sin/samba-mirror
1
0
Fork 0
mirror of https://github.com/samba-team/samba.git synced 2024-12-22 13:34:15 +03:00
Code
9312bdd271
samba-mirror/pidl/idl.yp

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

697 lines
12 KiB
Plaintext
Raw Normal View History

re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
########################
# IDL Parse::Yapp parser
# Copyright (C) Andrew Tridgell <tridge@samba.org>
Convert some more files to GPLv3. (This used to be commit ebe5e8399422eb7e2ff4deb546338823e2718907)
2008-03-28 09:08:49 +03:00
# released under the GNU GPL version 3 or later
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
r8878: Spelling fixes. (-: (This used to be commit f850faca9797c9818a46fec72dba069cac5a6db7)
2005-08-01 05:39:24 +04:00
# the precedence actually doesn't matter at all for this grammar, but
* reduced the number of grammer conflicts a lot using (arbitrary) precedence rules * build a standalone parser. When we come to distributing Samba4 we can just include idl.pm and other developers won't need Parse::Yapp installed * avoid the recursive make in most cases in build_idl.sh (This used to be commit be2c2be459d1bed41c113590e70711cb89ad12b9)
2003-12-08 05:24:40 +03:00
# by providing a precedence we reduce the number of conflicts
# enormously
%left '-' '+' '&' '|' '*' '>' '.' '/' '(' ')' '[' ',' ';'
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
################
r8878: Spelling fixes. (-: (This used to be commit f850faca9797c9818a46fec72dba069cac5a6db7)
2005-08-01 05:39:24 +04:00
# grammar
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
%%
r1736: - Pidl updates: - Support for "object oriented" interfaces in pidl - Support for inherited interfaces in pidl - Simplification of the support for properties on an interface - Start on dcom rpc torture tests (This used to be commit 45c3d0036b8510102816f9cdff9210098259cc5f)
2004-08-11 23:48:36 +04:00
idl:
#empty { {} }
r4533: parsing support for: typedef [bitmap16bit] bitmap { FLAG1 = 0x0001, FLAG2 = 0x8000 } fooflags; metze (This used to be commit 192f2495468d5d0c5889d94ca791aa8f371514e8)
2005-01-05 18:26:48 +03:00
|
idl interface { push(@{$_[1]}, $_[2]); $_[1] }
r3440: Add support for "coclass" to the pidl parser - the idl keyword that describes a COM class. A coclass is the implementation of one or more interfaces. It has a UUID referred to as it's CLSID (Class ID). Also adding an example coclass called "CoffeeMachine". You can give it a string (or a cup, whatever you like ;-) and it will fill it with "COFFEE" (kind of the like the echo pipe is for regular RPC). CoffeeMachine's Windows implementation already works, a torture test for Samba will follow soon. (This used to be commit 05dd840b6ffba0d38d98e2e80d856e2f2b5d877c)
2004-11-01 23:00:02 +03:00
|
idl coclass { push(@{$_[1]}, $_[2]); $_[1] }
r19585: Add support for some more standard IDL instructions: - `include' (replaces helper()) - `import' (replaces depends()) Add support for parsing importlib() - importlib() is now ignored (with a warning), but no longer causes syntax errors. helper() and depends() are now marked deprecated and will cause warnings. (This used to be commit 1ccab71cb8a9e3db9448b6679d01ad00e1c1e9a3)
2006-11-07 00:54:19 +03:00
|
idl import { push(@{$_[1]}, $_[2]); $_[1] }
|
idl include { push(@{$_[1]}, $_[2]); $_[1] }
|
idl importlib { push(@{$_[1]}, $_[2]); $_[1] }
r24815: Support cpp_quote(). (This used to be commit 30c1de30bb4ded16e79316c0ab43809e0e19f75d)
2007-08-31 04:03:54 +04:00
|
idl cpp_quote { push(@{$_[1]}, $_[2]); $_[1] }
r19585: Add support for some more standard IDL instructions: - `include' (replaces helper()) - `import' (replaces depends()) Add support for parsing importlib() - importlib() is now ignored (with a warning), but no longer causes syntax errors. helper() and depends() are now marked deprecated and will cause warnings. (This used to be commit 1ccab71cb8a9e3db9448b6679d01ad00e1c1e9a3)
2006-11-07 00:54:19 +03:00
;
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
r19588: Use include and import statements rather than depends() and helper(). (This used to be commit 347ae9628202ca4de4318ef8156999239aad9192)
2006-11-07 01:54:49 +03:00
import:
'import' commalist ';'
{{
r19585: Add support for some more standard IDL instructions: - `include' (replaces helper()) - `import' (replaces depends()) Add support for parsing importlib() - importlib() is now ignored (with a warning), but no longer causes syntax errors. helper() and depends() are now marked deprecated and will cause warnings. (This used to be commit 1ccab71cb8a9e3db9448b6679d01ad00e1c1e9a3)
2006-11-07 00:54:19 +03:00
"TYPE" => "IMPORT",
r19588: Use include and import statements rather than depends() and helper(). (This used to be commit 347ae9628202ca4de4318ef8156999239aad9192)
2006-11-07 01:54:49 +03:00
"PATHS" => $_[2],
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
"FILE" => $_[0]->YYData->{FILE},
r19585: Add support for some more standard IDL instructions: - `include' (replaces helper()) - `import' (replaces depends()) Add support for parsing importlib() - importlib() is now ignored (with a warning), but no longer causes syntax errors. helper() and depends() are now marked deprecated and will cause warnings. (This used to be commit 1ccab71cb8a9e3db9448b6679d01ad00e1c1e9a3)
2006-11-07 00:54:19 +03:00
"LINE" => $_[0]->YYData->{LINE},
}}
;
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
r19588: Use include and import statements rather than depends() and helper(). (This used to be commit 347ae9628202ca4de4318ef8156999239aad9192)
2006-11-07 01:54:49 +03:00
include:
'include' commalist ';'
{{
r19585: Add support for some more standard IDL instructions: - `include' (replaces helper()) - `import' (replaces depends()) Add support for parsing importlib() - importlib() is now ignored (with a warning), but no longer causes syntax errors. helper() and depends() are now marked deprecated and will cause warnings. (This used to be commit 1ccab71cb8a9e3db9448b6679d01ad00e1c1e9a3)
2006-11-07 00:54:19 +03:00
"TYPE" => "INCLUDE",
r19588: Use include and import statements rather than depends() and helper(). (This used to be commit 347ae9628202ca4de4318ef8156999239aad9192)
2006-11-07 01:54:49 +03:00
"PATHS" => $_[2],
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
"FILE" => $_[0]->YYData->{FILE},
r19585: Add support for some more standard IDL instructions: - `include' (replaces helper()) - `import' (replaces depends()) Add support for parsing importlib() - importlib() is now ignored (with a warning), but no longer causes syntax errors. helper() and depends() are now marked deprecated and will cause warnings. (This used to be commit 1ccab71cb8a9e3db9448b6679d01ad00e1c1e9a3)
2006-11-07 00:54:19 +03:00
"LINE" => $_[0]->YYData->{LINE},
}}
;
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
r19588: Use include and import statements rather than depends() and helper(). (This used to be commit 347ae9628202ca4de4318ef8156999239aad9192)
2006-11-07 01:54:49 +03:00
importlib:
'importlib' commalist ';'
{{
r19585: Add support for some more standard IDL instructions: - `include' (replaces helper()) - `import' (replaces depends()) Add support for parsing importlib() - importlib() is now ignored (with a warning), but no longer causes syntax errors. helper() and depends() are now marked deprecated and will cause warnings. (This used to be commit 1ccab71cb8a9e3db9448b6679d01ad00e1c1e9a3)
2006-11-07 00:54:19 +03:00
"TYPE" => "IMPORTLIB",
r19588: Use include and import statements rather than depends() and helper(). (This used to be commit 347ae9628202ca4de4318ef8156999239aad9192)
2006-11-07 01:54:49 +03:00
"PATHS" => $_[2],
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
"FILE" => $_[0]->YYData->{FILE},
r19585: Add support for some more standard IDL instructions: - `include' (replaces helper()) - `import' (replaces depends()) Add support for parsing importlib() - importlib() is now ignored (with a warning), but no longer causes syntax errors. helper() and depends() are now marked deprecated and will cause warnings. (This used to be commit 1ccab71cb8a9e3db9448b6679d01ad00e1c1e9a3)
2006-11-07 00:54:19 +03:00
"LINE" => $_[0]->YYData->{LINE},
}}
r3440: Add support for "coclass" to the pidl parser - the idl keyword that describes a COM class. A coclass is the implementation of one or more interfaces. It has a UUID referred to as it's CLSID (Class ID). Also adding an example coclass called "CoffeeMachine". You can give it a string (or a cup, whatever you like ;-) and it will fill it with "COFFEE" (kind of the like the echo pipe is for regular RPC). CoffeeMachine's Windows implementation already works, a torture test for Samba will follow soon. (This used to be commit 05dd840b6ffba0d38d98e2e80d856e2f2b5d877c)
2004-11-01 23:00:02 +03:00
;
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
r19588: Use include and import statements rather than depends() and helper(). (This used to be commit 347ae9628202ca4de4318ef8156999239aad9192)
2006-11-07 01:54:49 +03:00
commalist:
text { [ $_[1] ] }
|
commalist ',' text { push(@{$_[1]}, $_[3]); $_[1] }
;
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
r5486: - Use references to interfaces in coclasses rather then full interfaces. - Couple of bugfixes in odl.pm - First use of multiple pointers in dcom.idl (This used to be commit 3110c2ec44a872a53e7d6015b85c68fdb604b4de)
2005-02-21 15:13:42 +03:00
coclass:
property_list 'coclass' identifier '{' interface_names '}' optional_semicolon
r19585: Add support for some more standard IDL instructions: - `include' (replaces helper()) - `import' (replaces depends()) Add support for parsing importlib() - importlib() is now ignored (with a warning), but no longer causes syntax errors. helper() and depends() are now marked deprecated and will cause warnings. (This used to be commit 1ccab71cb8a9e3db9448b6679d01ad00e1c1e9a3)
2006-11-07 00:54:19 +03:00
{{
r3440: Add support for "coclass" to the pidl parser - the idl keyword that describes a COM class. A coclass is the implementation of one or more interfaces. It has a UUID referred to as it's CLSID (Class ID). Also adding an example coclass called "CoffeeMachine". You can give it a string (or a cup, whatever you like ;-) and it will fill it with "COFFEE" (kind of the like the echo pipe is for regular RPC). CoffeeMachine's Windows implementation already works, a torture test for Samba will follow soon. (This used to be commit 05dd840b6ffba0d38d98e2e80d856e2f2b5d877c)
2004-11-01 23:00:02 +03:00
"TYPE" => "COCLASS",
