asterisk/res/ael/ael.flex

916 lines
25 KiB
Plaintext
Raw Normal View History

/*
* Asterisk -- An open source telephony toolkit.
*
* Copyright (C) 2006, Digium, Inc.
*
* Steve Murphy <murf@parsetree.com>
*
* See http://www.asterisk.org for more information about
* the Asterisk project. Please do not directly contact
* any of the maintainers of this project for assistance;
* the project provides a web site, mailing lists and IRC
* channels for your use.
*
* This program is free software, distributed under the terms of
* the GNU General Public License Version 2. See the LICENSE file
* at the top of the source tree.
*/
/*! \file
*
* \brief Flex scanner description of tokens used in AEL2 .
*
*/
/*
* Start with flex options:
*
* %x describes the contexts we have: paren, semic and argg, plus INITIAL
*/
Merged revisions 141094 via svnmerge from https://origsvn.digium.com/svn/asterisk/branches/1.4 ........ r141094 | murf | 2008-09-04 17:15:07 -0600 (Thu, 04 Sep 2008) | 70 lines (closes issue #13357) Reported by: pj Tested by: murf (closes issue #13416) Reported by: yarns Tested by: murf If you find this message overly verbose, relax, it's probably not meant for you. This message is meant for probably only two people in the whole world: me, or the poor schnook that has to maintain this code because I'm either dead or unavailable at the moment. This fix solves two reports, both having to do with embedding a function call in a ${} construct. It was tricky because the funccall syntax has parenthesis () in it. And up till now, the 'word' token in the flex stuff didn't allow that, because it would tend to steal the LP and RP tokens. To be truthful, the "word" token was the trickiest, most unstable thing in the whole lexer. I was lucky it made this long without complaints. I had to choose every character in the pattern with extreme care, and I knew that someday I'd have to revisit it. Well, the day has come. So, my brilliant idea (and I'm being modest), was to use the surrounding ${} construct to make a state machine and capture everything in it, no matter what it contains. But, I have to now treat the word token like I did with comments, in that I turn the whole thing into a state-machine sort of spec, with new contexts "curlystate", "wordstate", and "brackstate". Wait a minute, "brackstate"? Yes, well, it didn't take very many regression tests to point out if I do this for ${} constructs, I also have to do it with the $[] constructs, too. I had to create a separate pcbstack2 and pcbstack3 because these constructs can occur inside macro argument lists, and when we have two state machines operating on the same structures we'd get problems otherwise. I guess I could have stopped at pcbstack2 and had the brackstate stuff share it, but it doesn't hurt to be safe. So, the pcbpush and pcbpop routines also now have versions for "2" and "3". I had to add the {KEYWORD} construct to the initial pattern for "word", because previously word would match stuff like "default7", because it was a longer match than the keyword "default". But, not any more, because the word pattern only matches only one or two characters now, and it will always lose. So, I made it the winner again by making an optional match on any of the keywords before it's normal pattern. I added another regression test to make sure we don't lose this in future edits, and had to fix just one regression, where it no longer reports a 'cascaded' error, which I guess is a plus. I've given some thought as to whether to apply these fixes to 1.4 and the 1.6.x releases, vs trunk; I decided to put it in 1.4 because one of the bug reports was against 1.4; and it is unexpected that AEL cannot handle this situation. It actually reduced the amount of useless "cascade" error messages that appeared in the regressions (by one line, ehhem). There is a possible side-effect in that it does now do more careful checking of what's in those ${} constructs, as far as matching parens, and brackets are concerned. Some users may find a an insidious problem and correct it this way. This should be exceedingly rare, I hope. ........ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@141115 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2008-09-04 23:31:41 +00:00
%x paren semic argg comment curlystate wordstate brackstate
/* prefix used for various globally-visible functions and variables.
* This renames also yywrap, but since we do not use it, we just
* add option noyywrap to remove it.
*/
%option prefix="ael_yy"
%option noyywrap 8bit
/* I specify this option to suppress flex generating code with ECHO
in it. This generates compiler warnings in some systems; We've
seen the fwrite generate Unused variable warnings with 4.1.2 gcc.
Some systems have tweaked flex ECHO macro to keep the compiler
happy. To keep the warning message from getting output, I added
a default rule at the end of the patterns section */
%option nodefault
/* yyfree normally just frees its arg. It can be null sometimes,
which some systems will complain about, so, we'll define our own version */
%option noyyfree
/* batch gives a bit more performance if we are using it in
* a non-interactive mode. We probably don't care much.
*/
%option batch
/* outfile is the filename to be used instead of lex.yy.c */
%option outfile="ael_lex.c"
/*
* These are not supported in flex 2.5.4, but we need them
* at the moment:
* reentrant produces a thread-safe parser. Not 100% sure that
* we require it, though.
* bison-bridge passes an additional yylval argument to yylex().
* bison-locations is probably not needed.
*/
%option reentrant
%option bison-bridge
%option bison-locations
%{
(closes issue #6002) Reported by: rizzo Tested by: murf Proposal of the changes to be made, and then an announcement of how they were accomplished: http://lists.digium.com/pipermail/asterisk-dev/2008-February/032065.html and: http://lists.digium.com/pipermail/asterisk-dev/2008-March/032124.html Here is a recap, file by file, of what I have done: pbx/pbx_config.c pbx/pbx_ael.c All funcs that were passed a ptr to the context list, now will ALSO be passed a hashtab ptr to the same set. Why? because (for the time being), the dialplan is stored in both, to facilitate a quick, low-cost move to hash-tables to speed up dialplan processing. If it was deemed necessary to pass the context LIST, well, it is just as necessary to have the TABLE available. This is because the list/table in question might not be the global one, but temporary ones we would use to stage the dialplan on, and then swap into the global position when things are ready. We now have one external function for apps to use, "ast_context_find_or_create()" instead of the pre-existing "find" and "create", as all existing usages used both in tandem anyway. pbx_config, and pbx_ael, will stage the reloaded dialplan into local lists and tables, and then call merge_contexts_and_delete, which will merge (now) existing contexts and priorities from other registrars into this local set by copying them. Then, merge_contexts_and_delete will lock down the contexts, swap the lists and tables, and unlock (real quick), and then destroy the old dialplan. chan_sip.c chan_iax.c chan_skinny.c All the channel drivers that would add regcontexts now use the ast_context_find_or_create now. chan_sip also includes a small fix to get rid of warnings about removing priorities that never got entered. apps/app_meetme.c apps/app_dial.c apps/app_queue.c All the apps that added a context/exten/priority were also modified to use ast_context_find_or_create instead. include/asterisk/pbx.h ast_context_create() is removed. Find_or_create_ is the new method. ast_context_find_or_create() interface gets the hashtab added. ast_merge_contexts_and_delete() gets the local hashtab arg added. ast_wrlock_contexts_version() is added so you can detect if someone else got a writelock between your readlocking and writelocking. ast_hashtab_compare_contexts was made public for use in pbx_config/pbx_ael ast_hashtab_hash_contexts was in like fashion make public. include/asterisk/pval.h ast_compile_ael2() interface changed to include the local hashtab table ptr. main/features.c For the sake of the parking context, we use ast_context_find_or_create(). main/pbx.c I changed all the "tree" names to "table" instead. That's because the original implementation was based on binary trees. (had a free library). Then I moved to hashtabs. Now, the names move forward too. refcount field added to contexts, so you can keep track of how many modules wanted this context to exist. Some log messages that are warnings were inflated from LOG_NOTICE to LOG_WARNING. Added some calls to ast_verb(3,...) for debug messages Lots of little mods to ast_context_remove_extension2, which is now excersized in ways it was not previously; one definite bug fixed. find_or_create was upgraded to handle both local lists/tables as well as the globals. context_merge() was added to do the per-context merging of the old/present contexts/extens/prios into the new/proposed local list/tables ast_merge_contexts_and_delete() was heavily modified. ast_add_extension2() was also upgraded to handle changes. the context_destroy() code was re-engineered to handle the new way of doing things, by exten/prio instead of by context. res/ael/pval.c res/ael/ael.tab.c res/ael/ael.tab.h res/ael/ael.y res/ael/ael_lex.c res/ael/ael.flex utils/ael_main.c utils/extconf.c utils/conf2ael.c utils/Makefile Had to change the interface to ast_compile_ael2(), to include the hashtab ptr. This ended up involving several external apps. The main gotcha was I had to include lock.h and hashtab.h in several places. As a side note, I tested this stuff pretty thoroughly, I replicated the problems originally reported by Luigi, and made triply sure that reloads worked, and everything worked thru "stop gracefully". I found a and fixed a few bugs as I was merging into trunk, that did not appear in my tests of bug6002. How's this for verbose commit messages? git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@106757 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2008-03-07 18:57:57 +00:00
#include "asterisk.h"
ASTERISK_FILE_VERSION(__FILE__, "$Revision$")
#include <sys/types.h>
#include <sys/stat.h>
#include <unistd.h>
#include <glob.h>
#if !defined(GLOB_ABORTED)
#define GLOB_ABORTED GLOB_ABEND
#endif
#include "asterisk/logger.h"
#include "asterisk/utils.h"
