Skip to content

V
V8
Commit d048600c authored by Frank Tang's avatar Frank Tang Committed by Commit Bot
Browse files
Options

[Intl] mv code from builtins/builtins-intl.cc to objects/js-number-format.* 

Bug: v8:5751
Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng
Change-Id: I334a6eca81c02bdb90b2054f085bf57dec5f9a9d
Reviewed-on: https://chromium-review.googlesource.com/1215645Reviewed-by: 's avatarSathya Gunasekaran <gsathya@chromium.org>
Commit-Queue: Frank Tang <ftang@chromium.org>
Cr-Commit-Position: refs/heads/master@{#55775}
parent eec50fa3
Show whitespace changes
Inline Side-by-side
Showing with 226 additions and 242 deletions
BUILD.gn
View file @ d048600c
......@@ -1618,7 +1618,6 @@ v8_source_set("v8_base") {
"src/builtins/builtins-internal.cc",
"src/builtins/builtins-interpreter.cc",
"src/builtins/builtins-intl.cc",
"src/builtins/builtins-intl.h",
"src/builtins/builtins-json.cc",
"src/builtins/builtins-math.cc",
"src/builtins/builtins-number.cc",
......@@ -2920,7 +2919,6 @@ v8_source_set("v8_base") {
} else {
sources -= [
"src/builtins/builtins-intl.cc",
"src/builtins/builtins-intl.h",
"src/char-predicates.cc",
"src/intl.cc",
"src/intl.h",
......
src/builtins/builtins-intl.cc
View file @ d048600c
......@@ -10,7 +10,6 @@
#include <list>
#include <memory>
#include "src/builtins/builtins-intl.h"
#include "src/builtins/builtins-utils-inl.h"
#include "src/builtins/builtins.h"
#include "src/date.h"
......@@ -132,50 +131,6 @@ BUILTIN(StringPrototypeNormalizeIntl) {
namespace {
// The list comes from third_party/icu/source/i18n/unicode/unum.h.
// They're mapped to NumberFormat part types mentioned throughout
// https://tc39.github.io/ecma402/#sec-partitionnumberpattern .
Handle<String> IcuNumberFieldIdToNumberType(int32_t field_id, double number,
Isolate* isolate) {
switch (static_cast<UNumberFormatFields>(field_id)) {
case UNUM_INTEGER_FIELD:
if (std::isfinite(number)) return isolate->factory()->integer_string();
if (std::isnan(number)) return isolate->factory()->nan_string();
return isolate->factory()->infinity_string();
case UNUM_FRACTION_FIELD:
return isolate->factory()->fraction_string();
case UNUM_DECIMAL_SEPARATOR_FIELD:
return isolate->factory()->decimal_string();
case UNUM_GROUPING_SEPARATOR_FIELD:
return isolate->factory()->group_string();
case UNUM_CURRENCY_FIELD:
return isolate->factory()->currency_string();
case UNUM_PERCENT_FIELD:
return isolate->factory()->percentSign_string();
case UNUM_SIGN_FIELD:
return number < 0 ? isolate->factory()->minusSign_string()
: isolate->factory()->plusSign_string();
case UNUM_EXPONENT_SYMBOL_FIELD:
case UNUM_EXPONENT_SIGN_FIELD:
case UNUM_EXPONENT_FIELD:
// We should never get these because we're not using any scientific
// formatter.
UNREACHABLE();
return Handle<String>();
case UNUM_PERMILL_FIELD:
// We're not creating any permill formatter, and it's not even clear how
// that would be possible with the ICU API.
UNREACHABLE();
return Handle<String>();
default:
UNREACHABLE();
return Handle<String>();
}
}
// The list comes from third_party/icu/source/i18n/unicode/udat.h.
// They're mapped to DateTimeFormat components listed at
// https://tc39.github.io/ecma402/#sec-datetimeformat-abstracts .
