Diff
Modified: trunk/Source/WebCore/ChangeLog (181725 => 181726)
--- trunk/Source/WebCore/ChangeLog 2015-03-19 04:05:51 UTC (rev 181725)
+++ trunk/Source/WebCore/ChangeLog 2015-03-19 04:57:20 UTC (rev 181726)
@@ -1,3 +1,53 @@
+2015-03-18 Alex Christensen <[email protected]>
+
+ [ContentExtensions] Prepare for compiling stylesheets of selectors to be used on every page.
+ https://bugs.webkit.org/show_bug.cgi?id=142799
+
+ Reviewed by Brady Eidson.
+
+ * WebCore.xcodeproj/project.pbxproj:
+ Make private headers to use with API tests.
+ * contentextensions/CompiledContentExtension.cpp:
+ (WebCore::ContentExtensions::CompiledContentExtension::globalDisplayNoneSelectors):
+ * contentextensions/CompiledContentExtension.h:
+ Added method to get only the selectors from the root of the DFA, which apply to all URLs.
+ * contentextensions/ContentExtensionCompiler.cpp:
+ (WebCore::ContentExtensions::compileRuleList):
+ Added checking if the trigger will match everything.
+ These actions can be put directly on the root of the DFA without adding extra epsilon transitions to the NFA.
+ * contentextensions/DFA.h:
+ (WebCore::ContentExtensions::DFA::nodeAt):
+ * contentextensions/DFABytecodeInterpreter.cpp:
+ (WebCore::ContentExtensions::DFABytecodeInterpreter::actionsFromDFARoot):
+ (WebCore::ContentExtensions::DFABytecodeInterpreter::interpret):
+ * contentextensions/DFABytecodeInterpreter.h:
+ * contentextensions/NFA.h:
+ * contentextensions/URLFilterParser.cpp:
+ (WebCore::ContentExtensions::Term::quantifier):
+ Sink terms to a vector then add nodes to NFA when finalizing after checking for regexes that match everything.
+ (WebCore::ContentExtensions::GraphBuilder::GraphBuilder):
+ (WebCore::ContentExtensions::GraphBuilder::finalize):
+ (WebCore::ContentExtensions::GraphBuilder::parseStatus):
+ (WebCore::ContentExtensions::GraphBuilder::atomPatternCharacter):
+ (WebCore::ContentExtensions::GraphBuilder::atomBuiltInCharacterClass):
+ (WebCore::ContentExtensions::GraphBuilder::quantifyAtom):
+ (WebCore::ContentExtensions::GraphBuilder::atomBackReference):
+ (WebCore::ContentExtensions::GraphBuilder::assertionBOL):
+ (WebCore::ContentExtensions::GraphBuilder::assertionWordBoundary):
+ (WebCore::ContentExtensions::GraphBuilder::atomCharacterClassAtom):
+ (WebCore::ContentExtensions::GraphBuilder::atomCharacterClassRange):
+ (WebCore::ContentExtensions::GraphBuilder::atomCharacterClassBuiltIn):
+ (WebCore::ContentExtensions::GraphBuilder::atomParentheticalAssertionBegin):
+ (WebCore::ContentExtensions::GraphBuilder::disjunction):
+ (WebCore::ContentExtensions::GraphBuilder::hasError):
+ (WebCore::ContentExtensions::GraphBuilder::fail):
+ (WebCore::ContentExtensions::GraphBuilder::sinkFloatingTermIfNecessary):
+ (WebCore::ContentExtensions::URLFilterParser::addPattern):
+ (WebCore::ContentExtensions::URLFilterParser::statusString):
+ (WebCore::ContentExtensions::GraphBuilder::errorMessage): Deleted.
+ * contentextensions/URLFilterParser.h:
+ Use an enum instead of strings for the status to avoid checking strings when we have a regex that matches everything.
