This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
perliol tweaks.
[perl5.git] / pod / perliol.pod
CommitLineData
50b80e25
NIS
1=head1 NAME
2
3perliol - C API for Perl's implementation of IO in Layers.
4
5=head1 SYNOPSIS
6
7 /* Defining a layer ... */
8 #include <perliol.h>
9
50b80e25
NIS
10=head1 DESCRIPTION
11
9d799145
NIS
12This document describes the behavior and implementation of the PerlIO
13abstraction described in L<perlapio> when C<USE_PERLIO> is defined (and
14C<USE_SFIO> is not).
50b80e25
NIS
15
16=head2 History and Background
17
9d799145
NIS
18The PerlIO abstraction was introduced in perl5.003_02 but languished as
19just an abstraction until perl5.7.0. However during that time a number
d1be9408 20of perl extensions switched to using it, so the API is mostly fixed to
9d799145 21maintain (source) compatibility.
50b80e25 22
9d799145
NIS
23The aim of the implementation is to provide the PerlIO API in a flexible
24and platform neutral manner. It is also a trial of an "Object Oriented
25C, with vtables" approach which may be applied to perl6.
50b80e25 26
cc83745d
JH
27=head2 Basic Structure
28
cc7ef057 29PerlIO is a stack of layers.
cc83745d
JH
30
31The low levels of the stack work with the low-level operating system
32calls (file descriptors in C) getting bytes in and out, the higher
cc7ef057
JH
33layers of the stack buffer, filter, and otherwise manipulate the I/O,
34and return characters (or bytes) to Perl. Terms I<above> and I<below>
35are used to refer to the relative positioning of the stack layers.
cc83745d
JH
36
37A layer contains a "vtable", the table of I/O operations (at C level
38a table of function pointers), and status flags. The functions in the
39vtable implement operations like "open", "read", and "write".
40
41When I/O, for example "read", is requested, the request goes from Perl
42first down the stack using "read" functions of each layer, then at the
43bottom the input is requested from the operating system services, then
44the result is returned up the stack, finally being interpreted as Perl
45data.
46
cc7ef057
JH
47The requests do not necessarily go always all the way down to the
48operating system: that's where PerlIO buffering comes into play.
49
cc83745d
JH
50When you do an open() and specify extra PerlIO layers to be deployed,
51the layers you specify are "pushed" on top of the already existing
cc7ef057
JH
52default stack. One way to see it is that "operating system is
53on the left" and "Perl is on the right".
54
55What exact layers are in this default stack depends on a lot of
56things: your operating system, Perl version, Perl compile time
57configuration, and Perl runtime configuration. See L<PerlIO>,
cc83745d
JH
58L<perlrun/PERLIO>, and L<open> for more information.
59
60binmode() operates similarly to open(): by default the specified
61layers are pushed on top of the existing stack.
62
63However, note that even as the specified layers are "pushed on top"
64for open() and binmode(), this doesn't mean that the effects are
65limited to the "top": PerlIO layers can be very 'active' and inspect
66and affect layers also deeper in the stack. As an example there
67is a layer called "raw" which repeatedly "pops" layers until
68it reaches the first layer that has declared itself capable of
69handling binary data. The "pushed" layers are processed in left-to-right
70order.
71
72sysopen() operates (unsurprisingly) at a lower level in the stack than
73open(). For example in UNIX or UNIX-like systems sysopen() operates
74directly at the level of file descriptors: in the terms of PerlIO
75layers, it uses only the "unix" layer, which is a rather thin wrapper
76on top of the UNIX file descriptors.
77
50b80e25
NIS
78=head2 Layers vs Disciplines
79
9d799145
NIS
80Initial discussion of the ability to modify IO streams behaviour used
81the term "discipline" for the entities which were added. This came (I
82believe) from the use of the term in "sfio", which in turn borrowed it
83from "line disciplines" on Unix terminals. However, this document (and
84the C code) uses the term "layer".
85
1d11c889
JH
86This is, I hope, a natural term given the implementation, and should
87avoid connotations that are inherent in earlier uses of "discipline"
88for things which are rather different.
50b80e25
NIS
89
90=head2 Data Structures
91
92The basic data structure is a PerlIOl:
93
94 typedef struct _PerlIO PerlIOl;
95 typedef struct _PerlIO_funcs PerlIO_funcs;
96 typedef PerlIOl *PerlIO;
97
98 struct _PerlIO
99 {
100 PerlIOl * next; /* Lower layer */
101 PerlIO_funcs * tab; /* Functions for this layer */
102 IV flags; /* Various flags for state */
103 };
104
1d11c889
JH
105A C<PerlIOl *> is a pointer to the struct, and the I<application>
106level C<PerlIO *> is a pointer to a C<PerlIOl *> - i.e. a pointer
107to a pointer to the struct. This allows the application level C<PerlIO *>
108to remain constant while the actual C<PerlIOl *> underneath
109changes. (Compare perl's C<SV *> which remains constant while its
110C<sv_any> field changes as the scalar's type changes.) An IO stream is
111then in general represented as a pointer to this linked-list of
112"layers".
50b80e25 113
9d799145 114It should be noted that because of the double indirection in a C<PerlIO *>,
d4165bde 115a C<< &(perlio->next) >> "is" a C<PerlIO *>, and so to some degree
11e1c8f2 116at least one layer can use the "standard" API on the next layer down.
50b80e25
NIS
117
118A "layer" is composed of two parts:
119
120=over 4
121
210b36aa 122=item 1.
50b80e25 123
210b36aa
AMS
124The functions and attributes of the "layer class".
