1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096
|
/* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
/*
* This file is part of the LibreOffice project.
*
* This Source Code Form is subject to the terms of the Mozilla Public
* License, v. 2.0. If a copy of the MPL was not distributed with this
* file, You can obtain one at http://mozilla.org/MPL/2.0/.
*
* This file incorporates work covered by the following license notice:
*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed
* with this work for additional information regarding copyright
* ownership. The ASF licenses this file to you under the Apache
* License, Version 2.0 (the "License"); you may not use this file
* except in compliance with the License. You may obtain a copy of
* the License at http://www.apache.org/licenses/LICENSE-2.0 .
*/
#include <stdlib.h>
#include <basic/sbxform.hxx>
#include <rtl/ustrbuf.hxx>
/*
TODO: are there any Star-Basic characteristics unconsidered?
what means: * as placeholder
COMMENT: Visual-Basic treats the following (invalid) format-strings
as shown:
##0##.##0## --> ##000.000##
(this class behaves the same way)
*/
#include <stdio.h>
#include <float.h>
#include <math.h>
#define NO_DIGIT_ -1
#define MAX_NO_OF_DIGITS DBL_DIG
#define MAX_DOUBLE_BUFFER_LENGTH MAX_NO_OF_DIGITS + 9
// +1 for leading sign
// +1 for digit before the decimal point
// +1 for decimal point
// +2 for exponent E and exp. leading sign
// +3 for the exponent's value
// +1 for closing 0
// Defines for the digits:
#define ASCII_0 '0' // 48
#define ASCII_9 '9' // 57
#define CREATE_1000SEP_CHAR '@'
#define FORMAT_SEPARATOR ';'
// predefined formats for the Format$()-command:
#define BASICFORMAT_GENERALNUMBER "General Number"
#define BASICFORMAT_CURRENCY "Currency"
#define BASICFORMAT_FIXED "Fixed"
#define BASICFORMAT_STANDARD "Standard"
#define BASICFORMAT_PERCENT "Percent"
#define BASICFORMAT_SCIENTIFIC "Scientific"
#define BASICFORMAT_YESNO "Yes/No"
#define BASICFORMAT_TRUEFALSE "True/False"
#define BASICFORMAT_ONOFF "On/Off"
// Comment: Visual-Basic has a maximum of 12 positions after the
// decimal point for floating-point-numbers.
// all format-strings are compatible to Visual-Basic:
#define GENERALNUMBER_FORMAT "0.############"
#define FIXED_FORMAT "0.00"
#define STANDARD_FORMAT "@0.00"
#define PERCENT_FORMAT "0.00%"
#define SCIENTIFIC_FORMAT "#.00E+00"
// Comment: the character @ means that thousand-separators shall
// be generated. That's a StarBasic 'extension'.
double get_number_of_digits( double dNumber )
//double floor_log10_fabs( double dNumber )
{
if( dNumber==0.0 )
return 0.0; // used to be 1.0, now 0.0 because of #40025;
else
return floor( log10( fabs( dNumber ) ) );
}
SbxBasicFormater::SbxBasicFormater( sal_Unicode _cDecPoint, sal_Unicode _cThousandSep,
const OUString& _sOnStrg,
const OUString& _sOffStrg,
const OUString& _sYesStrg,
const OUString& _sNoStrg,
const OUString& _sTrueStrg,
const OUString& _sFalseStrg,
const OUString& _sCurrencyStrg,
const OUString& _sCurrencyFormatStrg )
: cDecPoint(_cDecPoint)
, cThousandSep(_cThousandSep)
, sOnStrg(_sOnStrg)
, sOffStrg(_sOffStrg)
, sYesStrg(_sYesStrg)
, sNoStrg(_sNoStrg)
, sTrueStrg(_sTrueStrg)
, sFalseStrg(_sFalseStrg)
, sCurrencyStrg(_sCurrencyStrg)
, sCurrencyFormatStrg(_sCurrencyFormatStrg)
, dNum(0.0)
, nNumExp(0)
, nExpExp(0)
{
}
// function for ouput of a error-text (for debugging)
// displaces all characters of the string, starting from nStartPos
// for one position to larger indexes, i. e. place for a new
// character (which is to be inserted) is created.
