/** 先頭パラメータの発見 @param[in] end buf末尾 @param[in] p 解析の現在位置 パラメータが0個と1個以上の判別のために状態を設けている. */ const wchar_t* COutlineErlang::ScanArgs1( const wchar_t* end, const wchar_t* p ) { assert( m_state == STATE_FUNC_ARGS1 ); while( IS_SPACE( *p ) && p < end ) p++; if( p >= end ) return end; if( *p == /* ( */ L')' ){ // no argument m_state = STATE_FUNC_ARGS_FIN; p++; } else if( IS_COMMENT( *p )){ return end; } else { // argument found m_state = STATE_FUNC_ARGS; ++m_argcount; } return p; }
/** パラメータの発見 @param[in] end buf末尾 @param[in] p 解析の現在位置 関数名の取得が完了し,パラメータ先頭の括弧を探す. */ const wchar_t* COutlineErlang::EnterArgs( const wchar_t* end, const wchar_t* p ) { assert( m_state == STATE_FUNC_CANDIDATE_FIN ); while( IS_SPACE( *p ) && p < end ) p++; if( p >= end ) return end; if( IS_COMMENT( *p )){ return end; } else if( *p == L'(' ){ // ) m_state = STATE_FUNC_ARGS1; m_argcount = 0; m_parenthesis_ptr = 1; m_parenthesis[0] = *p; ++p; return p; } // not a function m_state = STATE_NORMAL; return end; }
static void clear_comments(CONF *conf, char *p) { for (;;) { if (IS_FCOMMENT(conf,*p)) { *p='\0'; return; } if (!IS_WS(conf,*p)) { break; } p++; } for (;;) { if (IS_COMMENT(conf,*p)) { *p='\0'; return; } if (IS_DQUOTE(conf,*p)) { p=scan_dquote(conf, p); continue; } if (IS_QUOTE(conf,*p)) { p=scan_quote(conf, p); continue; } if (IS_ESC(conf,*p)) { p=scan_esc(conf,p); continue; } if (IS_EOF(conf,*p)) return; else p++; } }
/** 関数本体の区切り,または条件文の検出 @param[in] end buf末尾 @param[in] p 解析の現在位置 パラメータ本体を表す記号(->)か条件文の開始キーワード(when)を 見つけたら,関数発見とする. それ以外の場合は関数ではなかったと考える. */ const wchar_t* COutlineErlang::EnterCond( const wchar_t* end, const wchar_t* p ) { while( IS_SPACE( *p ) && p < end ) p++; if( p >= end ) return end; if( p + 1 < end && wcsncmp( p, L"->", 2 ) == 0){ p += 2; m_state = STATE_FUNC_FOUND; } else if( p + 3 < end && wcsncmp( p, L"when", 4 ) == 0 ){ m_state = STATE_FUNC_FOUND; p += 4; } else if( IS_COMMENT( *p )){ return end; } else { m_state = STATE_NORMAL; } return end; }
/** * Check hash sums in a hash file. * Lines beginning with ';' and '#' are ignored. * * @param hash_file_path - the path of the file with hash sums to verify. * @param chdir - true if function should emulate chdir to directory of filepath before checking it. * @return zero on success, -1 on fail */ int check_hash_file(file_t* file, int chdir) { FILE *fd; char buf[2048]; size_t pos; const char *ralign; timedelta_t timer; struct file_info info; const char* hash_file_path = file->path; int res = 0, line_num = 0; double time; /* process --check-embedded option */ if(opt.mode & MODE_CHECK_EMBEDDED) { unsigned crc32_be; if(find_embedded_crc32(hash_file_path, &crc32_be)) { /* initialize file_info structure */ memset(&info, 0, sizeof(info)); info.full_path = rsh_strdup(hash_file_path); info.file = file; file_info_set_print_path(&info, info.full_path); info.sums_flags = info.hc.hash_mask = RHASH_CRC32; info.hc.flags = HC_HAS_EMBCRC32; info.hc.embedded_crc32_be = crc32_be; res = verify_sums(&info); fflush(rhash_data.out); if(!rhash_data.interrupted) { if(res == 0) rhash_data.ok++; else if(res == -1 && errno == ENOENT) rhash_data.miss++; rhash_data.processed++; } free(info.full_path); file_info_destroy(&info); } else { log_warning(_("file name doesn't contain a CRC32: %s\n"), hash_file_path); return -1; } return 0; } /* initialize statistics */ rhash_data.processed = rhash_data.ok = rhash_data.miss = 0; rhash_data.total_size = 0; if(file->mode & FILE_IFSTDIN) { fd = stdin; hash_file_path = "<stdin>"; } else if( !(fd = rsh_fopen_bin(hash_file_path, "rb") )) { log_file_error(hash_file_path); return -1; } pos = strlen(hash_file_path)+16; ralign = str_set(buf, '-', (pos < 80 ? 80 - (int)pos : 2)); fprintf(rhash_data.out, _("\n--( Verifying %s )%s\n"), hash_file_path, ralign); fflush(rhash_data.out); rhash_timer_start(&timer); /* mark the directory part of the path, by setting the pos index */ if(chdir) { pos = strlen(hash_file_path); for(; pos > 0 && !IS_PATH_SEPARATOR(hash_file_path[pos]); pos--); if(IS_PATH_SEPARATOR(hash_file_path[pos])) pos++; } else pos = 0; /* read crc file line by line */ for(line_num = 0; fgets(buf, 2048, fd); line_num++) { char* line = buf; char* path_without_ext = NULL; /* skip unicode BOM */ if(line_num == 0 && buf[0] == (char)0xEF && buf[1] == (char)0xBB && buf[2] == (char)0xBF) line += 3; if(*line == 0) continue; /* skip empty lines */ if(is_binary_string(line)) { log_error(_("file is binary: %s\n"), hash_file_path); if(fd != stdin) fclose(fd); return -1; } /* skip comments and empty lines */ if(IS_COMMENT(*line) || *line == '\r' || *line == '\n') continue; memset(&info, 0, sizeof(info)); if(!hash_check_parse_line(line, &info.hc, !feof(fd))) continue; if(info.hc.hash_mask == 0) continue; info.print_path = info.hc.file_path; info.sums_flags = info.hc.hash_mask; /* see if crc file contains a hash sum without a filename */ if(info.print_path == NULL) { char* point; path_without_ext = rsh_strdup(hash_file_path); point = strrchr(path_without_ext, '.'); if(point) { *point = '\0'; file_info_set_print_path(&info, path_without_ext); } } if(info.print_path != NULL) { file_t file_to_check; int is_absolute = IS_PATH_SEPARATOR(info.print_path[0]); IF_WINDOWS(is_absolute = is_absolute || (info.print_path[0] && info.print_path[1] == ':')); /* if filename shall be prepended by a directory path */ if(pos && !is_absolute) { size_t len = strlen(info.print_path); info.full_path = (char*)rsh_malloc(pos + len + 1); memcpy(info.full_path, hash_file_path, pos); strcpy(info.full_path + pos, info.print_path); } else { info.full_path = rsh_strdup(info.print_path); } memset(&file_to_check, 0, sizeof(file_t)); file_to_check.path = info.full_path; rsh_file_stat(&file_to_check); info.file = &file_to_check; /* verify hash sums of the file */ res = verify_sums(&info); fflush(rhash_data.out); rsh_file_cleanup(&file_to_check); file_info_destroy(&info); if(rhash_data.interrupted) { free(path_without_ext); break; } /* update statistics */ if(res == 0) rhash_data.ok++; else if(res == -1 && errno == ENOENT) rhash_data.miss++; rhash_data.processed++; } free(path_without_ext); } time = rhash_timer_stop(&timer); fprintf(rhash_data.out, "%s\n", str_set(buf, '-', 80)); print_check_stats(); if(rhash_data.processed != rhash_data.ok) rhash_data.error_flag = 1; if(opt.flags & OPT_SPEED && rhash_data.processed > 1) { print_time_stats(time, rhash_data.total_size, 1); } rhash_data.processed = 0; res = ferror(fd); /* check that crc file has been read without errors */ if(fd != stdin) fclose(fd); return (res == 0 ? 0 : -1); }