125
126=item 2.
127
128The per-instance data for a particular handle.
50b80e25
NIS
129
130=back
131
132=head2 Functions and Attributes
133
9d799145
NIS
134The functions and attributes are accessed via the "tab" (for table)
135member of C<PerlIOl>. The functions (methods of the layer "class") are
136fixed, and are defined by the C<PerlIO_funcs> type. They are broadly the
137same as the public C<PerlIO_xxxxx> functions:
50b80e25 138
b76cc8ba
NIS
139 struct _PerlIO_funcs
140 {
2dc2558e 141 Size_t fsize;
b76cc8ba
NIS
142 char * name;
143 Size_t size;
144 IV kind;
2dc2558e 145 IV (*Pushed)(pTHX_ PerlIO *f,const char *mode,SV *arg, PerlIO_funcs *tab);
d4165bde 146 IV (*Popped)(pTHX_ PerlIO *f);
b76cc8ba
NIS
147 PerlIO * (*Open)(pTHX_ PerlIO_funcs *tab,
148 AV *layers, IV n,
149 const char *mode,
150 int fd, int imode, int perm,
151 PerlIO *old,
152 int narg, SV **args);
86e05cf2 153 IV (*Binmode)(pTHX_ PerlIO *f);
d4165bde
SB
154 SV * (*Getarg)(pTHX_ PerlIO *f, CLONE_PARAMS *param, int flags)
155 IV (*Fileno)(pTHX_ PerlIO *f);
156 PerlIO * (*Dup)(pTHX_ PerlIO *f, PerlIO *o, CLONE_PARAMS *param, int flags)
b76cc8ba 157 /* Unix-like functions - cf sfio line disciplines */
d4165bde
SB
158 SSize_t (*Read)(pTHX_ PerlIO *f, void *vbuf, Size_t count);
159 SSize_t (*Unread)(pTHX_ PerlIO *f, const void *vbuf, Size_t count);
160 SSize_t (*Write)(pTHX_ PerlIO *f, const void *vbuf, Size_t count);
161 IV (*Seek)(pTHX_ PerlIO *f, Off_t offset, int whence);
162 Off_t (*Tell)(pTHX_ PerlIO *f);
163 IV (*Close)(pTHX_ PerlIO *f);
b76cc8ba 164 /* Stdio-like buffered IO functions */
d4165bde
SB
165 IV (*Flush)(pTHX_ PerlIO *f);
166 IV (*Fill)(pTHX_ PerlIO *f);
167 IV (*Eof)(pTHX_ PerlIO *f);
168 IV (*Error)(pTHX_ PerlIO *f);
169 void (*Clearerr)(pTHX_ PerlIO *f);
170 void (*Setlinebuf)(pTHX_ PerlIO *f);
b76cc8ba 171 /* Perl's snooping functions */
d4165bde
SB
172 STDCHAR * (*Get_base)(pTHX_ PerlIO *f);
173 Size_t (*Get_bufsiz)(pTHX_ PerlIO *f);
174 STDCHAR * (*Get_ptr)(pTHX_ PerlIO *f);
175 SSize_t (*Get_cnt)(pTHX_ PerlIO *f);
176 void (*Set_ptrcnt)(pTHX_ PerlIO *f,STDCHAR *ptr,SSize_t cnt);
b76cc8ba
NIS
177 };
178
2dc2558e
NIS
179The first few members of the struct give a function table size for
180compatibility check "name" for the layer, the size to C<malloc> for the per-instance data,
181and some flags which are attributes of the class as whole (such as whether it is a buffering
9d799145 182layer), then follow the functions which fall into four basic groups:
50b80e25
NIS
183
184=over 4
185
aa500c9e 186=item 1.
50b80e25 187
aa500c9e 188Opening and setup functions
50b80e25 189
aa500c9e 190=item 2.
50b80e25 191
aa500c9e
RB
192Basic IO operations
193
194=item 3.
195
196Stdio class buffering options.
197
198=item 4.
199
200Functions to support Perl's traditional "fast" access to the buffer.
50b80e25
NIS
201
202=back
203
1d11c889
JH
204A layer does not have to implement all the functions, but the whole
205table has to be present. Unimplemented slots can be NULL (which will
206result in an error when called) or can be filled in with stubs to
207"inherit" behaviour from a "base class". This "inheritance" is fixed
208for all instances of the layer, but as the layer chooses which stubs
209to populate the table, limited "multiple inheritance" is possible.
50b80e25
NIS
210
211=head2 Per-instance Data
212
1d11c889
JH
213The per-instance data are held in memory beyond the basic PerlIOl
214struct, by making a PerlIOl the first member of the layer's struct
215thus:
50b80e25
NIS
216
217 typedef struct
218 {
219 struct _PerlIO base; /* Base "class" info */
220 STDCHAR * buf; /* Start of buffer */
221 STDCHAR * end; /* End of valid part of buffer */
222 STDCHAR * ptr; /* Current position in buffer */
223 Off_t posn; /* Offset of buf into the file */
224 Size_t bufsiz; /* Real size of buffer */
225 IV oneword; /* Emergency buffer */
226 } PerlIOBuf;
227
1d11c889
JH
228In this way (as for perl's scalars) a pointer to a PerlIOBuf can be
229treated as a pointer to a PerlIOl.
50b80e25
NIS
230
231=head2 Layers in action.