(closes issue #6002) Reported by: rizzo Tested by: murf Proposal of the changes to be made, and then an announcement of how they were accomplished: http://lists.digium.com/pipermail/asterisk-dev/2008-February/032065.html and: http://lists.digium.com/pipermail/asterisk-dev/2008-March/032124.html Here is a recap, file by file, of what I have done: pbx/pbx_config.c pbx/pbx_ael.c All funcs that were passed a ptr to the context list, now will ALSO be passed a hashtab ptr to the same set. Why? because (for the time being), the dialplan is stored in both, to facilitate a quick, low-cost move to hash-tables to speed up dialplan processing. If it was deemed necessary to pass the context LIST, well, it is just as necessary to have the TABLE available. This is because the list/table in question might not be the global one, but temporary ones we would use to stage the dialplan on, and then swap into the global position when things are ready. We now have one external function for apps to use, "ast_context_find_or_create()" instead of the pre-existing "find" and "create", as all existing usages used both in tandem anyway. pbx_config, and pbx_ael, will stage the reloaded dialplan into local lists and tables, and then call merge_contexts_and_delete, which will merge (now) existing contexts and priorities from other registrars into this local set by copying them. Then, merge_contexts_and_delete will lock down the contexts, swap the lists and tables, and unlock (real quick), and then destroy the old dialplan. chan_sip.c chan_iax.c chan_skinny.c All the channel drivers that would add regcontexts now use the ast_context_find_or_create now. chan_sip also includes a small fix to get rid of warnings about removing priorities that never got entered. apps/app_meetme.c apps/app_dial.c apps/app_queue.c All the apps that added a context/exten/priority were also modified to use ast_context_find_or_create instead. include/asterisk/pbx.h ast_context_create() is removed. Find_or_create_ is the new method. ast_context_find_or_create() interface gets the hashtab added. ast_merge_contexts_and_delete() gets the local hashtab arg added. ast_wrlock_contexts_version() is added so you can detect if someone else got a writelock between your readlocking and writelocking. ast_hashtab_compare_contexts was made public for use in pbx_config/pbx_ael ast_hashtab_hash_contexts was in like fashion make public. include/asterisk/pval.h ast_compile_ael2() interface changed to include the local hashtab table ptr. main/features.c For the sake of the parking context, we use ast_context_find_or_create(). main/pbx.c I changed all the "tree" names to "table" instead. That's because the original implementation was based on binary trees. (had a free library). Then I moved to hashtabs. Now, the names move forward too. refcount field added to contexts, so you can keep track of how many modules wanted this context to exist. Some log messages that are warnings were inflated from LOG_NOTICE to LOG_WARNING. Added some calls to ast_verb(3,...) for debug messages Lots of little mods to ast_context_remove_extension2, which is now excersized in ways it was not previously; one definite bug fixed. find_or_create was upgraded to handle both local lists/tables as well as the globals. context_merge() was added to do the per-context merging of the old/present contexts/extens/prios into the new/proposed local list/tables ast_merge_contexts_and_delete() was heavily modified. ast_add_extension2() was also upgraded to handle changes. the context_destroy() code was re-engineered to handle the new way of doing things, by exten/prio instead of by context. res/ael/pval.c res/ael/ael.tab.c res/ael/ael.tab.h res/ael/ael.y res/ael/ael_lex.c res/ael/ael.flex utils/ael_main.c utils/extconf.c utils/conf2ael.c utils/Makefile Had to change the interface to ast_compile_ael2(), to include the hashtab ptr. This ended up involving several external apps. The main gotcha was I had to include lock.h and hashtab.h in several places. As a side note, I tested this stuff pretty thoroughly, I replicated the problems originally reported by Luigi, and made triply sure that reloads worked, and everything worked thru "stop gracefully". I found a and fixed a few bugs as I was merging into trunk, that did not appear in my tests of bug6002. How's this for verbose commit messages? git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@106757 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2008-03-07 18:57:57 +00:00
#include "asterisk/lock.h"
#include "asterisk/hashtab.h"
#include "ael/ael.tab.h"
#include "asterisk/ael_structs.h"
/*
* A stack to keep track of matching brackets ( [ { } ] )
*/
static char pbcstack[400]; /* XXX missing size checks */
static int pbcpos = 0;
static void pbcpush(char x);
static int pbcpop(char x);
static int parencount = 0;
Merged revisions 141094 via svnmerge from https://origsvn.digium.com/svn/asterisk/branches/1.4 ........ r141094 | murf | 2008-09-04 17:15:07 -0600 (Thu, 04 Sep 2008) | 70 lines (closes issue #13357) Reported by: pj Tested by: murf (closes issue #13416) Reported by: yarns Tested by: murf If you find this message overly verbose, relax, it's probably not meant for you. This message is meant for probably only two people in the whole world: me, or the poor schnook that has to maintain this code because I'm either dead or unavailable at the moment. This fix solves two reports, both having to do with embedding a function call in a ${} construct. It was tricky because the funccall syntax has parenthesis () in it. And up till now, the 'word' token in the flex stuff didn't allow that, because it would tend to steal the LP and RP tokens. To be truthful, the "word" token was the trickiest, most unstable thing in the whole lexer. I was lucky it made this long without complaints. I had to choose every character in the pattern with extreme care, and I knew that someday I'd have to revisit it. Well, the day has come. So, my brilliant idea (and I'm being modest), was to use the surrounding ${} construct to make a state machine and capture everything in it, no matter what it contains. But, I have to now treat the word token like I did with comments, in that I turn the whole thing into a state-machine sort of spec, with new contexts "curlystate", "wordstate", and "brackstate". Wait a minute, "brackstate"? Yes, well, it didn't take very many regression tests to point out if I do this for ${} constructs, I also have to do it with the $[] constructs, too. I had to create a separate pcbstack2 and pcbstack3 because these constructs can occur inside macro argument lists, and when we have two state machines operating on the same structures we'd get problems otherwise. I guess I could have stopped at pcbstack2 and had the brackstate stuff share it, but it doesn't hurt to be safe. So, the pcbpush and pcbpop routines also now have versions for "2" and "3". I had to add the {KEYWORD} construct to the initial pattern for "word", because previously word would match stuff like "default7", because it was a longer match than the keyword "default". But, not any more, because the word pattern only matches only one or two characters now, and it will always lose. So, I made it the winner again by making an optional match on any of the keywords before it's normal pattern. I added another regression test to make sure we don't lose this in future edits, and had to fix just one regression, where it no longer reports a 'cascaded' error, which I guess is a plus. I've given some thought as to whether to apply these fixes to 1.4 and the 1.6.x releases, vs trunk; I decided to put it in 1.4 because one of the bug reports was against 1.4; and it is unexpected that AEL cannot handle this situation. It actually reduced the amount of useless "cascade" error messages that appeared in the regressions (by one line, ehhem). There is a possible side-effect in that it does now do more careful checking of what's in those ${} constructs, as far as matching parens, and brackets are concerned. Some users may find a an insidious problem and correct it this way. This should be exceedingly rare, I hope. ........ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@141115 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2008-09-04 23:31:41 +00:00
/*
* A similar stack to keep track of matching brackets ( [ { } ] ) in word tokens surrounded by ${ ... }
*/
static char pbcstack2[400]; /* XXX missing size checks */
static int pbcpos2 = 0;
static void pbcpush2(char x);
static int pbcpop2(char x);
static int parencount2 = 0;
/*
* A similar stack to keep track of matching brackets ( [ { } ] ) in word tokens surrounded by $[ ... ]
*/
static char pbcstack3[400]; /* XXX missing size checks */
static int pbcpos3 = 0;
static void pbcpush3(char x);
static int pbcpop3(char x);
static int parencount3 = 0;
/*
* current line, column and filename, updated as we read the input.
*/
static int my_lineno = 1; /* current line in the source */
static int my_col = 1; /* current column in the source */
char *my_file = 0; /* used also in the bison code */
char *prev_word; /* XXX document it */
#define MAX_INCLUDE_DEPTH 50
/*
* flex is not too smart, and generates global functions
* without prototypes so the compiler may complain.
* To avoid that, we declare the prototypes here,
* even though these functions are not used.
*/
int ael_yyget_column (yyscan_t yyscanner);
void ael_yyset_column (int column_no , yyscan_t yyscanner);
int ael_yyparse (struct parse_io *);
/*
* A stack to process include files.
* As we switch into the new file we need to store the previous
* state to restore it later.
*/
struct stackelement {
char *fname;
int lineno;
int colno;
glob_t globbuf; /* the current globbuf */
int globbuf_pos; /* where we are in the current globbuf */
YY_BUFFER_STATE bufstate;
};
static struct stackelement include_stack[MAX_INCLUDE_DEPTH];
static int include_stack_index = 0;
static void setup_filestack(char *fnamebuf, int fnamebuf_siz, glob_t *globbuf, int globpos, yyscan_t xscan, int create);
/*
* if we use the @n feature of bison, we must supply the start/end
* location of tokens in the structure pointed by yylloc.
* Simple tokens are just assumed to be on the same line, so
* the line number is constant, and the column is incremented
* by the length of the token.
*/
#ifdef FLEX_BETA /* set for 2.5.33 */
/* compute the total number of lines and columns in the text
* passed as argument.
*/
static void pbcwhere(const char *text, int *line, int *col )
{
int loc_line = *line;
int loc_col = *col;
char c;
while ( (c = *text++) ) {
if ( c == '\t' ) {
loc_col += 8 - (loc_col % 8);
} else if ( c == '\n' ) {
loc_line++;
loc_col = 1;
} else
loc_col++;
}
*line = loc_line;
*col = loc_col;
}
#define STORE_POS do { \
yylloc->first_line = yylloc->last_line = my_lineno; \
yylloc->first_column=my_col; \
yylloc->last_column=my_col+yyleng-1; \
my_col+=yyleng; \
} while (0)
#define STORE_LOC do { \
yylloc->first_line = my_lineno; \
yylloc->first_column=my_col; \
pbcwhere(yytext, &my_lineno, &my_col); \
yylloc->last_line = my_lineno; \
yylloc->last_column = my_col - 1; \
} while (0)
#else
#define STORE_POS
#define STORE_LOC
#endif
%}
Merged revisions 141094 via svnmerge from https://origsvn.digium.com/svn/asterisk/branches/1.4 ........ r141094 | murf | 2008-09-04 17:15:07 -0600 (Thu, 04 Sep 2008) | 70 lines (closes issue #13357) Reported by: pj Tested by: murf (closes issue #13416) Reported by: yarns Tested by: murf If you find this message overly verbose, relax, it's probably not meant for you. This message is meant for probably only two people in the whole world: me, or the poor schnook that has to maintain this code because I'm either dead or unavailable at the moment. This fix solves two reports, both having to do with embedding a function call in a ${} construct. It was tricky because the funccall syntax has parenthesis () in it. And up till now, the 'word' token in the flex stuff didn't allow that, because it would tend to steal the LP and RP tokens. To be truthful, the "word" token was the trickiest, most unstable thing in the whole lexer. I was lucky it made this long without complaints. I had to choose every character in the pattern with extreme care, and I knew that someday I'd have to revisit it. Well, the day has come. So, my brilliant idea (and I'm being modest), was to use the surrounding ${} construct to make a state machine and capture everything in it, no matter what it contains. But, I have to now treat the word token like I did with comments, in that I turn the whole thing into a state-machine sort of spec, with new contexts "curlystate", "wordstate", and "brackstate". Wait a minute, "brackstate"? Yes, well, it didn't take very many regression tests to point out if I do this for ${} constructs, I also have to do it with the $[] constructs, too. I had to create a separate pcbstack2 and pcbstack3 because these constructs can occur inside macro argument lists, and when we have two state machines operating on the same structures we'd get problems otherwise. I guess I could have stopped at pcbstack2 and had the brackstate stuff share it, but it doesn't hurt to be safe. So, the pcbpush and pcbpop routines also now have versions for "2" and "3". I had to add the {KEYWORD} construct to the initial pattern for "word", because previously word would match stuff like "default7", because it was a longer match than the keyword "default". But, not any more, because the word pattern only matches only one or two characters now, and it will always lose. So, I made it the winner again by making an optional match on any of the keywords before it's normal pattern. I added another regression test to make sure we don't lose this in future edits, and had to fix just one regression, where it no longer reports a 'cascaded' error, which I guess is a plus. I've given some thought as to whether to apply these fixes to 1.4 and the 1.6.x releases, vs trunk; I decided to put it in 1.4 because one of the bug reports was against 1.4; and it is unexpected that AEL cannot handle this situation. It actually reduced the amount of useless "cascade" error messages that appeared in the regressions (by one line, ehhem). There is a possible side-effect in that it does now do more careful checking of what's in those ${} constructs, as far as matching parens, and brackets are concerned. Some users may find a an insidious problem and correct it this way. This should be exceedingly rare, I hope. ........ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@141115 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2008-09-04 23:31:41 +00:00
KEYWORD (context|abstract|extend|macro|globals|local|ignorepat|switch|if|ifTime|random|regexten|hint|else|goto|jump|return|break|continue|for|while|case|default|pattern|catch|switches|eswitches|includes)
NOPARENS ([^()\[\]\{\}]|\\[()\[\]\{\}])*
NOARGG ([^(),\{\}\[\]]|\\[,()\[\]\{\}])*
NOSEMIC ([^;()\{\}\[\]]|\\[;()\[\]\{\}])*
HIBIT [\x80-\xff]
%%
\{ { STORE_POS; return LC;}
\} { STORE_POS; return RC;}
\( { STORE_POS; return LP;}
\) { STORE_POS; return RP;}
\; { STORE_POS; return SEMI;}
\= { STORE_POS; return EQ;}
\, { STORE_POS; return COMMA;}
\: { STORE_POS; return COLON;}
\& { STORE_POS; return AMPER;}
\| { STORE_POS; return BAR;}
\=\> { STORE_POS; return EXTENMARK;}
\@ { STORE_POS; return AT;}
\/\/[^\n]* {/*comment*/}
context { STORE_POS; return KW_CONTEXT;}
abstract { STORE_POS; return KW_ABSTRACT;}
extend { STORE_POS; return KW_EXTEND;}
macro { STORE_POS; return KW_MACRO;};
globals { STORE_POS; return KW_GLOBALS;}
local { STORE_POS; return KW_LOCAL;}
ignorepat { STORE_POS; return KW_IGNOREPAT;}
switch { STORE_POS; return KW_SWITCH;}
if { STORE_POS; return KW_IF;}
ifTime { STORE_POS; return KW_IFTIME;}
random { STORE_POS; return KW_RANDOM;}
regexten { STORE_POS; return KW_REGEXTEN;}
hint { STORE_POS; return KW_HINT;}
else { STORE_POS; return KW_ELSE;}
goto { STORE_POS; return KW_GOTO;}
jump { STORE_POS; return KW_JUMP;}
return { STORE_POS; return KW_RETURN;}
break { STORE_POS; return KW_BREAK;}
continue { STORE_POS; return KW_CONTINUE;}
for { STORE_POS; return KW_FOR;}
while { STORE_POS; return KW_WHILE;}
case { STORE_POS; return KW_CASE;}
default { STORE_POS; return KW_DEFAULT;}
pattern { STORE_POS; return KW_PATTERN;}
catch { STORE_POS; return KW_CATCH;}
switches { STORE_POS; return KW_SWITCHES;}
eswitches { STORE_POS; return KW_ESWITCHES;}
includes { STORE_POS; return KW_INCLUDES;}
"/*" { BEGIN(comment); my_col += 2; }
<comment>[^*\n]* { my_col += yyleng; }
<comment>[^*\n]*\n { ++my_lineno; my_col=1;}
<comment>"*"+[^*/\n]* { my_col += yyleng; }
<comment>"*"+[^*/\n]*\n { ++my_lineno; my_col=1;}
Merged revisions 141094 via svnmerge from https://origsvn.digium.com/svn/asterisk/branches/1.4 ........ r141094 | murf | 2008-09-04 17:15:07 -0600 (Thu, 04 Sep 2008) | 70 lines (closes issue #13357) Reported by: pj Tested by: murf (closes issue #13416) Reported by: yarns Tested by: murf If you find this message overly verbose, relax, it's probably not meant for you. This message is meant for probably only two people in the whole world: me, or the poor schnook that has to maintain this code because I'm either dead or unavailable at the moment. This fix solves two reports, both having to do with embedding a function call in a ${} construct. It was tricky because the funccall syntax has parenthesis () in it. And up till now, the 'word' token in the flex stuff didn't allow that, because it would tend to steal the LP and RP tokens. To be truthful, the "word" token was the trickiest, most unstable thing in the whole lexer. I was lucky it made this long without complaints. I had to choose every character in the pattern with extreme care, and I knew that someday I'd have to revisit it. Well, the day has come. So, my brilliant idea (and I'm being modest), was to use the surrounding ${} construct to make a state machine and capture everything in it, no matter what it contains. But, I have to now treat the word token like I did with comments, in that I turn the whole thing into a state-machine sort of spec, with new contexts "curlystate", "wordstate", and "brackstate". Wait a minute, "brackstate"? Yes, well, it didn't take very many regression tests to point out if I do this for ${} constructs, I also have to do it with the $[] constructs, too. I had to create a separate pcbstack2 and pcbstack3 because these constructs can occur inside macro argument lists, and when we have two state machines operating on the same structures we'd get problems otherwise. I guess I could have stopped at pcbstack2 and had the brackstate stuff share it, but it doesn't hurt to be safe. So, the pcbpush and pcbpop routines also now have versions for "2" and "3". I had to add the {KEYWORD} construct to the initial pattern for "word", because previously word would match stuff like "default7", because it was a longer match than the keyword "default". But, not any more, because the word pattern only matches only one or two characters now, and it will always lose. So, I made it the winner again by making an optional match on any of the keywords before it's normal pattern. I added another regression test to make sure we don't lose this in future edits, and had to fix just one regression, where it no longer reports a 'cascaded' error, which I guess is a plus. I've given some thought as to whether to apply these fixes to 1.4 and the 1.6.x releases, vs trunk; I decided to put it in 1.4 because one of the bug reports was against 1.4; and it is unexpected that AEL cannot handle this situation. It actually reduced the amount of useless "cascade" error messages that appeared in the regressions (by one line, ehhem). There is a possible side-effect in that it does now do more careful checking of what's in those ${} constructs, as far as matching parens, and brackets are concerned. Some users may find a an insidious problem and correct it this way. This should be exceedingly rare, I hope. ........ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@141115 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2008-09-04 23:31:41 +00:00
<comment>"*/" { my_col += 2; BEGIN(INITIAL); } /* the nice thing about comments is that you know exactly what ends them */
\n { my_lineno++; my_col = 1; }
