Package org.jcodings

Examples of org.jcodings.Encoding

386
387
388
389
390
391
392
393
394
395
396
397
    static Regex preparePattern(Ruby runtime, Regex pattern, RubyString str) {
        if (str.scanForCodeRange() == StringSupport.CR_BROKEN) {
            throw runtime.newArgumentError("invalid byte sequence in " + str.getEncoding());
        }
        Encoding enc = str.getEncoding();
        if (!enc.isAsciiCompatible()) {
            if (enc != pattern.getEncoding()) encodingMatchError(runtime, pattern, enc);
        }
        // TODO: check for isKCodeDefault() somehow
//        if (warn && isEncodingNone() && enc != ASCIIEncoding.INSTANCE && str.scanForCodeRange() != StringSupport.CR_7BIT) {
//            getRuntime().getWarnings().warn(ID.REGEXP_MATCH_AGAINST_STRING, "regexp match /.../n against to " + enc + " string");
View Full Code Here
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
    static ByteList quote19(ByteList bs, boolean asciiOnly) {
        int p = bs.getBegin();
        int end = p + bs.getRealSize();
        byte[]bytes = bs.getUnsafeBytes();
        Encoding enc = bs.getEncoding();

        metaFound: do {
            while (p < end) {
                final int c;
                final int cl;
                if (enc.isAsciiCompatible()) {
                    cl = 1;
                    c = bytes[p] & 0xff;
                } else {
                    cl = StringSupport.preciseLength(enc, bytes, p, end);
                    c = enc.mbcToCode(bytes, p, end);
                }

                if (!Encoding.isAscii(c)) {
                    p += StringSupport.length(enc, bytes, p, end);
                    continue;
                }
               
                switch (c) {
                case '[': case ']': case '{': case '}':
                case '(': case ')': case '|': case '-':
                case '*': case '.': case '\\':
                case '?': case '+': case '^': case '$':
                case ' ': case '#':
                case '\t': case '\f': case '\n': case '\r':
                    break metaFound;
                }
                p += cl;
            }
            if (asciiOnly) {
                ByteList tmp = bs.shallowDup();
                tmp.setEncoding(USASCIIEncoding.INSTANCE);
                return tmp;
            }
            return bs;
        } while (false);

        ByteList result = new ByteList(end * 2);
        result.setEncoding(asciiOnly ? USASCIIEncoding.INSTANCE : bs.getEncoding());
        byte[]obytes = result.getUnsafeBytes();
        int op = p - bs.getBegin();
        System.arraycopy(bytes, bs.getBegin(), obytes, 0, op);

        while (p < end) {
            final int c;
            final int cl;
            if (enc.isAsciiCompatible()) {
                cl = 1;
                c = bytes[p] & 0xff;
            } else {
                cl = StringSupport.preciseLength(enc, bytes, p, end);
                c = enc.mbcToCode(bytes, p, end);
            }

            if (!Encoding.isAscii(c)) {
                int n = StringSupport.length(enc, bytes, p, end);
                while (n-- > 0) obytes[op++] = bytes[p++];
                continue;
            }
            p += cl;
            switch (c) {
            case '[': case ']': case '{': case '}':
            case '(': case ')': case '|': case '-':
            case '*': case '.': case '\\':
            case '?': case '+': case '^': case '$':
            case '#':
                op += enc.codeToMbc('\\', obytes, op);
                break;
            case ' ':
                op += enc.codeToMbc('\\', obytes, op);
                op += enc.codeToMbc(' ', obytes, op);
                continue;
            case '\t':
                op += enc.codeToMbc('\\', obytes, op);
                op += enc.codeToMbc('t', obytes, op);
                continue;
            case '\n':
                op += enc.codeToMbc('\\', obytes, op);
                op += enc.codeToMbc('n', obytes, op);
                continue;
            case '\r':
                op += enc.codeToMbc('\\', obytes, op);
                op += enc.codeToMbc('r', obytes, op);
                continue;
            case '\f':
                op += enc.codeToMbc('\\', obytes, op);
                op += enc.codeToMbc('f', obytes, op);
                continue;
            }
            op += enc.codeToMbc(c, obytes, op);
        }

        result.setRealSize(op);
        return result;
    }
View Full Code Here
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
            }
        }

        boolean hasAsciiOnly = false;
        RubyString source = runtime.newString();
        Encoding hasAsciiCompatFixed = null;
        Encoding hasAsciiIncompat = null;       

        for (int i = 0; i < realArgs.length; i++) {
            if (0 < i) source.cat((byte)'|');
           
            Encoding enc;
            IRubyObject v = TypeConverter.convertToTypeWithCheck(realArgs[i], runtime.getRegexp(), "to_regexp");
            if (!v.isNil()) {
                RubyRegexp regex = (RubyRegexp) v;
                enc = regex.getEncoding();

                if (!enc.isAsciiCompatible()) {
                    if (hasAsciiIncompat == null) { // First regexp of union sets kcode.
                        hasAsciiIncompat = enc;
                    } else if (hasAsciiIncompat != enc) { // n kcode doesn't match first one
                        throw runtime.newArgumentError("incompatible encodings: " + hasAsciiIncompat + " and " + enc);
                    }
                } else if (regex.getOptions().isFixed()) {
                    if (hasAsciiCompatFixed == null) { // First regexp of union sets kcode.
                        hasAsciiCompatFixed = enc;
                    } else if (hasAsciiCompatFixed != enc) { // n kcode doesn't match first one
                        throw runtime.newArgumentError("incompatible encodings: " + hasAsciiCompatFixed + " and " + enc);
                    }
                } else {
                    hasAsciiOnly = true;
                }
                v = regex.to_s();
            } else {
                RubyString str = realArgs[i].convertToString();
                enc = str.getEncoding();
               
                if (!enc.isAsciiCompatible()) {
                    if (hasAsciiIncompat == null) { // First regexp of union sets kcode.
                        hasAsciiIncompat = enc;
                    } else if (hasAsciiIncompat != enc) { // n kcode doesn't match first one
                        throw runtime.newArgumentError("incompatible encodings: " + hasAsciiIncompat + " and " + enc);
                    }
View Full Code Here
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
    }

