/* +----------------------------------------------------------------------+ | Zend Engine | +----------------------------------------------------------------------+ | Copyright (c) 1998-2014 Zend Technologies Ltd. (http://www.zend.com) | +----------------------------------------------------------------------+ | This source file is subject to version 2.00 of the Zend license, | | that is bundled with this package in the file LICENSE, and is | | available through the world-wide-web at the following url: | | http://www.zend.com/license/2_00.txt. | | If you did not receive a copy of the Zend license and are unable to | | obtain it through the world-wide-web, please send a note to | | license@zend.com so we can mail you a copy immediately. | +----------------------------------------------------------------------+ | Authors: Zeev Suraski | | Jani Taskinen | | Marcus Boerger | | Nuno Lopes | | Scott MacVicar | +----------------------------------------------------------------------+ */ /* $Id: zend_ini_scanner.l,v 1.1.1.4 2014/06/15 20:04:03 misho Exp $ */ #include #include "zend.h" #include "zend_globals.h" #include #include "zend_ini_scanner.h" #if 0 # define YYDEBUG(s, c) printf("state: %d char: %c\n", s, c) #else # define YYDEBUG(s, c) #endif #include "zend_ini_scanner_defs.h" #define YYCTYPE unsigned char /* allow the scanner to read one null byte after the end of the string (from ZEND_MMAP_AHEAD) * so that if will be able to terminate to match the current token (e.g. non-enclosed string) */ #define YYFILL(n) { if (YYCURSOR > YYLIMIT) return 0; } #define YYCURSOR SCNG(yy_cursor) #define YYLIMIT SCNG(yy_limit) #define YYMARKER SCNG(yy_marker) #define YYGETCONDITION() SCNG(yy_state) #define YYSETCONDITION(s) SCNG(yy_state) = s #define STATE(name) yyc##name /* emulate flex constructs */ #define BEGIN(state) YYSETCONDITION(STATE(state)) #define YYSTATE YYGETCONDITION() #define yytext ((char*)SCNG(yy_text)) #define yyleng SCNG(yy_leng) #define yyless(x) do { YYCURSOR = (unsigned char*)yytext + x; \ yyleng = (unsigned int)x; } while(0) /* #define yymore() goto yymore_restart */ /* perform sanity check. If this message is triggered you should increase the ZEND_MMAP_AHEAD value in the zend_streams.h file */ /*!max:re2c */ #if ZEND_MMAP_AHEAD < (YYMAXFILL + 1) # error ZEND_MMAP_AHEAD should be greater than YYMAXFILL #endif /* How it works (for the core ini directives): * =========================================== * * 1. Scanner scans file for tokens and passes them to parser. * 2. Parser parses the tokens and passes the name/value pairs to the callback * function which stores them in the configuration hash table. * 3. Later REGISTER_INI_ENTRIES() is called which triggers the actual * registering of ini entries and uses zend_get_configuration_directive() * to fetch the previously stored name/value pair from configuration hash table * and registers the static ini entries which match the name to the value * into EG(ini_directives) hash table. * 4. PATH section entries are used per-request from down to top, each overriding * previous if one exists. zend_alter_ini_entry() is called for each entry. * Settings in PATH section are ZEND_INI_SYSTEM accessible and thus mimics the * php_admin_* directives used within Apache httpd.conf when PHP is compiled as * module for Apache. * 5. User defined ini files (like .htaccess for apache) are parsed for each request and * stored in separate hash defined by SAPI. */ /* TODO: (ordered by importance :-) * =============================================================================== * * - Separate constant lookup totally from plain strings (using CONSTANT pattern) * - Add #if .. #else .. #endif and ==, !=, <, > , <=, >= operators * - Add #include "some.ini" * - Allow variables to refer to options also when using parse_ini_file() * */ /* Globals Macros */ #define SCNG INI_SCNG #ifdef ZTS ZEND_API ts_rsrc_id ini_scanner_globals_id; #else ZEND_API zend_ini_scanner_globals ini_scanner_globals; #endif /* Eat leading whitespace */ #define EAT_LEADING_WHITESPACE() \ while (yytext[0]) { \ if (yytext[0] == ' ' || yytext[0] == '\t') { \ SCNG(yy_text)++; \ yyleng--; \ } else { \ break; \ } \ } /* Eat trailing whitespace + extra char */ #define EAT_TRAILING_WHITESPACE_EX(ch) \ while (yyleng > 0 && ( \ (ch != 'X' && yytext[yyleng - 1] == ch) || \ yytext[yyleng - 1] == '\n' || \ yytext[yyleng - 1] == '\r' || \ yytext[yyleng - 1] == '\t' || \ yytext[yyleng - 1] == ' ') \ ) { \ yyleng--; \ } /* Eat trailing whitespace */ #define EAT_TRAILING_WHITESPACE() EAT_TRAILING_WHITESPACE_EX('X') #define zend_ini_copy_value(retval, str, len) { \ Z_STRVAL_P(retval) = zend_strndup(str, len); \ Z_STRLEN_P(retval) = len; \ Z_TYPE_P(retval) = IS_STRING; \ } #define RETURN_TOKEN(type, str, len) { \ zend_ini_copy_value(ini_lval, str, len); \ return type; \ } static void _yy_push_state(int new_state TSRMLS_DC) { zend_stack_push(&SCNG(state_stack), (void *) &YYGETCONDITION(), sizeof(int)); YYSETCONDITION(new_state); } #define yy_push_state(state_and_tsrm) _yy_push_state(yyc##state_and_tsrm) static void yy_pop_state(TSRMLS_D) { int *stack_state; zend_stack_top(&SCNG(state_stack), (void **) &stack_state); YYSETCONDITION(*stack_state); zend_stack_del_top(&SCNG(state_stack)); } static void yy_scan_buffer(char *str, unsigned int len TSRMLS_DC) { YYCURSOR = (YYCTYPE*)str; SCNG(yy_start) = YYCURSOR; YYLIMIT = YYCURSOR + len; } #define ini_filename SCNG(filename) /* {{{ init_ini_scanner() */ static int init_ini_scanner(int scanner_mode, zend_file_handle *fh TSRMLS_DC) { /* Sanity check */ if (scanner_mode != ZEND_INI_SCANNER_NORMAL && scanner_mode != ZEND_INI_SCANNER_RAW) { zend_error(E_WARNING, "Invalid scanner mode"); return FAILURE; } SCNG(lineno) = 1; SCNG(scanner_mode) = scanner_mode; SCNG(yy_in) = fh; if (fh != NULL) { ini_filename = zend_strndup(fh->filename, strlen(fh->filename)); } else { ini_filename = NULL; } zend_stack_init(&SCNG(state_stack)); BEGIN(INITIAL); return SUCCESS; } /* }}} */ /* {{{ shutdown_ini_scanner() */ void shutdown_ini_scanner(TSRMLS_D) { zend_stack_destroy(&SCNG(state_stack)); if (ini_filename) { free(ini_filename); } } /* }}} */ /* {{{ zend_ini_scanner_get_lineno() */ int zend_ini_scanner_get_lineno(TSRMLS_D) { return SCNG(lineno); } /* }}} */ /* {{{ zend_ini_scanner_get_filename() */ char *zend_ini_scanner_get_filename(TSRMLS_D) { return ini_filename ? ini_filename : "Unknown"; } /* }}} */ /* {{{ zend_ini_open_file_for_scanning() */ int zend_ini_open_file_for_scanning(zend_file_handle *fh, int scanner_mode TSRMLS_DC) { char *buf; size_t size; if (zend_stream_fixup(fh, &buf, &size TSRMLS_CC) == FAILURE) { return FAILURE; } if (init_ini_scanner(scanner_mode, fh TSRMLS_CC) == FAILURE) { zend_file_handle_dtor(fh TSRMLS_CC); return FAILURE; } yy_scan_buffer(buf, size TSRMLS_CC); return SUCCESS; } /* }}} */ /* {{{ zend_ini_prepare_string_for_scanning() */ int zend_ini_prepare_string_for_scanning(char *str, int scanner_mode TSRMLS_DC) { int len = strlen(str); if (init_ini_scanner(scanner_mode, NULL TSRMLS_CC) == FAILURE) { return FAILURE; } yy_scan_buffer(str, len TSRMLS_CC); return SUCCESS; } /* }}} */ /* {{{ zend_ini_escape_string() */ static void zend_ini_escape_string(zval *lval, char *str, int len, char quote_type TSRMLS_DC) { register char *s, *t; char *end; zend_ini_copy_value(lval, str, len); /* convert escape sequences */ s = t = Z_STRVAL_P(lval); end = s + Z_STRLEN_P(lval); while (s < end) { if (*s == '\\') { s++; if (s >= end) { *t++ = '\\'; continue; } switch (*s) { case '"': if (*s != quote_type) { *t++ = '\\'; *t++ = *s; break; } case '\\': case '$': *t++ = *s; Z_STRLEN_P(lval)--; break; default: *t++ = '\\'; *t++ = *s; break; } } else { *t++ = *s; } if (*s == '\n' || (*s == '\r' && (*(s+1) != '\n'))) { SCNG(lineno)++; } s++; } *t = 0; } /* }}} */ int ini_lex(zval *ini_lval TSRMLS_DC) { restart: SCNG(yy_text) = YYCURSOR; /* yymore_restart: */ /* detect EOF */ if (YYCURSOR >= YYLIMIT) { if (YYSTATE == STATE(ST_VALUE) || YYSTATE == STATE(ST_RAW)) { BEGIN(INITIAL); return END_OF_LINE; } return 0; } /* Eat any UTF-8 BOM we find in the first 3 bytes */ if (YYCURSOR == SCNG(yy_start) && YYCURSOR + 3 < YYLIMIT) { if (memcmp(YYCURSOR, "\xef\xbb\xbf", 3) == 0) { YYCURSOR += 3; goto restart; } } /*!re2c re2c:yyfill:check = 0; LNUM [0-9]+ DNUM ([0-9]*[\.][0-9]+)|([0-9]+[\.][0-9]*) NUMBER [-]?{LNUM}|{DNUM} ANY_CHAR (.|[\n\t]) NEWLINE ("\r"|"\n"|"\r\n") TABS_AND_SPACES [ \t] WHITESPACE [ \t]+ CONSTANT [a-zA-Z_][a-zA-Z0-9_]* LABEL [^=\n\r\t;|&$~(){}!"\[]+ TOKENS [:,.\[\]"'()|^&+-/*=%$!~<>?@{}] OPERATORS [&|~()!] DOLLAR_CURLY "${" SECTION_RAW_CHARS [^\]\n\r] SINGLE_QUOTED_CHARS [^'] RAW_VALUE_CHARS [^\n\r;\000] LITERAL_DOLLAR ("$"([^{\000]|("\\"{ANY_CHAR}))) VALUE_CHARS ([^$= \t\n\r;&|~()!"'\000]|{LITERAL_DOLLAR}) SECTION_VALUE_CHARS ([^$\n\r;"'\]\\]|("\\"{ANY_CHAR})|{LITERAL_DOLLAR}) := yyleng = YYCURSOR - SCNG(yy_text); "[" { /* Section start */ /* Enter section data lookup state */ if (SCNG(scanner_mode) == ZEND_INI_SCANNER_RAW) { yy_push_state(ST_SECTION_RAW TSRMLS_CC); } else { yy_push_state(ST_SECTION_VALUE TSRMLS_CC); } return TC_SECTION; } "'"{SINGLE_QUOTED_CHARS}+"'" { /* Raw string */ /* Eat leading and trailing single quotes */ if (yytext[0] == '\'' && yytext[yyleng - 1] == '\'') { SCNG(yy_text)++; yyleng = yyleng - 2; } RETURN_TOKEN(TC_RAW, yytext, yyleng); } "]"{TABS_AND_SPACES}*{NEWLINE}? { /* End of section */ BEGIN(INITIAL); SCNG(lineno)++; return ']'; } {LABEL}"["{TABS_AND_SPACES}* { /* Start of option with offset */ /* Eat leading whitespace */ EAT_LEADING_WHITESPACE(); /* Eat trailing whitespace and [ */ EAT_TRAILING_WHITESPACE_EX('['); /* Enter offset lookup state */ yy_push_state(ST_OFFSET TSRMLS_CC); RETURN_TOKEN(TC_OFFSET, yytext, yyleng); } {TABS_AND_SPACES}*"]" { /* End of section or an option offset */ BEGIN(INITIAL); return ']'; } {DOLLAR_CURLY} { /* Variable start */ yy_push_state(ST_VARNAME TSRMLS_CC); return TC_DOLLAR_CURLY; } {LABEL} { /* Variable name */ /* Eat leading whitespace */ EAT_LEADING_WHITESPACE(); /* Eat trailing whitespace */ EAT_TRAILING_WHITESPACE(); RETURN_TOKEN(TC_VARNAME, yytext, yyleng); } "}" { /* Variable end */ yy_pop_state(TSRMLS_C); return '}'; } ("true"|"on"|"yes"){TABS_AND_SPACES}* { /* TRUE value (when used outside option value/offset this causes parse error!) */ RETURN_TOKEN(BOOL_TRUE, "1", 1); } ("false"|"off"|"no"|"none"|"null"){TABS_AND_SPACES}* { /* FALSE value (when used outside option value/offset this causes parse error!)*/ RETURN_TOKEN(BOOL_FALSE, "", 0); } {LABEL} { /* Get option name */ /* Eat leading whitespace */ EAT_LEADING_WHITESPACE(); /* Eat trailing whitespace */ EAT_TRAILING_WHITESPACE(); RETURN_TOKEN(TC_LABEL, yytext, yyleng); } {TABS_AND_SPACES}*[=]{TABS_AND_SPACES}* { /* Start option value */ if (SCNG(scanner_mode) == ZEND_INI_SCANNER_RAW) { yy_push_state(ST_RAW TSRMLS_CC); } else { yy_push_state(ST_VALUE TSRMLS_CC); } return '='; } {RAW_VALUE_CHARS} { /* Raw value, only used when SCNG(scanner_mode) == ZEND_INI_SCANNER_RAW. */ char *sc = NULL; while (YYCURSOR < YYLIMIT) { switch (*YYCURSOR) { case '\n': case '\r': goto end_raw_value_chars; break; case ';': if (sc == NULL) { sc = YYCURSOR; } /* no break */ default: YYCURSOR++; break; } } end_raw_value_chars: yyleng = YYCURSOR - SCNG(yy_text); /* Eat trailing semicolons */ while (yytext[yyleng - 1] == ';') { yyleng--; } /* Eat leading and trailing double quotes */ if (yytext[0] == '"' && yytext[yyleng - 1] == '"') { SCNG(yy_text)++; yyleng = yyleng - 2; } else if (sc) { YYCURSOR = sc; yyleng = YYCURSOR - SCNG(yy_text); } RETURN_TOKEN(TC_RAW, yytext, yyleng); } {SECTION_RAW_CHARS}+ { /* Raw value, only used when SCNG(scanner_mode) == ZEND_INI_SCANNER_RAW. */ RETURN_TOKEN(TC_RAW, yytext, yyleng); } {TABS_AND_SPACES}*{NEWLINE} { /* End of option value */ BEGIN(INITIAL); SCNG(lineno)++; return END_OF_LINE; } {CONSTANT} { /* Get constant option value */ RETURN_TOKEN(TC_CONSTANT, yytext, yyleng); } {NUMBER} { /* Get number option value as string */ RETURN_TOKEN(TC_NUMBER, yytext, yyleng); } {TOKENS} { /* Disallow these chars outside option values */ return yytext[0]; } {OPERATORS}{TABS_AND_SPACES}* { /* Boolean operators */ return yytext[0]; } [=] { /* Make = used in option value to trigger error */ yyless(0); BEGIN(INITIAL); return END_OF_LINE; } {VALUE_CHARS}+ { /* Get everything else as option/offset value */ RETURN_TOKEN(TC_STRING, yytext, yyleng); } {SECTION_VALUE_CHARS}+ { /* Get rest as section/offset value */ RETURN_TOKEN(TC_STRING, yytext, yyleng); } {TABS_AND_SPACES}*["] { /* Double quoted '"' string start */ yy_push_state(ST_DOUBLE_QUOTES TSRMLS_CC); return '"'; } ["]{TABS_AND_SPACES}* { /* Double quoted '"' string ends */ yy_pop_state(TSRMLS_C); return '"'; } [^] { /* Escape double quoted string contents */ if (YYCURSOR > YYLIMIT) { return 0; } while (YYCURSOR < YYLIMIT) { switch (*YYCURSOR++) { case '"': if (YYCURSOR < YYLIMIT && YYCURSOR[-2] == '\\' && *YYCURSOR != '\r' && *YYCURSOR != '\n') { continue; } break; case '$': if (*YYCURSOR == '{') { break; } continue; case '\\': if (YYCURSOR < YYLIMIT && *YYCURSOR != '"') { YYCURSOR++; } /* fall through */ default: continue; } YYCURSOR--; break; } yyleng = YYCURSOR - SCNG(yy_text); zend_ini_escape_string(ini_lval, yytext, yyleng, '"' TSRMLS_CC); return TC_QUOTED_STRING; } {WHITESPACE} { RETURN_TOKEN(TC_WHITESPACE, yytext, yyleng); } {TABS_AND_SPACES}+ { /* eat whitespace */ goto restart; } {TABS_AND_SPACES}*{NEWLINE} { SCNG(lineno)++; return END_OF_LINE; } {TABS_AND_SPACES}*[;][^\r\n]*{NEWLINE} { /* Comment */ BEGIN(INITIAL); SCNG(lineno)++; return END_OF_LINE; } {TABS_AND_SPACES}*[#][^\r\n]*{NEWLINE} { /* #Comment */ zend_error(E_DEPRECATED, "Comments starting with '#' are deprecated in %s on line %d", zend_ini_scanner_get_filename(TSRMLS_C), SCNG(lineno)); BEGIN(INITIAL); SCNG(lineno)++; return END_OF_LINE; } [^] { /* End of option value (if EOF is reached before EOL */ BEGIN(INITIAL); return END_OF_LINE; } <*>[^] { return 0; } */ }