[ ]+ { my_col += yyleng; }
[\t]+ { my_col += (yyleng*8)-(my_col%8); }
Merged revisions 141094 via svnmerge from https://origsvn.digium.com/svn/asterisk/branches/1.4 ........ r141094 | murf | 2008-09-04 17:15:07 -0600 (Thu, 04 Sep 2008) | 70 lines (closes issue #13357) Reported by: pj Tested by: murf (closes issue #13416) Reported by: yarns Tested by: murf If you find this message overly verbose, relax, it's probably not meant for you. This message is meant for probably only two people in the whole world: me, or the poor schnook that has to maintain this code because I'm either dead or unavailable at the moment. This fix solves two reports, both having to do with embedding a function call in a ${} construct. It was tricky because the funccall syntax has parenthesis () in it. And up till now, the 'word' token in the flex stuff didn't allow that, because it would tend to steal the LP and RP tokens. To be truthful, the "word" token was the trickiest, most unstable thing in the whole lexer. I was lucky it made this long without complaints. I had to choose every character in the pattern with extreme care, and I knew that someday I'd have to revisit it. Well, the day has come. So, my brilliant idea (and I'm being modest), was to use the surrounding ${} construct to make a state machine and capture everything in it, no matter what it contains. But, I have to now treat the word token like I did with comments, in that I turn the whole thing into a state-machine sort of spec, with new contexts "curlystate", "wordstate", and "brackstate". Wait a minute, "brackstate"? Yes, well, it didn't take very many regression tests to point out if I do this for ${} constructs, I also have to do it with the $[] constructs, too. I had to create a separate pcbstack2 and pcbstack3 because these constructs can occur inside macro argument lists, and when we have two state machines operating on the same structures we'd get problems otherwise. I guess I could have stopped at pcbstack2 and had the brackstate stuff share it, but it doesn't hurt to be safe. So, the pcbpush and pcbpop routines also now have versions for "2" and "3". I had to add the {KEYWORD} construct to the initial pattern for "word", because previously word would match stuff like "default7", because it was a longer match than the keyword "default". But, not any more, because the word pattern only matches only one or two characters now, and it will always lose. So, I made it the winner again by making an optional match on any of the keywords before it's normal pattern. I added another regression test to make sure we don't lose this in future edits, and had to fix just one regression, where it no longer reports a 'cascaded' error, which I guess is a plus. I've given some thought as to whether to apply these fixes to 1.4 and the 1.6.x releases, vs trunk; I decided to put it in 1.4 because one of the bug reports was against 1.4; and it is unexpected that AEL cannot handle this situation. It actually reduced the amount of useless "cascade" error messages that appeared in the regressions (by one line, ehhem). There is a possible side-effect in that it does now do more careful checking of what's in those ${} constructs, as far as matching parens, and brackets are concerned. Some users may find a an insidious problem and correct it this way. This should be exceedingly rare, I hope. ........ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@141115 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2008-09-04 23:31:41 +00:00
({KEYWORD}?[-a-zA-Z0-9'"_/.\<\>\*\+!$#\[\]]|{HIBIT}|(\\.)|(\$\{)|(\$\[)) {
/* boy did I open a can of worms when I changed the lexical token "word".
all the above keywords can be used as a beginning to a "word".-
before, a "word" would match a longer sequence than the above
keywords, and all would be well. But now "word" is a single char
and feeds into a statemachine sort of sequence from there on. So...
I added the {KEYWORD}? to the beginning of the word match sequence */
if (!strcmp(yytext,"${")) {
parencount2 = 0;
pbcpos2 = 0;
pbcpush2('{'); /* push '{' so the last pcbpop (parencount2 = -1) will succeed */
BEGIN(curlystate);
yymore();
} else if (!strcmp(yytext,"$[")) {
parencount3 = 0;
pbcpos3 = 0;
pbcpush3('['); /* push '[' so the last pcbpop (parencount3 = -1) will succeed */
BEGIN(brackstate);
yymore();
} else {
BEGIN(wordstate);
yymore();
}
}
<wordstate>[-a-zA-Z0-9'"_/.\<\>\*\+!$#\[\]] { yymore(); /* Keep going */ }
<wordstate>{HIBIT} { yymore(); /* Keep going */ }
<wordstate>(\\.) { yymore(); /* Keep Going */ }
<wordstate>(\$\{) { /* the beginning of a ${} construct. prepare and pop into curlystate */
parencount2 = 0;
pbcpos2 = 0;
pbcpush2('{'); /* push '{' so the last pcbpop (parencount2 = -1) will succeed */
BEGIN(curlystate);
yymore();
}
<wordstate>(\$\[) { /* the beginning of a $[] construct. prepare and pop into brackstate */
parencount3 = 0;
pbcpos3 = 0;
pbcpush3('['); /* push '[' so the last pcbpop (parencount3 = -1) will succeed */
BEGIN(brackstate);
yymore();
}
<wordstate>([^a-zA-Z0-9\x80-\xff\x2d'"_/.\<\>\*\+!$#\[\]]) {
/* a non-word constituent char, like a space, tab, curly, paren, etc */
char c = yytext[yyleng-1];
STORE_POS;
yylval->str = malloc(yyleng);
strncpy(yylval->str, yytext, yyleng);
yylval->str[yyleng-1] = 0;
Merged revisions 141094 via svnmerge from https://origsvn.digium.com/svn/asterisk/branches/1.4 ........ r141094 | murf | 2008-09-04 17:15:07 -0600 (Thu, 04 Sep 2008) | 70 lines (closes issue #13357) Reported by: pj Tested by: murf (closes issue #13416) Reported by: yarns Tested by: murf If you find this message overly verbose, relax, it's probably not meant for you. This message is meant for probably only two people in the whole world: me, or the poor schnook that has to maintain this code because I'm either dead or unavailable at the moment. This fix solves two reports, both having to do with embedding a function call in a ${} construct. It was tricky because the funccall syntax has parenthesis () in it. And up till now, the 'word' token in the flex stuff didn't allow that, because it would tend to steal the LP and RP tokens. To be truthful, the "word" token was the trickiest, most unstable thing in the whole lexer. I was lucky it made this long without complaints. I had to choose every character in the pattern with extreme care, and I knew that someday I'd have to revisit it. Well, the day has come. So, my brilliant idea (and I'm being modest), was to use the surrounding ${} construct to make a state machine and capture everything in it, no matter what it contains. But, I have to now treat the word token like I did with comments, in that I turn the whole thing into a state-machine sort of spec, with new contexts "curlystate", "wordstate", and "brackstate". Wait a minute, "brackstate"? Yes, well, it didn't take very many regression tests to point out if I do this for ${} constructs, I also have to do it with the $[] constructs, too. I had to create a separate pcbstack2 and pcbstack3 because these constructs can occur inside macro argument lists, and when we have two state machines operating on the same structures we'd get problems otherwise. I guess I could have stopped at pcbstack2 and had the brackstate stuff share it, but it doesn't hurt to be safe. So, the pcbpush and pcbpop routines also now have versions for "2" and "3". I had to add the {KEYWORD} construct to the initial pattern for "word", because previously word would match stuff like "default7", because it was a longer match than the keyword "default". But, not any more, because the word pattern only matches only one or two characters now, and it will always lose. So, I made it the winner again by making an optional match on any of the keywords before it's normal pattern. I added another regression test to make sure we don't lose this in future edits, and had to fix just one regression, where it no longer reports a 'cascaded' error, which I guess is a plus. I've given some thought as to whether to apply these fixes to 1.4 and the 1.6.x releases, vs trunk; I decided to put it in 1.4 because one of the bug reports was against 1.4; and it is unexpected that AEL cannot handle this situation. It actually reduced the amount of useless "cascade" error messages that appeared in the regressions (by one line, ehhem). There is a possible side-effect in that it does now do more careful checking of what's in those ${} constructs, as far as matching parens, and brackets are concerned. Some users may find a an insidious problem and correct it this way. This should be exceedingly rare, I hope. ........ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@141115 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2008-09-04 23:31:41 +00:00
unput(c); /* put this ending char back in the stream */
BEGIN(0);
return word;
}
