Apache
/
lucene
mirror of https://github.com/apache/lucene.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

LUCENE-10238: Update icu4j to 70.1. (#447)

This commit is contained in:
Dawid Weiss 2021-11-17 18:13:40 +01:00 committed by GitHub
parent 556c7c5fb5
commit 0eeba8d37c
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
29 changed files with 37030 additions and 36936 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

36
gradle/generation/icu.gradle
View File

@ -33,21 +33,26 @@ def resources = scriptResources(buildscript)
// Configure different icu4j dependencies.
configure(rootProject) {
configurations {
icu_62
icu_68
// icu_xyz
icu_current
}
dependencies {
icu_62 "com.ibm.icu:icu4j:62.2"
icu_68 "com.ibm.icu:icu4j:68.2"
// icu_xyz "com.ibm.icu:icu4j:xyz"
icu_current 'com.ibm.icu:icu4j'
}
// Exclude ICU config from palantir's version unification.
// Exclude explicit ICU configs from palantir's version unification.
versionRecommendations {
excludeConfigurations "icu_68", "icu_62"
// excludeConfigurations "icu_xyz"
}
}
// This retrieves the module name/ version for ICU from the given
// configuration (if present).
def icuVersionFromConfiguration(Configuration configuration) {
return configuration.resolvedConfiguration.getFirstLevelModuleDependencies({ dep -> dep.group.startsWith("com.ibm.icu") }).collect { dep -> dep.module.id }.join(", ")
}
configure(project(":lucene:analysis:icu")) {
def utr30DataDir = file("src/data/utr30")
@ -75,6 +80,9 @@ configure(project(":lucene:analysis:icu")) {
def icuCompileTask = Os.isFamily(Os.FAMILY_WINDOWS) ? "compileIcuWindows" : "compileIcuLinux"
task genUtr30DataFilesInternal() {
def icuConfig = rootProject.configurations.icu_current
dependsOn icuConfig
dependsOn icuCompileTask
// May be undefined yet, so use a provider.
@ -85,7 +93,7 @@ configure(project(":lucene:analysis:icu")) {
def outputFile = file("src/resources/org/apache/lucene/analysis/icu/utr30.nrm")
inputs.files inputFiles
inputs.property "icu4j", icu4jVersionProvider
inputs.property "icuConfig", provider { icuVersionFromConfiguration(icuConfig) }
outputs.file outputFile
doFirst {
@ -125,6 +133,10 @@ configure(project(":lucene:analysis:icu")) {
}
task genRbbiInternal() {
def icuConfig = rootProject.configurations.icu_current
dependsOn icuConfig
// May be undefined yet, so use a provider.
dependsOn { sourceSets.tools.runtimeClasspath }
@ -132,7 +144,7 @@ configure(project(":lucene:analysis:icu")) {
def targetDir = file("src/resources/org/apache/lucene/analysis/icu/segmentation")
inputs.files fileTree(dir: sourceDir, include: "*.rbbi")
inputs.property "icu4j", icu4jVersionProvider
inputs.property "icuConfig", provider { icuVersionFromConfiguration(icuConfig) }
outputs.files fileTree(dir: targetDir, include: "*.brk")
doFirst {
@ -244,7 +256,7 @@ configure(project(":lucene:analysis:icu")) {
// Regenerates UnicodeProps.java
configure(project(":lucene:analysis:common")) {
task generateUnicodePropsInternal() {
def icuConfig = rootProject.configurations.icu_68
def icuConfig = rootProject.configurations.icu_current
def outputFile = file("src/java/org/apache/lucene/analysis/util/UnicodeProps.java")
description "Regenerate ${outputFile} (with ${icuConfig.name})"
@ -252,7 +264,7 @@ configure(project(":lucene:analysis:common")) {
dependsOn icuConfig
inputs.property "icuConfig", icuConfig.name
inputs.property "icuConfig", provider { icuVersionFromConfiguration(icuConfig) }
outputs.file outputFile
doFirst {
@ -276,7 +288,7 @@ configure(project(":lucene:analysis:common")) {
// UnicodeEmojiProperties.jflex
configure(project(":lucene:core")) {
task generateEmojiPropertiesInternal() {
def icuConfig = rootProject.configurations.icu_62
def icuConfig = rootProject.configurations.icu_current
def outputFile = file("src/data/jflex/UnicodeEmojiProperties.jflex")
description "Regenerate ${outputFile} (with ${icuConfig.name})"
@ -284,7 +296,7 @@ configure(project(":lucene:core")) {
dependsOn icuConfig
inputs.property "icuConfig", icuConfig.name
inputs.property "icuConfig", provider { icuVersionFromConfiguration(icuConfig) }
outputs.file outputFile
doFirst {

10
gradle/generation/regenerate.gradle
View File

@ -178,7 +178,7 @@ configure([
expected = expected - same
throw new GradleException("Checksums mismatch for derived resources; you might have" +
" modified a generated resource (regenerate task: ${sourceTask.name}):\n" +
" modified a generated resource (regenerate task: ${sourceTaskName}):\n" +
"Current:\n ${current.entrySet().join('\n ')}\n\n" +
"Expected:\n ${expected.entrySet().join('\n ')}"
)
@ -222,7 +222,13 @@ configure([
})
// Load checksums before the source task executes, otherwise it's always ignored.
sourceTask.dependsOn checksumLoadTask
project.afterEvaluate {
resolveTaskRefs([sourceTask, *ignoreWithSource]).each { t ->
t.configure {
dependsOn checksumLoadTask
}
}
}
// Copy the description and group from the source task.
project.afterEvaluate {

3
lucene/CHANGES.txt
View File

@ -47,7 +47,8 @@ New Features
Improvements
---------------------
(No changes)
* LUCENE-10238: Upgrade icu4j dependency to 70.1. (Dawid Weiss)
Optimizations
---------------------

4
lucene/analysis/common/src/generated/checksums/generateTlds.json
View File

@ -1,5 +1,5 @@
{
"lucene/analysis/common/src/java/org/apache/lucene/analysis/email/ASCIITLD.jflex": "bb3878ea10f85f124a0a9e4ea614d3400d664dae",
"lucene/analysis/common/src/test/org/apache/lucene/analysis/email/TLDs.txt": "800f0f70d7ce4723d493cd5da88a19bfc8e532e5",
"lucene/analysis/common/src/java/org/apache/lucene/analysis/email/ASCIITLD.jflex": "b79e24f254afcd138287049394f22bac9e094e13",
"lucene/analysis/common/src/test/org/apache/lucene/analysis/email/TLDs.txt": "c65096becfadf92ef811f50fbe783201c6d186ea",
"property:tldZones": "https://data.iana.org/TLD/tlds-alpha-by-domain.txt"
}

6
lucene/analysis/common/src/generated/checksums/generateUAX29URLEmailTokenizer.json
View File

@ -1,7 +1,7 @@
{
"gradle/generation/jflex/skeleton.disable.buffer.expansion.txt": "1424f4df33c977bb150d7377c3bd61f819113091",
"lucene/analysis/common/src/java/org/apache/lucene/analysis/email/ASCIITLD.jflex": "bb3878ea10f85f124a0a9e4ea614d3400d664dae",
"lucene/analysis/common/src/java/org/apache/lucene/analysis/email/UAX29URLEmailTokenizerImpl.java": "b88c349d24028f557f2c014437f3f60c968ad9de",
"lucene/analysis/common/src/java/org/apache/lucene/analysis/email/ASCIITLD.jflex": "b79e24f254afcd138287049394f22bac9e094e13",
"lucene/analysis/common/src/java/org/apache/lucene/analysis/email/UAX29URLEmailTokenizerImpl.java": "d7e19b7a7c06173750f3d16b1d8ade1950727481",
"lucene/analysis/common/src/java/org/apache/lucene/analysis/email/UAX29URLEmailTokenizerImpl.jflex": "56a751d27e481fb55388f91ebf34f5a0cb8cb1b2",
"lucene/core/src/data/jflex/UnicodeEmojiProperties.jflex": "7491dd535debc6e9e9ce367c4d3a7217e466dcae"
"lucene/core/src/data/jflex/UnicodeEmojiProperties.jflex": "704643e507c77a66142dd5b1c84ac4273559371a"
}

4
lucene/analysis/common/src/generated/checksums/generateUnicodeProps.json
View File

@ -1,4 +1,4 @@
{
"lucene/analysis/common/src/java/org/apache/lucene/analysis/util/UnicodeProps.java": "7d2cf5f959c2dfc5b83295e359212a1228f761c4",
"property:icuConfig": "icu_68"
"lucene/analysis/common/src/java/org/apache/lucene/analysis/util/UnicodeProps.java": "0c95238312bae4a7234e1aa539c09b20a8a4dc7f",
"property:icuConfig": "com.ibm.icu:icu4j:70.1"
}

7
lucene/analysis/common/src/java/org/apache/lucene/analysis/email/ASCIITLD.jflex
View File

