| 1 | .\" Automatically generated by Pod::Man v1.37, Pod::Parser v1.32 |
| 2 | .\" |
| 3 | .\" Standard preamble: |
| 4 | .\" ======================================================================== |
| 5 | .de Sh \" Subsection heading |
| 6 | .br |
| 7 | .if t .Sp |
| 8 | .ne 5 |
| 9 | .PP |
| 10 | \fB\\$1\fR |
| 11 | .PP |
| 12 | .. |
| 13 | .de Sp \" Vertical space (when we can't use .PP) |
| 14 | .if t .sp .5v |
| 15 | .if n .sp |
| 16 | .. |
| 17 | .de Vb \" Begin verbatim text |
| 18 | .ft CW |
| 19 | .nf |
| 20 | .ne \\$1 |
| 21 | .. |
| 22 | .de Ve \" End verbatim text |
| 23 | .ft R |
| 24 | .fi |
| 25 | .. |
| 26 | .\" Set up some character translations and predefined strings. \*(-- will |
| 27 | .\" give an unbreakable dash, \*(PI will give pi, \*(L" will give a left |
| 28 | .\" double quote, and \*(R" will give a right double quote. | will give a |
| 29 | .\" real vertical bar. \*(C+ will give a nicer C++. Capital omega is used to |
| 30 | .\" do unbreakable dashes and therefore won't be available. \*(C` and \*(C' |
| 31 | .\" expand to `' in nroff, nothing in troff, for use with C<>. |
| 32 | .tr \(*W-|\(bv\*(Tr |
| 33 | .ds C+ C\v'-.1v'\h'-1p'\s-2+\h'-1p'+\s0\v'.1v'\h'-1p' |
| 34 | .ie n \{\ |
| 35 | . ds -- \(*W- |
| 36 | . ds PI pi |
| 37 | . if (\n(.H=4u)&(1m=24u) .ds -- \(*W\h'-12u'\(*W\h'-12u'-\" diablo 10 pitch |
| 38 | . if (\n(.H=4u)&(1m=20u) .ds -- \(*W\h'-12u'\(*W\h'-8u'-\" diablo 12 pitch |
| 39 | . ds L" "" |
| 40 | . ds R" "" |
| 41 | . ds C` "" |
| 42 | . ds C' "" |
| 43 | 'br\} |
| 44 | .el\{\ |
| 45 | . ds -- \|\(em\| |
| 46 | . ds PI \(*p |
| 47 | . ds L" `` |
| 48 | . ds R" '' |
| 49 | 'br\} |
| 50 | .\" |
| 51 | .\" If the F register is turned on, we'll generate index entries on stderr for |
| 52 | .\" titles (.TH), headers (.SH), subsections (.Sh), items (.Ip), and index |
| 53 | .\" entries marked with X<> in POD. Of course, you'll have to process the |
| 54 | .\" output yourself in some meaningful fashion. |
| 55 | .if \nF \{\ |
| 56 | . de IX |
| 57 | . tm Index:\\$1\t\\n%\t"\\$2" |
| 58 | .. |
| 59 | . nr % 0 |
| 60 | . rr F |
| 61 | .\} |
| 62 | .\" |
| 63 | .\" For nroff, turn off justification. Always turn off hyphenation; it makes |
| 64 | .\" way too many mistakes in technical documents. |
| 65 | .hy 0 |
| 66 | .if n .na |
| 67 | .\" |
| 68 | .\" Accent mark definitions (@(#)ms.acc 1.5 88/02/08 SMI; from UCB 4.2). |
| 69 | .\" Fear. Run. Save yourself. No user-serviceable parts. |
| 70 | . \" fudge factors for nroff and troff |
| 71 | .if n \{\ |
| 72 | . ds #H 0 |
| 73 | . ds #V .8m |
| 74 | . ds #F .3m |
| 75 | . ds #[ \f1 |
| 76 | . ds #] \fP |
| 77 | .\} |
| 78 | .if t \{\ |
| 79 | . ds #H ((1u-(\\\\n(.fu%2u))*.13m) |
| 80 | . ds #V .6m |
| 81 | . ds #F 0 |
| 82 | . ds #[ \& |
| 83 | . ds #] \& |
| 84 | .\} |
| 85 | . \" simple accents for nroff and troff |
| 86 | .if n \{\ |
| 87 | . ds ' \& |
| 88 | . ds ` \& |
| 89 | . ds ^ \& |
| 90 | . ds , \& |
| 91 | . ds ~ ~ |
| 92 | . ds / |
| 93 | .\} |
| 94 | .if t \{\ |
| 95 | . ds ' \\k:\h'-(\\n(.wu*8/10-\*(#H)'\'\h"|\\n:u" |
| 96 | . ds ` \\k:\h'-(\\n(.wu*8/10-\*(#H)'\`\h'|\\n:u' |
| 97 | . ds ^ \\k:\h'-(\\n(.wu*10/11-\*(#H)'^\h'|\\n:u' |
| 98 | . ds , \\k:\h'-(\\n(.wu*8/10)',\h'|\\n:u' |
| 99 | . ds ~ \\k:\h'-(\\n(.wu-\*(#H-.1m)'~\h'|\\n:u' |
| 100 | . ds / \\k:\h'-(\\n(.wu*8/10-\*(#H)'\z\(sl\h'|\\n:u' |
| 101 | .\} |
| 102 | . \" troff and (daisy-wheel) nroff accents |
| 103 | .ds : \\k:\h'-(\\n(.wu*8/10-\*(#H+.1m+\*(#F)'\v'-\*(#V'\z.\h'.2m+\*(#F'.\h'|\\n:u'\v'\*(#V' |
| 104 | .ds 8 \h'\*(#H'\(*b\h'-\*(#H' |
| 105 | .ds o \\k:\h'-(\\n(.wu+\w'\(de'u-\*(#H)/2u'\v'-.3n'\*(#[\z\(de\v'.3n'\h'|\\n:u'\*(#] |
| 106 | .ds d- \h'\*(#H'\(pd\h'-\w'~'u'\v'-.25m'\f2\(hy\fP\v'.25m'\h'-\*(#H' |
| 107 | .ds D- D\\k:\h'-\w'D'u'\v'-.11m'\z\(hy\v'.11m'\h'|\\n:u' |
| 108 | .ds th \*(#[\v'.3m'\s+1I\s-1\v'-.3m'\h'-(\w'I'u*2/3)'\s-1o\s+1\*(#] |
| 109 | .ds Th \*(#[\s+2I\s-2\h'-\w'I'u*3/5'\v'-.3m'o\v'.3m'\*(#] |
| 110 | .ds ae a\h'-(\w'a'u*4/10)'e |
| 111 | .ds Ae A\h'-(\w'A'u*4/10)'E |
| 112 | . \" corrections for vroff |
| 113 | .if v .ds ~ \\k:\h'-(\\n(.wu*9/10-\*(#H)'\s-2\u~\d\s+2\h'|\\n:u' |
| 114 | .if v .ds ^ \\k:\h'-(\\n(.wu*10/11-\*(#H)'\v'-.4m'^\v'.4m'\h'|\\n:u' |
| 115 | . \" for low resolution devices (crt and lpr) |
| 116 | .if \n(.H>23 .if \n(.V>19 \ |
| 117 | \{\ |
| 118 | . ds : e |
| 119 | . ds 8 ss |
| 120 | . ds o a |
| 121 | . ds d- d\h'-1'\(ga |
| 122 | . ds D- D\h'-1'\(hy |
| 123 | . ds th \o'bp' |
| 124 | . ds Th \o'LP' |
| 125 | . ds ae ae |
| 126 | . ds Ae AE |
| 127 | .\} |
| 128 | .rm #[ #] #H #V #F C |
| 129 | .\" ======================================================================== |
| 130 | .\" |
| 131 | .IX Title "PERLTIE 1" |
| 132 | .TH PERLTIE 1 "2006-01-07" "perl v5.8.8" "Perl Programmers Reference Guide" |
| 133 | .SH "NAME" |
| 134 | .IX Xref "tie" |
| 135 | perltie \- how to hide an object class in a simple variable |
| 136 | .SH "SYNOPSIS" |
| 137 | .IX Header "SYNOPSIS" |
| 138 | .Vb 1 |
| 139 | \& tie VARIABLE, CLASSNAME, LIST |
| 140 | .Ve |
| 141 | .PP |
| 142 | .Vb 1 |
| 143 | \& $object = tied VARIABLE |
| 144 | .Ve |
| 145 | .PP |
| 146 | .Vb 1 |
| 147 | \& untie VARIABLE |
| 148 | .Ve |
| 149 | .SH "DESCRIPTION" |
| 150 | .IX Header "DESCRIPTION" |
| 151 | Prior to release 5.0 of Perl, a programmer could use \fIdbmopen()\fR |
| 152 | to connect an on-disk database in the standard Unix \fIdbm\fR\|(3x) |
| 153 | format magically to a \f(CW%HASH\fR in their program. However, their Perl was either |
| 154 | built with one particular dbm library or another, but not both, and |
| 155 | you couldn't extend this mechanism to other packages or types of variables. |
| 156 | .PP |
| 157 | Now you can. |
| 158 | .PP |
| 159 | The \fItie()\fR function binds a variable to a class (package) that will provide |
| 160 | the implementation for access methods for that variable. Once this magic |
| 161 | has been performed, accessing a tied variable automatically triggers |
| 162 | method calls in the proper class. The complexity of the class is |
| 163 | hidden behind magic methods calls. The method names are in \s-1ALL\s0 \s-1CAPS\s0, |
| 164 | which is a convention that Perl uses to indicate that they're called |
| 165 | implicitly rather than explicitly\*(--just like the \s-1\fIBEGIN\s0()\fR and \s-1\fIEND\s0()\fR |
| 166 | functions. |
| 167 | .PP |
| 168 | In the \fItie()\fR call, \f(CW\*(C`VARIABLE\*(C'\fR is the name of the variable to be |
| 169 | enchanted. \f(CW\*(C`CLASSNAME\*(C'\fR is the name of a class implementing objects of |
| 170 | the correct type. Any additional arguments in the \f(CW\*(C`LIST\*(C'\fR are passed to |
| 171 | the appropriate constructor method for that class\*(--meaning \s-1\fITIESCALAR\s0()\fR, |
| 172 | \&\s-1\fITIEARRAY\s0()\fR, \s-1\fITIEHASH\s0()\fR, or \s-1\fITIEHANDLE\s0()\fR. (Typically these are arguments |
| 173 | such as might be passed to the \fIdbminit()\fR function of C.) The object |
| 174 | returned by the \*(L"new\*(R" method is also returned by the \fItie()\fR function, |
| 175 | which would be useful if you wanted to access other methods in |
| 176 | \&\f(CW\*(C`CLASSNAME\*(C'\fR. (You don't actually have to return a reference to a right |
| 177 | \&\*(L"type\*(R" (e.g., \s-1HASH\s0 or \f(CW\*(C`CLASSNAME\*(C'\fR) so long as it's a properly blessed |
| 178 | object.) You can also retrieve a reference to the underlying object |
