1 /*
2 * Copyright (C) 2010 The Android Open Source Project
3 *
4 * Licensed under the Apache License, Version 2.0 (the "License");
5 * you may not use this file except in compliance with the License.
6 * You may obtain a copy of the License at
7 *
8 * http://www.apache.org/licenses/LICENSE-2.0
9 *
10 * Unless required by applicable law or agreed to in writing, software
11 * distributed under the License is distributed on an "AS IS" BASIS,
12 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13 * See the License for the specific language governing permissions and
14 * limitations under the License.
15 */
16
17 #define LOG_TAG "PatternNative"
18
19 #include <stdlib.h>
20
21 #include <nativehelper/JNIHelp.h>
22 #include <nativehelper/jni_macros.h>
23
24 #include "IcuUtilities.h"
25 #include "MatcherState.h"
26 #include "unicode/parseerr.h"
27 #include "unicode/regex.h"
28
29 #include "JniConstants.h"
30 #include "ScopedJavaUnicodeString.h"
31
32 // ICU documentation: http://icu-project.org/apiref/icu4c/classRegexPattern.html
33
regexDetailMessage(UErrorCode status)34 static const char* regexDetailMessage(UErrorCode status) {
35 // These human-readable error messages were culled from "utypes.h", and then slightly tuned
36 // to make more sense in context.
37 // If we don't have a special-case, we'll just return the textual name of
38 // the enum value (such as U_REGEX_RULE_SYNTAX), which is better than nothing.
39 switch (status) {
40 case U_REGEX_INTERNAL_ERROR: return "An internal error was detected";
41 case U_REGEX_RULE_SYNTAX: return "Syntax error in regexp pattern";
42 case U_REGEX_INVALID_STATE: return "Matcher in invalid state for requested operation";
43 case U_REGEX_BAD_ESCAPE_SEQUENCE: return "Unrecognized backslash escape sequence in pattern";
44 case U_REGEX_PROPERTY_SYNTAX: return "Incorrect Unicode property";
45 case U_REGEX_UNIMPLEMENTED: return "Use of unimplemented feature";
46 case U_REGEX_MISMATCHED_PAREN: return "Incorrectly nested parentheses in regexp pattern";
47 case U_REGEX_NUMBER_TOO_BIG: return "Decimal number is too large";
48 case U_REGEX_BAD_INTERVAL: return "Error in {min,max} interval";
49 case U_REGEX_MAX_LT_MIN: return "In {min,max}, max is less than min";
50 case U_REGEX_INVALID_BACK_REF: return "Back-reference to a non-existent capture group";
51 case U_REGEX_INVALID_FLAG: return "Invalid value for match mode flags";
52 case U_REGEX_LOOK_BEHIND_LIMIT: return "Look-behind pattern matches must have a bounded maximum length";
53 case U_REGEX_SET_CONTAINS_STRING: return "Regular expressions cannot have UnicodeSets containing strings";
54 case U_REGEX_OCTAL_TOO_BIG: return "Octal character constants must be <= 0377.";
55 case U_REGEX_MISSING_CLOSE_BRACKET: return "Missing closing bracket in character class";
56 case U_REGEX_INVALID_RANGE: return "In a character range [x-y], x is greater than y";
57 case U_REGEX_STACK_OVERFLOW: return "Regular expression backtrack stack overflow";
58 case U_REGEX_TIME_OUT: return "Maximum allowed match time exceeded";
59 case U_REGEX_STOPPED_BY_CALLER: return "Matching operation aborted by user callback function";
60 default:
61 return u_errorName(status);
62 }
63 }
64
throwPatternSyntaxException(JNIEnv * env,UErrorCode status,jstring pattern,UParseError error)65 static void throwPatternSyntaxException(JNIEnv* env, UErrorCode status, jstring pattern, UParseError error) {