+
2015-03-18 Yusuke Suzuki <[email protected]>
Fix build failure due to FALLTHROUGH in unreachable code
Modified: trunk/Source/WebCore/WebCore.xcodeproj/project.pbxproj (181725 => 181726)
--- trunk/Source/WebCore/WebCore.xcodeproj/project.pbxproj 2015-03-19 04:05:51 UTC (rev 181725)
+++ trunk/Source/WebCore/WebCore.xcodeproj/project.pbxproj 2015-03-19 04:57:20 UTC (rev 181726)
@@ -993,7 +993,7 @@
24F54EAD101FE914000AE741 /* ApplicationCacheHost.h in Headers */ = {isa = PBXBuildFile; fileRef = 24F54EAB101FE914000AE741 /* ApplicationCacheHost.h */; settings = {ATTRIBUTES = (Private, ); }; };
2542F4DA1166C25A00E89A86 /* UserGestureIndicator.cpp in Sources */ = {isa = PBXBuildFile; fileRef = 2542F4D81166C25A00E89A86 /* UserGestureIndicator.cpp */; };
2542F4DB1166C25A00E89A86 /* UserGestureIndicator.h in Headers */ = {isa = PBXBuildFile; fileRef = 2542F4D91166C25A00E89A86 /* UserGestureIndicator.h */; settings = {ATTRIBUTES = (Private, ); }; };
- 262391361A648CEE007251A3 /* ContentExtensionsDebugging.h in Headers */ = {isa = PBXBuildFile; fileRef = 262391351A648CEE007251A3 /* ContentExtensionsDebugging.h */; };
+ 262391361A648CEE007251A3 /* ContentExtensionsDebugging.h in Headers */ = {isa = PBXBuildFile; fileRef = 262391351A648CEE007251A3 /* ContentExtensionsDebugging.h */; settings = {ATTRIBUTES = (Private, ); }; };
26255F0018878DFF0006E1FD /* UserAgentIOS.mm in Sources */ = {isa = PBXBuildFile; fileRef = 26255EFF18878DFF0006E1FD /* UserAgentIOS.mm */; };
26255F0318878E110006E1FD /* UserAgent.h in Headers */ = {isa = PBXBuildFile; fileRef = 26255F0118878E110006E1FD /* UserAgent.h */; settings = {ATTRIBUTES = (Private, ); }; };
26255F0418878E110006E1FD /* UserAgentMac.mm in Sources */ = {isa = PBXBuildFile; fileRef = 26255F0218878E110006E1FD /* UserAgentMac.mm */; };
@@ -1008,10 +1008,10 @@
267726001A5B3AD9003C24DD /* NFAToDFA.cpp in Sources */ = {isa = PBXBuildFile; fileRef = 267725FA1A5B3AD9003C24DD /* NFAToDFA.cpp */; };
267726011A5B3AD9003C24DD /* NFAToDFA.h in Headers */ = {isa = PBXBuildFile; fileRef = 267725FB1A5B3AD9003C24DD /* NFAToDFA.h */; };
267726041A5DF6F2003C24DD /* URLFilterParser.cpp in Sources */ = {isa = PBXBuildFile; fileRef = 267726021A5DF6F2003C24DD /* URLFilterParser.cpp */; };
- 267726051A5DF6F2003C24DD /* URLFilterParser.h in Headers */ = {isa = PBXBuildFile; fileRef = 267726031A5DF6F2003C24DD /* URLFilterParser.h */; };
+ 267726051A5DF6F2003C24DD /* URLFilterParser.h in Headers */ = {isa = PBXBuildFile; fileRef = 267726031A5DF6F2003C24DD /* URLFilterParser.h */; settings = {ATTRIBUTES = (Private, ); }; };
269239961505E1AA009E57FC /* JSIDBVersionChangeEvent.h in Headers */ = {isa = PBXBuildFile; fileRef = 269239921505E1AA009E57FC /* JSIDBVersionChangeEvent.h */; };
- 269397221A4A412F00E8349D /* NFANode.h in Headers */ = {isa = PBXBuildFile; fileRef = 269397201A4A412F00E8349D /* NFANode.h */; };
- 269397241A4A5B6400E8349D /* NFA.h in Headers */ = {isa = PBXBuildFile; fileRef = 269397231A4A5B6400E8349D /* NFA.h */; };
+ 269397221A4A412F00E8349D /* NFANode.h in Headers */ = {isa = PBXBuildFile; fileRef = 269397201A4A412F00E8349D /* NFANode.h */; settings = {ATTRIBUTES = (Private, ); }; };
+ 269397241A4A5B6400E8349D /* NFA.h in Headers */ = {isa = PBXBuildFile; fileRef = 269397231A4A5B6400E8349D /* NFA.h */; settings = {ATTRIBUTES = (Private, ); }; };
269397261A4A5FBD00E8349D /* NFA.cpp in Sources */ = {isa = PBXBuildFile; fileRef = 269397251A4A5FBD00E8349D /* NFA.cpp */; };
26AA0F9E18D2A18B00419381 /* SelectorPseudoElementTypeMap.cpp in Sources */ = {isa = PBXBuildFile; fileRef = 26AA0F9D18D2A18B00419381 /* SelectorPseudoElementTypeMap.cpp */; };
26B9998F1803AE7200D01121 /* RegisterAllocator.h in Headers */ = {isa = PBXBuildFile; fileRef = 26B9998E1803AE7200D01121 /* RegisterAllocator.h */; };
Modified: trunk/Source/WebCore/contentextensions/CompiledContentExtension.cpp (181725 => 181726)
--- trunk/Source/WebCore/contentextensions/CompiledContentExtension.cpp 2015-03-19 04:05:51 UTC (rev 181725)
+++ trunk/Source/WebCore/contentextensions/CompiledContentExtension.cpp 2015-03-19 04:57:20 UTC (rev 181726)
@@ -25,6 +25,7 @@
#include "config.h"
#include "CompiledContentExtension.h"
+#include "DFABytecodeInterpreter.h"
#if ENABLE(CONTENT_EXTENSIONS)
@@ -35,6 +36,24 @@
{
}
+Vector<String> CompiledContentExtension::globalDisplayNoneSelectors()
+{
+ DFABytecodeInterpreter interpreter(bytecode(), bytecodeLength());
+ DFABytecodeInterpreter::Actions actionLocations = interpreter.actionsFromDFARoot();
+
+ Vector<Action> globalActions;
+ for (uint64_t actionLocation : actionLocations)
+ globalActions.append(Action::deserialize(actions(), actionsLength(), static_cast<unsigned>(actionLocation)));
+
+ Vector<String> selectors;
+ for (Action& action : globalActions) {
+ if (action.cssSelector().length())
+ selectors.append(action.cssSelector());
+ }
+
+ return selectors;
+}
+
} // namespace ContentExtensions
} // namespace WebCore
Modified: trunk/Source/WebCore/contentextensions/CompiledContentExtension.h (181725 => 181726)
--- trunk/Source/WebCore/contentextensions/CompiledContentExtension.h 2015-03-19 04:05:51 UTC (rev 181725)
+++ trunk/Source/WebCore/contentextensions/CompiledContentExtension.h 2015-03-19 04:57:20 UTC (rev 181726)
@@ -44,6 +44,7 @@
virtual unsigned bytecodeLength() const = 0;
virtual const SerializedActionByte* actions() const = 0;
virtual unsigned actionsLength() const = 0;
+ Vector<String> globalDisplayNoneSelectors();
};
} // namespace ContentExtensions
Modified: trunk/Source/WebCore/contentextensions/ContentExtensionCompiler.cpp (181725 => 181726)
--- trunk/Source/WebCore/contentextensions/ContentExtensionCompiler.cpp 2015-03-19 04:05:51 UTC (rev 181725)
+++ trunk/Source/WebCore/contentextensions/ContentExtensionCompiler.cpp 2015-03-19 04:57:20 UTC (rev 181726)
@@ -108,19 +108,29 @@
Vector<SerializedActionByte> actions;
Vector<unsigned> actionLocations = serializeActions(parsedRuleList, actions);
+ Vector<uint64_t> universalActionLocations;
NFA nfa;
URLFilterParser urlFilterParser(nfa);
+ bool nonUniversalActionSeen = false;
for (unsigned ruleIndex = 0; ruleIndex < parsedRuleList.size(); ++ruleIndex) {
const ContentExtensionRule& contentExtensionRule = parsedRuleList[ruleIndex];
const Trigger& trigger = contentExtensionRule.trigger();
ASSERT(trigger.urlFilter.length());
// High bits are used for flags. This should match how they are used in DFABytecodeCompiler::compileNode.
- String error = urlFilterParser.addPattern(trigger.urlFilter, trigger.urlFilterIsCaseSensitive, (static_cast<uint64_t>(trigger.flags) << 32) | static_cast<uint64_t>(actionLocations[ruleIndex]));
+ uint64_t actionLocationAndFlags =(static_cast<uint64_t>(trigger.flags) << 32) | static_cast<uint64_t>(actionLocations[ruleIndex]);
+ URLFilterParser::ParseStatus status = urlFilterParser.addPattern(trigger.urlFilter, trigger.urlFilterIsCaseSensitive, actionLocationAndFlags);
- if (!error.isNull()) {
- dataLogF("Error while parsing %s: %s\n", trigger.urlFilter.utf8().data(), error.utf8().data());
+ if (status == URLFilterParser::MatchesEverything) {
+ if (nonUniversalActionSeen)
+ dataLogF("Trigger matching everything found not at beginning. This may cause incorrect behavior with ignore-previous-rules");
+ universalActionLocations.append(actionLocationAndFlags);
+ } else
+ nonUniversalActionSeen = true;
+
+ if (status != URLFilterParser::Ok && status != URLFilterParser::MatchesEverything) {
+ dataLogF("Error while parsing %s: %s\n", trigger.urlFilter.utf8().data(), URLFilterParser::statusString(status).utf8().data());
continue;
}
}
@@ -138,7 +148,9 @@
double dfaBuildTimeStart = monotonicallyIncreasingTime();
#endif
- const DFA dfa = NFAToDFA::convert(nfa);
+ DFA dfa = NFAToDFA::convert(nfa);
+ for (uint64_t actionLocation : universalActionLocations)
+ dfa.nodeAt(dfa.root()).actions.append(actionLocation);
#if CONTENT_EXTENSIONS_PERFORMANCE_REPORTING
double dfaBuildTimeEnd = monotonicallyIncreasingTime();
Modified: trunk/Source/WebCore/contentextensions/DFA.h (181725 => 181726)
--- trunk/Source/WebCore/contentextensions/DFA.h 2015-03-19 04:05:51 UTC (rev 181725)
+++ trunk/Source/WebCore/contentextensions/DFA.h 2015-03-19 04:57:20 UTC (rev 181726)
@@ -48,6 +48,7 @@
unsigned root() const { return m_root; }
unsigned size() const { return m_nodes.size(); }
const DFANode& nodeAt(unsigned i) const { return m_nodes[i]; }
+ DFANode& nodeAt(unsigned i) { return m_nodes[i]; }
#if CONTENT_EXTENSIONS_STATE_MACHINE_DEBUGGING
void debugPrintDot() const;
Modified: trunk/Source/WebCore/contentextensions/DFABytecodeInterpreter.cpp (181725 => 181726)
--- trunk/Source/WebCore/contentextensions/DFABytecodeInterpreter.cpp 2015-03-19 04:05:51 UTC (rev 181725)
+++ trunk/Source/WebCore/contentextensions/DFABytecodeInterpreter.cpp 2015-03-19 04:57:20 UTC (rev 181726)
@@ -40,7 +40,18 @@
ASSERT_UNUSED(bytecodeLength, index + sizeof(IntType) <= bytecodeLength);
return *reinterpret_cast<const IntType*>(&bytecode[index]);
}
-
+
+DFABytecodeInterpreter::Actions DFABytecodeInterpreter::actionsFromDFARoot()
+{
+ unsigned programCounter = 0;
+ DFABytecodeInterpreter::Actions globalActionLocations;
+ while (static_cast<DFABytecodeInstruction>(m_bytecode[programCounter]) == DFABytecodeInstruction::AppendAction) {
+ globalActionLocations.add(static_cast<uint64_t>(getBits<unsigned>(m_bytecode, m_bytecodeLength, programCounter + sizeof(DFABytecode))));
+ programCounter += instructionSizeWithArguments(DFABytecodeInstruction::AppendAction);
+ }
+ return globalActionLocations;
+}
+
DFABytecodeInterpreter::Actions DFABytecodeInterpreter::interpret(const CString& urlCString, uint16_t flags)
{
const char* url = ""
@@ -51,6 +62,8 @@
bool urlIndexIsAfterEndOfString = false;
Actions actions;
+ // FIXME: Skip the actions from the root once they are used through actionsFromDFARoot. Change AppendAction to AppendActions to make this faster.
+
// This should always terminate if interpreting correctly compiled bytecode.
while (true) {
ASSERT(programCounter <= m_bytecodeLength);
Modified: trunk/Source/WebCore/contentextensions/DFABytecodeInterpreter.h (181725 => 181726)
--- trunk/Source/WebCore/contentextensions/DFABytecodeInterpreter.h 2015-03-19 04:05:51 UTC (rev 181725)
+++ trunk/Source/WebCore/contentextensions/DFABytecodeInterpreter.h 2015-03-19 04:57:20 UTC (rev 181726)
@@ -48,6 +48,7 @@
typedef HashSet<uint64_t, DefaultHash<uint64_t>::Hash, WTF::UnsignedWithZeroKeyHashTraits<uint64_t>> Actions;
Actions interpret(const CString&, uint16_t flags);
+ Actions actionsFromDFARoot();
private:
const DFABytecode* m_bytecode;
Modified: trunk/Source/WebCore/contentextensions/NFA.h (181725 => 181726)
--- trunk/Source/WebCore/contentextensions/NFA.h 2015-03-19 04:05:51 UTC (rev 181725)
+++ trunk/Source/WebCore/contentextensions/NFA.h 2015-03-19 04:57:20 UTC (rev 181726)
@@ -42,7 +42,7 @@
// The nodes are accessed through an identifier.
class NFA {
public:
- NFA();
+ WEBCORE_EXPORT NFA();
unsigned root() const { return m_root; }
unsigned createNode();
Modified: trunk/Source/WebCore/contentextensions/URLFilterParser.cpp (181725 => 181726)
--- trunk/Source/WebCore/contentextensions/URLFilterParser.cpp 2015-03-19 04:05:51 UTC (rev 181725)
+++ trunk/Source/WebCore/contentextensions/URLFilterParser.cpp 2015-03-19 04:57:20 UTC (rev 181726)
@@ -32,6 +32,7 @@
#include <_javascript_Core/YarrParser.h>
#include <wtf/BitVector.h>
#include <wtf/Deque.h>
+#include <wtf/text/CString.h>
namespace WebCore {
@@ -174,6 +175,7 @@
ASSERT_WITH_MESSAGE(m_quantifier == AtomQuantifier::One, "Transition to quantified term should only happen once.");
m_quantifier = quantifier;
}
+ AtomQuantifier quantifier() const { return m_quantifier; }
unsigned generateGraph(NFA& nfa, uint64_t patternId, unsigned start) const
{
@@ -428,6 +430,7 @@
, m_subtreeStart(nfa.root())
, m_subtreeEnd(nfa.root())
, m_lastPrefixTreeEntry(&prefixTreeRoot)
+ , m_parseStatus(URLFilterParser::Ok)
{
}
@@ -438,20 +441,56 @@
sinkFloatingTermIfNecessary();
+ // Check to see if there are any terms without ? or *.
+ bool matchesEverything = true;
+ for (const auto& term : m_sunkTerms) {
+ if (term.quantifier() == AtomQuantifier::One || term.quantifier() == AtomQuantifier::OneOrMore) {
+ matchesEverything = false;
+ break;
+ }
+ }
+ if (matchesEverything)
+ fail(URLFilterParser::MatchesEverything);
+
+ for (const auto& term : m_sunkTerms) {
+ ASSERT(m_lastPrefixTreeEntry);
+ auto nextEntry = m_lastPrefixTreeEntry->nextPattern.find(term);
+ if (nextEntry != m_lastPrefixTreeEntry->nextPattern.end()) {
+ m_lastPrefixTreeEntry = nextEntry->value.get();
+ m_nfa.addRuleId(m_lastPrefixTreeEntry->nfaNode, m_patternId);
+ } else {
+ std::unique_ptr<PrefixTreeEntry> nextPrefixTreeEntry = std::make_unique<PrefixTreeEntry>();
+
+ unsigned newEnd = term.generateGraph(m_nfa, m_patternId, m_lastPrefixTreeEntry->nfaNode);
+ nextPrefixTreeEntry->nfaNode = newEnd;
+
+ auto addResult = m_lastPrefixTreeEntry->nextPattern.set(term, WTF::move(nextPrefixTreeEntry));
+ ASSERT(addResult.isNewEntry);
+
+ if (!m_newPrefixSubtreeRoot) {
+ m_newPrefixSubtreeRoot = m_lastPrefixTreeEntry;
+ m_newPrefixStaringPoint = term;
+ }
+
+ m_lastPrefixTreeEntry = addResult.iterator->value.get();
+ }
+ m_subtreeEnd = m_lastPrefixTreeEntry->nfaNode;
+ }
+
if (!m_openGroups.isEmpty()) {
- fail(ASCIILiteral("The _expression_ has unclosed groups."));
+ fail(URLFilterParser::UnclosedGroups);
return;
}
if (m_subtreeStart != m_subtreeEnd)
m_nfa.setFinal(m_subtreeEnd, m_patternId);
else
- fail(ASCIILiteral("The pattern cannot match anything."));
+ fail(URLFilterParser::CannotMatchAnything);
}
- const String& errorMessage() const
+ URLFilterParser::ParseStatus parseStatus() const
{
- return m_errorMessage;
+ return m_parseStatus;
}
void atomPatternCharacter(UChar character)
@@ -460,7 +499,7 @@
return;
if (!isASCII(character)) {
- fail(ASCIILiteral("Only ASCII characters are supported in pattern."));
+ fail(URLFilterParser::NonASCII);
return;
}
@@ -482,7 +521,7 @@
if (builtInCharacterClassID == JSC::Yarr::NewlineClassID && inverted)
m_floatingTerm = Term(Term::UniversalTransition);
else
- fail(ASCIILiteral("Character class is not supported."));
+ fail(URLFilterParser::UnsupportedCharacterClass);
}
void quantifyAtom(unsigned minimum, unsigned maximum, bool)
@@ -491,7 +530,7 @@
return;
if (!m_floatingTerm.isValid())