| 179 | using the \fItied()\fR function. |
| 180 | .PP |
| 181 | Unlike \fIdbmopen()\fR, the \fItie()\fR function will not \f(CW\*(C`use\*(C'\fR or \f(CW\*(C`require\*(C'\fR a module |
| 182 | for you\*(--you need to do that explicitly yourself. |
| 183 | .Sh "Tying Scalars" |
| 184 | .IX Xref "scalar, tying" |
| 185 | .IX Subsection "Tying Scalars" |
| 186 | A class implementing a tied scalar should define the following methods: |
| 187 | \&\s-1TIESCALAR\s0, \s-1FETCH\s0, \s-1STORE\s0, and possibly \s-1UNTIE\s0 and/or \s-1DESTROY\s0. |
| 188 | .PP |
| 189 | Let's look at each in turn, using as an example a tie class for |
| 190 | scalars that allows the user to do something like: |
| 191 | .PP |
| 192 | .Vb 2 |
| 193 | \& tie $his_speed, 'Nice', getppid(); |
| 194 | \& tie $my_speed, 'Nice', $$; |
| 195 | .Ve |
| 196 | .PP |
| 197 | And now whenever either of those variables is accessed, its current |
| 198 | system priority is retrieved and returned. If those variables are set, |
| 199 | then the process's priority is changed! |
| 200 | .PP |
| 201 | We'll use Jarkko Hietaniemi <\fIjhi@iki.fi\fR>'s BSD::Resource class (not |
| 202 | included) to access the \s-1PRIO_PROCESS\s0, \s-1PRIO_MIN\s0, and \s-1PRIO_MAX\s0 constants |
| 203 | from your system, as well as the \fIgetpriority()\fR and \fIsetpriority()\fR system |
| 204 | calls. Here's the preamble of the class. |
| 205 | .PP |
| 206 | .Vb 5 |
| 207 | \& package Nice; |
| 208 | \& use Carp; |
| 209 | \& use BSD::Resource; |
| 210 | \& use strict; |
| 211 | \& $Nice::DEBUG = 0 unless defined $Nice::DEBUG; |
| 212 | .Ve |
| 213 | .IP "\s-1TIESCALAR\s0 classname, \s-1LIST\s0" 4 |
| 214 | .IX Xref "TIESCALAR" |
| 215 | .IX Item "TIESCALAR classname, LIST" |
| 216 | This is the constructor for the class. That means it is |
| 217 | expected to return a blessed reference to a new scalar |
| 218 | (probably anonymous) that it's creating. For example: |
| 219 | .Sp |
| 220 | .Vb 3 |
| 221 | \& sub TIESCALAR { |
| 222 | \& my $class = shift; |
| 223 | \& my $pid = shift || $$; # 0 means me |
| 224 | .Ve |
| 225 | .Sp |
| 226 | .Vb 4 |
| 227 | \& if ($pid !~ /^\ed+$/) { |
| 228 | \& carp "Nice::Tie::Scalar got non-numeric pid $pid" if $^W; |
| 229 | \& return undef; |
| 230 | \& } |
| 231 | .Ve |
| 232 | .Sp |
| 233 | .Vb 4 |
| 234 | \& unless (kill 0, $pid) { # EPERM or ERSCH, no doubt |
| 235 | \& carp "Nice::Tie::Scalar got bad pid $pid: $!" if $^W; |
| 236 | \& return undef; |
| 237 | \& } |
| 238 | .Ve |
| 239 | .Sp |
| 240 | .Vb 2 |
| 241 | \& return bless \e$pid, $class; |
| 242 | \& } |
| 243 | .Ve |
| 244 | .Sp |
| 245 | This tie class has chosen to return an error rather than raising an |
| 246 | exception if its constructor should fail. While this is how \fIdbmopen()\fR works, |
| 247 | other classes may well not wish to be so forgiving. It checks the global |
| 248 | variable \f(CW$^W\fR to see whether to emit a bit of noise anyway. |
| 249 | .IP "\s-1FETCH\s0 this" 4 |
| 250 | .IX Xref "FETCH" |
| 251 | .IX Item "FETCH this" |
| 252 | This method will be triggered every time the tied variable is accessed |
| 253 | (read). It takes no arguments beyond its self reference, which is the |
| 254 | object representing the scalar we're dealing with. Because in this case |
| 255 | we're using just a \s-1SCALAR\s0 ref for the tied scalar object, a simple $$self |
| 256 | allows the method to get at the real value stored there. In our example |
| 257 | below, that real value is the process \s-1ID\s0 to which we've tied our variable. |
| 258 | .Sp |
| 259 | .Vb 10 |
| 260 | \& sub FETCH { |
| 261 | \& my $self = shift; |
| 262 | \& confess "wrong type" unless ref $self; |
| 263 | \& croak "usage error" if @_; |
| 264 | \& my $nicety; |
| 265 | \& local($!) = 0; |
| 266 | \& $nicety = getpriority(PRIO_PROCESS, $$self); |
| 267 | \& if ($!) { croak "getpriority failed: $!" } |
| 268 | \& return $nicety; |
| 269 | \& } |
| 270 | .Ve |
| 271 | .Sp |
| 272 | This time we've decided to blow up (raise an exception) if the renice |
| 273 | fails\*(--there's no place for us to return an error otherwise, and it's |
| 274 | probably the right thing to do. |
| 275 | .IP "\s-1STORE\s0 this, value" 4 |
| 276 | .IX Xref "STORE" |
| 277 | .IX Item "STORE this, value" |
| 278 | This method will be triggered every time the tied variable is set |
| 279 | (assigned). Beyond its self reference, it also expects one (and only one) |
| 280 | argument\*(--the new value the user is trying to assign. Don't worry about |
| 281 | returning a value from \s-1STORE\s0 \*(-- the semantic of assignment returning the |
| 282 | assigned value is implemented with \s-1FETCH\s0. |
| 283 | .Sp |
| 284 | .Vb 5 |
| 285 | \& sub STORE { |
| 286 | \& my $self = shift; |
| 287 | \& confess "wrong type" unless ref $self; |
| 288 | \& my $new_nicety = shift; |
| 289 | \& croak "usage error" if @_; |
| 290 | .Ve |
| 291 | .Sp |
| 292 | .Vb 6 |
| 293 | \& if ($new_nicety < PRIO_MIN) { |
| 294 | \& carp sprintf |
| 295 | \& "WARNING: priority %d less than minimum system priority %d", |
| 296 | \& $new_nicety, PRIO_MIN if $^W; |
| 297 | \& $new_nicety = PRIO_MIN; |
| 298 | \& } |
| 299 | .Ve |
| 300 | .Sp |
| 301 | .Vb 6 |
| 302 | \& if ($new_nicety > PRIO_MAX) { |
| 303 | \& carp sprintf |
| 304 | \& "WARNING: priority %d greater than maximum system priority %d", |
| 305 | \& $new_nicety, PRIO_MAX if $^W; |
| 306 | \& $new_nicety = PRIO_MAX; |
| 307 | \& } |
| 308 | .Ve |
| 309 | .Sp |
| 310 | .Vb 4 |
| 311 | \& unless (defined setpriority(PRIO_PROCESS, $$self, $new_nicety)) { |
| 312 | \& confess "setpriority failed: $!"; |
| 313 | \& } |
| 314 | \& } |
| 315 | .Ve |
| 316 | .IP "\s-1UNTIE\s0 this" 4 |
| 317 | .IX Xref "UNTIE" |
| 318 | .IX Item "UNTIE this" |
| 319 | This method will be triggered when the \f(CW\*(C`untie\*(C'\fR occurs. This can be useful |
| 320 | if the class needs to know when no further calls will be made. (Except \s-1DESTROY\s0 |
| 321 | of course.) See "The \f(CW\*(C`untie\*(C'\fR Gotcha" below for more details. |
| 322 | .IP "\s-1DESTROY\s0 this" 4 |
| 323 | .IX Xref "DESTROY" |
| 324 | .IX Item "DESTROY this" |
| 325 | This method will be triggered when the tied variable needs to be destructed. |
| 326 | As with other object classes, such a method is seldom necessary, because Perl |
| 327 | deallocates its moribund object's memory for you automatically\*(--this isn't |
| 328 | \&\*(C+, you know. We'll use a \s-1DESTROY\s0 method here for debugging purposes only. |
| 329 | .Sp |
| 330 | .Vb 5 |
| 331 | \& sub DESTROY { |
| 332 | \& my $self = shift; |
| 333 | \& confess "wrong type" unless ref $self; |
| 334 | \& carp "[ Nice::DESTROY pid $$self ]" if $Nice::DEBUG; |
| 335 | \& } |
| 336 | .Ve |
| 337 | .PP |
| 338 | That's about all there is to it. Actually, it's more than all there |
| 339 | is to it, because we've done a few nice things here for the sake |
| 340 | of completeness, robustness, and general aesthetics. Simpler |
| 341 | \&\s-1TIESCALAR\s0 classes are certainly possible. |
| 342 | .Sh "Tying Arrays" |
| 343 | .IX Xref "array, tying" |
| 344 | .IX Subsection "Tying Arrays" |
| 345 | A class implementing a tied ordinary array should define the following |
| 346 | methods: \s-1TIEARRAY\s0, \s-1FETCH\s0, \s-1STORE\s0, \s-1FETCHSIZE\s0, \s-1STORESIZE\s0 and perhaps \s-1UNTIE\s0 and/or \s-1DESTROY\s0. |
| 347 | .PP |
| 348 | \&\s-1FETCHSIZE\s0 and \s-1STORESIZE\s0 are used to provide \f(CW$#array\fR and |