......@@ -227,74 +182,6 @@ Handle<String> IcuDateFieldIdToDateType(int32_t field_id, Isolate* isolate) {
}
}
bool cmp_NumberFormatSpan(const NumberFormatSpan& a,
const NumberFormatSpan& b) {
// Regions that start earlier should be encountered earlier.
if (a.begin_pos < b.begin_pos) return true;
if (a.begin_pos > b.begin_pos) return false;
// For regions that start in the same place, regions that last longer should
// be encountered earlier.
if (a.end_pos < b.end_pos) return false;
if (a.end_pos > b.end_pos) return true;
// For regions that are exactly the same, one of them must be the "literal"
// backdrop we added, which has a field_id of -1, so consider higher field_ids
// to be later.
return a.field_id < b.field_id;
}
MaybeHandle<Object> FormatNumberToParts(Isolate* isolate,
icu::NumberFormat* fmt, double number) {
Factory* factory = isolate->factory();
icu::UnicodeString formatted;
icu::FieldPositionIterator fp_iter;
UErrorCode status = U_ZERO_ERROR;
fmt->format(number, formatted, &fp_iter, status);
if (U_FAILURE(status)) {
THROW_NEW_ERROR(isolate, NewTypeError(MessageTemplate::kIcuError), Object);
}
Handle<JSArray> result = factory->NewJSArray(0);
int32_t length = formatted.length();
if (length == 0) return result;
std::vector<NumberFormatSpan> regions;
// Add a "literal" backdrop for the entire string. This will be used if no
// other region covers some part of the formatted string. It's possible
// there's another field with exactly the same begin and end as this backdrop,
// in which case the backdrop's field_id of -1 will give it lower priority.
regions.push_back(NumberFormatSpan(-1, 0, formatted.length()));
{
icu::FieldPosition fp;
while (fp_iter.next(fp)) {
regions.push_back(NumberFormatSpan(fp.getField(), fp.getBeginIndex(),
fp.getEndIndex()));
}
}
std::vector<NumberFormatSpan> parts = FlattenRegionsToParts(&regions);
int index = 0;
for (auto it = parts.begin(); it < parts.end(); it++) {
NumberFormatSpan part = *it;
Handle<String> field_type_string =
part.field_id == -1
? isolate->factory()->literal_string()
: IcuNumberFieldIdToNumberType(part.field_id, number, isolate);
Handle<String> substring;
ASSIGN_RETURN_ON_EXCEPTION(
isolate, substring,
Intl::ToString(isolate, formatted, part.begin_pos, part.end_pos),
Object);
Intl::AddElement(isolate, result, index, field_type_string, substring);
++index;
}
JSObject::ValidateElements(*result);
return result;
}
MaybeHandle<Object> FormatDateToParts(Isolate* isolate, icu::DateFormat* format,
double date_value) {
Factory* factory = isolate->factory();
......@@ -380,96 +267,6 @@ BUILTIN(V8BreakIteratorSupportedLocalesOf) {
isolate, SupportedLocalesOfCommon(isolate, "breakiterator", args));
}
// Flattens a list of possibly-overlapping "regions" to a list of
// non-overlapping "parts". At least one of the input regions must span the
// entire space of possible indexes. The regions parameter will sorted in-place
// according to some criteria; this is done for performance to avoid copying the
// input.
std::vector<NumberFormatSpan> FlattenRegionsToParts(
std::vector<NumberFormatSpan>* regions) {
// The intention of this algorithm is that it's used to translate ICU "fields"
// to JavaScript "parts" of a formatted string. Each ICU field and JavaScript
// part has an integer field_id, which corresponds to something like "grouping
// separator", "fraction", or "percent sign", and has a begin and end
// position. Here's a diagram of:
// var nf = new Intl.NumberFormat(['de'], {style:'currency',currency:'EUR'});
// nf.formatToParts(123456.78);
// : 6
// input regions: 0000000211 7
// ('-' means -1): ------------
// formatted string: "123.456,78 €"
// output parts: 0006000211-7
// To illustrate the requirements of this algorithm, here's a contrived and
// convoluted example of inputs and expected outputs:
// : 4
// : 22 33 3
// : 11111 22
// input regions: 0000000 111
// : ------------
// formatted string: "abcdefghijkl"
// output parts: 0221340--231
// (The characters in the formatted string are irrelevant to this function.)
// We arrange the overlapping input regions like a mountain range where
// smaller regions are "on top" of larger regions, and we output a birds-eye
// view of the mountains, so that smaller regions take priority over larger
// regions.
std::sort(regions->begin(), regions->end(), cmp_NumberFormatSpan);
std::vector<size_t> overlapping_region_index_stack;
// At least one item in regions must be a region spanning the entire string.
