| 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 "PERLXSTUT 1" |
| 132 | .TH PERLXSTUT 1 "2006-01-07" "perl v5.8.8" "Perl Programmers Reference Guide" |
| 133 | .SH "NAME" |
| 134 | perlXStut \- Tutorial for writing XSUBs |
| 135 | .SH "DESCRIPTION" |
| 136 | .IX Header "DESCRIPTION" |
| 137 | This tutorial will educate the reader on the steps involved in creating |
| 138 | a Perl extension. The reader is assumed to have access to perlguts, |
| 139 | perlapi and perlxs. |
| 140 | .PP |
| 141 | This tutorial starts with very simple examples and becomes more complex, |
| 142 | with each new example adding new features. Certain concepts may not be |
| 143 | completely explained until later in the tutorial in order to slowly ease |
| 144 | the reader into building extensions. |
| 145 | .PP |
| 146 | This tutorial was written from a Unix point of view. Where I know them |
| 147 | to be otherwise different for other platforms (e.g. Win32), I will list |
| 148 | them. If you find something that was missed, please let me know. |
| 149 | .SH "SPECIAL NOTES" |
| 150 | .IX Header "SPECIAL NOTES" |
| 151 | .Sh "make" |
| 152 | .IX Subsection "make" |
| 153 | This tutorial assumes that the make program that Perl is configured to |
| 154 | use is called \f(CW\*(C`make\*(C'\fR. Instead of running \*(L"make\*(R" in the examples that |
| 155 | follow, you may have to substitute whatever make program Perl has been |
| 156 | configured to use. Running \fBperl \-V:make\fR should tell you what it is. |
| 157 | .Sh "Version caveat" |
| 158 | .IX Subsection "Version caveat" |
| 159 | When writing a Perl extension for general consumption, one should expect that |
| 160 | the extension will be used with versions of Perl different from the |
| 161 | version available on your machine. Since you are reading this document, |
| 162 | the version of Perl on your machine is probably 5.005 or later, but the users |
| 163 | of your extension may have more ancient versions. |
| 164 | .PP |
| 165 | To understand what kinds of incompatibilities one may expect, and in the rare |
| 166 | case that the version of Perl on your machine is older than this document, |
| 167 | see the section on \*(L"Troubleshooting these Examples\*(R" for more information. |
| 168 | .PP |
| 169 | If your extension uses some features of Perl which are not available on older |
| 170 | releases of Perl, your users would appreciate an early meaningful warning. |
| 171 | You would probably put this information into the \fI\s-1README\s0\fR file, but nowadays |
| 172 | installation of extensions may be performed automatically, guided by \fI\s-1CPAN\s0.pm\fR |
| 173 | module or other tools. |
| 174 | .PP |
| 175 | In MakeMaker-based installations, \fIMakefile.PL\fR provides the earliest |
| 176 | opportunity to perform version checks. One can put something like this |
| 177 | in \fIMakefile.PL\fR for this purpose: |
| 178 | .PP |
| 179 | .Vb 7 |
| 180 | \& eval { require 5.007 } |
| 181 | \& or die <<EOD; |
| 182 | \& ############ |
| 183 | \& ### This module uses frobnication framework which is not available before |
| 184 | \& ### version 5.007 of Perl. Upgrade your Perl before installing Kara::Mba. |
| 185 | \& ############ |
| 186 | \& EOD |
| 187 | .Ve |
| 188 | .Sh "Dynamic Loading versus Static Loading" |
| 189 | .IX Subsection "Dynamic Loading versus Static Loading" |
| 190 | It is commonly thought that if a system does not have the capability to |
| 191 | dynamically load a library, you cannot build XSUBs. This is incorrect. |
| 192 | You \fIcan\fR build them, but you must link the XSUBs subroutines with the |
| 193 | rest of Perl, creating a new executable. This situation is similar to |
| 194 | Perl 4. |
| 195 | .PP |
| 196 | This tutorial can still be used on such a system. The \s-1XSUB\s0 build mechanism |
| 197 | will check the system and build a dynamically-loadable library if possible, |
| 198 | or else a static library and then, optionally, a new statically-linked |
| 199 | executable with that static library linked in. |
| 200 | .PP |
| 201 | Should you wish to build a statically-linked executable on a system which |
| 202 | can dynamically load libraries, you may, in all the following examples, |
| 203 | where the command "\f(CW\*(C`make\*(C'\fR\*(L" with no arguments is executed, run the command |
| 204 | \&\*(R"\f(CW\*(C`make perl\*(C'\fR" instead. |
| 205 | .PP |
| 206 | If you have generated such a statically-linked executable by choice, then |
| 207 | instead of saying "\f(CW\*(C`make test\*(C'\fR\*(L", you should say \*(R"\f(CW\*(C`make test_static\*(C'\fR\*(L". |
| 208 | On systems that cannot build dynamically-loadable libraries at all, simply |
| 209 | saying \*(R"\f(CW\*(C`make test\*(C'\fR" is sufficient. |
| 210 | .SH "TUTORIAL" |
| 211 | .IX Header "TUTORIAL" |
| 212 | Now let's go on with the show! |
| 213 | .Sh "\s-1EXAMPLE\s0 1" |
| 214 | .IX Subsection "EXAMPLE 1" |
| 215 | Our first extension will be very simple. When we call the routine in the |
| 216 | extension, it will print out a well-known message and return. |
| 217 | .PP |
| 218 | Run "\f(CW\*(C`h2xs \-A \-n Mytest\*(C'\fR". This creates a directory named Mytest, |
| 219 | possibly under ext/ if that directory exists in the current working |
| 220 | directory. Several files will be created in the Mytest dir, including |
| 221 | \&\s-1MANIFEST\s0, Makefile.PL, Mytest.pm, Mytest.xs, test.pl, and Changes. |
| 222 | .PP |
| 223 | The \s-1MANIFEST\s0 file contains the names of all the files just created in the |
| 224 | Mytest directory. |
| 225 | .PP |
| 226 | The file Makefile.PL should look something like this: |
| 227 | .PP |
| 228 | .Vb 10 |
| 229 | \& use ExtUtils::MakeMaker; |
| 230 | \& # See lib/ExtUtils/MakeMaker.pm for details of how to influence |
| 231 | \& # the contents of the Makefile that is written. |
| 232 | \& WriteMakefile( |
| 233 | \& NAME => 'Mytest', |
| 234 | \& VERSION_FROM => 'Mytest.pm', # finds $VERSION |
| 235 | \& LIBS => [''], # e.g., '-lm' |
| 236 | \& DEFINE => '', # e.g., '-DHAVE_SOMETHING' |
| 237 | \& INC => '', # e.g., '-I/usr/include/other' |
| 238 | \& ); |
| 239 | .Ve |
| 240 | .PP |
| 241 | The file Mytest.pm should start with something like this: |
| 242 | .PP |
| 243 | .Vb 1 |
| 244 | \& package Mytest; |
| 245 | .Ve |
| 246 | .PP |
| 247 | .Vb 2 |
| 248 | \& use strict; |
| 249 | \& use warnings; |
| 250 | .Ve |
| 251 | .PP |
| 252 | .Vb 2 |
| 253 | \& require Exporter; |
| 254 | \& require DynaLoader; |
| 255 | .Ve |
| 256 | .PP |
| 257 | .Vb 5 |
| 258 | \& our @ISA = qw(Exporter DynaLoader); |
| 259 | \& # Items to export into callers namespace by default. Note: do not export |
| 260 | \& # names by default without a very good reason. Use EXPORT_OK instead. |
| 261 | \& # Do not simply export all your public functions/methods/constants. |
| 262 | \& our @EXPORT = qw( |
| 263 | .Ve |
| 264 | .PP |
| 265 | .Vb 2 |
| 266 | \& ); |
| 267 | \& our $VERSION = '0.01'; |
| 268 | .Ve |
| 269 | .PP |
| 270 | .Vb 1 |
| 271 | \& bootstrap Mytest $VERSION; |
| 272 | .Ve |
| 273 | .PP |
| 274 | .Vb 1 |
| 275 | \& # Preloaded methods go here. |
| 276 | .Ve |
| 277 | .PP |
| 278 | .Vb 1 |
| 279 | \& # Autoload methods go after __END__, and are processed by the autosplit program. |
| 280 | .Ve |
| 281 | .PP |
| 282 | .Vb 3 |
| 283 | \& 1; |
| 284 | \& __END__ |
| 285 | \& # Below is the stub of documentation for your module. You better edit it! |
| 286 | .Ve |
| 287 | .PP |
| 288 | The rest of the .pm file contains sample code for providing documentation for |
| 289 | the extension. |
| 290 | .PP |
| 291 | Finally, the Mytest.xs file should look something like this: |
| 292 | .PP |
| 293 | .Vb 3 |
| 294 | \& #include "EXTERN.h" |
| 295 | \& #include "perl.h" |
| 296 | \& #include "XSUB.h" |
| 297 | .Ve |
| 298 | .PP |
| 299 | .Vb 1 |
| 300 | \& MODULE = Mytest PACKAGE = Mytest |
| 301 | .Ve |
| 302 | .PP |
| 303 | Let's edit the .xs file by adding this to the end of the file: |
| 304 | .PP |
| 305 | .Vb 4 |
| 306 | \& void |
| 307 | \& hello() |
| 308 | \& CODE: |
| 309 | \& printf("Hello, world!\en"); |
| 310 | .Ve |
| 311 | .PP |
| 312 | It is okay for the lines starting at the \*(L"\s-1CODE:\s0\*(R" line to not be indented. |
| 313 | However, for readability purposes, it is suggested that you indent \s-1CODE:\s0 |
| 314 | one level and the lines following one more level. |
| 315 | .PP |
| 316 | Now we'll run "\f(CW\*(C`perl Makefile.PL\*(C'\fR". This will create a real Makefile, |
| 317 | which make needs. Its output looks something like: |
| 318 | .PP |
| 319 | .Vb 5 |
| 320 | \& % perl Makefile.PL |
| 321 | \& Checking if your kit is complete... |
| 322 | \& Looks good |
| 323 | \& Writing Makefile for Mytest |
| 324 | \& % |
| 325 | .Ve |
| 326 | .PP |
