1=head1 NAME 2 3perlform - Perl formats 4 5=head1 DESCRIPTION 6 7Perl has a mechanism to help you generate simple reports and charts. To 8facilitate this, Perl helps you code up your output page close to how it 9will look when it's printed. It can keep track of things like how many 10lines are on a page, what page you're on, when to print page headers, 11etc. Keywords are borrowed from FORTRAN: format() to declare and write() 12to execute; see their entries in L<perlfunc>. Fortunately, the layout is 13much more legible, more like BASIC's PRINT USING statement. Think of it 14as a poor man's nroff(1). 15 16Formats, like packages and subroutines, are declared rather than 17executed, so they may occur at any point in your program. (Usually it's 18best to keep them all together though.) They have their own namespace 19apart from all the other "types" in Perl. This means that if you have a 20function named "Foo", it is not the same thing as having a format named 21"Foo". However, the default name for the format associated with a given 22filehandle is the same as the name of the filehandle. Thus, the default 23format for STDOUT is named "STDOUT", and the default format for filehandle 24TEMP is named "TEMP". They just look the same. They aren't. 25 26Output record formats are declared as follows: 27 28 format NAME = 29 FORMLIST 30 . 31 32If the name is omitted, format "STDOUT" is defined. A single "." in 33column 1 is used to terminate a format. FORMLIST consists of a sequence 34of lines, each of which may be one of three types: 35 36=over 4 37 38=item 1. 39 40A comment, indicated by putting a '#' in the first column. 41 42=item 2. 43 44A "picture" line giving the format for one output line. 45 46=item 3. 47 48An argument line supplying values to plug into the previous picture line. 49 50=back 51 52Picture lines contain output field definitions, intermingled with 53literal text. These lines do not undergo any kind of variable interpolation. 54Field definitions are made up from a set of characters, for starting and 55extending a field to its desired width. This is the complete set of 56characters for field definitions: 57 58 @ start of regular field 59 ^ start of special field 60 < pad character for left adjustification 61 | pad character for centering 62 > pad character for right adjustificat 63 # pad character for a right justified numeric field 64 0 instead of first #: pad number with leading zeroes 65 . decimal point within a numeric field 66 ... terminate a text field, show "..." as truncation evidence 67 @* variable width field for a multi-line value 68 ^* variable width field for next line of a multi-line value 69 ~ suppress line with all fields empty 70 ~~ repeat line until all fields are exhausted 71 72Each field in a picture line starts with either "@" (at) or "^" (caret), 73indicating what we'll call, respectively, a "regular" or "special" field. 74The choice of pad characters determines whether a field is textual or 75numeric. The tilde operators are not part of a field. Let's look at 76the various possibilities in detail. 77 78 79=head2 Text Fields 80 81The length of the field is supplied by padding out the field with multiple 82"E<lt>", "E<gt>", or "|" characters to specify a non-numeric field with, 83respectively, left justification, right justification, or centering. 84For a regular field, the value (up to the first newline) is taken and 85printed according to the selected justification, truncating excess characters. 86If you terminate a text field with "...", three dots will be shown if 87the value is truncated. A special text field may be used to do rudimentary 88multi-line text block filling; see L</Using Fill Mode> for details. 89 90 Example: 91 format STDOUT = 92 @<<<<<< @|||||| @>>>>>> 93 "left", "middle", "right" 94 . 95 Output: 96 left middle right 97 98 99=head2 Numeric Fields 100 101Using "#" as a padding character specifies a numeric field, with 102right justification. An optional "." defines the position of the 103decimal point. With a "0" (zero) instead of the first "#", the 104formatted number will be padded with leading zeroes if necessary. 