232
233 table perlio unix
234 | |
235 +-----------+ +----------+ +--------+
236 PerlIO ->| |--->| next |--->| NULL |
237 +-----------+ +----------+ +--------+
238 | | | buffer | | fd |
239 +-----------+ | | +--------+
240 | | +----------+
241
242
243The above attempts to show how the layer scheme works in a simple case.
9d799145
NIS
244The application's C<PerlIO *> points to an entry in the table(s)
245representing open (allocated) handles. For example the first three slots
246in the table correspond to C<stdin>,C<stdout> and C<stderr>. The table
247in turn points to the current "top" layer for the handle - in this case
248an instance of the generic buffering layer "perlio". That layer in turn
249points to the next layer down - in this case the lowlevel "unix" layer.
50b80e25 250
9d799145
NIS
251The above is roughly equivalent to a "stdio" buffered stream, but with
252much more flexibility:
50b80e25
NIS
253
254=over 4
255
256=item *
257
9d799145
NIS
258If Unix level C<read>/C<write>/C<lseek> is not appropriate for (say)
259sockets then the "unix" layer can be replaced (at open time or even
260dynamically) with a "socket" layer.
50b80e25
NIS
261
262=item *
263
1d11c889
JH
264Different handles can have different buffering schemes. The "top"
265layer could be the "mmap" layer if reading disk files was quicker
266using C<mmap> than C<read>. An "unbuffered" stream can be implemented
267simply by not having a buffer layer.
50b80e25
NIS
268
269=item *
270
271Extra layers can be inserted to process the data as it flows through.
9d799145 272This was the driving need for including the scheme in perl 5.7.0+ - we
d1be9408 273needed a mechanism to allow data to be translated between perl's
9d799145
NIS
274internal encoding (conceptually at least Unicode as UTF-8), and the
275"native" format used by the system. This is provided by the
276":encoding(xxxx)" layer which typically sits above the buffering layer.
50b80e25
NIS
277
278=item *
279
1d11c889
JH
280A layer can be added that does "\n" to CRLF translation. This layer
281can be used on any platform, not just those that normally do such
282things.
50b80e25
NIS
283
284=back
285
286=head2 Per-instance flag bits
287
1d11c889
JH
288The generic flag bits are a hybrid of C<O_XXXXX> style flags deduced
289from the mode string passed to C<PerlIO_open()>, and state bits for
290typical buffer layers.
50b80e25 291
9d799145 292=over 4
50b80e25
NIS
293
294=item PERLIO_F_EOF
295
296End of file.
297
298=item PERLIO_F_CANWRITE
299
3039a93d 300Writes are permitted, i.e. opened as "w" or "r+" or "a", etc.
50b80e25
NIS
301
302=item PERLIO_F_CANREAD
303
3039a93d 304Reads are permitted i.e. opened "r" or "w+" (or even "a+" - ick).
50b80e25
NIS
305
306=item PERLIO_F_ERROR
307
d4165bde 308An error has occurred (for C<PerlIO_error()>).
50b80e25
NIS
309
310=item PERLIO_F_TRUNCATE
311
312Truncate file suggested by open mode.
313
314=item PERLIO_F_APPEND
315
316All writes should be appends.
317
318=item PERLIO_F_CRLF
319
11e1c8f2
PP
320Layer is performing Win32-like "\n" mapped to CR,LF for output and CR,LF
321mapped to "\n" for input. Normally the provided "crlf" layer is the only
322layer that need bother about this. C<PerlIO_binmode()> will mess with this
9d799145
NIS
323flag rather than add/remove layers if the C<PERLIO_K_CANCRLF> bit is set
324for the layers class.
50b80e25
NIS
325
326=item PERLIO_F_UTF8
327
3039a93d 328Data written to this layer should be UTF-8 encoded; data provided
50b80e25
NIS
329by this layer should be considered UTF-8 encoded. Can be set on any layer
330by ":utf8" dummy layer. Also set on ":encoding" layer.
331
332=item PERLIO_F_UNBUF
333
334Layer is unbuffered - i.e. write to next layer down should occur for
335each write to this layer.
336
337=item PERLIO_F_WRBUF
338
339The buffer for this layer currently holds data written to it but not sent
340to next layer.
341
342=item PERLIO_F_RDBUF
343
344The buffer for this layer currently holds unconsumed data read from
345layer below.
346
347=item PERLIO_F_LINEBUF
348
9d799145
NIS
349Layer is line buffered. Write data should be passed to next layer down
350whenever a "\n" is seen. Any data beyond the "\n" should then be
351processed.
50b80e25
NIS
352
353=item PERLIO_F_TEMP
354
9d799145 355File has been C<unlink()>ed, or should be deleted on C<close()>.
50b80e25
NIS
356
357=item PERLIO_F_OPEN
358
359Handle is open.
360
361=item PERLIO_F_FASTGETS
362
9d799145
NIS
363This instance of this layer supports the "fast C<gets>" interface.
364Normally set based on C<PERLIO_K_FASTGETS> for the class and by the
d1be9408 365existence of the function(s) in the table. However a class that
50b80e25
NIS
366normally provides that interface may need to avoid it on a
367particular instance. The "pending" layer needs to do this when
d1be9408 368it is pushed above a layer which does not support the interface.
9d799145 369(Perl's C<sv_gets()> does not expect the streams fast C<gets> behaviour
50b80e25
NIS
370to change during one "get".)
371
372=back
373
374=head2 Methods in Detail
375
376=over 4
377
e2d9456f 378=item fsize
2dc2558e
NIS
379
380 Size_t fsize;
381
a489db4d
JH
382Size of the function table. This is compared against the value PerlIO
383code "knows" as a compatibility check. Future versions I<may> be able
384to tolerate layers compiled against an old version of the headers.