| 327 | Now, running make will produce output that looks something like this (some |
| 328 | long lines have been shortened for clarity and some extraneous lines have |
| 329 | been deleted): |
| 330 | .PP |
| 331 | .Vb 13 |
| 332 | \& % make |
| 333 | \& umask 0 && cp Mytest.pm ./blib/Mytest.pm |
| 334 | \& perl xsubpp -typemap typemap Mytest.xs >Mytest.tc && mv Mytest.tc Mytest.c |
| 335 | \& Please specify prototyping behavior for Mytest.xs (see perlxs manual) |
| 336 | \& cc -c Mytest.c |
| 337 | \& Running Mkbootstrap for Mytest () |
| 338 | \& chmod 644 Mytest.bs |
| 339 | \& LD_RUN_PATH="" ld -o ./blib/PA-RISC1.1/auto/Mytest/Mytest.sl -b Mytest.o |
| 340 | \& chmod 755 ./blib/PA-RISC1.1/auto/Mytest/Mytest.sl |
| 341 | \& cp Mytest.bs ./blib/PA-RISC1.1/auto/Mytest/Mytest.bs |
| 342 | \& chmod 644 ./blib/PA-RISC1.1/auto/Mytest/Mytest.bs |
| 343 | \& Manifying ./blib/man3/Mytest.3 |
| 344 | \& % |
| 345 | .Ve |
| 346 | .PP |
| 347 | You can safely ignore the line about \*(L"prototyping behavior\*(R" \- it is |
| 348 | explained in the section \*(L"The \s-1PROTOTYPES:\s0 Keyword\*(R" in perlxs. |
| 349 | .PP |
| 350 | If you are on a Win32 system, and the build process fails with linker |
| 351 | errors for functions in the C library, check if your Perl is configured |
| 352 | to use PerlCRT (running \fBperl \-V:libc\fR should show you if this is the |
| 353 | case). If Perl is configured to use PerlCRT, you have to make sure |
| 354 | PerlCRT.lib is copied to the same location that msvcrt.lib lives in, |
| 355 | so that the compiler can find it on its own. msvcrt.lib is usually |
| 356 | found in the Visual C compiler's lib directory (e.g. C:/DevStudio/VC/lib). |
| 357 | .PP |
| 358 | Perl has its own special way of easily writing test scripts, but for this |
| 359 | example only, we'll create our own test script. Create a file called hello |
| 360 | that looks like this: |
| 361 | .PP |
| 362 | .Vb 1 |
| 363 | \& #! /opt/perl5/bin/perl |
| 364 | .Ve |
| 365 | .PP |
| 366 | .Vb 1 |
| 367 | \& use ExtUtils::testlib; |
| 368 | .Ve |
| 369 | .PP |
| 370 | .Vb 1 |
| 371 | \& use Mytest; |
| 372 | .Ve |
| 373 | .PP |
| 374 | .Vb 1 |
| 375 | \& Mytest::hello(); |
| 376 | .Ve |
| 377 | .PP |
| 378 | Now we make the script executable (\f(CW\*(C`chmod +x hello\*(C'\fR), run the script |
| 379 | and we should see the following output: |
| 380 | .PP |
| 381 | .Vb 3 |
| 382 | \& % ./hello |
| 383 | \& Hello, world! |
| 384 | \& % |
| 385 | .Ve |
| 386 | .Sh "\s-1EXAMPLE\s0 2" |
| 387 | .IX Subsection "EXAMPLE 2" |
| 388 | Now let's add to our extension a subroutine that will take a single numeric |
| 389 | argument as input and return 0 if the number is even or 1 if the number |
| 390 | is odd. |
| 391 | .PP |
| 392 | Add the following to the end of Mytest.xs: |
| 393 | .PP |
| 394 | .Vb 7 |
| 395 | \& int |
| 396 | \& is_even(input) |
| 397 | \& int input |
| 398 | \& CODE: |
| 399 | \& RETVAL = (input % 2 == 0); |
| 400 | \& OUTPUT: |
| 401 | \& RETVAL |
| 402 | .Ve |
| 403 | .PP |
| 404 | There does not need to be whitespace at the start of the "\f(CW\*(C`int input\*(C'\fR\*(L" |
| 405 | line, but it is useful for improving readability. Placing a semi-colon at |
| 406 | the end of that line is also optional. Any amount and kind of whitespace |
| 407 | may be placed between the \*(R"\f(CW\*(C`int\*(C'\fR\*(L" and \*(R"\f(CW\*(C`input\*(C'\fR". |
| 408 | .PP |
| 409 | Now re-run make to rebuild our new shared library. |
| 410 | .PP |
| 411 | Now perform the same steps as before, generating a Makefile from the |
| 412 | Makefile.PL file, and running make. |
| 413 | .PP |
| 414 | In order to test that our extension works, we now need to look at the |
| 415 | file test.pl. This file is set up to imitate the same kind of testing |
| 416 | structure that Perl itself has. Within the test script, you perform a |
| 417 | number of tests to confirm the behavior of the extension, printing \*(L"ok\*(R" |
| 418 | when the test is correct, \*(L"not ok\*(R" when it is not. Change the print |
| 419 | statement in the \s-1BEGIN\s0 block to print \*(L"1..4\*(R", and add the following code |
| 420 | to the end of the file: |
| 421 | .PP |
| 422 | .Vb 3 |
| 423 | \& print &Mytest::is_even(0) == 1 ? "ok 2" : "not ok 2", "\en"; |
| 424 | \& print &Mytest::is_even(1) == 0 ? "ok 3" : "not ok 3", "\en"; |
| 425 | \& print &Mytest::is_even(2) == 1 ? "ok 4" : "not ok 4", "\en"; |
| 426 | .Ve |
| 427 | .PP |
| 428 | We will be calling the test script through the command "\f(CW\*(C`make test\*(C'\fR". You |
| 429 | should see output that looks something like this: |
| 430 | .PP |
| 431 | .Vb 8 |
| 432 | \& % make test |
| 433 | \& PERL_DL_NONLAZY=1 /opt/perl5.004/bin/perl (lots of -I arguments) test.pl |
| 434 | \& 1..4 |
| 435 | \& ok 1 |
| 436 | \& ok 2 |
| 437 | \& ok 3 |
| 438 | \& ok 4 |
| 439 | \& % |
| 440 | .Ve |
| 441 | .Sh "What has gone on?" |
| 442 | .IX Subsection "What has gone on?" |
| 443 | The program h2xs is the starting point for creating extensions. In later |
| 444 | examples we'll see how we can use h2xs to read header files and generate |
| 445 | templates to connect to C routines. |
| 446 | .PP |
| 447 | h2xs creates a number of files in the extension directory. The file |
| 448 | Makefile.PL is a perl script which will generate a true Makefile to build |
| 449 | the extension. We'll take a closer look at it later. |
| 450 | .PP |
| 451 | The .pm and .xs files contain the meat of the extension. The .xs file holds |
| 452 | the C routines that make up the extension. The .pm file contains routines |
| 453 | that tell Perl how to load your extension. |
| 454 | .PP |
| 455 | Generating the Makefile and running \f(CW\*(C`make\*(C'\fR created a directory called blib |
| 456 | (which stands for \*(L"build library\*(R") in the current working directory. This |
| 457 | directory will contain the shared library that we will build. Once we have |
| 458 | tested it, we can install it into its final location. |
| 459 | .PP |
| 460 | Invoking the test script via "\f(CW\*(C`make test\*(C'\fR" did something very important. |
| 461 | It invoked perl with all those \f(CW\*(C`\-I\*(C'\fR arguments so that it could find the |
| 462 | various files that are part of the extension. It is \fIvery\fR important that |
| 463 | while you are still testing extensions that you use "\f(CW\*(C`make test\*(C'\fR\*(L". If you |
| 464 | try to run the test script all by itself, you will get a fatal error. |
| 465 | Another reason it is important to use \*(R"\f(CW\*(C`make test\*(C'\fR\*(L" to run your test |
| 466 | script is that if you are testing an upgrade to an already-existing version, |
| 467 | using \*(R"\f(CW\*(C`make test\*(C'\fR" insures that you will test your new extension, not the |
| 468 | already-existing version. |
| 469 | .PP |
| 470 | When Perl sees a \f(CW\*(C`use extension;\*(C'\fR, it searches for a file with the same name |
| 471 | as the \f(CW\*(C`use\*(C'\fR'd extension that has a .pm suffix. If that file cannot be found, |
| 472 | Perl dies with a fatal error. The default search path is contained in the |
| 473 | \&\f(CW@INC\fR array. |
| 474 | .PP |
| 475 | In our case, Mytest.pm tells perl that it will need the Exporter and Dynamic |
| 476 | Loader extensions. It then sets the \f(CW@ISA\fR and \f(CW@EXPORT\fR arrays and the |
| 477 | \&\f(CW$VERSION\fR scalar; finally it tells perl to bootstrap the module. Perl |
| 478 | will call its dynamic loader routine (if there is one) and load the shared |
| 479 | library. |
| 480 | .PP |
| 481 | The two arrays \f(CW@ISA\fR and \f(CW@EXPORT\fR are very important. The \f(CW@ISA\fR |
| 482 | array contains a list of other packages in which to search for methods (or |
| 483 | subroutines) that do not exist in the current package. This is usually |
| 484 | only important for object-oriented extensions (which we will talk about |
| 485 | much later), and so usually doesn't need to be modified. |
| 486 | .PP |
| 487 | The \f(CW@EXPORT\fR array tells Perl which of the extension's variables and |
| 488 | subroutines should be placed into the calling package's namespace. Because |
| 489 | you don't know if the user has already used your variable and subroutine |
| 490 | names, it's vitally important to carefully select what to export. Do \fInot\fR |
| 491 | export method or variable names \fIby default\fR without a good reason. |
| 492 | .PP |
| 493 | As a general rule, if the module is trying to be object-oriented then don't |
| 494 | export anything. If it's just a collection of functions and variables, then |
| 495 | you can export them via another array, called \f(CW@EXPORT_OK\fR. This array |
| 496 | does not automatically place its subroutine and variable names into the |