- fail(ASCIILiteral("Quantifier without corresponding term to quantify."));
+ fail(URLFilterParser::MisplacedQuantifier);
if (!minimum && maximum == 1)
m_floatingTerm.quantify(AtomQuantifier::ZeroOrOne);
@@ -500,12 +539,12 @@
else if (minimum == 1 && maximum == JSC::Yarr::quantifyInfinite)
m_floatingTerm.quantify(AtomQuantifier::OneOrMore);
else
- fail(ASCIILiteral("Arbitrary atom repetitions are not supported."));
+ fail(URLFilterParser::InvalidQuantifier);
}
void atomBackReference(unsigned)
{
- fail(ASCIILiteral("Patterns cannot contain backreferences."));
+ fail(URLFilterParser::BackReference);
}
void assertionBOL()
@@ -514,7 +553,7 @@
return;
if (m_subtreeStart != m_subtreeEnd || m_floatingTerm.isValid() || !m_openGroups.isEmpty())
- fail(ASCIILiteral("Start of line assertion can only appear as the first term in a filter."));
+ fail(URLFilterParser::MisplacedStartOfLine);
}
void assertionEOL()
@@ -530,7 +569,7 @@
void assertionWordBoundary(bool)
{
- fail(ASCIILiteral("Word boundaries assertions are not supported yet."));
+ fail(URLFilterParser::WordBoundary);
}
void atomCharacterClassBegin(bool inverted = false)
@@ -549,10 +588,7 @@
if (hasError())
return;
- if (!isASCII(character)) {
- fail(ASCIILiteral("Non ASCII Character in a character set."));
- return;
- }
+ ASSERT(isASCII(character));
m_floatingTerm.addCharacter(character, m_patternIsCaseSensitive);
}
@@ -562,10 +598,10 @@
if (hasError())
return;
- if (!a || !b || !isASCII(a) || !isASCII(b)) {
- fail(ASCIILiteral("Non ASCII Character in a character range of a character set."));
- return;
- }
+ ASSERT(a);
+ ASSERT(b);
+ ASSERT(isASCII(a));
+ ASSERT(isASCII(b));
for (unsigned i = a; i <= b; ++i)
m_floatingTerm.addCharacter(static_cast<UChar>(i), m_patternIsCaseSensitive);
@@ -578,7 +614,7 @@
void atomCharacterClassBuiltIn(JSC::Yarr::BuiltInCharacterClassID, bool)
{
- fail(ASCIILiteral("Builtins character class atoms are not supported yet."));
+ fail(URLFilterParser::AtomCharacter);
}
void atomParenthesesSubpatternBegin(bool = true)
@@ -593,7 +629,7 @@
void atomParentheticalAssertionBegin(bool = false)
{
- fail(ASCIILiteral("Groups are not supported yet."));
+ fail(URLFilterParser::Group);
}
void atomParenthesesEnd()
@@ -609,16 +645,16 @@
void disjunction()
{
- fail(ASCIILiteral("Disjunctions are not supported yet."));
+ fail(URLFilterParser::Disjunction);
}
private:
bool hasError() const
{
- return !m_errorMessage.isNull();
+ return m_parseStatus != URLFilterParser::Ok;
}
- void fail(const String& errorMessage)
+ void fail(URLFilterParser::ParseStatus reason)
{
if (hasError())
return;
@@ -626,7 +662,7 @@
if (m_newPrefixSubtreeRoot)
m_newPrefixSubtreeRoot->nextPattern.remove(m_newPrefixStaringPoint);
- m_errorMessage = errorMessage;
+ m_parseStatus = reason;
}
void sinkFloatingTermIfNecessary()
@@ -634,10 +670,8 @@
if (!m_floatingTerm.isValid())
return;
- ASSERT(m_lastPrefixTreeEntry);
-
if (m_hasProcessedEndOfLineAssertion) {
- fail(ASCIILiteral("The end of line assertion must be the last term in an _expression_."));
+ fail(URLFilterParser::MisplacedEndOfLine);
m_floatingTerm = Term();
return;
}
@@ -651,30 +685,8 @@
return;
}
- auto nextEntry = m_lastPrefixTreeEntry->nextPattern.find(m_floatingTerm);
- if (nextEntry != m_lastPrefixTreeEntry->nextPattern.end()) {
- m_lastPrefixTreeEntry = nextEntry->value.get();
- m_nfa.addRuleId(m_lastPrefixTreeEntry->nfaNode, m_patternId);
- } else {