// ATTENTION: the string MUST be long enough!
inline void SbxBasicFormater::ShiftString( OUStringBuffer& sStrg, sal_uInt16 nStartPos )
{
sStrg.remove(nStartPos,1);
}
void SbxBasicFormater::AppendDigit( OUStringBuffer& sStrg, short nDigit )
{
if( nDigit>=0 && nDigit<=9 )
{
sStrg.append((sal_Unicode)(nDigit+ASCII_0));
}
}
void SbxBasicFormater::LeftShiftDecimalPoint( OUStringBuffer& sStrg )
{
sal_Int32 nPos = -1;
for(sal_Int32 i = 0; i < sStrg.getLength(); i++)
{
if(sStrg[i] == cDecPoint)
{
nPos = i;
break;
}
}
if( nPos >= 0 )
{
sStrg[nPos] = sStrg[nPos - 1];
sStrg[nPos - 1] = cDecPoint;
}
}
// returns a flag if rounding a 9
void SbxBasicFormater::StrRoundDigit( OUStringBuffer& sStrg, short nPos, bool& bOverflow )
{
if( nPos<0 )
{
return;
}
bOverflow = false;
sal_Unicode c = sStrg[nPos];
if( nPos > 0 && (c == cDecPoint || c == cThousandSep) )
{
StrRoundDigit( sStrg, nPos - 1, bOverflow );
// CHANGE from 9.3.1997: end the method immediately after recursive call!
return;
}
// skip non-digits:
// COMMENT:
// in a valid format-string the number's output should be done
// in one piece, i. e. special characters should ONLY be in
// front OR behind the number and not right in the middle of
// the format information for the number
while( nPos >= 0 && ( sStrg[nPos] < ASCII_0 || sStrg[nPos] > ASCII_9 ))
{
nPos--;
}
if( nPos==-1 )
{
ShiftString( sStrg, 0 );
sStrg[0] = (sal_Unicode)'1';
bOverflow = true;
}
else
{
sal_Unicode c2 = sStrg[nPos];
if( c2 >= ASCII_0 && c2 <= ASCII_9 )
{
if( c2 == ASCII_9 )
{
sStrg[nPos] = (sal_Unicode)'0';
StrRoundDigit( sStrg, nPos - 1, bOverflow );
}
else
{
sStrg[nPos] = c2 + 1;
}
}
else
{
ShiftString( sStrg,nPos+1 );
sStrg[nPos + 1] = (sal_Unicode)'1';
bOverflow = true;
}
}
}
void SbxBasicFormater::StrRoundDigit( OUStringBuffer& sStrg, short nPos )
{
bool bOverflow;
StrRoundDigit( sStrg, nPos, bOverflow );
}
void SbxBasicFormater::ParseBack( OUStringBuffer& sStrg, const OUString& sFormatStrg,
short nFormatPos )
{
for( sal_Int32 i = nFormatPos;
i>0 && sFormatStrg[ i ] == (sal_Unicode)'#' && sStrg[sStrg.getLength() - 1] == (sal_Unicode)'0';
i-- )
{
sStrg.setLength(sStrg.getLength() - 1 );
}
}
#ifdef with_sprintf_
void SbxBasicFormater::InitScan( double _dNum )
{
char sBuffer[ MAX_DOUBLE_BUFFER_LENGTH ];
dNum = _dNum;
InitExp( get_number_of_digits( dNum ) );
// maximum of 15 positions behind the decimal point, example: -1.234000000000000E-001
/*int nCount =*/ sprintf( sBuffer,"%+22.15lE",dNum );
sSciNumStrg = OUString::createFromAscii( sBuffer );
}
void SbxBasicFormater::InitExp( double _dNewExp )
{
char sBuffer[ MAX_DOUBLE_BUFFER_LENGTH ];
nNumExp = (short)_dNewExp;
/*int nCount =*/ sprintf( sBuffer,"%+i",nNumExp );
sNumExpStrg = OUString::createFromAscii( sBuffer );
nExpExp = (short)get_number_of_digits( (double)nNumExp );
}
short SbxBasicFormater::GetDigitAtPosScan( short nPos, bool& bFoundFirstDigit )
{
// trying to read a higher digit,
// e. g. position 4 in 1.234,
// or to read a digit outside of the
// number's dissolution (double)
if( nPos>nNumExp || abs(nNumExp-nPos)>MAX_NO_OF_DIGITS )
{
return NO_DIGIT_;
}
// determine the index of the position in the number-string:
// skip the leading sign
sal_uInt16 no = 1;
// skip the decimal point if necessary
if( nPos<nNumExp )
no++;
no += nNumExp-nPos;
// query of the number's first valid digit --> set flag
if( nPos==nNumExp )
bFoundFirstDigit = true;
return (short)(sSciNumStrg[ no ] - ASCII_0);
}
short SbxBasicFormater::GetDigitAtPosExpScan( short nPos, bool& bFoundFirstDigit )
{
if( nPos>nExpExp )
return -1;
sal_uInt16 no = 1;
no += nExpExp-nPos;
if( nPos==nExpExp )
bFoundFirstDigit = true;
return (short)(sNumExpStrg[ no ] - ASCII_0);
}
// a value for the exponent can be given because the number maybe shall
// not be displayed in a normed way (e. g. 1.2345e-03) but maybe 123.345e-3 !
short SbxBasicFormater::GetDigitAtPosExpScan( double dNewExponent, short nPos,
bool& bFoundFirstDigit )
{
InitExp( dNewExponent );
return GetDigitAtPosExpScan( nPos,bFoundFirstDigit );
}
#else
/* Problems with the following method:
TODO: an 'intelligent' peek-parser might be needed to detect rounding
mistakes at double-numbers - e. g. for 0.00115 #.#e-000
problem with: format( 0.3345 , "0.000" )
problem with: format( 0.00115 , "0.0000" )
*/
// returns the digit at the given '10 system'-position,
// i. e. positive nPos for positions before the decimal
// point and negative for positions after.
// nPos==0 means first position after the decimalpoint, so 10^0.
// returns 0..9 for valid digits and -1 for not existing,
// i. e. if the passed number is too small
// (e. g. position 5 of dNumber=123).
// Furthermore in dNextNumber the number shorted by leading
// positions (till nPos) is returned, e. g.
// GetDigitAtPos( 3434.565 , 2 , dNewNumber ) --> dNewNumber = 434.565
// In bFoundFirstDigit a flag is set if a digit has been found,
// this is used to prevent 'errors' on parsing 202
// ATTENTION: apparently there are sometimes still problems with rounding mistakes!
short SbxBasicFormater::GetDigitAtPos( double dNumber, short nPos,
double& dNextNumber, bool& bFoundFirstDigit )
{
double dDigit;
short nMaxDigit;
dNumber = fabs( dNumber );
nMaxDigit = (short)get_number_of_digits( dNumber );
// error only at numbers > 0, i. e. for digits before
// the decimal point
if( nMaxDigit<nPos && !bFoundFirstDigit && nPos>=0 )
return NO_DIGIT_;
bFoundFirstDigit = true;
for( short i=nMaxDigit; i>=nPos; i-- )
{
double dI = (double)i;
double dTemp1 = pow( 10.0,dI );
dDigit = floor( pow( 10.0,log10( fabs( dNumber ) )-dI ) );
dNumber -= dTemp1 * dDigit;
}
// for optimized loop run
dNextNumber = dNumber;
return RoundDigit( dDigit );
}
short SbxBasicFormater::RoundDigit( double dNumber )
{
if( dNumber<0.0 || dNumber>10.0 )
return -1;
short nTempHigh = (short)(dNumber+0.5); // maybe floor( )
return nTempHigh;
}
#endif
// Copies the respective part of the format-string, if existing, and returns it.
// So a new string is created, which has to be freed by the caller later.
OUString SbxBasicFormater::GetPosFormatString( const OUString& sFormatStrg, bool & bFound )
{
bFound = false; // default...
sal_Int32 nPos = sFormatStrg.indexOf( FORMAT_SEPARATOR );
if( nPos >= 0 )
{
bFound = true;
// the format-string for positive numbers is
// everything before the first ';'
return sFormatStrg.copy( 0,nPos );
}
return OUString();
}
// see also GetPosFormatString()
OUString SbxBasicFormater::GetNegFormatString( const OUString& sFormatStrg, bool & bFound )
{
bFound = false; // default...