| 497 | namespace unless the user specifically requests that this be done. |
| 498 | .PP |
| 499 | See perlmod for more information. |
| 500 | .PP |
| 501 | The \f(CW$VERSION\fR variable is used to ensure that the .pm file and the shared |
| 502 | library are \*(L"in sync\*(R" with each other. Any time you make changes to |
| 503 | the .pm or .xs files, you should increment the value of this variable. |
| 504 | .Sh "Writing good test scripts" |
| 505 | .IX Subsection "Writing good test scripts" |
| 506 | The importance of writing good test scripts cannot be overemphasized. You |
| 507 | should closely follow the \*(L"ok/not ok\*(R" style that Perl itself uses, so that |
| 508 | it is very easy and unambiguous to determine the outcome of each test case. |
| 509 | When you find and fix a bug, make sure you add a test case for it. |
| 510 | .PP |
| 511 | By running "\f(CW\*(C`make test\*(C'\fR\*(L", you ensure that your test.pl script runs and uses |
| 512 | the correct version of your extension. If you have many test cases, you |
| 513 | might want to copy Perl's test style. Create a directory named \*(R"t\*(L" in the |
| 514 | extension's directory and append the suffix \*(R".t\*(L" to the names of your test |
| 515 | files. When you run \*(R"\f(CW\*(C`make test\*(C'\fR", all of these test files will be executed. |
| 516 | .Sh "\s-1EXAMPLE\s0 3" |
| 517 | .IX Subsection "EXAMPLE 3" |
| 518 | Our third extension will take one argument as its input, round off that |
| 519 | value, and set the \fIargument\fR to the rounded value. |
| 520 | .PP |
| 521 | Add the following to the end of Mytest.xs: |
| 522 | .PP |
| 523 | .Vb 13 |
| 524 | \& void |
| 525 | \& round(arg) |
| 526 | \& double arg |
| 527 | \& CODE: |
| 528 | \& if (arg > 0.0) { |
| 529 | \& arg = floor(arg + 0.5); |
| 530 | \& } else if (arg < 0.0) { |
| 531 | \& arg = ceil(arg - 0.5); |
| 532 | \& } else { |
| 533 | \& arg = 0.0; |
| 534 | \& } |
| 535 | \& OUTPUT: |
| 536 | \& arg |
| 537 | .Ve |
| 538 | .PP |
| 539 | Edit the Makefile.PL file so that the corresponding line looks like this: |
| 540 | .PP |
| 541 | .Vb 1 |
| 542 | \& 'LIBS' => ['-lm'], # e.g., '-lm' |
| 543 | .Ve |
| 544 | .PP |
| 545 | Generate the Makefile and run make. Change the \s-1BEGIN\s0 block to print |
| 546 | \&\*(L"1..9\*(R" and add the following to test.pl: |
| 547 | .PP |
| 548 | .Vb 5 |
| 549 | \& $i = -1.5; &Mytest::round($i); print $i == -2.0 ? "ok 5" : "not ok 5", "\en"; |
| 550 | \& $i = -1.1; &Mytest::round($i); print $i == -1.0 ? "ok 6" : "not ok 6", "\en"; |
| 551 | \& $i = 0.0; &Mytest::round($i); print $i == 0.0 ? "ok 7" : "not ok 7", "\en"; |
| 552 | \& $i = 0.5; &Mytest::round($i); print $i == 1.0 ? "ok 8" : "not ok 8", "\en"; |
| 553 | \& $i = 1.2; &Mytest::round($i); print $i == 1.0 ? "ok 9" : "not ok 9", "\en"; |
| 554 | .Ve |
| 555 | .PP |
| 556 | Running "\f(CW\*(C`make test\*(C'\fR" should now print out that all nine tests are okay. |
| 557 | .PP |
| 558 | Notice that in these new test cases, the argument passed to round was a |
| 559 | scalar variable. You might be wondering if you can round a constant or |
| 560 | literal. To see what happens, temporarily add the following line to test.pl: |
| 561 | .PP |
| 562 | .Vb 1 |
| 563 | \& &Mytest::round(3); |
| 564 | .Ve |
| 565 | .PP |
| 566 | Run "\f(CW\*(C`make test\*(C'\fR" and notice that Perl dies with a fatal error. Perl won't |
| 567 | let you change the value of constants! |
| 568 | .Sh "What's new here?" |
| 569 | .IX Subsection "What's new here?" |
| 570 | .IP "\(bu" 4 |
| 571 | We've made some changes to Makefile.PL. In this case, we've specified an |
| 572 | extra library to be linked into the extension's shared library, the math |
| 573 | library libm in this case. We'll talk later about how to write XSUBs that |
| 574 | can call every routine in a library. |
| 575 | .IP "\(bu" 4 |
| 576 | The value of the function is not being passed back as the function's return |
| 577 | value, but by changing the value of the variable that was passed into the |
| 578 | function. You might have guessed that when you saw that the return value |
| 579 | of round is of type \*(L"void\*(R". |
| 580 | .Sh "Input and Output Parameters" |
| 581 | .IX Subsection "Input and Output Parameters" |
| 582 | You specify the parameters that will be passed into the \s-1XSUB\s0 on the line(s) |
| 583 | after you declare the function's return value and name. Each input parameter |
| 584 | line starts with optional whitespace, and may have an optional terminating |
| 585 | semicolon. |
| 586 | .PP |
| 587 | The list of output parameters occurs at the very end of the function, just |
| 588 | before after the \s-1OUTPUT:\s0 directive. The use of \s-1RETVAL\s0 tells Perl that you |
| 589 | wish to send this value back as the return value of the \s-1XSUB\s0 function. In |
| 590 | Example 3, we wanted the \*(L"return value\*(R" placed in the original variable |
| 591 | which we passed in, so we listed it (and not \s-1RETVAL\s0) in the \s-1OUTPUT:\s0 section. |
| 592 | .Sh "The \s-1XSUBPP\s0 Program" |
| 593 | .IX Subsection "The XSUBPP Program" |
| 594 | The \fBxsubpp\fR program takes the \s-1XS\s0 code in the .xs file and translates it into |
| 595 | C code, placing it in a file whose suffix is .c. The C code created makes |
| 596 | heavy use of the C functions within Perl. |
| 597 | .Sh "The \s-1TYPEMAP\s0 file" |
| 598 | .IX Subsection "The TYPEMAP file" |
| 599 | The \fBxsubpp\fR program uses rules to convert from Perl's data types (scalar, |
| 600 | array, etc.) to C's data types (int, char, etc.). These rules are stored |
| 601 | in the typemap file ($PERLLIB/ExtUtils/typemap). This file is split into |
| 602 | three parts. |
| 603 | .PP |
| 604 | The first section maps various C data types to a name, which corresponds |
| 605 | somewhat with the various Perl types. The second section contains C code |
| 606 | which \fBxsubpp\fR uses to handle input parameters. The third section contains |
| 607 | C code which \fBxsubpp\fR uses to handle output parameters. |
| 608 | .PP |
| 609 | Let's take a look at a portion of the .c file created for our extension. |
| 610 | The file name is Mytest.c: |
| 611 | .PP |
| 612 | .Vb 18 |
| 613 | \& XS(XS_Mytest_round) |
| 614 | \& { |
| 615 | \& dXSARGS; |
| 616 | \& if (items != 1) |
| 617 | \& croak("Usage: Mytest::round(arg)"); |
| 618 | \& { |
| 619 | \& double arg = (double)SvNV(ST(0)); /* XXXXX */ |
| 620 | \& if (arg > 0.0) { |
| 621 | \& arg = floor(arg + 0.5); |
| 622 | \& } else if (arg < 0.0) { |
| 623 | \& arg = ceil(arg - 0.5); |
| 624 | \& } else { |
| 625 | \& arg = 0.0; |
| 626 | \& } |
| 627 | \& sv_setnv(ST(0), (double)arg); /* XXXXX */ |
| 628 | \& } |
| 629 | \& XSRETURN(1); |
| 630 | \& } |
| 631 | .Ve |
| 632 | .PP |
| 633 | Notice the two lines commented with \*(L"\s-1XXXXX\s0\*(R". If you check the first section |
| 634 | of the typemap file, you'll see that doubles are of type T_DOUBLE. In the |
| 635 | \&\s-1INPUT\s0 section, an argument that is T_DOUBLE is assigned to the variable |
| 636 | arg by calling the routine SvNV on something, then casting it to double, |
| 637 | then assigned to the variable arg. Similarly, in the \s-1OUTPUT\s0 section, |
| 638 | once arg has its final value, it is passed to the sv_setnv function to |
| 639 | be passed back to the calling subroutine. These two functions are explained |
| 640 | in perlguts; we'll talk more later about what that \*(L"\s-1\fIST\s0\fR\|(0)\*(R" means in the |
| 641 | section on the argument stack. |
| 642 | .Sh "Warning about Output Arguments" |
| 643 | .IX Subsection "Warning about Output Arguments" |
| 644 | In general, it's not a good idea to write extensions that modify their input |
| 645 | parameters, as in Example 3. Instead, you should probably return multiple |
| 646 | values in an array and let the caller handle them (we'll do this in a later |
| 647 | example). However, in order to better accommodate calling pre-existing C |
| 648 | routines, which often do modify their input parameters, this behavior is |
| 649 | tolerated. |
| 650 | .Sh "\s-1EXAMPLE\s0 4" |
| 651 | .IX Subsection "EXAMPLE 4" |
| 652 | In this example, we'll now begin to write XSUBs that will interact with |
| 653 | pre-defined C libraries. To begin with, we will build a small library of |
| 654 | our own, then let h2xs write our .pm and .xs files for us. |
| 655 | .PP |
| 656 | Create a new directory called Mytest2 at the same level as the directory |
| 657 | Mytest. In the Mytest2 directory, create another directory called mylib, |
| 658 | and cd into that directory. |
| 659 | .PP |
| 660 | Here we'll create some files that will generate a test library. These will |
| 661 | include a C source file and a header file. We'll also create a Makefile.PL |