| 349 | equivalent \f(CW\*(C`scalar(@array)\*(C'\fR access. |
| 350 | .PP |
| 351 | The methods \s-1POP\s0, \s-1PUSH\s0, \s-1SHIFT\s0, \s-1UNSHIFT\s0, \s-1SPLICE\s0, \s-1DELETE\s0, and \s-1EXISTS\s0 are |
| 352 | required if the perl operator with the corresponding (but lowercase) name |
| 353 | is to operate on the tied array. The \fBTie::Array\fR class can be used as a |
| 354 | base class to implement the first five of these in terms of the basic |
| 355 | methods above. The default implementations of \s-1DELETE\s0 and \s-1EXISTS\s0 in |
| 356 | \&\fBTie::Array\fR simply \f(CW\*(C`croak\*(C'\fR. |
| 357 | .PP |
| 358 | In addition \s-1EXTEND\s0 will be called when perl would have pre-extended |
| 359 | allocation in a real array. |
| 360 | .PP |
| 361 | For this discussion, we'll implement an array whose elements are a fixed |
| 362 | size at creation. If you try to create an element larger than the fixed |
| 363 | size, you'll take an exception. For example: |
| 364 | .PP |
| 365 | .Vb 4 |
| 366 | \& use FixedElem_Array; |
| 367 | \& tie @array, 'FixedElem_Array', 3; |
| 368 | \& $array[0] = 'cat'; # ok. |
| 369 | \& $array[1] = 'dogs'; # exception, length('dogs') > 3. |
| 370 | .Ve |
| 371 | .PP |
| 372 | The preamble code for the class is as follows: |
| 373 | .PP |
| 374 | .Vb 3 |
| 375 | \& package FixedElem_Array; |
| 376 | \& use Carp; |
| 377 | \& use strict; |
| 378 | .Ve |
| 379 | .IP "\s-1TIEARRAY\s0 classname, \s-1LIST\s0" 4 |
| 380 | .IX Xref "TIEARRAY" |
| 381 | .IX Item "TIEARRAY classname, LIST" |
| 382 | This is the constructor for the class. That means it is expected to |
| 383 | return a blessed reference through which the new array (probably an |
| 384 | anonymous \s-1ARRAY\s0 ref) will be accessed. |
| 385 | .Sp |
| 386 | In our example, just to show you that you don't \fIreally\fR have to return an |
| 387 | \&\s-1ARRAY\s0 reference, we'll choose a \s-1HASH\s0 reference to represent our object. |
| 388 | A \s-1HASH\s0 works out well as a generic record type: the \f(CW\*(C`{ELEMSIZE}\*(C'\fR field will |
| 389 | store the maximum element size allowed, and the \f(CW\*(C`{ARRAY}\*(C'\fR field will hold the |
| 390 | true \s-1ARRAY\s0 ref. If someone outside the class tries to dereference the |
| 391 | object returned (doubtless thinking it an \s-1ARRAY\s0 ref), they'll blow up. |
| 392 | This just goes to show you that you should respect an object's privacy. |
| 393 | .Sp |
| 394 | .Vb 11 |
| 395 | \& sub TIEARRAY { |
| 396 | \& my $class = shift; |
| 397 | \& my $elemsize = shift; |
| 398 | \& if ( @_ || $elemsize =~ /\eD/ ) { |
| 399 | \& croak "usage: tie ARRAY, '" . __PACKAGE__ . "', elem_size"; |
| 400 | \& } |
| 401 | \& return bless { |
| 402 | \& ELEMSIZE => $elemsize, |
| 403 | \& ARRAY => [], |
| 404 | \& }, $class; |
| 405 | \& } |
| 406 | .Ve |
| 407 | .IP "\s-1FETCH\s0 this, index" 4 |
| 408 | .IX Xref "FETCH" |
| 409 | .IX Item "FETCH this, index" |
| 410 | This method will be triggered every time an individual element the tied array |
| 411 | is accessed (read). It takes one argument beyond its self reference: the |
| 412 | index whose value we're trying to fetch. |
| 413 | .Sp |
| 414 | .Vb 5 |
| 415 | \& sub FETCH { |
| 416 | \& my $self = shift; |
| 417 | \& my $index = shift; |
| 418 | \& return $self->{ARRAY}->[$index]; |
| 419 | \& } |
| 420 | .Ve |
| 421 | .Sp |
| 422 | If a negative array index is used to read from an array, the index |
| 423 | will be translated to a positive one internally by calling \s-1FETCHSIZE\s0 |
| 424 | before being passed to \s-1FETCH\s0. You may disable this feature by |
| 425 | assigning a true value to the variable \f(CW$NEGATIVE_INDICES\fR in the |
| 426 | tied array class. |
| 427 | .Sp |
| 428 | As you may have noticed, the name of the \s-1FETCH\s0 method (et al.) is the same |
| 429 | for all accesses, even though the constructors differ in names (\s-1TIESCALAR\s0 |
| 430 | vs \s-1TIEARRAY\s0). While in theory you could have the same class servicing |
| 431 | several tied types, in practice this becomes cumbersome, and it's easiest |
| 432 | to keep them at simply one tie type per class. |
| 433 | .IP "\s-1STORE\s0 this, index, value" 4 |
| 434 | .IX Xref "STORE" |
| 435 | .IX Item "STORE this, index, value" |
| 436 | This method will be triggered every time an element in the tied array is set |
| 437 | (written). It takes two arguments beyond its self reference: the index at |
| 438 | which we're trying to store something and the value we're trying to put |
| 439 | there. |
| 440 | .Sp |
| 441 | In our example, \f(CW\*(C`undef\*(C'\fR is really \f(CW\*(C`$self\->{ELEMSIZE}\*(C'\fR number of |
| 442 | spaces so we have a little more work to do here: |
| 443 | .Sp |
| 444 | .Vb 11 |
| 445 | \& sub STORE { |
| 446 | \& my $self = shift; |
| 447 | \& my( $index, $value ) = @_; |
| 448 | \& if ( length $value > $self->{ELEMSIZE} ) { |
| 449 | \& croak "length of $value is greater than $self->{ELEMSIZE}"; |
| 450 | \& } |
| 451 | \& # fill in the blanks |
| 452 | \& $self->EXTEND( $index ) if $index > $self->FETCHSIZE(); |
| 453 | \& # right justify to keep element size for smaller elements |
| 454 | \& $self->{ARRAY}->[$index] = sprintf "%$self->{ELEMSIZE}s", $value; |
| 455 | \& } |
| 456 | .Ve |
| 457 | .Sp |
| 458 | Negative indexes are treated the same as with \s-1FETCH\s0. |
| 459 | .IP "\s-1FETCHSIZE\s0 this" 4 |
| 460 | .IX Xref "FETCHSIZE" |
| 461 | .IX Item "FETCHSIZE this" |
| 462 | Returns the total number of items in the tied array associated with |
| 463 | object \fIthis\fR. (Equivalent to \f(CW\*(C`scalar(@array)\*(C'\fR). For example: |
| 464 | .Sp |
| 465 | .Vb 4 |
| 466 | \& sub FETCHSIZE { |
| 467 | \& my $self = shift; |
| 468 | \& return scalar @{$self->{ARRAY}}; |
| 469 | \& } |
| 470 | .Ve |
| 471 | .IP "\s-1STORESIZE\s0 this, count" 4 |
| 472 | .IX Xref "STORESIZE" |
| 473 | .IX Item "STORESIZE this, count" |
| 474 | Sets the total number of items in the tied array associated with |
| 475 | object \fIthis\fR to be \fIcount\fR. If this makes the array larger then |
| 476 | class's mapping of \f(CW\*(C`undef\*(C'\fR should be returned for new positions. |
| 477 | If the array becomes smaller then entries beyond count should be |
| 478 | deleted. |
| 479 | .Sp |
| 480 | In our example, 'undef' is really an element containing |
| 481 | \&\f(CW\*(C`$self\->{ELEMSIZE}\*(C'\fR number of spaces. Observe: |
| 482 | .Sp |
| 483 | .Vb 13 |
| 484 | \& sub STORESIZE { |
| 485 | \& my $self = shift; |
| 486 | \& my $count = shift; |
| 487 | \& if ( $count > $self->FETCHSIZE() ) { |
| 488 | \& foreach ( $count - $self->FETCHSIZE() .. $count ) { |
| 489 | \& $self->STORE( $_, '' ); |
| 490 | \& } |
| 491 | \& } elsif ( $count < $self->FETCHSIZE() ) { |
| 492 | \& foreach ( 0 .. $self->FETCHSIZE() - $count - 2 ) { |
| 493 | \& $self->POP(); |
| 494 | \& } |
| 495 | \& } |
| 496 | \& } |
| 497 | .Ve |
| 498 | .IP "\s-1EXTEND\s0 this, count" 4 |
| 499 | .IX Xref "EXTEND" |
| 500 | .IX Item "EXTEND this, count" |
| 501 | Informative call that array is likely to grow to have \fIcount\fR entries. |
| 502 | Can be used to optimize allocation. This method need do nothing. |
| 503 | .Sp |
| 504 | In our example, we want to make sure there are no blank (\f(CW\*(C`undef\*(C'\fR) |
| 505 | entries, so \f(CW\*(C`EXTEND\*(C'\fR will make use of \f(CW\*(C`STORESIZE\*(C'\fR to fill elements |
| 506 | as needed: |