- std::unique_ptr<PrefixTreeEntry> nextPrefixTreeEntry = std::make_unique<PrefixTreeEntry>();
-
- unsigned newEnd = m_floatingTerm.generateGraph(m_nfa, m_patternId, m_lastPrefixTreeEntry->nfaNode);
- nextPrefixTreeEntry->nfaNode = newEnd;
-
- auto addResult = m_lastPrefixTreeEntry->nextPattern.set(m_floatingTerm, WTF::move(nextPrefixTreeEntry));
- ASSERT(addResult.isNewEntry);
-
- if (!m_newPrefixSubtreeRoot) {
- m_newPrefixSubtreeRoot = m_lastPrefixTreeEntry;
- m_newPrefixStaringPoint = m_floatingTerm;
- }
-
- m_lastPrefixTreeEntry = addResult.iterator->value.get();
- }
- m_subtreeEnd = m_lastPrefixTreeEntry->nfaNode;
-
+ m_sunkTerms.append(m_floatingTerm);
m_floatingTerm = Term();
- ASSERT(m_lastPrefixTreeEntry);
}
NFA& m_nfa;
@@ -686,13 +698,14 @@
PrefixTreeEntry* m_lastPrefixTreeEntry;
Deque<Term> m_openGroups;
+ Vector<Term> m_sunkTerms;
Term m_floatingTerm;
bool m_hasProcessedEndOfLineAssertion { false };
PrefixTreeEntry* m_newPrefixSubtreeRoot = nullptr;
Term m_newPrefixStaringPoint;
- String m_errorMessage;
+ URLFilterParser::ParseStatus m_parseStatus;
};
URLFilterParser::URLFilterParser(NFA& nfa)
@@ -706,33 +719,74 @@
{
}
-String URLFilterParser::addPattern(const String& pattern, bool patternIsCaseSensitive, uint64_t patternId)
+URLFilterParser::ParseStatus URLFilterParser::addPattern(const String& pattern, bool patternIsCaseSensitive, uint64_t patternId)
{
if (!pattern.containsOnlyASCII())
- return ASCIILiteral("URLFilterParser only supports ASCII patterns.");
+ return NonASCII;
ASSERT(!pattern.isEmpty());
if (pattern.isEmpty())
- return ASCIILiteral("Empty pattern.");
+ return EmptyPattern;
unsigned oldSize = m_nfa.graphSize();
- String error;
-
+ ParseStatus status = Ok;
GraphBuilder graphBuilder(m_nfa, *m_prefixTreeRoot, patternIsCaseSensitive, patternId);
- error = String(JSC::Yarr::parse(graphBuilder, pattern, 0));
+ String error = String(JSC::Yarr::parse(graphBuilder, pattern, 0));
if (error.isNull())
graphBuilder.finalize();
+ else
+ status = YarrError;
+
+ if (status == Ok)
+ status = graphBuilder.parseStatus();
- if (error.isNull())
- error = graphBuilder.errorMessage();
-
- if (!error.isNull())
+ if (status != Ok)
m_nfa.restoreToGraphSize(oldSize);
- return error;
+ return status;
}
+String URLFilterParser::statusString(ParseStatus status)
+{
+ switch (status) {
+ case Ok:
+ return "Ok";
+ case MatchesEverything:
+ return "Matches everything.";
+ case UnclosedGroups:
+ return "The _expression_ has unclosed groups.";
+ case CannotMatchAnything:
+ return "The pattern cannot match anything.";
+ case NonASCII:
+ return "Only ASCII characters are supported in pattern.";
+ case UnsupportedCharacterClass:
+ return "Character class is not supported.";
+ case MisplacedQuantifier:
+ return "Quantifier without corresponding term to quantify.";
+ case BackReference:
+ return "Patterns cannot contain backreferences.";
+ case MisplacedStartOfLine:
+ return "Start of line assertion can only appear as the first term in a filter.";
+ case WordBoundary:
+ return "Word boundaries assertions are not supported yet.";
+ case AtomCharacter:
+ return "Builtins character class atoms are not supported yet.";
+ case Group:
+ return "Groups are not supported yet.";
+ case Disjunction:
+ return "Disjunctions are not supported yet.";
+ case MisplacedEndOfLine:
+ return "The end of line assertion must be the last term in an _expression_.";
+ case EmptyPattern:
+ return "Empty pattern.";
+ case YarrError:
