
2784 PeerPort => $port[4] * 256 + $port[5],
2788 The only explicit control on vertical alignment is to turn it off using
2789 B<-novalign>, a flag mainly intended for debugging. However, vertical
2790 alignment can be forced to stop and restart by selectively introducing blank
2791 lines. For example, a blank has been inserted in the following code
2792 to keep somewhat similar things aligned.
2795 'format' => [ 'tidy', 'html', 'user' ],
2796 'output-line-ending' => [ 'dos', 'win', 'mac', 'unix' ],
2797 'character-encoding' => [ 'none', 'utf8' ],
2799 'block-brace-tightness' => [ 0, 2 ],
2800 'brace-tightness' => [ 0, 2 ],
2801 'paren-tightness' => [ 0, 2 ],
2802 'square-bracket-tightness' => [ 0, 2 ],
2806 =head2 Other Controls
2810 =item Deleting selected text
2812 Perltidy can selectively delete comments and/or pod documentation. The
2813 command B<-dac> or B<--delete-all-comments> will delete all comments
2814 B<and> all pod documentation, leaving just code and any leading system
2817 The command B<-dp> or B<--delete-pod> will remove all pod documentation
2820 Two commands which remove comments (but not pod) are: B<-dbc> or
2821 B<--delete-block-comments> and B<-dsc> or B<--delete-side-comments>.
2822 (Hanging side comments will be deleted with block comments here.)
2824 The negatives of these commands also work, and are the defaults. When
2825 block comments are deleted, any leading 'hash-bang' will be retained.
2826 Also, if the B<-x> flag is used, any system commands before a leading
2827 hash-bang will be retained (even if they are in the form of comments).
2829 =item Writing selected text to a file
2831 When perltidy writes a formatted text file, it has the ability to also
2832 send selected text to a file with a F<.TEE> extension. This text can
2833 include comments and pod documentation.
2835 The command B<-tac> or B<--tee-all-comments> will write all comments
2836 B<and> all pod documentation.
2838 The command B<-tp> or B<--tee-pod> will write all pod documentation (but
2841 The commands which write comments (but not pod) are: B<-tbc> or
2842 B<--tee-block-comments> and B<-tsc> or B<--tee-side-comments>.
2843 (Hanging side comments will be written with block comments here.)
2845 The negatives of these commands also work, and are the defaults.
2847 =item Using a F<.perltidyrc> command file
2849 If you use perltidy frequently, you probably won't be happy until you
2850 create a F<.perltidyrc> file to avoid typing commonly-used parameters.
2851 Perltidy will first look in your current directory for a command file
2852 named F<.perltidyrc>. If it does not find one, it will continue looking
2853 for one in other standard locations.
2855 These other locations are system-dependent, and may be displayed with
2856 the command C<perltidy -dpro>. Under Unix systems, it will first look
2857 for an environment variable B<PERLTIDY>. Then it will look for a
2858 F<.perltidyrc> file in the home directory, and then for a system-wide
2859 file F</usr/local/etc/perltidyrc>, and then it will look for
2860 F</etc/perltidyrc>. Note that these last two system-wide files do not
2861 have a leading dot. Further system-dependent information will be found
2862 in the INSTALL file distributed with perltidy.
2864 Under Windows, perltidy will also search for a configuration file named perltidy.ini since Windows does not allow files with a leading period (.).
2865 Use C<perltidy -dpro> to see the possible locations for your system.
2866 An example might be F<C:\Documents and Settings\All Users\perltidy.ini>.
2868 Another option is the use of the PERLTIDY environment variable.
2869 The method for setting environment variables depends upon the version of
2870 Windows that you are using. Instructions for Windows 95 and later versions can
2873 http://www.netmanage.com/000/20021101_005_tcm21-6336.pdf
2875 Under Windows NT / 2000 / XP the PERLTIDY environment variable can be placed in
2876 either the user section or the system section. The later makes the
2877 configuration file common to all users on the machine. Be sure to enter the
2878 full path of the configuration file in the value of the environment variable.
2879 Ex. PERLTIDY=C:\Documents and Settings\perltidy.ini
2881 The configuration file is free format, and simply a list of parameters, just as
2882 they would be entered on a command line. Any number of lines may be used, with
2883 any number of parameters per line, although it may be easiest to read with one
2884 parameter per line. Comment text begins with a #, and there must
2885 also be a space before the # for side comments. It is a good idea to
2886 put complex parameters in either single or double quotes.