| 507 | .Sp |
| 508 | .Vb 5 |
| 509 | \& sub EXTEND { |
| 510 | \& my $self = shift; |
| 511 | \& my $count = shift; |
| 512 | \& $self->STORESIZE( $count ); |
| 513 | \& } |
| 514 | .Ve |
| 515 | .IP "\s-1EXISTS\s0 this, key" 4 |
| 516 | .IX Xref "EXISTS" |
| 517 | .IX Item "EXISTS this, key" |
| 518 | Verify that the element at index \fIkey\fR exists in the tied array \fIthis\fR. |
| 519 | .Sp |
| 520 | In our example, we will determine that if an element consists of |
| 521 | \&\f(CW\*(C`$self\->{ELEMSIZE}\*(C'\fR spaces only, it does not exist: |
| 522 | .Sp |
| 523 | .Vb 7 |
| 524 | \& sub EXISTS { |
| 525 | \& my $self = shift; |
| 526 | \& my $index = shift; |
| 527 | \& return 0 if ! defined $self->{ARRAY}->[$index] || |
| 528 | \& $self->{ARRAY}->[$index] eq ' ' x $self->{ELEMSIZE}; |
| 529 | \& return 1; |
| 530 | \& } |
| 531 | .Ve |
| 532 | .IP "\s-1DELETE\s0 this, key" 4 |
| 533 | .IX Xref "DELETE" |
| 534 | .IX Item "DELETE this, key" |
| 535 | Delete the element at index \fIkey\fR from the tied array \fIthis\fR. |
| 536 | .Sp |
| 537 | In our example, a deleted item is \f(CW\*(C`$self\->{ELEMSIZE}\*(C'\fR spaces: |
| 538 | .Sp |
| 539 | .Vb 5 |
| 540 | \& sub DELETE { |
| 541 | \& my $self = shift; |
| 542 | \& my $index = shift; |
| 543 | \& return $self->STORE( $index, '' ); |
| 544 | \& } |
| 545 | .Ve |
| 546 | .IP "\s-1CLEAR\s0 this" 4 |
| 547 | .IX Xref "CLEAR" |
| 548 | .IX Item "CLEAR this" |
| 549 | Clear (remove, delete, ...) all values from the tied array associated with |
| 550 | object \fIthis\fR. For example: |
| 551 | .Sp |
| 552 | .Vb 4 |
| 553 | \& sub CLEAR { |
| 554 | \& my $self = shift; |
| 555 | \& return $self->{ARRAY} = []; |
| 556 | \& } |
| 557 | .Ve |
| 558 | .IP "\s-1PUSH\s0 this, \s-1LIST\s0" 4 |
| 559 | .IX Xref "PUSH" |
| 560 | .IX Item "PUSH this, LIST" |
| 561 | Append elements of \fI\s-1LIST\s0\fR to the array. For example: |
| 562 | .Sp |
| 563 | .Vb 7 |
| 564 | \& sub PUSH { |
| 565 | \& my $self = shift; |
| 566 | \& my @list = @_; |
| 567 | \& my $last = $self->FETCHSIZE(); |
| 568 | \& $self->STORE( $last + $_, $list[$_] ) foreach 0 .. $#list; |
| 569 | \& return $self->FETCHSIZE(); |
| 570 | \& } |
| 571 | .Ve |
| 572 | .IP "\s-1POP\s0 this" 4 |
| 573 | .IX Xref "POP" |
| 574 | .IX Item "POP this" |
| 575 | Remove last element of the array and return it. For example: |
| 576 | .Sp |
| 577 | .Vb 4 |
| 578 | \& sub POP { |
| 579 | \& my $self = shift; |
| 580 | \& return pop @{$self->{ARRAY}}; |
| 581 | \& } |
| 582 | .Ve |
| 583 | .IP "\s-1SHIFT\s0 this" 4 |
| 584 | .IX Xref "SHIFT" |
| 585 | .IX Item "SHIFT this" |
| 586 | Remove the first element of the array (shifting other elements down) |
| 587 | and return it. For example: |
| 588 | .Sp |
| 589 | .Vb 4 |
| 590 | \& sub SHIFT { |
| 591 | \& my $self = shift; |
| 592 | \& return shift @{$self->{ARRAY}}; |
| 593 | \& } |
| 594 | .Ve |
| 595 | .IP "\s-1UNSHIFT\s0 this, \s-1LIST\s0" 4 |
| 596 | .IX Xref "UNSHIFT" |
| 597 | .IX Item "UNSHIFT this, LIST" |
| 598 | Insert \s-1LIST\s0 elements at the beginning of the array, moving existing elements |
| 599 | up to make room. For example: |
| 600 | .Sp |
| 601 | .Vb 9 |
| 602 | \& sub UNSHIFT { |
| 603 | \& my $self = shift; |
| 604 | \& my @list = @_; |
| 605 | \& my $size = scalar( @list ); |
| 606 | \& # make room for our list |
| 607 | \& @{$self->{ARRAY}}[ $size .. $#{$self->{ARRAY}} + $size ] |
| 608 | \& = @{$self->{ARRAY}}; |
| 609 | \& $self->STORE( $_, $list[$_] ) foreach 0 .. $#list; |
| 610 | \& } |
| 611 | .Ve |
| 612 | .IP "\s-1SPLICE\s0 this, offset, length, \s-1LIST\s0" 4 |
| 613 | .IX Xref "SPLICE" |
| 614 | .IX Item "SPLICE this, offset, length, LIST" |
| 615 | Perform the equivalent of \f(CW\*(C`splice\*(C'\fR on the array. |
| 616 | .Sp |
| 617 | \&\fIoffset\fR is optional and defaults to zero, negative values count back |
| 618 | from the end of the array. |
| 619 | .Sp |
| 620 | \&\fIlength\fR is optional and defaults to rest of the array. |
| 621 | .Sp |
| 622 | \&\fI\s-1LIST\s0\fR may be empty. |
| 623 | .Sp |
| 624 | Returns a list of the original \fIlength\fR elements at \fIoffset\fR. |
| 625 | .Sp |
| 626 | In our example, we'll use a little shortcut if there is a \fI\s-1LIST\s0\fR: |
| 627 | .Sp |
| 628 | .Vb 11 |
| 629 | \& sub SPLICE { |
| 630 | \& my $self = shift; |
| 631 | \& my $offset = shift || 0; |
| 632 | \& my $length = shift || $self->FETCHSIZE() - $offset; |
| 633 | \& my @list = (); |
| 634 | \& if ( @_ ) { |
| 635 | \& tie @list, __PACKAGE__, $self->{ELEMSIZE}; |
| 636 | \& @list = @_; |
| 637 | \& } |
| 638 | \& return splice @{$self->{ARRAY}}, $offset, $length, @list; |
| 639 | \& } |
| 640 | .Ve |
| 641 | .IP "\s-1UNTIE\s0 this" 4 |
| 642 | .IX Xref "UNTIE" |
| 643 | .IX Item "UNTIE this" |
| 644 | Will be called when \f(CW\*(C`untie\*(C'\fR happens. (See "The \f(CW\*(C`untie\*(C'\fR Gotcha" below.) |
| 645 | .IP "\s-1DESTROY\s0 this" 4 |
| 646 | .IX Xref "DESTROY" |
| 647 | .IX Item "DESTROY this" |
| 648 | This method will be triggered when the tied variable needs to be destructed. |
| 649 | As with the scalar tie class, this is almost never needed in a |
| 650 | language that does its own garbage collection, so this time we'll |
| 651 | just leave it out. |
| 652 | .Sh "Tying Hashes" |
| 653 | .IX Xref "hash, tying" |
| 654 | .IX Subsection "Tying Hashes" |
| 655 | Hashes were the first Perl data type to be tied (see \fIdbmopen()\fR). A class |
| 656 | implementing a tied hash should define the following methods: \s-1TIEHASH\s0 is |
| 657 | the constructor. \s-1FETCH\s0 and \s-1STORE\s0 access the key and value pairs. \s-1EXISTS\s0 |
| 658 | reports whether a key is present in the hash, and \s-1DELETE\s0 deletes one. |
| 659 | \&\s-1CLEAR\s0 empties the hash by deleting all the key and value pairs. \s-1FIRSTKEY\s0 |
| 660 | and \s-1NEXTKEY\s0 implement the \fIkeys()\fR and \fIeach()\fR functions to iterate over all |
| 661 | the keys. \s-1SCALAR\s0 is triggered when the tied hash is evaluated in scalar |
| 662 | context. \s-1UNTIE\s0 is called when \f(CW\*(C`untie\*(C'\fR happens, and \s-1DESTROY\s0 is called when |
| 663 | the tied variable is garbage collected. |
| 664 | .PP |
| 665 | If this seems like a lot, then feel free to inherit from merely the |
| 666 | standard Tie::StdHash module for most of your methods, redefining only the |
| 667 | interesting ones. See Tie::Hash for details. |
| 668 | .PP |
| 669 | Remember that Perl distinguishes between a key not existing in the hash, |
| 670 | and the key existing in the hash but having a corresponding value of |
| 671 | \&\f(CW\*(C`undef\*(C'\fR. The two possibilities can be tested with the \f(CW\*(C`exists()\*(C'\fR and |
| 672 | \&\f(CW\*(C`defined()\*(C'\fR functions. |
| 673 | .PP |
| 674 | Here's an example of a somewhat interesting tied hash class: it gives you |
| 675 | a hash representing a particular user's dot files. You index into the hash |
| 676 | with the name of the file (minus the dot) and you get back that dot file's |
| 677 | contents. For example: |
| 678 | .PP |
| 679 | .Vb 8 |
| 680 | \& use DotFiles; |
| 681 | \& tie %dot, 'DotFiles'; |
| 682 | \& if ( $dot{profile} =~ /MANPATH/ || |
| 683 | \& $dot{login} =~ /MANPATH/ || |
| 684 | \& $dot{cshrc} =~ /MANPATH/ ) |
| 685 | \& { |
| 686 | \& print "you seem to set your MANPATH\en"; |
| 687 | \& } |
| 688 | .Ve |
| 689 | .PP |
| 690 | Or here's another sample of using our tied class: |
| 691 | .PP |
| 692 | .Vb 5 |
| 693 | \& tie %him, 'DotFiles', 'daemon'; |
| 694 | \& foreach $f ( keys %him ) { |
| 695 | \& printf "daemon dot file %s is size %d\en", |
| 696 | \& $f, length $him{$f}; |
| 697 | \& } |
| 698 | .Ve |
| 699 | .PP |
| 700 | In our tied hash DotFiles example, we use a regular |
| 701 | hash for the object containing several important |
| 702 | fields, of which only the \f(CW\*(C`{LIST}\*(C'\fR field will be what the |
| 703 | user thinks of as the real hash. |
| 704 | .IP "\s-1USER\s0" 5 |
| 705 | .IX Item "USER" |
| 706 | whose dot files this object represents |
| 707 | .IP "\s-1HOME\s0" 5 |
| 708 | .IX Item "HOME" |
| 709 | where those dot files live |
| 710 | .IP "\s-1CLOBBER\s0" 5 |
| 711 | .IX Item "CLOBBER" |
| 712 | whether we should try to change or remove those dot files |
| 713 | .IP "\s-1LIST\s0" 5 |
| 714 | .IX Item "LIST" |
| 715 | the hash of dot file names and content mappings |
| 716 | .PP |
| 717 | Here's the start of \fIDotfiles.pm\fR: |
| 718 | .PP |
| 719 | .Vb 5 |
| 720 | \& package DotFiles; |
| 721 | \& use Carp; |
| 722 | \& sub whowasi { (caller(1))[3] . '()' } |
| 723 | \& my $DEBUG = 0; |
| 724 | \& sub debug { $DEBUG = @_ ? shift : 1 } |
| 725 | .Ve |
| 726 | .PP |
| 727 | For our example, we want to be able to emit debugging info to help in tracing |
| 728 | during development. We keep also one convenience function around |
| 729 | internally to help print out warnings; \fIwhowasi()\fR returns the function name |
| 730 | that calls it. |
| 731 | .PP |
| 732 | Here are the methods for the DotFiles tied hash. |
| 733 | .IP "\s-1TIEHASH\s0 classname, \s-1LIST\s0" 4 |
| 734 | .IX Xref "TIEHASH" |
| 735 | .IX Item "TIEHASH classname, LIST" |
| 736 | This is the constructor for the class. That means it is expected to |
| 737 | return a blessed reference through which the new object (probably but not |
| 738 | necessarily an anonymous hash) will be accessed. |
| 739 | .Sp |
| 740 | Here's the constructor: |
| 741 | .Sp |
| 742 | .Vb 9 |
| 743 | \& sub TIEHASH { |
| 744 | \& my $self = shift; |
| 745 | \& my $user = shift || $>; |
| 746 | \& my $dotdir = shift || ''; |
| 747 | \& croak "usage: @{[&whowasi]} [USER [DOTDIR]]" if @_; |
| 748 | \& $user = getpwuid($user) if $user =~ /^\ed+$/; |
| 749 | \& my $dir = (getpwnam($user))[7] |
| 750 | \& || croak "@{[&whowasi]}: no user $user"; |
| 751 | \& $dir .= "/$dotdir" if $dotdir; |
| 752 | .Ve |
| 753 | .Sp |
| 754 | .Vb 6 |
| 755 | \& my $node = { |
| 756 | \& USER => $user, |
| 757 | \& HOME => $dir, |
| 758 | \& LIST => {}, |
| 759 | \& CLOBBER => 0, |
| 760 | \& }; |
| 761 | .Ve |
| 762 | .Sp |
| 763 | .Vb 9 |
| 764 | \& opendir(DIR, $dir) |
| 765 | \& || croak "@{[&whowasi]}: can't opendir $dir: $!"; |
| 766 | \& foreach $dot ( grep /^\e./ && -f "$dir/$_", readdir(DIR)) { |
| 767 | \& $dot =~ s/^\e.//; |
| 768 | \& $node->{LIST}{$dot} = undef; |
| 769 | \& } |
| 770 | \& closedir DIR; |
| 771 | \& return bless $node, $self; |
| 772 | \& } |
| 773 | .Ve |
| 774 | .Sp |
| 775 | It's probably worth mentioning that if you're going to filetest the |
| 776 | return values out of a readdir, you'd better prepend the directory |
| 777 | in question. Otherwise, because we didn't \fIchdir()\fR there, it would |
| 778 | have been testing the wrong file. |
| 779 | .IP "\s-1FETCH\s0 this, key" 4 |
| 780 | .IX Xref "FETCH" |
| 781 | .IX Item "FETCH this, key" |
| 782 | This method will be triggered every time an element in the tied hash is |
| 783 | accessed (read). It takes one argument beyond its self reference: the key |
| 784 | whose value we're trying to fetch. |
| 785 | .Sp |
| 786 | Here's the fetch for our DotFiles example. |
| 787 | .Sp |
| 788 | .Vb 6 |
| 789 | \& sub FETCH { |
| 790 | \& carp &whowasi if $DEBUG; |
| 791 | \& my $self = shift; |
| 792 | \& my $dot = shift; |
| 793 | \& my $dir = $self->{HOME}; |
| 794 | \& my $file = "$dir/.$dot"; |
| 795 | .Ve |
| 796 | .Sp |
| 797 | .Vb 4 |
| 798 | \& unless (exists $self->{LIST}->{$dot} || -f $file) { |
| 799 | \& carp "@{[&whowasi]}: no $dot file" if $DEBUG; |
| 800 | \& return undef; |
| 801 | \& } |
| 802 | .Ve |
| 803 | .Sp |
| 804 | .Vb 6 |
| 805 | \& if (defined $self->{LIST}->{$dot}) { |
| 806 | \& return $self->{LIST}->{$dot}; |
| 807 | \& } else { |
| 808 | \& return $self->{LIST}->{$dot} = `cat $dir/.$dot`; |
| 809 | \& } |
| 810 | \& } |
| 811 | .Ve |
| 812 | .Sp |
| 813 | It was easy to write by having it call the Unix \fIcat\fR\|(1) command, but it |
| 814 | would probably be more portable to open the file manually (and somewhat |
| 815 | more efficient). Of course, because dot files are a Unixy concept, we're |
| 816 | not that concerned. |
| 817 | .IP "\s-1STORE\s0 this, key, value" 4 |
| 818 | .IX Xref "STORE" |
| 819 | .IX Item "STORE this, key, value" |
| 820 | This method will be triggered every time an element in the tied hash is set |
| 821 | (written). It takes two arguments beyond its self reference: the index at |
| 822 | which we're trying to store something, and the value we're trying to put |
| 823 | there. |
| 824 | .Sp |
| 825 | Here in our DotFiles example, we'll be careful not to let |
| 826 | them try to overwrite the file unless they've called the \fIclobber()\fR |
| 827 | method on the original object reference returned by \fItie()\fR. |
| 828 | .Sp |
| 829 | .Vb 7 |
| 830 | \& sub STORE { |
| 831 | \& carp &whowasi if $DEBUG; |
| 832 | \& my $self = shift; |
| 833 | \& my $dot = shift; |
| 834 | \& my $value = shift; |
| 835 | \& my $file = $self->{HOME} . "/.$dot"; |
| 836 | \& my $user = $self->{USER}; |
| 837 | .Ve |
| 838 | .Sp |
| 839 | .Vb 2 |
| 840 | \& croak "@{[&whowasi]}: $file not clobberable" |
| 841 | \& unless $self->{CLOBBER}; |
| 842 | .Ve |
| 843 | .Sp |
| 844 | .Vb 4 |
| 845 | \& open(F, "> $file") || croak "can't open $file: $!"; |
| 846 | \& print F $value; |
| 847 | \& close(F); |
| 848 | \& } |
| 849 | .Ve |
| 850 | .Sp |
| 851 | If they wanted to clobber something, they might say: |
| 852 | .Sp |
| 853 | .Vb 3 |
| 854 | \& $ob = tie %daemon_dots, 'daemon'; |
| 855 | \& $ob->clobber(1); |
| 856 | \& $daemon_dots{signature} = "A true daemon\en"; |
| 857 | .Ve |
| 858 | .Sp |
| 859 | Another way to lay hands on a reference to the underlying object is to |
| 860 | use the \fItied()\fR function, so they might alternately have set clobber |
| 861 | using: |
| 862 | .Sp |
| 863 | .Vb 2 |
| 864 | \& tie %daemon_dots, 'daemon'; |
| 865 | \& tied(%daemon_dots)->clobber(1); |
| 866 | .Ve |
| 867 | .Sp |
| 868 | The clobber method is simply: |
| 869 | .Sp |
| 870 | .Vb 4 |
| 871 | \& sub clobber { |
| 872 | \& my $self = shift; |
| 873 | \& $self->{CLOBBER} = @_ ? shift : 1; |
| 874 | \& } |
| 875 | .Ve |
| 876 | .IP "\s-1DELETE\s0 this, key" 4 |
| 877 | .IX Xref "DELETE" |
| 878 | .IX Item "DELETE this, key" |
| 879 | This method is triggered when we remove an element from the hash, |
| 880 | typically by using the \fIdelete()\fR function. Again, we'll |
| 881 | be careful to check whether they really want to clobber files. |
| 882 | .Sp |
| 883 | .Vb 2 |
| 884 | \& sub DELETE { |
| 885 | \& carp &whowasi if $DEBUG; |
| 886 | .Ve |
| 887 | .Sp |
| 888 | .Vb 10 |
| 889 | \& my $self = shift; |
| 890 | \& my $dot = shift; |
| 891 | \& my $file = $self->{HOME} . "/.$dot"; |
| 892 | \& croak "@{[&whowasi]}: won't remove file $file" |
| 893 | \& unless $self->{CLOBBER}; |