| 662 | in this directory. Then we'll make sure that running make at the Mytest2 |
| 663 | level will automatically run this Makefile.PL file and the resulting Makefile. |
| 664 | .PP |
| 665 | In the mylib directory, create a file mylib.h that looks like this: |
| 666 | .PP |
| 667 | .Vb 1 |
| 668 | \& #define TESTVAL 4 |
| 669 | .Ve |
| 670 | .PP |
| 671 | .Vb 1 |
| 672 | \& extern double foo(int, long, const char*); |
| 673 | .Ve |
| 674 | .PP |
| 675 | Also create a file mylib.c that looks like this: |
| 676 | .PP |
| 677 | .Vb 2 |
| 678 | \& #include <stdlib.h> |
| 679 | \& #include "./mylib.h" |
| 680 | .Ve |
| 681 | .PP |
| 682 | .Vb 5 |
| 683 | \& double |
| 684 | \& foo(int a, long b, const char *c) |
| 685 | \& { |
| 686 | \& return (a + b + atof(c) + TESTVAL); |
| 687 | \& } |
| 688 | .Ve |
| 689 | .PP |
| 690 | And finally create a file Makefile.PL that looks like this: |
| 691 | .PP |
| 692 | .Vb 7 |
| 693 | \& use ExtUtils::MakeMaker; |
| 694 | \& $Verbose = 1; |
| 695 | \& WriteMakefile( |
| 696 | \& NAME => 'Mytest2::mylib', |
| 697 | \& SKIP => [qw(all static static_lib dynamic dynamic_lib)], |
| 698 | \& clean => {'FILES' => 'libmylib$(LIB_EXT)'}, |
| 699 | \& ); |
| 700 | .Ve |
| 701 | .PP |
| 702 | .Vb 3 |
| 703 | \& sub MY::top_targets { |
| 704 | \& ' |
| 705 | \& all :: static |
| 706 | .Ve |
| 707 | .PP |
| 708 | .Vb 1 |
| 709 | \& pure_all :: static |
| 710 | .Ve |
| 711 | .PP |
| 712 | .Vb 1 |
| 713 | \& static :: libmylib$(LIB_EXT) |
| 714 | .Ve |
| 715 | .PP |
| 716 | .Vb 3 |
| 717 | \& libmylib$(LIB_EXT): $(O_FILES) |
| 718 | \& $(AR) cr libmylib$(LIB_EXT) $(O_FILES) |
| 719 | \& $(RANLIB) libmylib$(LIB_EXT) |
| 720 | .Ve |
| 721 | .PP |
| 722 | .Vb 2 |
| 723 | \& '; |
| 724 | \& } |
| 725 | .Ve |
| 726 | .PP |
| 727 | Make sure you use a tab and not spaces on the lines beginning with \*(L"$(\s-1AR\s0)\*(R" |
| 728 | and \*(L"$(\s-1RANLIB\s0)\*(R". Make will not function properly if you use spaces. |
| 729 | It has also been reported that the \*(L"cr\*(R" argument to $(\s-1AR\s0) is unnecessary |
| 730 | on Win32 systems. |
| 731 | .PP |
| 732 | We will now create the main top-level Mytest2 files. Change to the directory |
| 733 | above Mytest2 and run the following command: |
| 734 | .PP |
| 735 | .Vb 1 |
| 736 | \& % h2xs -O -n Mytest2 ./Mytest2/mylib/mylib.h |
| 737 | .Ve |
| 738 | .PP |
| 739 | This will print out a warning about overwriting Mytest2, but that's okay. |
| 740 | Our files are stored in Mytest2/mylib, and will be untouched. |
| 741 | .PP |
| 742 | The normal Makefile.PL that h2xs generates doesn't know about the mylib |
| 743 | directory. We need to tell it that there is a subdirectory and that we |
| 744 | will be generating a library in it. Let's add the argument \s-1MYEXTLIB\s0 to |
| 745 | the WriteMakefile call so that it looks like this: |
| 746 | .PP |
| 747 | .Vb 8 |
| 748 | \& WriteMakefile( |
| 749 | \& 'NAME' => 'Mytest2', |
| 750 | \& 'VERSION_FROM' => 'Mytest2.pm', # finds $VERSION |
| 751 | \& 'LIBS' => [''], # e.g., '-lm' |
| 752 | \& 'DEFINE' => '', # e.g., '-DHAVE_SOMETHING' |
| 753 | \& 'INC' => '', # e.g., '-I/usr/include/other' |
| 754 | \& 'MYEXTLIB' => 'mylib/libmylib$(LIB_EXT)', |
| 755 | \& ); |
| 756 | .Ve |
| 757 | .PP |
| 758 | and then at the end add a subroutine (which will override the pre-existing |
| 759 | subroutine). Remember to use a tab character to indent the line beginning |
| 760 | with \*(L"cd\*(R"! |
| 761 | .PP |
| 762 | .Vb 6 |
| 763 | \& sub MY::postamble { |
| 764 | \& ' |
| 765 | \& $(MYEXTLIB): mylib/Makefile |
| 766 | \& cd mylib && $(MAKE) $(PASSTHRU) |
| 767 | \& '; |
| 768 | \& } |
| 769 | .Ve |
| 770 | .PP |
| 771 | Let's also fix the \s-1MANIFEST\s0 file so that it accurately reflects the contents |
| 772 | of our extension. The single line that says \*(L"mylib\*(R" should be replaced by |
| 773 | the following three lines: |
| 774 | .PP |
| 775 | .Vb 3 |
| 776 | \& mylib/Makefile.PL |
| 777 | \& mylib/mylib.c |
| 778 | \& mylib/mylib.h |
| 779 | .Ve |
| 780 | .PP |
| 781 | To keep our namespace nice and unpolluted, edit the .pm file and change |
| 782 | the variable \f(CW@EXPORT\fR to \f(CW@EXPORT_OK\fR. Finally, in the |
| 783 | \&.xs file, edit the #include line to read: |
| 784 | .PP |
| 785 | .Vb 1 |
| 786 | \& #include "mylib/mylib.h" |
| 787 | .Ve |
| 788 | .PP |
| 789 | And also add the following function definition to the end of the .xs file: |
| 790 | .PP |
| 791 | .Vb 7 |
| 792 | \& double |
| 793 | \& foo(a,b,c) |
| 794 | \& int a |
| 795 | \& long b |
| 796 | \& const char * c |
| 797 | \& OUTPUT: |
| 798 | \& RETVAL |
| 799 | .Ve |
| 800 | .PP |
| 801 | Now we also need to create a typemap file because the default Perl doesn't |
| 802 | currently support the const char * type. Create a file called typemap in |
| 803 | the Mytest2 directory and place the following in it: |
| 804 | .PP |
| 805 | .Vb 1 |
| 806 | \& const char * T_PV |
| 807 | .Ve |
| 808 | .PP |
| 809 | Now run perl on the top-level Makefile.PL. Notice that it also created a |
| 810 | Makefile in the mylib directory. Run make and watch that it does cd into |
| 811 | the mylib directory and run make in there as well. |
| 812 | .PP |
| 813 | Now edit the test.pl script and change the \s-1BEGIN\s0 block to print \*(L"1..4\*(R", |
| 814 | and add the following lines to the end of the script: |
| 815 | .PP |
| 816 | .Vb 3 |
| 817 | \& print &Mytest2::foo(1, 2, "Hello, world!") == 7 ? "ok 2\en" : "not ok 2\en"; |
| 818 | \& print &Mytest2::foo(1, 2, "0.0") == 7 ? "ok 3\en" : "not ok 3\en"; |
| 819 | \& print abs(&Mytest2::foo(0, 0, "-3.4") - 0.6) <= 0.01 ? "ok 4\en" : "not ok 4\en"; |
| 820 | .Ve |
| 821 | .PP |
| 822 | (When dealing with floating-point comparisons, it is best to not check for |
| 823 | equality, but rather that the difference between the expected and actual |
| 824 | result is below a certain amount (called epsilon) which is 0.01 in this case) |
| 825 | .PP |
| 826 | Run "\f(CW\*(C`make test\*(C'\fR" and all should be well. |
| 827 | .Sh "What has happened here?" |
| 828 | .IX Subsection "What has happened here?" |
| 829 | Unlike previous examples, we've now run h2xs on a real include file. This |
| 830 | has caused some extra goodies to appear in both the .pm and .xs files. |
| 831 | .IP "\(bu" 4 |
| 832 | In the .xs file, there's now a #include directive with the absolute path to |
| 833 | the mylib.h header file. We changed this to a relative path so that we |
| 834 | could move the extension directory if we wanted to. |
| 835 | .IP "\(bu" 4 |
| 836 | There's now some new C code that's been added to the .xs file. The purpose |
| 837 | of the \f(CW\*(C`constant\*(C'\fR routine is to make the values that are #define'd in the |
| 838 | header file accessible by the Perl script (by calling either \f(CW\*(C`TESTVAL\*(C'\fR or |
| 839 | \&\f(CW&Mytest2::TESTVAL\fR). There's also some \s-1XS\s0 code to allow calls to the |
| 840 | \&\f(CW\*(C`constant\*(C'\fR routine. |
| 841 | .IP "\(bu" 4 |
| 842 | The .pm file originally exported the name \f(CW\*(C`TESTVAL\*(C'\fR in the \f(CW@EXPORT\fR array. |
| 843 | This could lead to name clashes. A good rule of thumb is that if the #define |
| 844 | is only going to be used by the C routines themselves, and not by the user, |
| 845 | they should be removed from the \f(CW@EXPORT\fR array. Alternately, if you don't |
| 846 | mind using the \*(L"fully qualified name\*(R" of a variable, you could move most |
| 847 | or all of the items from the \f(CW@EXPORT\fR array into the \f(CW@EXPORT_OK\fR array. |
| 848 | .IP "\(bu" 4 |
| 849 | If our include file had contained #include directives, these would not have |
| 850 | been processed by h2xs. There is no good solution to this right now. |
| 851 | .IP "\(bu" 4 |
| 852 | We've also told Perl about the library that we built in the mylib |
| 853 | subdirectory. That required only the addition of the \f(CW\*(C`MYEXTLIB\*(C'\fR variable |
| 854 | to the WriteMakefile call and the replacement of the postamble subroutine |
| 855 | to cd into the subdirectory and run make. The Makefile.PL for the |
| 856 | library is a bit more complicated, but not excessively so. Again we |
| 857 | replaced the postamble subroutine to insert our own code. This code |
| 858 | simply specified that the library to be created here was a static archive |
| 859 | library (as opposed to a dynamically loadable library) and provided the |
| 860 | commands to build it. |
| 861 | .Sh "Anatomy of .xs file" |
| 862 | .IX Subsection "Anatomy of .xs file" |
| 863 | The .xs file of \*(L"\s-1EXAMPLE\s0 4\*(R" contained some new elements. To understand |
| 864 | the meaning of these elements, pay attention to the line which reads |
| 865 | .PP |
| 866 | .Vb 1 |
| 867 | \& MODULE = Mytest2 PACKAGE = Mytest2 |