2888 Here is an example of a F<.perltidyrc> file:
2890 # This is a simple of a .perltidyrc configuration file
2891 # This implements a highly spaced style
2892 -se # errors to standard error output
2893 -w # show all warnings
2894 -bl # braces on new lines
2895 -pt=0 # parens not tight at all
2896 -bt=0 # braces not tight
2897 -sbt=0 # square brackets not tight
2899 The parameters in the F<.perltidyrc> file are installed first, so any
2900 parameters given on the command line will have priority over them.
2902 To avoid confusion, perltidy ignores any command in the .perltidyrc
2903 file which would cause some kind of dump and an exit. These are:
2905 -h -v -ddf -dln -dop -dsn -dtt -dwls -dwrs -ss
2907 There are several options may be helpful in debugging a F<.perltidyrc>
2914 A very helpful command is B<--dump-profile> or B<-dpro>. It writes a
2915 list of all configuration filenames tested to standard output, and
2916 if a file is found, it dumps the content to standard output before
2917 exiting. So, to find out where perltidy looks for its configuration
2918 files, and which one if any it selects, just enter
2924 It may be simplest to develop and test configuration files with
2925 alternative names, and invoke them with B<-pro=filename> on the command
2926 line. Then rename the desired file to F<.perltidyrc> when finished.
2930 The parameters in the F<.perltidyrc> file can be switched off with
2931 the B<-npro> option.
2935 The commands B<--dump-options>, B<--dump-defaults>, B<--dump-long-names>,
2936 and B<--dump-short-names>, all described below, may all be helpful.
2940 =item Creating a new abbreviation
2942 A special notation is available for use in a F<.perltidyrc> file
2943 for creating an abbreviation for a group
2944 of options. This can be used to create a
2945 shorthand for one or more styles which are frequently, but not always,
2946 used. The notation is to group the options within curly braces which
2947 are preceded by the name of the alias (without leading dashes), like this:
2954 where B<newword> is the abbreviation, and B<opt1>, etc, are existing parameters
2955 I<or other abbreviations>. The main syntax requirement is that the new
2956 abbreviation along with its opening curly brace must begin on a new line.
2957 Space before and after the curly braces is optional.
2959 specific example, the following line
2961 airy {-bl -pt=0 -bt=0 -sbt=0}
2963 could be placed in a F<.perltidyrc> file, and then invoked at will with
2965 perltidy -airy somefile.pl
2967 (Either C<-airy> or C<--airy> may be used).
2969 =item Skipping leading non-perl commands with B<-x> or B<--look-for-hash-bang>
2971 If your script has leading lines of system commands or other text which
2972 are not valid perl code, and which are separated from the start of the
2973 perl code by a "hash-bang" line, ( a line of the form C<#!...perl> ),
2974 you must use the B<-x> flag to tell perltidy not to parse and format any
2975 lines before the "hash-bang" line. This option also invokes perl with a
2976 -x flag when checking the syntax. This option was originally added to
2977 allow perltidy to parse interactive VMS scripts, but it should be used
2978 for any script which is normally invoked with C<perl -x>.
2980 =item Making a file unreadable
2982 The goal of perltidy is to improve the readability of files, but there
2983 are two commands which have the opposite effect, B<--mangle> and
2984 B<--extrude>. They are actually
2985 merely aliases for combinations of other parameters. Both of these
2986 strip all possible whitespace, but leave comments and pod documents,
2987 so that they are essentially reversible. The
2988 difference between these is that B<--mangle> puts the fewest possible
2989 line breaks in a script while B<--extrude> puts the maximum possible.
2990 Note that these options do not provided any meaningful obfuscation, because
2991 perltidy can be used to reformat the files. They were originally
2992 developed to help test the tokenization logic of perltidy, but they
2994 One use for B<--mangle> is the following:
2996 perltidy --mangle myfile.pl -st | perltidy -o myfile.pl.new
2998 This will form the maximum possible number of one-line blocks (see next
2999 section), and can sometimes help clean up a badly formatted script.
3001 A similar technique can be used with B<--extrude> instead of B<--mangle>
3002 to make the minimum number of one-line blocks.
3004 Another use for B<--mangle> is to combine it with B<-dac> to reduce
3005 the file size of a perl script.
3007 =item One-line blocks
3009 There are a few points to note regarding one-line blocks. A one-line
3010 block is something like this,
3012 if ($x > 0) { $y = 1 / $x }
3014 where the contents within the curly braces is short enough to fit
3017 With few exceptions, perltidy retains existing one-line blocks, if it
3018 is possible within the line-length constraint, but it does not attempt
3019 to form new ones. In other words, perltidy will try to follow the
3020 one-line block style of the input file.