    // rb_reg_initialize_str
    private RubyRegexp initializeCommon19(RubyString str, RegexpOptions options) {
        ByteList bytes = str.getByteList();
        Encoding enc = bytes.getEncoding();
        if (options.isEncodingNone()) {
            if (enc != ASCIIEncoding.INSTANCE) {
                if (str.scanForCodeRange() != StringSupport.CR_7BIT) {
                    raiseRegexpError19(getRuntime(), bytes, enc, options, "/.../n has a non escaped non ASCII character in non ASCII-8BIT script");
                }
View Full Code Here
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
        ByteList bs = str.getByteList();
        int p = bs.getBegin();
        int s = p;
        int end = p + bs.getRealSize();
        byte[]bytes = bs.getUnsafeBytes();
        Encoding strEnc = bs.getEncoding();

        ByteList srcbs = src.getByteList();
        Encoding srcEnc = srcbs.getEncoding();

        RubyString val = null;

        while (s < end) {
            int c, cl;
View Full Code Here
542
543
544
545
546
547
548
549
550
551
552
553
554
555
    private boolean isPrintable() {
        Ruby runtime = getRuntime();
        int p = symbolBytes.getBegin();
        int end = p + symbolBytes.getRealSize();
        byte[]bytes = symbolBytes.getUnsafeBytes();
        Encoding enc = symbolBytes.getEncoding();

        while (p < end) {
            int c = codePoint(runtime, enc, bytes, p, end);
            if (!enc.isPrint(c)) return false;
            p += codeLength(runtime, enc, c);
        }
        return true;
    }
View Full Code Here
76
77
78
79
80
81
82
83
84
85
86
        }
    }

    private static Map getAsMapOfRubyStrings(Ruby runtime, Set<Map.Entry<Object, Object>> entrySet) {
        Map envs = new HashMap();
        Encoding encoding = runtime.getEncodingService().getLocaleEncoding();

        // On Windows, entrySet doesn't have corresponding keys for these
        if (Platform.IS_WINDOWS) {
            addRubyKeyValuePair(runtime, envs, "HOME", System.getProperty("user.home"), encoding);
            addRubyKeyValuePair(runtime, envs, "USER", System.getProperty("user.name"), encoding);
View Full Code Here
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
        ByteList value = str.getByteList();

        if (pos >= value.getRealSize()) return runtime.getNil();
        int len;
        if (is1_9) {
            Encoding enc = str.getEncoding();
            len = enc.isSingleByte() ? 1 : StringSupport.length(enc, value.getUnsafeBytes(), value.getBegin() + pos, value.getBegin() + value.getRealSize());
        } else {
            Encoding enc = runtime.getKCode().getEncoding();
            len = enc.isSingleByte() ? 1 : enc.length(value.getUnsafeBytes(), value.getBegin() + pos, value.getBegin() + value.getRealSize());
        }

        if (pos + len > value.getRealSize()) len = value.getRealSize() - pos;
        lastPos = pos;
        pos += len;
View Full Code Here
780
781
782
783
784
785
786
787
788
789
790
     *           </tr>
     *         </table>
     *
     **/
    public static RubyArray unpack(Ruby runtime, ByteList encodedString, ByteList formatString) {
        Encoding encoding = encodedString.getEncoding();
        RubyArray result = runtime.newArray();
        // FIXME: potentially could just use ByteList here?
        ByteBuffer format = ByteBuffer.wrap(formatString.getUnsafeBytes(), formatString.begin(), formatString.length());
        ByteBuffer encode = ByteBuffer.wrap(encodedString.getUnsafeBytes(), encodedString.begin(), encodedString.length());
        int type = 0;
View Full Code Here
372
373
374
375
376
377
378
379
380
381
382
        ByteList r = new ByteList(arr, start, buf.limit() - start);

        EncodingDB.Entry entry = EncodingDB.getEncodings().get(displayName.getBytes());
        if (entry != null) {
            Encoding charset = entry.getEncoding();
            r.setEncoding(charset);
        }

        return getRuntime().newString(r);
    }
View Full Code Here
TOP

Related Classes of org.jcodings.Encoding

  • com.redcareditor.onig.Match
  • data_objects.drivers.AbstractDriverDefinition
  • org.joni.ApplyCaseFold
  • org.joni.ast.CClassNode
  • org.jruby.compiler.ASTCompiler
  • org.jruby.ext.iconv.RubyIconv
  • org.jruby.ext.nkf.RubyNKF$Converter
  • org.jruby.ext.psych.PsychEmitter
  • org.jruby.ext.psych.PsychParser
  • org.jruby.ext.ripper.HeredocTerm
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.