r4522: PROPERTIES are now handled at the typedef level (This used to be commit 73d5a033e8e9de536239c53e8a063e9be0eff701)
2005-01-05 03:14:29 +03:00
"PROPERTIES" => $_[1],
"NAME" => $_[3],
"DATA" => $_[5],
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
"FILE" => $_[0]->YYData->{FILE},
r6973: Merge new version of pidl into the main SAMBA_4_0 branch. The main difference in this new version is the extra data structure generated between the IDL data structure and the NDR parser: IDL -> NDR -> { ndr_parser, ndr_header, eparser, etc } This makes the ndr_parser.pm internals much more sane. Other changes include: - Remove unnecessary calls with NDR_BUFFERS (for example, GUID doesn't have any buffers, just scalars) as well as some (unnecessary) nested setting of flags. - Parse array loops in the C code rather then calling ndr_pull_array(). This allows us to have, for example, arrays of pointers or arrays of pointers to arrays, etc.. - Use if() {} rather then if () goto foo; everywhere - NDR_IN no longer implies LIBNDR_FLAG_REF_ALLOC - By default, top level pointers are now "ref" (as is the default in most other IDL compilers). This can be overridden using the default_pointer_top() property. - initial work on new ethereal parser generators by Alan DeKok and me - pidl now writes errors in the standard format used by compilers, which is parsable by most editors - ability to warn about the fact that pidl extension(s) have been used, useful for making sure IDL files work with other IDL compilers. oh, and there's probably some other things I can't think of right now.. (This used to be commit 13cf227615f6b9e0e5fa62e59197024410254f01)
2005-05-25 17:50:27 +04:00
"LINE" => $_[0]->YYData->{LINE},
r3440: Add support for "coclass" to the pidl parser - the idl keyword that describes a COM class. A coclass is the implementation of one or more interfaces. It has a UUID referred to as it's CLSID (Class ID). Also adding an example coclass called "CoffeeMachine". You can give it a string (or a cup, whatever you like ;-) and it will fill it with "COFFEE" (kind of the like the echo pipe is for regular RPC). CoffeeMachine's Windows implementation already works, a torture test for Samba will follow soon. (This used to be commit 05dd840b6ffba0d38d98e2e80d856e2f2b5d877c)
2004-11-01 23:00:02 +03:00
}}
;
r5486: - Use references to interfaces in coclasses rather then full interfaces. - Couple of bugfixes in odl.pm - First use of multiple pointers in dcom.idl (This used to be commit 3110c2ec44a872a53e7d6015b85c68fdb604b4de)
2005-02-21 15:13:42 +03:00
interface_names:
r3440: Add support for "coclass" to the pidl parser - the idl keyword that describes a COM class. A coclass is the implementation of one or more interfaces. It has a UUID referred to as it's CLSID (Class ID). Also adding an example coclass called "CoffeeMachine". You can give it a string (or a cup, whatever you like ;-) and it will fill it with "COFFEE" (kind of the like the echo pipe is for regular RPC). CoffeeMachine's Windows implementation already works, a torture test for Samba will follow soon. (This used to be commit 05dd840b6ffba0d38d98e2e80d856e2f2b5d877c)
2004-11-01 23:00:02 +03:00
#empty { {} }
r5486: - Use references to interfaces in coclasses rather then full interfaces. - Couple of bugfixes in odl.pm - First use of multiple pointers in dcom.idl (This used to be commit 3110c2ec44a872a53e7d6015b85c68fdb604b4de)
2005-02-21 15:13:42 +03:00
|
interface_names 'interface' identifier ';' { push(@{$_[1]}, $_[2]); $_[1] }
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
;
Reimport pidl support for COM.
2008-05-07 19:18:18 +04:00
interface:
property_list 'interface' identifier base_interface '{' definitions '}' optional_semicolon
r19585: Add support for some more standard IDL instructions: - `include' (replaces helper()) - `import' (replaces depends()) Add support for parsing importlib() - importlib() is now ignored (with a warning), but no longer causes syntax errors. helper() and depends() are now marked deprecated and will cause warnings. (This used to be commit 1ccab71cb8a9e3db9448b6679d01ad00e1c1e9a3)
2006-11-07 00:54:19 +03:00
{{
r1736: - Pidl updates: - Support for "object oriented" interfaces in pidl - Support for inherited interfaces in pidl - Simplification of the support for properties on an interface - Start on dcom rpc torture tests (This used to be commit 45c3d0036b8510102816f9cdff9210098259cc5f)
2004-08-11 23:48:36 +04:00
"TYPE" => "INTERFACE",
r4522: PROPERTIES are now handled at the typedef level (This used to be commit 73d5a033e8e9de536239c53e8a063e9be0eff701)
2005-01-05 03:14:29 +03:00
"PROPERTIES" => $_[1],
"NAME" => $_[3],
Reimport pidl support for COM.
2008-05-07 19:18:18 +04:00
"BASE" => $_[4],
"DATA" => $_[6],
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
"FILE" => $_[0]->YYData->{FILE},
r6973: Merge new version of pidl into the main SAMBA_4_0 branch. The main difference in this new version is the extra data structure generated between the IDL data structure and the NDR parser: IDL -> NDR -> { ndr_parser, ndr_header, eparser, etc } This makes the ndr_parser.pm internals much more sane. Other changes include: - Remove unnecessary calls with NDR_BUFFERS (for example, GUID doesn't have any buffers, just scalars) as well as some (unnecessary) nested setting of flags. - Parse array loops in the C code rather then calling ndr_pull_array(). This allows us to have, for example, arrays of pointers or arrays of pointers to arrays, etc.. - Use if() {} rather then if () goto foo; everywhere - NDR_IN no longer implies LIBNDR_FLAG_REF_ALLOC - By default, top level pointers are now "ref" (as is the default in most other IDL compilers). This can be overridden using the default_pointer_top() property. - initial work on new ethereal parser generators by Alan DeKok and me - pidl now writes errors in the standard format used by compilers, which is parsable by most editors - ability to warn about the fact that pidl extension(s) have been used, useful for making sure IDL files work with other IDL compilers. oh, and there's probably some other things I can't think of right now.. (This used to be commit 13cf227615f6b9e0e5fa62e59197024410254f01)
2005-05-25 17:50:27 +04:00
"LINE" => $_[0]->YYData->{LINE},
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
}}
;
Reimport pidl support for COM.
2008-05-07 19:18:18 +04:00
base_interface:
#empty
|
':' identifier { $_[2] }
;
r24815: Support cpp_quote(). (This used to be commit 30c1de30bb4ded16e79316c0ab43809e0e19f75d)
2007-08-31 04:03:54 +04:00
cpp_quote:
'cpp_quote' '(' text ')'
{{
"TYPE" => "CPP_QUOTE",
r4518: added proper support for "typedef enum" in pidl. We can now use enums as types in switch statements and variable types, and the value will be printed with its string name and value (This used to be commit 1852033f2ec7584657d586e44d25a280959e4d81)
2005-01-05 02:25:25 +03:00
"DATA" => $_[3],
r24815: Support cpp_quote(). (This used to be commit 30c1de30bb4ded16e79316c0ab43809e0e19f75d)
2007-08-31 04:03:54 +04:00
"FILE" => $_[0]->YYData->{FILE},
"LINE" => $_[0]->YYData->{LINE},
}}
;
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
definitions:
definition { [ $_[1] ] }
|
definitions definition { push(@{$_[1]}, $_[2]); $_[1] }
;
r11097: Allow pointers in "const" (This used to be commit 88fcd56fb55ac3f7e1eef3b3c92a2593c79599d3)
2005-10-16 21:17:42 +04:00
definition:
function
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
|
r11097: Allow pointers in "const" (This used to be commit 88fcd56fb55ac3f7e1eef3b3c92a2593c79599d3)
2005-10-16 21:17:42 +04:00
const
|
r3998: allow const arrays (see next commit to drsuapi.idl) metze (This used to be commit 7264a2227d7a1881db01964d0d713c144eb77152)
2004-11-29 14:08:15 +03:00
typedef
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
|
typedecl
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
;
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
const:
r11097: Allow pointers in "const" (This used to be commit 88fcd56fb55ac3f7e1eef3b3c92a2593c79599d3)
2005-10-16 21:17:42 +04:00
'const' identifier pointers identifier '=' anytext ';'
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
{{
r3998: allow const arrays (see next commit to drsuapi.idl) metze (This used to be commit 7264a2227d7a1881db01964d0d713c144eb77152)
2004-11-29 14:08:15 +03:00
"TYPE" => "CONST",
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
"DTYPE" => $_[2],
r11097: Allow pointers in "const" (This used to be commit 88fcd56fb55ac3f7e1eef3b3c92a2593c79599d3)
2005-10-16 21:17:42 +04:00
"POINTERS" => $_[3],
r4518: added proper support for "typedef enum" in pidl. We can now use enums as types in switch statements and variable types, and the value will be printed with its string name and value (This used to be commit 1852033f2ec7584657d586e44d25a280959e4d81)
2005-01-05 02:25:25 +03:00
"NAME" => $_[4],
r11097: Allow pointers in "const" (This used to be commit 88fcd56fb55ac3f7e1eef3b3c92a2593c79599d3)
2005-10-16 21:17:42 +04:00
"VALUE" => $_[6],
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
"FILE" => $_[0]->YYData->{FILE},
r6973: Merge new version of pidl into the main SAMBA_4_0 branch. The main difference in this new version is the extra data structure generated between the IDL data structure and the NDR parser: IDL -> NDR -> { ndr_parser, ndr_header, eparser, etc } This makes the ndr_parser.pm internals much more sane. Other changes include: - Remove unnecessary calls with NDR_BUFFERS (for example, GUID doesn't have any buffers, just scalars) as well as some (unnecessary) nested setting of flags. - Parse array loops in the C code rather then calling ndr_pull_array(). This allows us to have, for example, arrays of pointers or arrays of pointers to arrays, etc.. - Use if() {} rather then if () goto foo; everywhere - NDR_IN no longer implies LIBNDR_FLAG_REF_ALLOC - By default, top level pointers are now "ref" (as is the default in most other IDL compilers). This can be overridden using the default_pointer_top() property. - initial work on new ethereal parser generators by Alan DeKok and me - pidl now writes errors in the standard format used by compilers, which is parsable by most editors - ability to warn about the fact that pidl extension(s) have been used, useful for making sure IDL files work with other IDL compilers. oh, and there's probably some other things I can't think of right now.. (This used to be commit 13cf227615f6b9e0e5fa62e59197024410254f01)
2005-05-25 17:50:27 +04:00
"LINE" => $_[0]->YYData->{LINE},
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
}}
|