// Due to the sorting above, the first item in the vector will be one of them.
overlapping_region_index_stack.push_back(0);
NumberFormatSpan top_region = regions->at(0);
size_t region_iterator = 1;
int32_t entire_size = top_region.end_pos;
std::vector<NumberFormatSpan> out_parts;
// The "climber" is a cursor that advances from left to right climbing "up"
// and "down" the mountains. Whenever the climber moves to the right, that
// represents an item of output.
int32_t climber = 0;
while (climber < entire_size) {
int32_t next_region_begin_pos;
if (region_iterator < regions->size()) {
next_region_begin_pos = regions->at(region_iterator).begin_pos;
} else {
// finish off the rest of the input by proceeding to the end.
next_region_begin_pos = entire_size;
}
if (climber < next_region_begin_pos) {
while (top_region.end_pos < next_region_begin_pos) {
if (climber < top_region.end_pos) {
// step down
out_parts.push_back(NumberFormatSpan(top_region.field_id, climber,
top_region.end_pos));
climber = top_region.end_pos;
} else {
// drop down
}
overlapping_region_index_stack.pop_back();
top_region = regions->at(overlapping_region_index_stack.back());
}
if (climber < next_region_begin_pos) {
// cross a plateau/mesa/valley
out_parts.push_back(NumberFormatSpan(top_region.field_id, climber,
next_region_begin_pos));
climber = next_region_begin_pos;
}
}
if (region_iterator < regions->size()) {
overlapping_region_index_stack.push_back(region_iterator++);
top_region = regions->at(overlapping_region_index_stack.back());
}
}
return out_parts;
}
BUILTIN(NumberFormatSupportedLocalesOf) {
HandleScope scope(isolate);
RETURN_RESULT_OR_FAILURE(
......@@ -489,12 +286,8 @@ BUILTIN(NumberFormatPrototypeFormatToParts) {
x = isolate->factory()->nan_value();
}
icu::NumberFormat* icu_number_format =
number_format->icu_number_format()->raw();
CHECK_NOT_NULL(icu_number_format);
RETURN_RESULT_OR_FAILURE(
isolate, FormatNumberToParts(isolate, icu_number_format, x->Number()));
RETURN_RESULT_OR_FAILURE(isolate, JSNumberFormat::FormatToParts(
isolate, number_format, x->Number()));
}
BUILTIN(DateTimeFormatSupportedLocalesOf) {
......
src/builtins/builtins-intl.h deleted 100644 → 0
View file @ eec50fa3
// Copyright 2017 the V8 project authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#ifndef V8_BUILTINS_BUILTINS_INTL_H_
#define V8_BUILTINS_BUILTINS_INTL_H_
#include <stdint.h>
#include <vector>
namespace v8 {
namespace internal {
struct NumberFormatSpan {
int32_t field_id;
int32_t begin_pos;
int32_t end_pos;
NumberFormatSpan() {}
NumberFormatSpan(int32_t field_id, int32_t begin_pos, int32_t end_pos)
: field_id(field_id), begin_pos(begin_pos), end_pos(end_pos) {}
};
std::vector<NumberFormatSpan> FlattenRegionsToParts(
std::vector<NumberFormatSpan>* regions);
} // namespace internal
} // namespace v8
#endif // V8_BUILTINS_BUILTINS_INTL_H_
src/objects/js-number-format.cc
View file @ d048600c
......@@ -501,5 +501,212 @@ MaybeHandle<String> JSNumberFormat::FormatNumber(
reinterpret_cast<const uint16_t*>(result.getBuffer()), result.length()));
}
namespace {
bool cmp_NumberFormatSpan(const NumberFormatSpan& a,
const NumberFormatSpan& b) {
// Regions that start earlier should be encountered earlier.
if (a.begin_pos < b.begin_pos) return true;
if (a.begin_pos > b.begin_pos) return false;
// For regions that start in the same place, regions that last longer should
// be encountered earlier.
if (a.end_pos < b.end_pos) return false;
if (a.end_pos > b.end_pos) return true;
// For regions that are exactly the same, one of them must be the "literal"
// backdrop we added, which has a field_id of -1, so consider higher field_ids
// to be later.
return a.field_id < b.field_id;
}
// The list comes from third_party/icu/source/i18n/unicode/unum.h.
// They're mapped to NumberFormat part types mentioned throughout
// https://tc39.github.io/ecma402/#sec-partitionnumberpattern .