@ -15,7 +15,7 @@
* limitations under the License.
*/
// Generated from IANA TLD Database <https://data.iana.org/TLD/tlds-alpha-by-domain.txt>
// file version from 2021 May 1, Sat 07:07:01 Coordinated Universal Time
// file version from 2021 Nov 17, Wed 07:07:01 Coordinated Universal Time
// generated by org.apache.lucene.analysis.standard.GenerateJflexTLDMacros
// LUCENE-8278: None of the TLDs in {ASCIITLD} is a 1-character-shorter prefix of another TLD
@ -767,6 +767,7 @@ ASCIITLD = "." (
| [mM][tT][rR]
| [mM][uU]
| [mM][uU][sS][eE][uU][mM]
| [mM][uU][sS][iI][cC]
| [mM][uU][tT][uU][aA][lL]
| [mM][vV]
| [mM][wW]
@ -906,7 +907,6 @@ ASCIITLD = "." (
| [qQ][pP][oO][nN]
| [qQ][uU][eE][bB][eE][cC]
| [qQ][uU][eE][sS][tT]
| [qQ][vV][cC]
| [rR][aA][cC][iI][nN][gG]
| [rR][aA][dD][iI][oO]
| [rR][aA][iI][dD]
@ -937,7 +937,6 @@ ASCIITLD = "." (
| [rR][iI][lL]
| [rR][iI][oO]
| [rR][iI][pP]
| [rR][mM][iI][tT]
| [rR][oO]
| [rR][oO][cC][hH][eE][rR]
| [rR][oO][cC][kK][sS]
@ -1064,7 +1063,6 @@ ASCIITLD = "." (
| [sS][uU][zZ][uU][kK][iI]
| [sS][vV]
| [sS][wW][aA][tT][cC][hH]
| [sS][wW][iI][fF][tT][cC][oO][vV][eE][rR]
| [sS][wW][iI][sS][sS]
| [sS][xX]
| [sS][yY]
@ -1231,7 +1229,6 @@ ASCIITLD = "." (
| [xX][nN]--3[dD][sS]443[gG]
| [xX][nN]--3[eE]0[bB]707[eE]
| [xX][nN]--3[hH][cC][rR][jJ]9[cC]
| [xX][nN]--3[oO][qQ]18[vV][lL]8[pP][nN]36[aA]
| [xX][nN]--3[pP][xX][uU]8[kK]
| [xX][nN]--42[cC]2[dD]9[aA]
| [xX][nN]--45[bB][rR]5[cC][yY][lL]

73393
lucene/analysis/common/src/java/org/apache/lucene/analysis/email/UAX29URLEmailTokenizerImpl.java
View File

File diff suppressed because it is too large Load Diff

4
lucene/analysis/common/src/java/org/apache/lucene/analysis/util/UnicodeProps.java
View File

@ -24,13 +24,13 @@ import org.apache.lucene.util.SparseFixedBitSet;
/**
* This file contains unicode properties used by various {@link CharTokenizer}s. The data was
* generated using ICU4J v68.2.0.0, unicode version: 13.0.0.0.
* generated using ICU4J v70.1.0.0, unicode version: 14.0.0.0.
*/
public final class UnicodeProps {
private UnicodeProps() {}
/** Unicode version that was used to generate this file: {@value} */
public static final String UNICODE_VERSION = "13.0.0.0";
public static final String UNICODE_VERSION = "14.0.0.0";
/** Bitset with Unicode WHITESPACE code points. */
public static final Bits WHITESPACE =

5
lucene/analysis/common/src/test/org/apache/lucene/analysis/email/TLDs.txt
View File

@ -820,6 +820,7 @@ mtn
mtr
mu
museum
music
mutual
mv
mw
@ -975,7 +976,6 @@ qa
qpon
quebec
quest
qvc
racing
radio
raid
@ -1010,7 +1010,6 @@ ricoh
ril
rio
rip
rmit
ro
rocher
rocks
@ -1149,7 +1148,6 @@ surgery
suzuki
sv
swatch
swiftcover
swiss
sx
sy
@ -1328,7 +1326,6 @@ xn--3bst00m
xn--3ds443g
xn--3e0b707e
xn--3hcrj9c
xn--3oq18vl8pn36a
xn--3pxu8k
xn--42c2d9a
xn--45br5cyl

1
lucene/analysis/icu/src/data/utr30/BasicFoldings.txt
View File

@ -50,6 +50,7 @@
2E1A>002D
2E3A..2E3B>002D
2E40>002D
2E5D>002D
301C>002D
3030>002D
30A0>002D

24
lucene/analysis/icu/src/data/utr30/DiacriticFolding.txt
View File

@ -62,6 +62,8 @@
07A6..07B0>
07EB..07F5>
0818..0819>
0898..089F>
08C9..08D2>
08E3..08FE>
093C>
094D>
@ -78,6 +80,7 @@
0B4D>
0B55>
0BCD>
0C3C>
0C4D>
0CBC>
0CCD>
@ -104,12 +107,14 @@
108F>
109A..109B>
135D..135F>
1714..1715>
17C9..17D3>
17DD>
1939..193B>
1A75..1A7C>
1A7F>
1AB0..1ABD>
1AB0..1ABE>
1AC1..1ACB>
1B34>
1B44>
1B6B..1B73>
@ -122,8 +127,7 @@
1CF7..1CF9>
1D2C..1D6A>
1DC4..1DCF>
1DF5..1DF9>
1DFD..1DFF>
1DF5..1DFF>
1FBD>
1FBF..1FC1>
1FCD..1FCF>
@ -164,9 +168,15 @@ FF70>
FF9E..FF9F>
FFE3>
102E0>
10780..10785>
10787..107B0>
107B2..107BA>
10AE5..10AE6>
10D22..10D27>
10F46..10F50>
10F82..10F85>
11046>
11070>
110B9..110BA>
11133..11134>
11173>
@ -200,12 +210,18 @@ FFE3>
16B30..16B36>
16F8F..16F9F>
16FF0..16FF1>
1AFF0..1AFF3>
1AFF5..1AFFB>
1AFFD..1AFFE>
1CF00..1CF2D>
1CF30..1CF46>
1D167..1D169>
1D16D..1D172>
1D17B..1D182>
1D185..1D18B>
1D1AA..1D1AD>
1E130..1E136>
1E2AE>
1E2EC..1E2EF>
1E8D0..1E8D6>
1E944..1E946>
@ -604,7 +620,6 @@ A7FF>004D
0829>
082A..082D>
0900>0901
1714>
1734>
1DC0..1DC3>
1DD0..1DE6>
@ -613,7 +628,6 @@ A7FF>004D
A670..A672>
A802>
10A3F>
11046>
1D165..1D166>
1D242..1D244>

10
lucene/analysis/icu/src/data/utr30/NativeDigitFolding.txt
View File

@ -630,6 +630,16 @@ ABF9>0039 # MEETEI MAYEK DIGIT NINE
16A67>0037 # MRO DIGIT SEVEN
16A68>0038 # MRO DIGIT EIGHT
16A69>0039 # MRO DIGIT NINE
16AC0>0030 # TANGSA DIGIT ZERO
16AC1>0031 # TANGSA DIGIT ONE
16AC2>0032 # TANGSA DIGIT TWO
16AC3>0033 # TANGSA DIGIT THREE
16AC4>0034 # TANGSA DIGIT FOUR
16AC5>0035 # TANGSA DIGIT FIVE
16AC6>0036 # TANGSA DIGIT SIX
16AC7>0037 # TANGSA DIGIT SEVEN
16AC8>0038 # TANGSA DIGIT EIGHT
16AC9>0039 # TANGSA DIGIT NINE
16B50>0030 # PAHAWH HMONG DIGIT ZERO
16B51>0031 # PAHAWH HMONG DIGIT ONE
16B52>0032 # PAHAWH HMONG DIGIT TWO

23
lucene/analysis/icu/src/data/utr30/nfc.txt
View File

@ -9,7 +9,7 @@
#
# Complete data for Unicode NFC normalization.
* Unicode 13.0.0
* Unicode 14.0.0
# Canonical_Combining_Class (ccc) values
0300..0314:230
@ -132,7 +132,11 @@
0825..0827:230
0829..082D:230
0859..085B:220
08D3:220
0898:230
0899..089B:220
089C..089F:230
08CA..08CE:230
08CF..08D3:220
08D4..08E1:230
08E3:220
08E4..08E5:230
@ -164,6 +168,7 @@
0B3C:7
0B4D:9
0BCD:9
0C3C:7
0C4D:9
0C55:84
0C56:91
@ -195,7 +200,7 @@
1039..103A:9
108D:220
135D..135F:230
1714:9
1714..1715:9
1734:9
17D2:9
17DD:230
@ -213,6 +218,11 @@
1ABB..1ABC:230
1ABD:220
1ABF..1AC0:220
1AC1..1AC2:230
1AC3..1AC4:220
1AC5..1AC9:230
1ACA:220
1ACB..1ACE:230
1B34:7
1B44:9
1B6B:230
@ -245,6 +255,7 @@
1DF6:232
1DF7..1DF8:228
1DF9:220
1DFA:218
1DFB:230
1DFC:233
1DFD:220
@ -314,7 +325,12 @@ FE2E..FE2F:230
10F4B:220
10F4C:230
10F4D..10F50:220
10F82:230
10F83:220
10F84:230
10F85:220
11046:9
11070:9
1107F:9
110B9:9
110BA:7
@ -373,6 +389,7 @@ FE2E..FE2F:230
1E023..1E024:230
1E026..1E02A:230
1E130..1E136:230
1E2AE:230
1E2EC..1E2EF:230
1E8D0..1E8D6:220
1E944..1E949:230