r11097: Allow pointers in "const" (This used to be commit 88fcd56fb55ac3f7e1eef3b3c92a2593c79599d3)
2005-10-16 21:17:42 +04:00
'const' identifier pointers identifier array_len '=' anytext ';'
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
{{
r3998: allow const arrays (see next commit to drsuapi.idl) metze (This used to be commit 7264a2227d7a1881db01964d0d713c144eb77152)
2004-11-29 14:08:15 +03:00
"TYPE" => "CONST",
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
"DTYPE" => $_[2],
r11097: Allow pointers in "const" (This used to be commit 88fcd56fb55ac3f7e1eef3b3c92a2593c79599d3)
2005-10-16 21:17:42 +04:00
"POINTERS" => $_[3],
r4518: added proper support for "typedef enum" in pidl. We can now use enums as types in switch statements and variable types, and the value will be printed with its string name and value (This used to be commit 1852033f2ec7584657d586e44d25a280959e4d81)
2005-01-05 02:25:25 +03:00
"NAME" => $_[4],
r6973: Merge new version of pidl into the main SAMBA_4_0 branch. The main difference in this new version is the extra data structure generated between the IDL data structure and the NDR parser: IDL -> NDR -> { ndr_parser, ndr_header, eparser, etc } This makes the ndr_parser.pm internals much more sane. Other changes include: - Remove unnecessary calls with NDR_BUFFERS (for example, GUID doesn't have any buffers, just scalars) as well as some (unnecessary) nested setting of flags. - Parse array loops in the C code rather then calling ndr_pull_array(). This allows us to have, for example, arrays of pointers or arrays of pointers to arrays, etc.. - Use if() {} rather then if () goto foo; everywhere - NDR_IN no longer implies LIBNDR_FLAG_REF_ALLOC - By default, top level pointers are now "ref" (as is the default in most other IDL compilers). This can be overridden using the default_pointer_top() property. - initial work on new ethereal parser generators by Alan DeKok and me - pidl now writes errors in the standard format used by compilers, which is parsable by most editors - ability to warn about the fact that pidl extension(s) have been used, useful for making sure IDL files work with other IDL compilers. oh, and there's probably some other things I can't think of right now.. (This used to be commit 13cf227615f6b9e0e5fa62e59197024410254f01)
2005-05-25 17:50:27 +04:00
"ARRAY_LEN" => $_[5],
r11097: Allow pointers in "const" (This used to be commit 88fcd56fb55ac3f7e1eef3b3c92a2593c79599d3)
2005-10-16 21:17:42 +04:00
"VALUE" => $_[7],
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
"FILE" => $_[0]->YYData->{FILE},
r6973: Merge new version of pidl into the main SAMBA_4_0 branch. The main difference in this new version is the extra data structure generated between the IDL data structure and the NDR parser: IDL -> NDR -> { ndr_parser, ndr_header, eparser, etc } This makes the ndr_parser.pm internals much more sane. Other changes include: - Remove unnecessary calls with NDR_BUFFERS (for example, GUID doesn't have any buffers, just scalars) as well as some (unnecessary) nested setting of flags. - Parse array loops in the C code rather then calling ndr_pull_array(). This allows us to have, for example, arrays of pointers or arrays of pointers to arrays, etc.. - Use if() {} rather then if () goto foo; everywhere - NDR_IN no longer implies LIBNDR_FLAG_REF_ALLOC - By default, top level pointers are now "ref" (as is the default in most other IDL compilers). This can be overridden using the default_pointer_top() property. - initial work on new ethereal parser generators by Alan DeKok and me - pidl now writes errors in the standard format used by compilers, which is parsable by most editors - ability to warn about the fact that pidl extension(s) have been used, useful for making sure IDL files work with other IDL compilers. oh, and there's probably some other things I can't think of right now.. (This used to be commit 13cf227615f6b9e0e5fa62e59197024410254f01)
2005-05-25 17:50:27 +04:00
"LINE" => $_[0]->YYData->{LINE},
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
}}
;
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
function:
property_list type identifier '(' element_list2 ')' ';'
{{
"TYPE" => "FUNCTION",
"NAME" => $_[3],
"RETURN_TYPE" => $_[2],
"PROPERTIES" => $_[1],
r6973: Merge new version of pidl into the main SAMBA_4_0 branch. The main difference in this new version is the extra data structure generated between the IDL data structure and the NDR parser: IDL -> NDR -> { ndr_parser, ndr_header, eparser, etc } This makes the ndr_parser.pm internals much more sane. Other changes include: - Remove unnecessary calls with NDR_BUFFERS (for example, GUID doesn't have any buffers, just scalars) as well as some (unnecessary) nested setting of flags. - Parse array loops in the C code rather then calling ndr_pull_array(). This allows us to have, for example, arrays of pointers or arrays of pointers to arrays, etc.. - Use if() {} rather then if () goto foo; everywhere - NDR_IN no longer implies LIBNDR_FLAG_REF_ALLOC - By default, top level pointers are now "ref" (as is the default in most other IDL compilers). This can be overridden using the default_pointer_top() property. - initial work on new ethereal parser generators by Alan DeKok and me - pidl now writes errors in the standard format used by compilers, which is parsable by most editors - ability to warn about the fact that pidl extension(s) have been used, useful for making sure IDL files work with other IDL compilers. oh, and there's probably some other things I can't think of right now.. (This used to be commit 13cf227615f6b9e0e5fa62e59197024410254f01)
2005-05-25 17:50:27 +04:00
"ELEMENTS" => $_[5],
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
"FILE" => $_[0]->YYData->{FILE},
r6973: Merge new version of pidl into the main SAMBA_4_0 branch. The main difference in this new version is the extra data structure generated between the IDL data structure and the NDR parser: IDL -> NDR -> { ndr_parser, ndr_header, eparser, etc } This makes the ndr_parser.pm internals much more sane. Other changes include: - Remove unnecessary calls with NDR_BUFFERS (for example, GUID doesn't have any buffers, just scalars) as well as some (unnecessary) nested setting of flags. - Parse array loops in the C code rather then calling ndr_pull_array(). This allows us to have, for example, arrays of pointers or arrays of pointers to arrays, etc.. - Use if() {} rather then if () goto foo; everywhere - NDR_IN no longer implies LIBNDR_FLAG_REF_ALLOC - By default, top level pointers are now "ref" (as is the default in most other IDL compilers). This can be overridden using the default_pointer_top() property. - initial work on new ethereal parser generators by Alan DeKok and me - pidl now writes errors in the standard format used by compilers, which is parsable by most editors - ability to warn about the fact that pidl extension(s) have been used, useful for making sure IDL files work with other IDL compilers. oh, and there's probably some other things I can't think of right now.. (This used to be commit 13cf227615f6b9e0e5fa62e59197024410254f01)
2005-05-25 17:50:27 +04:00
"LINE" => $_[0]->YYData->{LINE},
}}
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
;
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
r21430: Support tagged types without typedef. This means: struct foo { ... }; in IDL will now work. This is the first step towards nested types and using typedefs for partial types (such as "typedef int *bar;"), a requirement for complex uses of represent_as(). (This used to be commit a716aa70f0c90898e6fcf57d63a2cf4c40e7d4df)
2007-02-18 19:21:28 +03:00
typedef:
pidl: add support for pointers in typedefs metze
2009-08-05 15:43:49 +04:00
property_list 'typedef' type pointers identifier array_len ';'
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
{{
r4518: added proper support for "typedef enum" in pidl. We can now use enums as types in switch statements and variable types, and the value will be printed with its string name and value (This used to be commit 1852033f2ec7584657d586e44d25a280959e4d81)
2005-01-05 02:25:25 +03:00
"TYPE" => "TYPEDEF",
r21430: Support tagged types without typedef. This means: struct foo { ... }; in IDL will now work. This is the first step towards nested types and using typedefs for partial types (such as "typedef int *bar;"), a requirement for complex uses of represent_as(). (This used to be commit a716aa70f0c90898e6fcf57d63a2cf4c40e7d4df)
2007-02-18 19:21:28 +03:00
"PROPERTIES" => $_[1],
pidl: add support for pointers in typedefs metze
2009-08-05 15:43:49 +04:00
"NAME" => $_[5],
r4518: added proper support for "typedef enum" in pidl. We can now use enums as types in switch statements and variable types, and the value will be printed with its string name and value (This used to be commit 1852033f2ec7584657d586e44d25a280959e4d81)
2005-01-05 02:25:25 +03:00
"DATA" => $_[3],
pidl: add support for pointers in typedefs metze
2009-08-05 15:43:49 +04:00
"POINTERS" => $_[4],
"ARRAY_LEN" => $_[6],
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
"FILE" => $_[0]->YYData->{FILE},
r6973: Merge new version of pidl into the main SAMBA_4_0 branch. The main difference in this new version is the extra data structure generated between the IDL data structure and the NDR parser: IDL -> NDR -> { ndr_parser, ndr_header, eparser, etc } This makes the ndr_parser.pm internals much more sane. Other changes include: - Remove unnecessary calls with NDR_BUFFERS (for example, GUID doesn't have any buffers, just scalars) as well as some (unnecessary) nested setting of flags. - Parse array loops in the C code rather then calling ndr_pull_array(). This allows us to have, for example, arrays of pointers or arrays of pointers to arrays, etc.. - Use if() {} rather then if () goto foo; everywhere - NDR_IN no longer implies LIBNDR_FLAG_REF_ALLOC - By default, top level pointers are now "ref" (as is the default in most other IDL compilers). This can be overridden using the default_pointer_top() property. - initial work on new ethereal parser generators by Alan DeKok and me - pidl now writes errors in the standard format used by compilers, which is parsable by most editors - ability to warn about the fact that pidl extension(s) have been used, useful for making sure IDL files work with other IDL compilers. oh, and there's probably some other things I can't think of right now.. (This used to be commit 13cf227615f6b9e0e5fa62e59197024410254f01)
2005-05-25 17:50:27 +04:00
"LINE" => $_[0]->YYData->{LINE},
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
}}
;