Merged revisions 141094 via svnmerge from https://origsvn.digium.com/svn/asterisk/branches/1.4 ........ r141094 | murf | 2008-09-04 17:15:07 -0600 (Thu, 04 Sep 2008) | 70 lines (closes issue #13357) Reported by: pj Tested by: murf (closes issue #13416) Reported by: yarns Tested by: murf If you find this message overly verbose, relax, it's probably not meant for you. This message is meant for probably only two people in the whole world: me, or the poor schnook that has to maintain this code because I'm either dead or unavailable at the moment. This fix solves two reports, both having to do with embedding a function call in a ${} construct. It was tricky because the funccall syntax has parenthesis () in it. And up till now, the 'word' token in the flex stuff didn't allow that, because it would tend to steal the LP and RP tokens. To be truthful, the "word" token was the trickiest, most unstable thing in the whole lexer. I was lucky it made this long without complaints. I had to choose every character in the pattern with extreme care, and I knew that someday I'd have to revisit it. Well, the day has come. So, my brilliant idea (and I'm being modest), was to use the surrounding ${} construct to make a state machine and capture everything in it, no matter what it contains. But, I have to now treat the word token like I did with comments, in that I turn the whole thing into a state-machine sort of spec, with new contexts "curlystate", "wordstate", and "brackstate". Wait a minute, "brackstate"? Yes, well, it didn't take very many regression tests to point out if I do this for ${} constructs, I also have to do it with the $[] constructs, too. I had to create a separate pcbstack2 and pcbstack3 because these constructs can occur inside macro argument lists, and when we have two state machines operating on the same structures we'd get problems otherwise. I guess I could have stopped at pcbstack2 and had the brackstate stuff share it, but it doesn't hurt to be safe. So, the pcbpush and pcbpop routines also now have versions for "2" and "3". I had to add the {KEYWORD} construct to the initial pattern for "word", because previously word would match stuff like "default7", because it was a longer match than the keyword "default". But, not any more, because the word pattern only matches only one or two characters now, and it will always lose. So, I made it the winner again by making an optional match on any of the keywords before it's normal pattern. I added another regression test to make sure we don't lose this in future edits, and had to fix just one regression, where it no longer reports a 'cascaded' error, which I guess is a plus. I've given some thought as to whether to apply these fixes to 1.4 and the 1.6.x releases, vs trunk; I decided to put it in 1.4 because one of the bug reports was against 1.4; and it is unexpected that AEL cannot handle this situation. It actually reduced the amount of useless "cascade" error messages that appeared in the regressions (by one line, ehhem). There is a possible side-effect in that it does now do more careful checking of what's in those ${} constructs, as far as matching parens, and brackets are concerned. Some users may find a an insidious problem and correct it this way. This should be exceedingly rare, I hope. ........ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@141115 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2008-09-04 23:31:41 +00:00
<curlystate>{NOPARENS}\} {
if ( pbcpop2('}') ) { /* error */
STORE_LOC;
ast_log(LOG_ERROR,"File=%s, line=%d, column=%d: Mismatched ')' in expression: %s !\n", my_file, my_lineno, my_col, yytext);
BEGIN(0);
yylval->str = malloc(yyleng+1);
strncpy(yylval->str, yytext, yyleng);
yylval->str[yyleng] = 0;
Merged revisions 141094 via svnmerge from https://origsvn.digium.com/svn/asterisk/branches/1.4 ........ r141094 | murf | 2008-09-04 17:15:07 -0600 (Thu, 04 Sep 2008) | 70 lines (closes issue #13357) Reported by: pj Tested by: murf (closes issue #13416) Reported by: yarns Tested by: murf If you find this message overly verbose, relax, it's probably not meant for you. This message is meant for probably only two people in the whole world: me, or the poor schnook that has to maintain this code because I'm either dead or unavailable at the moment. This fix solves two reports, both having to do with embedding a function call in a ${} construct. It was tricky because the funccall syntax has parenthesis () in it. And up till now, the 'word' token in the flex stuff didn't allow that, because it would tend to steal the LP and RP tokens. To be truthful, the "word" token was the trickiest, most unstable thing in the whole lexer. I was lucky it made this long without complaints. I had to choose every character in the pattern with extreme care, and I knew that someday I'd have to revisit it. Well, the day has come. So, my brilliant idea (and I'm being modest), was to use the surrounding ${} construct to make a state machine and capture everything in it, no matter what it contains. But, I have to now treat the word token like I did with comments, in that I turn the whole thing into a state-machine sort of spec, with new contexts "curlystate", "wordstate", and "brackstate". Wait a minute, "brackstate"? Yes, well, it didn't take very many regression tests to point out if I do this for ${} constructs, I also have to do it with the $[] constructs, too. I had to create a separate pcbstack2 and pcbstack3 because these constructs can occur inside macro argument lists, and when we have two state machines operating on the same structures we'd get problems otherwise. I guess I could have stopped at pcbstack2 and had the brackstate stuff share it, but it doesn't hurt to be safe. So, the pcbpush and pcbpop routines also now have versions for "2" and "3". I had to add the {KEYWORD} construct to the initial pattern for "word", because previously word would match stuff like "default7", because it was a longer match than the keyword "default". But, not any more, because the word pattern only matches only one or two characters now, and it will always lose. So, I made it the winner again by making an optional match on any of the keywords before it's normal pattern. I added another regression test to make sure we don't lose this in future edits, and had to fix just one regression, where it no longer reports a 'cascaded' error, which I guess is a plus. I've given some thought as to whether to apply these fixes to 1.4 and the 1.6.x releases, vs trunk; I decided to put it in 1.4 because one of the bug reports was against 1.4; and it is unexpected that AEL cannot handle this situation. It actually reduced the amount of useless "cascade" error messages that appeared in the regressions (by one line, ehhem). There is a possible side-effect in that it does now do more careful checking of what's in those ${} constructs, as far as matching parens, and brackets are concerned. Some users may find a an insidious problem and correct it this way. This should be exceedingly rare, I hope. ........ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@141115 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2008-09-04 23:31:41 +00:00
return word;
}
parencount2--;
if ( parencount2 >= 0) {
yymore();
} else {
BEGIN(wordstate); /* Finished with the current ${} construct. Return to word gathering state */
yymore();
}
}
<curlystate>{NOPARENS}[\(\[\{] {
char c = yytext[yyleng-1];
if (c == '{')
parencount2++;
pbcpush2(c);
yymore();
}
<curlystate>{NOPARENS}[\]\)] {
char c = yytext[yyleng-1];
if ( pbcpop2(c)) { /* error */
STORE_LOC;
ast_log(LOG_ERROR,"File=%s, line=%d, column=%d: Mismatched '%c' in expression!\n",
my_file, my_lineno, my_col, c);
BEGIN(0);
yylval->str = malloc(yyleng+1);
strncpy(yylval->str, yytext, yyleng);
yylval->str[yyleng] = 0;
Merged revisions 141094 via svnmerge from https://origsvn.digium.com/svn/asterisk/branches/1.4 ........ r141094 | murf | 2008-09-04 17:15:07 -0600 (Thu, 04 Sep 2008) | 70 lines (closes issue #13357) Reported by: pj Tested by: murf (closes issue #13416) Reported by: yarns Tested by: murf If you find this message overly verbose, relax, it's probably not meant for you. This message is meant for probably only two people in the whole world: me, or the poor schnook that has to maintain this code because I'm either dead or unavailable at the moment. This fix solves two reports, both having to do with embedding a function call in a ${} construct. It was tricky because the funccall syntax has parenthesis () in it. And up till now, the 'word' token in the flex stuff didn't allow that, because it would tend to steal the LP and RP tokens. To be truthful, the "word" token was the trickiest, most unstable thing in the whole lexer. I was lucky it made this long without complaints. I had to choose every character in the pattern with extreme care, and I knew that someday I'd have to revisit it. Well, the day has come. So, my brilliant idea (and I'm being modest), was to use the surrounding ${} construct to make a state machine and capture everything in it, no matter what it contains. But, I have to now treat the word token like I did with comments, in that I turn the whole thing into a state-machine sort of spec, with new contexts "curlystate", "wordstate", and "brackstate". Wait a minute, "brackstate"? Yes, well, it didn't take very many regression tests to point out if I do this for ${} constructs, I also have to do it with the $[] constructs, too. I had to create a separate pcbstack2 and pcbstack3 because these constructs can occur inside macro argument lists, and when we have two state machines operating on the same structures we'd get problems otherwise. I guess I could have stopped at pcbstack2 and had the brackstate stuff share it, but it doesn't hurt to be safe. So, the pcbpush and pcbpop routines also now have versions for "2" and "3". I had to add the {KEYWORD} construct to the initial pattern for "word", because previously word would match stuff like "default7", because it was a longer match than the keyword "default". But, not any more, because the word pattern only matches only one or two characters now, and it will always lose. So, I made it the winner again by making an optional match on any of the keywords before it's normal pattern. I added another regression test to make sure we don't lose this in future edits, and had to fix just one regression, where it no longer reports a 'cascaded' error, which I guess is a plus. I've given some thought as to whether to apply these fixes to 1.4 and the 1.6.x releases, vs trunk; I decided to put it in 1.4 because one of the bug reports was against 1.4; and it is unexpected that AEL cannot handle this situation. It actually reduced the amount of useless "cascade" error messages that appeared in the regressions (by one line, ehhem). There is a possible side-effect in that it does now do more careful checking of what's in those ${} constructs, as far as matching parens, and brackets are concerned. Some users may find a an insidious problem and correct it this way. This should be exceedingly rare, I hope. ........ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@141115 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2008-09-04 23:31:41 +00:00