| 894 | \& delete $self->{LIST}->{$dot}; |
| 895 | \& my $success = unlink($file); |
| 896 | \& carp "@{[&whowasi]}: can't unlink $file: $!" unless $success; |
| 897 | \& $success; |
| 898 | \& } |
| 899 | .Ve |
| 900 | .Sp |
| 901 | The value returned by \s-1DELETE\s0 becomes the return value of the call |
| 902 | to \fIdelete()\fR. If you want to emulate the normal behavior of \fIdelete()\fR, |
| 903 | you should return whatever \s-1FETCH\s0 would have returned for this key. |
| 904 | In this example, we have chosen instead to return a value which tells |
| 905 | the caller whether the file was successfully deleted. |
| 906 | .IP "\s-1CLEAR\s0 this" 4 |
| 907 | .IX Xref "CLEAR" |
| 908 | .IX Item "CLEAR this" |
| 909 | This method is triggered when the whole hash is to be cleared, usually by |
| 910 | assigning the empty list to it. |
| 911 | .Sp |
| 912 | In our example, that would remove all the user's dot files! It's such a |
| 913 | dangerous thing that they'll have to set \s-1CLOBBER\s0 to something higher than |
| 914 | 1 to make it happen. |
| 915 | .Sp |
| 916 | .Vb 10 |
| 917 | \& sub CLEAR { |
| 918 | \& carp &whowasi if $DEBUG; |
| 919 | \& my $self = shift; |
| 920 | \& croak "@{[&whowasi]}: won't remove all dot files for $self->{USER}" |
| 921 | \& unless $self->{CLOBBER} > 1; |
| 922 | \& my $dot; |
| 923 | \& foreach $dot ( keys %{$self->{LIST}}) { |
| 924 | \& $self->DELETE($dot); |
| 925 | \& } |
| 926 | \& } |
| 927 | .Ve |
| 928 | .IP "\s-1EXISTS\s0 this, key" 4 |
| 929 | .IX Xref "EXISTS" |
| 930 | .IX Item "EXISTS this, key" |
| 931 | This method is triggered when the user uses the \fIexists()\fR function |
| 932 | on a particular hash. In our example, we'll look at the \f(CW\*(C`{LIST}\*(C'\fR |
| 933 | hash element for this: |
| 934 | .Sp |
| 935 | .Vb 6 |
| 936 | \& sub EXISTS { |
| 937 | \& carp &whowasi if $DEBUG; |
| 938 | \& my $self = shift; |
| 939 | \& my $dot = shift; |
| 940 | \& return exists $self->{LIST}->{$dot}; |
| 941 | \& } |
| 942 | .Ve |
| 943 | .IP "\s-1FIRSTKEY\s0 this" 4 |
| 944 | .IX Xref "FIRSTKEY" |
| 945 | .IX Item "FIRSTKEY this" |
| 946 | This method will be triggered when the user is going |
| 947 | to iterate through the hash, such as via a \fIkeys()\fR or \fIeach()\fR |
| 948 | call. |
| 949 | .Sp |
| 950 | .Vb 6 |
| 951 | \& sub FIRSTKEY { |
| 952 | \& carp &whowasi if $DEBUG; |
| 953 | \& my $self = shift; |
| 954 | \& my $a = keys %{$self->{LIST}}; # reset each() iterator |
| 955 | \& each %{$self->{LIST}} |
| 956 | \& } |
| 957 | .Ve |
| 958 | .IP "\s-1NEXTKEY\s0 this, lastkey" 4 |
| 959 | .IX Xref "NEXTKEY" |
| 960 | .IX Item "NEXTKEY this, lastkey" |
| 961 | This method gets triggered during a \fIkeys()\fR or \fIeach()\fR iteration. It has a |
| 962 | second argument which is the last key that had been accessed. This is |
| 963 | useful if you're carrying about ordering or calling the iterator from more |
| 964 | than one sequence, or not really storing things in a hash anywhere. |
| 965 | .Sp |
| 966 | For our example, we're using a real hash so we'll do just the simple |
| 967 | thing, but we'll have to go through the \s-1LIST\s0 field indirectly. |
| 968 | .Sp |
| 969 | .Vb 5 |
| 970 | \& sub NEXTKEY { |
| 971 | \& carp &whowasi if $DEBUG; |
| 972 | \& my $self = shift; |
| 973 | \& return each %{ $self->{LIST} } |
| 974 | \& } |
| 975 | .Ve |
| 976 | .IP "\s-1SCALAR\s0 this" 4 |
| 977 | .IX Xref "SCALAR" |
| 978 | .IX Item "SCALAR this" |
| 979 | This is called when the hash is evaluated in scalar context. In order |
| 980 | to mimic the behaviour of untied hashes, this method should return a |
| 981 | false value when the tied hash is considered empty. If this method does |
| 982 | not exist, perl will make some educated guesses and return true when |
| 983 | the hash is inside an iteration. If this isn't the case, \s-1FIRSTKEY\s0 is |
| 984 | called, and the result will be a false value if \s-1FIRSTKEY\s0 returns the empty |
| 985 | list, true otherwise. |
| 986 | .Sp |
| 987 | However, you should \fBnot\fR blindly rely on perl always doing the right |
| 988 | thing. Particularly, perl will mistakenly return true when you clear the |
| 989 | hash by repeatedly calling \s-1DELETE\s0 until it is empty. You are therefore |
| 990 | advised to supply your own \s-1SCALAR\s0 method when you want to be absolutely |
| 991 | sure that your hash behaves nicely in scalar context. |
| 992 | .Sp |
| 993 | In our example we can just call \f(CW\*(C`scalar\*(C'\fR on the underlying hash |
| 994 | referenced by \f(CW\*(C`$self\->{LIST}\*(C'\fR: |
| 995 | .Sp |
| 996 | .Vb 5 |
| 997 | \& sub SCALAR { |
| 998 | \& carp &whowasi if $DEBUG; |
| 999 | \& my $self = shift; |
| 1000 | \& return scalar %{ $self->{LIST} } |
| 1001 | \& } |
| 1002 | .Ve |
| 1003 | .IP "\s-1UNTIE\s0 this" 4 |
| 1004 | .IX Xref "UNTIE" |
| 1005 | .IX Item "UNTIE this" |
| 1006 | This is called when \f(CW\*(C`untie\*(C'\fR occurs. See "The \f(CW\*(C`untie\*(C'\fR Gotcha" below. |
| 1007 | .IP "\s-1DESTROY\s0 this" 4 |
| 1008 | .IX Xref "DESTROY" |
| 1009 | .IX Item "DESTROY this" |
| 1010 | This method is triggered when a tied hash is about to go out of |
| 1011 | scope. You don't really need it unless you're trying to add debugging |
| 1012 | or have auxiliary state to clean up. Here's a very simple function: |
| 1013 | .Sp |
| 1014 | .Vb 3 |
| 1015 | \& sub DESTROY { |
| 1016 | \& carp &whowasi if $DEBUG; |
| 1017 | \& } |
| 1018 | .Ve |
| 1019 | .PP |
| 1020 | Note that functions such as \fIkeys()\fR and \fIvalues()\fR may return huge lists |
| 1021 | when used on large objects, like \s-1DBM\s0 files. You may prefer to use the |
| 1022 | \&\fIeach()\fR function to iterate over such. Example: |
| 1023 | .PP |
| 1024 | .Vb 7 |
| 1025 | \& # print out history file offsets |
| 1026 | \& use NDBM_File; |
| 1027 | \& tie(%HIST, 'NDBM_File', '/usr/lib/news/history', 1, 0); |
| 1028 | \& while (($key,$val) = each %HIST) { |
| 1029 | \& print $key, ' = ', unpack('L',$val), "\en"; |
| 1030 | \& } |
| 1031 | \& untie(%HIST); |
| 1032 | .Ve |
| 1033 | .Sh "Tying FileHandles" |
| 1034 | .IX Xref "filehandle, tying" |
| 1035 | .IX Subsection "Tying FileHandles" |
| 1036 | This is partially implemented now. |
| 1037 | .PP |
| 1038 | A class implementing a tied filehandle should define the following |
| 1039 | methods: \s-1TIEHANDLE\s0, at least one of \s-1PRINT\s0, \s-1PRINTF\s0, \s-1WRITE\s0, \s-1READLINE\s0, \s-1GETC\s0, |
| 1040 | \&\s-1READ\s0, and possibly \s-1CLOSE\s0, \s-1UNTIE\s0 and \s-1DESTROY\s0. The class can also provide: \s-1BINMODE\s0, |
| 1041 | \&\s-1OPEN\s0, \s-1EOF\s0, \s-1FILENO\s0, \s-1SEEK\s0, \s-1TELL\s0 \- if the corresponding perl operators are |
| 1042 | used on the handle. |
| 1043 | .PP |
| 1044 | When \s-1STDERR\s0 is tied, its \s-1PRINT\s0 method will be called to issue warnings |
| 1045 | and error messages. This feature is temporarily disabled during the call, |
| 1046 | which means you can use \f(CW\*(C`warn()\*(C'\fR inside \s-1PRINT\s0 without starting a recursive |
| 1047 | loop. And just like \f(CW\*(C`_\|_WARN_\|_\*(C'\fR and \f(CW\*(C`_\|_DIE_\|_\*(C'\fR handlers, \s-1STDERR\s0's \s-1PRINT\s0 |
| 1048 | method may be called to report parser errors, so the caveats mentioned under |
| 1049 | \&\*(L"%SIG\*(R" in perlvar apply. |
| 1050 | .PP |
| 1051 | All of this is especially useful when perl is embedded in some other |
| 1052 | program, where output to \s-1STDOUT\s0 and \s-1STDERR\s0 may have to be redirected |
