opensource
/
fine-third
1
0
Fork 1
Code Issues Pull Requests Releases Wiki Activity
Browse Source

Merge pull request #10054 in CORE/base-third from bugfix/10.0 to feature/10.0 

* commit '3a3dc27c330aaee37fc91fa4e78894493ce240bf':
  REPORT-110010 删除lucene里的py腳本
feature/10.0
superman 1 year ago
parent
9d59790b85 3a3dc27c33
commit
04fa2327eb
7 changed files with 0 additions and 2367 deletions
Whitespace
Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL
Split View
Diff Options
Show Stats Download Patch File Download Diff File
  1. 539
      fine-lucene/src/main/java/com/fr/third/org/apache/lucene/analysis/charfilter/htmlentity.py
  2. 366
      fine-lucene/src/main/java/com/fr/third/org/apache/lucene/util/automaton/UTF32ToUTF8.py
  3. 500
      fine-lucene/src/main/java/com/fr/third/org/apache/lucene/util/automaton/createLevAutomata.py
  4. 335
      fine-lucene/src/main/java/com/fr/third/org/apache/lucene/util/packed/gen_BulkOperation.py
  5. 175
      fine-lucene/src/main/java/com/fr/third/org/apache/lucene/util/packed/gen_Direct.py
  6. 291
      fine-lucene/src/main/java/com/fr/third/org/apache/lucene/util/packed/gen_Packed64SingleBlock.py
  7. 161
      fine-lucene/src/main/java/com/fr/third/org/apache/lucene/util/packed/gen_PackedThreeBlocks.py

539
fine-lucene/src/main/java/com/fr/third/org/apache/lucene/analysis/charfilter/htmlentity.py
Unescape View File

@ -1,539 +0,0 @@
# Licensed to the Apache Software Foundation (ASF) under one or more
# contributor license agreements. See the NOTICE file distributed with
# this work for additional information regarding copyright ownership.
# The ASF licenses this file to You under the Apache License, Version 2.0
# (the "License"); you may not use this file except in compliance with
# the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import re
# A simple python script to generate an HTML entity map and a regex alternation
# for inclusion in HTMLStripCharFilter.jflex.
def main():
print get_apache_license()
codes = {}
regex = re.compile(r'\s*<!ENTITY\s+(\S+)\s+"&(?:#38;)?#(\d+);"')
for line in get_entity_text().split('\n'):
match = regex.match(line)
if match:
key = match.group(1)
if key == 'quot': codes[key] = r'\"'
elif key == 'nbsp': codes[key] = ' ';
else : codes[key] = r'\u%04X' % int(match.group(2))
keys = sorted(codes)
first_entry = True
output_line = 'CharacterEntities = ( '
for key in keys:
new_entry = ('"%s"' if first_entry else ' | "%s"') % key
first_entry = False
if len(output_line) + len(new_entry) >= 80:
print output_line
output_line = ' '
output_line += new_entry
if key in ('quot','copy','gt','lt','reg','amp'):
new_entry = ' | "%s"' % key.upper()
if len(output_line) + len(new_entry) >= 80:
print output_line
output_line = ' '
output_line += new_entry
print output_line, ')'
print '%{'
print ' private static final Map<String,String> upperCaseVariantsAccepted'
print ' = new HashMap<String,String>();'
print ' static {'
print ' upperCaseVariantsAccepted.put("quot", "QUOT");'
print ' upperCaseVariantsAccepted.put("copy", "COPY");'
print ' upperCaseVariantsAccepted.put("gt", "GT");'
print ' upperCaseVariantsAccepted.put("lt", "LT");'
print ' upperCaseVariantsAccepted.put("reg", "REG");'
print ' upperCaseVariantsAccepted.put("amp", "AMP");'
print ' }'
print ' private static final CharArrayMap<Character> entityValues'
print ' = new CharArrayMap<Character>(Version.LUCENE_40, %i, false);' % len(keys)
print ' static {'
print ' String[] entities = {'
output_line = ' '
for key in keys:
new_entry = ' "%s", "%s",' % (key, codes[key])
if len(output_line) + len(new_entry) >= 80:
print output_line
output_line = ' '
output_line += new_entry
print output_line[:-1]
print ' };'
print ' for (int i = 0 ; i < entities.length ; i += 2) {'
print ' Character value = entities[i + 1].charAt(0);'
print ' entityValues.put(entities[i], value);'
print ' String upperCaseVariant = upperCaseVariantsAccepted.get(entities[i]);'
print ' if (upperCaseVariant != null) {'
print ' entityValues.put(upperCaseVariant, value);'
print ' }'
print ' }'
print " }"
print "%}"
def get_entity_text():
# The text below is taken verbatim from
# <http://www.w3.org/TR/REC-html40/sgml/entities.html>:
text = r"""
F.1. XHTML Character Entities
XHTML DTDs make available a standard collection of named character entities. Those entities are defined in this section.