66 static jmethodID method = env->GetMethodID(JniConstants::GetPatternSyntaxExceptionClass(env),
67 "<init>", "(Ljava/lang/String;Ljava/lang/String;I)V");
68 jstring message = env->NewStringUTF(regexDetailMessage(status));
69 jclass exceptionClass = JniConstants::GetPatternSyntaxExceptionClass(env);
70 jobject exception = env->NewObject(exceptionClass, method, message, pattern, error.offset);
71 env->Throw(reinterpret_cast<jthrowable>(exception));
72 }
73
PatternNative_free(void * addr)74 static void PatternNative_free(void* addr) {
75 delete reinterpret_cast<icu::RegexPattern*>(addr);
76 }
77
PatternNative_getNativeFinalizer(JNIEnv *,jclass)78 static jlong PatternNative_getNativeFinalizer(JNIEnv*, jclass) {
79 return reinterpret_cast<jlong>(&PatternNative_free);
80 }
81
PatternNative_compileImpl(JNIEnv * env,jclass,jstring javaRegex,jint flags)82 static jlong PatternNative_compileImpl(JNIEnv* env, jclass, jstring javaRegex, jint flags) {
83 flags |= UREGEX_ERROR_ON_UNKNOWN_ESCAPES;
84
85 UErrorCode status = U_ZERO_ERROR;
86 UParseError error;
87 error.offset = -1;
88
89 ScopedJavaUnicodeString regex(env, javaRegex);
90 if (!regex.valid()) {
91 return 0;
92 }
93 icu::UnicodeString& regexString(regex.unicodeString());
94 icu::RegexPattern* result = icu::RegexPattern::compile(regexString, flags, error, status);
95 if (!U_SUCCESS(status)) {
96 throwPatternSyntaxException(env, status, javaRegex, error);
97 }
98 return static_cast<jlong>(reinterpret_cast<uintptr_t>(result));
99 }
100
PatternNative_openMatcherImpl(JNIEnv * env,jclass,jlong addr)101 static jlong PatternNative_openMatcherImpl(JNIEnv* env, jclass, jlong addr) {
102 icu::RegexPattern* pattern = reinterpret_cast<icu::RegexPattern*>(static_cast<uintptr_t>(addr));
103 UErrorCode status = U_ZERO_ERROR;
104 icu::RegexMatcher* result = pattern->matcher(status);
105 if (maybeThrowIcuException(env, "RegexPattern::matcher", status)) {
106 return 0;
107 }
108
109 return reinterpret_cast<uintptr_t>(new MatcherState(result));
110 }
111
PatternNative_getMatchedGroupIndexImpl(JNIEnv * env,jclass,jlong addr,jstring javaGroupName)112 static jint PatternNative_getMatchedGroupIndexImpl(JNIEnv* env, jclass, jlong addr, jstring javaGroupName) {
113 icu::RegexPattern* pattern = reinterpret_cast<icu::RegexPattern*>(static_cast<uintptr_t>(addr));
114 ScopedJavaUnicodeString groupName(env, javaGroupName);
115 UErrorCode status = U_ZERO_ERROR;
116
117 jint result = pattern->groupNumberFromName(groupName.unicodeString(), status);
118 if (U_SUCCESS(status)) {
119 return result;
120 }
121 if (status == U_REGEX_INVALID_CAPTURE_GROUP_NAME) {
122 return -1;
123 }
124 maybeThrowIcuException(env, "RegexPattern::groupNumberFromName", status);
125 return -1;
126 }
127
128
129 static JNINativeMethod gMethods[] = {
130 NATIVE_METHOD(PatternNative, compileImpl, "(Ljava/lang/String;I)J"),
131 NATIVE_METHOD(PatternNative, getNativeFinalizer, "()J"),
132 NATIVE_METHOD(PatternNative, openMatcherImpl, "(J)J"),
133 NATIVE_METHOD(PatternNative, getMatchedGroupIndexImpl, "(JLjava/lang/String;)I"),
134 };
135
register_com_android_icu_util_regex_PatternNative(JNIEnv * env)136 void register_com_android_icu_util_regex_PatternNative(JNIEnv* env) {
137 jniRegisterNativeMethods(env, "com/android/icu/util/regex/PatternNative", gMethods, NELEM(gMethods));
138 }
139