sal_Int32 nPos = sFormatStrg.indexOf( FORMAT_SEPARATOR );
if( nPos >= 0)
{
// the format-string for negative numbers is
// everything between the first and the second ';'
OUString sTempStrg = sFormatStrg.copy( nPos+1 );
nPos = sTempStrg.indexOf( FORMAT_SEPARATOR );
bFound = true;
if( nPos < 0 )
{
return sTempStrg;
}
else
{
return sTempStrg.copy( 0,nPos );
}
}
return OUString();
}
// see also GetPosFormatString()
OUString SbxBasicFormater::Get0FormatString( const OUString& sFormatStrg, bool & bFound )
{
bFound = false; // default...
sal_Int32 nPos = sFormatStrg.indexOf( FORMAT_SEPARATOR );
if( nPos >= 0 )
{
// the format string for the zero is
// everything after the second ';'
OUString sTempStrg = sFormatStrg.copy( nPos+1 );
nPos = sTempStrg.indexOf( FORMAT_SEPARATOR );
if( nPos >= 0 )
{
bFound = true;
sTempStrg = sTempStrg.copy( nPos+1 );
nPos = sTempStrg.indexOf( FORMAT_SEPARATOR );
if( nPos < 0 )
{
return sTempStrg;
}
else
{
return sTempStrg.copy( 0,nPos );
}
}
}
return OUString();
}
// see also GetPosFormatString()
OUString SbxBasicFormater::GetNullFormatString( const OUString& sFormatStrg, bool & bFound )
{
bFound = false; // default...
sal_Int32 nPos = sFormatStrg.indexOf( FORMAT_SEPARATOR );
if( nPos >= 0 )
{
// the format-string for the Null is
// everything after the third ';'
OUString sTempStrg = sFormatStrg.copy( nPos+1 );
nPos = sTempStrg.indexOf( FORMAT_SEPARATOR );
if( nPos >= 0 )
{
sTempStrg = sTempStrg.copy( nPos+1 );
nPos = sTempStrg.indexOf( FORMAT_SEPARATOR );
if( nPos >= 0 )
{
bFound = true;
return sTempStrg.copy( nPos+1 );
}
}
}
return OUString();
}
// returns value <> 0 in case of an error
void SbxBasicFormater::AnalyseFormatString( const OUString& sFormatStrg,
short& nNoOfDigitsLeft, short& nNoOfDigitsRight,
short& nNoOfOptionalDigitsLeft,
short& nNoOfExponentDigits, short& nNoOfOptionalExponentDigits,
bool& bPercent, bool& bCurrency, bool& bScientific,
bool& bGenerateThousandSeparator,
short& nMultipleThousandSeparators )
{
sal_Int32 nLen;
short nState = 0;
nLen = sFormatStrg.getLength();
nNoOfDigitsLeft = 0;
nNoOfDigitsRight = 0;
nNoOfOptionalDigitsLeft = 0;
nNoOfExponentDigits = 0;
nNoOfOptionalExponentDigits = 0;
bPercent = false;
bCurrency = false;
bScientific = false;
// from 11.7.97: as soon as a comma (point?) is found in the format string,
// all three decimal powers are marked (i. e. thousand, million, ...)
bGenerateThousandSeparator = sFormatStrg.indexOf( ',' ) >= 0;
nMultipleThousandSeparators = 0;
for( sal_Int32 i = 0; i < nLen; i++ )
{
sal_Unicode c = sFormatStrg[ i ];
switch( c )
{
case '#':
case '0':
if( nState==0 )
{
nNoOfDigitsLeft++;
// TODO here maybe better error inspection of the mantissa for valid syntax (see grammar)h
// ATTENTION: 'undefined' behaviour if # and 0 are combined!
// REMARK: #-placeholders are actually useless for
// scientific display before the decimal point!
if( c=='#' )
{
nNoOfOptionalDigitsLeft++;
}
}
else if( nState==1 )
{
nNoOfDigitsRight++;
}
else if( nState==-1 ) // search 0 in the exponent
{
if( c=='#' ) // # switches on the condition
{
nNoOfOptionalExponentDigits++;
nState = -2;
}
nNoOfExponentDigits++;
}
else if( nState==-2 ) // search # in the exponent
{
if( c=='0' )
{
// ERROR: 0 after # in the exponent is NOT allowed!!