| 1053 | in some special way. See nvi and the Apache module for examples. |
| 1054 | .PP |
| 1055 | In our example we're going to create a shouting handle. |
| 1056 | .PP |
| 1057 | .Vb 1 |
| 1058 | \& package Shout; |
| 1059 | .Ve |
| 1060 | .IP "\s-1TIEHANDLE\s0 classname, \s-1LIST\s0" 4 |
| 1061 | .IX Xref "TIEHANDLE" |
| 1062 | .IX Item "TIEHANDLE classname, LIST" |
| 1063 | This is the constructor for the class. That means it is expected to |
| 1064 | return a blessed reference of some sort. The reference can be used to |
| 1065 | hold some internal information. |
| 1066 | .Sp |
| 1067 | .Vb 1 |
| 1068 | \& sub TIEHANDLE { print "<shout>\en"; my $i; bless \e$i, shift } |
| 1069 | .Ve |
| 1070 | .IP "\s-1WRITE\s0 this, \s-1LIST\s0" 4 |
| 1071 | .IX Xref "WRITE" |
| 1072 | .IX Item "WRITE this, LIST" |
| 1073 | This method will be called when the handle is written to via the |
| 1074 | \&\f(CW\*(C`syswrite\*(C'\fR function. |
| 1075 | .Sp |
| 1076 | .Vb 5 |
| 1077 | \& sub WRITE { |
| 1078 | \& $r = shift; |
| 1079 | \& my($buf,$len,$offset) = @_; |
| 1080 | \& print "WRITE called, \e$buf=$buf, \e$len=$len, \e$offset=$offset"; |
| 1081 | \& } |
| 1082 | .Ve |
| 1083 | .IP "\s-1PRINT\s0 this, \s-1LIST\s0" 4 |
| 1084 | .IX Xref "PRINT" |
| 1085 | .IX Item "PRINT this, LIST" |
| 1086 | This method will be triggered every time the tied handle is printed to |
| 1087 | with the \f(CW\*(C`print()\*(C'\fR function. |
| 1088 | Beyond its self reference it also expects the list that was passed to |
| 1089 | the print function. |
| 1090 | .Sp |
| 1091 | .Vb 1 |
| 1092 | \& sub PRINT { $r = shift; $$r++; print join($,,map(uc($_),@_)),$\e } |
| 1093 | .Ve |
| 1094 | .IP "\s-1PRINTF\s0 this, \s-1LIST\s0" 4 |
| 1095 | .IX Xref "PRINTF" |
| 1096 | .IX Item "PRINTF this, LIST" |
| 1097 | This method will be triggered every time the tied handle is printed to |
| 1098 | with the \f(CW\*(C`printf()\*(C'\fR function. |
| 1099 | Beyond its self reference it also expects the format and list that was |
| 1100 | passed to the printf function. |
| 1101 | .Sp |
| 1102 | .Vb 5 |
| 1103 | \& sub PRINTF { |
| 1104 | \& shift; |
| 1105 | \& my $fmt = shift; |
| 1106 | \& print sprintf($fmt, @_); |
| 1107 | \& } |
| 1108 | .Ve |
| 1109 | .IP "\s-1READ\s0 this, \s-1LIST\s0" 4 |
| 1110 | .IX Xref "READ" |
| 1111 | .IX Item "READ this, LIST" |
| 1112 | This method will be called when the handle is read from via the \f(CW\*(C`read\*(C'\fR |
| 1113 | or \f(CW\*(C`sysread\*(C'\fR functions. |
| 1114 | .Sp |
| 1115 | .Vb 8 |
| 1116 | \& sub READ { |
| 1117 | \& my $self = shift; |
| 1118 | \& my $bufref = \e$_[0]; |
| 1119 | \& my(undef,$len,$offset) = @_; |
| 1120 | \& print "READ called, \e$buf=$bufref, \e$len=$len, \e$offset=$offset"; |
| 1121 | \& # add to $$bufref, set $len to number of characters read |
| 1122 | \& $len; |
| 1123 | \& } |
| 1124 | .Ve |
| 1125 | .IP "\s-1READLINE\s0 this" 4 |
| 1126 | .IX Xref "READLINE" |
| 1127 | .IX Item "READLINE this" |
| 1128 | This method will be called when the handle is read from via <\s-1HANDLE\s0>. |
| 1129 | The method should return undef when there is no more data. |
| 1130 | .Sp |
| 1131 | .Vb 1 |
| 1132 | \& sub READLINE { $r = shift; "READLINE called $$r times\en"; } |
| 1133 | .Ve |
| 1134 | .IP "\s-1GETC\s0 this" 4 |
| 1135 | .IX Xref "GETC" |
| 1136 | .IX Item "GETC this" |
| 1137 | This method will be called when the \f(CW\*(C`getc\*(C'\fR function is called. |
| 1138 | .Sp |
| 1139 | .Vb 1 |
| 1140 | \& sub GETC { print "Don't GETC, Get Perl"; return "a"; } |
| 1141 | .Ve |
| 1142 | .IP "\s-1CLOSE\s0 this" 4 |
| 1143 | .IX Xref "CLOSE" |
| 1144 | .IX Item "CLOSE this" |
| 1145 | This method will be called when the handle is closed via the \f(CW\*(C`close\*(C'\fR |
| 1146 | function. |
| 1147 | .Sp |
| 1148 | .Vb 1 |
| 1149 | \& sub CLOSE { print "CLOSE called.\en" } |
| 1150 | .Ve |
| 1151 | .IP "\s-1UNTIE\s0 this" 4 |
| 1152 | .IX Xref "UNTIE" |
| 1153 | .IX Item "UNTIE this" |
| 1154 | As with the other types of ties, this method will be called when \f(CW\*(C`untie\*(C'\fR happens. |
| 1155 | It may be appropriate to \*(L"auto \s-1CLOSE\s0\*(R" when this occurs. See |
| 1156 | "The \f(CW\*(C`untie\*(C'\fR Gotcha" below. |
| 1157 | .IP "\s-1DESTROY\s0 this" 4 |
| 1158 | .IX Xref "DESTROY" |
| 1159 | .IX Item "DESTROY this" |
| 1160 | As with the other types of ties, this method will be called when the |
| 1161 | tied handle is about to be destroyed. This is useful for debugging and |
| 1162 | possibly cleaning up. |
| 1163 | .Sp |
| 1164 | .Vb 1 |
| 1165 | \& sub DESTROY { print "</shout>\en" } |
| 1166 | .Ve |
| 1167 | .PP |
| 1168 | Here's how to use our little example: |
| 1169 | .PP |
| 1170 | .Vb 5 |
| 1171 | \& tie(*FOO,'Shout'); |
| 1172 | \& print FOO "hello\en"; |
| 1173 | \& $a = 4; $b = 6; |
| 1174 | \& print FOO $a, " plus ", $b, " equals ", $a + $b, "\en"; |
| 1175 | \& print <FOO>; |
| 1176 | .Ve |
| 1177 | .Sh "\s-1UNTIE\s0 this" |
| 1178 | .IX Xref "UNTIE" |
| 1179 | .IX Subsection "UNTIE this" |
| 1180 | You can define for all tie types an \s-1UNTIE\s0 method that will be called |
| 1181 | at \fIuntie()\fR. See "The \f(CW\*(C`untie\*(C'\fR Gotcha" below. |
| 1182 | .ie n .Sh "The ""untie"" Gotcha" |
| 1183 | .el .Sh "The \f(CWuntie\fP Gotcha" |
| 1184 | .IX Xref "untie" |
| 1185 | .IX Subsection "The untie Gotcha" |
| 1186 | If you intend making use of the object returned from either \fItie()\fR or |
| 1187 | \&\fItied()\fR, and if the tie's target class defines a destructor, there is a |
| 1188 | subtle gotcha you \fImust\fR guard against. |
| 1189 | .PP |
| 1190 | As setup, consider this (admittedly rather contrived) example of a |
| 1191 | tie; all it does is use a file to keep a log of the values assigned to |
| 1192 | a scalar. |
| 1193 | .PP |
| 1194 | .Vb 1 |
| 1195 | \& package Remember; |
| 1196 | .Ve |
| 1197 | .PP |
| 1198 | .Vb 3 |
| 1199 | \& use strict; |
| 1200 | \& use warnings; |
| 1201 | \& use IO::File; |
| 1202 | .Ve |
| 1203 | .PP |
| 1204 | .Vb 5 |
| 1205 | \& sub TIESCALAR { |
| 1206 | \& my $class = shift; |
| 1207 | \& my $filename = shift; |
| 1208 | \& my $handle = new IO::File "> $filename" |
| 1209 | \& or die "Cannot open $filename: $!\en"; |
| 1210 | .Ve |
| 1211 | .PP |
| 1212 | .Vb 3 |
| 1213 | \& print $handle "The Start\en"; |
| 1214 | \& bless {FH => $handle, Value => 0}, $class; |
| 1215 | \& } |
| 1216 | .Ve |
| 1217 | .PP |
| 1218 | .Vb 4 |
| 1219 | \& sub FETCH { |
| 1220 | \& my $self = shift; |
| 1221 | \& return $self->{Value}; |
| 1222 | \& } |
| 1223 | .Ve |
| 1224 | .PP |
| 1225 | .Vb 7 |
| 1226 | \& sub STORE { |
| 1227 | \& my $self = shift; |
| 1228 | \& my $value = shift; |
| 1229 | \& my $handle = $self->{FH}; |
| 1230 | \& print $handle "$value\en"; |
| 1231 | \& $self->{Value} = $value; |
| 1232 | \& } |
| 1233 | .Ve |
| 1234 | .PP |
| 1235 | .Vb 6 |
| 1236 | \& sub DESTROY { |
| 1237 | \& my $self = shift; |
| 1238 | \& my $handle = $self->{FH}; |
| 1239 | \& print $handle "The End\en"; |
| 1240 | \& close $handle; |
| 1241 | \& } |
| 1242 | .Ve |
| 1243 | .PP |
| 1244 | .Vb 1 |
| 1245 | \& 1; |
| 1246 | .Ve |
| 1247 | .PP |
| 1248 | Here is an example that makes use of this tie: |
| 1249 | .PP |
| 1250 | .Vb 2 |
| 1251 | \& use strict; |
| 1252 | \& use Remember; |
| 1253 | .Ve |
| 1254 | .PP |
| 1255 | .Vb 7 |
| 1256 | \& my $fred; |
| 1257 | \& tie $fred, 'Remember', 'myfile.txt'; |
