DCausse has uploaded a new change for review.
https://gerrit.wikimedia.org/r/285380
Change subject: Fix security violations
......................................................................
Fix security violations
Plugins are now loaded in their own ClassLoader. It is no more possible to use
the package trick to access package visibles members/methods owned by
elasticsearch/lucene. This patch removes all org.elasticsearch and
org.apache.lucene packages. Unfortunately this was not seen by integration
tests because we use the same classloader.
Bug: T133618
Change-Id: I16c09ec2ca03555e3c13814fd86fb1506ca95657
---
R
src/main/java/org/wikimedia/search/extra/regex/ContainsCharacterRunAutomaton.java
D src/main/java/org/wikimedia/search/extra/regex/RegexTooComplexException.java
M src/main/java/org/wikimedia/search/extra/regex/SourceRegexQuery.java
3 files changed, 10 insertions(+), 36 deletions(-)
git pull ssh://gerrit.wikimedia.org:29418/search/extra
refs/changes/80/285380/1
diff --git
a/src/main/java/org/apache/lucene/util/automaton/ContainsCharacterRunAutomaton.java
b/src/main/java/org/wikimedia/search/extra/regex/ContainsCharacterRunAutomaton.java
similarity index 92%
rename from
src/main/java/org/apache/lucene/util/automaton/ContainsCharacterRunAutomaton.java
rename to
src/main/java/org/wikimedia/search/extra/regex/ContainsCharacterRunAutomaton.java
index ff4182e..9a8f9a1 100644
---
a/src/main/java/org/apache/lucene/util/automaton/ContainsCharacterRunAutomaton.java
+++
b/src/main/java/org/wikimedia/search/extra/regex/ContainsCharacterRunAutomaton.java
@@ -1,6 +1,9 @@
-package org.apache.lucene.util.automaton;
+package org.wikimedia.search.extra.regex;
-public class ContainsCharacterRunAutomaton extends RunAutomaton {
+import org.apache.lucene.util.automaton.Automaton;
+import org.apache.lucene.util.automaton.RunAutomaton;
+
+class ContainsCharacterRunAutomaton extends RunAutomaton {
public ContainsCharacterRunAutomaton(Automaton a) {
super(a, Character.MAX_CODE_POINT, true);
}
@@ -13,6 +16,8 @@
public boolean contains(String s) {
int end = s.length();
int offset = 0;
+ // super.initial is final
+ final int initial = getInitialState();
while (offset < end) {
int cp = s.codePointAt(offset);
offset += Character.charCount(cp);
@@ -20,7 +25,7 @@
if (p == -1) {
continue;
}
- if (accept[p]) {
+ if (isAccept(p)) {
return true;
}
/*
@@ -36,7 +41,7 @@
if (p == -1) {
break;
}
- if (accept[p]) {
+ if (isAccept(p)) {
return true;
}
}
diff --git
a/src/main/java/org/wikimedia/search/extra/regex/RegexTooComplexException.java
b/src/main/java/org/wikimedia/search/extra/regex/RegexTooComplexException.java
deleted file mode 100644
index 3161bb9..0000000
---
a/src/main/java/org/wikimedia/search/extra/regex/RegexTooComplexException.java
+++ /dev/null
@@ -1,15 +0,0 @@
-package org.wikimedia.search.extra.regex;
-
-import org.apache.lucene.util.automaton.TooComplexToDeterminizeException;
-
-/**
- * Wraps Lucene's XTooComplexToDeterminizeException to be serializable to be
- * thrown over the wire.
- */
-public class RegexTooComplexException extends RuntimeException {
- private static final long serialVersionUID = -41975279199116247L;
-
- public RegexTooComplexException(TooComplexToDeterminizeException e) {
- super(e.getMessage());
- }
-}
diff --git
a/src/main/java/org/wikimedia/search/extra/regex/SourceRegexQuery.java
b/src/main/java/org/wikimedia/search/extra/regex/SourceRegexQuery.java
index bcb9e9d..1eac220 100644
--- a/src/main/java/org/wikimedia/search/extra/regex/SourceRegexQuery.java
+++ b/src/main/java/org/wikimedia/search/extra/regex/SourceRegexQuery.java
@@ -10,11 +10,7 @@
import org.apache.lucene.search.TwoPhaseIterator;
import org.apache.lucene.util.automaton.Automaton;
import org.apache.lucene.util.automaton.CharacterRunAutomaton;
-import org.apache.lucene.util.automaton.ContainsCharacterRunAutomaton;
import org.apache.lucene.util.automaton.RegExp;
-import org.apache.lucene.util.automaton.TooComplexToDeterminizeException;
-import org.elasticsearch.common.logging.ESLogger;
-import org.elasticsearch.common.logging.ESLoggerFactory;
import org.elasticsearch.common.lucene.search.Queries;
import org.wikimedia.search.extra.regex.expression.Expression;
import org.wikimedia.search.extra.regex.ngram.AutomatonTooComplexException;
@@ -23,7 +19,6 @@
@EqualsAndHashCode
public class SourceRegexQuery extends Query {
- private static final ESLogger log =
ESLoggerFactory.getLogger(SourceRegexQuery.class.getPackage().getName());
private final String fieldPath;
private final String ngramFieldPath;
private final String regex;
@@ -87,18 +82,7 @@
}
private static Automaton regexToAutomaton(RegExp regex, int
maxDeterminizedStates) {
- try {
- return regex.toAutomaton(maxDeterminizedStates);
- } catch (TooComplexToDeterminizeException e) {
- /*
- * Since we're going to lose the stack trace we give our future
- * selves an opportunity to log it in case we need it.
- */
- if (log.isDebugEnabled()) {
- log.debug("Regex too complex to determinize", e);
- }
- throw new RegexTooComplexException(e);
- }
+ return regex.toAutomaton(maxDeterminizedStates);
}
/**
--
To view, visit https://gerrit.wikimedia.org/r/285380
To unsubscribe, visit https://gerrit.wikimedia.org/r/settings
Gerrit-MessageType: newchange
Gerrit-Change-Id: I16c09ec2ca03555e3c13814fd86fb1506ca95657
Gerrit-PatchSet: 1
Gerrit-Project: search/extra
Gerrit-Branch: master
Gerrit-Owner: DCausse <[email protected]>
_______________________________________________
MediaWiki-commits mailing list
[email protected]
https://lists.wikimedia.org/mailman/listinfo/mediawiki-commits
