Warning, /utilities/kregexpeditor/src/qregexpparser.l is written in an unsupported language. File is not indexed.
0001 /*
0002 * Copyright (c) 2002-2003 Jesper K. Pedersen <blackie@kde.org>
0003 *
0004 * This library is free software; you can redistribute it and/or
0005 * modify it under the terms of the GNU Library General Public
0006 * License version 2 as published by the Free Software Foundation.
0007 *
0008 * This library is distributed in the hope that it will be useful,
0009 * but WITHOUT ANY WARRANTY; without even the implied warranty of
0010 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
0011 * Library General Public License for more details.
0012 *
0013 * You should have received a copy of the GNU Library General Public License
0014 * along with this library; see the file COPYING.LIB. If not, write to
0015 * the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
0016 * Boston, MA 02110-1301, USA.
0017 **/
0018 %option noyywrap
0019 %option nounput
0020 %option reentrant
0021 %option bison-bridge
0022
0023 %{
0024
0025 #include "textrangeregexp.h"
0026 #include "gen_qregexpparser.hh"
0027 #include "qregexpparsercommon.h"
0028
0029 void parseRange( const char* txt, int* min, int* max );
0030 RegExp* parseCharClass( const char* match );
0031 %}
0032
0033 Escape \\.
0034 BackRef \\[1-9][0-9]*
0035 CharClass \[^?\]?[^]]*\]
0036 Range \{[0-9]*(,[0-9]*)?\}
0037 HexChar \\x[0-9a-fA-F]{1,4}
0038 OctChar \\0[0-7]{1,4}
0039 SpecialEsc \\[afnrtv]
0040 %%
0041 "\\b" return TOK_PosWordChar;
0042 "\\B" return TOK_PosNonWordChar;
0043 "\\d" {
0044 TextRangeRegExp* regexp = new TextRangeRegExp( false );
0045 regexp->setDigit( true );
0046 yylval->regexp = regexp;
0047 return TOK_CharClass;
0048 }
0049 "\\D" {
0050 TextRangeRegExp* regexp = new TextRangeRegExp( false );
0051 regexp->setNonDigit( true );
0052 yylval->regexp = regexp;
0053 return TOK_CharClass;
0054 }
0055 "\\s" {
0056 TextRangeRegExp* regexp = new TextRangeRegExp( false );
0057 regexp->setSpace( true );
0058 yylval->regexp = regexp;
0059 return TOK_CharClass;
0060 }
0061 "\\S" {
0062 TextRangeRegExp* regexp = new TextRangeRegExp( false );
0063 regexp->setNonSpace( true );
0064 yylval->regexp = regexp;
0065 return TOK_CharClass;
0066 }
0067 "\\w" {
0068 TextRangeRegExp* regexp = new TextRangeRegExp( false );
0069 regexp->setWordChar( true );
0070 yylval->regexp = regexp;
0071 return TOK_CharClass;
0072 }
0073 "\\W" {
0074 TextRangeRegExp* regexp = new TextRangeRegExp( false );
0075 regexp->setNonWordChar( true );
0076 yylval->regexp = regexp;
0077 return TOK_CharClass;
0078 }
0079 {SpecialEsc} {
0080 TextRangeRegExp* regexp = new TextRangeRegExp( false );
0081 regexp->addCharacter( QString::fromLocal8Bit( yytext ) );
0082 yylval->regexp = regexp;
0083 return TOK_CharClass;
0084 }
0085
0086 {HexChar} {
0087 TextRangeRegExp* regexp = new TextRangeRegExp( false );
0088 regexp->addCharacter( QString::fromLocal8Bit(yytext) );
0089 yylval->regexp = regexp;
0090 return TOK_CharClass;
0091 }
0092 {OctChar} {
0093 TextRangeRegExp* regexp = new TextRangeRegExp( false );
0094 regexp->addCharacter( QString::fromLocal8Bit(yytext) );
0095 yylval->regexp = regexp;
0096 return TOK_CharClass;
0097 }
0098 "." return TOK_Dot;
0099 "$" return TOK_Dollar;
0100 "^" return TOK_Carat;
0101 "(?:" return TOK_MagicLeftParent;
0102 "(?=" return TOK_PosLookAhead;
0103 "(?!" return TOK_NegLookAhead;
0104 "(" return TOK_LeftParen;
0105 ")" return TOK_RightParent;
0106 "|" return TOK_Bar;
0107 "*" { yylval->range.min = 0; yylval->range.max=-1; return TOK_Quantifier; }