105A special numeric field is blanked out if the value is undefined. 106If the resulting value would exceed the width specified the field is 107filled with "#" as overflow evidence. 108 109 Example: 110 format STDOUT = 111 @### @.### @##.### @### @### ^#### 112 42, 3.1415, undef, 0, 10000, undef 113 . 114 Output: 115 42 3.142 0.000 0 #### 116 117 118=head2 The Field @* for Variable Width Multi-Line Text 119 120The field "@*" can be used for printing multi-line, nontruncated 121values; it should (but need not) appear by itself on a line. A final 122line feed is chomped off, but all other characters are emitted verbatim. 123 124 125=head2 The Field ^* for Variable Width One-line-at-a-time Text 126 127Like "@*", this is a variable width field. The value supplied must be a 128scalar variable. Perl puts the first line (up to the first "\n") of the 129text into the field, and then chops off the front of the string so that 130the next time the variable is referenced, more of the text can be printed. 131The variable will I<not> be restored. 132 133 Example: 134 $text = "line 1\nline 2\nline 3"; 135 format STDOUT = 136 Text: ^* 137 $text 138 ~~ ^* 139 $text 140 . 141 Output: 142 Text: line 1 143 line 2 144 line 3 145 146 147=head2 Specifying Values 148 149The values are specified on the following format line in the same order as 150the picture fields. The expressions providing the values must be 151separated by commas. They are all evaluated in a list context 152before the line is processed, so a single list expression could produce 153multiple list elements. The expressions may be spread out to more than 154one line if enclosed in braces. If so, the opening brace must be the first 155token on the first line. If an expression evaluates to a number with a 156decimal part, and if the corresponding picture specifies that the decimal 157part should appear in the output (that is, any picture except multiple "#" 158characters B<without> an embedded "."), the character used for the decimal 159point is B<always> determined by the current LC_NUMERIC locale. This 160means that, if, for example, the run-time environment happens to specify a 161German locale, "," will be used instead of the default ".". See 162L<perllocale> and L<"WARNINGS"> for more information. 163 164 165=head2 Using Fill Mode 166 167On text fields the caret enables a kind of fill mode. Instead of an 168arbitrary expression, the value supplied must be a scalar variable 169that contains a text string. Perl puts the next portion of the text into 170the field, and then chops off the front of the string so that the next time 171the variable is referenced, more of the text can be printed. (Yes, this 172means that the variable itself is altered during execution of the write() 173call, and is not restored.) The next portion of text is determined by 174a crude line breaking algorithm. You may use the carriage return character 175(C<\r>) to force a line break. You can change which characters are legal 176to break on by changing the variable C<$:> (that's 177$FORMAT_LINE_BREAK_CHARACTERS if you're using the English module) to a 178list of the desired characters. 179 180Normally you would use a sequence of fields in a vertical stack associated 181with the same scalar variable to print out a block of text. You might wish 182to end the final field with the text "...", which will appear in the output 183if the text was too long to appear in its entirety. 184 185 186=head2 Suppressing Lines Where All Fields Are Void 187 188Using caret fields can produce lines where all fields are blank. You can 189suppress such lines by putting a "~" (tilde) character anywhere in the 190line. The tilde will be translated to a space upon output. 191 192 193=head2 Repeating Format Lines 194 195If you put two contiguous tilde characters "~~" anywhere into a line, 196the line will be repeated until all the fields on the line are exhausted, 197i.e. undefined. For special (caret) text fields this will occur sooner or 198later, but if you use a text field of the at variety, the expression you 199supply had better not give the same value every time forever! (C<shift(@f)> 200is a simple example that would work.) Don't use a regular (at) numeric 201field in such lines, because it will never go blank. 