61
lucene/analysis/icu/src/data/utr30/nfkc.txt
View File

@ -13,7 +13,7 @@
# to NFKC one-way mappings.
# Use this file as the second gennorm2 input file after nfc.txt.
* Unicode 13.0.0
* Unicode 14.0.0
00A0>0020
00A8>0020 0308
@ -1367,6 +1367,9 @@
A69C>044A
A69D>044C
A770>A76F
A7F2>0043
A7F3>0046
A7F4>0051
A7F8>0126
A7F9>0153
AB5C>A727
@ -2419,6 +2422,62 @@ FFEB>2192
FFEC>2193
FFED>25A0
FFEE>25CB
10781>02D0
10782>02D1
10783>00E6
10784>0299
10785>0253
10787>02A3
10788>AB66
10789>02A5
1078A>02A4
1078B>0256
1078C>0257
1078D>1D91
1078E>0258
1078F>025E
10790>02A9
10791>0264
10792>0262
10793>0260
10794>029B
10795>0127
10796>029C
10797>0267
10798>0284
10799>02AA
1079A>02AB
1079B>026C
1079C>1DF04
1079D>A78E
1079E>026E
1079F>1DF05
107A0>028E
107A1>1DF06
107A2>00F8
107A3>0276
107A4>0277
107A5>0071
107A6>027A
107A7>1DF08
107A8>027D
107A9>027E
107AA>0280
107AB>02A8
107AC>02A6
107AD>AB67
107AE>02A7
107AF>0288
107B0>2C71
107B2>028F
107B3>02A1
107B4>02A2
107B5>0298
107B6>01C0
107B7>01C1
107B8>01C2
107B9>1DF0A
107BA>1DF1E
1D400>0041
1D401>0042
1D402>0043

103
lucene/analysis/icu/src/data/utr30/nfkc_cf.txt
View File

@ -12,7 +12,7 @@
# and reformatted into syntax for the gennorm2 Normalizer2 data generator tool.
# Use this file as the third gennorm2 input file after nfc.txt and nfkc.txt.
* Unicode 13.0.0
* Unicode 14.0.0
0041>0061
0042>0062
@ -639,7 +639,7 @@
13FC>13F4
13FD>13F5
17B4..17B5>
180B..180E>
180B..180F>
1C80>0432
1C81>0434
1C82>043E
@ -1433,6 +1433,7 @@
2C2C>2C5C
2C2D>2C5D
2C2E>2C5E
2C2F>2C5F
2C60>2C61
2C62>026B
2C63>1D7D
@ -2454,12 +2455,19 @@ A7B8>A7B9
A7BA>A7BB
A7BC>A7BD
A7BE>A7BF
A7C0>A7C1
A7C2>A7C3
A7C4>A794
A7C5>0282
A7C6>1D8E
A7C7>A7C8
A7C9>A7CA
A7D0>A7D1
A7D6>A7D7
A7D8>A7D9
A7F2>0063
A7F3>0066
A7F4>0071
A7F5>A7F6
A7F8>0127
A7F9>0153
@ -3960,6 +3968,97 @@ FFF0..FFF8>
104D1>104F9
104D2>104FA
104D3>104FB
10570>10597
10571>10598
10572>10599
10573>1059A
10574>1059B
10575>1059C
10576>1059D
10577>1059E
10578>1059F
10579>105A0
1057A>105A1
1057C>105A3
1057D>105A4
1057E>105A5
1057F>105A6
10580>105A7
10581>105A8
10582>105A9
10583>105AA
10584>105AB
10585>105AC
10586>105AD
10587>105AE
10588>105AF
10589>105B0
1058A>105B1
1058C>105B3
1058D>105B4
1058E>105B5
1058F>105B6
10590>105B7
10591>105B8
10592>105B9
10594>105BB
10595>105BC
10781>02D0
10782>02D1
10783>00E6
10784>0299
10785>0253
10787>02A3
10788>AB66
10789>02A5
1078A>02A4
1078B>0256
1078C>0257
1078D>1D91
1078E>0258
1078F>025E
10790>02A9
10791>0264
10792>0262
10793>0260
10794>029B
10795>0127
10796>029C
10797>0267
10798>0284
10799>02AA
1079A>02AB
1079B>026C
1079C>1DF04
1079D>A78E
1079E>026E
1079F>1DF05
107A0>028E
107A1>1DF06
107A2>00F8
107A3>0276
107A4>0277
107A5>0071
107A6>027A
107A7>1DF08
107A8>027D
107A9>027E
107AA>0280
107AB>02A8
107AC>02A6
107AD>AB67
107AE>02A7
107AF>0288
107B0>2C71
107B2>028F
107B3>02A1
107B4>02A2
107B5>0298
107B6>01C0
107B7>01C1
107B8>01C2
107B9>1DF0A
107BA>1DF1E
10C80>10CC0
10C81>10CC1
10C82>10CC2

6
lucene/analysis/icu/src/generated/checksums/genRbbi.json
View File

@ -1,7 +1,7 @@
{
"lucene/analysis/icu/src/data/uax29/Default.rbbi": "71bfaee5e81ac272aff828d1e44d0612be1b8363",
"lucene/analysis/icu/src/data/uax29/MyanmarSyllable.rbbi": "4c6817658b454add5ec1f9ac8c0015ce8eb3b5f2",
"lucene/analysis/icu/src/resources/org/apache/lucene/analysis/icu/segmentation/Default.brk": "1b9013b7ef4ba32a851a330c58a8fa820b9dda79",
"lucene/analysis/icu/src/resources/org/apache/lucene/analysis/icu/segmentation/MyanmarSyllable.brk": "cc023ec17e0148518086098691785a32b88ee09a",
"property:icu4j": "68.2"
"lucene/analysis/icu/src/resources/org/apache/lucene/analysis/icu/segmentation/Default.brk": "50fa00b5e1f40da5dfe3f0b9e49ab5612dfd4a0e",
"lucene/analysis/icu/src/resources/org/apache/lucene/analysis/icu/segmentation/MyanmarSyllable.brk": "1034111656380fbeb08e80ad98983f575a6f07fd",
"property:icuConfig": "com.ibm.icu:icu4j:70.1"
}

16
lucene/analysis/icu/src/generated/checksums/genUtr30DataFiles.json
View File

@ -1,12 +1,12 @@
{
"lucene/analysis/icu/src/data/utr30/BasicFoldings.txt": "fc5badae307c5b740f8dd81407e4be04e3e2fde6",
"lucene/analysis/icu/src/data/utr30/DiacriticFolding.txt": "2d48c24758ed9322b49eee84c23c74b7f11337c6",
"lucene/analysis/icu/src/data/utr30/BasicFoldings.txt": "c8c10df0fb555092c206416a33e3f4d56174bd28",
"lucene/analysis/icu/src/data/utr30/DiacriticFolding.txt": "0cd67967ac94aa2bf134e2ec55c8c26b13dd7d23",
"lucene/analysis/icu/src/data/utr30/DingbatFolding.txt": "ca562b2029e208bdf07cb2e5f689e509a336265e",
"lucene/analysis/icu/src/data/utr30/HanRadicalFolding.txt": "f5848aa38cbf1a11131d59384f31477f3b7f077b",
"lucene/analysis/icu/src/data/utr30/NativeDigitFolding.txt": "434ad7eb17fc32a441385795e4a4aa7e589ec30a",
"lucene/analysis/icu/src/data/utr30/nfc.txt": "ec95d7f7e5910791717234dd09efc4b13dc32d35",
"lucene/analysis/icu/src/data/utr30/nfkc.txt": "b8e91bc64e354af505d51f6072a43c90c6b4d1b5",
"lucene/analysis/icu/src/data/utr30/nfkc_cf.txt": "22d90ea4a7771e6ddebaaeb9438e98ce625e16f5",
"lucene/analysis/icu/src/resources/org/apache/lucene/analysis/icu/utr30.nrm": "32a6a3198039883c93f9ebef31fe24c1029f2b07",
"property:icu4j": "68.2"
"lucene/analysis/icu/src/data/utr30/NativeDigitFolding.txt": "d01d190007e6bbd1bf97473a2bedbe65d64d0b31",
"lucene/analysis/icu/src/data/utr30/nfc.txt": "95dba9e4992693e851e5012ca17862e50600c2c1",
"lucene/analysis/icu/src/data/utr30/nfkc.txt": "960d000f7f0b20a01d6eae915649f8a5d75bc590",
"lucene/analysis/icu/src/data/utr30/nfkc_cf.txt": "bc71ce253a290fba8bb32bf55b543b49fabe4271",
"lucene/analysis/icu/src/resources/org/apache/lucene/analysis/icu/utr30.nrm": "188d75c1faf8f3212a772a93c809b7fe69d3e57f",
"property:icuConfig": "com.ibm.icu:icu4j:70.1"
}

BIN
lucene/analysis/icu/src/resources/org/apache/lucene/analysis/icu/segmentation/Default.brk
View File

Binary file not shown.