2dc2558e 385
5cb3728c
RB
386=item name
387
388 char * name;
d4165bde
SB
389
390The name of the layer whose open() method Perl should invoke on
391open(). For example if the layer is called APR, you will call:
392
393 open $fh, ">:APR", ...
394
395and Perl knows that it has to invoke the PerlIOAPR_open() method
396implemented by the APR layer.
397
5cb3728c
RB
398=item size
399
400 Size_t size;
d4165bde
SB
401
402The size of the per-instance data structure, e.g.:
403
404 sizeof(PerlIOAPR)
405
a489db4d
JH
406If this field is zero then C<PerlIO_pushed> does not malloc anything
407and assumes layer's Pushed function will do any required layer stack
408manipulation - used to avoid malloc/free overhead for dummy layers.
2dc2558e
NIS
409If the field is non-zero it must be at least the size of C<PerlIOl>,
410C<PerlIO_pushed> will allocate memory for the layer's data structures
411and link new layer onto the stream's stack. (If the layer's Pushed
412method returns an error indication the layer is popped again.)
413
5cb3728c
RB
414=item kind
415
416 IV kind;
d4165bde 417
d4165bde
SB
418=over 4
419
420=item * PERLIO_K_BUFFERED
421
86e05cf2
NIS
422The layer is buffered.
423
424=item * PERLIO_K_RAW
425
426The layer is acceptable to have in a binmode(FH) stack - i.e. it does not
427(or will configure itself not to) transform bytes passing through it.
428
d4165bde
SB
429=item * PERLIO_K_CANCRLF
430
86e05cf2
NIS
431Layer can translate between "\n" and CRLF line ends.
432
d4165bde
SB
433=item * PERLIO_K_FASTGETS
434
86e05cf2
NIS
435Layer allows buffer snooping.
436
d4165bde
SB
437=item * PERLIO_K_MULTIARG
438
439Used when the layer's open() accepts more arguments than usual. The
440extra arguments should come not before the C<MODE> argument. When this
441flag is used it's up to the layer to validate the args.
442
d4165bde
SB
443=back
444
5cb3728c
RB
445=item Pushed
446
447 IV (*Pushed)(pTHX_ PerlIO *f,const char *mode, SV *arg);
50b80e25 448
1d11c889
JH
449The only absolutely mandatory method. Called when the layer is pushed
450onto the stack. The C<mode> argument may be NULL if this occurs
451post-open. The C<arg> will be non-C<NULL> if an argument string was
452passed. In most cases this should call C<PerlIOBase_pushed()> to
453convert C<mode> into the appropriate C<PERLIO_F_XXXXX> flags in
454addition to any actions the layer itself takes. If a layer is not
455expecting an argument it need neither save the one passed to it, nor
456provide C<Getarg()> (it could perhaps C<Perl_warn> that the argument
457was un-expected).
50b80e25 458
d4165bde
SB
459Returns 0 on success. On failure returns -1 and should set errno.
460
5cb3728c
RB
461=item Popped
462
463 IV (*Popped)(pTHX_ PerlIO *f);
50b80e25 464
1d11c889
JH
465Called when the layer is popped from the stack. A layer will normally
466be popped after C<Close()> is called. But a layer can be popped
467without being closed if the program is dynamically managing layers on
468the stream. In such cases C<Popped()> should free any resources
469(buffers, translation tables, ...) not held directly in the layer's
470struct. It should also C<Unread()> any unconsumed data that has been
471read and buffered from the layer below back to that layer, so that it
472can be re-provided to what ever is now above.
b76cc8ba 473
3077d0b1
JH
474Returns 0 on success and failure. If C<Popped()> returns I<true> then
475I<perlio.c> assumes that either the layer has popped itself, or the
476layer is super special and needs to be retained for other reasons.
477In most cases it should return I<false>.
d4165bde 478
5cb3728c
RB
479=item Open
480
481 PerlIO * (*Open)(...);
b76cc8ba 482
1d11c889
JH
483The C<Open()> method has lots of arguments because it combines the
484functions of perl's C<open>, C<PerlIO_open>, perl's C<sysopen>,
485C<PerlIO_fdopen> and C<PerlIO_reopen>. The full prototype is as
486follows:
b76cc8ba
NIS
487
488 PerlIO * (*Open)(pTHX_ PerlIO_funcs *tab,
489 AV *layers, IV n,
490 const char *mode,
491 int fd, int imode, int perm,
492 PerlIO *old,
493 int narg, SV **args);
494
1d11c889
JH
495Open should (perhaps indirectly) call C<PerlIO_allocate()> to allocate
496a slot in the table and associate it with the layers information for
497the opened file, by calling C<PerlIO_push>. The I<layers> AV is an
498array of all the layers destined for the C<PerlIO *>, and any
499arguments passed to them, I<n> is the index into that array of the
500layer being called. The macro C<PerlIOArg> will return a (possibly
501C<NULL>) SV * for the argument passed to the layer.
502
503The I<mode> string is an "C<fopen()>-like" string which would match
504the regular expression C</^[I#]?[rwa]\+?[bt]?$/>.
505
506The C<'I'> prefix is used during creation of C<stdin>..C<stderr> via
507special C<PerlIO_fdopen> calls; the C<'#'> prefix means that this is
508C<sysopen> and that I<imode> and I<perm> should be passed to
509C<PerlLIO_open3>; C<'r'> means B<r>ead, C<'w'> means B<w>rite and
510C<'a'> means B<a>ppend. The C<'+'> suffix means that both reading and
a489db4d
JH
511writing/appending are permitted. The C<'b'> suffix means file should
512be binary, and C<'t'> means it is text. (Almost all layers should do
513the IO in binary mode, and ignore the b/t bits. The C<:crlf> layer
514should be pushed to handle the distinction.)