int do_scrub_chars (int (*get) (char *, int), char *tostart, int tolen) { char *to = tostart; char *toend = tostart + tolen; char *from; char *fromend; int fromlen; register int ch, ch2 = 0; /* Character that started the string we're working on. */ static char quotechar; /*State 0: beginning of normal line 1: After first whitespace on line (flush more white) 2: After first non-white (opcode) on line (keep 1white) 3: after second white on line (into operands) (flush white) 4: after putting out a .linefile, put out digits 5: parsing a string, then go to old-state 6: putting out \ escape in a "d string. 7: no longer used 8: no longer used 9: After seeing symbol char in state 3 (keep 1white after symchar) 10: After seeing whitespace in state 9 (keep white before symchar) 11: After seeing a symbol character in state 0 (eg a label definition) -1: output string in out_string and go to the state in old_state -2: flush text until a '*' '/' is seen, then go to state old_state #ifdef TC_V850 12: After seeing a dash, looking for a second dash as a start of comment. #endif #ifdef DOUBLEBAR_PARALLEL 13: After seeing a vertical bar, looking for a second vertical bar as a parallel expression separator. #endif #ifdef TC_IA64 14: After seeing a `(' at state 0, looking for a `)' as predicate. 15: After seeing a `(' at state 1, looking for a `)' as predicate. #endif #ifdef TC_Z80 16: After seeing an 'a' or an 'A' at the start of a symbol 17: After seeing an 'f' or an 'F' in state 16 #endif */ /* I added states 9 and 10 because the MIPS ECOFF assembler uses constructs like ``.loc 1 20''. This was turning into ``.loc 120''. States 9 and 10 ensure that a space is never dropped in between characters which could appear in an identifier. Ian Taylor, [email protected]. I added state 11 so that something like "Lfoo add %r25,%r26,%r27" works correctly on the PA (and any other target where colons are optional). Jeff Law, [email protected]. I added state 13 so that something like "cmp r1, r2 || trap #1" does not get squashed into "cmp r1,r2||trap#1", with the all important space between the 'trap' and the '#1' being eliminated. [email protected] */ /* This macro gets the next input character. */ #define GET() \ (from < fromend \ ? * (unsigned char *) (from++) \ : (saved_input = NULL, \ fromlen = (*get) (input_buffer, sizeof input_buffer), \ from = input_buffer, \ fromend = from + fromlen, \ (fromlen == 0 \ ? EOF \ : * (unsigned char *) (from++)))) /* This macro pushes a character back on the input stream. */ #define UNGET(uch) (*--from = (uch)) /* This macro puts a character into the output buffer. If this character fills the output buffer, this macro jumps to the label TOFULL. We use this rather ugly approach because we need to handle two different termination conditions: EOF on the input stream, and a full output buffer. It would be simpler if we always read in the entire input stream before processing it, but I don't want to make such a significant change to the assembler's memory usage. */ #define PUT(pch) \ do \ { \ *to++ = (pch); \ if (to >= toend) \ goto tofull; \ } \ while (0) if (saved_input != NULL) { from = saved_input; fromend = from + saved_input_len; } else { fromlen = (*get) (input_buffer, sizeof input_buffer); if (fromlen == 0) return 0; from = input_buffer; fromend = from + fromlen; } while (1) { /* The cases in this switch end with continue, in order to branch back to the top of this while loop and generate the next output character in the appropriate state. */ switch (state) { case -1: ch = *out_string++; if (*out_string == '\0') { state = old_state; old_state = 3; } PUT (ch); continue; case -2: for (;;) { do { ch = GET (); if (ch == EOF) { as_warn (_("end of file in comment")); goto fromeof; } if (ch == '\n') PUT ('\n'); } while (ch != '*'); while ((ch = GET ()) == '*') ; if (ch == EOF) { as_warn (_("end of file in comment")); goto fromeof; } if (ch == '/') break; UNGET (ch); } state = old_state; UNGET (' '); continue; case 4: ch = GET (); if (ch == EOF) goto fromeof; else if (ch >= '0' && ch <= '9') PUT (ch); else { while (ch != EOF && IS_WHITESPACE (ch)) ch = GET (); if (ch == '"') { quotechar = ch; state = 5; old_state = 3; PUT (ch); } else { while (ch != EOF && ch != '\n') ch = GET (); state = 0; PUT (ch); } } continue; case 5: /* We are going to copy everything up to a quote character, with special handling for a backslash. We try to optimize the copying in the simple case without using the GET and PUT macros. */ { char *s; int len; for (s = from; s < fromend; s++) { ch = *s; if (ch == '\\' || ch == quotechar || ch == '\n') break; } len = s - from; if (len > toend - to) len = toend - to; if (len > 0) { memcpy (to, from, len); to += len; from += len; if (to >= toend) goto tofull; } } ch = GET (); if (ch == EOF) { as_warn (_("end of file in string; '%c' inserted"), quotechar); state = old_state; UNGET ('\n'); PUT (quotechar); } else if (ch == quotechar) { state = old_state; PUT (ch); } #ifndef NO_STRING_ESCAPES else if (ch == '\\') { state = 6; PUT (ch); } #endif else if (scrub_m68k_mri && ch == '\n') { /* Just quietly terminate the string. This permits lines like bne label loop if we haven't reach end yet. */ state = old_state; UNGET (ch); PUT ('\''); } else { PUT (ch); } continue; case 6: state = 5; ch = GET (); switch (ch) { /* Handle strings broken across lines, by turning '\n' into '\\' and 'n'. */ case '\n': UNGET ('n'); add_newlines++; PUT ('\\'); continue; case EOF: as_warn (_("end of file in string; '%c' inserted"), quotechar); PUT (quotechar); continue; case '"': case '\\': case 'b': case 'f': case 'n': case 'r': case 't': case 'v': case 'x': case 'X': case '0': case '1': case '2': case '3': case '4': case '5': case '6': case '7': break; default: #ifdef ONLY_STANDARD_ESCAPES as_warn (_("unknown escape '\\%c' in string; ignored"), ch); #endif break; } PUT (ch); continue; #ifdef DOUBLEBAR_PARALLEL case 13: ch = GET (); if (ch != '|') abort (); /* Reset back to state 1 and pretend that we are parsing a line from just after the first white space. */ state = 1; PUT ('|'); continue; #endif #ifdef TC_Z80 case 16: /* We have seen an 'a' at the start of a symbol, look for an 'f'. */ ch = GET (); if (ch == 'f' || ch == 'F') { state = 17; PUT (ch); } else { state = 9; break; } case 17: /* We have seen "af" at the start of a symbol, a ' here is a part of that symbol. */ ch = GET (); state = 9; if (ch == '\'') /* Change to avoid warning about unclosed string. */ PUT ('`'); else UNGET (ch); break; #endif } /* OK, we are somewhere in states 0 through 4 or 9 through 11. */ /* flushchar: */ ch = GET (); #ifdef TC_IA64 if (ch == '(' && (state == 0 || state == 1)) { state += 14; PUT (ch); continue; } else if (state == 14 || state == 15) { if (ch == ')') { state -= 14; PUT (ch); ch = GET (); } else { PUT (ch); continue; } } #endif recycle: #if defined TC_ARM && defined OBJ_ELF /* We need to watch out for .symver directives. See the comment later in this function. */ if (symver_state == NULL) { if ((state == 0 || state == 1) && ch == symver_pseudo[0]) symver_state = symver_pseudo + 1; } else { /* We advance to the next state if we find the right character. */ if (ch != '\0' && (*symver_state == ch)) ++symver_state; else if (*symver_state != '\0') /* We did not get the expected character, or we didn't get a valid terminating character after seeing the entire pseudo-op, so we must go back to the beginning. */ symver_state = NULL; else { /* We've read the entire pseudo-op. If this is the end of the line, go back to the beginning. */ if (IS_NEWLINE (ch)) symver_state = NULL; } } #endif /* TC_ARM && OBJ_ELF */ #ifdef TC_M68K /* We want to have pseudo-ops which control whether we are in MRI mode or not. Unfortunately, since m68k MRI mode affects the scrubber, that means that we need a special purpose recognizer here. */ if (mri_state == NULL) { if ((state == 0 || state == 1) && ch == mri_pseudo[0]) mri_state = mri_pseudo + 1; } else { /* We advance to the next state if we find the right character, or if we need a space character and we get any whitespace character, or if we need a '0' and we get a '1' (this is so that we only need one state to handle ``.mri 0'' and ``.mri 1''). */ if (ch != '\0' && (*mri_state == ch || (*mri_state == ' ' && lex[ch] == LEX_IS_WHITESPACE) || (*mri_state == '0' && ch == '1'))) { mri_last_ch = ch; ++mri_state; } else if (*mri_state != '\0' || (lex[ch] != LEX_IS_WHITESPACE && lex[ch] != LEX_IS_NEWLINE)) { /* We did not get the expected character, or we didn't get a valid terminating character after seeing the entire pseudo-op, so we must go back to the beginning. */ mri_state = NULL; } else { /* We've read the entire pseudo-op. mips_last_ch is either '0' or '1' indicating whether to enter or leave MRI mode. */ do_scrub_begin (mri_last_ch == '1'); mri_state = NULL; /* We continue handling the character as usual. The main gas reader must also handle the .mri pseudo-op to control expression parsing and the like. */ } } #endif if (ch == EOF) { if (state != 0) { as_warn (_("end of file not at end of a line; newline inserted")); state = 0; PUT ('\n'); } goto fromeof; } switch (lex[ch]) { case LEX_IS_WHITESPACE: do { ch = GET (); } while (ch != EOF && IS_WHITESPACE (ch)); if (ch == EOF) goto fromeof; if (state == 0) { /* Preserve a single whitespace character at the beginning of a line. */ state = 1; UNGET (ch); PUT (' '); break; } #ifdef KEEP_WHITE_AROUND_COLON if (lex[ch] == LEX_IS_COLON) { /* Only keep this white if there's no white *after* the colon. */ ch2 = GET (); UNGET (ch2); if (!IS_WHITESPACE (ch2)) { state = 9; UNGET (ch); PUT (' '); break; } } #endif if (IS_COMMENT (ch) || ch == '/' || IS_LINE_SEPARATOR (ch) || IS_PARALLEL_SEPARATOR (ch)) { if (scrub_m68k_mri) { /* In MRI mode, we keep these spaces. */ UNGET (ch); PUT (' '); break; } goto recycle; } /* If we're in state 2 or 11, we've seen a non-white character followed by whitespace. If the next character is ':', this is whitespace after a label name which we normally must ignore. In MRI mode, though, spaces are not permitted between the label and the colon. */ if ((state == 2 || state == 11) && lex[ch] == LEX_IS_COLON && ! scrub_m68k_mri) { state = 1; PUT (ch); break; } switch (state) { case 1: /* We can arrive here if we leave a leading whitespace character at the beginning of a line. */ goto recycle; case 2: state = 3; if (to + 1 < toend) { /* Optimize common case by skipping UNGET/GET. */ PUT (' '); /* Sp after opco */ goto recycle; } UNGET (ch); PUT (' '); break; case 3: if (scrub_m68k_mri) { /* In MRI mode, we keep these spaces. */ UNGET (ch); PUT (' '); break; } goto recycle; /* Sp in operands */ case 9: case 10: if (scrub_m68k_mri) { /* In MRI mode, we keep these spaces. */ state = 3; UNGET (ch); PUT (' '); break; } state = 10; /* Sp after symbol char */ goto recycle; case 11: if (LABELS_WITHOUT_COLONS || flag_m68k_mri) state = 1; else { /* We know that ch is not ':', since we tested that case above. Therefore this is not a label, so it must be the opcode, and we've just seen the whitespace after it. */ state = 3; } UNGET (ch); PUT (' '); /* Sp after label definition. */ break; default: BAD_CASE (state); } break; case LEX_IS_TWOCHAR_COMMENT_1ST: ch2 = GET (); if (ch2 == '*') { for (;;) { do { ch2 = GET (); if (ch2 != EOF && IS_NEWLINE (ch2)) add_newlines++; } while (ch2 != EOF && ch2 != '*'); while (ch2 == '*') ch2 = GET (); if (ch2 == EOF || ch2 == '/') break; /* This UNGET will ensure that we count newlines correctly. */ UNGET (ch2); } if (ch2 == EOF) as_warn (_("end of file in multiline comment")); ch = ' '; goto recycle; } #ifdef DOUBLESLASH_LINE_COMMENTS else if (ch2 == '/') { do { ch = GET (); } while (ch != EOF && !IS_NEWLINE (ch)); if (ch == EOF) as_warn ("end of file in comment; newline inserted"); state = 0; PUT ('\n'); break; } #endif else { if (ch2 != EOF) UNGET (ch2); if (state == 9 || state == 10) state = 3; PUT (ch); } break; case LEX_IS_STRINGQUOTE: quotechar = ch; if (state == 10) { /* Preserve the whitespace in foo "bar". */ UNGET (ch); state = 3; PUT (' '); /* PUT didn't jump out. We could just break, but we know what will happen, so optimize a bit. */ ch = GET (); old_state = 3; } else if (state == 9) old_state = 3; else old_state = state; state = 5; PUT (ch); break; #ifndef IEEE_STYLE case LEX_IS_ONECHAR_QUOTE: if (state == 10) { /* Preserve the whitespace in foo 'b'. */ UNGET (ch); state = 3; PUT (' '); break; } ch = GET (); if (ch == EOF) { as_warn (_("end of file after a one-character quote; \\0 inserted")); ch = 0; } if (ch == '\\') { ch = GET (); if (ch == EOF) { as_warn (_("end of file in escape character")); ch = '\\'; } else ch = process_escape (ch); } sprintf (out_buf, "%d", (int) (unsigned char) ch); /* None of these 'x constants for us. We want 'x'. */ if ((ch = GET ()) != '\'') { #ifdef REQUIRE_CHAR_CLOSE_QUOTE as_warn (_("missing close quote; (assumed)")); #else if (ch != EOF) UNGET (ch); #endif } if (strlen (out_buf) == 1) { PUT (out_buf[0]); break; } if (state == 9) old_state = 3; else old_state = state; state = -1; out_string = out_buf; PUT (*out_string++); break; #endif case LEX_IS_COLON: #ifdef KEEP_WHITE_AROUND_COLON state = 9; #else if (state == 9 || state == 10) state = 3; else if (state != 3) state = 1; #endif PUT (ch); break; case LEX_IS_NEWLINE: /* Roll out a bunch of newlines from inside comments, etc. */ if (add_newlines) { --add_newlines; UNGET (ch); } /* Fall through. */ case LEX_IS_LINE_SEPARATOR: state = 0; PUT (ch); break; case LEX_IS_PARALLEL_SEPARATOR: state = 1; PUT (ch); break; #ifdef TC_V850 case LEX_IS_DOUBLEDASH_1ST: ch2 = GET (); if (ch2 != '-') { UNGET (ch2); goto de_fault; } /* Read and skip to end of line. */ do { ch = GET (); } while (ch != EOF && ch != '\n'); if (ch == EOF) as_warn (_("end of file in comment; newline inserted")); state = 0; PUT ('\n'); break; #endif #ifdef DOUBLEBAR_PARALLEL case LEX_IS_DOUBLEBAR_1ST: ch2 = GET (); UNGET (ch2); if (ch2 != '|') goto de_fault; /* Handle '||' in two states as invoking PUT twice might result in the first one jumping out of this loop. We'd then lose track of the state and one '|' char. */ state = 13; PUT ('|'); break; #endif case LEX_IS_LINE_COMMENT_START: /* FIXME-someday: The two character comment stuff was badly thought out. On i386, we want '/' as line comment start AND we want C style comments. hence this hack. The whole lexical process should be reworked. xoxorich. */ if (ch == '/') { ch2 = GET (); if (ch2 == '*') { old_state = 3; state = -2; break; } else { UNGET (ch2); } } if (state == 0 || state == 1) /* Only comment at start of line. */ { int startch; startch = ch; do { ch = GET (); } while (ch != EOF && IS_WHITESPACE (ch)); if (ch == EOF) { as_warn (_("end of file in comment; newline inserted")); PUT ('\n'); break; } if (ch < '0' || ch > '9' || state != 0 || startch != '#') { /* Not a cpp line. */ while (ch != EOF && !IS_NEWLINE (ch)) ch = GET (); if (ch == EOF) as_warn (_("end of file in comment; newline inserted")); state = 0; PUT ('\n'); break; } /* Looks like `# 123 "filename"' from cpp. */ UNGET (ch); old_state = 4; state = -1; if (scrub_m68k_mri) out_string = "\tlinefile "; else out_string = "\t.linefile "; PUT (*out_string++); break; } #ifdef TC_D10V /* All insns end in a char for which LEX_IS_SYMBOL_COMPONENT is true. Trap is the only short insn that has a first operand that is neither register nor label. We must prevent exef0f ||trap #1 to degenerate to exef0f ||trap#1 . We can't make '#' LEX_IS_SYMBOL_COMPONENT because it is already LEX_IS_LINE_COMMENT_START. However, it is the only character in line_comment_chars for d10v, hence we can recognize it as such. */ /* An alternative approach would be to reset the state to 1 when we see '||', '<'- or '->', but that seems to be overkill. */ if (state == 10) PUT (' '); #endif /* We have a line comment character which is not at the start of a line. If this is also a normal comment character, fall through. Otherwise treat it as a default character. */ if (strchr (tc_comment_chars, ch) == NULL && (! scrub_m68k_mri || (ch != '!' && ch != '*'))) goto de_fault; if (scrub_m68k_mri && (ch == '!' || ch == '*' || ch == '#') && state != 1 && state != 10) goto de_fault; /* Fall through. */ case LEX_IS_COMMENT_START: #if defined TC_ARM && defined OBJ_ELF /* On the ARM, `@' is the comment character. Unfortunately this is also a special character in ELF .symver directives (and .type, though we deal with those another way). So we check if this line is such a directive, and treat the character as default if so. This is a hack. */ if ((symver_state != NULL) && (*symver_state == 0)) goto de_fault; #endif #ifdef TC_ARM /* For the ARM, care is needed not to damage occurrences of \@ by stripping the @ onwards. Yuck. */ if (to > tostart && *(to - 1) == '\\') /* Do not treat the @ as a start-of-comment. */ goto de_fault; #endif #ifdef WARN_COMMENTS if (!found_comment) as_where (&found_comment_file, &found_comment); #endif do { ch = GET (); } while (ch != EOF && !IS_NEWLINE (ch)); if (ch == EOF) as_warn (_("end of file in comment; newline inserted")); state = 0; PUT ('\n'); break; case LEX_IS_SYMBOL_COMPONENT: if (state == 10) { /* This is a symbol character following another symbol character, with whitespace in between. We skipped the whitespace earlier, so output it now. */ UNGET (ch); state = 3; PUT (' '); break; } #ifdef TC_Z80 /* "af'" is a symbol containing '\''. */ if (state == 3 && (ch == 'a' || ch == 'A')) { state = 16; PUT (ch); ch = GET (); if (ch == 'f' || ch == 'F') { state = 17; PUT (ch); break; } else { state = 9; if (!IS_SYMBOL_COMPONENT (ch)) { UNGET (ch); break; } } } #endif if (state == 3) state = 9; /* This is a common case. Quickly copy CH and all the following symbol component or normal characters. */ if (to + 1 < toend && mri_state == NULL #if defined TC_ARM && defined OBJ_ELF && symver_state == NULL #endif ) { char *s; int len; for (s = from; s < fromend; s++) { int type; ch2 = *(unsigned char *) s; type = lex[ch2]; if (type != 0 && type != LEX_IS_SYMBOL_COMPONENT) break; } if (s > from) /* Handle the last character normally, for simplicity. */ --s; len = s - from; if (len > (toend - to) - 1) len = (toend - to) - 1; if (len > 0) { PUT (ch); memcpy (to, from, len); to += len; from += len; if (to >= toend) goto tofull; ch = GET (); } } /* Fall through. */ default: de_fault: /* Some relatively `normal' character. */ if (state == 0) { state = 11; /* Now seeing label definition. */ } else if (state == 1) { state = 2; /* Ditto. */ } else if (state == 9) { if (!IS_SYMBOL_COMPONENT (ch)) state = 3; } else if (state == 10) { if (ch == '\\') { /* Special handling for backslash: a backslash may be the beginning of a formal parameter (of a macro) following another symbol character, with whitespace in between. If that is the case, we output a space before the parameter. Strictly speaking, correct handling depends upon what the macro parameter expands into; if the parameter expands into something which does not start with an operand character, then we don't want to keep the space. We don't have enough information to make the right choice, so here we are making the choice which is more likely to be correct. */ if (to + 1 >= toend) { /* If we're near the end of the buffer, save the character for the next time round. Otherwise we'll lose our state. */ UNGET (ch); goto tofull; } *to++ = ' '; } state = 3; } PUT (ch); break; } } /*NOTREACHED*/ fromeof: /* We have reached the end of the input. */ return to - tostart; tofull: /* The output buffer is full. Save any input we have not yet processed. */ if (fromend > from) { saved_input = from; saved_input_len = fromend - from; } else saved_input = NULL; return to - tostart; }
// process a line of the sass text string process (string& sass, converter& converter) { // resulting string string scss = ""; // strip multi line comments if (STRIP_COMMENT(converter)) { sass = removeMultilineComment(sass); } // right trim input sass = rtrim(sass); // get postion of first meaningfull character in string size_t pos_left = sass.find_first_not_of(SASS2SCSS_FIND_WHITESPACE); // special case for final run if (converter.end_of_file) pos_left = 0; // maybe has only whitespace if (pos_left == string::npos) { // just add complete whitespace converter.whitespace += sass + "\n"; } // have meaningfull first char else { // extract and store indentation string string indent = sass.substr(0, pos_left); // check if current line starts a comment string open = sass.substr(pos_left, 2); // line has less or same indentation // finalize previous open parser context if (indent.length() <= INDENT(converter).length()) { // close multilinie comment if (IS_CSS_COMMENT(converter)) { // check if comments will be stripped anyway if (!STRIP_COMMENT(converter)) scss += " */"; } // close src comment comment else if (IS_SRC_COMMENT(converter)) { // add a newline to avoid closer on same line // this would put the bracket in the comment node // no longer needed since we parse them correctly // if (KEEP_COMMENT(converter)) scss += "\n"; } // close css properties else if (converter.property) { // add closer unless in concat mode if (!converter.comma) { // if there was no colon we have a selector // looks like there were no inner properties if (converter.selector) scss += " {}"; // add final semicolon else if (!converter.semicolon) scss += ";"; } } // reset comment state converter.comment = ""; } // make sure we close every "higher" block while (indent.length() < INDENT(converter).length()) { // pop stacked context converter.indents.pop(); // print close bracket if (IS_PARSING(converter)) { scss += closer(converter); } else { scss += " */"; } // reset comment state converter.comment = ""; } // reset converter state converter.selector = false; // check if we have sass property syntax if (sass.substr(pos_left, 1) == ":" && sass.substr(pos_left, 2) != "::") { // default to a selector // change back if property found converter.selector = true; // get postion of first whitespace char size_t pos_wspace = sass.find_first_of(SASS2SCSS_FIND_WHITESPACE, pos_left); // assertion check for valid result if (pos_wspace != string::npos) { // get the possible pseudo selector string pseudo = sass.substr(pos_left, pos_wspace - pos_left); // get position of the first real property value char // pseudo selectors get this far, but have no actual value size_t pos_value = sass.find_first_not_of(SASS2SCSS_FIND_WHITESPACE, pos_wspace); // assertion check for valid result if (pos_value != string::npos) { // only process if not (fallowed by a semicolon or is a pseudo selector) if (!(sass.at(pos_value) == ':' || isPseudoSelector(pseudo))) { // create new string by interchanging the colon sign for property and value sass = indent + sass.substr(pos_left + 1, pos_wspace - pos_left - 1) + ":" + sass.substr(pos_wspace); // try to find a colon in the current line, but only ... size_t pos_colon = sass.find_first_not_of(":", pos_left); // assertion for valid result if (pos_colon != string::npos) { // ... after the first word (skip begining colons) pos_colon = sass.find_first_of(":", pos_colon); // it is a selector if there was no colon found converter.selector = pos_colon == string::npos; } } } } } // terminate warn and debug statements immediately else if (sass.substr(pos_left, 5) == "@warn" || sass.substr(pos_left, 6) == "@debug") { sass = indent + sass.substr(pos_left); } // replace some specific sass shorthand directives (if not fallowed by a white space character) else if (sass.substr(pos_left, 1) == "=" && sass.find_first_of(SASS2SCSS_FIND_WHITESPACE, pos_left) != pos_left + 1) { sass = indent + "@mixin " + sass.substr(pos_left + 1); } else if (sass.substr(pos_left, 1) == "+" && sass.find_first_of(SASS2SCSS_FIND_WHITESPACE, pos_left) != pos_left + 1) { sass = indent + "@include " + sass.substr(pos_left + 1); } // add quotes for import if needed else if (sass.substr(pos_left, 7) == "@import") { // get positions for the actual import url size_t pos_import = sass.find_first_of(SASS2SCSS_FIND_WHITESPACE, pos_left + 7); size_t pos_quote = sass.find_first_not_of(SASS2SCSS_FIND_WHITESPACE, pos_import); // leave proper urls untouched if (sass.substr(pos_quote, 4) != "url(") { // check if the url appears to be already quoted if (sass.substr(pos_quote, 1) != "\"" && sass.substr(pos_quote, 1) != "\'") { // get position of the last char on the line size_t pos_end = sass.find_last_not_of(SASS2SCSS_FIND_WHITESPACE); // assertion check for valid result if (pos_end != string::npos) { // add quotes around the full line after the import statement sass = sass.substr(0, pos_quote) + "\"" + sass.substr(pos_quote, pos_end - pos_quote + 1) + "\""; } } } } else if ( sass.substr(pos_left, 7) != "@return" && sass.substr(pos_left, 7) != "@extend" && sass.substr(pos_left, 8) != "@content" ) { // try to find a colon in the current line, but only ... size_t pos_colon = sass.find_first_not_of(":", pos_left); // assertion for valid result if (pos_colon != string::npos) { // ... after the first word (skip begining colons) pos_colon = sass.find_first_of(":", pos_colon); // it is a selector if there was no colon found converter.selector = pos_colon == string::npos; } } // current line has more indentation if (indent.length() >= INDENT(converter).length()) { // not in comment mode if (IS_PARSING(converter)) { // has meaningfull chars if (hasCharData(sass)) { // is probably a property // also true for selectors converter.property = true; } } } // current line has more indentation if (indent.length() > INDENT(converter).length()) { // not in comment mode if (IS_PARSING(converter)) { // had meaningfull chars if (converter.property) { // print block opener scss += opener(converter); // push new stack context converter.indents.push(""); // store block indentation INDENT(converter) = indent; } } // is and will be a src comment else if (!IS_CSS_COMMENT(converter)) { // scss does not allow multiline src comments // therefore add forward slashes to all lines sass.at(INDENT(converter).length()+0) = '/'; // there is an edge case here if indentation // is minimal (will overwrite the fist char) sass.at(INDENT(converter).length()+1) = '/'; // could code around that, but I dont' think // this will ever be the cause for any trouble } } // line is opening a new comment if (open == "/*" || open == "//") { // reset the property state converter.property = false; // close previous comment if (IS_CSS_COMMENT(converter) && open != "") { if (!STRIP_COMMENT(converter) && !CONVERT_COMMENT(converter)) scss += " */"; } // force single line comments // into a correct css comment if (CONVERT_COMMENT(converter)) { if (IS_PARSING(converter)) { sass.at(pos_left + 1) = '*'; } } // set comment flag converter.comment = open; } // flush data only under certain conditions if (!( // strip css and src comments if option is set (IS_COMMENT(converter) && STRIP_COMMENT(converter)) || // strip src comment even if strip option is not set // but only if the keep src comment option is not set (IS_SRC_COMMENT(converter) && ! KEEP_COMMENT(converter)) )) { // flush data and buffer whitespace scss += flush(sass, converter); } // get postion of last meaningfull char size_t pos_right = sass.find_last_not_of(SASS2SCSS_FIND_WHITESPACE); // check for invalid result if (pos_right != string::npos) { // get the last meaningfull char string close = sass.substr(pos_right, 1); // check if next line should be concatenated (list mode) converter.comma = IS_PARSING(converter) && close == ","; converter.semicolon = IS_PARSING(converter) && close == ";"; // check if we have more than // one meaningfull char if (pos_right > 0) { // get the last two chars from string string close = sass.substr(pos_right - 1, 2); // update parser status for expicitly closed comment if (close == "*/") converter.comment = ""; } } // EO have meaningfull chars from end } // EO have meaningfull chars from start // return scss return scss; }
// flush whitespace and print additional text, but // only print additional chars and buffer whitespace string flush (string& sass, converter& converter) { // return flushed string scss = ""; // print whitespace buffer scss += PRETTIFY(converter) > 0 ? converter.whitespace : ""; // reset whitespace buffer converter.whitespace = ""; // remove possible newlines from string size_t pos_right = sass.find_last_not_of("\n\r"); if (pos_right == string::npos) return scss; // get the linefeeds from the string string lfs = sass.substr(pos_right + 1); sass = sass.substr(0, pos_right + 1); // find some source comment opener size_t comment_pos = findCommentOpener(sass); // check if there was a source comment if (comment_pos != string::npos) { // convert comment (but only outside other coments) if (CONVERT_COMMENT(converter) && !IS_COMMENT(converter)) { // convert to multiline comment sass.at(comment_pos + 1) = '*'; // add comment node to the whitespace sass += " */"; } // not at line start if (comment_pos > 0) { // also include whitespace before the actual comment opener size_t ws_pos = sass.find_last_not_of(SASS2SCSS_FIND_WHITESPACE, comment_pos - 1); comment_pos = ws_pos == string::npos ? 0 : ws_pos + 1; } if (!STRIP_COMMENT(converter)) { // add comment node to the whitespace converter.whitespace += sass.substr(comment_pos); } else { // sass = removeMultilineComments(sass); } // update the actual sass code sass = sass.substr(0, comment_pos); } // add newline as getline discharged it converter.whitespace += lfs + "\n"; // maybe remove any leading whitespace if (PRETTIFY(converter) == 0) { // remove leading whitespace and update string size_t pos_left = sass.find_first_not_of(SASS2SCSS_FIND_WHITESPACE); if (pos_left != string::npos) sass = sass.substr(pos_left); } // add flushed data scss += sass; // return string return scss; }
std::string read_user_format_number(std::ifstream &ifs, UserFormatColumn<DATA_T> &ufc, std::string &line, bool &need_readline, encode_result &encres, char* argv[] = 0) { std::string ret, fm_buffer; DATA_T val = 0; char* c; ret.clear(); while (std::getline(ifs, line)) { encres.line_nr++; encres.line_in_blk++; c = const_cast<char*>(line.c_str()); fm_buffer.clear(); if (IS_COMMENT(line)) continue; /* end of FORMATs of a user block */ if (IS_BLANK(line)) { need_readline = false; break; } while (!IS_EOL(*c) && !IS_EOS(*c)) { /* Meeting the next FORMAT, which means this format is done */ if (IS_LETTER(*c)) { need_readline = false; break; } else if (IS_PLACEHOLD(*c)) { if (!TEMPLATE_MODE) throw E2promValueException("Placehold should be used in template file", to_string<int>(encres.line_nr), usage_allowed_user_format_type); E2promMsg("Reading parameter for user "+to_string<uint16_t>(ufc.uheader.ID)+ " @ line "+to_string<int>(encres.line_nr)); read_user_parameters(fm_buffer, ufc.uheader.ID, encres, argv); break; } else if (IS_DIGIT(*c) || IS_PERIOD(*c)) { fm_buffer += *c++; } else if (IS_SPACE(*c) || (IS_COMMENT_SIGN(*c))) { break; } else { throw E2promValueException( "Unexpected characters found for type", to_string<int>(encres.line_nr), usage_allowed_user_format_type ); } } /* end of read-char while-loop */ // if (TEMPLATE_MODE) // process_fm_buffer(fm_buffer, ufc.uheader.ID); if (!fm_buffer.empty()) { val = to_digits<DATA_T>(fm_buffer); ufc += val; } if (!IS_LETTER(*c) && !IS_BLANK(line)) { need_readline = true; } else { break; } } /* end of read-line while-loop */ // ret += strize_formats<DATA_T>(ufc); std::stringstream ss; // ss.width(decres.GLOBAL_ALIGNMENT); ss << ufc; return ss.str(); }
char *get_script_next_line(char **line_buf_ptr) { char *line_buf; char *next_line; int i; UBOOT_TRACE("IN\n"); line_buf = (*line_buf_ptr); // strip '\r', '\n' and comment while (1) { // strip '\r' & '\n' & ' ' & '\t' if (IS_IGNORED_CHAR(line_buf[0])) { line_buf++; } // strip comment else if (IS_COMMENT(line_buf[0])) { for (i = 0; !IS_LINE_END(line_buf[0]) && i <= MAX_LINE_SIZE; i++) { line_buf++; } if (i > MAX_LINE_SIZE) { line_buf[0] = SCRIPT_FILE_END; UBOOT_ERROR("Error: the max size of one line is %d!!!\n", MAX_LINE_SIZE); // <-@@@ break; } } else { break; } } // get next line if (IS_FILE_END(line_buf[0])) { next_line = NULL; } else { next_line = line_buf; for (i = 0; !IS_LINE_END(line_buf[0]) && i <= MAX_LINE_SIZE; i++) { line_buf++; } if (i > MAX_LINE_SIZE) { next_line = NULL; UBOOT_ERROR("Error: the max size of one line is %d!!!\n", MAX_LINE_SIZE); // <-@@@ } else { line_buf[0] = '\0'; *line_buf_ptr = line_buf + 1; } } UBOOT_TRACE("OK\n"); return next_line; }
int cfg_get_token(cfg_token_t* token, cfg_parser_t* st, unsigned int flags) { static int look_ahead = EOF; int c; enum st state; state = ST_S; token->val.s = token->buf; token->val.len = 0; if (look_ahead != EOF) { c = look_ahead; look_ahead = EOF; } else { READ_CHAR; } while(c != EOF) { switch(state) { case ST_S: if (flags & CFG_EXTENDED_ALPHA) { if (IS_WHITESPACE(c)) { /* Do nothing */ } else if (IS_ALPHA(c) || IS_ESCAPE(c) || IS_DELIM(c)) { PUSH(c); state = ST_A; } else if (IS_QUOTE(c)) { state = ST_Q; } else if (IS_COMMENT(c)) { state = ST_C; } else if (IS_EOL(c)) { PUSH(c); RETURN(c); } else { ERR("%s:%d:%d: Invalid character 0x%x\n", st->file, st->line, st->col, c); return -1; } } else { if (IS_WHITESPACE(c)) { /* Do nothing */ } else if (IS_ALPHA(c)) { PUSH(c); state = ST_A; } else if (IS_QUOTE(c)) { state = ST_Q; } else if (IS_COMMENT(c)) { state = ST_C; } else if (IS_ESCAPE(c)) { state = ST_E; } else if (IS_DELIM(c) || IS_EOL(c)) { PUSH(c); RETURN(c); } else { ERR("%s:%d:%d: Invalid character 0x%x\n", st->file, st->line, st->col, c); return -1; } } break; case ST_A: if (flags & CFG_EXTENDED_ALPHA) { if (IS_ALPHA(c) || IS_DELIM(c) || IS_QUOTE(c)) { PUSH(c); } else if (IS_ESCAPE(c)) { state = ST_AE; } else if (IS_COMMENT(c) || IS_EOL(c) || IS_WHITESPACE(c)) { look_ahead = c; RETURN(CFG_TOKEN_ALPHA); } else { ERR("%s:%d:%d: Invalid character 0x%x\n", st->file, st->line, st->col, c); return -1; } } else { if (IS_ALPHA(c)) { PUSH(c); } else if (IS_ESCAPE(c)) { state = ST_AE; } else if (IS_WHITESPACE(c) || IS_DELIM(c) || IS_QUOTE(c) || IS_COMMENT(c) || IS_EOL(c)) { look_ahead = c; RETURN(CFG_TOKEN_ALPHA); } else { ERR("%s:%d:%d: Invalid character 0x%x\n", st->file, st->line, st->col, c); return -1; } } break; case ST_AE: if (IS_COMMENT(c) || IS_QUOTE(c) || IS_ESCAPE(c)) { PUSH(c); } else if (c == 'r') { PUSH('\r'); } else if (c == 'n') { PUSH('\n'); } else if (c == 't') { PUSH('\t'); } else if (c == ' ') { PUSH(' '); } else if (IS_EOL(c)) { /* Do nothing */ } else { ERR("%s:%d:%d: Unsupported escape character 0x%x\n", st->file, st->line, st->col, c); return -1; } state = ST_A; break; case ST_Q: if (IS_QUOTE(c)) { RETURN(CFG_TOKEN_STRING); } else if (IS_ESCAPE(c)) { state = ST_QE; break; } else { PUSH(c); } break; case ST_QE: if (IS_ESCAPE(c) || IS_QUOTE(c)) { PUSH(c); } else if (c == 'n') { PUSH('\n'); } else if (c == 'r') { PUSH('\r'); } else if (c == 't') { PUSH('\t'); } else if (IS_EOL(c)) { /* Do nothing */ } else { ERR("%s:%d:%d: Unsupported escape character 0x%x\n", st->file, st->line, st->col, c); return -1; } state = ST_Q; break; case ST_C: if (IS_ESCAPE(c)) { state = ST_CE; } else if (IS_EOL(c)) { state = ST_S; continue; /* Do not read a new char, return EOL */ } else { /* Do nothing */ } break; case ST_CE: state = ST_C; break; case ST_E: if (IS_COMMENT(c) || IS_QUOTE(c) || IS_ESCAPE(c)) { PUSH(c); RETURN(c); } else if (c == 'r') { PUSH('\r'); RETURN('\r'); } else if (c == 'n') { PUSH('\n'); RETURN('\n'); } else if (c == 't') { PUSH('\t'); RETURN('\t'); } else if (c == ' ') { PUSH(' '); RETURN(' '); } else if (IS_EOL(c)) { /* Escped eol means no eol */ state = ST_S; } else { ERR("%s:%d:%d: Unsupported escape character 0x%x\n", st->file, st->line, st->col, c); return -1; } break; } READ_CHAR; }; switch(state) { case ST_S: case ST_C: case ST_CE: return 1; case ST_A: RETURN(CFG_TOKEN_ALPHA); case ST_Q: ERR("%s:%d:%d: Premature end of file, missing closing quote in" " string constant\n", st->file, st->line, st->col); return -1; case ST_QE: case ST_E: case ST_AE: ERR("%s:%d:%d: Premature end of file, missing escaped character\n", st->file, st->line, st->col); return -1; } BUG("%s:%d:%d: Invalid state %d\n", st->file, st->line, st->col, state); return -1; }
/** * Parse config file of the program. * * @return 0 on success, -1 on fail */ static int read_config(void) { #define LINE_BUF_SIZE 2048 char buf[LINE_BUF_SIZE]; FILE* fd; parsed_option_t option; int res; /* initialize conf_opt and opt structures */ memset(&conf_opt, 0, sizeof(opt)); conf_opt.find_max_depth = -1; if(!find_conf_file()) return 0; fd = fopen(conf_opt.config_file, "r"); if(!fd) return -1; while(fgets(buf, LINE_BUF_SIZE, fd)) { size_t index; cmdline_opt_t* t; char* line = str_trim(buf); char *name, *value; if(*line == 0 || IS_COMMENT(*line)) continue; /* search for '=' */ index = strcspn(line, "="); if(line[index] == 0) { log_warning(_("%s: can't parse line \"%s\"\n"), conf_opt.config_file, line); continue; } line[index] = 0; name = str_trim(line); for(t = cmdline_opt; t->type; t++) { if(strcmp(name, t->long_name) == 0) { break; } } if(!t->type) { log_warning(_("%s: unknown option \"%s\"\n"), conf_opt.config_file, line); continue; } value = str_trim(line + index + 1); /* process a long option */ if(is_param_required(t->type)) { rsh_vector_add_ptr(opt.mem, (value = rsh_strdup(value)));; } else { /* possible boolean values for a config file variable */ static const char* strings[] = {"1", "on", "yes", 0}; const char** cmp; for(cmp = strings; *cmp && strcmp(value, *cmp); cmp++); if(*cmp == 0) continue; } option.name = name; option.parameter = value; option.o = t; apply_option(&conf_opt, &option); } res = fclose(fd); #ifdef _WIN32 if( (opt.flags & OPT_ENCODING) == 0 ) opt.flags |= (conf_opt.flags & OPT_ENCODING); #endif return (res == 0 ? 0 : -1); }
/** パラメータの解析とカウント @param[in] end buf末尾 @param[in] p 解析の現在位置 パラメータを解析する.パラメータの数と末尾の閉じ括弧を正しく判別するために, 引用符,括弧,パラメータの区切りのカンマに着目する. 引用符は改行を含むことができない. */ const wchar_t* COutlineErlang::ScanArgs( const wchar_t* end, const wchar_t* p ) { assert( m_state == STATE_FUNC_ARGS ); const int parptr_max = sizeof( m_parenthesis ) / sizeof( m_parenthesis[0] ); wchar_t quote = L'\0'; // 先頭位置を保存 for(const wchar_t* head = p ; p < end ; p++ ){ if( quote ){ if( *p == quote ) quote = L'\0'; } else { if( wcschr( L"([{", *p )){ //) // level up if( m_parenthesis_ptr < parptr_max ){ m_parenthesis[ m_parenthesis_ptr ] = *p; } m_parenthesis_ptr++; } else if( wcschr( L")]}", *p )){ //) wchar_t op; switch( *p ){ case L')': op = L'('; break; case L']': op = L'['; break; case L'}': op = L'{'; break; default: PleaseReportToAuthor(NULL, LS(STR_OUTLINE_ERLANG_SCANARGS)); op = 0; break; } // level down --m_parenthesis_ptr; while( 1 <= m_parenthesis_ptr && m_parenthesis_ptr < parptr_max ){ if( m_parenthesis[ m_parenthesis_ptr ] != op ){ // if unmatch then skip --m_parenthesis_ptr; } else{ break; } } // check level if( m_parenthesis_ptr == 0 ){ m_state = STATE_FUNC_ARGS_FIN; ++p; return p; } } else if( *p == L',' && m_parenthesis_ptr == 1 ){ ++m_argcount; } else if( *p == L';' ){ // セミコロンは複数の文の区切り. // パラメータ中には現れないので,解析が失敗している // 括弧の閉じ忘れが考えられるので,仕切り直し m_state = STATE_NORMAL; return end; } else if( *p == L'.' ){ // ピリオドは式の末尾か,小数点として使われる. if( p > head && ( L'0' <= p[-1] && p[-1] <= L'9' )){ // 小数点かもしれないので,そのままにする } else { // 引数の途中で文末が現れたのは解析が失敗している // 括弧の閉じ忘れが考えられるので,仕切り直し m_state = STATE_NORMAL; return end; } } else if( *p == L'"' ){ quote = L'"'; } else if( *p == L'\'' ){ quote = L'\''; } else if( IS_COMMENT( *p )){ return end; } } } return p; }
int do_scrub_next_char_from_string() { /* State 0: beginning of normal line 1: After first whitespace on normal line (flush more white) 2: After first non-white on normal line (keep 1white) 3: after second white on normal line (flush white) 4: after putting out a .line, put out digits 5: parsing a string, then go to old-state 6: putting out \ escape in a "d string. 7: After putting out a .file, put out string. 8: After putting out a .file string, flush until newline. -1: output string in out_string and go to the state in old_state -2: flush text until a '*' '/' is seen, then go to state old_state */ #ifndef NeXT_MOD /* .include feature */ static state; static old_state; static char *out_string; static char out_buf[20]; static add_newlines = 0; #endif /* NeXT_MOD .include feature */ int ch; if(state==-1) { ch= *out_string++; if(*out_string==0) { state=old_state; old_state=3; } return ch; } if(state==-2) { for(;;) { do ch=scrub_from_string(); while(ch!=EOF && ch!='\n' && ch!='*'); if(ch=='\n' || ch==EOF) return ch; ch=scrub_from_string(); if(ch==EOF || ch=='/') break; scrub_to_string(ch); } state=old_state; return ' '; } if(state==4) { ch=scrub_from_string(); if(ch==EOF || (ch>='0' && ch<='9')) return ch; else { while(ch!=EOF && IS_WHITESPACE(ch)) ch=scrub_from_string(); if(ch=='"') { scrub_to_string(ch); #if defined(M88K) || defined(PPC) || defined(HPPA) out_string="@ .file "; #else out_string="; .file "; #endif old_state=7; state= -1; return *out_string++; } else { while(ch!=EOF && ch!='\n') ch=scrub_from_string(); #ifdef NeXT_MOD /* bug fix for bug #8918, which was when * a full line comment line this: * # 40 MP1 = M + 1 * got confused with a cpp output like: * # 1 "hello.c" 1 */ state = 0; #endif /* NeXT_MOD */ return ch; } } } if(state==5) { ch=scrub_from_string(); #ifdef PPC if(flagseen[(int)'p'] == TRUE && ch=='\'') { state=old_state; return '\''; } else #endif /* PPC */ if(ch=='"') { state=old_state; return '"'; } else if(ch=='\\') { state=6; return ch; } else if(ch==EOF) { state=old_state; scrub_to_string('\n'); #ifdef PPC if(flagseen[(int)'p'] == TRUE){ as_warn("End of file in string: inserted '\''"); return '\''; } #endif /* PPC */ as_warn("End of file in string: inserted '\"'"); return '"'; } else { return ch; } } if(state==6) { state=5; ch=scrub_from_string(); switch(ch) { /* This is neet. Turn "string more string" into "string\n more string" */ case '\n': scrub_to_string('n'); add_newlines++; return '\\'; case '"': case '\\': case 'b': case 'f': case 'n': case 'r': case 't': case '0': case '1': case '2': case '3': case '4': case '5': case '6': case '7': break; default: as_warn("Unknown escape '\\%c' in string: Ignored",ch); break; case EOF: as_warn("End of file in string: '\"' inserted"); return '"'; } return ch; } if(state==7) { ch=scrub_from_string(); state=5; old_state=8; return ch; } if(state==8) { do ch= scrub_from_string(); while(ch!='\n'); state=0; return ch; } flushchar: ch=scrub_from_string(); switch(ch) { case ' ': case '\t': do ch=scrub_from_string(); while(ch!=EOF && IS_WHITESPACE(ch)); if(ch==EOF) return ch; if(IS_COMMENT(ch) || (state==0 && IS_LINE_COMMENT(ch)) || ch=='/' || IS_LINE_SEPERATOR(ch)) { scrub_to_string(ch); goto flushchar; } scrub_to_string(ch); if(state==0 || state==2) { state++; return ' '; } #ifdef ARM if (state == 3) return ' '; /* stupid stupid stupid. */ #endif #ifdef PPC if(flagseen[(int)'p'] == TRUE && state == 3){ return ' '; } #endif else goto flushchar; case '/': ch=scrub_from_string(); if(ch=='*') { for(;;) { do { ch=scrub_from_string(); if(ch=='\n') add_newlines++; } while(ch!=EOF && ch!='*'); ch=scrub_from_string(); if(ch==EOF || ch=='/') break; scrub_to_string(ch); } if(ch==EOF) as_warn("End of file in '/' '*' string: */ inserted"); scrub_to_string(' '); goto flushchar; } else { #if defined(I860) || defined(M88K) || defined(PPC) || defined(I386) || \ defined(HPPA) || defined (SPARC) if (ch == '/') { do { ch=scrub_from_string(); } while (ch != EOF && (ch != '\n')); if (ch == EOF) as_warn("End of file before newline in // comment"); if ( ch == '\n' ) /* Push NL back so we can complete state */ scrub_to_string(ch); goto flushchar; } #endif if(IS_COMMENT('/') || (state==0 && IS_LINE_COMMENT('/'))) { scrub_to_string(ch); ch='/'; goto deal_misc; } if(ch!=EOF) scrub_to_string(ch); return '/'; } break; case '"': old_state=state; state=5; return '"'; break; case '\'': #ifdef PPC if(flagseen[(int)'p'] == TRUE){ old_state=state; state=5; return '\''; break; } #endif ch=scrub_from_string(); if(ch==EOF) { as_warn("End-of-file after a ': \\000 inserted"); ch=0; } sprintf(out_buf,"(%d)",ch&0xff); old_state=state; state= -1; out_string=out_buf; return *out_string++; case ':': if(state!=3) state=0; return ch; case '\n': if(add_newlines) { --add_newlines; scrub_to_string(ch); } /* Fall through. */ #if defined(M88K) || defined(PPC) || defined(HPPA) case '@': #else case ';': #endif state=0; return ch; default: deal_misc: if(state==0 && IS_LINE_COMMENT(ch)) { do ch=scrub_from_string(); while(ch!=EOF && IS_WHITESPACE(ch)); if(ch==EOF) { as_warn("EOF in comment: Newline inserted"); return '\n'; } if(ch<'0' || ch>'9') { if(ch!='\n'){ do ch=scrub_from_string(); while(ch!=EOF && ch!='\n'); } if(ch==EOF) as_warn("EOF in Comment: Newline inserted"); state=0; return '\n'; } scrub_to_string(ch); old_state=4; state= -1; out_string=".line "; return *out_string++; } else if(IS_COMMENT(ch)) { do ch=scrub_from_string(); while(ch!=EOF && ch!='\n'); if(ch==EOF) as_warn("EOF in comment: Newline inserted"); state=0; return '\n'; } else if(state==0) { state=2; return ch; } else if(state==1) { state=2; return ch; } else { return ch; } case EOF: if(state==0) return ch; as_warn("End-of-File not at end of a line"); } return -1; }