3022 If an existing one-line block is longer than the maximum line length,
3023 however, it will be broken into multiple lines. When this happens, perltidy
3024 checks for and adds any optional terminating semicolon (unless the B<-nasc>
3025 option is used) if the block is a code block.
3027 The main exception is that perltidy will attempt to form new one-line
3028 blocks following the keywords C<map>, C<eval>, and C<sort>, because
3029 these code blocks are often small and most clearly displayed in a single
3032 One-line block rules can conflict with the cuddled-else option. When
3033 the cuddled-else option is used, perltidy retains existing one-line
3034 blocks, even if they do not obey cuddled-else formatting.
3036 Occasionally, when one-line blocks get broken because they exceed the
3037 available line length, the formatting will violate the requested brace style.
3038 If this happens, reformatting the script a second time should correct
3043 The following flags are available for debugging:
3045 B<--dump-cuddled-block-list> or B<-dcbl> will dump to standard output the
3046 internal hash of cuddled block types created by a B<-cuddled-block-list> input
3049 B<--dump-defaults> or B<-ddf> will write the default option set to standard output and quit
3051 B<--dump-profile> or B<-dpro> will write the name of the current
3052 configuration file and its contents to standard output and quit.
3054 B<--dump-options> or B<-dop> will write current option set to standard
3057 B<--dump-long-names> or B<-dln> will write all command line long names (passed
3058 to Get_options) to standard output and quit.
3060 B<--dump-short-names> or B<-dsn> will write all command line short names
3061 to standard output and quit.
3063 B<--dump-token-types> or B<-dtt> will write a list of all token types
3064 to standard output and quit.
3066 B<--dump-want-left-space> or B<-dwls> will write the hash %want_left_space
3067 to standard output and quit. See the section on controlling whitespace
3070 B<--dump-want-right-space> or B<-dwrs> will write the hash %want_right_space
3071 to standard output and quit. See the section on controlling whitespace
3074 B<--no-memoize> or B<-nmem> will turn of memoizing.
3075 Memoization can reduce run time when running perltidy repeatedly in a
3076 single process. It is on by default but can be deactivated for
3077 testing with B<-nmem>.
3079 B<--file-size-order> or B<-fso> will cause files to be processed in order of
3080 increasing size, when multiple files are being processed. This is useful
3081 during program development, when large numbers of files with varying sizes are
3082 processed, because it can reduce virtual memory usage.
3084 B<-DEBUG> will write a file with extension F<.DEBUG> for each input file
3085 showing the tokenization of all lines of code.
3087 =item Working with MakeMaker, AutoLoader and SelfLoader
3089 The first $VERSION line of a file which might be eval'd by MakeMaker
3090 is passed through unchanged except for indentation.
3091 Use B<--nopass-version-line>, or B<-npvl>, to deactivate this feature.
3093 If the AutoLoader module is used, perltidy will continue formatting
3094 code after seeing an __END__ line.
3095 Use B<--nolook-for-autoloader>, or B<-nlal>, to deactivate this feature.
3097 Likewise, if the SelfLoader module is used, perltidy will continue formatting
3098 code after seeing a __DATA__ line.
3099 Use B<--nolook-for-selfloader>, or B<-nlsl>, to deactivate this feature.
3101 =item Working around problems with older version of Perl
3103 Perltidy contains a number of rules which help avoid known subtleties
3104 and problems with older versions of perl, and these rules always
3105 take priority over whatever formatting flags have been set. For example,
3106 perltidy will usually avoid starting a new line with a bareword, because
3107 this might cause problems if C<use strict> is active.
3109 There is no way to override these rules.
3117 =item The B<-html> master switch
3119 The flag B<-html> causes perltidy to write an html file with extension
3120 F<.html>. So, for example, the following command
3122 perltidy -html somefile.pl
3124 will produce a syntax-colored html file named F<somefile.pl.html>
3125 which may be viewed with a browser.
3127 B<Please Note>: In this case, perltidy does not do any formatting to the
3128 input file, and it does not write a formatted file with extension
3129 F<.tdy>. This means that two perltidy runs are required to create a
3130 fully reformatted, html copy of a script.