1d11c889
JH
515
516If I<old> is not C<NULL> then this is a C<PerlIO_reopen>. Perl itself
517does not use this (yet?) and semantics are a little vague.
518
519If I<fd> not negative then it is the numeric file descriptor I<fd>,
520which will be open in a manner compatible with the supplied mode
521string, the call is thus equivalent to C<PerlIO_fdopen>. In this case
522I<nargs> will be zero.
523
524If I<nargs> is greater than zero then it gives the number of arguments
525passed to C<open>, otherwise it will be 1 if for example
526C<PerlIO_open> was called. In simple cases SvPV_nolen(*args) is the
527pathname to open.
528
529Having said all that translation-only layers do not need to provide
530C<Open()> at all, but rather leave the opening to a lower level layer
531and wait to be "pushed". If a layer does provide C<Open()> it should
532normally call the C<Open()> method of next layer down (if any) and
533then push itself on top if that succeeds.
b76cc8ba 534
3077d0b1
JH
535If C<PerlIO_push> was performed and open has failed, it must
536C<PerlIO_pop> itself, since if it's not, the layer won't be removed
537and may cause bad problems.
538
d4165bde
SB
539Returns C<NULL> on failure.
540
86e05cf2
NIS
541=item Binmode
542
543 IV (*Binmode)(pTHX_ PerlIO *f);
544
545Optional. Used when C<:raw> layer is pushed (explicitly or as a result
546of binmode(FH)). If not present layer will be popped. If present
547should configure layer as binary (or pop itself) and return 0.
548If it returns -1 for error C<binmode> will fail with layer
549still on the stack.
550
5cb3728c
RB
551=item Getarg
552
553 SV * (*Getarg)(pTHX_ PerlIO *f,
554 CLONE_PARAMS *param, int flags);
b76cc8ba 555
d4165bde
SB
556Optional. If present should return an SV * representing the string
557argument passed to the layer when it was
558pushed. e.g. ":encoding(ascii)" would return an SvPV with value
559"ascii". (I<param> and I<flags> arguments can be ignored in most
560cases)
b76cc8ba 561
5cb3728c
RB
562=item Fileno
563
564 IV (*Fileno)(pTHX_ PerlIO *f);
b76cc8ba 565
d1be9408 566Returns the Unix/Posix numeric file descriptor for the handle. Normally
b76cc8ba
NIS
567C<PerlIOBase_fileno()> (which just asks next layer down) will suffice
568for this.
50b80e25 569
a489db4d
JH
570Returns -1 on error, which is considered to include the case where the
571layer cannot provide such a file descriptor.
d4165bde 572
5cb3728c
RB
573=item Dup
574
575 PerlIO * (*Dup)(pTHX_ PerlIO *f, PerlIO *o,
576 CLONE_PARAMS *param, int flags);
d4165bde 577
2dc2558e
NIS
578XXX: Needs more docs.
579
a489db4d
JH
580Used as part of the "clone" process when a thread is spawned (in which
581case param will be non-NULL) and when a stream is being duplicated via
582'&' in the C<open>.
d4165bde
SB
583
584Similar to C<Open>, returns PerlIO* on success, C<NULL> on failure.
585
5cb3728c
RB
586=item Read
587
588 SSize_t (*Read)(pTHX_ PerlIO *f, void *vbuf, Size_t count);
d4165bde
SB
589
590Basic read operation.
50b80e25 591
d4165bde
SB
592Typically will call C<Fill> and manipulate pointers (possibly via the
593API). C<PerlIOBuf_read()> may be suitable for derived classes which
594provide "fast gets" methods.
50b80e25 595
d4165bde
SB
596Returns actual bytes read, or -1 on an error.
597
5cb3728c
RB
598=item Unread
599
600 SSize_t (*Unread)(pTHX_ PerlIO *f,
601 const void *vbuf, Size_t count);
50b80e25 602
9d799145
NIS
603A superset of stdio's C<ungetc()>. Should arrange for future reads to
604see the bytes in C<vbuf>. If there is no obviously better implementation
605then C<PerlIOBase_unread()> provides the function by pushing a "fake"
606"pending" layer above the calling layer.
50b80e25 607
d4165bde
SB
608Returns the number of unread chars.
609
5cb3728c
RB
610=item Write
611
612 SSize_t (*Write)(PerlIO *f, const void *vbuf, Size_t count);
50b80e25 613
d4165bde 614Basic write operation.
50b80e25 615
d4165bde
SB
616Returns bytes written or -1 on an error.
617
5cb3728c
RB
618=item Seek
619
620 IV (*Seek)(pTHX_ PerlIO *f, Off_t offset, int whence);
50b80e25 621
1d11c889
JH
622Position the file pointer. Should normally call its own C<Flush>
623method and then the C<Seek> method of next layer down.
50b80e25 624
d4165bde
SB
625Returns 0 on success, -1 on failure.
626
5cb3728c
RB
627=item Tell
628
629 Off_t (*Tell)(pTHX_ PerlIO *f);
50b80e25 630
9d799145
NIS
631Return the file pointer. May be based on layers cached concept of
632position to avoid overhead.
50b80e25 633
d4165bde
SB
634Returns -1 on failure to get the file pointer.