0108 "?" { yylval->range.min = 0; yylval->range.max=1; return TOK_Quantifier; }
0109 "+" { yylval->range.min = 1; yylval->range.max=-1; return TOK_Quantifier; }
0110 {Range} { parseRange( yytext, &yylval->range.min, &yylval->range.max ); return TOK_Quantifier; }
0111 {CharClass} { yylval->regexp = parseCharClass(yytext); return TOK_CharClass; }
0112 {BackRef} { yylval->backRef = atoi( yytext+1 ); return TOK_BackRef; }
0113 {Escape} { yylval->ch = yytext[1]; return TOK_EscapeChar; }
0114 . { yylval->ch = yytext[0]; return TOK_Char; }
0115
0116 %%
0117
0118 void scannerInit( yyscan_t *scanner, struct parse_context *context, const QString& qstr )
0119 {
0120 yylex_init( scanner );
0121 yyset_extra( context, *scanner );
0122 QByteArray cstr;
0123 if ( !qstr.isNull() )
0124 cstr = qstr.toLatin1();
0125 yy_switch_to_buffer( yy_scan_string( cstr.constData(), *scanner ), *scanner );
0126 }
0127
0128 void scannerDestroy( yyscan_t scanner )
0129 {
0130 yylex_destroy( scanner );
0131 }
0132
0133 /**
0134 This function parses a range in a form similar to "{3,4}", "{,7}"
0135 etc. and returns the value in the integers pointed to by min and max.
0136 */
0137 void parseRange( const char* txt, int* min, int* max )
0138 {
0139
0140 /*
0141 case txt min max
0142 1 {} 0 -1
0143 2 {,} 0 -1
0144 3 {5} 5 5
0145 4 {5,} 5 -1
0146 5 {,7} 0 7
0147 6 {5,7} 5 7
0148 */
0149 char c;
0150 int i = 1;
0151 int minimum=0, maximum=0;
0152 int minFound=0, maxFound=0, commaFound = 0;
0153
0154 while ( (c = txt[i++]) != ',' && c != '}') {
0155 minimum = minimum*10+ c-'0';
0156 minFound=1;
0157 }
0158
0159 if ( c == ',' )
0160 commaFound = 1;
0161
0162 if ( c != '}' ) {
0163 while ( (c = txt[i++]) != '}') {
0164 maximum = maximum*10+ c-'0';
0165 maxFound = 1;
0166 }
0167 }
0168
0169 *min = minimum;
0170 if ( maxFound )
0171 *max = maximum; /* case 5,6 */
0172 else if ( !minFound )
0173 *max = -1; /* case 1,2 */
0174 else if ( commaFound )
0175 *max = -1; /* case 4 */
0176 else
0177 *max = minimum; /* case 3 */
0178 }
0179
0180
0181 /**
0182 This function parses a character range like "[^ab1-4]".
0183 */
0184 RegExp* parseCharClass( const char* match )
0185 {
0186 TextRangeRegExp* res = new TextRangeRegExp( false );
0187 QString txt = QString::fromLocal8Bit( match );
0188 if(txt.length() <= 2)
0189 return res;
0190 txt = txt.mid(1,txt.length()-2);
0191
0192 int i = 0;
0193 QChar ch = txt.at(i++);
0194 QString pendingChar;
0195 QString thisChar;
0196 bool charPending = false;
0197 bool rangePending = false;
0198 bool flushPending = false;
0199
0200 if ( i < txt.length() && ch == QLatin1Char('^') ) {
0201 res->setNegate( true );
0202 ch = txt.at(i++);
0203 }
0204
0205 do {
0206 // If a character is pending, and the next char is '-' then we are
0207 // possible looking at a range.
0208 if ( i < txt.length() && ch == QLatin1Char('-') && charPending ) {
0209 rangePending = true;
0210 ch = txt.at(i++);
0211 continue;
0212 }
0213
0214 // If we have a pending character, but do not also have a pending
0215 // range, then the pending character was not part of a range, and
0216 // should therefore just be added as a single character.
0217 if ( charPending && !rangePending ) {
0218 res->addCharacter( pendingChar );
0219 charPending = false;
0220 }
0221
0222 if ( ch == QLatin1Char('\\') ) {
0223 // Handle the cases where an escape character is specified.
0224 ch = txt.at(i++);
0225
0226 if ( ch == QLatin1Char('a') || ch == QLatin1Char('f') || ch == QLatin1Char('n') || ch == QLatin1Char('r') || ch == QLatin1Char('t') || ch == QLatin1Char('v') ) {
0227 // These are just seen as normal characters.