3132 =item The B<-pre> flag for code snippets
3134 When the B<-pre> flag is given, only the pre-formatted section, within
3135 the <PRE> and </PRE> tags, will be output. This simplifies inclusion
3136 of the output in other files. The default is to output a complete
3139 =item The B<-nnn> flag for line numbering
3141 When the B<-nnn> flag is given, the output lines will be numbered.
3143 =item The B<-toc>, or B<--html-table-of-contents> flag
3145 By default, a table of contents to packages and subroutines will be
3146 written at the start of html output. Use B<-ntoc> to prevent this.
3147 This might be useful, for example, for a pod document which contains a
3148 number of unrelated code snippets. This flag only influences the code
3149 table of contents; it has no effect on any table of contents produced by
3150 pod2html (see next item).
3152 =item The B<-pod>, or B<--pod2html> flag
3154 There are two options for formatting pod documentation. The default is
3155 to pass the pod through the Pod::Html module (which forms the basis of
3156 the pod2html utility). Any code sections are formatted by perltidy, and
3157 the results then merged. Note: perltidy creates a temporary file when
3158 Pod::Html is used; see L<"FILES">. Also, Pod::Html creates temporary
3159 files for its cache.
3161 NOTE: Perltidy counts the number of C<=cut> lines, and either moves the
3162 pod text to the top of the html file if there is one C<=cut>, or leaves
3163 the pod text in its original order (interleaved with code) otherwise.
3165 Most of the flags accepted by pod2html may be included in the perltidy
3166 command line, and they will be passed to pod2html. In some cases,
3167 the flags have a prefix C<pod> to emphasize that they are for the
3168 pod2html, and this prefix will be removed before they are passed to
3169 pod2html. The flags which have the additional C<pod> prefix are:
3171 --[no]podheader --[no]podindex --[no]podrecurse --[no]podquiet
3172 --[no]podverbose --podflush
3174 The flags which are unchanged from their use in pod2html are:
3176 --backlink=s --cachedir=s --htmlroot=s --libpods=s --title=s
3177 --podpath=s --podroot=s
3179 where 's' is an appropriate character string. Not all of these flags are
3180 available in older versions of Pod::Html. See your Pod::Html documentation for
3183 The alternative, indicated with B<-npod>, is not to use Pod::Html, but
3184 rather to format pod text in italics (or whatever the stylesheet
3185 indicates), without special html markup. This is useful, for example,
3186 if pod is being used as an alternative way to write comments.
3188 =item The B<-frm>, or B<--frames> flag
3190 By default, a single html output file is produced. This can be changed
3191 with the B<-frm> option, which creates a frame holding a table of
3192 contents in the left panel and the source code in the right side. This
3193 simplifies code browsing. Assume, for example, that the input file is
3194 F<MyModule.pm>. Then, for default file extension choices, these three
3195 files will be created:
3197 MyModule.pm.html - the frame
3198 MyModule.pm.toc.html - the table of contents
3199 MyModule.pm.src.html - the formatted source code
3201 Obviously this file naming scheme requires that output be directed to a real
3202 file (as opposed to, say, standard output). If this is not the
3203 case, or if the file extension is unknown, the B<-frm> option will be
3206 =item The B<-text=s>, or B<--html-toc-extension> flag
3208 Use this flag to specify the extra file extension of the table of contents file
3209 when html frames are used. The default is "toc".
3210 See L<Specifying File Extensions>.
3212 =item The B<-sext=s>, or B<--html-src-extension> flag
3214 Use this flag to specify the extra file extension of the content file when html
3215 frames are used. The default is "src".
3216 See L<Specifying File Extensions>.
3218 =item The B<-hent>, or B<--html-entities> flag
3220 This flag controls the use of Html::Entities for html formatting. By
3221 default, the module Html::Entities is used to encode special symbols.
3222 This may not be the right thing for some browser/language
3223 combinations. Use --nohtml-entities or -nhent to prevent this.
3227 Style sheets make it very convenient to control and adjust the
3228 appearance of html pages. The default behavior is to write a page of
3229 html with an embedded style sheet.
3231 An alternative to an embedded style sheet is to create a page with a
3232 link to an external style sheet. This is indicated with the
3233 B<-css=filename>, where the external style sheet is F<filename>. The
3234 external style sheet F<filename> will be created if and only if it does
3235 not exist. This option is useful for controlling multiple pages from a
3238 To cause perltidy to write a style sheet to standard output and exit,
3239 use the B<-ss>, or B<--stylesheet>, flag. This is useful if the style
3240 sheet could not be written for some reason, such as if the B<-pre> flag
3241 was used. Thus, for example,
3243 perltidy -html -ss >mystyle.css
3245 will write a style sheet with the default properties to file
3248 The use of style sheets is encouraged, but a web page without a style
3249 sheets can be created with the flag B<-nss>. Use this option if you
3250 must to be sure that older browsers (roughly speaking, versions prior to
3251 4.0 of Netscape Navigator and Internet Explorer) can display the
3252 syntax-coloring of the html files.