635
5cb3728c
RB
636=item Close
637
638 IV (*Close)(pTHX_ PerlIO *f);
50b80e25 639
9d799145
NIS
640Close the stream. Should normally call C<PerlIOBase_close()> to flush
641itself and close layers below, and then deallocate any data structures
642(buffers, translation tables, ...) not held directly in the data
643structure.
50b80e25 644
d4165bde
SB
645Returns 0 on success, -1 on failure.
646
5cb3728c
RB
647=item Flush
648
649 IV (*Flush)(pTHX_ PerlIO *f);
50b80e25 650
9d799145
NIS
651Should make stream's state consistent with layers below. That is, any
652buffered write data should be written, and file position of lower layers
d1be9408 653adjusted for data read from below but not actually consumed.
b76cc8ba 654(Should perhaps C<Unread()> such data to the lower layer.)
50b80e25 655
d4165bde
SB
656Returns 0 on success, -1 on failure.
657
5cb3728c
RB
658=item Fill
659
660 IV (*Fill)(pTHX_ PerlIO *f);
d4165bde
SB
661
662The buffer for this layer should be filled (for read) from layer
663below. When you "subclass" PerlIOBuf layer, you want to use its
664I<_read> method and to supply your own fill method, which fills the
665PerlIOBuf's buffer.
50b80e25 666
d4165bde 667Returns 0 on success, -1 on failure.
50b80e25 668
5cb3728c
RB
669=item Eof
670
671 IV (*Eof)(pTHX_ PerlIO *f);
50b80e25 672
9d799145 673Return end-of-file indicator. C<PerlIOBase_eof()> is normally sufficient.
50b80e25 674
d4165bde
SB
675Returns 0 on end-of-file, 1 if not end-of-file, -1 on error.
676
5cb3728c
RB
677=item Error
678
679 IV (*Error)(pTHX_ PerlIO *f);
50b80e25 680
9d799145 681Return error indicator. C<PerlIOBase_error()> is normally sufficient.
50b80e25 682
d4165bde
SB
683Returns 1 if there is an error (usually when C<PERLIO_F_ERROR> is set,
6840 otherwise.
685
5cb3728c
RB
686=item Clearerr
687
688 void (*Clearerr)(pTHX_ PerlIO *f);
50b80e25 689
9d799145
NIS
690Clear end-of-file and error indicators. Should call C<PerlIOBase_clearerr()>
691to set the C<PERLIO_F_XXXXX> flags, which may suffice.
50b80e25 692
5cb3728c
RB
693=item Setlinebuf
694
695 void (*Setlinebuf)(pTHX_ PerlIO *f);
50b80e25 696
b76cc8ba
NIS
697Mark the stream as line buffered. C<PerlIOBase_setlinebuf()> sets the
698PERLIO_F_LINEBUF flag and is normally sufficient.
50b80e25 699
5cb3728c
RB
700=item Get_base
701
702 STDCHAR * (*Get_base)(pTHX_ PerlIO *f);
50b80e25
NIS
703
704Allocate (if not already done so) the read buffer for this layer and
d4165bde 705return pointer to it. Return NULL on failure.
50b80e25 706
5cb3728c
RB
707=item Get_bufsiz
708
709 Size_t (*Get_bufsiz)(pTHX_ PerlIO *f);
50b80e25 710
9d799145 711Return the number of bytes that last C<Fill()> put in the buffer.
50b80e25 712
5cb3728c
RB
713=item Get_ptr
714
715 STDCHAR * (*Get_ptr)(pTHX_ PerlIO *f);
50b80e25 716
3039a93d 717Return the current read pointer relative to this layer's buffer.
50b80e25 718
5cb3728c
RB
719=item Get_cnt
720
721 SSize_t (*Get_cnt)(pTHX_ PerlIO *f);
50b80e25
NIS
722
723Return the number of bytes left to be read in the current buffer.
724
5cb3728c
RB
725=item Set_ptrcnt
726
727 void (*Set_ptrcnt)(pTHX_ PerlIO *f,
728 STDCHAR *ptr, SSize_t cnt);
50b80e25
NIS
729
730Adjust the read pointer and count of bytes to match C<ptr> and/or C<cnt>.
731The application (or layer above) must ensure they are consistent.
732(Checking is allowed by the paranoid.)
733
734=back
735
b0a2be78
JH
736=head2 Utilities
737
738To ask for the next layer down use PerlIONext(PerlIO *f).
739
740To check that a PerlIO* is valid use PerlIOValid(PerlIO *f). (All
741this does is really just to check that the pointer is non-NULL and
742that the pointer behind that is non-NULL.)
743
744PerlIOBase(PerlIO *f) returns the "Base" pointer, or in other words,
745the C<PerlIOl*> pointer.
746
747PerlIOSelf(PerlIO* f, type) return the PerlIOBase cast to a type.
748
749Perl_PerlIO_or_Base(PerlIO* f, callback, base, failure, args) either
750calls the I<callback> from the functions of the layer I<f> (just by
751the name of the IO function, like "Read") with the I<args>, or if
752there is no such callback, calls the I<base> version of the callback
753with the same args, or if the f is invalid, set errno to EBADF and
754return I<failure>.
755
756Perl_PerlIO_or_fail(PerlIO* f, callback, failure, args) either calls
757the I<callback> of the functions of the layer I<f> with the I<args>,
758or if there is no such callback, set errno to EINVAL. Or if the f is
759invalid, set errno to EBADF and return I<failure>.
