1+ /*-
2+ * #%L
3+ * Java ECS logging
4+ * %%
5+ * Copyright (C) 2019 Elastic and contributors
6+ * %%
7+ * Licensed to Elasticsearch B.V. under one or more contributor
8+ * license agreements. See the NOTICE file distributed with
9+ * this work for additional information regarding copyright
10+ * ownership. Elasticsearch B.V. licenses this file to you under
11+ * the Apache License, Version 2.0 (the "License"); you may
12+ * not use this file except in compliance with the License.
13+ * You may obtain a copy of the License at
14+ *
15+ * http://www.apache.org/licenses/LICENSE-2.0
16+ *
17+ * Unless required by applicable law or agreed to in writing,
18+ * software distributed under the License is distributed on an
19+ * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
20+ * KIND, either express or implied. See the License for the
21+ * specific language governing permissions and limitations
22+ * under the License.
23+ * #L%
24+ */
125package co .elastic .logging ;
226
327/**
4- * This class is borrowed from <a href="https://github.com/FasterXML/jackson-core">Jackson</a>.
28+ * This class is based on com.fasterxml.jackson.core.io.CharTypes,
29+ * under Apache License 2.0
530 */
631public final class JsonUtils {
732
8- private static final char [] HC = "0123456789ABCDEF" .toCharArray ();
33+ private final static char [] HC = "0123456789ABCDEF" .toCharArray ();
934
1035 /**
11- * Read-only encoding table for first 128 Unicode code points (single-byte UTF-8 characters).
12- * Value of 0 means "no escaping"; other positive values that value is character
13- * to use after backslash; and negative values that generic (backslash - u)
14- * escaping is to be used.
36+ * Lookup table used for determining which output characters in
37+ * 7-bit ASCII range need to be quoted.
1538 */
16- private static final int [] ESC_CODES ;
39+ private final static int [] sOutputEscapes128 ;
40+
1741 static {
18- final int [] table = new int [128 ];
42+ int [] table = new int [128 ];
1943 // Control chars need generic escape sequence
2044 for (int i = 0 ; i < 32 ; ++i ) {
2145 // 04-Mar-2011, tatu: Used to use "-(i + 1)", replaced with constant
2246 table [i ] = -1 ;
2347 }
24- /* Others (and some within that range too) have explicit shorter
25- * sequences
26- */
48+ // Others (and some within that range too) have explicit shorter sequences
2749 table ['"' ] = '"' ;
2850 table ['\\' ] = '\\' ;
2951 // Escaping of slash is optional, so let's not add it
@@ -32,72 +54,40 @@ public final class JsonUtils {
3254 table [0x0C ] = 'f' ;
3355 table [0x0A ] = 'n' ;
3456 table [0x0D ] = 'r' ;
35- ESC_CODES = table ;
36- }
37-
38- /**
39- * Temporary buffer used for composing quote/escape sequences
40- */
41- private final static ThreadLocal <char []> _qbufLocal = new ThreadLocal <>();
42-
43- private static char [] getQBuf () {
44- char [] _qbuf = _qbufLocal .get ();
45- if (_qbuf == null ) {
46- _qbuf = new char [6 ];
47- _qbuf [0 ] = '\\' ;
48- _qbuf [2 ] = '0' ;
49- _qbuf [3 ] = '0' ;
50-
51- _qbufLocal .set (_qbuf );
52- }
53- return _qbuf ;
57+ sOutputEscapes128 = table ;
5458 }
5559
56- /**
57- * Quote text contents using JSON standard quoting, and append results to a supplied {@link StringBuilder}.
58- */
59- public static void quoteAsString (final CharSequence input , final StringBuilder output ) {
60- final char [] qbuf = getQBuf ();
61- final int escCodeCount = ESC_CODES .length ;
62- int inPtr = 0 ;
63- final int inputLen = input .length ();
64-
65- outer :
66- while (inPtr < inputLen ) {
67- tight_loop :
68- while (true ) {
69- final char c = input .charAt (inPtr );
70- if (c < escCodeCount && ESC_CODES [c ] != 0 ) {
71- break tight_loop ;
72- }
73- output .append (c );
74- if (++inPtr >= inputLen ) {
75- break outer ;
76- }
60+ public static void quoteAsString (CharSequence content , StringBuilder sb ) {
61+ final int [] escCodes = sOutputEscapes128 ;
62+ final int escLen = escCodes .length ;
63+ for (int i = 0 , len = content .length (); i < len ; ++i ) {
64+ char c = content .charAt (i );
65+ if (c >= escLen || escCodes [c ] == 0 ) {
66+ sb .append (c );
67+ continue ;
7768 }
78- // something to escape; 2 or 6-char variant?
79- final char d = input .charAt (inPtr ++);
80- final int escCode = ESC_CODES [d ];
81- final int length = (escCode < 0 )
82- ? _appendNumeric (d , qbuf )
83- : _appendNamed (escCode , qbuf );
69+ sb .append ('\\' );
70+ int escCode = escCodes [c ];
71+ if (escCode < 0 ) { // generic quoting (hex value)
72+ // The only negative value sOutputEscapes128 returns
73+ // is CharacterEscapes.ESCAPE_STANDARD, which mean
74+ // appendQuotes should encode using the Unicode encoding;
75+ // not sure if this is the right way to encode for
76+ // CharacterEscapes.ESCAPE_CUSTOM or other (future)
77+ // CharacterEscapes.ESCAPE_XXX values.
8478
85- output .append (qbuf , 0 , length );
79+ // We know that it has to fit in just 2 hex chars
80+ sb .append ('u' );
81+ sb .append ('0' );
82+ sb .append ('0' );
83+ int value = c ; // widening
84+ sb .append (HC [value >> 4 ]);
85+ sb .append (HC [value & 0xF ]);
86+ } else { // "named", i.e. prepend with slash
87+ sb .append ((char ) escCode );
88+ }
8689 }
8790 }
8891
89- private static int _appendNumeric (final int value , final char [] qbuf ) {
90- qbuf [1 ] = 'u' ;
91- // We know it's a control char, so only the last 2 chars are non-0
92- qbuf [4 ] = HC [value >> 4 ];
93- qbuf [5 ] = HC [value & 0xF ];
94- return 6 ;
95- }
96-
97- private static int _appendNamed (final int esc , final char [] qbuf ) {
98- qbuf [1 ] = (char ) esc ;
99- return 2 ;
100- }
101-
10292}
10393
0 commit comments