r8622: Support parsing new structs/unions/enums/bitmaps without typedef in (This used to be commit cd33b5efc544345411db498b178fa532a9a2c82f)
2005-07-20 03:34:02 +04:00
usertype:
struct
|
union
|
enum
|
bitmap
pidl: parse idl 'pipe' typedefs, but print out a not supported message for now metze
2009-08-05 15:40:38 +04:00
|
pipe
r8622: Support parsing new structs/unions/enums/bitmaps without typedef in (This used to be commit cd33b5efc544345411db498b178fa532a9a2c82f)
2005-07-20 03:34:02 +04:00
;
typedecl:
usertype ';' { $_[1] }
;
r12481: Move parser-specific utility functions to idl.yp, remove some unused functions Allow the use of non-typedef structs and unions when declaring variables. Allow the use of the 'signed' and 'unsigned' qualifiers for integer types (This used to be commit bc6b45e242c8d7b2ef1a6e6d3eb172c27afd952d)
2005-12-25 17:59:21 +03:00
sign:
'signed'
|
'unsigned'
;
existingtype:
r20625: Fix couple of warnings. (This used to be commit 203076129b967ccc6258e807280dc1b75583a064)
2007-01-09 09:02:41 +03:00
sign identifier { ($_[1]?$_[1]:"signed") ." $_[2]" }
r12481: Move parser-specific utility functions to idl.yp, remove some unused functions Allow the use of non-typedef structs and unions when declaring variables. Allow the use of the 'signed' and 'unsigned' qualifiers for integer types (This used to be commit bc6b45e242c8d7b2ef1a6e6d3eb172c27afd952d)
2005-12-25 17:59:21 +03:00
|
identifier
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
;
r12481: Move parser-specific utility functions to idl.yp, remove some unused functions Allow the use of non-typedef structs and unions when declaring variables. Allow the use of the 'signed' and 'unsigned' qualifiers for integer types (This used to be commit bc6b45e242c8d7b2ef1a6e6d3eb172c27afd952d)
2005-12-25 17:59:21 +03:00
type:
usertype
|
existingtype
|
void { "void" }
;
enum_body:
'{' enum_elements '}' { $_[2] }
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
;
opt_enum_body:
#empty
|
enum_body
;
enum:
property_list 'enum' optional_identifier opt_enum_body
{{
"TYPE" => "ENUM",
"PROPERTIES" => $_[1],
"NAME" => $_[3],
"ELEMENTS" => $_[4],
"FILE" => $_[0]->YYData->{FILE},
r6973: Merge new version of pidl into the main SAMBA_4_0 branch. The main difference in this new version is the extra data structure generated between the IDL data structure and the NDR parser: IDL -> NDR -> { ndr_parser, ndr_header, eparser, etc } This makes the ndr_parser.pm internals much more sane. Other changes include: - Remove unnecessary calls with NDR_BUFFERS (for example, GUID doesn't have any buffers, just scalars) as well as some (unnecessary) nested setting of flags. - Parse array loops in the C code rather then calling ndr_pull_array(). This allows us to have, for example, arrays of pointers or arrays of pointers to arrays, etc.. - Use if() {} rather then if () goto foo; everywhere - NDR_IN no longer implies LIBNDR_FLAG_REF_ALLOC - By default, top level pointers are now "ref" (as is the default in most other IDL compilers). This can be overridden using the default_pointer_top() property. - initial work on new ethereal parser generators by Alan DeKok and me - pidl now writes errors in the standard format used by compilers, which is parsable by most editors - ability to warn about the fact that pidl extension(s) have been used, useful for making sure IDL files work with other IDL compilers. oh, and there's probably some other things I can't think of right now.. (This used to be commit 13cf227615f6b9e0e5fa62e59197024410254f01)
2005-05-25 17:50:27 +04:00
"LINE" => $_[0]->YYData->{LINE},
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
}}
;
r18194: Allow empty bitmap {} declarations. (This used to be commit 084b5002b9a8c9790ce81480fe26d57e52789dbb)
2006-09-07 02:25:54 +04:00
enum_elements:
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
enum_element { [ $_[1] ] }
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
|
r18194: Allow empty bitmap {} declarations. (This used to be commit 084b5002b9a8c9790ce81480fe26d57e52789dbb)
2006-09-07 02:25:54 +04:00
enum_elements ',' enum_element { push(@{$_[1]}, $_[3]); $_[1] }
r4533: parsing support for: typedef [bitmap16bit] bitmap { FLAG1 = 0x0001, FLAG2 = 0x8000 } fooflags; metze (This used to be commit 192f2495468d5d0c5889d94ca791aa8f371514e8)
2005-01-05 18:26:48 +03:00
;
enum_element:
identifier
|
identifier '=' anytext { "$_[1]$_[2]$_[3]" }
;
r18194: Allow empty bitmap {} declarations. (This used to be commit 084b5002b9a8c9790ce81480fe26d57e52789dbb)
2006-09-07 02:25:54 +04:00
bitmap_body:
'{' opt_bitmap_elements '}' { $_[2] }
;
r4533: parsing support for: typedef [bitmap16bit] bitmap { FLAG1 = 0x0001, FLAG2 = 0x8000 } fooflags; metze (This used to be commit 192f2495468d5d0c5889d94ca791aa8f371514e8)
2005-01-05 18:26:48 +03:00
opt_bitmap_body:
#empty
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
|
r4533: parsing support for: typedef [bitmap16bit] bitmap { FLAG1 = 0x0001, FLAG2 = 0x8000 } fooflags; metze (This used to be commit 192f2495468d5d0c5889d94ca791aa8f371514e8)
2005-01-05 18:26:48 +03:00
bitmap_body
;
r12481: Move parser-specific utility functions to idl.yp, remove some unused functions Allow the use of non-typedef structs and unions when declaring variables. Allow the use of the 'signed' and 'unsigned' qualifiers for integer types (This used to be commit bc6b45e242c8d7b2ef1a6e6d3eb172c27afd952d)
2005-12-25 17:59:21 +03:00
bitmap:
property_list 'bitmap' optional_identifier opt_bitmap_body
{{
r8622: Support parsing new structs/unions/enums/bitmaps without typedef in (This used to be commit cd33b5efc544345411db498b178fa532a9a2c82f)
2005-07-20 03:34:02 +04:00
"TYPE" => "BITMAP",
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
"PROPERTIES" => $_[1],
r21430: Support tagged types without typedef. This means: struct foo { ... }; in IDL will now work. This is the first step towards nested types and using typedefs for partial types (such as "typedef int *bar;"), a requirement for complex uses of represent_as(). (This used to be commit a716aa70f0c90898e6fcf57d63a2cf4c40e7d4df)
2007-02-18 19:21:28 +03:00
"NAME" => $_[3],
r12481: Move parser-specific utility functions to idl.yp, remove some unused functions Allow the use of non-typedef structs and unions when declaring variables. Allow the use of the 'signed' and 'unsigned' qualifiers for integer types (This used to be commit bc6b45e242c8d7b2ef1a6e6d3eb172c27afd952d)
2005-12-25 17:59:21 +03:00
"ELEMENTS" => $_[4],
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
"FILE" => $_[0]->YYData->{FILE},
r6973: Merge new version of pidl into the main SAMBA_4_0 branch. The main difference in this new version is the extra data structure generated between the IDL data structure and the NDR parser: IDL -> NDR -> { ndr_parser, ndr_header, eparser, etc } This makes the ndr_parser.pm internals much more sane. Other changes include: - Remove unnecessary calls with NDR_BUFFERS (for example, GUID doesn't have any buffers, just scalars) as well as some (unnecessary) nested setting of flags. - Parse array loops in the C code rather then calling ndr_pull_array(). This allows us to have, for example, arrays of pointers or arrays of pointers to arrays, etc.. - Use if() {} rather then if () goto foo; everywhere - NDR_IN no longer implies LIBNDR_FLAG_REF_ALLOC - By default, top level pointers are now "ref" (as is the default in most other IDL compilers). This can be overridden using the default_pointer_top() property. - initial work on new ethereal parser generators by Alan DeKok and me - pidl now writes errors in the standard format used by compilers, which is parsable by most editors - ability to warn about the fact that pidl extension(s) have been used, useful for making sure IDL files work with other IDL compilers. oh, and there's probably some other things I can't think of right now.. (This used to be commit 13cf227615f6b9e0e5fa62e59197024410254f01)
2005-05-25 17:50:27 +04:00
"LINE" => $_[0]->YYData->{LINE},
r12481: Move parser-specific utility functions to idl.yp, remove some unused functions Allow the use of non-typedef structs and unions when declaring variables. Allow the use of the 'signed' and 'unsigned' qualifiers for integer types (This used to be commit bc6b45e242c8d7b2ef1a6e6d3eb172c27afd952d)
2005-12-25 17:59:21 +03:00
}}
;
r21430: Support tagged types without typedef. This means: struct foo { ... }; in IDL will now work. This is the first step towards nested types and using typedefs for partial types (such as "typedef int *bar;"), a requirement for complex uses of represent_as(). (This used to be commit a716aa70f0c90898e6fcf57d63a2cf4c40e7d4df)
2007-02-18 19:21:28 +03:00
bitmap_elements:
bitmap_element { [ $_[1] ] }
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
|
r21430: Support tagged types without typedef. This means: struct foo { ... }; in IDL will now work. This is the first step towards nested types and using typedefs for partial types (such as "typedef int *bar;"), a requirement for complex uses of represent_as(). (This used to be commit a716aa70f0c90898e6fcf57d63a2cf4c40e7d4df)
2007-02-18 19:21:28 +03:00
bitmap_elements ',' bitmap_element { push(@{$_[1]}, $_[3]); $_[1] }
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
;
r4533: parsing support for: typedef [bitmap16bit] bitmap { FLAG1 = 0x0001, FLAG2 = 0x8000 } fooflags; metze (This used to be commit 192f2495468d5d0c5889d94ca791aa8f371514e8)
2005-01-05 18:26:48 +03:00
opt_bitmap_elements:
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
#empty
|
r4533: parsing support for: typedef [bitmap16bit] bitmap { FLAG1 = 0x0001, FLAG2 = 0x8000 } fooflags; metze (This used to be commit 192f2495468d5d0c5889d94ca791aa8f371514e8)
2005-01-05 18:26:48 +03:00
bitmap_elements
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
;
r5320: Treat structs and unions somewhat more similarly: - use same names in hashes (DATA -> ELEMENTS, etc) - [case()] and [default] are no longer special case, they're just regular properties (This used to be commit 6a0f599f822f5c42d93e78c13765804ab2223968)
2005-02-11 05:05:47 +03:00
bitmap_element:
r4533: parsing support for: typedef [bitmap16bit] bitmap { FLAG1 = 0x0001, FLAG2 = 0x8000 } fooflags; metze (This used to be commit 192f2495468d5d0c5889d94ca791aa8f371514e8)
2005-01-05 18:26:48 +03:00
identifier '=' anytext { "$_[1] ( $_[3] )" }
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
;
r12481: Move parser-specific utility functions to idl.yp, remove some unused functions Allow the use of non-typedef structs and unions when declaring variables. Allow the use of the 'signed' and 'unsigned' qualifiers for integer types (This used to be commit bc6b45e242c8d7b2ef1a6e6d3eb172c27afd952d)
2005-12-25 17:59:21 +03:00
struct_body:
r5320: Treat structs and unions somewhat more similarly: - use same names in hashes (DATA -> ELEMENTS, etc) - [case()] and [default] are no longer special case, they're just regular properties (This used to be commit 6a0f599f822f5c42d93e78c13765804ab2223968)
2005-02-11 05:05:47 +03:00
'{' element_list1 '}' { $_[2] }
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
;
r12481: Move parser-specific utility functions to idl.yp, remove some unused functions Allow the use of non-typedef structs and unions when declaring variables. Allow the use of the 'signed' and 'unsigned' qualifiers for integer types (This used to be commit bc6b45e242c8d7b2ef1a6e6d3eb172c27afd952d)
2005-12-25 17:59:21 +03:00
opt_struct_body:
r5320: Treat structs and unions somewhat more similarly: - use same names in hashes (DATA -> ELEMENTS, etc) - [case()] and [default] are no longer special case, they're just regular properties (This used to be commit 6a0f599f822f5c42d93e78c13765804ab2223968)
2005-02-11 05:05:47 +03:00
#empty
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
|
r21430: Support tagged types without typedef. This means: struct foo { ... }; in IDL will now work. This is the first step towards nested types and using typedefs for partial types (such as "typedef int *bar;"), a requirement for complex uses of represent_as(). (This used to be commit a716aa70f0c90898e6fcf57d63a2cf4c40e7d4df)
2007-02-18 19:21:28 +03:00
struct_body
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
;
struct:
r5320: Treat structs and unions somewhat more similarly: - use same names in hashes (DATA -> ELEMENTS, etc) - [case()] and [default] are no longer special case, they're just regular properties (This used to be commit 6a0f599f822f5c42d93e78c13765804ab2223968)
2005-02-11 05:05:47 +03:00
property_list 'struct' optional_identifier opt_struct_body
{{
"TYPE" => "STRUCT",
r6098: fix parsing of empty union cases metze (This used to be commit f3c64120a16289472bdc56329d39c7221d00b558)
2005-03-28 22:22:45 +04:00
"PROPERTIES" => $_[1],
r6973: Merge new version of pidl into the main SAMBA_4_0 branch. The main difference in this new version is the extra data structure generated between the IDL data structure and the NDR parser: IDL -> NDR -> { ndr_parser, ndr_header, eparser, etc } This makes the ndr_parser.pm internals much more sane. Other changes include: - Remove unnecessary calls with NDR_BUFFERS (for example, GUID doesn't have any buffers, just scalars) as well as some (unnecessary) nested setting of flags. - Parse array loops in the C code rather then calling ndr_pull_array(). This allows us to have, for example, arrays of pointers or arrays of pointers to arrays, etc.. - Use if() {} rather then if () goto foo; everywhere - NDR_IN no longer implies LIBNDR_FLAG_REF_ALLOC - By default, top level pointers are now "ref" (as is the default in most other IDL compilers). This can be overridden using the default_pointer_top() property. - initial work on new ethereal parser generators by Alan DeKok and me - pidl now writes errors in the standard format used by compilers, which is parsable by most editors - ability to warn about the fact that pidl extension(s) have been used, useful for making sure IDL files work with other IDL compilers. oh, and there's probably some other things I can't think of right now.. (This used to be commit 13cf227615f6b9e0e5fa62e59197024410254f01)
2005-05-25 17:50:27 +04:00
"NAME" => $_[3],
r7022: Add support for parsing definitions of multi-dimension arrays. This will also be required for supporting parsing pointers to arrays and arrays of pointers simultaneously. (This used to be commit a34f848b02a6e8284d62532a792a5136e846fe8f)
2005-05-27 18:01:22 +04:00
"ELEMENTS" => $_[4],
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
"FILE" => $_[0]->YYData->{FILE},
r6973: Merge new version of pidl into the main SAMBA_4_0 branch. The main difference in this new version is the extra data structure generated between the IDL data structure and the NDR parser: IDL -> NDR -> { ndr_parser, ndr_header, eparser, etc } This makes the ndr_parser.pm internals much more sane. Other changes include: - Remove unnecessary calls with NDR_BUFFERS (for example, GUID doesn't have any buffers, just scalars) as well as some (unnecessary) nested setting of flags. - Parse array loops in the C code rather then calling ndr_pull_array(). This allows us to have, for example, arrays of pointers or arrays of pointers to arrays, etc.. - Use if() {} rather then if () goto foo; everywhere - NDR_IN no longer implies LIBNDR_FLAG_REF_ALLOC - By default, top level pointers are now "ref" (as is the default in most other IDL compilers). This can be overridden using the default_pointer_top() property. - initial work on new ethereal parser generators by Alan DeKok and me - pidl now writes errors in the standard format used by compilers, which is parsable by most editors - ability to warn about the fact that pidl extension(s) have been used, useful for making sure IDL files work with other IDL compilers. oh, and there's probably some other things I can't think of right now.. (This used to be commit 13cf227615f6b9e0e5fa62e59197024410254f01)
2005-05-25 17:50:27 +04:00
"LINE" => $_[0]->YYData->{LINE},
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
}}
;
r5320: Treat structs and unions somewhat more similarly: - use same names in hashes (DATA -> ELEMENTS, etc) - [case()] and [default] are no longer special case, they're just regular properties (This used to be commit 6a0f599f822f5c42d93e78c13765804ab2223968)
2005-02-11 05:05:47 +03:00
empty_element:
r12481: Move parser-specific utility functions to idl.yp, remove some unused functions Allow the use of non-typedef structs and unions when declaring variables. Allow the use of the 'signed' and 'unsigned' qualifiers for integer types (This used to be commit bc6b45e242c8d7b2ef1a6e6d3eb172c27afd952d)
2005-12-25 17:59:21 +03:00
property_list ';'
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
{{
r5320: Treat structs and unions somewhat more similarly: - use same names in hashes (DATA -> ELEMENTS, etc) - [case()] and [default] are no longer special case, they're just regular properties (This used to be commit 6a0f599f822f5c42d93e78c13765804ab2223968)
2005-02-11 05:05:47 +03:00
"NAME" => "",
r6973: Merge new version of pidl into the main SAMBA_4_0 branch. The main difference in this new version is the extra data structure generated between the IDL data structure and the NDR parser: IDL -> NDR -> { ndr_parser, ndr_header, eparser, etc } This makes the ndr_parser.pm internals much more sane. Other changes include: - Remove unnecessary calls with NDR_BUFFERS (for example, GUID doesn't have any buffers, just scalars) as well as some (unnecessary) nested setting of flags. - Parse array loops in the C code rather then calling ndr_pull_array(). This allows us to have, for example, arrays of pointers or arrays of pointers to arrays, etc.. - Use if() {} rather then if () goto foo; everywhere - NDR_IN no longer implies LIBNDR_FLAG_REF_ALLOC - By default, top level pointers are now "ref" (as is the default in most other IDL compilers). This can be overridden using the default_pointer_top() property. - initial work on new ethereal parser generators by Alan DeKok and me - pidl now writes errors in the standard format used by compilers, which is parsable by most editors - ability to warn about the fact that pidl extension(s) have been used, useful for making sure IDL files work with other IDL compilers. oh, and there's probably some other things I can't think of right now.. (This used to be commit 13cf227615f6b9e0e5fa62e59197024410254f01)
2005-05-25 17:50:27 +04:00
"TYPE" => "EMPTY",
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
"PROPERTIES" => $_[1],
"POINTERS" => 0,
r6973: Merge new version of pidl into the main SAMBA_4_0 branch. The main difference in this new version is the extra data structure generated between the IDL data structure and the NDR parser: IDL -> NDR -> { ndr_parser, ndr_header, eparser, etc } This makes the ndr_parser.pm internals much more sane. Other changes include: - Remove unnecessary calls with NDR_BUFFERS (for example, GUID doesn't have any buffers, just scalars) as well as some (unnecessary) nested setting of flags. - Parse array loops in the C code rather then calling ndr_pull_array(). This allows us to have, for example, arrays of pointers or arrays of pointers to arrays, etc.. - Use if() {} rather then if () goto foo; everywhere - NDR_IN no longer implies LIBNDR_FLAG_REF_ALLOC - By default, top level pointers are now "ref" (as is the default in most other IDL compilers). This can be overridden using the default_pointer_top() property. - initial work on new ethereal parser generators by Alan DeKok and me - pidl now writes errors in the standard format used by compilers, which is parsable by most editors - ability to warn about the fact that pidl extension(s) have been used, useful for making sure IDL files work with other IDL compilers. oh, and there's probably some other things I can't think of right now.. (This used to be commit 13cf227615f6b9e0e5fa62e59197024410254f01)
2005-05-25 17:50:27 +04:00
"ARRAY_LEN" => [],
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
"FILE" => $_[0]->YYData->{FILE},
r6973: Merge new version of pidl into the main SAMBA_4_0 branch. The main difference in this new version is the extra data structure generated between the IDL data structure and the NDR parser: IDL -> NDR -> { ndr_parser, ndr_header, eparser, etc } This makes the ndr_parser.pm internals much more sane. Other changes include: - Remove unnecessary calls with NDR_BUFFERS (for example, GUID doesn't have any buffers, just scalars) as well as some (unnecessary) nested setting of flags. - Parse array loops in the C code rather then calling ndr_pull_array(). This allows us to have, for example, arrays of pointers or arrays of pointers to arrays, etc.. - Use if() {} rather then if () goto foo; everywhere - NDR_IN no longer implies LIBNDR_FLAG_REF_ALLOC - By default, top level pointers are now "ref" (as is the default in most other IDL compilers). This can be overridden using the default_pointer_top() property. - initial work on new ethereal parser generators by Alan DeKok and me - pidl now writes errors in the standard format used by compilers, which is parsable by most editors - ability to warn about the fact that pidl extension(s) have been used, useful for making sure IDL files work with other IDL compilers. oh, and there's probably some other things I can't think of right now.. (This used to be commit 13cf227615f6b9e0e5fa62e59197024410254f01)
2005-05-25 17:50:27 +04:00
"LINE" => $_[0]->YYData->{LINE},
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
}}
r5320: Treat structs and unions somewhat more similarly: - use same names in hashes (DATA -> ELEMENTS, etc) - [case()] and [default] are no longer special case, they're just regular properties (This used to be commit 6a0f599f822f5c42d93e78c13765804ab2223968)
2005-02-11 05:05:47 +03:00
;
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
r5320: Treat structs and unions somewhat more similarly: - use same names in hashes (DATA -> ELEMENTS, etc) - [case()] and [default] are no longer special case, they're just regular properties (This used to be commit 6a0f599f822f5c42d93e78c13765804ab2223968)
2005-02-11 05:05:47 +03:00
base_or_empty:
base_element ';'
|
empty_element;
optional_base_element:
r12481: Move parser-specific utility functions to idl.yp, remove some unused functions Allow the use of non-typedef structs and unions when declaring variables. Allow the use of the 'signed' and 'unsigned' qualifiers for integer types (This used to be commit bc6b45e242c8d7b2ef1a6e6d3eb172c27afd952d)
2005-12-25 17:59:21 +03:00
property_list base_or_empty { $_[2]->{PROPERTIES} = FlattenHash([$_[1],$_[2]->{PROPERTIES}]); $_[2] }
r5320: Treat structs and unions somewhat more similarly: - use same names in hashes (DATA -> ELEMENTS, etc) - [case()] and [default] are no longer special case, they're just regular properties (This used to be commit 6a0f599f822f5c42d93e78c13765804ab2223968)
2005-02-11 05:05:47 +03:00
;
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
union_elements:
r5320: Treat structs and unions somewhat more similarly: - use same names in hashes (DATA -> ELEMENTS, etc) - [case()] and [default] are no longer special case, they're just regular properties (This used to be commit 6a0f599f822f5c42d93e78c13765804ab2223968)
2005-02-11 05:05:47 +03:00
#empty
|
union_elements optional_base_element { push(@{$_[1]}, $_[2]); $_[1] }
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
;
r12481: Move parser-specific utility functions to idl.yp, remove some unused functions Allow the use of non-typedef structs and unions when declaring variables. Allow the use of the 'signed' and 'unsigned' qualifiers for integer types (This used to be commit bc6b45e242c8d7b2ef1a6e6d3eb172c27afd952d)
2005-12-25 17:59:21 +03:00
union_body:
'{' union_elements '}' { $_[2] }
;
r21430: Support tagged types without typedef. This means: struct foo { ... }; in IDL will now work. This is the first step towards nested types and using typedefs for partial types (such as "typedef int *bar;"), a requirement for complex uses of represent_as(). (This used to be commit a716aa70f0c90898e6fcf57d63a2cf4c40e7d4df)
2007-02-18 19:21:28 +03:00
opt_union_body:
#empty
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
|
r21430: Support tagged types without typedef. This means: struct foo { ... }; in IDL will now work. This is the first step towards nested types and using typedefs for partial types (such as "typedef int *bar;"), a requirement for complex uses of represent_as(). (This used to be commit a716aa70f0c90898e6fcf57d63a2cf4c40e7d4df)
2007-02-18 19:21:28 +03:00
union_body
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
;
union:
property_list 'union' optional_identifier opt_union_body
{{
"TYPE" => "UNION",
"PROPERTIES" => $_[1],
"NAME" => $_[3],
r6973: Merge new version of pidl into the main SAMBA_4_0 branch. The main difference in this new version is the extra data structure generated between the IDL data structure and the NDR parser: IDL -> NDR -> { ndr_parser, ndr_header, eparser, etc } This makes the ndr_parser.pm internals much more sane. Other changes include: - Remove unnecessary calls with NDR_BUFFERS (for example, GUID doesn't have any buffers, just scalars) as well as some (unnecessary) nested setting of flags. - Parse array loops in the C code rather then calling ndr_pull_array(). This allows us to have, for example, arrays of pointers or arrays of pointers to arrays, etc.. - Use if() {} rather then if () goto foo; everywhere - NDR_IN no longer implies LIBNDR_FLAG_REF_ALLOC - By default, top level pointers are now "ref" (as is the default in most other IDL compilers). This can be overridden using the default_pointer_top() property. - initial work on new ethereal parser generators by Alan DeKok and me - pidl now writes errors in the standard format used by compilers, which is parsable by most editors - ability to warn about the fact that pidl extension(s) have been used, useful for making sure IDL files work with other IDL compilers. oh, and there's probably some other things I can't think of right now.. (This used to be commit 13cf227615f6b9e0e5fa62e59197024410254f01)
2005-05-25 17:50:27 +04:00
"ELEMENTS" => $_[4],
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
"FILE" => $_[0]->YYData->{FILE},
r6973: Merge new version of pidl into the main SAMBA_4_0 branch. The main difference in this new version is the extra data structure generated between the IDL data structure and the NDR parser: IDL -> NDR -> { ndr_parser, ndr_header, eparser, etc } This makes the ndr_parser.pm internals much more sane. Other changes include: - Remove unnecessary calls with NDR_BUFFERS (for example, GUID doesn't have any buffers, just scalars) as well as some (unnecessary) nested setting of flags. - Parse array loops in the C code rather then calling ndr_pull_array(). This allows us to have, for example, arrays of pointers or arrays of pointers to arrays, etc.. - Use if() {} rather then if () goto foo; everywhere - NDR_IN no longer implies LIBNDR_FLAG_REF_ALLOC - By default, top level pointers are now "ref" (as is the default in most other IDL compilers). This can be overridden using the default_pointer_top() property. - initial work on new ethereal parser generators by Alan DeKok and me - pidl now writes errors in the standard format used by compilers, which is parsable by most editors - ability to warn about the fact that pidl extension(s) have been used, useful for making sure IDL files work with other IDL compilers. oh, and there's probably some other things I can't think of right now.. (This used to be commit 13cf227615f6b9e0e5fa62e59197024410254f01)
2005-05-25 17:50:27 +04:00
"LINE" => $_[0]->YYData->{LINE},
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
}}
;
base_element:
r11097: Allow pointers in "const" (This used to be commit 88fcd56fb55ac3f7e1eef3b3c92a2593c79599d3)
2005-10-16 21:17:42 +04:00
property_list type pointers identifier array_len
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
{{
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
"NAME" => $_[4],
"TYPE" => $_[2],
r21430: Support tagged types without typedef. This means: struct foo { ... }; in IDL will now work. This is the first step towards nested types and using typedefs for partial types (such as "typedef int *bar;"), a requirement for complex uses of represent_as(). (This used to be commit a716aa70f0c90898e6fcf57d63a2cf4c40e7d4df)
2007-02-18 19:21:28 +03:00
"PROPERTIES" => $_[1],
r11097: Allow pointers in "const" (This used to be commit 88fcd56fb55ac3f7e1eef3b3c92a2593c79599d3)
2005-10-16 21:17:42 +04:00
"POINTERS" => $_[3],
"ARRAY_LEN" => $_[5],
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
"FILE" => $_[0]->YYData->{FILE},
r6973: Merge new version of pidl into the main SAMBA_4_0 branch. The main difference in this new version is the extra data structure generated between the IDL data structure and the NDR parser: IDL -> NDR -> { ndr_parser, ndr_header, eparser, etc } This makes the ndr_parser.pm internals much more sane. Other changes include: - Remove unnecessary calls with NDR_BUFFERS (for example, GUID doesn't have any buffers, just scalars) as well as some (unnecessary) nested setting of flags. - Parse array loops in the C code rather then calling ndr_pull_array(). This allows us to have, for example, arrays of pointers or arrays of pointers to arrays, etc.. - Use if() {} rather then if () goto foo; everywhere - NDR_IN no longer implies LIBNDR_FLAG_REF_ALLOC - By default, top level pointers are now "ref" (as is the default in most other IDL compilers). This can be overridden using the default_pointer_top() property. - initial work on new ethereal parser generators by Alan DeKok and me - pidl now writes errors in the standard format used by compilers, which is parsable by most editors - ability to warn about the fact that pidl extension(s) have been used, useful for making sure IDL files work with other IDL compilers. oh, and there's probably some other things I can't think of right now.. (This used to be commit 13cf227615f6b9e0e5fa62e59197024410254f01)
2005-05-25 17:50:27 +04:00
"LINE" => $_[0]->YYData->{LINE},
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
}}
;
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
pointers:
#empty
{ 0 }
|
pointers '*' { $_[1]+1 }
;
pidl: parse idl 'pipe' typedefs, but print out a not supported message for now metze
2009-08-05 15:40:38 +04:00
pipe:
property_list 'pipe' type
{{
"TYPE" => "PIPE",
"PROPERTIES" => $_[1],
pidl/idl.yp: fill 'pipe' elements with array and count metze
2010-09-21 21:30:08 +04:00
"NAME" => undef,
"DATA" => {
"TYPE" => "STRUCT",
"PROPERTIES" => $_[1],
"NAME" => undef,
"ELEMENTS" => [{
"NAME" => "count",
"PROPERTIES" => $_[1],
"POINTERS" => 0,
"ARRAY_LEN" => [],
"TYPE" => "uint3264",
"FILE" => $_[0]->YYData->{FILE},
"LINE" => $_[0]->YYData->{LINE},
},{
"NAME" => "array",
"PROPERTIES" => $_[1],
"POINTERS" => 0,
"ARRAY_LEN" => [ "count" ],
"TYPE" => $_[3],
"FILE" => $_[0]->YYData->{FILE},
"LINE" => $_[0]->YYData->{LINE},
}],
"FILE" => $_[0]->YYData->{FILE},
"LINE" => $_[0]->YYData->{LINE},
},
pidl: parse idl 'pipe' typedefs, but print out a not supported message for now metze
2009-08-05 15:40:38 +04:00
"FILE" => $_[0]->YYData->{FILE},
"LINE" => $_[0]->YYData->{LINE},
}}
;
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
element_list1:
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
#empty
r21681: Fix bug in the parsing code that parsed "struct foo;" the same as "struct foo {};". Reported by one of the OpenChange folks, thanks! (This used to be commit d65b520f08ea4ee82c35ff334a58aa6ffc403d67)
2007-03-04 17:16:52 +03:00
{ [] }
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
|
element_list1 base_element ';' { push(@{$_[1]}, $_[2]); $_[1] }
;
Import proper IDL for interface pointers.
2008-09-16 06:15:39 +04:00
optional_const:
#empty
|
'const'
;
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
element_list2:
#empty
* reduced the number of grammer conflicts a lot using (arbitrary) precedence rules * build a standalone parser. When we come to distributing Samba4 we can just include idl.pm and other developers won't need Parse::Yapp installed * avoid the recursive make in most cases in build_idl.sh (This used to be commit be2c2be459d1bed41c113590e70711cb89ad12b9)
2003-12-08 05:24:40 +03:00
|
'void'
Import proper IDL for interface pointers.
2008-09-16 06:15:39 +04:00
|
optional_const base_element { [ $_[2] ] }
|
element_list2 ',' optional_const base_element { push(@{$_[1]}, $_[4]); $_[1] }
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
;
array_len:
r7022: Add support for parsing definitions of multi-dimension arrays. This will also be required for supporting parsing pointers to arrays and arrays of pointers simultaneously. (This used to be commit a34f848b02a6e8284d62532a792a5136e846fe8f)
2005-05-27 18:01:22 +04:00
#empty { [] }
|
'[' ']' array_len { push(@{$_[3]}, "*"); $_[3] }
|
'[' anytext ']' array_len { push(@{$_[4]}, "$_[2]"); $_[4] }
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
;
property_list:
* reduced the number of grammer conflicts a lot using (arbitrary) precedence rules * build a standalone parser. When we come to distributing Samba4 we can just include idl.pm and other developers won't need Parse::Yapp installed * avoid the recursive make in most cases in build_idl.sh (This used to be commit be2c2be459d1bed41c113590e70711cb89ad12b9)
2003-12-08 05:24:40 +03:00
#empty
r12481: Move parser-specific utility functions to idl.yp, remove some unused functions Allow the use of non-typedef structs and unions when declaring variables. Allow the use of the 'signed' and 'unsigned' qualifiers for integer types (This used to be commit bc6b45e242c8d7b2ef1a6e6d3eb172c27afd952d)
2005-12-25 17:59:21 +03:00
|
property_list '[' properties ']' { FlattenHash([$_[1],$_[3]]); }
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
;
commit idl.pm now, as many build farm machines don't have 'yapp' (This used to be commit 2b1300386dc8a3e3e802c8b1dafdbea286cfbb46)
2003-12-08 05:55:28 +03:00
properties:
property { $_[1] }
r12481: Move parser-specific utility functions to idl.yp, remove some unused functions Allow the use of non-typedef structs and unions when declaring variables. Allow the use of the 'signed' and 'unsigned' qualifiers for integer types (This used to be commit bc6b45e242c8d7b2ef1a6e6d3eb172c27afd952d)
2005-12-25 17:59:21 +03:00
|
properties ',' property { FlattenHash([$_[1], $_[3]]); }
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
;
commit idl.pm now, as many build farm machines don't have 'yapp' (This used to be commit 2b1300386dc8a3e3e802c8b1dafdbea286cfbb46)
2003-12-08 05:55:28 +03:00
property:
identifier {{ "$_[1]" => "1" }}
pidl/IDL: don't strip ',' from the properties content metze (This used to be commit fdf9bcb163516f7d96675ae0dce2917afb8f86d3)
2008-01-31 17:30:50 +03:00
|
identifier '(' commalisttext ')' {{ "$_[1]" => "$_[3]" }}
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
;
r2973: Allow comma's inside parentheses in property arguments (This used to be commit ab2a788fe75ddaf8ff493477f2006a03959e6ab5)
2004-10-14 14:30:08 +04:00
commalisttext:
anytext
|
commalisttext ',' anytext { "$_[1],$_[3]" }
;
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
anytext:
#empty
{ "" }
|
identifier
|
constant
|
text
|
anytext '-' anytext { "$_[1]$_[2]$_[3]" }
|
anytext '.' anytext { "$_[1]$_[2]$_[3]" }
|
anytext '*' anytext { "$_[1]$_[2]$_[3]" }
|
anytext '>' anytext { "$_[1]$_[2]$_[3]" }
r4884: - 2nd part of support ndr_size_ generation on unions as well as structures - added "nopull" and "nopush" flags, to allow for externally written parsers for sub-structures (This used to be commit f65f239978425de795a0e188aaad3d5d1167da32)
2005-01-21 09:46:07 +03:00
|
anytext '<' anytext { "$_[1]$_[2]$_[3]" }
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
|
anytext '|' anytext { "$_[1]$_[2]$_[3]" }
|
anytext '&' anytext { "$_[1]$_[2]$_[3]" }
|
anytext '/' anytext { "$_[1]$_[2]$_[3]" }
r10014: Support ?, : and = in attribute arguments. (This used to be commit 6a549f35583e71531bea56ce717635ad9499441a)
2005-09-04 02:57:30 +04:00
|
anytext '?' anytext { "$_[1]$_[2]$_[3]" }
|
anytext ':' anytext { "$_[1]$_[2]$_[3]" }
|
anytext '=' anytext { "$_[1]$_[2]$_[3]" }
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
|
anytext '+' anytext { "$_[1]$_[2]$_[3]" }
r7036: Allow more operations in several properties (such as size_is, length_is, switch_is, etc) and simplify the code involved (This used to be commit 86de98ee09348297f2c30ce15888ba5e5637c078)
2005-05-27 22:14:43 +04:00
|
anytext '~' anytext { "$_[1]$_[2]$_[3]" }
r2973: Allow comma's inside parentheses in property arguments (This used to be commit ab2a788fe75ddaf8ff493477f2006a03959e6ab5)
2004-10-14 14:30:08 +04:00
|
anytext '(' commalisttext ')' anytext { "$_[1]$_[2]$_[3]$_[4]$_[5]" }
r3998: allow const arrays (see next commit to drsuapi.idl) metze (This used to be commit 7264a2227d7a1881db01964d0d713c144eb77152)
2004-11-29 14:08:15 +03:00
|
anytext '{' commalisttext '}' anytext { "$_[1]$_[2]$_[3]$_[4]$_[5]" }
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
;
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
identifier:
r8622: Support parsing new structs/unions/enums/bitmaps without typedef in (This used to be commit cd33b5efc544345411db498b178fa532a9a2c82f)
2005-07-20 03:34:02 +04:00
IDENTIFIER
;
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
optional_identifier:
r8622: Support parsing new structs/unions/enums/bitmaps without typedef in (This used to be commit cd33b5efc544345411db498b178fa532a9a2c82f)
2005-07-20 03:34:02 +04:00
#empty { undef }
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
|
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
IDENTIFIER
;
constant:
CONSTANT
;
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
text:
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
TEXT { "\"$_[1]\"" }
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
;
r3431: Allow optional semicolon after interface definitions (This used to be commit edbd789abb505accd3e72863d13acd2603d991f6)
2004-11-01 15:26:59 +03:00
optional_semicolon:
#empty
|
';'
;
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
#####################################
# start code
%%
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
use Parse::Pidl qw(error);
r12481: Move parser-specific utility functions to idl.yp, remove some unused functions Allow the use of non-typedef structs and unions when declaring variables. Allow the use of the 'signed' and 'unsigned' qualifiers for integer types (This used to be commit bc6b45e242c8d7b2ef1a6e6d3eb172c27afd952d)
2005-12-25 17:59:21 +03:00
#####################################################################
# flatten an array of hashes into a single hash
sub FlattenHash($)
{
my $a = shift;
my %b;
for my $d (@{$a}) {
for my $k (keys %{$d}) {
$b{$k} = $d->{$k};
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
}
r12481: Move parser-specific utility functions to idl.yp, remove some unused functions Allow the use of non-typedef structs and unions when declaring variables. Allow the use of the 'signed' and 'unsigned' qualifiers for integer types (This used to be commit bc6b45e242c8d7b2ef1a6e6d3eb172c27afd952d)
2005-12-25 17:59:21 +03:00
}
return \%b;
}
r8274: Export some more symbols. (This used to be commit d1f754a0a34c5938579a605b4f113100e14bac3d)
2005-07-10 05:16:02 +04:00
#####################################################################
# traverse a perl data structure removing any empty arrays or
# hashes and any hash elements that map to undef
sub CleanData($)
{
sub CleanData($);
my($v) = shift;
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
r12464: Add simple IDL parsing tests for pidl using the standard perl testing framework (Test::Simple, distributed with perl itself). Run these tests from 'make test' (This used to be commit 975d8816db6697dab828941b69a740e3a0a2c272)
2005-12-25 02:32:50 +03:00
return undef if (not defined($v));
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
r8274: Export some more symbols. (This used to be commit d1f754a0a34c5938579a605b4f113100e14bac3d)
2005-07-10 05:16:02 +04:00
if (ref($v) eq "ARRAY") {
foreach my $i (0 .. $#{$v}) {
CleanData($v->[$i]);
}
# this removes any undefined elements from the array
@{$v} = grep { defined $_ } @{$v};
} elsif (ref($v) eq "HASH") {
foreach my $x (keys %{$v}) {
CleanData($v->{$x});
if (!defined $v->{$x}) {
delete($v->{$x});
next;
}
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
}
r8274: Export some more symbols. (This used to be commit d1f754a0a34c5938579a605b4f113100e14bac3d)
2005-07-10 05:16:02 +04:00
}
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
r8274: Export some more symbols. (This used to be commit d1f754a0a34c5938579a605b4f113100e14bac3d)
2005-07-10 05:16:02 +04:00
return $v;
}
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
sub _Error {