BIN
lucene/analysis/icu/src/resources/org/apache/lucene/analysis/icu/segmentation/MyanmarSyllable.brk
View File

Binary file not shown.

BIN
lucene/analysis/icu/src/resources/org/apache/lucene/analysis/icu/utr30.nrm
View File

Binary file not shown.

8
lucene/core/src/data/jflex/UnicodeEmojiProperties.jflex
View File

@ -1,6 +1,6 @@
// DO NOT EDIT THIS FILE! Use "gradlew generateEmojiProperties" to recreate.
// The data was generated using ICU4J v62.2.0.0, unicode version: 11.0.0.0.
// The data was generated using ICU4J v70.1.0.0, unicode version: 14.0.0.0.
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
@ -19,8 +19,8 @@
* limitations under the License.
*/
Emoji = [\u{23}\u{2A}\u{30}-\u{39}\u{A9}\u{AE}\u{203C}\u{2049}\u{2122}\u{2139}\u{2194}-\u{2199}\u{21A9}-\u{21AA}\u{231A}-\u{231B}\u{2328}\u{23CF}\u{23E9}-\u{23F3}\u{23F8}-\u{23FA}\u{24C2}\u{25AA}-\u{25AB}\u{25B6}\u{25C0}\u{25FB}-\u{25FE}\u{2600}-\u{2604}\u{260E}\u{2611}\u{2614}-\u{2615}\u{2618}\u{261D}\u{2620}\u{2622}-\u{2623}\u{2626}\u{262A}\u{262E}-\u{262F}\u{2638}-\u{263A}\u{2640}\u{2642}\u{2648}-\u{2653}\u{265F}-\u{2660}\u{2663}\u{2665}-\u{2666}\u{2668}\u{267B}\u{267E}-\u{267F}\u{2692}-\u{2697}\u{2699}\u{269B}-\u{269C}\u{26A0}-\u{26A1}\u{26AA}-\u{26AB}\u{26B0}-\u{26B1}\u{26BD}-\u{26BE}\u{26C4}-\u{26C5}\u{26C8}\u{26CE}-\u{26CF}\u{26D1}\u{26D3}-\u{26D4}\u{26E9}-\u{26EA}\u{26F0}-\u{26F5}\u{26F7}-\u{26FA}\u{26FD}\u{2702}\u{2705}\u{2708}-\u{270D}\u{270F}\u{2712}\u{2714}\u{2716}\u{271D}\u{2721}\u{2728}\u{2733}-\u{2734}\u{2744}\u{2747}\u{274C}\u{274E}\u{2753}-\u{2755}\u{2757}\u{2763}-\u{2764}\u{2795}-\u{2797}\u{27A1}\u{27B0}\u{27BF}\u{2934}-\u{2935}\u{2B05}-\u{2B07}\u{2B1B}-\u{2B1C}\u{2B50}\u{2B55}\u{3030}\u{303D}\u{3297}\u{3299}\u{1F004}\u{1F0CF}\u{1F170}-\u{1F171}\u{1F17E}-\u{1F17F}\u{1F18E}\u{1F191}-\u{1F19A}\u{1F1E6}-\u{1F1FF}\u{1F201}-\u{1F202}\u{1F21A}\u{1F22F}\u{1F232}-\u{1F23A}\u{1F250}-\u{1F251}\u{1F300}-\u{1F321}\u{1F324}-\u{1F393}\u{1F396}-\u{1F397}\u{1F399}-\u{1F39B}\u{1F39E}-\u{1F3F0}\u{1F3F3}-\u{1F3F5}\u{1F3F7}-\u{1F4FD}\u{1F4FF}-\u{1F53D}\u{1F549}-\u{1F54E}\u{1F550}-\u{1F567}\u{1F56F}-\u{1F570}\u{1F573}-\u{1F57A}\u{1F587}\u{1F58A}-\u{1F58D}\u{1F590}\u{1F595}-\u{1F596}\u{1F5A4}-\u{1F5A5}\u{1F5A8}\u{1F5B1}-\u{1F5B2}\u{1F5BC}\u{1F5C2}-\u{1F5C4}\u{1F5D1}-\u{1F5D3}\u{1F5DC}-\u{1F5DE}\u{1F5E1}\u{1F5E3}\u{1F5E8}\u{1F5EF}\u{1F5F3}\u{1F5FA}-\u{1F64F}\u{1F680}-\u{1F6C5}\u{1F6CB}-\u{1F6D2}\u{1F6E0}-\u{1F6E5}\u{1F6E9}\u{1F6EB}-\u{1F6EC}\u{1F6F0}\u{1F6F3}-\u{1F6F9}\u{1F910}-\u{1F93A}\u{1F93C}-\u{1F93E}\u{1F940}-\u{1F945}\u{1F947}-\u{1F970}\u{1F973}-\u{1F976}\u{1F97A}\u{1F97C}-\u{1F9A2}\u{1F9B0}-\u{1F9B9}\u{1F9C0}-\u{1F9C2}\u{1F9D0}-\u{1F9FF}]
Emoji = [\u{23}\u{2A}\u{30}-\u{39}\u{A9}\u{AE}\u{203C}\u{2049}\u{2122}\u{2139}\u{2194}-\u{2199}\u{21A9}-\u{21AA}\u{231A}-\u{231B}\u{2328}\u{23CF}\u{23E9}-\u{23F3}\u{23F8}-\u{23FA}\u{24C2}\u{25AA}-\u{25AB}\u{25B6}\u{25C0}\u{25FB}-\u{25FE}\u{2600}-\u{2604}\u{260E}\u{2611}\u{2614}-\u{2615}\u{2618}\u{261D}\u{2620}\u{2622}-\u{2623}\u{2626}\u{262A}\u{262E}-\u{262F}\u{2638}-\u{263A}\u{2640}\u{2642}\u{2648}-\u{2653}\u{265F}-\u{2660}\u{2663}\u{2665}-\u{2666}\u{2668}\u{267B}\u{267E}-\u{267F}\u{2692}-\u{2697}\u{2699}\u{269B}-\u{269C}\u{26A0}-\u{26A1}\u{26A7}\u{26AA}-\u{26AB}\u{26B0}-\u{26B1}\u{26BD}-\u{26BE}\u{26C4}-\u{26C5}\u{26C8}\u{26CE}-\u{26CF}\u{26D1}\u{26D3}-\u{26D4}\u{26E9}-\u{26EA}\u{26F0}-\u{26F5}\u{26F7}-\u{26FA}\u{26FD}\u{2702}\u{2705}\u{2708}-\u{270D}\u{270F}\u{2712}\u{2714}\u{2716}\u{271D}\u{2721}\u{2728}\u{2733}-\u{2734}\u{2744}\u{2747}\u{274C}\u{274E}\u{2753}-\u{2755}\u{2757}\u{2763}-\u{2764}\u{2795}-\u{2797}\u{27A1}\u{27B0}\u{27BF}\u{2934}-\u{2935}\u{2B05}-\u{2B07}\u{2B1B}-\u{2B1C}\u{2B50}\u{2B55}\u{3030}\u{303D}\u{3297}\u{3299}\u{1F004}\u{1F0CF}\u{1F170}-\u{1F171}\u{1F17E}-\u{1F17F}\u{1F18E}\u{1F191}-\u{1F19A}\u{1F1E6}-\u{1F1FF}\u{1F201}-\u{1F202}\u{1F21A}\u{1F22F}\u{1F232}-\u{1F23A}\u{1F250}-\u{1F251}\u{1F300}-\u{1F321}\u{1F324}-\u{1F393}\u{1F396}-\u{1F397}\u{1F399}-\u{1F39B}\u{1F39E}-\u{1F3F0}\u{1F3F3}-\u{1F3F5}\u{1F3F7}-\u{1F4FD}\u{1F4FF}-\u{1F53D}\u{1F549}-\u{1F54E}\u{1F550}-\u{1F567}\u{1F56F}-\u{1F570}\u{1F573}-\u{1F57A}\u{1F587}\u{1F58A}-\u{1F58D}\u{1F590}\u{1F595}-\u{1F596}\u{1F5A4}-\u{1F5A5}\u{1F5A8}\u{1F5B1}-\u{1F5B2}\u{1F5BC}\u{1F5C2}-\u{1F5C4}\u{1F5D1}-\u{1F5D3}\u{1F5DC}-\u{1F5DE}\u{1F5E1}\u{1F5E3}\u{1F5E8}\u{1F5EF}\u{1F5F3}\u{1F5FA}-\u{1F64F}\u{1F680}-\u{1F6C5}\u{1F6CB}-\u{1F6D2}\u{1F6D5}-\u{1F6D7}\u{1F6DD}-\u{1F6E5}\u{1F6E9}\u{1F6EB}-\u{1F6EC}\u{1F6F0}\u{1F6F3}-\u{1F6FC}\u{1F7E0}-\u{1F7EB}\u{1F7F0}\u{1F90C}-\u{1F93A}\u{1F93C}-\u{1F945}\u{1F947}-\u{1F9FF}\u{1FA70}-\u{1FA74}\u{1FA78}-\u{1FA7C}\u{1FA80}-\u{1FA86}\u{1FA90}-\u{1FAAC}\u{1FAB0}-\u{1FABA}\u{1FAC0}-\u{1FAC5}\u{1FAD0}-\u{1FAD9}\u{1FAE0}-\u{1FAE7}\u{1FAF0}-\u{1FAF6}]
Emoji_Modifier = [\u{1F3FB}-\u{1F3FF}]
Emoji_Modifier_Base = [\u{261D}\u{26F9}\u{270A}-\u{270D}\u{1F385}\u{1F3C2}-\u{1F3C4}\u{1F3C7}\u{1F3CA}-\u{1F3CC}\u{1F442}-\u{1F443}\u{1F446}-\u{1F450}\u{1F466}-\u{1F469}\u{1F46E}\u{1F470}-\u{1F478}\u{1F47C}\u{1F481}-\u{1F483}\u{1F485}-\u{1F487}\u{1F4AA}\u{1F574}-\u{1F575}\u{1F57A}\u{1F590}\u{1F595}-\u{1F596}\u{1F645}-\u{1F647}\u{1F64B}-\u{1F64F}\u{1F6A3}\u{1F6B4}-\u{1F6B6}\u{1F6C0}\u{1F6CC}\u{1F918}-\u{1F91C}\u{1F91E}-\u{1F91F}\u{1F926}\u{1F930}-\u{1F939}\u{1F93D}-\u{1F93E}\u{1F9B5}-\u{1F9B6}\u{1F9B8}-\u{1F9B9}\u{1F9D1}-\u{1F9DD}]
Extended_Pictographic = [\u{A9}\u{AE}\u{203C}\u{2049}\u{2122}\u{2139}\u{2194}-\u{2199}\u{21A9}-\u{21AA}\u{231A}-\u{231B}\u{2328}\u{2388}\u{23CF}\u{23E9}-\u{23F3}\u{23F8}-\u{23FA}\u{24C2}\u{25AA}-\u{25AB}\u{25B6}\u{25C0}\u{25FB}-\u{25FE}\u{2600}-\u{2605}\u{2607}-\u{2612}\u{2614}-\u{2685}\u{2690}-\u{2705}\u{2708}-\u{2712}\u{2714}\u{2716}\u{271D}\u{2721}\u{2728}\u{2733}-\u{2734}\u{2744}\u{2747}\u{274C}\u{274E}\u{2753}-\u{2755}\u{2757}\u{2763}-\u{2767}\u{2795}-\u{2797}\u{27A1}\u{27B0}\u{27BF}\u{2934}-\u{2935}\u{2B05}-\u{2B07}\u{2B1B}-\u{2B1C}\u{2B50}\u{2B55}\u{3030}\u{303D}\u{3297}\u{3299}\u{1F000}-\u{1F0FF}\u{1F10D}-\u{1F10F}\u{1F12F}\u{1F16C}-\u{1F171}\u{1F17E}-\u{1F17F}\u{1F18E}\u{1F191}-\u{1F19A}\u{1F1AD}-\u{1F1E5}\u{1F201}-\u{1F20F}\u{1F21A}\u{1F22F}\u{1F232}-\u{1F23A}\u{1F23C}-\u{1F23F}\u{1F249}-\u{1F3FA}\u{1F400}-\u{1F53D}\u{1F546}-\u{1F64F}\u{1F680}-\u{1F6FF}\u{1F774}-\u{1F77F}\u{1F7D5}-\u{1F7FF}\u{1F80C}-\u{1F80F}\u{1F848}-\u{1F84F}\u{1F85A}-\u{1F85F}\u{1F888}-\u{1F88F}\u{1F8AE}-\u{1F8FF}\u{1F90C}-\u{1F93A}\u{1F93C}-\u{1F945}\u{1F947}-\u{1FFFD}]
Emoji_Modifier_Base = [\u{261D}\u{26F9}\u{270A}-\u{270D}\u{1F385}\u{1F3C2}-\u{1F3C4}\u{1F3C7}\u{1F3CA}-\u{1F3CC}\u{1F442}-\u{1F443}\u{1F446}-\u{1F450}\u{1F466}-\u{1F478}\u{1F47C}\u{1F481}-\u{1F483}\u{1F485}-\u{1F487}\u{1F48F}\u{1F491}\u{1F4AA}\u{1F574}-\u{1F575}\u{1F57A}\u{1F590}\u{1F595}-\u{1F596}\u{1F645}-\u{1F647}\u{1F64B}-\u{1F64F}\u{1F6A3}\u{1F6B4}-\u{1F6B6}\u{1F6C0}\u{1F6CC}\u{1F90C}\u{1F90F}\u{1F918}-\u{1F91F}\u{1F926}\u{1F930}-\u{1F939}\u{1F93C}-\u{1F93E}\u{1F977}\u{1F9B5}-\u{1F9B6}\u{1F9B8}-\u{1F9B9}\u{1F9BB}\u{1F9CD}-\u{1F9CF}\u{1F9D1}-\u{1F9DD}\u{1FAC3}-\u{1FAC5}\u{1FAF0}-\u{1FAF6}]
Extended_Pictographic = [\u{A9}\u{AE}\u{203C}\u{2049}\u{2122}\u{2139}\u{2194}-\u{2199}\u{21A9}-\u{21AA}\u{231A}-\u{231B}\u{2328}\u{2388}\u{23CF}\u{23E9}-\u{23F3}\u{23F8}-\u{23FA}\u{24C2}\u{25AA}-\u{25AB}\u{25B6}\u{25C0}\u{25FB}-\u{25FE}\u{2600}-\u{2605}\u{2607}-\u{2612}\u{2614}-\u{2685}\u{2690}-\u{2705}\u{2708}-\u{2712}\u{2714}\u{2716}\u{271D}\u{2721}\u{2728}\u{2733}-\u{2734}\u{2744}\u{2747}\u{274C}\u{274E}\u{2753}-\u{2755}\u{2757}\u{2763}-\u{2767}\u{2795}-\u{2797}\u{27A1}\u{27B0}\u{27BF}\u{2934}-\u{2935}\u{2B05}-\u{2B07}\u{2B1B}-\u{2B1C}\u{2B50}\u{2B55}\u{3030}\u{303D}\u{3297}\u{3299}\u{1F000}-\u{1F0FF}\u{1F10D}-\u{1F10F}\u{1F12F}\u{1F16C}-\u{1F171}\u{1F17E}-\u{1F17F}\u{1F18E}\u{1F191}-\u{1F19A}\u{1F1AD}-\u{1F1E5}\u{1F201}-\u{1F20F}\u{1F21A}\u{1F22F}\u{1F232}-\u{1F23A}\u{1F23C}-\u{1F23F}\u{1F249}-\u{1F3FA}\u{1F400}-\u{1F53D}\u{1F546}-\u{1F64F}\u{1F680}-\u{1F6FF}\u{1F774}-\u{1F77F}\u{1F7D5}-\u{1F7FF}\u{1F80C}-\u{1F80F}\u{1F848}-\u{1F84F}\u{1F85A}-\u{1F85F}\u{1F888}-\u{1F88F}\u{1F8AE}-\u{1F8FF}\u{1F90C}-\u{1F93A}\u{1F93C}-\u{1F945}\u{1F947}-\u{1FAFF}\u{1FC00}-\u{1FFFD}]