| 868 | .Ve |
| 869 | .PP |
| 870 | Anything before this line is plain C code which describes which headers |
| 871 | to include, and defines some convenience functions. No translations are |
| 872 | performed on this part, apart from having embedded \s-1POD\s0 documentation |
| 873 | skipped over (see perlpod) it goes into the generated output C file as is. |
| 874 | .PP |
| 875 | Anything after this line is the description of \s-1XSUB\s0 functions. |
| 876 | These descriptions are translated by \fBxsubpp\fR into C code which |
| 877 | implements these functions using Perl calling conventions, and which |
| 878 | makes these functions visible from Perl interpreter. |
| 879 | .PP |
| 880 | Pay a special attention to the function \f(CW\*(C`constant\*(C'\fR. This name appears |
| 881 | twice in the generated .xs file: once in the first part, as a static C |
| 882 | function, then another time in the second part, when an \s-1XSUB\s0 interface to |
| 883 | this static C function is defined. |
| 884 | .PP |
| 885 | This is quite typical for .xs files: usually the .xs file provides |
| 886 | an interface to an existing C function. Then this C function is defined |
| 887 | somewhere (either in an external library, or in the first part of .xs file), |
| 888 | and a Perl interface to this function (i.e. \*(L"Perl glue\*(R") is described in the |
| 889 | second part of .xs file. The situation in \*(L"\s-1EXAMPLE\s0 1\*(R", \*(L"\s-1EXAMPLE\s0 2\*(R", |
| 890 | and \*(L"\s-1EXAMPLE\s0 3\*(R", when all the work is done inside the \*(L"Perl glue\*(R", is |
| 891 | somewhat of an exception rather than the rule. |
| 892 | .Sh "Getting the fat out of XSUBs" |
| 893 | .IX Subsection "Getting the fat out of XSUBs" |
| 894 | In \*(L"\s-1EXAMPLE\s0 4\*(R" the second part of .xs file contained the following |
| 895 | description of an \s-1XSUB:\s0 |
| 896 | .PP |
| 897 | .Vb 7 |
| 898 | \& double |
| 899 | \& foo(a,b,c) |
| 900 | \& int a |
| 901 | \& long b |
| 902 | \& const char * c |
| 903 | \& OUTPUT: |
| 904 | \& RETVAL |
| 905 | .Ve |
| 906 | .PP |
| 907 | Note that in contrast with \*(L"\s-1EXAMPLE\s0 1\*(R", \*(L"\s-1EXAMPLE\s0 2\*(R" and \*(L"\s-1EXAMPLE\s0 3\*(R", |
| 908 | this description does not contain the actual \fIcode\fR for what is done |
| 909 | is done during a call to Perl function \fIfoo()\fR. To understand what is going |
| 910 | on here, one can add a \s-1CODE\s0 section to this \s-1XSUB:\s0 |
| 911 | .PP |
| 912 | .Vb 9 |
| 913 | \& double |
| 914 | \& foo(a,b,c) |
| 915 | \& int a |
| 916 | \& long b |
| 917 | \& const char * c |
| 918 | \& CODE: |
| 919 | \& RETVAL = foo(a,b,c); |
| 920 | \& OUTPUT: |
| 921 | \& RETVAL |
| 922 | .Ve |
| 923 | .PP |
| 924 | However, these two XSUBs provide almost identical generated C code: \fBxsubpp\fR |
| 925 | compiler is smart enough to figure out the \f(CW\*(C`CODE:\*(C'\fR section from the first |
| 926 | two lines of the description of \s-1XSUB\s0. What about \f(CW\*(C`OUTPUT:\*(C'\fR section? In |
| 927 | fact, that is absolutely the same! The \f(CW\*(C`OUTPUT:\*(C'\fR section can be removed |
| 928 | as well, \fIas far as \f(CI\*(C`CODE:\*(C'\fI section or \f(CI\*(C`PPCODE:\*(C'\fI section\fR is not |
| 929 | specified: \fBxsubpp\fR can see that it needs to generate a function call |
| 930 | section, and will autogenerate the \s-1OUTPUT\s0 section too. Thus one can |
| 931 | shortcut the \s-1XSUB\s0 to become: |
| 932 | .PP |
| 933 | .Vb 5 |
| 934 | \& double |
| 935 | \& foo(a,b,c) |
| 936 | \& int a |
| 937 | \& long b |
| 938 | \& const char * c |
| 939 | .Ve |
| 940 | .PP |
| 941 | Can we do the same with an \s-1XSUB\s0 |
| 942 | .PP |
| 943 | .Vb 7 |
| 944 | \& int |
| 945 | \& is_even(input) |
| 946 | \& int input |
| 947 | \& CODE: |
| 948 | \& RETVAL = (input % 2 == 0); |
| 949 | \& OUTPUT: |
| 950 | \& RETVAL |
| 951 | .Ve |
| 952 | .PP |
| 953 | of \*(L"\s-1EXAMPLE\s0 2\*(R"? To do this, one needs to define a C function \f(CW\*(C`int |
| 954 | is_even(int input)\*(C'\fR. As we saw in \*(L"Anatomy of .xs file\*(R", a proper place |
| 955 | for this definition is in the first part of .xs file. In fact a C function |
| 956 | .PP |
| 957 | .Vb 5 |
| 958 | \& int |
| 959 | \& is_even(int arg) |
| 960 | \& { |
| 961 | \& return (arg % 2 == 0); |
| 962 | \& } |
| 963 | .Ve |
| 964 | .PP |
| 965 | is probably overkill for this. Something as simple as a \f(CW\*(C`#define\*(C'\fR will |
| 966 | do too: |
| 967 | .PP |
| 968 | .Vb 1 |
| 969 | \& #define is_even(arg) ((arg) % 2 == 0) |
| 970 | .Ve |
| 971 | .PP |
| 972 | After having this in the first part of .xs file, the \*(L"Perl glue\*(R" part becomes |
| 973 | as simple as |
| 974 | .PP |
| 975 | .Vb 3 |
| 976 | \& int |
| 977 | \& is_even(input) |
| 978 | \& int input |
| 979 | .Ve |
| 980 | .PP |
| 981 | This technique of separation of the glue part from the workhorse part has |
| 982 | obvious tradeoffs: if you want to change a Perl interface, you need to |
| 983 | change two places in your code. However, it removes a lot of clutter, |
| 984 | and makes the workhorse part independent from idiosyncrasies of Perl calling |
| 985 | convention. (In fact, there is nothing Perl-specific in the above description, |
| 986 | a different version of \fBxsubpp\fR might have translated this to \s-1TCL\s0 glue or |
| 987 | Python glue as well.) |
| 988 | .Sh "More about \s-1XSUB\s0 arguments" |
| 989 | .IX Subsection "More about XSUB arguments" |
| 990 | With the completion of Example 4, we now have an easy way to simulate some |
| 991 | real-life libraries whose interfaces may not be the cleanest in the world. |
| 992 | We shall now continue with a discussion of the arguments passed to the |
| 993 | \&\fBxsubpp\fR compiler. |
| 994 | .PP |
| 995 | When you specify arguments to routines in the .xs file, you are really |
| 996 | passing three pieces of information for each argument listed. The first |
| 997 | piece is the order of that argument relative to the others (first, second, |
| 998 | etc). The second is the type of argument, and consists of the type |
| 999 | declaration of the argument (e.g., int, char*, etc). The third piece is |
| 1000 | the calling convention for the argument in the call to the library function. |
| 1001 | .PP |
| 1002 | While Perl passes arguments to functions by reference, |
| 1003 | C passes arguments by value; to implement a C function which modifies data |
| 1004 | of one of the \*(L"arguments\*(R", the actual argument of this C function would be |
| 1005 | a pointer to the data. Thus two C functions with declarations |
| 1006 | .PP |
| 1007 | .Vb 2 |
| 1008 | \& int string_length(char *s); |
| 1009 | \& int upper_case_char(char *cp); |
| 1010 | .Ve |
| 1011 | .PP |
| 1012 | may have completely different semantics: the first one may inspect an array |
| 1013 | of chars pointed by s, and the second one may immediately dereference \f(CW\*(C`cp\*(C'\fR |
| 1014 | and manipulate \f(CW*cp\fR only (using the return value as, say, a success |
| 1015 | indicator). From Perl one would use these functions in |
| 1016 | a completely different manner. |
| 1017 | .PP |
| 1018 | One conveys this info to \fBxsubpp\fR by replacing \f(CW\*(C`*\*(C'\fR before the |
| 1019 | argument by \f(CW\*(C`&\*(C'\fR. \f(CW\*(C`&\*(C'\fR means that the argument should be passed to a library |
| 1020 | function by its address. The above two function may be XSUB-ified as |
| 1021 | .PP |
| 1022 | .Vb 3 |
| 1023 | \& int |
| 1024 | \& string_length(s) |
| 1025 | \& char * s |
| 1026 | .Ve |
| 1027 | .PP |
| 1028 | .Vb 3 |
| 1029 | \& int |
| 1030 | \& upper_case_char(cp) |
| 1031 | \& char &cp |
| 1032 | .Ve |
| 1033 | .PP |
| 1034 | For example, consider: |
| 1035 | .PP |
| 1036 | .Vb 4 |
| 1037 | \& int |
| 1038 | \& foo(a,b) |
| 1039 | \& char &a |
| 1040 | \& char * b |
| 1041 | .Ve |
| 1042 | .PP |
| 1043 | The first Perl argument to this function would be treated as a char and assigned |
| 1044 | to the variable a, and its address would be passed into the function foo. |
| 1045 | The second Perl argument would be treated as a string pointer and assigned to the |
| 1046 | variable b. The \fIvalue\fR of b would be passed into the function foo. The |
| 1047 | actual call to the function foo that \fBxsubpp\fR generates would look like this: |
| 1048 | .PP |
| 1049 | .Vb 1 |
| 1050 | \& foo(&a, b); |
| 1051 | .Ve |
| 1052 | .PP |
| 1053 | \&\fBxsubpp\fR will parse the following function argument lists identically: |
| 1054 | .PP |
| 1055 | .Vb 3 |
| 1056 | \& char &a |
| 1057 | \& char&a |
| 1058 | \& char & a |
| 1059 | .Ve |
| 1060 | .PP |
| 1061 | However, to help ease understanding, it is suggested that you place a \*(L"&\*(R" |
| 1062 | next to the variable name and away from the variable type), and place a |