0228 thisChar = QString::fromLocal8Bit("\\") + ch;
0229 }
0230 else if ( ch == QLatin1Char('d') ) {
0231 // The following characters represent character groups. If any of
0232 // these are seen in a range, then the range is ignored, thus [a-\s]
0233 // matches an 'a', a '-', and a space (\s means space).
0234 res->setDigit( true );
0235 flushPending = true;
0236 }
0237 else if ( ch == QLatin1Char('D') ) {
0238 res->setNonDigit( true );
0239 flushPending = true;
0240 }
0241 else if ( ch == QLatin1Char('s') ) {
0242 res->setSpace( true );
0243 flushPending = true;
0244 }
0245 else if ( ch == QLatin1Char('S') ) {
0246 res->setNonSpace( true );
0247 flushPending = true;
0248 }
0249 else if ( ch == QLatin1Char('w') ) {
0250 res->setWordChar( true );
0251 flushPending = true;
0252 }
0253 else if ( ch == QLatin1Char('W') ) {
0254 res->setNonWordChar( true );
0255 flushPending = true;
0256 }
0257 else if ( ch == QLatin1Char('x') || ch == QLatin1Char('X') ) {
0258 // This is a hexidecimal character: \xHHHH
0259 QString str;
0260 for ( int j=0; j<4; j++) {
0261 ch = txt.at(i++);
0262 if ( ch == QLatin1Char('a') || ch == QLatin1Char('A') || ch == QLatin1Char('b') || ch == QLatin1Char('B') ||
0263 ch == QLatin1Char('c') || ch == QLatin1Char('C') || ch == QLatin1Char('d') || ch == QLatin1Char('D') ||
0264 ch == QLatin1Char('e') || ch == QLatin1Char('E') || ch == QLatin1Char('f') || ch == QLatin1Char('F') ||
0265 ch == QLatin1Char('0') || ch == QLatin1Char('1') || ch == QLatin1Char('2') || ch == QLatin1Char('3') ||
0266 ch == QLatin1Char('4') || ch == QLatin1Char('5') || ch == QLatin1Char('6') || ch == QLatin1Char('7') ||
0267 ch == QLatin1Char('8') || ch == QLatin1Char('9') )
0268 str += ch;
0269 else
0270 i--;
0271 }
0272 thisChar = QString::fromLocal8Bit("\\x") + str;
0273 }
0274 else if ( ch == QLatin1Char('0') ) {
0275 // This is an octal character
0276 QString str;
0277 for ( int j=0; j<4; j++) {
0278 ch = txt.at(i++);
0279 if ( ch == QLatin1Char('0') || ch == QLatin1Char('1') || ch == QLatin1Char('2') || ch == QLatin1Char('3')
0280 || ch == QLatin1Char('4') || ch == QLatin1Char('5') || ch == QLatin1Char('6') || ch == QLatin1Char('7') )
0281 str += ch;
0282 else
0283 i--;
0284 }
0285 thisChar = QString::fromLocal8Bit("\\x") + str ;
0286 }
0287 else {
0288 // Anything else escaped just means the character itself.
0289 thisChar = ch;
0290 }
0291 }
0292 else {
0293 // A non escaped character.
0294 thisChar = ch;
0295 }
0296
0297 // The characters \s,\S,\w,\W,\d or \D, can not be part of a range,
0298 // thus if they are meet in what looks like a range, then the
0299 // characters of the range is justed seen as normal non range
0300 // characters. thus [a-\s] matches an 'a', a '-', and a space (\s means
0301 // space).
0302 if ( flushPending ) {
0303 if ( charPending )
0304 res->addCharacter( pendingChar );
0305 if ( rangePending )
0306 res->addCharacter( QString::fromLocal8Bit("-") );
0307 flushPending = false;
0308 charPending = false;
0309 rangePending = false;
0310 }
0311 else {
0312 if ( rangePending ) {
0313 res->addRange( pendingChar, thisChar );
0314 charPending = false;
0315 rangePending = false;
0316 }
0317 else {
0318 pendingChar = thisChar;
0319 charPending = true;
0320 }
0321 }
0322 if ( i == txt.length() )
0323 break;
0324 ch = txt.at(i++);
0325 }
0326 while ( ch != QLatin1Char(']') );
0327
0328 if ( charPending )
0329 res->addCharacter( pendingChar );
0330 if ( rangePending )
0331 res->addCharacter( QString::fromLocal8Bit("-") );
0332
0333 return res;
0334 }