Handle<String> IcuNumberFieldIdToNumberType(int32_t field_id, double number,
Isolate* isolate) {
switch (static_cast<UNumberFormatFields>(field_id)) {
case UNUM_INTEGER_FIELD:
if (std::isfinite(number)) return isolate->factory()->integer_string();
if (std::isnan(number)) return isolate->factory()->nan_string();
return isolate->factory()->infinity_string();
case UNUM_FRACTION_FIELD:
return isolate->factory()->fraction_string();
case UNUM_DECIMAL_SEPARATOR_FIELD:
return isolate->factory()->decimal_string();
case UNUM_GROUPING_SEPARATOR_FIELD:
return isolate->factory()->group_string();
case UNUM_CURRENCY_FIELD:
return isolate->factory()->currency_string();
case UNUM_PERCENT_FIELD:
return isolate->factory()->percentSign_string();
case UNUM_SIGN_FIELD:
return number < 0 ? isolate->factory()->minusSign_string()
: isolate->factory()->plusSign_string();
case UNUM_EXPONENT_SYMBOL_FIELD:
case UNUM_EXPONENT_SIGN_FIELD:
case UNUM_EXPONENT_FIELD:
// We should never get these because we're not using any scientific
// formatter.
UNREACHABLE();
return Handle<String>();
case UNUM_PERMILL_FIELD:
// We're not creating any permill formatter, and it's not even clear how
// that would be possible with the ICU API.
UNREACHABLE();
return Handle<String>();
default:
UNREACHABLE();
return Handle<String>();
}
}
} // namespace
// Flattens a list of possibly-overlapping "regions" to a list of
// non-overlapping "parts". At least one of the input regions must span the
// entire space of possible indexes. The regions parameter will sorted in-place
// according to some criteria; this is done for performance to avoid copying the
// input.
std::vector<NumberFormatSpan> FlattenRegionsToParts(
std::vector<NumberFormatSpan>* regions) {
// The intention of this algorithm is that it's used to translate ICU "fields"
// to JavaScript "parts" of a formatted string. Each ICU field and JavaScript
// part has an integer field_id, which corresponds to something like "grouping
// separator", "fraction", or "percent sign", and has a begin and end
// position. Here's a diagram of:
// var nf = new Intl.NumberFormat(['de'], {style:'currency',currency:'EUR'});
// nf.formatToParts(123456.78);
// : 6
// input regions: 0000000211 7
// ('-' means -1): ------------
// formatted string: "123.456,78 €"
// output parts: 0006000211-7
// To illustrate the requirements of this algorithm, here's a contrived and
// convoluted example of inputs and expected outputs:
// : 4
// : 22 33 3
// : 11111 22
// input regions: 0000000 111
// : ------------
// formatted string: "abcdefghijkl"
// output parts: 0221340--231
// (The characters in the formatted string are irrelevant to this function.)
// We arrange the overlapping input regions like a mountain range where
// smaller regions are "on top" of larger regions, and we output a birds-eye
// view of the mountains, so that smaller regions take priority over larger
// regions.
std::sort(regions->begin(), regions->end(), cmp_NumberFormatSpan);
std::vector<size_t> overlapping_region_index_stack;
// At least one item in regions must be a region spanning the entire string.
// Due to the sorting above, the first item in the vector will be one of them.
overlapping_region_index_stack.push_back(0);
NumberFormatSpan top_region = regions->at(0);
size_t region_iterator = 1;
int32_t entire_size = top_region.end_pos;
std::vector<NumberFormatSpan> out_parts;
// The "climber" is a cursor that advances from left to right climbing "up"
// and "down" the mountains. Whenever the climber moves to the right, that
// represents an item of output.
int32_t climber = 0;
while (climber < entire_size) {
int32_t next_region_begin_pos;
if (region_iterator < regions->size()) {
next_region_begin_pos = regions->at(region_iterator).begin_pos;
} else {
// finish off the rest of the input by proceeding to the end.