F.1.1. XHTML Latin 1 Character Entities
You can download this version of this file from http://www.w3.org/TR/2010/REC-xhtml-modularization/DTD/xhtml-lat1.ent. The latest version is available at http://www.w3.org/MarkUp/DTD/xhtml-lat1.ent.
<!-- ...................................................................... -->
<!-- XML-compatible ISO Latin 1 Character Entity Set for XHTML ............ -->
<!-- file: xhtml-lat1.ent
Typical invocation:
<!ENTITY % xhtml-lat1
PUBLIC "-//W3C//ENTITIES Latin 1 for XHTML//EN"
"xhtml-lat1.ent" >
%xhtml-lat1;
This DTD module is identified by the PUBLIC and SYSTEM identifiers:
PUBLIC "-//W3C//ENTITIES Latin 1 for XHTML//EN"
SYSTEM "http://www.w3.org/MarkUp/DTD/xhtml-lat1.ent"
Revision: $Id: xhtml-lat1.ent,v 4.1 2001/04/10 09:34:14 altheim Exp $ SMI
Portions (C) International Organization for Standardization 1986:
Permission to copy in any form is granted for use with conforming
SGML systems and applications as defined in ISO 8879, provided
this notice is included in all copies.
-->
<!ENTITY nbsp "&#160;" ><!-- no-break space = non-breaking space, U+00A0 ISOnum -->
<!ENTITY iexcl "&#161;" ><!-- inverted exclamation mark, U+00A1 ISOnum -->
<!ENTITY cent "&#162;" ><!-- cent sign, U+00A2 ISOnum -->
<!ENTITY pound "&#163;" ><!-- pound sign, U+00A3 ISOnum -->
<!ENTITY curren "&#164;" ><!-- currency sign, U+00A4 ISOnum -->
<!ENTITY yen "&#165;" ><!-- yen sign = yuan sign, U+00A5 ISOnum -->
<!ENTITY brvbar "&#166;" ><!-- broken bar = broken vertical bar, U+00A6 ISOnum -->
<!ENTITY sect "&#167;" ><!-- section sign, U+00A7 ISOnum -->
<!ENTITY uml "&#168;" ><!-- diaeresis = spacing diaeresis, U+00A8 ISOdia -->
<!ENTITY copy "&#169;" ><!-- copyright sign, U+00A9 ISOnum -->
<!ENTITY ordf "&#170;" ><!-- feminine ordinal indicator, U+00AA ISOnum -->
<!ENTITY laquo "&#171;" ><!-- left-pointing double angle quotation mark = left pointing guillemet, U+00AB ISOnum -->
<!ENTITY not "&#172;" ><!-- not sign, U+00AC ISOnum -->
<!ENTITY shy "&#173;" ><!-- soft hyphen = discretionary hyphen, U+00AD ISOnum -->
<!ENTITY reg "&#174;" ><!-- registered sign = registered trade mark sign, U+00AE ISOnum -->
<!ENTITY macr "&#175;" ><!-- macron = spacing macron = overline = APL overbar, U+00AF ISOdia -->
<!ENTITY deg "&#176;" ><!-- degree sign, U+00B0 ISOnum -->
<!ENTITY plusmn "&#177;" ><!-- plus-minus sign = plus-or-minus sign, U+00B1 ISOnum -->
<!ENTITY sup2 "&#178;" ><!-- superscript two = superscript digit two = squared, U+00B2 ISOnum -->
<!ENTITY sup3 "&#179;" ><!-- superscript three = superscript digit three = cubed, U+00B3 ISOnum -->
<!ENTITY acute "&#180;" ><!-- acute accent = spacing acute, U+00B4 ISOdia -->
<!ENTITY micro "&#181;" ><!-- micro sign, U+00B5 ISOnum -->
<!ENTITY para "&#182;" ><!-- pilcrow sign = paragraph sign, U+00B6 ISOnum -->
<!ENTITY middot "&#183;" ><!-- middle dot = Georgian comma = Greek middle dot, U+00B7 ISOnum -->
<!ENTITY cedil "&#184;" ><!-- cedilla = spacing cedilla, U+00B8 ISOdia -->
<!ENTITY sup1 "&#185;" ><!-- superscript one = superscript digit one, U+00B9 ISOnum -->
<!ENTITY ordm "&#186;" ><!-- masculine ordinal indicator, U+00BA ISOnum -->
<!ENTITY raquo "&#187;" ><!-- right-pointing double angle quotation mark = right pointing guillemet, U+00BB ISOnum -->
<!ENTITY frac14 "&#188;" ><!-- vulgar fraction one quarter = fraction one quarter, U+00BC ISOnum -->
<!ENTITY frac12 "&#189;" ><!-- vulgar fraction one half = fraction one half, U+00BD ISOnum -->
<!ENTITY frac34 "&#190;" ><!-- vulgar fraction three quarters = fraction three quarters, U+00BE ISOnum -->
<!ENTITY iquest "&#191;" ><!-- inverted question mark = turned question mark, U+00BF ISOnum -->
<!ENTITY Agrave "&#192;" ><!-- latin capital A with grave = latin capital A grave, U+00C0 ISOlat1 -->
<!ENTITY Aacute "&#193;" ><!-- latin capital A with acute, U+00C1 ISOlat1 -->
<!ENTITY Acirc "&#194;" ><!-- latin capital A with circumflex, U+00C2 ISOlat1 -->
<!ENTITY Atilde "&#195;" ><!-- latin capital A with tilde, U+00C3 ISOlat1 -->
<!ENTITY Auml "&#196;" ><!-- latin capital A with diaeresis, U+00C4 ISOlat1 -->
<!ENTITY Aring "&#197;" ><!-- latin capital A with ring above = latin capital A ring, U+00C5 ISOlat1 -->
<!ENTITY AElig "&#198;" ><!-- latin capital AE = latin capital ligature AE, U+00C6 ISOlat1 -->
<!ENTITY Ccedil "&#199;" ><!-- latin capital C with cedilla, U+00C7 ISOlat1 -->
<!ENTITY Egrave "&#200;" ><!-- latin capital E with grave, U+00C8 ISOlat1 -->
<!ENTITY Eacute "&#201;" ><!-- latin capital E with acute, U+00C9 ISOlat1 -->
<!ENTITY Ecirc "&#202;" ><!-- latin capital E with circumflex, U+00CA ISOlat1 -->
<!ENTITY Euml "&#203;" ><!-- latin capital E with diaeresis, U+00CB ISOlat1 -->
<!ENTITY Igrave "&#204;" ><!-- latin capital I with grave, U+00CC ISOlat1 -->
<!ENTITY Iacute "&#205;" ><!-- latin capital I with acute, U+00CD ISOlat1 -->
<!ENTITY Icirc "&#206;" ><!-- latin capital I with circumflex, U+00CE ISOlat1 -->
<!ENTITY Iuml "&#207;" ><!-- latin capital I with diaeresis, U+00CF ISOlat1 -->
<!ENTITY ETH "&#208;" ><!-- latin capital ETH, U+00D0 ISOlat1 -->
<!ENTITY Ntilde "&#209;" ><!-- latin capital N with tilde, U+00D1 ISOlat1 -->
<!ENTITY Ograve "&#210;" ><!-- latin capital O with grave, U+00D2 ISOlat1 -->
<!ENTITY Oacute "&#211;" ><!-- latin capital O with acute, U+00D3 ISOlat1 -->
<!ENTITY Ocirc "&#212;" ><!-- latin capital O with circumflex, U+00D4 ISOlat1 -->
<!ENTITY Otilde "&#213;" ><!-- latin capital O with tilde, U+00D5 ISOlat1 -->
<!ENTITY Ouml "&#214;" ><!-- latin capital O with diaeresis, U+00D6 ISOlat1 -->
<!ENTITY times "&#215;" ><!-- multiplication sign, U+00D7 ISOnum -->
<!ENTITY Oslash "&#216;" ><!-- latin capital O with stroke = latin capital O slash, U+00D8 ISOlat1 -->
<!ENTITY Ugrave "&#217;" ><!-- latin capital U with grave, U+00D9 ISOlat1 -->
<!ENTITY Uacute "&#218;" ><!-- latin capital U with acute, U+00DA ISOlat1 -->
<!ENTITY Ucirc "&#219;" ><!-- latin capital U with circumflex, U+00DB ISOlat1 -->
<!ENTITY Uuml "&#220;" ><!-- latin capital U with diaeresis, U+00DC ISOlat1 -->