4
lucene/core/src/generated/checksums/generateEmojiProperties.json
View File

@ -1,4 +1,4 @@
{
"lucene/core/src/data/jflex/UnicodeEmojiProperties.jflex": "7491dd535debc6e9e9ce367c4d3a7217e466dcae",
"property:icuConfig": "icu_62"
"lucene/core/src/data/jflex/UnicodeEmojiProperties.jflex": "704643e507c77a66142dd5b1c84ac4273559371a",
"property:icuConfig": "com.ibm.icu:icu4j:70.1"
}

4
lucene/core/src/generated/checksums/generateStandardTokenizer.json
View File

@ -1,6 +1,6 @@
{
"gradle/generation/jflex/skeleton.disable.buffer.expansion.txt": "1424f4df33c977bb150d7377c3bd61f819113091",
"lucene/core/src/data/jflex/UnicodeEmojiProperties.jflex": "7491dd535debc6e9e9ce367c4d3a7217e466dcae",
"lucene/core/src/java/org/apache/lucene/analysis/standard/StandardTokenizerImpl.java": "47efe8b0d39974f41707b415668136d4c0cf425b",
"lucene/core/src/data/jflex/UnicodeEmojiProperties.jflex": "704643e507c77a66142dd5b1c84ac4273559371a",
"lucene/core/src/java/org/apache/lucene/analysis/standard/StandardTokenizerImpl.java": "dd169c562d18ab106f4a056ca8860c62b0e53475",
"lucene/core/src/java/org/apache/lucene/analysis/standard/StandardTokenizerImpl.jflex": "6158aeb8dd11cd9100623608b2dcce51b2df9d0b"
}

228
lucene/core/src/java/org/apache/lucene/analysis/standard/StandardTokenizerImpl.java
View File