| 1063 | \&\*(L"*\*(R" near the variable type, but away from the variable name (as in the |
| 1064 | call to foo above). By doing so, it is easy to understand exactly what |
| 1065 | will be passed to the C function \*(-- it will be whatever is in the \*(L"last |
| 1066 | column\*(R". |
| 1067 | .PP |
| 1068 | You should take great pains to try to pass the function the type of variable |
| 1069 | it wants, when possible. It will save you a lot of trouble in the long run. |
| 1070 | .Sh "The Argument Stack" |
| 1071 | .IX Subsection "The Argument Stack" |
| 1072 | If we look at any of the C code generated by any of the examples except |
| 1073 | example 1, you will notice a number of references to \s-1ST\s0(n), where n is |
| 1074 | usually 0. \*(L"\s-1ST\s0\*(R" is actually a macro that points to the n'th argument |
| 1075 | on the argument stack. \s-1\fIST\s0\fR\|(0) is thus the first argument on the stack and |
| 1076 | therefore the first argument passed to the \s-1XSUB\s0, \s-1\fIST\s0\fR\|(1) is the second |
| 1077 | argument, and so on. |
| 1078 | .PP |
| 1079 | When you list the arguments to the \s-1XSUB\s0 in the .xs file, that tells \fBxsubpp\fR |
| 1080 | which argument corresponds to which of the argument stack (i.e., the first |
| 1081 | one listed is the first argument, and so on). You invite disaster if you |
| 1082 | do not list them in the same order as the function expects them. |
| 1083 | .PP |
| 1084 | The actual values on the argument stack are pointers to the values passed |
| 1085 | in. When an argument is listed as being an \s-1OUTPUT\s0 value, its corresponding |
| 1086 | value on the stack (i.e., \s-1\fIST\s0\fR\|(0) if it was the first argument) is changed. |
| 1087 | You can verify this by looking at the C code generated for Example 3. |
| 1088 | The code for the \fIround()\fR \s-1XSUB\s0 routine contains lines that look like this: |
| 1089 | .PP |
| 1090 | .Vb 3 |
| 1091 | \& double arg = (double)SvNV(ST(0)); |
| 1092 | \& /* Round the contents of the variable arg */ |
| 1093 | \& sv_setnv(ST(0), (double)arg); |
| 1094 | .Ve |
| 1095 | .PP |
| 1096 | The arg variable is initially set by taking the value from \s-1\fIST\s0\fR\|(0), then is |
| 1097 | stored back into \s-1\fIST\s0\fR\|(0) at the end of the routine. |
| 1098 | .PP |
| 1099 | XSUBs are also allowed to return lists, not just scalars. This must be |
| 1100 | done by manipulating stack values \s-1\fIST\s0\fR\|(0), \s-1\fIST\s0\fR\|(1), etc, in a subtly |
| 1101 | different way. See perlxs for details. |
| 1102 | .PP |
| 1103 | XSUBs are also allowed to avoid automatic conversion of Perl function arguments |
| 1104 | to C function arguments. See perlxs for details. Some people prefer |
| 1105 | manual conversion by inspecting \f(CWST(i)\fR even in the cases when automatic |
| 1106 | conversion will do, arguing that this makes the logic of an \s-1XSUB\s0 call clearer. |
| 1107 | Compare with \*(L"Getting the fat out of XSUBs\*(R" for a similar tradeoff of |
| 1108 | a complete separation of \*(L"Perl glue\*(R" and \*(L"workhorse\*(R" parts of an \s-1XSUB\s0. |
| 1109 | .PP |
| 1110 | While experts may argue about these idioms, a novice to Perl guts may |
| 1111 | prefer a way which is as little Perl-guts-specific as possible, meaning |
| 1112 | automatic conversion and automatic call generation, as in |
| 1113 | \&\*(L"Getting the fat out of XSUBs\*(R". This approach has the additional |
| 1114 | benefit of protecting the \s-1XSUB\s0 writer from future changes to the Perl \s-1API\s0. |
| 1115 | .Sh "Extending your Extension" |
| 1116 | .IX Subsection "Extending your Extension" |
| 1117 | Sometimes you might want to provide some extra methods or subroutines |
| 1118 | to assist in making the interface between Perl and your extension simpler |
| 1119 | or easier to understand. These routines should live in the .pm file. |
| 1120 | Whether they are automatically loaded when the extension itself is loaded |
| 1121 | or only loaded when called depends on where in the .pm file the subroutine |
| 1122 | definition is placed. You can also consult AutoLoader for an alternate |
| 1123 | way to store and load your extra subroutines. |
| 1124 | .Sh "Documenting your Extension" |
| 1125 | .IX Subsection "Documenting your Extension" |
| 1126 | There is absolutely no excuse for not documenting your extension. |
| 1127 | Documentation belongs in the .pm file. This file will be fed to pod2man, |
| 1128 | and the embedded documentation will be converted to the manpage format, |
| 1129 | then placed in the blib directory. It will be copied to Perl's |
| 1130 | manpage directory when the extension is installed. |
| 1131 | .PP |
| 1132 | You may intersperse documentation and Perl code within the .pm file. |
| 1133 | In fact, if you want to use method autoloading, you must do this, |
| 1134 | as the comment inside the .pm file explains. |
| 1135 | .PP |
| 1136 | See perlpod for more information about the pod format. |
| 1137 | .Sh "Installing your Extension" |
| 1138 | .IX Subsection "Installing your Extension" |
| 1139 | Once your extension is complete and passes all its tests, installing it |
| 1140 | is quite simple: you simply run \*(L"make install\*(R". You will either need |
| 1141 | to have write permission into the directories where Perl is installed, |
| 1142 | or ask your system administrator to run the make for you. |
| 1143 | .PP |
| 1144 | Alternately, you can specify the exact directory to place the extension's |
| 1145 | files by placing a \*(L"PREFIX=/destination/directory\*(R" after the make install. |
| 1146 | (or in between the make and install if you have a brain-dead version of make). |
| 1147 | This can be very useful if you are building an extension that will eventually |
| 1148 | be distributed to multiple systems. You can then just archive the files in |
| 1149 | the destination directory and distribute them to your destination systems. |
| 1150 | .Sh "\s-1EXAMPLE\s0 5" |
| 1151 | .IX Subsection "EXAMPLE 5" |
| 1152 | In this example, we'll do some more work with the argument stack. The |
| 1153 | previous examples have all returned only a single value. We'll now |
| 1154 | create an extension that returns an array. |
| 1155 | .PP |
| 1156 | This extension is very Unix-oriented (struct statfs and the statfs system |
| 1157 | call). If you are not running on a Unix system, you can substitute for |
| 1158 | statfs any other function that returns multiple values, you can hard-code |
| 1159 | values to be returned to the caller (although this will be a bit harder |
| 1160 | to test the error case), or you can simply not do this example. If you |
| 1161 | change the \s-1XSUB\s0, be sure to fix the test cases to match the changes. |
| 1162 | .PP |
| 1163 | Return to the Mytest directory and add the following code to the end of |
| 1164 | Mytest.xs: |
| 1165 | .PP |
| 1166 | .Vb 6 |
| 1167 | \& void |
| 1168 | \& statfs(path) |
| 1169 | \& char * path |
| 1170 | \& INIT: |
| 1171 | \& int i; |
| 1172 | \& struct statfs buf; |
| 1173 | .Ve |
| 1174 | .PP |
| 1175 | .Vb 15 |
| 1176 | \& PPCODE: |
| 1177 | \& i = statfs(path, &buf); |
| 1178 | \& if (i == 0) { |
| 1179 | \& XPUSHs(sv_2mortal(newSVnv(buf.f_bavail))); |
| 1180 | \& XPUSHs(sv_2mortal(newSVnv(buf.f_bfree))); |
| 1181 | \& XPUSHs(sv_2mortal(newSVnv(buf.f_blocks))); |
| 1182 | \& XPUSHs(sv_2mortal(newSVnv(buf.f_bsize))); |
| 1183 | \& XPUSHs(sv_2mortal(newSVnv(buf.f_ffree))); |
| 1184 | \& XPUSHs(sv_2mortal(newSVnv(buf.f_files))); |
| 1185 | \& XPUSHs(sv_2mortal(newSVnv(buf.f_type))); |
| 1186 | \& XPUSHs(sv_2mortal(newSVnv(buf.f_fsid[0]))); |
| 1187 | \& XPUSHs(sv_2mortal(newSVnv(buf.f_fsid[1]))); |
| 1188 | \& } else { |
| 1189 | \& XPUSHs(sv_2mortal(newSVnv(errno))); |
| 1190 | \& } |
| 1191 | .Ve |
| 1192 | .PP |
| 1193 | You'll also need to add the following code to the top of the .xs file, just |
| 1194 | after the include of \*(L"\s-1XSUB\s0.h\*(R": |
| 1195 | .PP |
| 1196 | .Vb 1 |
| 1197 | \& #include <sys/vfs.h> |
| 1198 | .Ve |
| 1199 | .PP |
| 1200 | Also add the following code segment to test.pl while incrementing the \*(L"1..9\*(R" |
| 1201 | string in the \s-1BEGIN\s0 block to \*(L"1..11\*(R": |
| 1202 | .PP |
| 1203 | .Vb 4 |
| 1204 | \& @a = &Mytest::statfs("/blech"); |
| 1205 | \& print ((scalar(@a) == 1 && $a[0] == 2) ? "ok 10\en" : "not ok 10\en"); |
| 1206 | \& @a = &Mytest::statfs("/"); |
| 1207 | \& print scalar(@a) == 9 ? "ok 11\en" : "not ok 11\en"; |
| 1208 | .Ve |
| 1209 | .Sh "New Things in this Example" |
| 1210 | .IX Subsection "New Things in this Example" |
| 1211 | This example added quite a few new concepts. We'll take them one at a time. |
| 1212 | .IP "\(bu" 4 |
| 1213 | The \s-1INIT:\s0 directive contains code that will be placed immediately after |