return;
}
nNoOfOptionalExponentDigits++;
nNoOfExponentDigits++;
}
break;
case '.':
nState++;
if( nState>1 )
{
return; // ERROR: too many decimal points
}
break;
case '%':
bPercent = true;
break;
case '(':
bCurrency = true;
break;
case ',':
{
sal_Unicode ch = sFormatStrg[ i+1 ];
if( ch!=0 && (ch==',' || ch=='.') )
{
nMultipleThousandSeparators++;
}
}
break;
case 'e':
case 'E':
// #i13821 not when no digits before
if( nNoOfDigitsLeft > 0 || nNoOfDigitsRight > 0 )
{
nState = -1; // abort counting digits
bScientific = true;
}
break;
// OWN command-character which turns on
// the creation of thousand-separators
case '\\':
// Ignore next char
i++;
break;
case CREATE_1000SEP_CHAR:
bGenerateThousandSeparator = true;
break;
}
}
}
// the flag bCreateSign says that at the mantissa a leading sign
// shall be created
void SbxBasicFormater::ScanFormatString( double dNumber,
const OUString& sFormatStrg, OUString& sReturnStrgFinal,
bool bCreateSign )
{
short /*nErr,*/nNoOfDigitsLeft,nNoOfDigitsRight,nNoOfOptionalDigitsLeft,
nNoOfExponentDigits,nNoOfOptionalExponentDigits,
nMultipleThousandSeparators;
bool bPercent,bCurrency,bScientific,bGenerateThousandSeparator;
OUStringBuffer sReturnStrg = OUStringBuffer();
// analyse the format-string, i. e. determine the following values:
/*
- number of digits before decimal point
- number of digits after decimal point
- optional digits before decimal point
- number of digits in the exponent
- optional digits in the exponent
- percent-character found?
- () for negative leading sign?
- exponential-notation?
- shall thousand-separators be generated?
- is a percent-character being found? --> dNumber *= 100.0;
- are there thousand-separators in a row?
,, or ,. --> dNumber /= 1000.0;
- other errors? multiple decimal points, E's, etc.
--> errors are simply ignored at the moment
*/
AnalyseFormatString( sFormatStrg, nNoOfDigitsLeft, nNoOfDigitsRight,
nNoOfOptionalDigitsLeft, nNoOfExponentDigits,
nNoOfOptionalExponentDigits,
bPercent, bCurrency, bScientific,
bGenerateThousandSeparator, nMultipleThousandSeparators );
// special handling for special characters
if( bPercent )
{
dNumber *= 100.0;
}
// TODO: this condition (,, or ,.) is NOT Visual-Basic compatible!
// Question: shall this stay here (requirements)?
if( nMultipleThousandSeparators )
{
dNumber /= 1000.0;
}
double dExponent;
short i,nLen;
short nState,nDigitPos,nExponentPos,nMaxDigit,nMaxExponentDigit;
bool bFirstDigit,bFirstExponentDigit,bFoundFirstDigit,
bIsNegative,bZeroSpaceOn, bSignHappend,bDigitPosNegative;
bSignHappend = false;
bFoundFirstDigit = false;
bIsNegative = dNumber < 0.0;
nLen = sFormatStrg.getLength();
dExponent = get_number_of_digits( dNumber );
nExponentPos = 0;
nMaxExponentDigit = 0;
nMaxDigit = (short)dExponent;
bDigitPosNegative = false;
if( bScientific )
{
dExponent = dExponent - (double)(nNoOfDigitsLeft-1);
nDigitPos = nMaxDigit;
nMaxExponentDigit = (short)get_number_of_digits( dExponent );
nExponentPos = nNoOfExponentDigits - 1 - nNoOfOptionalExponentDigits;
}
else
{
nDigitPos = nNoOfDigitsLeft - 1; // counting starts at 0, 10^0
// no exponent-data is needed here!