+ return "Internal error in YARR.";
+ case InvalidQuantifier:
+ return "Arbitrary atom repetitions are not supported.";
+ }
+}
+
} // namespace ContentExtensions
} // namespace WebCore
Modified: trunk/Source/WebCore/contentextensions/URLFilterParser.h (181725 => 181726)
--- trunk/Source/WebCore/contentextensions/URLFilterParser.h 2015-03-19 04:05:51 UTC (rev 181725)
+++ trunk/Source/WebCore/contentextensions/URLFilterParser.h 2015-03-19 04:57:20 UTC (rev 181726)
@@ -39,11 +39,31 @@
struct PrefixTreeEntry;
-class URLFilterParser {
+class WEBCORE_EXPORT URLFilterParser {
public:
+ enum ParseStatus {
+ Ok,
+ MatchesEverything,
+ UnclosedGroups,
+ CannotMatchAnything,
+ NonASCII,
+ UnsupportedCharacterClass,
+ MisplacedQuantifier,
+ BackReference,
+ MisplacedStartOfLine,
+ WordBoundary,
+ AtomCharacter,
+ Group,
+ Disjunction,
+ MisplacedEndOfLine,
+ EmptyPattern,
+ YarrError,
+ InvalidQuantifier,
+ };
+ static String statusString(ParseStatus);
explicit URLFilterParser(NFA&);
~URLFilterParser();
- String addPattern(const String& pattern, bool patternIsCaseSensitive, uint64_t patternId);
+ ParseStatus addPattern(const String& pattern, bool patternIsCaseSensitive, uint64_t patternId);
private:
NFA& m_nfa;
Modified: trunk/Tools/ChangeLog (181725 => 181726)
--- trunk/Tools/ChangeLog 2015-03-19 04:05:51 UTC (rev 181725)
+++ trunk/Tools/ChangeLog 2015-03-19 04:57:20 UTC (rev 181726)
@@ -1,3 +1,15 @@
+2015-03-18 Alex Christensen <[email protected]>
+
+ [ContentExtensions] Prepare for compiling stylesheets of selectors to be used on every page.
+ https://bugs.webkit.org/show_bug.cgi?id=142799
+
+ Reviewed by Brady Eidson.
+
+ * TestWebKitAPI/Tests/WebCore/ContentExtensions.cpp:
+ (TestWebKitAPI::testPattern):
+ (TestWebKitAPI::TEST_F):
+ Start testing regex failures.
+
2015-03-18 Dhi Aurrahman <[email protected]>
Fix StringView typos after r181525 and r181558
Modified: trunk/Tools/TestWebKitAPI/Tests/WebCore/ContentExtensions.cpp (181725 => 181726)
--- trunk/Tools/TestWebKitAPI/Tests/WebCore/ContentExtensions.cpp 2015-03-19 04:05:51 UTC (rev 181725)
+++ trunk/Tools/TestWebKitAPI/Tests/WebCore/ContentExtensions.cpp 2015-03-19 04:57:20 UTC (rev 181726)
@@ -29,10 +29,13 @@
#include <_javascript_Core/InitializeThreading.h>
#include <WebCore/ContentExtensionCompiler.h>
#include <WebCore/ContentExtensionsBackend.h>
+#include <WebCore/NFA.h>
#include <WebCore/ResourceLoadInfo.h>
#include <WebCore/URL.h>
+#include <WebCore/URLFilterParser.h>
#include <wtf/MainThread.h>
#include <wtf/RunLoop.h>
+#include <wtf/text/CString.h>
namespace WebCore {
namespace ContentExtensions {
@@ -335,4 +338,19 @@
testRequest(backend, mainDocumentRequest("http://block_only_images.org", ResourceType::Document), { });
}
+static void testPatternStatus(const char* pattern, ContentExtensions::URLFilterParser::ParseStatus status)
+{
+ ContentExtensions::NFA nfa;
+ ContentExtensions::URLFilterParser parser(nfa);
+ EXPECT_EQ(status, parser.addPattern(ASCIILiteral(pattern), false, 0));
+}
+
+TEST_F(ContentExtensionTest, ParsingFailures)
+{
+ testPatternStatus("a*b?.*.?[a-z]?[a-z]*", ContentExtensions::URLFilterParser::ParseStatus::MatchesEverything);
+ testPatternStatus("a*b?.*.?[a-z]?[a-z]+", ContentExtensions::URLFilterParser::ParseStatus::Ok);
+ testPatternStatus("a*b?.*.?[a-z]?[a-z]", ContentExtensions::URLFilterParser::ParseStatus::Ok);
+ // FIXME: Add regexes that cause each parse status.
+}
+
} // namespace TestWebKitAPI