@ -158,119 +158,121 @@ public final class StandardTokenizerImpl {
"\10\4\3\4\5\4\1\4\1\4\1\4\5\4\14\4\13\4\2\4"+
"\2\4\1\4\1\4\2\4\1\4\1\4\22\4\1\4\2\4\2\4"+
"\6\4\12\0\2\4\6\4\1\4\1\4\1\4\2\4\3\4\2\4"+
"\10\4\2\4\4\4\2\4\13\4\2\4\5\4\2\4\2\4\1\4"+
"\5\4\2\4\1\4\1\4\1\4\2\4\24\4\2\4\5\4\6\4"+
"\1\4\2\4\1\3\1\4\2\4\1\4\4\4\1\4\2\4\1\4"+
"\2\0\2\4\4\3\1\4\1\4\2\4\1\4\1\0\1\4\1\0"+
"\1\4\6\0\1\4\3\0\1\4\6\0\1\4\12\0\2\4\17\0"+
"\1\4\2\0\1\4\4\0\1\4\1\0\1\4\4\0\3\4\1\0"+
"\1\4\13\0\2\4\3\4\55\0\3\4\11\0\1\4\16\0\1\4"+
"\16\0\1\4\u0174\0\2\4\u01cf\0\3\4\23\0\2\4\63\0\1\4"+
"\4\0\1\4\252\0\57\15\1\0\57\15\1\0\205\15\6\0\4\15"+
"\3\5\2\15\14\0\46\15\1\0\1\15\5\0\1\15\2\0\70\15"+
"\7\0\1\15\17\0\1\5\27\15\11\0\7\15\1\0\7\15\1\0"+
"\5\4\1\4\2\4\2\4\4\4\2\4\13\4\2\4\5\4\2\4"+
"\2\4\1\4\5\4\2\4\1\4\1\4\1\4\2\4\24\4\2\4"+
"\5\4\6\4\1\4\2\4\1\3\1\4\2\4\1\4\4\4\1\4"+
"\2\4\1\4\2\0\2\4\4\3\1\4\1\4\2\4\1\4\1\0"+
"\1\4\1\0\1\4\6\0\1\4\3\0\1\4\6\0\1\4\12\0"+
"\2\4\17\0\1\4\2\0\1\4\4\0\1\4\1\0\1\4\4\0"+
"\3\4\1\0\1\4\13\0\2\4\3\4\55\0\3\4\11\0\1\4"+
"\16\0\1\4\16\0\1\4\u0174\0\2\4\u01cf\0\3\4\23\0\2\4"+
"\63\0\1\4\4\0\1\4\252\0\57\15\1\0\57\15\1\0\205\15"+
"\6\0\4\15\3\5\2\15\14\0\46\15\1\0\1\15\5\0\1\15"+
"\2\0\70\15\7\0\1\15\17\0\1\5\27\15\11\0\7\15\1\0"+
"\7\15\1\0\7\15\1\0\7\15\1\0\7\15\1\0\7\15\1\0"+
"\7\15\1\0\40\5\57\0\1\15\120\0\32\27\1\0\131\27\14\0"+
"\326\27\57\0\1\15\1\0\1\27\31\0\11\27\6\5\1\4\5\22"+
"\2\0\3\27\1\15\1\15\1\4\3\0\126\30\2\0\2\5\2\22"+
"\3\30\133\22\1\0\4\22\5\0\51\15\3\0\136\17\21\0\33\15"+
"\65\0\20\22\227\0\1\4\1\0\1\4\66\0\57\22\1\0\130\22"+
"\250\0\u19b6\27\112\0\u51d6\27\52\0\u048d\15\103\0\56\15\2\0\u010d\15"+
"\3\0\20\15\12\20\2\15\24\0\57\15\4\5\1\0\12\5\1\0"+
"\37\15\2\5\120\15\2\5\45\0\11\15\2\0\147\15\2\0\44\15"+
"\1\0\10\15\77\0\13\15\1\5\3\15\1\5\4\15\1\5\27\15"+
"\5\5\30\0\64\15\14\0\2\5\62\15\22\5\12\0\12\20\6\0"+
"\22\5\6\15\3\0\1\15\1\0\1\15\2\0\12\20\34\15\10\5"+
"\2\0\27\15\15\5\14\0\35\17\3\0\4\5\57\15\16\5\16\0"+
"\1\15\12\20\6\0\5\34\1\35\12\34\12\20\5\34\1\0\51\15"+
"\16\5\11\0\3\15\1\5\10\15\2\5\2\0\12\20\6\0\33\34"+
"\3\35\62\34\1\35\1\34\3\35\2\34\2\35\5\34\2\35\1\34"+
"\1\35\1\34\30\0\5\34\13\15\5\5\2\0\3\15\2\5\12\0"+
"\6\15\2\0\6\15\2\0\6\15\11\0\7\15\1\0\7\15\1\0"+
"\53\15\1\0\12\15\12\0\163\15\10\5\1\0\2\5\2\0\12\20"+
"\6\0\u2ba4\17\14\0\27\17\4\0\61\17\u2104\0\u016e\27\2\0\152\27"+
"\46\0\7\15\14\0\5\15\5\0\1\33\1\5\12\33\1\0\15\33"+
"\1\0\5\33\1\0\1\33\1\0\2\33\1\0\2\33\1\0\12\33"+
"\142\15\41\0\u016b\15\22\0\100\15\2\0\66\15\50\0\14\15\4\0"+
"\16\5\1\6\1\11\1\24\2\0\1\23\1\24\13\0\20\5\3\0"+
"\2\26\30\0\3\26\1\24\1\0\1\25\1\0\1\24\1\23\32\0"+
"\5\15\1\0\207\15\2\0\1\5\7\0\1\25\4\0\1\24\1\0"+
"\1\25\1\0\12\20\1\23\1\24\5\0\32\15\4\0\1\26\1\0"+
"\32\15\13\0\70\22\2\5\37\17\3\0\6\17\2\0\6\17\2\0"+
"\6\17\2\0\3\17\34\0\3\5\4\0\14\15\1\0\32\15\1\0"+
"\23\15\1\0\2\15\1\0\17\15\2\0\16\15\42\0\173\15\105\0"+
"\65\15\210\0\1\5\202\0\35\15\3\0\61\15\17\0\1\5\37\0"+
"\40\15\20\0\33\15\5\0\46\15\5\5\5\0\36\15\2\0\44\15"+
"\4\0\10\15\1\0\5\15\52\0\236\15\2\0\12\20\6\0\44\15"+
"\4\0\44\15\4\0\50\15\10\0\64\15\234\0\u0137\15\11\0\26\15"+
"\12\0\10\15\230\0\6\15\2\0\1\15\1\0\54\15\1\0\2\15"+
"\3\0\1\15\2\0\27\15\12\0\27\15\11\0\37\15\101\0\23\15"+
"\1\0\2\15\12\0\26\15\12\0\32\15\106\0\70\15\6\0\2\15"+
"\100\0\1\15\3\5\1\0\2\5\5\0\4\5\4\15\1\0\3\15"+
"\1\0\33\15\4\0\3\5\4\0\1\5\40\0\35\15\3\0\35\15"+
"\43\0\10\15\1\0\34\15\2\5\31\0\66\15\12\0\26\15\12\0"+
"\23\15\15\0\22\15\156\0\111\15\67\0\63\15\15\0\63\15\u030d\0"+
"\3\5\65\15\17\5\37\0\12\20\17\0\4\5\55\15\13\5\2\0"+
"\1\5\22\0\31\15\7\0\12\20\6\0\3\5\44\15\16\5\1\0"+
"\12\20\20\0\43\15\1\5\2\0\1\15\11\0\3\5\60\15\16\5"+
"\4\15\5\0\3\5\3\0\12\20\1\15\1\0\1\15\43\0\22\15"+
"\1\0\31\15\14\5\6\0\1\5\101\0\7\15\1\0\1\15\1\0"+
"\4\15\1\0\17\15\1\0\12\15\7\0\57\15\14\5\5\0\12\20"+
"\6\0\4\5\1\0\10\15\2\0\2\15\2\0\26\15\1\0\7\15"+
"\1\0\2\15\1\0\5\15\2\0\1\5\1\15\7\5\2\0\2\5"+
"\2\0\3\5\2\0\1\15\6\0\1\5\5\0\5\15\2\5\2\0"+
"\7\5\3\0\5\5\213\0\65\15\22\5\4\15\5\0\12\20\46\0"+
"\60\15\24\5\2\15\1\0\1\15\10\0\12\20\246\0\57\15\7\5"+
"\2\0\11\5\27\0\4\15\2\5\42\0\60\15\21\5\3\0\1\15"+
"\13\0\12\20\46\0\53\15\15\5\10\0\12\20\66\0\32\34\3\0"+
"\17\35\4\0\12\20\2\34\3\0\1\34\u0160\0\100\15\12\20\25\0"+
"\1\15\u01c0\0\71\15\u0107\0\11\15\1\0\45\15\10\5\1\0\10\5"+
"\1\15\17\0\12\20\30\0\36\15\2\0\26\5\1\0\16\5\u0349\0"+
"\u039a\15\146\0\157\15\21\0\304\15\u0abc\0\u042f\15\u0fd1\0\u0247\15\u21b9\0"+
"\u0239\15\7\0\37\15\1\0\12\20\146\0\36\15\2\0\5\5\13\0"+
"\60\15\7\5\11\0\4\15\14\0\12\20\11\0\25\15\5\0\23\15"+
"\u0370\0\105\15\13\0\1\15\56\5\20\0\4\5\15\15\100\0\1\15"+
"\u401f\0\1\22\1\30\u0bfe\0\153\15\5\0\15\15\3\0\11\15\7\0"+
"\12\15\3\0\2\5\1\0\4\5\u14c1\0\5\5\3\0\26\5\2\0"+
"\7\5\36\0\4\5\224\0\3\5\u01bb\0\125\15\1\0\107\15\1\0"+
"\2\15\2\0\1\15\2\0\2\15\2\0\4\15\1\0\14\15\1\0"+
"\1\15\1\0\7\15\1\0\101\15\1\0\4\15\2\0\10\15\1\0"+
"\7\15\1\0\34\15\1\0\4\15\1\0\5\15\1\0\1\15\3\0"+
"\7\15\1\0\u0154\15\2\0\31\15\1\0\31\15\1\0\37\15\1\0"+
"\31\15\1\0\37\15\1\0\31\15\1\0\37\15\1\0\31\15\1\0"+
"\37\15\1\0\31\15\1\0\10\15\2\0\62\20\u0200\0\67\5\4\0"+
"\62\5\10\0\1\5\16\0\1\5\26\0\5\5\1\0\17\5\u0550\0"+
"\7\5\1\0\21\5\2\0\7\5\1\0\2\5\1\0\5\5\u07d5\0"+
"\305\15\13\0\7\5\51\0\104\15\7\5\5\0\12\20\u04a6\0\4\15"+
"\1\0\33\15\1\0\2\15\1\0\1\15\2\0\1\15\1\0\12\15"+
"\1\0\4\15\1\0\1\15\1\0\1\15\6\0\1\15\4\0\1\15"+
"\1\0\1\15\1\0\1\15\1\0\3\15\1\0\2\15\1\0\1\15"+
"\2\0\1\15\1\0\1\15\1\0\1\15\1\0\1\15\1\0\1\15"+
"\1\0\2\15\1\0\1\15\2\0\4\15\1\0\7\15\1\0\4\15"+
"\1\0\4\15\1\0\1\15\1\0\12\15\1\0\21\15\5\0\3\15"+
"\1\0\5\15\1\0\21\15\u0144\0\4\4\1\4\312\4\1\4\60\4"+
"\15\0\3\4\37\0\1\4\32\15\6\0\32\15\2\0\4\4\2\16"+
"\14\15\2\16\12\15\4\0\1\4\2\0\12\4\22\0\71\4\32\1"+
"\1\30\2\4\15\4\12\0\1\4\24\0\1\4\2\0\11\4\1\0"+
"\4\4\11\0\7\4\2\4\256\4\42\4\2\4\141\4\1\3\16\4"+
"\2\4\2\4\1\4\3\4\2\4\44\4\3\3\2\4\1\3\2\4"+
"\3\3\44\4\2\4\3\4\1\4\4\4\5\2\102\4\2\3\2\4"+
"\13\3\25\4\4\3\4\4\1\3\1\4\11\3\3\4\1\3\4\4"+
"\3\3\1\4\3\3\42\4\1\3\123\4\1\4\77\4\10\0\3\4"+
"\6\4\1\4\30\4\7\4\2\4\2\4\1\4\2\3\4\4\1\3"+
"\14\4\1\4\2\4\4\4\2\4\1\3\4\4\2\3\15\4\2\4"+
"\2\4\1\4\10\4\2\4\11\4\1\4\5\4\3\4\14\4\3\4"+
"\10\4\3\4\2\4\1\4\1\4\1\4\4\4\1\4\6\4\1\4"+
"\3\4\1\4\6\4\113\4\3\3\3\4\5\3\60\0\43\4\1\3"+
"\20\4\3\3\11\4\1\3\5\4\5\4\1\4\1\3\6\4\15\4"+
"\6\4\3\4\1\4\1\4\2\4\3\4\1\4\2\4\7\4\6\4"+