bDigitPosNegative = (nDigitPos < 0);
}
bFirstDigit = true;
bFirstExponentDigit = true;
nState = 0; // 0 --> mantissa; 1 --> exponent
bZeroSpaceOn = false;
#ifdef with_sprintf_
InitScan( dNumber );
#endif
// scanning the format-string:
sal_Unicode cForce = 0;
for( i = 0; i < nLen; i++ )
{
sal_Unicode c;
if( cForce )
{
c = cForce;
cForce = 0;
}
else
{
c = sFormatStrg[ i ];
}
switch( c )
{
case '0':
case '#':
if( nState==0 )
{
// handling of the mantissa
if( bFirstDigit )
{
// remark: at bCurrency the negative
// leading sign shall be shown with ()
if( bIsNegative && !bCreateSign && !bSignHappend )
{
bSignHappend = true;
sReturnStrg.append('-');
}
// output redundant positions, i. e. those which
// are undocumented by the format-string
if( nMaxDigit > nDigitPos )
{
for( short j = nMaxDigit; j > nDigitPos; j-- )
{
short nTempDigit;
#ifdef with_sprintf_
AppendDigit( sReturnStrg, nTempDigit = GetDigitAtPosScan( j, bFoundFirstDigit ) );
#else
AppendDigit( sReturnStrg, nTempDigit = GetDigitAtPos( dNumber, j, dNumber, bFoundFirstDigit ) );
#endif
if( nTempDigit != NO_DIGIT_ )
{
bFirstDigit = false;
}
if( bGenerateThousandSeparator && ( c=='0' || nMaxDigit >= nDigitPos ) && j > 0 && (j % 3 == 0) )
{
sReturnStrg.append(cThousandSep );
}
}
}
}
if( nMaxDigit<nDigitPos && ( c=='0' || bZeroSpaceOn ) )
{
AppendDigit( sReturnStrg, 0 );
bFirstDigit = false;
bZeroSpaceOn = true;
// Remark: in Visual-Basic the first 0 turns on the 0 for
// all the following # (up to the decimal point),
// this behaviour is simulated here with the flag.
if( bGenerateThousandSeparator && ( c=='0' || nMaxDigit >= nDigitPos ) && nDigitPos > 0 && (nDigitPos % 3 == 0) )
{
sReturnStrg.append(cThousandSep);
}
}
else
{
short nTempDigit;
#ifdef with_sprintf_
AppendDigit( sReturnStrg, nTempDigit = GetDigitAtPosScan( nDigitPos, bFoundFirstDigit ) );
#else
AppendDigit( sReturnStrg, nTempDigit = GetDigitAtPos( dNumber, nDigitPos, dNumber, bFoundFirstDigit ) );
#endif
if( nTempDigit != NO_DIGIT_ )
{
bFirstDigit = false;
}
if( bGenerateThousandSeparator && ( c=='0' || nMaxDigit>=nDigitPos ) && nDigitPos>0 && (nDigitPos % 3 == 0) )
{
sReturnStrg.append(cThousandSep);
}
}
nDigitPos--;
}
else
{
// handling the exponent
if( bFirstExponentDigit )
{
// leading sign has been given out at e/E already
bFirstExponentDigit = false;
if( nMaxExponentDigit > nExponentPos )
// output redundant positions, i. e. those which
// are undocumented by the format-string
{
for( short j = nMaxExponentDigit; j > nExponentPos; j-- )
{
#ifdef with_sprintf_
AppendDigit( sReturnStrg, GetDigitAtPosExpScan( dExponent, j, bFoundFirstDigit ) );
#else
AppendDigit( sReturnStrg,GetDigitAtPos( dExponent, j, dExponent, bFoundFirstDigit ) );
#endif
}
}
}
if( nMaxExponentDigit < nExponentPos && c=='0' )
{
AppendDigit( sReturnStrg, 0 );
}
else
{
#ifdef with_sprintf_
AppendDigit( sReturnStrg, GetDigitAtPosExpScan( dExponent, nExponentPos, bFoundFirstDigit ) );
#else
AppendDigit( sReturnStrg, GetDigitAtPos( dExponent, nExponentPos, dExponent, bFoundFirstDigit ) );
#endif
}
nExponentPos--;
}
break;
case '.':
if( bDigitPosNegative ) // #i13821: If no digits before .
{
bDigitPosNegative = false;
nDigitPos = 0;
cForce = '#';
i-=2;
break;
}
sReturnStrg.append(cDecPoint);
break;
case '%':
// maybe remove redundant 0s, e. g. 4.500e4 in 0.0##e-00
ParseBack( sReturnStrg, sFormatStrg, i-1 );
sReturnStrg.append('%');
break;
case 'e':
case 'E':
// does mantissa have to be rounded, before the exponent is displayed?
{
// is there a mantissa at all?
if( bFirstDigit )
{
// apparently not, i. e. invalid format string, e. g. E000.00
// so ignore these e and E characters
// maybe output an error (like in Visual Basic)?