| 1214 | the argument stack is decoded. C does not allow variable declarations at |
| 1215 | arbitrary locations inside a function, |
| 1216 | so this is usually the best way to declare local variables needed by the \s-1XSUB\s0. |
| 1217 | (Alternatively, one could put the whole \f(CW\*(C`PPCODE:\*(C'\fR section into braces, and |
| 1218 | put these declarations on top.) |
| 1219 | .IP "\(bu" 4 |
| 1220 | This routine also returns a different number of arguments depending on the |
| 1221 | success or failure of the call to statfs. If there is an error, the error |
| 1222 | number is returned as a single-element array. If the call is successful, |
| 1223 | then a 9\-element array is returned. Since only one argument is passed into |
| 1224 | this function, we need room on the stack to hold the 9 values which may be |
| 1225 | returned. |
| 1226 | .Sp |
| 1227 | We do this by using the \s-1PPCODE:\s0 directive, rather than the \s-1CODE:\s0 directive. |
| 1228 | This tells \fBxsubpp\fR that we will be managing the return values that will be |
| 1229 | put on the argument stack by ourselves. |
| 1230 | .IP "\(bu" 4 |
| 1231 | When we want to place values to be returned to the caller onto the stack, |
| 1232 | we use the series of macros that begin with \*(L"\s-1XPUSH\s0\*(R". There are five |
| 1233 | different versions, for placing integers, unsigned integers, doubles, |
| 1234 | strings, and Perl scalars on the stack. In our example, we placed a |
| 1235 | Perl scalar onto the stack. (In fact this is the only macro which |
| 1236 | can be used to return multiple values.) |
| 1237 | .Sp |
| 1238 | The XPUSH* macros will automatically extend the return stack to prevent |
| 1239 | it from being overrun. You push values onto the stack in the order you |
| 1240 | want them seen by the calling program. |
| 1241 | .IP "\(bu" 4 |
| 1242 | The values pushed onto the return stack of the \s-1XSUB\s0 are actually mortal \s-1SV\s0's. |
| 1243 | They are made mortal so that once the values are copied by the calling |
| 1244 | program, the \s-1SV\s0's that held the returned values can be deallocated. |
| 1245 | If they were not mortal, then they would continue to exist after the \s-1XSUB\s0 |
| 1246 | routine returned, but would not be accessible. This is a memory leak. |
| 1247 | .IP "\(bu" 4 |
| 1248 | If we were interested in performance, not in code compactness, in the success |
| 1249 | branch we would not use \f(CW\*(C`XPUSHs\*(C'\fR macros, but \f(CW\*(C`PUSHs\*(C'\fR macros, and would |
| 1250 | pre-extend the stack before pushing the return values: |
| 1251 | .Sp |
| 1252 | .Vb 1 |
| 1253 | \& EXTEND(SP, 9); |
| 1254 | .Ve |
| 1255 | .Sp |
| 1256 | The tradeoff is that one needs to calculate the number of return values |
| 1257 | in advance (though overextending the stack will not typically hurt |
| 1258 | anything but memory consumption). |
| 1259 | .Sp |
| 1260 | Similarly, in the failure branch we could use \f(CW\*(C`PUSHs\*(C'\fR \fIwithout\fR extending |
| 1261 | the stack: the Perl function reference comes to an \s-1XSUB\s0 on the stack, thus |
| 1262 | the stack is \fIalways\fR large enough to take one return value. |
| 1263 | .Sh "\s-1EXAMPLE\s0 6" |
| 1264 | .IX Subsection "EXAMPLE 6" |
| 1265 | In this example, we will accept a reference to an array as an input |
| 1266 | parameter, and return a reference to an array of hashes. This will |
| 1267 | demonstrate manipulation of complex Perl data types from an \s-1XSUB\s0. |
| 1268 | .PP |
| 1269 | This extension is somewhat contrived. It is based on the code in |
| 1270 | the previous example. It calls the statfs function multiple times, |
| 1271 | accepting a reference to an array of filenames as input, and returning |
| 1272 | a reference to an array of hashes containing the data for each of the |
| 1273 | filesystems. |
| 1274 | .PP |
| 1275 | Return to the Mytest directory and add the following code to the end of |
| 1276 | Mytest.xs: |
| 1277 | .PP |
| 1278 | .Vb 8 |
| 1279 | \& SV * |
| 1280 | \& multi_statfs(paths) |
| 1281 | \& SV * paths |
| 1282 | \& INIT: |
| 1283 | \& AV * results; |
| 1284 | \& I32 numpaths = 0; |
| 1285 | \& int i, n; |
| 1286 | \& struct statfs buf; |
| 1287 | .Ve |
| 1288 | .PP |
| 1289 | .Vb 12 |
| 1290 | \& if ((!SvROK(paths)) |
| 1291 | \& || (SvTYPE(SvRV(paths)) != SVt_PVAV) |
| 1292 | \& || ((numpaths = av_len((AV *)SvRV(paths))) < 0)) |
| 1293 | \& { |
| 1294 | \& XSRETURN_UNDEF; |
| 1295 | \& } |
| 1296 | \& results = (AV *)sv_2mortal((SV *)newAV()); |
| 1297 | \& CODE: |
| 1298 | \& for (n = 0; n <= numpaths; n++) { |
| 1299 | \& HV * rh; |
| 1300 | \& STRLEN l; |
| 1301 | \& char * fn = SvPV(*av_fetch((AV *)SvRV(paths), n, 0), l); |
| 1302 | .Ve |
| 1303 | .PP |
| 1304 | .Vb 5 |
| 1305 | \& i = statfs(fn, &buf); |
| 1306 | \& if (i != 0) { |
| 1307 | \& av_push(results, newSVnv(errno)); |
| 1308 | \& continue; |
| 1309 | \& } |
| 1310 | .Ve |
| 1311 | .PP |
| 1312 | .Vb 1 |
| 1313 | \& rh = (HV *)sv_2mortal((SV *)newHV()); |
| 1314 | .Ve |
| 1315 | .PP |
| 1316 | .Vb 7 |
| 1317 | \& hv_store(rh, "f_bavail", 8, newSVnv(buf.f_bavail), 0); |
| 1318 | \& hv_store(rh, "f_bfree", 7, newSVnv(buf.f_bfree), 0); |
| 1319 | \& hv_store(rh, "f_blocks", 8, newSVnv(buf.f_blocks), 0); |
| 1320 | \& hv_store(rh, "f_bsize", 7, newSVnv(buf.f_bsize), 0); |
| 1321 | \& hv_store(rh, "f_ffree", 7, newSVnv(buf.f_ffree), 0); |
| 1322 | \& hv_store(rh, "f_files", 7, newSVnv(buf.f_files), 0); |
| 1323 | \& hv_store(rh, "f_type", 6, newSVnv(buf.f_type), 0); |
| 1324 | .Ve |
| 1325 | .PP |
| 1326 | .Vb 5 |
| 1327 | \& av_push(results, newRV((SV *)rh)); |
| 1328 | \& } |
| 1329 | \& RETVAL = newRV((SV *)results); |
| 1330 | \& OUTPUT: |
| 1331 | \& RETVAL |
| 1332 | .Ve |
| 1333 | .PP |
| 1334 | And add the following code to test.pl, while incrementing the \*(L"1..11\*(R" |
| 1335 | string in the \s-1BEGIN\s0 block to \*(L"1..13\*(R": |
| 1336 | .PP |
| 1337 | .Vb 3 |
| 1338 | \& $results = Mytest::multi_statfs([ '/', '/blech' ]); |
| 1339 | \& print ((ref $results->[0]) ? "ok 12\en" : "not ok 12\en"); |
| 1340 | \& print ((! ref $results->[1]) ? "ok 13\en" : "not ok 13\en"); |
| 1341 | .Ve |
| 1342 | .Sh "New Things in this Example" |
| 1343 | .IX Subsection "New Things in this Example" |
| 1344 | There are a number of new concepts introduced here, described below: |
| 1345 | .IP "\(bu" 4 |
| 1346 | This function does not use a typemap. Instead, we declare it as accepting |
| 1347 | one SV* (scalar) parameter, and returning an SV* value, and we take care of |
| 1348 | populating these scalars within the code. Because we are only returning |
| 1349 | one value, we don't need a \f(CW\*(C`PPCODE:\*(C'\fR directive \- instead, we use \f(CW\*(C`CODE:\*(C'\fR |
| 1350 | and \f(CW\*(C`OUTPUT:\*(C'\fR directives. |
| 1351 | .IP "\(bu" 4 |
| 1352 | When dealing with references, it is important to handle them with caution. |
| 1353 | The \f(CW\*(C`INIT:\*(C'\fR block first checks that |
| 1354 | \&\f(CW\*(C`SvROK\*(C'\fR returns true, which indicates that paths is a valid reference. It |
| 1355 | then verifies that the object referenced by paths is an array, using \f(CW\*(C`SvRV\*(C'\fR |
| 1356 | to dereference paths, and \f(CW\*(C`SvTYPE\*(C'\fR to discover its type. As an added test, |
| 1357 | it checks that the array referenced by paths is non\-empty, using the \f(CW\*(C`av_len\*(C'\fR |
| 1358 | function (which returns \-1 if the array is empty). The \s-1XSRETURN_UNDEF\s0 macro |
| 1359 | is used to abort the \s-1XSUB\s0 and return the undefined value whenever all three of |
| 1360 | these conditions are not met. |
| 1361 | .IP "\(bu" 4 |
| 1362 | We manipulate several arrays in this \s-1XSUB\s0. Note that an array is represented |
| 1363 | internally by an AV* pointer. The functions and macros for manipulating |
| 1364 | arrays are similar to the functions in Perl: \f(CW\*(C`av_len\*(C'\fR returns the highest |
| 1365 | index in an AV*, much like $#array; \f(CW\*(C`av_fetch\*(C'\fR fetches a single scalar value |
| 1366 | from an array, given its index; \f(CW\*(C`av_push\*(C'\fR pushes a scalar value onto the |
| 1367 | end of the array, automatically extending the array as necessary. |
| 1368 | .Sp |
| 1369 | Specifically, we read pathnames one at a time from the input array, and |
| 1370 | store the results in an output array (results) in the same order. If |
| 1371 | statfs fails, the element pushed onto the return array is the value of |
| 1372 | errno after the failure. If statfs succeeds, though, the value pushed |
| 1373 | onto the return array is a reference to a hash containing some of the |
| 1374 | information in the statfs structure. |
| 1375 | .Sp |
| 1376 | As with the return stack, it would be possible (and a small performance win) |