202 203 204=head2 Top of Form Processing 205 206Top-of-form processing is by default handled by a format with the 207same name as the current filehandle with "_TOP" concatenated to it. 208It's triggered at the top of each page. See L<perlfunc/write>. 209 210Examples: 211 212 # a report on the /etc/passwd file 213 format STDOUT_TOP = 214 Passwd File 215 Name Login Office Uid Gid Home 216 ------------------------------------------------------------------ 217 . 218 format STDOUT = 219 @<<<<<<<<<<<<<<<<<< @||||||| @<<<<<<@>>>> @>>>> @<<<<<<<<<<<<<<<<< 220 $name, $login, $office,$uid,$gid, $home 221 . 222 223 224 # a report from a bug report form 225 format STDOUT_TOP = 226 Bug Reports 227 @<<<<<<<<<<<<<<<<<<<<<<< @||| @>>>>>>>>>>>>>>>>>>>>>>> 228 $system, $%, $date 229 ------------------------------------------------------------------ 230 . 231 format STDOUT = 232 Subject: @<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<< 233 $subject 234 Index: @<<<<<<<<<<<<<<<<<<<<<<<<<<<< ^<<<<<<<<<<<<<<<<<<<<<<<<<<<< 235 $index, $description 236 Priority: @<<<<<<<<<< Date: @<<<<<<< ^<<<<<<<<<<<<<<<<<<<<<<<<<<<< 237 $priority, $date, $description 238 From: @<<<<<<<<<<<<<<<<<<<<<<<<<<<<< ^<<<<<<<<<<<<<<<<<<<<<<<<<<<< 239 $from, $description 240 Assigned to: @<<<<<<<<<<<<<<<<<<<<<< ^<<<<<<<<<<<<<<<<<<<<<<<<<<<< 241 $programmer, $description 242 ~ ^<<<<<<<<<<<<<<<<<<<<<<<<<<<< 243 $description 244 ~ ^<<<<<<<<<<<<<<<<<<<<<<<<<<<< 245 $description 246 ~ ^<<<<<<<<<<<<<<<<<<<<<<<<<<<< 247 $description 248 ~ ^<<<<<<<<<<<<<<<<<<<<<<<<<<<< 249 $description 250 ~ ^<<<<<<<<<<<<<<<<<<<<<<<... 251 $description 252 . 253 254It is possible to intermix print()s with write()s on the same output 255channel, but you'll have to handle C<$-> (C<$FORMAT_LINES_LEFT>) 256yourself. 257 258=head2 Format Variables 259 260The current format name is stored in the variable C<$~> (C<$FORMAT_NAME>), 261and the current top of form format name is in C<$^> (C<$FORMAT_TOP_NAME>). 262The current output page number is stored in C<$%> (C<$FORMAT_PAGE_NUMBER>), 263and the number of lines on the page is in C<$=> (C<$FORMAT_LINES_PER_PAGE>). 264Whether to autoflush output on this handle is stored in C<$|> 265(C<$OUTPUT_AUTOFLUSH>). The string output before each top of page (except 266the first) is stored in C<$^L> (C<$FORMAT_FORMFEED>). These variables are 267set on a per-filehandle basis, so you'll need to select() into a different 268one to affect them: 269 270 select((select(OUTF), 271 $~ = "My_Other_Format", 272 $^ = "My_Top_Format" 273 )[0]); 274 275Pretty ugly, eh? It's a common idiom though, so don't be too surprised 276when you see it. You can at least use a temporary variable to hold 277the previous filehandle: (this is a much better approach in general, 278because not only does legibility improve, you now have intermediary 279stage in the expression to single-step the debugger through): 280 281 $ofh = select(OUTF); 282 $~ = "My_Other_Format"; 283 $^ = "My_Top_Format"; 284 select($ofh); 285 286If you use the English module, you can even read the variable names: 287 288 use English '-no_match_vars'; 289 $ofh = select(OUTF); 290 $FORMAT_NAME = "My_Other_Format"; 291 $FORMAT_TOP_NAME = "My_Top_Format"; 292 select($ofh); 293 294But you still have those funny select()s. So just use the FileHandle 295module. Now, you can access these special variables using lowercase 296method names instead: 297 298 use FileHandle; 299 format_name OUTF "My_Other_Format"; 300 format_top_name OUTF "My_Top_Format"; 301 302Much better! 303 304=head1 NOTES 305 306Because the values line may contain arbitrary expressions (for at fields, 307not caret fields), you can farm out more sophisticated processing 308to other functions, like sprintf() or one of your own. For example: 309 310 format Ident = 311 @<<<<<<<<<<<<<<< 312 &commify($n) 313 . 314 315To get a real at or caret into the field, do this: 316 317 format Ident = 318 I have an @ here. 319 "@" 320 . 