return word;
}
yymore();
}
<brackstate>{NOPARENS}\] {
if ( pbcpop3(']') ) { /* error */
STORE_LOC;
ast_log(LOG_ERROR,"File=%s, line=%d, column=%d: Mismatched ')' in expression: %s !\n", my_file, my_lineno, my_col, yytext);
BEGIN(0);
yylval->str = malloc(yyleng+1);
strncpy(yylval->str, yytext, yyleng);
yylval->str[yyleng] = 0;
Merged revisions 141094 via svnmerge from https://origsvn.digium.com/svn/asterisk/branches/1.4 ........ r141094 | murf | 2008-09-04 17:15:07 -0600 (Thu, 04 Sep 2008) | 70 lines (closes issue #13357) Reported by: pj Tested by: murf (closes issue #13416) Reported by: yarns Tested by: murf If you find this message overly verbose, relax, it's probably not meant for you. This message is meant for probably only two people in the whole world: me, or the poor schnook that has to maintain this code because I'm either dead or unavailable at the moment. This fix solves two reports, both having to do with embedding a function call in a ${} construct. It was tricky because the funccall syntax has parenthesis () in it. And up till now, the 'word' token in the flex stuff didn't allow that, because it would tend to steal the LP and RP tokens. To be truthful, the "word" token was the trickiest, most unstable thing in the whole lexer. I was lucky it made this long without complaints. I had to choose every character in the pattern with extreme care, and I knew that someday I'd have to revisit it. Well, the day has come. So, my brilliant idea (and I'm being modest), was to use the surrounding ${} construct to make a state machine and capture everything in it, no matter what it contains. But, I have to now treat the word token like I did with comments, in that I turn the whole thing into a state-machine sort of spec, with new contexts "curlystate", "wordstate", and "brackstate". Wait a minute, "brackstate"? Yes, well, it didn't take very many regression tests to point out if I do this for ${} constructs, I also have to do it with the $[] constructs, too. I had to create a separate pcbstack2 and pcbstack3 because these constructs can occur inside macro argument lists, and when we have two state machines operating on the same structures we'd get problems otherwise. I guess I could have stopped at pcbstack2 and had the brackstate stuff share it, but it doesn't hurt to be safe. So, the pcbpush and pcbpop routines also now have versions for "2" and "3". I had to add the {KEYWORD} construct to the initial pattern for "word", because previously word would match stuff like "default7", because it was a longer match than the keyword "default". But, not any more, because the word pattern only matches only one or two characters now, and it will always lose. So, I made it the winner again by making an optional match on any of the keywords before it's normal pattern. I added another regression test to make sure we don't lose this in future edits, and had to fix just one regression, where it no longer reports a 'cascaded' error, which I guess is a plus. I've given some thought as to whether to apply these fixes to 1.4 and the 1.6.x releases, vs trunk; I decided to put it in 1.4 because one of the bug reports was against 1.4; and it is unexpected that AEL cannot handle this situation. It actually reduced the amount of useless "cascade" error messages that appeared in the regressions (by one line, ehhem). There is a possible side-effect in that it does now do more careful checking of what's in those ${} constructs, as far as matching parens, and brackets are concerned. Some users may find a an insidious problem and correct it this way. This should be exceedingly rare, I hope. ........ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@141115 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2008-09-04 23:31:41 +00:00
return word;
}
parencount3--;
if ( parencount3 >= 0) {
yymore();
} else {
BEGIN(wordstate); /* Finished with the current ${} construct. Return to word gathering state */
yymore();
}
}
<brackstate>{NOPARENS}[\(\[\{] {
char c = yytext[yyleng-1];
if (c == '[')
parencount3++;
pbcpush3(c);
yymore();
}
<brackstate>{NOPARENS}[\}\)] {
char c = yytext[yyleng-1];
if ( pbcpop3(c)) { /* error */
STORE_LOC;
ast_log(LOG_ERROR,"File=%s, line=%d, column=%d: Mismatched '%c' in expression!\n",
my_file, my_lineno, my_col, c);
BEGIN(0);
yylval->str = malloc(yyleng+1);
strncpy(yylval->str, yytext, yyleng);
yylval->str[yyleng] = 0;
Merged revisions 141094 via svnmerge from https://origsvn.digium.com/svn/asterisk/branches/1.4 ........ r141094 | murf | 2008-09-04 17:15:07 -0600 (Thu, 04 Sep 2008) | 70 lines (closes issue #13357) Reported by: pj Tested by: murf (closes issue #13416) Reported by: yarns Tested by: murf If you find this message overly verbose, relax, it's probably not meant for you. This message is meant for probably only two people in the whole world: me, or the poor schnook that has to maintain this code because I'm either dead or unavailable at the moment. This fix solves two reports, both having to do with embedding a function call in a ${} construct. It was tricky because the funccall syntax has parenthesis () in it. And up till now, the 'word' token in the flex stuff didn't allow that, because it would tend to steal the LP and RP tokens. To be truthful, the "word" token was the trickiest, most unstable thing in the whole lexer. I was lucky it made this long without complaints. I had to choose every character in the pattern with extreme care, and I knew that someday I'd have to revisit it. Well, the day has come. So, my brilliant idea (and I'm being modest), was to use the surrounding ${} construct to make a state machine and capture everything in it, no matter what it contains. But, I have to now treat the word token like I did with comments, in that I turn the whole thing into a state-machine sort of spec, with new contexts "curlystate", "wordstate", and "brackstate". Wait a minute, "brackstate"? Yes, well, it didn't take very many regression tests to point out if I do this for ${} constructs, I also have to do it with the $[] constructs, too. I had to create a separate pcbstack2 and pcbstack3 because these constructs can occur inside macro argument lists, and when we have two state machines operating on the same structures we'd get problems otherwise. I guess I could have stopped at pcbstack2 and had the brackstate stuff share it, but it doesn't hurt to be safe. So, the pcbpush and pcbpop routines also now have versions for "2" and "3". I had to add the {KEYWORD} construct to the initial pattern for "word", because previously word would match stuff like "default7", because it was a longer match than the keyword "default". But, not any more, because the word pattern only matches only one or two characters now, and it will always lose. So, I made it the winner again by making an optional match on any of the keywords before it's normal pattern. I added another regression test to make sure we don't lose this in future edits, and had to fix just one regression, where it no longer reports a 'cascaded' error, which I guess is a plus. I've given some thought as to whether to apply these fixes to 1.4 and the 1.6.x releases, vs trunk; I decided to put it in 1.4 because one of the bug reports was against 1.4; and it is unexpected that AEL cannot handle this situation. It actually reduced the amount of useless "cascade" error messages that appeared in the regressions (by one line, ehhem). There is a possible side-effect in that it does now do more careful checking of what's in those ${} constructs, as far as matching parens, and brackets are concerned. Some users may find a an insidious problem and correct it this way. This should be exceedingly rare, I hope. ........ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@141115 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2008-09-04 23:31:41 +00:00
return word;
}
yymore();
}
/*
* context used for arguments of if_head, random_head, switch_head,
* for (last statement), while (XXX why not iftime_head ?).
* End with the matching parentheses.
* A comma at the top level is valid here, unlike in argg where it
* is an argument separator so it must be returned as a token.
*/
<paren>{NOPARENS}\) {
if ( pbcpop(')') ) { /* error */
STORE_LOC;
ast_log(LOG_ERROR,"File=%s, line=%d, column=%d: Mismatched ')' in expression: %s !\n", my_file, my_lineno, my_col, yytext);
BEGIN(0);
yylval->str = malloc(yyleng+1);
strncpy(yylval->str, yytext, yyleng);
yylval->str[yyleng] = 0;
prev_word = 0;
return word;
}
parencount--;
if ( parencount >= 0) {
yymore();
} else {
STORE_LOC;
yylval->str = malloc(yyleng);
strncpy(yylval->str, yytext, yyleng);
yylval->str[yyleng-1] = 0;
unput(')');
BEGIN(0);
return word;
}
}
<paren>{NOPARENS}[\(\[\{] {
char c = yytext[yyleng-1];
if (c == '(')
parencount++;
pbcpush(c);
yymore();
}
<paren>{NOPARENS}[\]\}] {
char c = yytext[yyleng-1];
if ( pbcpop(c)) { /* error */
STORE_LOC;
ast_log(LOG_ERROR,"File=%s, line=%d, column=%d: Mismatched '%c' in expression!\n",
my_file, my_lineno, my_col, c);
BEGIN(0);
yylval->str = malloc(yyleng+1);
strncpy(yylval->str, yytext, yyleng);
yylval->str[yyleng] = 0;
return word;
}
yymore();
}
/*
* handlers for arguments to a macro or application calls.
* We enter this context when we find the initial '(' and
* stay here until we close all matching parentheses,
* and find the comma (argument separator) or the closing ')'
* of the (external) call, which happens when parencount == 0
* before the decrement.
