X-Git-Url: https://git.dlugolecki.net.pl/?a=blobdiff_plain;f=gedcom%2Fgedcom_lex_common.c;h=779af3bde742a0ab1366c89ecad485000d40b956;hb=809c37c2f87962cbb9fcb84c851d961e276fef72;hp=d0d329cab5e031d81a20bcab027e8bed6384ada1;hpb=2468a5aefafdfbd1982fe3214e3866e072892506;p=gedcom-parse.git diff --git a/gedcom/gedcom_lex_common.c b/gedcom/gedcom_lex_common.c index d0d329c..779af3b 100644 --- a/gedcom/gedcom_lex_common.c +++ b/gedcom/gedcom_lex_common.c @@ -35,14 +35,17 @@ static int current_level = -1; static int level_diff=MAXGEDCLEVEL; static size_t line_len = 0; -static char ptr_buf[MAXGEDCPTRLEN * UTF_FACTOR + 1]; -static char tag_buf[MAXGEDCTAGLEN * UTF_FACTOR + 1]; -static char str_buf[MAXGEDCLINELEN * UTF_FACTOR + 1]; +static struct conv_buffer* ptr_buffer = NULL; +static struct conv_buffer* tag_buffer = NULL; +static struct conv_buffer* str_buffer = NULL; + +#define INITIAL_PTR_BUFFER_LEN MAXGEDCPTRLEN * UTF_FACTOR + 1 +#define INITIAL_TAG_BUFFER_LEN MAXGEDCTAGLEN * UTF_FACTOR + 1 +#define INITIAL_STR_BUFFER_LEN MAXGEDCLINELEN * UTF_FACTOR + 1 #ifdef LEXER_TEST YYSTYPE gedcom_lval; int line_no = 1; -int compat_at = 0; int gedcom_lex(); @@ -136,10 +139,14 @@ static void error_unexpected_character(const char* str, char ch) gedcom_error(_("Unexpected character: '%s' (0x%02x)"), str, ch); } +/* This is to bypass the iconv conversion (if the input is UTF-8 coming + from the program) */ +static int dummy_conv = 0; + #elif LEX_SECTION == 2 #define TO_INTERNAL(STR,OUTBUF) \ - to_internal(STR, yyleng, OUTBUF, sizeof(OUTBUF)) + (dummy_conv ? STR : to_internal(STR, yyleng, OUTBUF)) #define INIT_LINE_LEN \ line_len = 0; @@ -157,9 +164,10 @@ static void error_unexpected_character(const char* str, char ch) #define MKTAGACTION(THETAG) \ { CHECK_LINE_LEN; \ - gedcom_lval.tag.string = TO_INTERNAL(yytext, tag_buf); \ + gedcom_lval.tag.string = TO_INTERNAL(yytext, tag_buffer); \ gedcom_lval.tag.value = TAG_##THETAG; \ BEGIN(NORMAL); \ + line_no++; \ return TAG_##THETAG; \ } @@ -227,10 +235,11 @@ static void error_unexpected_character(const char* str, char ch) #define ACTION_DIGITS \ - { int level = atoi(TO_INTERNAL(yytext, str_buf)); \ + { int level = atoi(TO_INTERNAL(yytext, str_buffer)); \ CHECK_LINE_LEN; \ if ((level < 0) || (level > MAXGEDCLEVEL)) { \ error_level_out_of_range(); \ + line_no++; \ return BADTOKEN; \ } \ level_diff = level - current_level; \ @@ -248,6 +257,7 @@ static void error_unexpected_character(const char* str, char ch) else { \ /* should never happen (error to GEDCOM spec) */ \ error_level_too_high(level_diff); \ + line_no++; \ return BADTOKEN; \ } \ } @@ -256,19 +266,21 @@ static void error_unexpected_character(const char* str, char ch) #define ACTION_ALPHANUM \ { if (strlen(yytext) > MAXGEDCTAGLEN * encoding_width) { \ error_tag_too_long(yytext); \ + line_no++; \ return BADTOKEN; \ } \ CHECK_LINE_LEN; \ - gedcom_lval.tag.string = TO_INTERNAL(yytext, tag_buf); \ + gedcom_lval.tag.string = TO_INTERNAL(yytext, tag_buffer); \ gedcom_lval.tag.value = USERTAG; \ BEGIN(NORMAL); \ + line_no++; \ return USERTAG; \ } #define ACTION_DELIM \ { CHECK_LINE_LEN; \ - gedcom_lval.string = TO_INTERNAL(yytext, str_buf); \ + gedcom_lval.string = TO_INTERNAL(yytext, str_buffer); \ return DELIM; \ } @@ -276,7 +288,7 @@ static void error_unexpected_character(const char* str, char ch) #define ACTION_ANY \ { char* tmp; \ CHECK_LINE_LEN; \ - tmp = TO_INTERNAL(yytext, str_buf); \ + tmp = TO_INTERNAL(yytext, str_buffer); \ if (!tmp) { \ /* Something went wrong during conversion... */ \ error_invalid_character(yytext, yytext[0]); \ @@ -298,7 +310,7 @@ static void error_unexpected_character(const char* str, char ch) #define ACTION_ESCAPE \ { CHECK_LINE_LEN; \ - gedcom_lval.string = TO_INTERNAL(yytext, str_buf); \ + gedcom_lval.string = TO_INTERNAL(yytext, str_buffer); \ return ESCAPE; \ } @@ -309,7 +321,7 @@ static void error_unexpected_character(const char* str, char ch) error_pointer_too_long(yytext); \ return BADTOKEN; \ } \ - gedcom_lval.string = TO_INTERNAL(yytext, ptr_buf); \ + gedcom_lval.string = TO_INTERNAL(yytext, ptr_buffer); \ return POINTER; \ } @@ -323,7 +335,6 @@ static void error_unexpected_character(const char* str, char ch) #define ACTION_TERMINATOR \ { CHECK_LINE_LEN; \ INIT_LINE_LEN; \ - line_no++; \ BEGIN(INITIAL); \ } @@ -349,7 +360,7 @@ static void error_unexpected_character(const char* str, char ch) } #define ACTION_NORMAL_AT \ - { if (compat_at) { \ + { if (compat_mode(C_NO_DOUBLE_AT)) { \ int i, j; \ char *yycopy = strdup(yytext); \ if (yycopy) { \ @@ -380,11 +391,28 @@ int yywrap() return 1; } +static void free_conv_buffers() +{ + free_conv_buffer(ptr_buffer); + free_conv_buffer(tag_buffer); + free_conv_buffer(str_buffer); +} + static void yylex_cleanup() { /* fix memory leak in lex */ yy_delete_buffer(yy_current_buffer); yy_current_buffer = NULL; + free_conv_buffers(); +} + +static void init_conv_buffers() +{ + if (!ptr_buffer) { + ptr_buffer = create_conv_buffer(INITIAL_PTR_BUFFER_LEN); + tag_buffer = create_conv_buffer(INITIAL_TAG_BUFFER_LEN); + str_buffer = create_conv_buffer(INITIAL_STR_BUFFER_LEN); + } } static int exitfuncregistered = 0; @@ -393,11 +421,13 @@ void yymyinit(FILE *f) { if (! exitfuncregistered && atexit(yylex_cleanup) == 0) exitfuncregistered = 1; + init_conv_buffers(); yyin = f; yyrestart(f); /* Reset our state */ current_level = -1; level_diff = MAXGEDCLEVEL; + BEGIN(INITIAL); } #endif