r6973: Merge new version of pidl into the main SAMBA_4_0 branch. The main difference in this new version is the extra data structure generated between the IDL data structure and the NDR parser: IDL -> NDR -> { ndr_parser, ndr_header, eparser, etc } This makes the ndr_parser.pm internals much more sane. Other changes include: - Remove unnecessary calls with NDR_BUFFERS (for example, GUID doesn't have any buffers, just scalars) as well as some (unnecessary) nested setting of flags. - Parse array loops in the C code rather then calling ndr_pull_array(). This allows us to have, for example, arrays of pointers or arrays of pointers to arrays, etc.. - Use if() {} rather then if () goto foo; everywhere - NDR_IN no longer implies LIBNDR_FLAG_REF_ALLOC - By default, top level pointers are now "ref" (as is the default in most other IDL compilers). This can be overridden using the default_pointer_top() property. - initial work on new ethereal parser generators by Alan DeKok and me - pidl now writes errors in the standard format used by compilers, which is parsable by most editors - ability to warn about the fact that pidl extension(s) have been used, useful for making sure IDL files work with other IDL compilers. oh, and there's probably some other things I can't think of right now.. (This used to be commit 13cf227615f6b9e0e5fa62e59197024410254f01)
2005-05-25 17:50:27 +04:00
if (exists $_[0]->YYData->{ERRMSG}) {
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
error($_[0]->YYData, $_[0]->YYData->{ERRMSG});
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
delete $_[0]->YYData->{ERRMSG};
return;
pidl: Remove superfluous semicolon in perl code. Michael (This used to be commit 85f707c6abc48739c0c3b423ac0b196d2b75e40c)
2008-02-26 17:38:45 +03:00
}
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
my $last_token = $_[0]->YYData->{LAST_TOKEN};
pidl: fix formatting in idl.yp metze
2009-08-04 16:59:33 +04:00
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
error($_[0]->YYData, "Syntax error near '$last_token'");
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
}
a bit more speed and better line matching in errors (This used to be commit 788f6c44dbae48e13c9eeaca3bfa3ac7efbbd4e2)
2003-12-07 16:51:23 +03:00
sub _Lexer($)
{
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
my($parser)=shift;
r8690: Code cleanups, improvements necessary for dependencies (not added yet) (This used to be commit 7384b3c4a1450ad63356dcf586e3a7e7335f8e6d)
2005-07-22 01:27:39 +04:00
$parser->YYData->{INPUT} or return('',undef);
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
again:
$parser->YYData->{INPUT} =~ s/^[ \t]*//;
for ($parser->YYData->{INPUT}) {
a bit more speed and better line matching in errors (This used to be commit 788f6c44dbae48e13c9eeaca3bfa3ac7efbbd4e2)
2003-12-07 16:51:23 +03:00
if (/^\#/) {
PIDL: fix parsing linemarkers in preprocessor output When PIDL calls out to C preprocessor to expand IDL files and parse the output, it filters out linemarkers and line control information as described in http://gcc.gnu.org/onlinedocs/cpp/Preprocessor-Output.html and http://gcc.gnu.org/onlinedocs/cpp/Line-Control.html#Line-Control With gcc 4.8 stdc-predef.h is included automatically and linemarker for the file has extended flags that PIDL couldn't parse ('system header that needs to be extern "C" protected for C++') Thanks to Jakub Jelinek <jakub@redhat.com> for explanation of the linemarker format. Fixes https://bugzilla.redhat.com/show_bug.cgi?id=906517 Reviewed-by: Andreas Schneider <asn@samba.org>
2013-02-06 12:17:57 +04:00
# Linemarker format is described at
# http://gcc.gnu.org/onlinedocs/cpp/Preprocessor-Output.html
if (s/^\# (\d+) \"(.*?)\"(( \d+){1,4}|)//) {
a bit more speed and better line matching in errors (This used to be commit 788f6c44dbae48e13c9eeaca3bfa3ac7efbbd4e2)
2003-12-07 16:51:23 +03:00
$parser->YYData->{LINE} = $1-1;
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
$parser->YYData->{FILE} = $2;
a bit more speed and better line matching in errors (This used to be commit 788f6c44dbae48e13c9eeaca3bfa3ac7efbbd4e2)
2003-12-07 16:51:23 +03:00
goto again;
}
cope with different cpp formatting on some systems, and make sure we don't loop forever if we get a cpp format line we don't understand (This used to be commit 003c6c493b1babadb1f7c4eb54babed394bd42a9)
2003-12-16 03:38:33 +03:00
if (s/^\#line (\d+) \"(.*?)\"( \d+|)//) {
$parser->YYData->{LINE} = $1-1;
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
$parser->YYData->{FILE} = $2;
cope with different cpp formatting on some systems, and make sure we don't loop forever if we get a cpp format line we don't understand (This used to be commit 003c6c493b1babadb1f7c4eb54babed394bd42a9)
2003-12-16 03:38:33 +03:00
goto again;
}
if (s/^(\#.*)$//m) {
goto again;
}
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
}
if (s/^(\n)//) {
$parser->YYData->{LINE}++;
goto again;
}
if (s/^\"(.*?)\"//) {
$parser->YYData->{LAST_TOKEN} = $1;
return('TEXT',$1);
}
if (s/^(\d+)(\W|$)/$2/) {
$parser->YYData->{LAST_TOKEN} = $1;
return('CONSTANT',$1);
}
if (s/^([\w_]+)//) {
$parser->YYData->{LAST_TOKEN} = $1;
if ($1 =~
pidl: Remove support for the "declare" keyword in the parser. (This used to be commit ea3371389226a756ba641a7fc02cfbee78663411)
2008-01-12 03:10:59 +03:00
/^(coclass|interface|import|importlib
|include|cpp_quote|typedef
pidl: parse idl 'pipe' typedefs, but print out a not supported message for now metze
2009-08-05 15:40:38 +04:00
|union|struct|enum|bitmap|pipe
r19585: Add support for some more standard IDL instructions: - `include' (replaces helper()) - `import' (replaces depends()) Add support for parsing importlib() - importlib() is now ignored (with a warning), but no longer causes syntax errors. helper() and depends() are now marked deprecated and will cause warnings. (This used to be commit 1ccab71cb8a9e3db9448b6679d01ad00e1c1e9a3)
2006-11-07 00:54:19 +03:00
|void|const|unsigned|signed)$/x) {
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
return $1;
}
return('IDENTIFIER',$1);
}
if (s/^(.)//s) {
$parser->YYData->{LAST_TOKEN} = $1;
return($1,$1);
}
}
}
r12464: Add simple IDL parsing tests for pidl using the standard perl testing framework (Test::Simple, distributed with perl itself). Run these tests from 'make test' (This used to be commit 975d8816db6697dab828941b69a740e3a0a2c272)
2005-12-25 02:32:50 +03:00
sub parse_string
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
{
r12464: Add simple IDL parsing tests for pidl using the standard perl testing framework (Test::Simple, distributed with perl itself). Run these tests from 'make test' (This used to be commit 975d8816db6697dab828941b69a740e3a0a2c272)
2005-12-25 02:32:50 +03:00
my ($data,$filename) = @_;
r12462: Hide oo magic from callers of the parser (This used to be commit 0c225d2978b11898d334ec2d6da910e0cca0f4cd)
2005-12-25 00:57:51 +03:00
my $self = new Parse::Pidl::IDL;
r20550: Use standard error mechanism in parser. Make sure errors are reported correctly. (This used to be commit 9f794c55a60b09536d6e0dbadfc172e1e74497ff)
2007-01-05 18:20:23 +03:00
$self->YYData->{FILE} = $filename;
r12462: Hide oo magic from callers of the parser (This used to be commit 0c225d2978b11898d334ec2d6da910e0cca0f4cd)
2005-12-25 00:57:51 +03:00
$self->YYData->{INPUT} = $data;
$self->YYData->{LINE} = 0;
$self->YYData->{LAST_TOKEN} = "NONE";
my $idl = $self->YYParse( yylex => \&_Lexer, yyerror => \&_Error );
return CleanData($idl);
}
r19580: Add --includedir option. (This used to be commit b076bfa39fefafcf30ac5be5056abfd629338fa5)
2006-11-06 23:01:22 +03:00
sub parse_file($$)
r12462: Hide oo magic from callers of the parser (This used to be commit 0c225d2978b11898d334ec2d6da910e0cca0f4cd)
2005-12-25 00:57:51 +03:00
{
r19580: Add --includedir option. (This used to be commit b076bfa39fefafcf30ac5be5056abfd629338fa5)
2006-11-06 23:01:22 +03:00
my ($filename,$incdirs) = @_;
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
my $saved_delim = $/;
undef $/;
use the @CPP@ value from configure if possible when determining cpp in pidl (This used to be commit c47e88dabfc62db7533d2d39c7f1010b4188d2ea)
2003-12-16 01:06:18 +03:00
my $cpp = $ENV{CPP};
pidl: sync idl.yp with the CPP changes This is taken from commit 21b0d5e3b10711e6ce3bfad0c1ee2da09a60e232. metze
2011-01-05 11:24:22 +03:00
my $options = "";
use the @CPP@ value from configure if possible when determining cpp in pidl (This used to be commit c47e88dabfc62db7533d2d39c7f1010b4188d2ea)
2003-12-16 01:06:18 +03:00
if (! defined $cpp) {
pidl: sync idl.yp with the CPP changes This is taken from commit 21b0d5e3b10711e6ce3bfad0c1ee2da09a60e232. metze
2011-01-05 11:24:22 +03:00
if (defined $ENV{CC}) {
$cpp = "$ENV{CC}";
$options = "-E";
} else {
$cpp = "cpp";
}
use the @CPP@ value from configure if possible when determining cpp in pidl (This used to be commit c47e88dabfc62db7533d2d39c7f1010b4188d2ea)
2003-12-16 01:06:18 +03:00
}
r19599: Fix --includedir. (This used to be commit 0da2bbcf766dc25805ad583fae185045bb390a5f)
2006-11-07 04:00:27 +03:00
my $includes = join('',map { " -I$_" } @$incdirs);
pidl: sync idl.yp with the CPP changes This is taken from commit 21b0d5e3b10711e6ce3bfad0c1ee2da09a60e232. metze
2011-01-05 11:24:22 +03:00
my $data = `$cpp $options -D__PIDL__$includes -xc "$filename"`;
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
$/ = $saved_delim;
r12464: Add simple IDL parsing tests for pidl using the standard perl testing framework (Test::Simple, distributed with perl itself). Run these tests from 'make test' (This used to be commit 975d8816db6697dab828941b69a740e3a0a2c272)
2005-12-25 02:32:50 +03:00
return parse_string($data, $filename);
re-wrote pidl to use Parse::Yapp instead of Parse::RecDescent, This makes pidl about 3x faster, and also gives us much better error reporting and a more standard grammer definition that will be much easier to code in lex/yacc if we want to do so at a later date. (Parse::Yapp uses essentially the same grammer file as lex/yacc) It also means we no longer need Parse::RecDescent, which should make pidl much more portable. (This used to be commit 4bbaffeb44dca99ad8c0245beb1fddbe01557215)
2003-12-07 16:38:07 +03:00
}
Copy Permalink