// #i13821: VB 6 behaviour
sReturnStrg.append(c);
break;
}
bool bOverflow = false;
#ifdef with_sprintf_
short nNextDigit = GetDigitAtPosScan( nDigitPos, bFoundFirstDigit );
#else
short nNextDigit = GetDigitAtPos( dNumber, nDigitPos, dNumber, bFoundFirstDigit );
#endif
if( nNextDigit>=5 )
{
StrRoundDigit( sReturnStrg, sReturnStrg.getLength() - 1, bOverflow );
}
if( bOverflow )
{
// a leading 9 has been rounded
LeftShiftDecimalPoint( sReturnStrg );
sReturnStrg[sReturnStrg.getLength() - 1] = 0;
dExponent += 1.0;
}
// maybe remove redundant 0s, e. g. 4.500e4 in 0.0##e-00
ParseBack( sReturnStrg, sFormatStrg, i-1 );
}
// change the scanner's condition
nState++;
// output exponent character
sReturnStrg.append(c);
// i++; // MANIPULATION of the loop-variable!
c = sFormatStrg[ ++i ];
// output leading sign / exponent
if( c != 0 )
{
if( c == '-' )
{
if( dExponent < 0.0 )
{
sReturnStrg.append('-');
}
}
else if( c == '+' )
{
if( dExponent < 0.0 )
{
sReturnStrg.append('-');
}
else
{
sReturnStrg.append('+');
}
}
}
break;
case ',':
break;
case ';':
break;
case '(':
case ')':
// maybe remove redundant 0s, e. g. 4.500e4 in 0.0##e-00
ParseBack( sReturnStrg, sFormatStrg, i-1 );
if( bIsNegative )
{
sReturnStrg.append(c);
}
break;
case '$':
// append the string for the currency:
sReturnStrg.append(sCurrencyStrg);
break;
case ' ':
case '-':
case '+':
ParseBack( sReturnStrg, sFormatStrg, i-1 );
sReturnStrg.append(c);
break;
case '\\':
ParseBack( sReturnStrg, sFormatStrg, i-1 );
// special character found, output next
// character directly (if existing)
c = sFormatStrg[ ++i ];
if( c!=0 )
{
sReturnStrg.append(c);
}
break;
case CREATE_1000SEP_CHAR:
// ignore here, action has already been
// executed in AnalyseFormatString
break;
default:
// output characters and digits, too (like in Visual-Basic)
if( ( c>='a' && c<='z' ) ||
( c>='A' && c<='Z' ) ||
( c>='1' && c<='9' ) )
{
sReturnStrg.append(c);
}
}
}
// scan completed - rounding necessary?
if( !bScientific )
{
#ifdef with_sprintf_
short nNextDigit = GetDigitAtPosScan( nDigitPos, bFoundFirstDigit );
#else
short nNextDigit = GetDigitAtPos( dNumber, nDigitPos, dNumber, bFoundFirstDigit );
#endif
if( nNextDigit>=5 )
{
StrRoundDigit( sReturnStrg, sReturnStrg.getLength() - 1 );
}
}
if( nNoOfDigitsRight>0 )
{
ParseBack( sReturnStrg, sFormatStrg, sFormatStrg.getLength()-1 );
}
sReturnStrgFinal = sReturnStrg.makeStringAndClear();
}
OUString SbxBasicFormater::BasicFormatNull( const OUString& sFormatStrg )
{
bool bNullFormatFound;
OUString sNullFormatStrg = GetNullFormatString( sFormatStrg, bNullFormatFound );
if( bNullFormatFound )
{
return sNullFormatStrg;
}
return OUString("null");
}
OUString SbxBasicFormater::BasicFormat( double dNumber, const OUString& _sFormatStrg )
{
bool bPosFormatFound,bNegFormatFound,b0FormatFound;
OUString sFormatStrg = _sFormatStrg;
// analyse format-string concerning predefined formats:
if( sFormatStrg.equalsIgnoreAsciiCase( BASICFORMAT_GENERALNUMBER ) )
{
sFormatStrg = GENERALNUMBER_FORMAT;
}