*/
<argg>{NOARGG}[\(\[\{] {
char c = yytext[yyleng-1];
if (c == '(')
parencount++;
pbcpush(c);
yymore();
}
<argg>{NOARGG}\) {
if ( pbcpop(')') ) { /* error */
STORE_LOC;
ast_log(LOG_ERROR,"File=%s, line=%d, column=%d: Mismatched ')' in expression!\n", my_file, my_lineno, my_col);
BEGIN(0);
yylval->str = malloc(yyleng+1);
strncpy(yylval->str, yytext, yyleng);
yylval->str[yyleng] = 0;
return word;
}
parencount--;
if( parencount >= 0){
yymore();
} else {
STORE_LOC;
BEGIN(0);
if ( !strcmp(yytext, ")") )
return RP;
yylval->str = malloc(yyleng);
strncpy(yylval->str, yytext, yyleng);
yylval->str[yyleng-1] = '\0'; /* trim trailing ')' */
unput(')');
return word;
}
}
<argg>{NOARGG}\, {
Merged revisions 162013 via svnmerge from https://origsvn.digium.com/svn/asterisk/branches/1.4 ........ r162013 | murf | 2008-12-09 09:31:55 -0700 (Tue, 09 Dec 2008) | 45 lines (closes issue #14019) Reported by: ckjohnsonme Patches: 14019.diff uploaded by murf (license 17) Tested by: ckjohnsonme, murf This crash was the result of a few small errors that would combine in 64-bit land to result in a crash. 32-bit land might have seen these combine to mysteriously drop the args to an application call, in certain circumstances. Also, in trying to find this bug, I spotted a situation in the flex input, where, in passing back a 'word' to the parser, it would allocate a buffer larger than necessary. I changed the usage in such situations, so that strdup was not used, but rather, an ast_malloc, followed by ast_copy_string. I removed a field from the pval struct, in u2, that was never getting used, and set in one spot in the code. I believe it was an artifact of a previous fix to make switch cases work invisibly with extens. And, for goto's I removed a '!' from before a strcmp, that has been there since the initial merging of AEL2, that might prevent the proper target of a goto from being found. This was pretty harmless on its own, as it would just louse up a consistency check for users. Many thanks to ckjohnsonme for providing a simplified and complete set of information about the bug, that helped considerably in finding and fixing the problem. Now, to get aelparse up and running again in trunk, and out of its "horribly broken" state, so I can run the regression suite! ........ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@162079 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2008-12-09 17:18:03 +00:00
if( parencount != 0) { /* ast_log(LOG_NOTICE,"Folding in a comma!\n"); */
yymore();
} else {
STORE_LOC;
if( !strcmp(yytext,"," ) )
return COMMA;
yylval->str = malloc(yyleng);
strncpy(yylval->str, yytext, yyleng);
yylval->str[yyleng-1] = '\0'; /* trim trailing ',' */
unput(',');
return word;
}
}
<argg>{NOARGG}[\]\}] {
char c = yytext[yyleng-1];
if ( pbcpop(c) ) { /* error */
STORE_LOC;
ast_log(LOG_ERROR,"File=%s, line=%d, column=%d: Mismatched '%c' in expression!\n", my_file, my_lineno, my_col, c);
BEGIN(0);
yylval->str = malloc(yyleng+1);
strncpy(yylval->str, yytext, yyleng);
yylval->str[yyleng] = '\0';
return word;
}
yymore();
}
/*
* context used to find tokens in the right hand side of assignments,
* or in the first and second operand of a 'for'. As above, match
* commas and use ';' as a separator (hence return it as a separate token).
*/
<semic>{NOSEMIC}[\(\[\{] {
char c = yytext[yyleng-1];
yymore();
pbcpush(c);
}
<semic>{NOSEMIC}[\)\]\}] {
char c = yytext[yyleng-1];
if ( pbcpop(c) ) { /* error */
STORE_LOC;
ast_log(LOG_ERROR,"File=%s, line=%d, column=%d: Mismatched '%c' in expression!\n", my_file, my_lineno, my_col, c);
BEGIN(0);
yylval->str = malloc(yyleng+1);
strncpy(yylval->str, yytext, yyleng);
yylval->str[yyleng] = '\0';
return word;
}
yymore();
}
<semic>{NOSEMIC}; {
STORE_LOC;
yylval->str = malloc(yyleng);
strncpy(yylval->str, yytext, yyleng);
yylval->str[yyleng-1] = '\0'; /* trim trailing ';' */
unput(';');
BEGIN(0);
return word;
}
\#include[ \t]+\"[^\"]+\" {
char fnamebuf[1024],*p1,*p2;
int glob_ret;
glob_t globbuf; /* the current globbuf */
int globbuf_pos = -1; /* where we are in the current globbuf */
globbuf.gl_offs = 0; /* initialize it to silence gcc */
p1 = strchr(yytext,'"');
p2 = strrchr(yytext,'"');
if ( include_stack_index >= MAX_INCLUDE_DEPTH ) {
ast_log(LOG_ERROR,"File=%s, line=%d, column=%d: Includes nested too deeply! Wow!!! How did you do that?\n", my_file, my_lineno, my_col);
} else if ( (int)(p2-p1) > sizeof(fnamebuf) - 1 ) {
ast_log(LOG_ERROR,"File=%s, line=%d, column=%d: Filename is incredibly way too long (%d chars!). Inclusion ignored!\n", my_file, my_lineno, my_col, yyleng - 10);
} else {
strncpy(fnamebuf, p1+1, p2-p1-1);
fnamebuf[p2-p1-1] = 0;
if (fnamebuf[0] != '/') {
char fnamebuf2[1024];
snprintf(fnamebuf2,sizeof(fnamebuf2), "%s/%s", (char *)ast_config_AST_CONFIG_DIR, fnamebuf);
ast_copy_string(fnamebuf,fnamebuf2,sizeof(fnamebuf));
}
#ifdef SOLARIS
glob_ret = glob(fnamebuf, GLOB_NOCHECK, NULL, &globbuf);
#else
glob_ret = glob(fnamebuf, GLOB_NOMAGIC|GLOB_BRACE, NULL, &globbuf);
#endif
if (glob_ret == GLOB_NOSPACE) {
ast_log(LOG_WARNING,
"Glob Expansion of pattern '%s' failed: Not enough memory\n", fnamebuf);
} else if (glob_ret == GLOB_ABORTED) {
ast_log(LOG_WARNING,
"Glob Expansion of pattern '%s' failed: Read error\n", fnamebuf);
} else if (glob_ret == GLOB_NOMATCH) {
ast_log(LOG_WARNING,
"Glob Expansion of pattern '%s' failed: No matches!\n", fnamebuf);
} else {
globbuf_pos = 0;
}
}
if (globbuf_pos > -1) {
setup_filestack(fnamebuf, sizeof(fnamebuf), &globbuf, 0, yyscanner, 1);
}
}
<<EOF>> {
char fnamebuf[2048];
if (include_stack_index > 0 && include_stack[include_stack_index-1].globbuf_pos < include_stack[include_stack_index-1].globbuf.gl_pathc-1) {
yy_delete_buffer( YY_CURRENT_BUFFER, yyscanner );
include_stack[include_stack_index-1].globbuf_pos++;
setup_filestack(fnamebuf, sizeof(fnamebuf), &include_stack[include_stack_index-1].globbuf, include_stack[include_stack_index-1].globbuf_pos, yyscanner, 0);
/* finish this */
} else {
if (include_stack[include_stack_index].fname) {
free(include_stack[include_stack_index].fname);
include_stack[include_stack_index].fname = 0;
}
if (my_file) {
free(my_file);
my_file = 0;
}
if ( --include_stack_index < 0 ) {
yyterminate();
} else {
globfree(&include_stack[include_stack_index].globbuf);
include_stack[include_stack_index].globbuf_pos = -1;
yy_delete_buffer( YY_CURRENT_BUFFER, yyscanner );
yy_switch_to_buffer(include_stack[include_stack_index].bufstate, yyscanner );
my_lineno = include_stack[include_stack_index].lineno;
my_col = include_stack[include_stack_index].colno;
my_file = strdup(include_stack[include_stack_index].fname);
}
}
}
<*>.|\n { /* default rule */ ast_log(LOG_ERROR,"Unhandled char(s): %s\n", yytext); }
%%
static void pbcpush(char x)
{
pbcstack[pbcpos++] = x;
}
void ael_yyfree(void *ptr, yyscan_t yyscanner)
{
if (ptr)
free( (char*) ptr );
}
static int pbcpop(char x)
{
if ( ( x == ')' && pbcstack[pbcpos-1] == '(' )
|| ( x == ']' && pbcstack[pbcpos-1] == '[' )
|| ( x == '}' && pbcstack[pbcpos-1] == '{' )) {
pbcpos--;
return 0;
}
return 1; /* error */
}