760
761Perl_PerlIO_or_Base_void(PerlIO* f, callback, base, args) either calls
762the I<callback> of the functions of the layer I<f> with the I<args>,
763or if there is no such callback, calls the I<base> version of the
764callback with the same args, or if the f is invalid, set errno to
765EBADF.
766
767Perl_PerlIO_or_fail_void(PerlIO* f, callback, args) either calls the
768I<callback> of the functions of the layer I<f> with the I<args>, or if
769there is no such callback, set errno to EINVAL. Or if the f is
770invalid, set errno to EBADF.
771
210e727c
JH
772=head2 Implementing PerlIO Layers
773
2535a4f7 774If you find the implementation document unclear or not sufficient,
b0a2be78 775look at the existing PerlIO layer implementations, which include:
2535a4f7
JH
776
777=over
778
779=item * C implementations
780
eae154c7
JH
781The F<perlio.c> and F<perliol.h> in the Perl core implement the
782"unix", "perlio", "stdio", "crlf", "utf8", "byte", "raw", "pending"
783layers, and also the "mmap" and "win32" layers if applicable.
784(The "win32" is currently unfinished and unused, to see what is used
785instead in Win32, see L<PerlIO/"Querying the layers of filehandles"> .)
786
2535a4f7
JH
787PerlIO::encoding, PerlIO::scalar, PerlIO::via in the Perl core.
788
789PerlIO::gzip and APR::PerlIO (mod_perl 2.0) on CPAN.
790
791=item * Perl implementations
792
793PerlIO::via::QuotedPrint in the Perl core and PerlIO::via::* on CPAN.
794
795=back
796
210e727c
JH
797If you are creating a PerlIO layer, you may want to be lazy, in other
798words, implement only the methods that interest you. The other methods
799you can either replace with the "blank" methods
800
801 PerlIOBase_noop_ok
802 PerlIOBase_noop_fail
803
804(which do nothing, and return zero and -1, respectively) or for
805certain methods you may assume a default behaviour by using a NULL
61bdadae
JH
806method. The Open method looks for help in the 'parent' layer.
807The following table summarizes the behaviour:
210e727c
JH
808
809 method behaviour with NULL
810
811 Clearerr PerlIOBase_clearerr
812 Close PerlIOBase_close
61bdadae 813 Dup PerlIOBase_dup
210e727c
JH
814 Eof PerlIOBase_eof
815 Error PerlIOBase_error
816 Fileno PerlIOBase_fileno
817 Fill FAILURE
818 Flush SUCCESS
61bdadae 819 Getarg SUCCESS
210e727c
JH
820 Get_base FAILURE
821 Get_bufsiz FAILURE
822 Get_cnt FAILURE
823 Get_ptr FAILURE
61bdadae
JH
824 Open INHERITED
825 Popped SUCCESS
826 Pushed SUCCESS
210e727c
JH
827 Read PerlIOBase_read
828 Seek FAILURE
829 Set_cnt FAILURE
830 Set_ptrcnt FAILURE
831 Setlinebuf PerlIOBase_setlinebuf
832 Tell FAILURE
833 Unread PerlIOBase_unread
834 Write FAILURE
50b80e25 835
61bdadae
JH
836 FAILURE Set errno (to EINVAL in UNIXish, to LIB$_INVARG in VMS) and
837 return -1 (for numeric return values) or NULL (for pointers)
838 INHERITED Inherited from the layer below
839 SUCCESS Return 0 (for numeric return values) or a pointer
840
50b80e25
NIS
841=head2 Core Layers
842
843The file C<perlio.c> provides the following layers:
844
845=over 4
846
847=item "unix"
848
9d799145
NIS
849A basic non-buffered layer which calls Unix/POSIX C<read()>, C<write()>,
850C<lseek()>, C<close()>. No buffering. Even on platforms that distinguish
851between O_TEXT and O_BINARY this layer is always O_BINARY.
50b80e25
NIS
852
853=item "perlio"
854
9d799145
NIS
855A very complete generic buffering layer which provides the whole of
856PerlIO API. It is also intended to be used as a "base class" for other
1d11c889
JH
857layers. (For example its C<Read()> method is implemented in terms of
858the C<Get_cnt()>/C<Get_ptr()>/C<Set_ptrcnt()> methods).
50b80e25 859
9d799145
NIS
860"perlio" over "unix" provides a complete replacement for stdio as seen
861via PerlIO API. This is the default for USE_PERLIO when system's stdio
1d11c889
JH
862does not permit perl's "fast gets" access, and which do not
863distinguish between C<O_TEXT> and C<O_BINARY>.
50b80e25
NIS
864
865=item "stdio"
866
9d799145
NIS
867A layer which provides the PerlIO API via the layer scheme, but
868implements it by calling system's stdio. This is (currently) the default
869if system's stdio provides sufficient access to allow perl's "fast gets"
870access and which do not distinguish between C<O_TEXT> and C<O_BINARY>.
50b80e25
NIS
871
872=item "crlf"
873
9d799145
NIS
874A layer derived using "perlio" as a base class. It provides Win32-like
875"\n" to CR,LF translation. Can either be applied above "perlio" or serve
876as the buffer layer itself. "crlf" over "unix" is the default if system
877distinguishes between C<O_TEXT> and C<O_BINARY> opens. (At some point
878"unix" will be replaced by a "native" Win32 IO layer on that platform,
879as Win32's read/write layer has various drawbacks.) The "crlf" layer is
880a reasonable model for a layer which transforms data in some way.