3254 =item Controlling HTML properties
3256 Note: It is usually more convenient to accept the default properties
3257 and then edit the stylesheet which is produced. However, this section
3258 shows how to control the properties with flags to perltidy.
3260 Syntax colors may be changed from their default values by flags of the either
3261 the long form, B<-html-color-xxxxxx=n>, or more conveniently the short form,
3262 B<-hcx=n>, where B<xxxxxx> is one of the following words, and B<x> is the
3263 corresponding abbreviation:
3266 ---------- -------- --
3269 identifier identifier i
3270 bareword, function bareword w
3272 quite, pattern quote q
3273 here doc text here-doc-text h
3274 here doc target here-doc-target hh
3275 punctuation punctuation pu
3277 structural braces structure s
3278 semicolon semicolon sc
3282 sub definition name subroutine m
3283 pod text pod-text pd
3285 A default set of colors has been defined, but they may be changed by providing
3286 values to any of the following parameters, where B<n> is either a 6 digit
3287 hex RGB color value or an ascii name for a color, such as 'red'.
3289 To illustrate, the following command will produce an html
3290 file F<somefile.pl.html> with "aqua" keywords:
3292 perltidy -html -hck=00ffff somefile.pl
3294 and this should be equivalent for most browsers:
3296 perltidy -html -hck=aqua somefile.pl
3298 Perltidy merely writes any non-hex names that it sees in the html file.
3299 The following 16 color names are defined in the HTML 3.2 standard:
3318 Many more names are supported in specific browsers, but it is safest
3319 to use the hex codes for other colors. Helpful color tables can be
3320 located with an internet search for "HTML color tables".
3322 Besides color, two other character attributes may be set: bold, and italics.
3323 To set a token type to use bold, use the flag
3324 B<--html-bold-xxxxxx> or B<-hbx>, where B<xxxxxx> or B<x> are the long
3325 or short names from the above table. Conversely, to set a token type to
3326 NOT use bold, use B<--nohtml-bold-xxxxxx> or B<-nhbx>.
3328 Likewise, to set a token type to use an italic font, use the flag
3329 B<--html-italic-xxxxxx> or B<-hix>, where again B<xxxxxx> or B<x> are the
3330 long or short names from the above table. And to set a token type to
3331 NOT use italics, use B<--nohtml-italic-xxxxxx> or B<-nhix>.
3333 For example, to use bold braces and lime color, non-bold, italics keywords the
3334 following command would be used:
3336 perltidy -html -hbs -hck=00FF00 -nhbk -hik somefile.pl
3338 The background color can be specified with B<--html-color-background=n>,
3339 or B<-hcbg=n> for short, where n is a 6 character hex RGB value. The
3340 default color of text is the value given to B<punctuation>, which is
3343 Here are some notes and hints:
3345 1. If you find a preferred set of these parameters, you may want
3346 to create a F<.perltidyrc> file containing them. See the perltidy man
3347 page for an explanation.
3349 2. Rather than specifying values for these parameters, it is probably
3350 easier to accept the defaults and then edit a style sheet. The style
3351 sheet contains comments which should make this easy.
3353 3. The syntax-colored html files can be very large, so it may be best to
3354 split large files into smaller pieces to improve download times.
3358 =head1 SOME COMMON INPUT CONVENTIONS
3360 =head2 Specifying Block Types
3362 Several parameters which refer to code block types may be customized by also
3363 specifying an associated list of block types. The type of a block is the name
3364 of the keyword which introduces that block, such as B<if>, B<else>, or B<sub>.
3365 An exception is a labeled block, which has no keyword, and should be specified
3366 with just a colon. To specify all blocks use B<'*'>.
3368 The keyword B<sub> indicates a named sub. For anonymous subs, use the special
3371 For example, the following parameter specifies C<sub>, labels, C<BEGIN>, and
3374 -cscl="sub : BEGIN END"
3376 (the meaning of the -cscl parameter is described above.) Note that
3377 quotes are required around the list of block types because of the
3378 spaces. For another example, the following list specifies all block types
3379 for vertical tightness:
3383 =head2 Specifying File Extensions
3385 Several parameters allow default file extensions to be overridden. For
3386 example, a backup file extension may be specified with B<-bext=ext>,
3387 where B<ext> is some new extension. In order to provides the user some
3388 flexibility, the following convention is used in all cases to decide if
3389 a leading '.' should be used. If the extension C<ext> begins with
3390 C<A-Z>, C<a-z>, or C<0-9>, then it will be appended to the filename with
3391 an intermediate '.' (or perhaps an '_' on VMS systems). Otherwise, it
3392 will be appended directly.
