Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(1162)

Issues Search
    My Issues | Starred     Open | Closed | All

Unified Diff: src/regexp/arm64/regexp-macro-assembler-arm64.cc

Issue 1418963009: Experimental support for RegExp lookbehind. (Closed) Base URL: https://chromium.googlesource.com/v8/v8.git@master
Patch Set: mips64 port Created 5 years, 1 month ago
Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.
Jump to:
View side-by-side diff with in-line comments
Download patch
« src/regexp/arm/regexp-macro-assembler-arm.cc ('K') | « src/regexp/arm64/regexp-macro-assembler-arm64.h ('k') | src/regexp/bytecodes-irregexp.h » ('j') | no next file with comments »
Expand Comments ('e') | Collapse Comments ('c') | Show Comments Hide Comments ('s')
Index: src/regexp/arm64/regexp-macro-assembler-arm64.cc
diff --git a/src/regexp/arm64/regexp-macro-assembler-arm64.cc b/src/regexp/arm64/regexp-macro-assembler-arm64.cc
index d440879e26f87f00c74dcbf8e8ea0e4a52440fd7..3c20956e6fbdb1d53384c3364c7ff9528d0dff12 100644
--- a/src/regexp/arm64/regexp-macro-assembler-arm64.cc
+++ b/src/regexp/arm64/regexp-macro-assembler-arm64.cc
@@ -210,22 +210,22 @@ void RegExpMacroAssemblerARM64::CheckCharacterGT(uc16 limit,
void RegExpMacroAssemblerARM64::CheckAtStart(Label* on_at_start) {
- Label not_at_start;
- // Did we start the match at the start of the input string?
- CompareAndBranchOrBacktrack(start_offset(), 0, ne, &not_at_start);
- // If we did, are we still at the start of the input string?
__ Add(x10, input_end(), Operand(current_input_offset(), SXTW));
+ for (int i = 0; i < char_size(); i++) {
+ __ Add(x10, x10, Operand(start_offset(), SXTW));
+ }
__ Cmp(x10, input_start());
BranchOrBacktrack(eq, on_at_start);
- __ Bind(&not_at_start);
}
-void RegExpMacroAssemblerARM64::CheckNotAtStart(Label* on_not_at_start) {
- // Did we start the match at the start of the input string?
- CompareAndBranchOrBacktrack(start_offset(), 0, ne, on_not_at_start);
- // If we did, are we still at the start of the input string?
+void RegExpMacroAssemblerARM64::CheckNotAtStart(int cp_offset,
+ Label* on_not_at_start) {
__ Add(x10, input_end(), Operand(current_input_offset(), SXTW));
+ __ Add(x10, x10, cp_offset * char_size());
+ for (int i = 0; i < char_size(); i++) {
+ __ Add(x10, x10, Operand(start_offset(), SXTW));
+ }
__ Cmp(x10, input_start());
BranchOrBacktrack(ne, on_not_at_start);
}
@@ -277,9 +277,9 @@ void RegExpMacroAssemblerARM64::CheckGreedyLoop(Label* on_equal) {
BranchOrBacktrack(eq, on_equal);
}
+
void RegExpMacroAssemblerARM64::CheckNotBackReferenceIgnoreCase(
- int start_reg,
- Label* on_no_match) {
+ int start_reg, bool read_backward, Label* on_no_match) {
Label fallthrough;
Register capture_start_offset = w10;
@@ -297,12 +297,24 @@ void RegExpMacroAssemblerARM64::CheckNotBackReferenceIgnoreCase(
__ Ldp(w11, capture_start_offset, capture_location(start_reg, x10));
}
__ Sub(capture_length, w11, capture_start_offset); // Length to check.
- // Succeed on empty capture (including no capture).
- __ Cbz(capture_length, &fallthrough);
+
+ // The length of the capture can only be negative if the end of the
+ // capture is not yet recorded. If the length is zero, the capture is
+ // either empty or uncaptured. In either of those cases, succeed.
+ __ CompareAndBranch(capture_length, Operand(0), le, &fallthrough);
// Check that there are enough characters left in the input.
- __ Cmn(capture_length, current_input_offset());
- BranchOrBacktrack(gt, on_no_match);
+ if (read_backward) {
+ __ Add(x12, input_end(), Operand(current_input_offset(), SXTW));
+ for (int i = 0; i < char_size(); i++) {
+ __ Add(x12, x12, Operand(start_offset(), SXTW));
+ }
+ __ Cmp(x12, input_start());
+ BranchOrBacktrack(lt, on_no_match);
+ } else {
+ __ Cmn(capture_length, current_input_offset());
+ BranchOrBacktrack(gt, on_no_match);
+ }
if (mode_ == LATIN1) {
Label success;
@@ -322,6 +334,11 @@ void RegExpMacroAssemblerARM64::CheckNotBackReferenceIgnoreCase(
__ Add(current_position_address,
input_end(),
Operand(current_input_offset(), SXTW));
+ if (read_backward) {
+ // Offset by length when matching backwards.
+ __ Sub(current_position_address, current_position_address,
+ Operand(capture_length, SXTW));
+ }
Label loop;
__ Bind(&loop);
@@ -361,6 +378,10 @@ void RegExpMacroAssemblerARM64::CheckNotBackReferenceIgnoreCase(
// The current input offset should be <= 0, and fit in a W register.
__ Check(le, kOffsetOutOfRange);
}
+ if (read_backward) {
+ __ Sub(current_input_offset(), current_input_offset(),
+ Operand(capture_length, SXTW));
+ }
} else {
DCHECK(mode_ == UC16);
int argument_count = 4;
@@ -383,6 +404,9 @@ void RegExpMacroAssemblerARM64::CheckNotBackReferenceIgnoreCase(