"\164\0\14\4\125\0\53\4\14\0\4\4\70\0\10\4\12\0\6\4"+
"\50\0\10\4\36\0\122\4\14\0\4\4\10\4\5\3\1\4\2\3"+
"\6\4\1\3\11\4\12\3\1\4\1\0\1\4\2\3\1\4\6\4"+
"\1\0\52\4\2\4\4\4\3\4\1\4\1\4\47\4\15\4\5\4"+
"\2\3\1\4\2\3\6\4\3\4\15\4\1\4\15\3\42\4\u05fe\4"+
"\2\0\ua6d7\27\51\0\u1035\27\13\0\336\27\2\0\u1682\27\u295e\0\u021e\27"+
"\uffff\0\uffff\0\uffff\0\uffff\0\uffff\0\uffff\0\uffff\0\uffff\0\uffff\0\uffff\0\uffff\0\u05ee\0"+
"\7\15\1\0\7\15\1\0\40\5\57\0\1\15\120\0\32\27\1\0"+
"\131\27\14\0\326\27\57\0\1\15\1\0\1\27\31\0\11\27\6\5"+
"\1\4\5\22\2\0\3\27\1\15\1\15\1\4\3\0\126\30\2\0"+
"\2\5\2\22\3\30\133\22\1\0\4\22\5\0\51\15\3\0\136\17"+
"\21\0\33\15\65\0\20\22\227\0\1\4\1\0\1\4\66\0\57\22"+
"\1\0\130\22\250\0\u19b6\27\112\0\u51d6\27\52\0\u048d\15\103\0\56\15"+
"\2\0\u010d\15\3\0\20\15\12\20\2\15\24\0\57\15\4\5\1\0"+
"\12\5\1\0\37\15\2\5\120\15\2\5\45\0\11\15\2\0\147\15"+
"\2\0\44\15\1\0\10\15\77\0\13\15\1\5\3\15\1\5\4\15"+
"\1\5\27\15\5\5\30\0\64\15\14\0\2\5\62\15\22\5\12\0"+
"\12\20\6\0\22\5\6\15\3\0\1\15\1\0\1\15\2\0\12\20"+
"\34\15\10\5\2\0\27\15\15\5\14\0\35\17\3\0\4\5\57\15"+
"\16\5\16\0\1\15\12\20\6\0\5\34\1\35\12\34\12\20\5\34"+
"\1\0\51\15\16\5\11\0\3\15\1\5\10\15\2\5\2\0\12\20"+
"\6\0\33\34\3\35\62\34\1\35\1\34\3\35\2\34\2\35\5\34"+
"\2\35\1\34\1\35\1\34\30\0\5\34\13\15\5\5\2\0\3\15"+
"\2\5\12\0\6\15\2\0\6\15\2\0\6\15\11\0\7\15\1\0"+
"\7\15\1\0\53\15\1\0\12\15\12\0\163\15\10\5\1\0\2\5"+
"\2\0\12\20\6\0\u2ba4\17\14\0\27\17\4\0\61\17\u2104\0\u016e\27"+
"\2\0\152\27\46\0\7\15\14\0\5\15\5\0\1\33\1\5\12\33"+
"\1\0\15\33\1\0\5\33\1\0\1\33\1\0\2\33\1\0\2\33"+
"\1\0\12\33\142\15\41\0\u016b\15\22\0\100\15\2\0\66\15\50\0"+
"\14\15\4\0\16\5\1\6\1\11\1\24\2\0\1\23\1\24\13\0"+
"\20\5\3\0\2\26\30\0\3\26\1\24\1\0\1\25\1\0\1\24"+
"\1\23\32\0\5\15\1\0\207\15\2\0\1\5\7\0\1\25\4\0"+
"\1\24\1\0\1\25\1\0\12\20\1\23\1\24\5\0\32\15\4\0"+
"\1\26\1\0\32\15\13\0\70\22\2\5\37\17\3\0\6\17\2\0"+
"\6\17\2\0\6\17\2\0\3\17\34\0\3\5\4\0\14\15\1\0"+
"\32\15\1\0\23\15\1\0\2\15\1\0\17\15\2\0\16\15\42\0"+
"\173\15\105\0\65\15\210\0\1\5\202\0\35\15\3\0\61\15\17\0"+
"\1\5\37\0\40\15\20\0\33\15\5\0\46\15\5\5\5\0\36\15"+
"\2\0\44\15\4\0\10\15\1\0\5\15\52\0\236\15\2\0\12\20"+
"\6\0\44\15\4\0\44\15\4\0\50\15\10\0\64\15\234\0\u0137\15"+
"\11\0\26\15\12\0\10\15\230\0\6\15\2\0\1\15\1\0\54\15"+
"\1\0\2\15\3\0\1\15\2\0\27\15\12\0\27\15\11\0\37\15"+
"\101\0\23\15\1\0\2\15\12\0\26\15\12\0\32\15\106\0\70\15"+
"\6\0\2\15\100\0\1\15\3\5\1\0\2\5\5\0\4\5\4\15"+
"\1\0\3\15\1\0\33\15\4\0\3\5\4\0\1\5\40\0\35\15"+
"\3\0\35\15\43\0\10\15\1\0\34\15\2\5\31\0\66\15\12\0"+
"\26\15\12\0\23\15\15\0\22\15\156\0\111\15\67\0\63\15\15\0"+
"\63\15\u030d\0\3\5\65\15\17\5\37\0\12\20\17\0\4\5\55\15"+
"\13\5\2\0\1\5\22\0\31\15\7\0\12\20\6\0\3\5\44\15"+
"\16\5\1\0\12\20\20\0\43\15\1\5\2\0\1\15\11\0\3\5"+
"\60\15\16\5\4\15\5\0\3\5\3\0\12\20\1\15\1\0\1\15"+
"\43\0\22\15\1\0\31\15\14\5\6\0\1\5\101\0\7\15\1\0"+
"\1\15\1\0\4\15\1\0\17\15\1\0\12\15\7\0\57\15\14\5"+
"\5\0\12\20\6\0\4\5\1\0\10\15\2\0\2\15\2\0\26\15"+
"\1\0\7\15\1\0\2\15\1\0\5\15\2\0\1\5\1\15\7\5"+
"\2\0\2\5\2\0\3\5\2\0\1\15\6\0\1\5\5\0\5\15"+
"\2\5\2\0\7\5\3\0\5\5\213\0\65\15\22\5\4\15\5\0"+
"\12\20\46\0\60\15\24\5\2\15\1\0\1\15\10\0\12\20\246\0"+
"\57\15\7\5\2\0\11\5\27\0\4\15\2\5\42\0\60\15\21\5"+
"\3\0\1\15\13\0\12\20\46\0\53\15\15\5\10\0\12\20\66\0"+
"\32\34\3\0\17\35\4\0\12\20\2\34\3\0\1\34\u0160\0\100\15"+
"\12\20\25\0\1\15\u01c0\0\71\15\u0107\0\11\15\1\0\45\15\10\5"+
"\1\0\10\5\1\15\17\0\12\20\30\0\36\15\2\0\26\5\1\0"+
"\16\5\u0349\0\u039a\15\146\0\157\15\21\0\304\15\u0abc\0\u042f\15\u0fd1\0"+
"\u0247\15\u21b9\0\u0239\15\7\0\37\15\1\0\12\20\146\0\36\15\2\0"+
"\5\5\13\0\60\15\7\5\11\0\4\15\14\0\12\20\11\0\25\15"+
"\5\0\23\15\u0370\0\105\15\13\0\1\15\56\5\20\0\4\5\15\15"+
"\100\0\1\15\u401f\0\1\22\1\30\u0bfe\0\153\15\5\0\15\15\3\0"+
"\11\15\7\0\12\15\3\0\2\5\1\0\4\5\u14c1\0\5\5\3\0"+
"\26\5\2\0\7\5\36\0\4\5\224\0\3\5\u01bb\0\125\15\1\0"+
"\107\15\1\0\2\15\2\0\1\15\2\0\2\15\2\0\4\15\1\0"+
"\14\15\1\0\1\15\1\0\7\15\1\0\101\15\1\0\4\15\2\0"+
"\10\15\1\0\7\15\1\0\34\15\1\0\4\15\1\0\5\15\1\0"+
"\1\15\3\0\7\15\1\0\u0154\15\2\0\31\15\1\0\31\15\1\0"+
"\37\15\1\0\31\15\1\0\37\15\1\0\31\15\1\0\37\15\1\0"+
"\31\15\1\0\37\15\1\0\31\15\1\0\10\15\2\0\62\20\u0200\0"+
"\67\5\4\0\62\5\10\0\1\5\16\0\1\5\26\0\5\5\1\0"+
"\17\5\u0550\0\7\5\1\0\21\5\2\0\7\5\1\0\2\5\1\0"+
"\5\5\u07d5\0\305\15\13\0\7\5\51\0\104\15\7\5\5\0\12\20"+
"\u04a6\0\4\15\1\0\33\15\1\0\2\15\1\0\1\15\2\0\1\15"+
"\1\0\12\15\1\0\4\15\1\0\1\15\1\0\1\15\6\0\1\15"+
"\4\0\1\15\1\0\1\15\1\0\1\15\1\0\3\15\1\0\2\15"+
"\1\0\1\15\2\0\1\15\1\0\1\15\1\0\1\15\1\0\1\15"+
"\1\0\1\15\1\0\2\15\1\0\1\15\2\0\4\15\1\0\7\15"+
"\1\0\4\15\1\0\4\15\1\0\1\15\1\0\12\15\1\0\21\15"+
"\5\0\3\15\1\0\5\15\1\0\21\15\u0144\0\4\4\1\4\312\4"+
"\1\4\60\4\15\0\3\4\37\0\1\4\32\15\6\0\32\15\2\0"+
"\4\4\2\16\14\15\2\16\12\15\4\0\1\4\2\0\12\4\22\0"+
"\71\4\32\1\1\30\2\4\15\4\12\0\1\4\24\0\1\4\2\0"+
"\11\4\1\0\4\4\11\0\7\4\2\4\256\4\42\4\2\4\141\4"+
"\1\3\16\4\2\4\2\4\1\4\3\4\2\4\44\4\3\3\2\4"+
"\1\3\2\4\3\3\44\4\2\4\3\4\1\4\4\4\5\2\102\4"+
"\2\3\2\4\13\3\25\4\23\3\3\4\1\3\4\4\3\3\1\4"+
"\3\3\7\4\1\3\1\4\1\3\30\4\1\3\123\4\1\4\77\4"+
"\10\0\3\4\6\4\1\4\30\4\7\4\2\4\2\4\1\4\2\3"+
"\4\4\1\3\14\4\1\4\2\4\4\4\2\4\1\3\4\4\2\3"+
"\15\4\2\4\2\4\1\4\10\4\2\4\11\4\1\4\5\4\3\4"+
"\14\4\3\4\10\4\3\4\2\4\1\4\1\4\1\4\4\4\1\4"+
"\6\4\1\4\3\4\1\4\6\4\113\4\3\3\3\4\5\3\60\0"+
"\43\4\1\3\20\4\3\3\11\4\1\3\5\4\5\4\1\4\1\3"+
"\6\4\2\4\3\4\5\4\11\4\3\4\1\4\1\4\2\4\3\4"+
"\1\4\2\4\12\4\3\4\164\0\14\4\125\0\13\4\14\4\4\4"+
"\1\4\17\4\14\0\4\4\70\0\10\4\12\0\6\4\50\0\10\4"+
"\36\0\122\4\14\0\1\3\2\4\1\3\10\4\10\3\6\4\1\3"+
"\11\4\12\3\1\4\1\0\3\3\7\4\1\0\60\4\1\3\75\4"+
"\2\3\1\4\2\3\1\4\1\3\21\4\3\3\1\4\15\3\42\4"+
"\160\4\5\4\3\4\5\4\3\4\7\4\11\4\35\4\3\4\13\4"+
"\5\4\3\4\3\3\12\4\12\4\6\4\10\4\10\4\7\3\11\4"+
"\u0100\0\u03fe\4\2\0\ua6d7\27\51\0\u1035\27\13\0\336\27\2\0\u1682\27"+
"\u295e\0\u021e\27\uffff\0\uffff\0\uffff\0\uffff\0\uffff\0\uffff\0\uffff\0\uffff\0\uffff\0\uffff\0\uffff\0\u05ee\0"+
"\1\5\36\0\137\13\1\14\200\0\360\5\uffff\0\uffff\0\ufe12\0";
/**
@ -631,7 +633,7 @@ public final class StandardTokenizerImpl {
char [] map = new char[0x110000];
int i = 0; /* index in packed string */
int j = 0; /* index in unpacked array */
while (i < 4122) {
while (i < 4166) {
int count = packed.charAt(i++);
char value = packed.charAt(i++);
do map[j++] = value; while (--count > 0);