if( sFormatStrg.equalsIgnoreAsciiCase( BASICFORMAT_CURRENCY ) )
{
sFormatStrg = sCurrencyFormatStrg;
}
if( sFormatStrg.equalsIgnoreAsciiCase( BASICFORMAT_FIXED ) )
{
sFormatStrg = FIXED_FORMAT;
}
if( sFormatStrg.equalsIgnoreAsciiCase( BASICFORMAT_STANDARD ) )
{
sFormatStrg = STANDARD_FORMAT;
}
if( sFormatStrg.equalsIgnoreAsciiCase( BASICFORMAT_PERCENT ) )
{
sFormatStrg = PERCENT_FORMAT;
}
if( sFormatStrg.equalsIgnoreAsciiCase( BASICFORMAT_SCIENTIFIC ) )
{
sFormatStrg = SCIENTIFIC_FORMAT;
}
if( sFormatStrg.equalsIgnoreAsciiCase( BASICFORMAT_YESNO ) )
{
return ( dNumber==0.0 ) ? sNoStrg : sYesStrg ;
}
if( sFormatStrg.equalsIgnoreAsciiCase( BASICFORMAT_TRUEFALSE ) )
{
return ( dNumber==0.0 ) ? sFalseStrg : sTrueStrg ;
}
if( sFormatStrg.equalsIgnoreAsciiCase( BASICFORMAT_ONOFF ) )
{
return ( dNumber==0.0 ) ? sOffStrg : sOnStrg ;
}
// analyse format-string concerning ';', i. e. format-strings for
// positive-, negative- and 0-values
OUString sPosFormatStrg = GetPosFormatString( sFormatStrg, bPosFormatFound );
OUString sNegFormatStrg = GetNegFormatString( sFormatStrg, bNegFormatFound );
OUString s0FormatStrg = Get0FormatString( sFormatStrg, b0FormatFound );
OUString sReturnStrg;
OUString sTempStrg;
if( dNumber==0.0 )
{
sTempStrg = sFormatStrg;
if( b0FormatFound )
{
if( s0FormatStrg.isEmpty() && bPosFormatFound )
{
sTempStrg = sPosFormatStrg;
}
else
{
sTempStrg = s0FormatStrg;
}
}
else if( bPosFormatFound )
{
sTempStrg = sPosFormatStrg;
}
ScanFormatString( dNumber, sTempStrg, sReturnStrg,/*bCreateSign=*/false );
}
else
{
if( dNumber<0.0 )
{
if( bNegFormatFound )
{
if( sNegFormatStrg.isEmpty() && bPosFormatFound )
{
sTempStrg = "-";
sTempStrg += sPosFormatStrg;
}
else
{
sTempStrg = sNegFormatStrg;
}
}
else
{
sTempStrg = sFormatStrg;
}
// if NO format-string especially for negative
// values is given, output the leading sign
ScanFormatString( dNumber, sTempStrg, sReturnStrg,/*bCreateSign=*/bNegFormatFound/*sNegFormatStrg!=EMPTYFORMATSTRING*/ );
}
else // if( dNumber>0.0 )
{
ScanFormatString( dNumber,
(/*sPosFormatStrg!=EMPTYFORMATSTRING*/bPosFormatFound ? sPosFormatStrg : sFormatStrg),
sReturnStrg,/*bCreateSign=*/false );
}
}
return sReturnStrg;
}
bool SbxBasicFormater::isBasicFormat( const OUString& sFormatStrg )
{
if( sFormatStrg.equalsIgnoreAsciiCase( BASICFORMAT_GENERALNUMBER ) )
{
return true;
}
if( sFormatStrg.equalsIgnoreAsciiCase( BASICFORMAT_CURRENCY ) )
{
return true;
}
if( sFormatStrg.equalsIgnoreAsciiCase( BASICFORMAT_FIXED ) )
{
return true;
}
if( sFormatStrg.equalsIgnoreAsciiCase( BASICFORMAT_STANDARD ) )
{
return true;
}
if( sFormatStrg.equalsIgnoreAsciiCase( BASICFORMAT_PERCENT ) )
{
return true;
}
if( sFormatStrg.equalsIgnoreAsciiCase( BASICFORMAT_SCIENTIFIC ) )
{
return true;
}
if( sFormatStrg.equalsIgnoreAsciiCase( BASICFORMAT_YESNO ) )
{
return true;
}
if( sFormatStrg.equalsIgnoreAsciiCase( BASICFORMAT_TRUEFALSE ) )
{
return true;
}
if( sFormatStrg.equalsIgnoreAsciiCase( BASICFORMAT_ONOFF ) )
{
return true;
}
return false;
}
/* vim:set shiftwidth=4 softtabstop=4 expandtab: */
|