Merged revisions 141094 via svnmerge from https://origsvn.digium.com/svn/asterisk/branches/1.4 ........ r141094 | murf | 2008-09-04 17:15:07 -0600 (Thu, 04 Sep 2008) | 70 lines (closes issue #13357) Reported by: pj Tested by: murf (closes issue #13416) Reported by: yarns Tested by: murf If you find this message overly verbose, relax, it's probably not meant for you. This message is meant for probably only two people in the whole world: me, or the poor schnook that has to maintain this code because I'm either dead or unavailable at the moment. This fix solves two reports, both having to do with embedding a function call in a ${} construct. It was tricky because the funccall syntax has parenthesis () in it. And up till now, the 'word' token in the flex stuff didn't allow that, because it would tend to steal the LP and RP tokens. To be truthful, the "word" token was the trickiest, most unstable thing in the whole lexer. I was lucky it made this long without complaints. I had to choose every character in the pattern with extreme care, and I knew that someday I'd have to revisit it. Well, the day has come. So, my brilliant idea (and I'm being modest), was to use the surrounding ${} construct to make a state machine and capture everything in it, no matter what it contains. But, I have to now treat the word token like I did with comments, in that I turn the whole thing into a state-machine sort of spec, with new contexts "curlystate", "wordstate", and "brackstate". Wait a minute, "brackstate"? Yes, well, it didn't take very many regression tests to point out if I do this for ${} constructs, I also have to do it with the $[] constructs, too. I had to create a separate pcbstack2 and pcbstack3 because these constructs can occur inside macro argument lists, and when we have two state machines operating on the same structures we'd get problems otherwise. I guess I could have stopped at pcbstack2 and had the brackstate stuff share it, but it doesn't hurt to be safe. So, the pcbpush and pcbpop routines also now have versions for "2" and "3". I had to add the {KEYWORD} construct to the initial pattern for "word", because previously word would match stuff like "default7", because it was a longer match than the keyword "default". But, not any more, because the word pattern only matches only one or two characters now, and it will always lose. So, I made it the winner again by making an optional match on any of the keywords before it's normal pattern. I added another regression test to make sure we don't lose this in future edits, and had to fix just one regression, where it no longer reports a 'cascaded' error, which I guess is a plus. I've given some thought as to whether to apply these fixes to 1.4 and the 1.6.x releases, vs trunk; I decided to put it in 1.4 because one of the bug reports was against 1.4; and it is unexpected that AEL cannot handle this situation. It actually reduced the amount of useless "cascade" error messages that appeared in the regressions (by one line, ehhem). There is a possible side-effect in that it does now do more careful checking of what's in those ${} constructs, as far as matching parens, and brackets are concerned. Some users may find a an insidious problem and correct it this way. This should be exceedingly rare, I hope. ........ git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@141115 65c4cc65-6c06-0410-ace0-fbb531ad65f3
2008-09-04 23:31:41 +00:00
static void pbcpush2(char x)
{
pbcstack2[pbcpos2++] = x;
}
static int pbcpop2(char x)
{
if ( ( x == ')' && pbcstack2[pbcpos2-1] == '(' )
|| ( x == ']' && pbcstack2[pbcpos2-1] == '[' )
|| ( x == '}' && pbcstack2[pbcpos2-1] == '{' )) {
pbcpos2--;
return 0;
}
return 1; /* error */
}
static void pbcpush3(char x)
{
pbcstack3[pbcpos3++] = x;
}
static int pbcpop3(char x)
{
if ( ( x == ')' && pbcstack3[pbcpos3-1] == '(' )
|| ( x == ']' && pbcstack3[pbcpos3-1] == '[' )
|| ( x == '}' && pbcstack3[pbcpos3-1] == '{' )) {
pbcpos3--;
return 0;
}
return 1; /* error */
}
static int c_prevword(void)
{
char *c = prev_word;
if (c == NULL)
return 0;
while ( *c ) {
switch (*c) {
case '{':
case '[':
case '(':
pbcpush(*c);
break;
case '}':
case ']':
case ')':
if (pbcpop(*c))
return 1;
break;
}
c++;
}
return 0;
}
/*
* The following three functions, reset_*, are used in the bison
* code to switch context. As a consequence, we need to
* declare them global and add a prototype so that the
* compiler does not complain.
*
* NOTE: yyg is declared because it is used in the BEGIN macros,
* though that should be hidden as the macro changes
* depending on the flex options that we use - in particular,
* %reentrant changes the way the macro is declared;
* without %reentrant, BEGIN uses yystart instead of yyg
*/
void reset_parencount(yyscan_t yyscanner );
void reset_parencount(yyscan_t yyscanner )
{
struct yyguts_t * yyg = (struct yyguts_t*)yyscanner;
parencount = 0;
pbcpos = 0;
pbcpush('('); /* push '(' so the last pcbpop (parencount= -1) will succeed */
c_prevword();
BEGIN(paren);
}
void reset_semicount(yyscan_t yyscanner );
void reset_semicount(yyscan_t yyscanner )
{
struct yyguts_t * yyg = (struct yyguts_t*)yyscanner;
pbcpos = 0;
BEGIN(semic);
}
void reset_argcount(yyscan_t yyscanner );
void reset_argcount(yyscan_t yyscanner )
{
struct yyguts_t * yyg = (struct yyguts_t*)yyscanner;
parencount = 0;
pbcpos = 0;
pbcpush('('); /* push '(' so the last pcbpop (parencount= -1) will succeed */
c_prevword();
BEGIN(argg);
}
/* used elsewhere, but some local vars */
struct pval *ael2_parse(char *filename, int *errors)
{
struct pval *pvalue;
struct parse_io *io;
char *buffer;
struct stat stats;
FILE *fin;
/* extern int ael_yydebug; */
io = calloc(sizeof(struct parse_io),1);
/* reset the global counters */
prev_word = 0;
my_lineno = 1;
include_stack_index=0;
my_col = 0;
/* ael_yydebug = 1; */
ael_yylex_init(&io->scanner);
fin = fopen(filename,"r");
if ( !fin ) {
ast_log(LOG_ERROR,"File %s could not be opened\n", filename);
*errors = 1;
return 0;
}
if (my_file)
free(my_file);
my_file = strdup(filename);
if (stat(filename, &stats)) {
ast_log(LOG_WARNING, "failed to populate stats from file '%s'\n", filename);
}
buffer = (char*)malloc(stats.st_size+2);
if (fread(buffer, 1, stats.st_size, fin) != stats.st_size) {
ast_log(LOG_ERROR, "fread() failed: %s\n", strerror(errno));
}
buffer[stats.st_size]=0;
fclose(fin);
ael_yy_scan_string (buffer ,io->scanner);
ael_yyset_lineno(1 , io->scanner);
/* ael_yyset_in (fin , io->scanner); OLD WAY */
ael_yyparse(io);
pvalue = io->pval;
*errors = io->syntax_error_count;
ael_yylex_destroy(io->scanner);
free(buffer);
free(io);
return pvalue;
}
static void setup_filestack(char *fnamebuf2, int fnamebuf_siz, glob_t *globbuf, int globpos, yyscan_t yyscanner, int create)
{
struct yyguts_t * yyg = (struct yyguts_t*)yyscanner;
int error, i;
FILE *in1;
char fnamebuf[2048];
if (globbuf && globbuf->gl_pathv && globbuf->gl_pathc > 0)
#if defined(STANDALONE) || defined(LOW_MEMORY) || defined(STANDALONE_AEL)
strncpy(fnamebuf, globbuf->gl_pathv[globpos], fnamebuf_siz);
#else
ast_copy_string(fnamebuf, globbuf->gl_pathv[globpos], fnamebuf_siz);
#endif
else {
ast_log(LOG_ERROR,"Include file name not present!\n");
return;
}
for (i=0; i<include_stack_index; i++) {
if ( !strcmp(fnamebuf,include_stack[i].fname )) {
ast_log(LOG_ERROR,"File=%s, line=%d, column=%d: Nice Try!!! But %s has already been included (perhaps by another file), and would cause an infinite loop of file inclusions!!! Include directive ignored\n",
my_file, my_lineno, my_col, fnamebuf);
break;
}
}
error = 1;
if (i == include_stack_index)
error = 0; /* we can use this file */
if ( !error ) { /* valid file name */
/* relative vs. absolute */
if (fnamebuf[0] != '/')
snprintf(fnamebuf2, fnamebuf_siz, "%s/%s", ast_config_AST_CONFIG_DIR, fnamebuf);
else
#if defined(STANDALONE) || defined(LOW_MEMORY) || defined(STANDALONE_AEL)
strncpy(fnamebuf2, fnamebuf, fnamebuf_siz);
#else
ast_copy_string(fnamebuf2, fnamebuf, fnamebuf_siz);
#endif
in1 = fopen( fnamebuf2, "r" );
if ( ! in1 ) {
ast_log(LOG_ERROR,"File=%s, line=%d, column=%d: Couldn't find the include file: %s; ignoring the Include directive!\n", my_file, my_lineno, my_col, fnamebuf2);
} else {
char *buffer;
struct stat stats;
if (stat(fnamebuf2, &stats)) {
ast_log(LOG_WARNING, "Failed to populate stats from file '%s'\n", fnamebuf2);
}
buffer = (char*)malloc(stats.st_size+1);
if (fread(buffer, 1, stats.st_size, in1) != stats.st_size) {
ast_log(LOG_ERROR, "fread() failed: %s\n", strerror(errno));
}
buffer[stats.st_size] = 0;
ast_log(LOG_NOTICE," --Read in included file %s, %d chars\n",fnamebuf2, (int)stats.st_size);
fclose(in1);
if (include_stack[include_stack_index].fname) {
free(include_stack[include_stack_index].fname);
include_stack[include_stack_index].fname = 0;
}
include_stack[include_stack_index].fname = strdup(S_OR(my_file, "<none>"));
include_stack[include_stack_index].lineno = my_lineno;
include_stack[include_stack_index].colno = my_col+yyleng;
if (my_file)
free(my_file);
my_file = strdup(fnamebuf2);
if (create)
include_stack[include_stack_index].globbuf = *globbuf;
include_stack[include_stack_index].globbuf_pos = 0;
include_stack[include_stack_index].bufstate = YY_CURRENT_BUFFER;
if (create)
include_stack_index++;
yy_switch_to_buffer(ael_yy_scan_string (buffer ,yyscanner),yyscanner);
free(buffer);
my_lineno = 1;
my_col = 1;
BEGIN(INITIAL);
}
}
}