1
lucene/licenses/icu4j-68.2.jar.sha1
View File

@ -1 +0,0 @@
76893e6000401ace133a65262254be0ebe556d46

1
lucene/licenses/icu4j-70.1.jar.sha1 Normal file
View File

@ -0,0 +1 @@
dfa3a1fbc55bf5db8c6e79fc0935ac7ab1202950

2
versions.lock
View File

@ -1,7 +1,7 @@
# Run ./gradlew --write-locks to regenerate this file
com.carrotsearch:hppc:0.9.0 (1 constraints: 0b050636)
com.carrotsearch.randomizedtesting:randomizedtesting-runner:2.7.6 (1 constraints: 11051036)
com.ibm.icu:icu4j:68.2 (1 constraints: e4041f31)
com.ibm.icu:icu4j:70.1 (1 constraints: dc040a31)
commons-codec:commons-codec:1.13 (1 constraints: d904f430)
commons-logging:commons-logging:1.2 (2 constraints: c8149e7f)
io.sgr:s2-geometry-library-java:1.0.0 (1 constraints: 0305f035)

2
versions.props
View File

@ -4,7 +4,7 @@ com.github.ben-manes.caffeine:caffeine=2.8.4
com.google.errorprone:*=2.4.0
com.google.guava:guava=25.1-jre
com.google.protobuf:protobuf-java=3.11.0
com.ibm.icu:icu4j=68.2
com.ibm.icu:icu4j=70.1
commons-cli:commons-cli=1.4
commons-codec:commons-codec=1.13
commons-collections:commons-collections=3.2.2