321 322To center a whole line of text, do something like this: 323 324 format Ident = 325 @||||||||||||||||||||||||||||||||||||||||||||||| 326 "Some text line" 327 . 328 329There is no builtin way to say "float this to the right hand side 330of the page, however wide it is." You have to specify where it goes. 331The truly desperate can generate their own format on the fly, based 332on the current number of columns, and then eval() it: 333 334 $format = "format STDOUT = \n" 335 . '^' . '<' x $cols . "\n" 336 . '$entry' . "\n" 337 . "\t^" . "<" x ($cols-8) . "~~\n" 338 . '$entry' . "\n" 339 . ".\n"; 340 print $format if $Debugging; 341 eval $format; 342 die $@ if $@; 343 344Which would generate a format looking something like this: 345 346 format STDOUT = 347 ^<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<< 348 $entry 349 ^<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<~~ 350 $entry 351 . 352 353Here's a little program that's somewhat like fmt(1): 354 355 format = 356 ^<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<< ~~ 357 $_ 358 359 . 360 361 $/ = ''; 362 while (<>) { 363 s/\s*\n\s*/ /g; 364 write; 365 } 366 367=head2 Footers 368 369While $FORMAT_TOP_NAME contains the name of the current header format, 370there is no corresponding mechanism to automatically do the same thing 371for a footer. Not knowing how big a format is going to be until you 372evaluate it is one of the major problems. It's on the TODO list. 373 374Here's one strategy: If you have a fixed-size footer, you can get footers 375by checking $FORMAT_LINES_LEFT before each write() and print the footer 376yourself if necessary. 377 378Here's another strategy: Open a pipe to yourself, using C<open(MYSELF, "|-")> 379(see L<perlfunc/open()>) and always write() to MYSELF instead of STDOUT. 380Have your child process massage its STDIN to rearrange headers and footers 381however you like. Not very convenient, but doable. 382 383=head2 Accessing Formatting Internals 384 385For low-level access to the formatting mechanism. you may use formline() 386and access C<$^A> (the $ACCUMULATOR variable) directly. 387 388For example: 389 390 $str = formline <<'END', 1,2,3; 391 @<<< @||| @>>> 392 END 393 394 print "Wow, I just stored `$^A' in the accumulator!\n"; 395 396Or to make an swrite() subroutine, which is to write() what sprintf() 397is to printf(), do this: 398 399 use Carp; 400 sub swrite { 401 croak "usage: swrite PICTURE ARGS" unless @_; 402 my $format = shift; 403 $^A = ""; 404 formline($format,@_); 405 return $^A; 406 } 407 408 $string = swrite(<<'END', 1, 2, 3); 409 Check me out 410 @<<< @||| @>>> 411 END 412 print $string; 413 414=head1 WARNINGS 415 416The lone dot that ends a format can also prematurely end a mail 417message passing through a misconfigured Internet mailer (and based on 418experience, such misconfiguration is the rule, not the exception). So 419when sending format code through mail, you should indent it so that 420the format-ending dot is not on the left margin; this will prevent 421SMTP cutoff. 422 423Lexical variables (declared with "my") are not visible within a 424format unless the format is declared within the scope of the lexical 425variable. (They weren't visible at all before version 5.001.) 426 427Formats are the only part of Perl that unconditionally use information 428from a program's locale; if a program's environment specifies an 429LC_NUMERIC locale, it is always used to specify the decimal point 430character in formatted output. Perl ignores all other aspects of locale 431handling unless the C<use locale> pragma is in effect. Formatted output 432cannot be controlled by C<use locale> because the pragma is tied to the 433block structure of the program, and, for historical reasons, formats 434exist outside that block structure. See L<perllocale> for further 435discussion of locale handling. 436 437Within strings that are to be displayed in a fixed length text field, 438each control character is substituted by a space. (But remember the 439special meaning of C<\r> when using fill mode.) This is done to avoid 440misalignment when control characters "disappear" on some output media. 441 442