next_region_begin_pos = entire_size;
}
if (climber < next_region_begin_pos) {
while (top_region.end_pos < next_region_begin_pos) {
if (climber < top_region.end_pos) {
// step down
out_parts.push_back(NumberFormatSpan(top_region.field_id, climber,
top_region.end_pos));
climber = top_region.end_pos;
} else {
// drop down
}
overlapping_region_index_stack.pop_back();
top_region = regions->at(overlapping_region_index_stack.back());
}
if (climber < next_region_begin_pos) {
// cross a plateau/mesa/valley
out_parts.push_back(NumberFormatSpan(top_region.field_id, climber,
next_region_begin_pos));
climber = next_region_begin_pos;
}
}
if (region_iterator < regions->size()) {
overlapping_region_index_stack.push_back(region_iterator++);
top_region = regions->at(overlapping_region_index_stack.back());
}
}
return out_parts;
}
MaybeHandle<JSArray> JSNumberFormat::FormatToParts(
Isolate* isolate, Handle<JSNumberFormat> number_format, double number) {
Factory* factory = isolate->factory();
icu::NumberFormat* fmt = number_format->icu_number_format()->raw();
CHECK_NOT_NULL(fmt);
icu::UnicodeString formatted;
icu::FieldPositionIterator fp_iter;
UErrorCode status = U_ZERO_ERROR;
fmt->format(number, formatted, &fp_iter, status);
if (U_FAILURE(status)) {
THROW_NEW_ERROR(isolate, NewTypeError(MessageTemplate::kIcuError), JSArray);
}
Handle<JSArray> result = factory->NewJSArray(0);
int32_t length = formatted.length();
if (length == 0) return result;
std::vector<NumberFormatSpan> regions;
// Add a "literal" backdrop for the entire string. This will be used if no
// other region covers some part of the formatted string. It's possible
// there's another field with exactly the same begin and end as this backdrop,
// in which case the backdrop's field_id of -1 will give it lower priority.
regions.push_back(NumberFormatSpan(-1, 0, formatted.length()));
{
icu::FieldPosition fp;
while (fp_iter.next(fp)) {
regions.push_back(NumberFormatSpan(fp.getField(), fp.getBeginIndex(),
fp.getEndIndex()));
}
}
std::vector<NumberFormatSpan> parts = FlattenRegionsToParts(&regions);
int index = 0;
for (auto it = parts.begin(); it < parts.end(); it++) {
NumberFormatSpan part = *it;
Handle<String> field_type_string =
part.field_id == -1
? isolate->factory()->literal_string()
: IcuNumberFieldIdToNumberType(part.field_id, number, isolate);
Handle<String> substring;
ASSIGN_RETURN_ON_EXCEPTION(
isolate, substring,
Intl::ToString(isolate, formatted, part.begin_pos, part.end_pos),
JSArray);
Intl::AddElement(isolate, result, index, field_type_string, substring);
++index;
}
JSObject::ValidateElements(*result);
return result;
}
} // namespace internal
} // namespace v8
src/objects/js-number-format.h
View file @ d048600c
......@@ -40,6 +40,9 @@ class JSNumberFormat : public JSObject {
static Handle<JSObject> ResolvedOptions(Isolate* isolate,
Handle<JSNumberFormat> number_format);
V8_WARN_UNUSED_RESULT static MaybeHandle<JSArray> FormatToParts(
Isolate* isolate, Handle<JSNumberFormat> number_format, double number);
V8_WARN_UNUSED_RESULT static MaybeHandle<String> FormatNumber(
Isolate* isolate, Handle<JSNumberFormat> number_format, double number);
......@@ -115,6 +118,19 @@ class JSNumberFormat : public JSObject {
DISALLOW_IMPLICIT_CONSTRUCTORS(JSNumberFormat);
};
struct NumberFormatSpan {
int32_t field_id;
int32_t begin_pos;
int32_t end_pos;
NumberFormatSpan() {}
NumberFormatSpan(int32_t field_id, int32_t begin_pos, int32_t end_pos)
: field_id(field_id), begin_pos(begin_pos), end_pos(end_pos) {}
};
std::vector<NumberFormatSpan> FlattenRegionsToParts(
std::vector<NumberFormatSpan>* regions);
} // namespace internal
} // namespace v8
......
test/cctest/test-intl.cc
View file @ d048600c
......@@ -4,10 +4,10 @@
#ifdef V8_INTL_SUPPORT
#include "src/builtins/builtins-intl.h"
#include "src/lookup.h"
#include "src/objects-inl.h"
#include "src/objects/intl-objects.h"
#include "src/objects/js-number-format.h"
#include "test/cctest/cctest.h"
namespace v8 {
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment