deps: ICU 59.1 bump

* No feature changes. * Bug fixes. * Details: http://site.icu-project.org/download/59 Fixes: nodejs#12077 PR-URL: nodejs#12486 Refs: nodejs#7844 Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com> Reviewed-By: Refael Ackermann <refack@gmail.com>
srl295 · May 9, 2017 · 5d0a770 · 5d0a770
1 parent 147048a
commit 5d0a770
Show file tree

Hide file tree

Showing 899 changed files with 11,192 additions and 6,834 deletions.
diff --git a/LICENSE b/LICENSE
@@ -100,7 +100,7 @@ The externally maintained libraries used by Node.js are:
   """
     COPYRIGHT AND PERMISSION NOTICE (ICU 58 and later)
 
-    Copyright © 1991-2016 Unicode, Inc. All rights reserved.
+    Copyright © 1991-2017 Unicode, Inc. All rights reserved.
     Distributed under the Terms of Use in http://www.unicode.org/copyright.html
 
     Permission is hereby granted, free of charge, to any person obtaining

diff --git a/configure b/configure
@@ -1043,8 +1043,8 @@ def glob_to_var(dir_base, dir_sub, patch_dir):
 def configure_intl(o):
   icus = [
     {
-      'url': 'https://ssl.icu-project.org/files/icu4c/58.2/icu4c-58_2-src.zip',
-      'md5': 'f4fca37508fc5d14390501cf17aef084',
+      'url': 'https://ssl.icu-project.org/files/icu4c/59.1/icu4c-59_1-src.zip',
+      'md5': '29a41f9bb576b06c7eef0487a84a7674',
     },
   ]
   def icu_download(path):

diff --git a/deps/icu-small/LICENSE b/deps/icu-small/LICENSE
@@ -1,6 +1,6 @@
 COPYRIGHT AND PERMISSION NOTICE (ICU 58 and later)
 
-Copyright © 1991-2016 Unicode, Inc. All rights reserved.
+Copyright © 1991-2017 Unicode, Inc. All rights reserved.
 Distributed under the Terms of Use in http://www.unicode.org/copyright.html
 
 Permission is hereby granted, free of charge, to any person obtaining

diff --git a/deps/icu-small/README-SMALL-ICU.txt b/deps/icu-small/README-SMALL-ICU.txt
@@ -1,8 +1,8 @@
 Small ICU sources - auto generated by shrink-icu-src.py
 
 This directory contains the ICU subset used by --with-intl=small-icu (the default)
-It is a strict subset of ICU 58 source files with the following exception(s):
-* deps/icu-small/source/data/in/icudt58l.dat : Reduced-size data file
+It is a strict subset of ICU 59 source files with the following exception(s):
+* deps/icu-small/source/data/in/icudt59l.dat : Reduced-size data file
 
 
 To rebuild this directory, see ../../tools/icu/README.md
diff --git a/deps/icu-small/source/common/appendable.cpp b/deps/icu-small/source/common/appendable.cpp
@@ -1,12 +1,12 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 *******************************************************************************
 *   Copyright (C) 2011-2012, International Business Machines
 *   Corporation and others.  All Rights Reserved.
 *******************************************************************************
 *   file name:  appendable.cpp
-*   encoding:   US-ASCII
+*   encoding:   UTF-8
 *   tab size:   8 (not used)
 *   indentation:4
 *

diff --git a/deps/icu-small/source/common/bmpset.cpp b/deps/icu-small/source/common/bmpset.cpp
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 ******************************************************************************
@@ -8,7 +8,7 @@
 *
 ******************************************************************************
 *   file name:  bmpset.cpp
-*   encoding:   US-ASCII
+*   encoding:   UTF-8
 *   tab size:   8 (not used)
 *   indentation:4
 *

diff --git a/deps/icu-small/source/common/bmpset.h b/deps/icu-small/source/common/bmpset.h
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 ******************************************************************************
@@ -8,7 +8,7 @@
 *
 ******************************************************************************
 *   file name:  bmpset.h
-*   encoding:   US-ASCII
+*   encoding:   UTF-8
 *   tab size:   8 (not used)
 *   indentation:4
 *

diff --git a/deps/icu-small/source/common/brkeng.cpp b/deps/icu-small/source/common/brkeng.cpp
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
  ************************************************************************************

diff --git a/deps/icu-small/source/common/brkeng.h b/deps/icu-small/source/common/brkeng.h
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /**
  ************************************************************************************

diff --git a/deps/icu-small/source/common/brkiter.cpp b/deps/icu-small/source/common/brkiter.cpp
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 *******************************************************************************

diff --git a/deps/icu-small/source/common/bytestream.cpp b/deps/icu-small/source/common/bytestream.cpp
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 // Copyright (C) 2009-2011, International Business Machines
 // Corporation and others. All Rights Reserved.

diff --git a/deps/icu-small/source/common/bytestrie.cpp b/deps/icu-small/source/common/bytestrie.cpp
@@ -1,12 +1,12 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 *******************************************************************************
 *   Copyright (C) 2010-2011, International Business Machines
 *   Corporation and others.  All Rights Reserved.
 *******************************************************************************
 *   file name:  bytestrie.cpp
-*   encoding:   US-ASCII
+*   encoding:   UTF-8
 *   tab size:   8 (not used)
 *   indentation:4
 *

diff --git a/deps/icu-small/source/common/bytestriebuilder.cpp b/deps/icu-small/source/common/bytestriebuilder.cpp
@@ -1,12 +1,12 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 *******************************************************************************
 *   Copyright (C) 2010-2012, International Business Machines
 *   Corporation and others.  All Rights Reserved.
 *******************************************************************************
 *   file name:  bytestriebuilder.cpp
-*   encoding:   US-ASCII
+*   encoding:   UTF-8
 *   tab size:   8 (not used)
 *   indentation:4
 *

diff --git a/deps/icu-small/source/common/bytestrieiterator.cpp b/deps/icu-small/source/common/bytestrieiterator.cpp
@@ -1,12 +1,12 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 *******************************************************************************
 *   Copyright (C) 2010-2012, International Business Machines
 *   Corporation and others.  All Rights Reserved.
 *******************************************************************************
 *   file name:  bytestrieiterator.cpp
-*   encoding:   US-ASCII
+*   encoding:   UTF-8
 *   tab size:   8 (not used)
 *   indentation:4
 *

diff --git a/deps/icu-small/source/common/caniter.cpp b/deps/icu-small/source/common/caniter.cpp
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
  *****************************************************************************
@@ -311,12 +311,12 @@ void U_EXPORT2 CanonicalIterator::permute(UnicodeString &source, UBool skipZeros
 
         // see what the permutations of the characters before and after this one are
         //Hashtable *subpermute = permute(source.substring(0,i) + source.substring(i + UTF16.getCharCount(cp)));
-        permute(subPermuteString.replace(i, U16_LENGTH(cp), NULL, 0), skipZeros, &subpermute, status);
+        permute(subPermuteString.remove(i, U16_LENGTH(cp)), skipZeros, &subpermute, status);
         /* Test for buffer overflows */
         if(U_FAILURE(status)) {
             return;
         }
-        // The upper replace is destructive. The question is do we have to make a copy, or we don't care about the contents
+        // The upper remove is destructive. The question is do we have to make a copy, or we don't care about the contents
         // of source at this point.
 
         // prefix this character to all of them

diff --git a/deps/icu-small/source/common/chariter.cpp b/deps/icu-small/source/common/chariter.cpp
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 **********************************************************************

diff --git a/deps/icu-small/source/common/charstr.cpp b/deps/icu-small/source/common/charstr.cpp
@@ -1,12 +1,12 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 *******************************************************************************
 *   Copyright (C) 2010-2015, International Business Machines
 *   Corporation and others.  All Rights Reserved.
 *******************************************************************************
 *   file name:  charstr.cpp
-*   encoding:   US-ASCII
+*   encoding:   UTF-8
 *   tab size:   8 (not used)
 *   indentation:4
 *
@@ -15,6 +15,7 @@
 */
 
 #include "unicode/utypes.h"
+#include "unicode/putil.h"
 #include "charstr.h"
 #include "cmemory.h"
 #include "cstring.h"

diff --git a/deps/icu-small/source/common/charstr.h b/deps/icu-small/source/common/charstr.h
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 **********************************************************************

diff --git a/deps/icu-small/source/common/cmemory.c → deps/icu-small/source/common/cmemory.cpp b/deps/icu-small/source/common/cmemory.c → deps/icu-small/source/common/cmemory.cpp
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 ******************************************************************************

diff --git a/deps/icu-small/source/common/cmemory.h b/deps/icu-small/source/common/cmemory.h
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 ******************************************************************************

diff --git a/deps/icu-small/source/common/cpputils.h b/deps/icu-small/source/common/cpputils.h
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 ******************************************************************************
@@ -8,7 +8,7 @@
 *
 ******************************************************************************
 *   file name:  cpputils.h
-*   encoding:   US-ASCII
+*   encoding:   UTF-8
 *   tab size:   8 (not used)
 *   indentation:4
 */

diff --git a/deps/icu-small/source/common/cstr.cpp b/deps/icu-small/source/common/cstr.cpp
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 *******************************************************************************

diff --git a/deps/icu-small/source/common/cstr.h b/deps/icu-small/source/common/cstr.h
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 ******************************************************************************

diff --git a/deps/icu-small/source/common/cstring.c → deps/icu-small/source/common/cstring.cpp b/deps/icu-small/source/common/cstring.c → deps/icu-small/source/common/cstring.cpp
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 ******************************************************************************

diff --git a/deps/icu-small/source/common/cstring.h b/deps/icu-small/source/common/cstring.h
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 ******************************************************************************

diff --git a/deps/icu-small/source/common/cwchar.c → deps/icu-small/source/common/cwchar.cpp b/deps/icu-small/source/common/cwchar.c → deps/icu-small/source/common/cwchar.cpp
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 ******************************************************************************
@@ -8,7 +8,7 @@
 *
 ******************************************************************************
 *   file name:  cwchar.c
-*   encoding:   US-ASCII
+*   encoding:   UTF-8
 *   tab size:   8 (not used)
 *   indentation:4
 *

diff --git a/deps/icu-small/source/common/cwchar.h b/deps/icu-small/source/common/cwchar.h
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 ******************************************************************************
@@ -8,7 +8,7 @@
 *
 ******************************************************************************
 *   file name:  cwchar.h
-*   encoding:   US-ASCII
+*   encoding:   UTF-8
 *   tab size:   8 (not used)
 *   indentation:4
 *

diff --git a/deps/icu-small/source/common/dictbe.cpp b/deps/icu-small/source/common/dictbe.cpp
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /**
  *******************************************************************************
@@ -1385,12 +1385,25 @@ CjkBreakEngine::divideUpDictionaryRange( UText *inText,
     // Now that we're done, convert positions in t_boundary[] (indices in
     // the normalized input string) back to indices in the original input UText
     // while reversing t_boundary and pushing values to foundBreaks.
+    int32_t prevCPPos = -1;
+    int32_t prevUTextPos = -1;
     for (int32_t i = numBreaks-1; i >= 0; i--) {
         int32_t cpPos = t_boundary.elementAti(i);
+        U_ASSERT(cpPos > prevCPPos);
         int32_t utextPos =  inputMap.isValid() ? inputMap->elementAti(cpPos) : cpPos + rangeStart;
-        // Boundaries are added to foundBreaks output in ascending order.
-        U_ASSERT(foundBreaks.size() == 0 ||foundBreaks.peeki() < utextPos);
-        foundBreaks.push(utextPos, status);
+        U_ASSERT(utextPos >= prevUTextPos);
+        if (utextPos > prevUTextPos) {
+            // Boundaries are added to foundBreaks output in ascending order.
+            U_ASSERT(foundBreaks.size() == 0 || foundBreaks.peeki() < utextPos);
+            foundBreaks.push(utextPos, status);
+        } else {
+            // Normalization expanded the input text, the dictionary found a boundary
+            // within the expansion, giving two boundaries with the same index in the
+            // original text. Ignore the second. See ticket #12918.
+            --numBreaks;
+        }
+        prevCPPos = cpPos;
+        prevUTextPos = utextPos;
     }
 
     // inString goes out of scope

diff --git a/deps/icu-small/source/common/dictbe.h b/deps/icu-small/source/common/dictbe.h
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /**
  *******************************************************************************

diff --git a/deps/icu-small/source/common/dictionarydata.cpp b/deps/icu-small/source/common/dictionarydata.cpp
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 *******************************************************************************

diff --git a/deps/icu-small/source/common/dictionarydata.h b/deps/icu-small/source/common/dictionarydata.h
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*
 *******************************************************************************

diff --git a/deps/icu-small/source/common/dtintrv.cpp b/deps/icu-small/source/common/dtintrv.cpp
@@ -1,4 +1,4 @@
-// Copyright (C) 2016 and later: Unicode, Inc. and others.
+// © 2016 and later: Unicode, Inc. and others.
 // License & terms of use: http://www.unicode.org/copyright.html
 /*******************************************************************************
 * Copyright (C) 2008, International Business Machines Corporation and