| 1258 | \& $fred = 1; |
| 1259 | \& $fred = 4; |
| 1260 | \& $fred = 5; |
| 1261 | \& untie $fred; |
| 1262 | \& system "cat myfile.txt"; |
| 1263 | .Ve |
| 1264 | .PP |
| 1265 | This is the output when it is executed: |
| 1266 | .PP |
| 1267 | .Vb 5 |
| 1268 | \& The Start |
| 1269 | \& 1 |
| 1270 | \& 4 |
| 1271 | \& 5 |
| 1272 | \& The End |
| 1273 | .Ve |
| 1274 | .PP |
| 1275 | So far so good. Those of you who have been paying attention will have |
| 1276 | spotted that the tied object hasn't been used so far. So lets add an |
| 1277 | extra method to the Remember class to allow comments to be included in |
| 1278 | the file \*(-- say, something like this: |
| 1279 | .PP |
| 1280 | .Vb 6 |
| 1281 | \& sub comment { |
| 1282 | \& my $self = shift; |
| 1283 | \& my $text = shift; |
| 1284 | \& my $handle = $self->{FH}; |
| 1285 | \& print $handle $text, "\en"; |
| 1286 | \& } |
| 1287 | .Ve |
| 1288 | .PP |
| 1289 | And here is the previous example modified to use the \f(CW\*(C`comment\*(C'\fR method |
| 1290 | (which requires the tied object): |
| 1291 | .PP |
| 1292 | .Vb 2 |
| 1293 | \& use strict; |
| 1294 | \& use Remember; |
| 1295 | .Ve |
| 1296 | .PP |
| 1297 | .Vb 8 |
| 1298 | \& my ($fred, $x); |
| 1299 | \& $x = tie $fred, 'Remember', 'myfile.txt'; |
| 1300 | \& $fred = 1; |
| 1301 | \& $fred = 4; |
| 1302 | \& comment $x "changing..."; |
| 1303 | \& $fred = 5; |
| 1304 | \& untie $fred; |
| 1305 | \& system "cat myfile.txt"; |
| 1306 | .Ve |
| 1307 | .PP |
| 1308 | When this code is executed there is no output. Here's why: |
| 1309 | .PP |
| 1310 | When a variable is tied, it is associated with the object which is the |
| 1311 | return value of the \s-1TIESCALAR\s0, \s-1TIEARRAY\s0, or \s-1TIEHASH\s0 function. This |
| 1312 | object normally has only one reference, namely, the implicit reference |
| 1313 | from the tied variable. When \fIuntie()\fR is called, that reference is |
| 1314 | destroyed. Then, as in the first example above, the object's |
| 1315 | destructor (\s-1DESTROY\s0) is called, which is normal for objects that have |
| 1316 | no more valid references; and thus the file is closed. |
| 1317 | .PP |
| 1318 | In the second example, however, we have stored another reference to |
| 1319 | the tied object in \f(CW$x\fR. That means that when \fIuntie()\fR gets called |
| 1320 | there will still be a valid reference to the object in existence, so |
| 1321 | the destructor is not called at that time, and thus the file is not |
| 1322 | closed. The reason there is no output is because the file buffers |
| 1323 | have not been flushed to disk. |
| 1324 | .PP |
| 1325 | Now that you know what the problem is, what can you do to avoid it? |
| 1326 | Prior to the introduction of the optional \s-1UNTIE\s0 method the only way |
| 1327 | was the good old \f(CW\*(C`\-w\*(C'\fR flag. Which will spot any instances where you call |
| 1328 | \&\fIuntie()\fR and there are still valid references to the tied object. If |
| 1329 | the second script above this near the top \f(CW\*(C`use warnings 'untie'\*(C'\fR |
| 1330 | or was run with the \f(CW\*(C`\-w\*(C'\fR flag, Perl prints this |
| 1331 | warning message: |
| 1332 | .PP |
| 1333 | .Vb 1 |
| 1334 | \& untie attempted while 1 inner references still exist |
| 1335 | .Ve |
| 1336 | .PP |
| 1337 | To get the script to work properly and silence the warning make sure |
| 1338 | there are no valid references to the tied object \fIbefore\fR \fIuntie()\fR is |
| 1339 | called: |
| 1340 | .PP |
| 1341 | .Vb 2 |
| 1342 | \& undef $x; |
| 1343 | \& untie $fred; |
| 1344 | .Ve |
| 1345 | .PP |
| 1346 | Now that \s-1UNTIE\s0 exists the class designer can decide which parts of the |
| 1347 | class functionality are really associated with \f(CW\*(C`untie\*(C'\fR and which with |
| 1348 | the object being destroyed. What makes sense for a given class depends |
| 1349 | on whether the inner references are being kept so that non-tie-related |
| 1350 | methods can be called on the object. But in most cases it probably makes |
| 1351 | sense to move the functionality that would have been in \s-1DESTROY\s0 to the \s-1UNTIE\s0 |
| 1352 | method. |
| 1353 | .PP |
| 1354 | If the \s-1UNTIE\s0 method exists then the warning above does not occur. Instead the |
| 1355 | \&\s-1UNTIE\s0 method is passed the count of \*(L"extra\*(R" references and can issue its own |
| 1356 | warning if appropriate. e.g. to replicate the no \s-1UNTIE\s0 case this method can |
| 1357 | be used: |
| 1358 | .PP |
| 1359 | .Vb 5 |
| 1360 | \& sub UNTIE |
| 1361 | \& { |
| 1362 | \& my ($obj,$count) = @_; |
| 1363 | \& carp "untie attempted while $count inner references still exist" if $count; |
| 1364 | \& } |
| 1365 | .Ve |
| 1366 | .SH "SEE ALSO" |
| 1367 | .IX Header "SEE ALSO" |
| 1368 | See DB_File or Config for some interesting \fItie()\fR implementations. |
| 1369 | A good starting point for many \fItie()\fR implementations is with one of the |
| 1370 | modules Tie::Scalar, Tie::Array, Tie::Hash, or Tie::Handle. |
| 1371 | .SH "BUGS" |
| 1372 | .IX Header "BUGS" |
| 1373 | The bucket usage information provided by \f(CW\*(C`scalar(%hash)\*(C'\fR is not |
| 1374 | available. What this means is that using \f(CW%tied_hash\fR in boolean |
| 1375 | context doesn't work right (currently this always tests false, |
| 1376 | regardless of whether the hash is empty or hash elements). |
| 1377 | .PP |
| 1378 | Localizing tied arrays or hashes does not work. After exiting the |
| 1379 | scope the arrays or the hashes are not restored. |
| 1380 | .PP |
| 1381 | Counting the number of entries in a hash via \f(CW\*(C`scalar(keys(%hash))\*(C'\fR |
| 1382 | or \f(CW\*(C`scalar(values(%hash)\*(C'\fR) is inefficient since it needs to iterate |
| 1383 | through all the entries with \s-1FIRSTKEY/NEXTKEY\s0. |
| 1384 | .PP |
| 1385 | Tied hash/array slices cause multiple \s-1FETCH/STORE\s0 pairs, there are no |
| 1386 | tie methods for slice operations. |
| 1387 | .PP |
| 1388 | You cannot easily tie a multilevel data structure (such as a hash of |
| 1389 | hashes) to a dbm file. The first problem is that all but \s-1GDBM\s0 and |
| 1390 | Berkeley \s-1DB\s0 have size limitations, but beyond that, you also have problems |
| 1391 | with how references are to be represented on disk. One experimental |
| 1392 | module that does attempt to address this need is DBM::Deep. Check your |
| 1393 | nearest \s-1CPAN\s0 site as described in perlmodlib for source code. Note |
| 1394 | that despite its name, DBM::Deep does not use dbm. Another earlier attempt |
| 1395 | at solving the problem is \s-1MLDBM\s0, which is also available on the \s-1CPAN\s0, but |
| 1396 | which has some fairly serious limitations. |
| 1397 | .PP |
| 1398 | Tied filehandles are still incomplete. \fIsysopen()\fR, \fItruncate()\fR, |
| 1399 | \&\fIflock()\fR, \fIfcntl()\fR, \fIstat()\fR and \-X can't currently be trapped. |
| 1400 | .SH "AUTHOR" |
| 1401 | .IX Header "AUTHOR" |
| 1402 | Tom Christiansen |
| 1403 | .PP |
| 1404 | \&\s-1TIEHANDLE\s0 by Sven Verdoolaege <\fIskimo@dns.ufsia.ac.be\fR> and Doug MacEachern <\fIdougm@osf.org\fR> |
| 1405 | .PP |
| 1406 | \&\s-1UNTIE\s0 by Nick Ing-Simmons <\fInick@ing\-simmons.net\fR> |
| 1407 | .PP |
| 1408 | \&\s-1SCALAR\s0 by Tassilo von Parseval <\fItassilo.von.parseval@rwth\-aachen.de\fR> |
| 1409 | .PP |
| 1410 | Tying Arrays by Casey West <\fIcasey@geeknest.com\fR> |