50b80e25
NIS
881
882=item "mmap"
883
9d799145
NIS
884If Configure detects C<mmap()> functions this layer is provided (with
885"perlio" as a "base") which does "read" operations by mmap()ing the
886file. Performance improvement is marginal on modern systems, so it is
887mainly there as a proof of concept. It is likely to be unbundled from
888the core at some point. The "mmap" layer is a reasonable model for a
889minimalist "derived" layer.
50b80e25
NIS
890
891=item "pending"
892
9d799145 893An "internal" derivative of "perlio" which can be used to provide
1d11c889
JH
894Unread() function for layers which have no buffer or cannot be
895bothered. (Basically this layer's C<Fill()> pops itself off the stack
896and so resumes reading from layer below.)
50b80e25
NIS
897
898=item "raw"
899
9d799145 900A dummy layer which never exists on the layer stack. Instead when
86e05cf2
NIS
901"pushed" it actually pops the stack removing itself, it then calls
902Binmode function table entry on all the layers in the stack - normally
903this (via PerlIOBase_binmode) removes any layers which do not have
904C<PERLIO_K_RAW> bit set. Layers can modify that behaviour by defining
905their own Binmode entry.
50b80e25
NIS
906
907=item "utf8"
908
9d799145 909Another dummy layer. When pushed it pops itself and sets the
1d11c889
JH
910C<PERLIO_F_UTF8> flag on the layer which was (and now is once more)
911the top of the stack.
50b80e25
NIS
912
913=back
914
9d799145
NIS
915In addition F<perlio.c> also provides a number of C<PerlIOBase_xxxx()>
916functions which are intended to be used in the table slots of classes
917which do not need to do anything special for a particular method.
50b80e25
NIS
918
919=head2 Extension Layers
920
1d11c889
JH
921Layers can made available by extension modules. When an unknown layer
922is encountered the PerlIO code will perform the equivalent of :
b76cc8ba
NIS
923
924 use PerlIO 'layer';
925
1d11c889 926Where I<layer> is the unknown layer. F<PerlIO.pm> will then attempt to:
b76cc8ba
NIS
927
928 require PerlIO::layer;
929
1d11c889
JH
930If after that process the layer is still not defined then the C<open>
931will fail.
b76cc8ba
NIS
932
933The following extension layers are bundled with perl:
50b80e25
NIS
934
935=over 4
936
b76cc8ba 937=item ":encoding"
50b80e25
NIS
938
939 use Encoding;
940
1d11c889
JH
941makes this layer available, although F<PerlIO.pm> "knows" where to
942find it. It is an example of a layer which takes an argument as it is
943called thus:
50b80e25 944
b31b80f9 945 open( $fh, "<:encoding(iso-8859-7)", $pathname );
50b80e25 946
385e1f9f 947=item ":scalar"
b76cc8ba 948
b31b80f9 949Provides support for reading data from and writing data to a scalar.
b76cc8ba 950
385e1f9f 951 open( $fh, "+<:scalar", \$scalar );
50b80e25 952
1d11c889
JH
953When a handle is so opened, then reads get bytes from the string value
954of I<$scalar>, and writes change the value. In both cases the position
955in I<$scalar> starts as zero but can be altered via C<seek>, and
956determined via C<tell>.
b76cc8ba 957
385e1f9f
EM
958Please note that this layer is implied when calling open() thus:
959
960 open( $fh, "+<", \$scalar );
961
962=item ":via"
b76cc8ba 963
4f7853f4
GA
964Provided to allow layers to be implemented as Perl code. For instance:
965
e934609f 966 use PerlIO::via::StripHTML;
385e1f9f 967 open( my $fh, "<:via(StripHTML)", "index.html" );
4f7853f4 968
e934609f 969See L<PerlIO::via> for details.
b76cc8ba
NIS
970
971=back
50b80e25 972
d4165bde
SB
973=head1 TODO
974
975Things that need to be done to improve this document.
976
977=over
978
979=item *
980
981Explain how to make a valid fh without going through open()(i.e. apply
982a layer). For example if the file is not opened through perl, but we
983want to get back a fh, like it was opened by Perl.
984
985How PerlIO_apply_layera fits in, where its docs, was it made public?
986
987Currently the example could be something like this:
988
989 PerlIO *foo_to_PerlIO(pTHX_ char *mode, ...)
990 {
991 char *mode; /* "w", "r", etc */
992 const char *layers = ":APR"; /* the layer name */
993 PerlIO *f = PerlIO_allocate(aTHX);
994 if (!f) {
995 return NULL;
996 }
997
998 PerlIO_apply_layers(aTHX_ f, mode, layers);
999
1000 if (f) {
1001 PerlIOAPR *st = PerlIOSelf(f, PerlIOAPR);
1002 /* fill in the st struct, as in _open() */
1003 st->file = file;
1004 PerlIOBase(f)->flags |= PERLIO_F_OPEN;
1005
1006 return f;
1007 }
1008 return NULL;
1009 }
1010
1011=item *
1012
1013fix/add the documentation in places marked as XXX.
1014
1015=item *
1016
1017The handling of errors by the layer is not specified. e.g. when $!
1018should be set explicitly, when the error handling should be just
1019delegated to the top layer.
1020
1021Probably give some hints on using SETERRNO() or pointers to where they
1022can be found.
1023
1024=item *
1025
1026I think it would help to give some concrete examples to make it easier
1027to understand the API. Of course I agree that the API has to be
1028concise, but since there is no second document that is more of a
1029guide, I think that it'd make it easier to start with the doc which is
1030an API, but has examples in it in places where things are unclear, to
1031a person who is not a PerlIO guru (yet).
1032
1033=back
1034
50b80e25 1035=cut