<!ENTITY Yacute "&#221;" ><!-- latin capital Y with acute, U+00DD ISOlat1 -->
<!ENTITY THORN "&#222;" ><!-- latin capital THORN, U+00DE ISOlat1 -->
<!ENTITY szlig "&#223;" ><!-- latin small sharp s = ess-zed, U+00DF ISOlat1 -->
<!ENTITY agrave "&#224;" ><!-- latin small a with grave = latin small a grave, U+00E0 ISOlat1 -->
<!ENTITY aacute "&#225;" ><!-- latin small a with acute, U+00E1 ISOlat1 -->
<!ENTITY acirc "&#226;" ><!-- latin small a with circumflex, U+00E2 ISOlat1 -->
<!ENTITY atilde "&#227;" ><!-- latin small a with tilde, U+00E3 ISOlat1 -->
<!ENTITY auml "&#228;" ><!-- latin small a with diaeresis, U+00E4 ISOlat1 -->
<!ENTITY aring "&#229;" ><!-- latin small a with ring above = latin small a ring, U+00E5 ISOlat1 -->
<!ENTITY aelig "&#230;" ><!-- latin small ae = latin small ligature ae, U+00E6 ISOlat1 -->
<!ENTITY ccedil "&#231;" ><!-- latin small c with cedilla, U+00E7 ISOlat1 -->
<!ENTITY egrave "&#232;" ><!-- latin small e with grave, U+00E8 ISOlat1 -->
<!ENTITY eacute "&#233;" ><!-- latin small e with acute, U+00E9 ISOlat1 -->
<!ENTITY ecirc "&#234;" ><!-- latin small e with circumflex, U+00EA ISOlat1 -->
<!ENTITY euml "&#235;" ><!-- latin small e with diaeresis, U+00EB ISOlat1 -->
<!ENTITY igrave "&#236;" ><!-- latin small i with grave, U+00EC ISOlat1 -->
<!ENTITY iacute "&#237;" ><!-- latin small i with acute, U+00ED ISOlat1 -->
<!ENTITY icirc "&#238;" ><!-- latin small i with circumflex, U+00EE ISOlat1 -->
<!ENTITY iuml "&#239;" ><!-- latin small i with diaeresis, U+00EF ISOlat1 -->
<!ENTITY eth "&#240;" ><!-- latin small eth, U+00F0 ISOlat1 -->
<!ENTITY ntilde "&#241;" ><!-- latin small n with tilde, U+00F1 ISOlat1 -->
<!ENTITY ograve "&#242;" ><!-- latin small o with grave, U+00F2 ISOlat1 -->
<!ENTITY oacute "&#243;" ><!-- latin small o with acute, U+00F3 ISOlat1 -->
<!ENTITY ocirc "&#244;" ><!-- latin small o with circumflex, U+00F4 ISOlat1 -->
<!ENTITY otilde "&#245;" ><!-- latin small o with tilde, U+00F5 ISOlat1 -->
<!ENTITY ouml "&#246;" ><!-- latin small o with diaeresis, U+00F6 ISOlat1 -->
<!ENTITY divide "&#247;" ><!-- division sign, U+00F7 ISOnum -->
<!ENTITY oslash "&#248;" ><!-- latin small o with stroke, = latin small o slash, U+00F8 ISOlat1 -->
<!ENTITY ugrave "&#249;" ><!-- latin small u with grave, U+00F9 ISOlat1 -->
<!ENTITY uacute "&#250;" ><!-- latin small u with acute, U+00FA ISOlat1 -->
<!ENTITY ucirc "&#251;" ><!-- latin small u with circumflex, U+00FB ISOlat1 -->
<!ENTITY uuml "&#252;" ><!-- latin small u with diaeresis, U+00FC ISOlat1 -->
<!ENTITY yacute "&#253;" ><!-- latin small y with acute, U+00FD ISOlat1 -->
<!ENTITY thorn "&#254;" ><!-- latin small thorn with, U+00FE ISOlat1 -->
<!ENTITY yuml "&#255;" ><!-- latin small y with diaeresis, U+00FF ISOlat1 -->
<!-- end of xhtml-lat1.ent -->
F.1.2. XHTML Special Characters
You can download this version of this file from http://www.w3.org/TR/2010/REC-xhtml-modularization/DTD/xhtml-special.ent. The latest version is available at http://www.w3.org/MarkUp/DTD/xhtml-special.ent.
