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 "Pattern"
18
19#include <stdlib.h>
20
21#include "JNIHelp.h"
22#include "JniConstants.h"
23#include "ScopedJavaUnicodeString.h"
24#include "jni.h"
25#include "unicode/parseerr.h"
26#include "unicode/regex.h"
27
28// ICU documentation: http://icu-project.org/apiref/icu4c/classRegexPattern.html
29
30static RegexPattern* toRegexPattern(jint addr) {
31    return reinterpret_cast<RegexPattern*>(static_cast<uintptr_t>(addr));
32}
33
34static 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
65static void throwPatternSyntaxException(JNIEnv* env, UErrorCode status, jstring pattern, UParseError error) {
66    static jmethodID method = env->GetMethodID(JniConstants::patternSyntaxExceptionClass,
67            "<init>", "(Ljava/lang/String;Ljava/lang/String;I)V");
68    jstring message = env->NewStringUTF(regexDetailMessage(status));
69    jclass exceptionClass = JniConstants::patternSyntaxExceptionClass;
70    jobject exception = env->NewObject(exceptionClass, method, message, pattern, error.offset);
71    env->Throw(reinterpret_cast<jthrowable>(exception));
72}
73
74static void Pattern_closeImpl(JNIEnv*, jclass, jint addr) {
75    delete toRegexPattern(addr);
76}
77
78static jint Pattern_compileImpl(JNIEnv* env, jclass, jstring javaRegex, jint flags) {
79    flags |= UREGEX_ERROR_ON_UNKNOWN_ESCAPES;
80
81    UErrorCode status = U_ZERO_ERROR;
82    UParseError error;
83    error.offset = -1;
84
85    ScopedJavaUnicodeString regex(env, javaRegex);
86    UnicodeString& regexString(regex.unicodeString());
87    RegexPattern* result = RegexPattern::compile(regexString, flags, error, status);
88    if (!U_SUCCESS(status)) {
89        throwPatternSyntaxException(env, status, javaRegex, error);
90    }
91    return static_cast<jint>(reinterpret_cast<uintptr_t>(result));
92}
93
94static JNINativeMethod gMethods[] = {
95    NATIVE_METHOD(Pattern, closeImpl, "(I)V"),
96    NATIVE_METHOD(Pattern, compileImpl, "(Ljava/lang/String;I)I"),
97};
98void register_java_util_regex_Pattern(JNIEnv* env) {
99    jniRegisterNativeMethods(env, "java/util/regex/Pattern", gMethods, NELEM(gMethods));
100}
101