__ Mov(w2, capture_length);
// Address of current input position.
__ Add(x1, input_end(), Operand(current_input_offset(), SXTW));
+ if (read_backward) {
+ __ Sub(x1, x1, Operand(capture_length, SXTW));
+ }
// Isolate.
__ Mov(x3, ExternalReference::isolate_address(isolate()));
@@ -400,16 +424,20 @@ void RegExpMacroAssemblerARM64::CheckNotBackReferenceIgnoreCase(
__ PopCPURegList(cached_registers);
BranchOrBacktrack(eq, on_no_match);
- // On success, increment position by length of capture.
- __ Add(current_input_offset(), current_input_offset(), capture_length);
+ // On success, advance position by length of capture.
+ if (read_backward) {
+ __ Sub(current_input_offset(), current_input_offset(), capture_length);
+ } else {
+ __ Add(current_input_offset(), current_input_offset(), capture_length);
+ }
}
__ Bind(&fallthrough);
}
-void RegExpMacroAssemblerARM64::CheckNotBackReference(
- int start_reg,
- Label* on_no_match) {
+void RegExpMacroAssemblerARM64::CheckNotBackReference(int start_reg,
+ bool read_backward,
+ Label* on_no_match) {
Label fallthrough;
Register capture_start_address = x12;
@@ -426,12 +454,23 @@ void RegExpMacroAssemblerARM64::CheckNotBackReference(
__ Ldp(w11, w10, capture_location(start_reg, x10));
}
__ Sub(capture_length, w11, w10); // Length to check.
- // Succeed on empty capture (including no capture).
- __ Cbz(capture_length, &fallthrough);
+ // The length of the capture can only be negative if the end of the
+ // capture is not yet recorded. If the length is zero, the capture is
+ // either empty or uncaptured. In either of those cases, succeed.
+ __ CompareAndBranch(capture_length, Operand(0), le, &fallthrough);
// Check that there are enough characters left in the input.
- __ Cmn(capture_length, current_input_offset());
- BranchOrBacktrack(gt, on_no_match);
+ if (read_backward) {
+ __ Add(x12, input_end(), Operand(current_input_offset(), SXTW));
+ for (int i = 0; i < char_size(); i++) {
+ __ Add(x12, x12, Operand(start_offset(), SXTW));
+ }
+ __ Cmp(x12, input_start());
+ BranchOrBacktrack(lt, on_no_match);
+ } else {
+ __ Cmn(capture_length, current_input_offset());
+ BranchOrBacktrack(gt, on_no_match);
+ }
// Compute pointers to match string and capture string
__ Add(capture_start_address, input_end(), Operand(w10, SXTW));
@@ -441,6 +480,11 @@ void RegExpMacroAssemblerARM64::CheckNotBackReference(
__ Add(current_position_address,
input_end(),
Operand(current_input_offset(), SXTW));
+ if (read_backward) {
+ // Offset by length when matching backwards.
+ __ Sub(current_position_address, current_position_address,
+ Operand(capture_length, SXTW));
+ }
Label loop;
__ Bind(&loop);
@@ -459,6 +503,11 @@ void RegExpMacroAssemblerARM64::CheckNotBackReference(
// Move current character position to position after match.
__ Sub(current_input_offset().X(), current_position_address, input_end());
+ if (read_backward) {
+ __ Sub(current_input_offset(), current_input_offset(),
+ Operand(capture_length, SXTW));
+ }
+
if (masm_->emit_debug_code()) {
__ Cmp(current_input_offset().X(), Operand(current_input_offset(), SXTW));
__ Ccmp(current_input_offset(), 0, NoFlag, eq);
@@ -1081,11 +1130,14 @@ void RegExpMacroAssemblerARM64::LoadCurrentCharacter(int cp_offset,
int characters) {
// TODO(pielan): Make sure long strings are caught before this, and not
// just asserted in debug mode.
- DCHECK(cp_offset >= -1); // ^ and \b can look behind one character.
// Be sane! (And ensure that an int32_t can be used to index the string)
DCHECK(cp_offset < (1<<30));
if (check_bounds) {
- CheckPosition(cp_offset + characters - 1, on_end_of_input);
+ if (cp_offset >= 0) {
+ CheckPosition(cp_offset + characters - 1, on_end_of_input);
+ } else {
+ CheckPosition(cp_offset, on_end_of_input);
+ }
}
LoadCurrentCharacterUnchecked(cp_offset, characters);
}
@@ -1301,10 +1353,18 @@ int RegExpMacroAssemblerARM64::CheckStackGuardState(
void RegExpMacroAssemblerARM64::CheckPosition(int cp_offset,
Label* on_outside_input) {
- CompareAndBranchOrBacktrack(current_input_offset(),
- -cp_offset * char_size(),
- ge,
- on_outside_input);
+ if (cp_offset >= 0) {
+ CompareAndBranchOrBacktrack(current_input_offset(),
+ -cp_offset * char_size(), ge, on_outside_input);
+ } else {
+ __ Add(x12, input_end(), Operand(current_input_offset(), SXTW));
+ __ Add(x12, x12, cp_offset * char_size());
+ for (int i = 0; i < char_size(); i++) {
+ __ Add(x12, x12, Operand(start_offset(), SXTW));
+ }
+ __ Cmp(x12, input_start());
+ BranchOrBacktrack(lt, on_outside_input);
+ }
}
« src/regexp/arm/regexp-macro-assembler-arm.cc ('K') | « src/regexp/arm64/regexp-macro-assembler-arm64.h ('k') | src/regexp/bytecodes-irregexp.h » ('j') | no next file with comments »

Powered by Google App Engine
This is Rietveld 408576698