| 1377 | to pre-extend the return array before pushing data into it, since we know |
| 1378 | how many elements we will return: |
| 1379 | .Sp |
| 1380 | .Vb 1 |
| 1381 | \& av_extend(results, numpaths); |
| 1382 | .Ve |
| 1383 | .IP "\(bu" 4 |
| 1384 | We are performing only one hash operation in this function, which is storing |
| 1385 | a new scalar under a key using \f(CW\*(C`hv_store\*(C'\fR. A hash is represented by an HV* |
| 1386 | pointer. Like arrays, the functions for manipulating hashes from an \s-1XSUB\s0 |
| 1387 | mirror the functionality available from Perl. See perlguts and perlapi |
| 1388 | for details. |
| 1389 | .IP "\(bu" 4 |
| 1390 | To create a reference, we use the \f(CW\*(C`newRV\*(C'\fR function. Note that you can |
| 1391 | cast an AV* or an HV* to type SV* in this case (and many others). This |
| 1392 | allows you to take references to arrays, hashes and scalars with the same |
| 1393 | function. Conversely, the \f(CW\*(C`SvRV\*(C'\fR function always returns an SV*, which may |
| 1394 | need to be cast to the appropriate type if it is something other than a |
| 1395 | scalar (check with \f(CW\*(C`SvTYPE\*(C'\fR). |
| 1396 | .IP "\(bu" 4 |
| 1397 | At this point, xsubpp is doing very little work \- the differences between |
| 1398 | Mytest.xs and Mytest.c are minimal. |
| 1399 | .Sh "\s-1EXAMPLE\s0 7 (Coming Soon)" |
| 1400 | .IX Subsection "EXAMPLE 7 (Coming Soon)" |
| 1401 | \&\s-1XPUSH\s0 args \s-1AND\s0 set \s-1RETVAL\s0 \s-1AND\s0 assign return value to array |
| 1402 | .Sh "\s-1EXAMPLE\s0 8 (Coming Soon)" |
| 1403 | .IX Subsection "EXAMPLE 8 (Coming Soon)" |
| 1404 | Setting $! |
| 1405 | .Sh "\s-1EXAMPLE\s0 9 Passing open files to XSes" |
| 1406 | .IX Subsection "EXAMPLE 9 Passing open files to XSes" |
| 1407 | You would think passing files to an \s-1XS\s0 is difficult, with all the |
| 1408 | typeglobs and stuff. Well, it isn't. |
| 1409 | .PP |
| 1410 | Suppose that for some strange reason we need a wrapper around the |
| 1411 | standard C library function \f(CW\*(C`fputs()\*(C'\fR. This is all we need: |
| 1412 | .PP |
| 1413 | .Vb 4 |
| 1414 | \& #define PERLIO_NOT_STDIO 0 |
| 1415 | \& #include "EXTERN.h" |
| 1416 | \& #include "perl.h" |
| 1417 | \& #include "XSUB.h" |
| 1418 | .Ve |
| 1419 | .PP |
| 1420 | .Vb 1 |
| 1421 | \& #include <stdio.h> |
| 1422 | .Ve |
| 1423 | .PP |
| 1424 | .Vb 4 |
| 1425 | \& int |
| 1426 | \& fputs(s, stream) |
| 1427 | \& char * s |
| 1428 | \& FILE * stream |
| 1429 | .Ve |
| 1430 | .PP |
| 1431 | The real work is done in the standard typemap. |
| 1432 | .PP |
| 1433 | \&\fBBut\fR you loose all the fine stuff done by the perlio layers. This |
| 1434 | calls the stdio function \f(CW\*(C`fputs()\*(C'\fR, which knows nothing about them. |
| 1435 | .PP |
| 1436 | The standard typemap offers three variants of PerlIO *: |
| 1437 | \&\f(CW\*(C`InputStream\*(C'\fR (T_IN), \f(CW\*(C`InOutStream\*(C'\fR (T_INOUT) and \f(CW\*(C`OutputStream\*(C'\fR |
| 1438 | (T_OUT). A bare \f(CW\*(C`PerlIO *\*(C'\fR is considered a T_INOUT. If it matters |
| 1439 | in your code (see below for why it might) #define or typedef |
| 1440 | one of the specific names and use that as the argument or result |
| 1441 | type in your \s-1XS\s0 file. |
| 1442 | .PP |
| 1443 | The standard typemap does not contain PerlIO * before perl 5.7, |
| 1444 | but it has the three stream variants. Using a PerlIO * directly |
| 1445 | is not backwards compatible unless you provide your own typemap. |
| 1446 | .PP |
| 1447 | For streams coming \fIfrom\fR perl the main difference is that |
| 1448 | \&\f(CW\*(C`OutputStream\*(C'\fR will get the output PerlIO * \- which may make |
| 1449 | a difference on a socket. Like in our example... |
| 1450 | .PP |
| 1451 | For streams being handed \fIto\fR perl a new file handle is created |
| 1452 | (i.e. a reference to a new glob) and associated with the PerlIO * |
| 1453 | provided. If the read/write state of the PerlIO * is not correct then you |
| 1454 | may get errors or warnings from when the file handle is used. |
| 1455 | So if you opened the PerlIO * as \*(L"w\*(R" it should really be an |
| 1456 | \&\f(CW\*(C`OutputStream\*(C'\fR if open as \*(L"r\*(R" it should be an \f(CW\*(C`InputStream\*(C'\fR. |
| 1457 | .PP |
| 1458 | Now, suppose you want to use perlio layers in your \s-1XS\s0. We'll use the |
| 1459 | perlio \f(CW\*(C`PerlIO_puts()\*(C'\fR function as an example. |
| 1460 | .PP |
| 1461 | In the C part of the \s-1XS\s0 file (above the first \s-1MODULE\s0 line) you |
| 1462 | have |
| 1463 | .PP |
| 1464 | .Vb 3 |
| 1465 | \& #define OutputStream PerlIO * |
| 1466 | \& or |
| 1467 | \& typedef PerlIO * OutputStream; |
| 1468 | .Ve |
| 1469 | .PP |
| 1470 | And this is the \s-1XS\s0 code: |
| 1471 | .PP |
| 1472 | .Vb 8 |
| 1473 | \& int |
| 1474 | \& perlioputs(s, stream) |
| 1475 | \& char * s |
| 1476 | \& OutputStream stream |
| 1477 | \& CODE: |
| 1478 | \& RETVAL = PerlIO_puts(stream, s); |
| 1479 | \& OUTPUT: |
| 1480 | \& RETVAL |
| 1481 | .Ve |
| 1482 | .PP |
| 1483 | We have to use a \f(CW\*(C`CODE\*(C'\fR section because \f(CW\*(C`PerlIO_puts()\*(C'\fR has the arguments |
| 1484 | reversed compared to \f(CW\*(C`fputs()\*(C'\fR, and we want to keep the arguments the same. |
| 1485 | .PP |
| 1486 | Wanting to explore this thoroughly, we want to use the stdio \f(CW\*(C`fputs()\*(C'\fR |
| 1487 | on a PerlIO *. This means we have to ask the perlio system for a stdio |
| 1488 | \&\f(CW\*(C`FILE *\*(C'\fR: |
| 1489 | .PP |
| 1490 | .Vb 14 |
| 1491 | \& int |
| 1492 | \& perliofputs(s, stream) |
| 1493 | \& char * s |
| 1494 | \& OutputStream stream |
| 1495 | \& PREINIT: |
| 1496 | \& FILE *fp = PerlIO_findFILE(stream); |
| 1497 | \& CODE: |
| 1498 | \& if (fp != (FILE*) 0) { |
| 1499 | \& RETVAL = fputs(s, fp); |
| 1500 | \& } else { |
| 1501 | \& RETVAL = -1; |
| 1502 | \& } |
| 1503 | \& OUTPUT: |
| 1504 | \& RETVAL |
| 1505 | .Ve |
| 1506 | .PP |
| 1507 | Note: \f(CW\*(C`PerlIO_findFILE()\*(C'\fR will search the layers for a stdio |
| 1508 | layer. If it can't find one, it will call \f(CW\*(C`PerlIO_exportFILE()\*(C'\fR to |
| 1509 | generate a new stdio \f(CW\*(C`FILE\*(C'\fR. Please only call \f(CW\*(C`PerlIO_exportFILE()\*(C'\fR if |
| 1510 | you want a \fInew\fR \f(CW\*(C`FILE\*(C'\fR. It will generate one on each call and push a |
| 1511 | new stdio layer. So don't call it repeatedly on the same |
| 1512 | file. \f(CW\*(C`PerlIO()\*(C'\fR_findFILE will retrieve the stdio layer once it has been |
| 1513 | generated by \f(CW\*(C`PerlIO_exportFILE()\*(C'\fR. |
| 1514 | .PP |
| 1515 | This applies to the perlio system only. For versions before 5.7, |
| 1516 | \&\f(CW\*(C`PerlIO_exportFILE()\*(C'\fR is equivalent to \f(CW\*(C`PerlIO_findFILE()\*(C'\fR. |
| 1517 | .Sh "Troubleshooting these Examples" |
| 1518 | .IX Subsection "Troubleshooting these Examples" |
| 1519 | As mentioned at the top of this document, if you are having problems with |
| 1520 | these example extensions, you might see if any of these help you. |
| 1521 | .IP "\(bu" 4 |
| 1522 | In versions of 5.002 prior to the gamma version, the test script in Example |
| 1523 | 1 will not function properly. You need to change the \*(L"use lib\*(R" line to |
| 1524 | read: |
| 1525 | .Sp |
| 1526 | .Vb 1 |
| 1527 | \& use lib './blib'; |
| 1528 | .Ve |
| 1529 | .IP "\(bu" 4 |
| 1530 | In versions of 5.002 prior to version 5.002b1h, the test.pl file was not |
| 1531 | automatically created by h2xs. This means that you cannot say \*(L"make test\*(R" |
| 1532 | to run the test script. You will need to add the following line before the |
| 1533 | \&\*(L"use extension\*(R" statement: |
| 1534 | .Sp |
| 1535 | .Vb 1 |
| 1536 | \& use lib './blib'; |
| 1537 | .Ve |
| 1538 | .IP "\(bu" 4 |
| 1539 | In versions 5.000 and 5.001, instead of using the above line, you will need |
| 1540 | to use the following line: |
| 1541 | .Sp |
| 1542 | .Vb 1 |
| 1543 | \& BEGIN { unshift(@INC, "./blib") } |
| 1544 | .Ve |
| 1545 | .IP "\(bu" 4 |
| 1546 | This document assumes that the executable named \*(L"perl\*(R" is Perl version 5. |
| 1547 | Some systems may have installed Perl version 5 as \*(L"perl5\*(R". |
| 1548 | .SH "See also" |
| 1549 | .IX Header "See also" |
| 1550 | For more information, consult perlguts, perlapi, perlxs, perlmod, |
| 1551 | and perlpod. |
| 1552 | .SH "Author" |
| 1553 | .IX Header "Author" |
| 1554 | Jeff Okamoto <\fIokamoto@corp.hp.com\fR> |
| 1555 | .PP |
| 1556 | Reviewed and assisted by Dean Roehrich, Ilya Zakharevich, Andreas Koenig, |
| 1557 | and Tim Bunce. |
| 1558 | .PP |
| 1559 | PerlIO material contributed by Lupe Christoph, with some clarification |
| 1560 | by Nick Ing\-Simmons. |
| 1561 | .Sh "Last Changed" |
| 1562 | .IX Subsection "Last Changed" |
| 1563 | 2002/05/08 |