3394 For example, suppose the file is F<somefile.pl>. For C<-bext=old>, a '.' is
3395 added to give F<somefile.pl.old>. For C<-bext=.old>, no additional '.' is
3396 added, so again the backup file is F<somefile.pl.old>. For C<-bext=~>, then no
3397 dot is added, and the backup file will be F<somefile.pl~> .
3399 =head1 SWITCHES WHICH MAY BE NEGATED
3401 The following list shows all short parameter names which allow a prefix
3402 'n' to produce the negated form:
3404 D anl asc aws b bbb bbc bbs bl bli boc bok bol bot ce
3405 csc dac dbc dcsc ddf dln dnl dop dp dpro dsc dsm dsn dtt dwls
3406 dwrs dws f fll frm fs hsc html ibc icb icp iob isbc lal log
3407 lp lsl ohbr okw ola oll opr opt osbr otr ple pod pvl q
3408 sbc sbl schb scp scsb sct se sfp sfs skp sob sohb sop sosb sot
3409 ssc st sts syn t tac tbc toc tp tqw tsc w x bar kis
3411 Equivalently, the prefix 'no' or 'no-' on the corresponding long names may be
3418 =item Parsing Limitations
3420 Perltidy should work properly on most perl scripts. It does a lot of
3421 self-checking, but still, it is possible that an error could be
3422 introduced and go undetected. Therefore, it is essential to make
3423 careful backups and to test reformatted scripts.
3425 The main current limitation is that perltidy does not scan modules
3426 included with 'use' statements. This makes it necessary to guess the
3427 context of any bare words introduced by such modules. Perltidy has good
3428 guessing algorithms, but they are not infallible. When it must guess,
3429 it leaves a message in the log file.
3431 If you encounter a bug, please report it.
3433 =item What perltidy does not parse and format
3435 Perltidy indents but does not reformat comments and C<qw> quotes.
3436 Perltidy does not in any way modify the contents of here documents or
3437 quoted text, even if they contain source code. (You could, however,
3438 reformat them separately). Perltidy does not format 'format' sections
3439 in any way. And, of course, it does not modify pod documents.
3447 =item Temporary files
3449 Under the -html option with the default --pod2html flag, a temporary file is
3450 required to pass text to Pod::Html. Unix systems will try to use the POSIX
3451 tmpnam() function. Otherwise the file F<perltidy.TMP> will be temporarily
3452 created in the current working directory.
3454 =item Special files when standard input is used
3456 When standard input is used, the log file, if saved, is F<perltidy.LOG>,
3457 and any errors are written to F<perltidy.ERR> unless the B<-se> flag is
3458 set. These are saved in the current working directory.
3460 =item Files overwritten
3462 The following file extensions are used by perltidy, and files with these
3463 extensions may be overwritten or deleted: F<.ERR>, F<.LOG>, F<.TEE>,
3464 and/or F<.tdy>, F<.html>, and F<.bak>, depending on the run type and
3467 =item Files extensions limitations
3469 Perltidy does not operate on files for which the run could produce a file with
3470 a duplicated file extension. These extensions include F<.LOG>, F<.ERR>,
3471 F<.TEE>, and perhaps F<.tdy> and F<.bak>, depending on the run type. The
3472 purpose of this rule is to prevent generating confusing filenames such as
3473 F<somefile.tdy.tdy.tdy>.
3479 perlstyle(1), Perl::Tidy(3)
3483 This man page documents perltidy version 20180220.
3487 A list of current bugs and issues can be found at the CPAN site
3489 https://rt.cpan.org/Public/Dist/Display.html?Name=Perl-Tidy
3491 To report a new bug or problem, use the link on this page.
3495 Copyright (c) 2000-2018 by Steve Hancock
3499 This package is free software; you can redistribute it and/or modify it
3500 under the terms of the "GNU General Public License".
3502 Please refer to the file "COPYING" for details.
3506 This package is distributed in the hope that it will be useful,
3507 but WITHOUT ANY WARRANTY; without even the implied warranty of
3508 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
3510 See the "GNU General Public License" for more details.