<!-- ...................................................................... -->
<!-- XML-compatible ISO Special Character Entity Set for XHTML ............ -->
<!-- file: xhtml-special.ent
Typical invocation:
<!ENTITY % xhtml-special
PUBLIC "-//W3C//ENTITIES Special for XHTML//EN"
"xhtml-special.ent" >
%xhtml-special;
This DTD module is identified by the PUBLIC and SYSTEM identifiers:
PUBLIC "-//W3C//ENTITIES Special for XHTML//EN"
SYSTEM "http://www.w3.org/MarkUp/DTD/xhtml-special.ent"
Revision: $Id: xhtml-special.ent,v 4.1 2001/04/10 09:34:14 altheim Exp $ SMI
Portions (C) International Organization for Standardization 1986:
Permission to copy in any form is granted for use with conforming
SGML systems and applications as defined in ISO 8879, provided
this notice is included in all copies.
Revisions:
2000-10-28: added &apos; and altered XML Predefined Entities for compatibility
-->
<!-- Relevant ISO entity set is given unless names are newly introduced.
New names (i.e., not in ISO 8879 [SGML] list) do not clash with
any existing ISO 8879 entity names. ISO 10646 [ISO10646] character
numbers are given for each character, in hex. Entity values are
decimal conversions of the ISO 10646 values and refer to the
document character set. Names are Unicode [UNICODE] names.
-->
<!-- C0 Controls and Basic Latin -->
<!ENTITY lt "&#38;#60;" ><!-- less-than sign, U+003C ISOnum -->
<!ENTITY gt "&#62;" ><!-- greater-than sign, U+003E ISOnum -->
<!ENTITY amp "&#38;#38;" ><!-- ampersand, U+0026 ISOnum -->
<!ENTITY apos "&#39;" ><!-- The Apostrophe (Apostrophe Quote, APL Quote), U+0027 ISOnum -->
<!ENTITY quot "&#34;" ><!-- quotation mark (Quote Double), U+0022 ISOnum -->
<!-- Latin Extended-A -->
<!ENTITY OElig "&#338;" ><!-- latin capital ligature OE, U+0152 ISOlat2 -->
<!ENTITY oelig "&#339;" ><!-- latin small ligature oe, U+0153 ISOlat2 -->
<!-- ligature is a misnomer, this is a separate character in some languages -->
<!ENTITY Scaron "&#352;" ><!-- latin capital letter S with caron, U+0160 ISOlat2 -->
<!ENTITY scaron "&#353;" ><!-- latin small letter s with caron, U+0161 ISOlat2 -->
<!ENTITY Yuml "&#376;" ><!-- latin capital letter Y with diaeresis, U+0178 ISOlat2 -->
<!-- Spacing Modifier Letters -->
<!ENTITY circ "&#710;" ><!-- modifier letter circumflex accent, U+02C6 ISOpub -->
<!ENTITY tilde "&#732;" ><!-- small tilde, U+02DC ISOdia -->
<!-- General Punctuation -->
<!ENTITY ensp "&#8194;" ><!-- en space, U+2002 ISOpub -->
<!ENTITY emsp "&#8195;" ><!-- em space, U+2003 ISOpub -->
<!ENTITY thinsp "&#8201;" ><!-- thin space, U+2009 ISOpub -->
<!ENTITY zwnj "&#8204;" ><!-- zero width non-joiner, U+200C NEW RFC 2070 -->
<!ENTITY zwj "&#8205;" ><!-- zero width joiner, U+200D NEW RFC 2070 -->
<!ENTITY lrm "&#8206;" ><!-- left-to-right mark, U+200E NEW RFC 2070 -->
<!ENTITY rlm "&#8207;" ><!-- right-to-left mark, U+200F NEW RFC 2070 -->
<!ENTITY ndash "&#8211;" ><!-- en dash, U+2013 ISOpub -->
<!ENTITY mdash "&#8212;" ><!-- em dash, U+2014 ISOpub -->
<!ENTITY lsquo "&#8216;" ><!-- left single quotation mark, U+2018 ISOnum -->
<!ENTITY rsquo "&#8217;" ><!-- right single quotation mark, U+2019 ISOnum -->
<!ENTITY sbquo "&#8218;" ><!-- single low-9 quotation mark, U+201A NEW -->
<!ENTITY ldquo "&#8220;" ><!-- left double quotation mark, U+201C ISOnum -->
<!ENTITY rdquo "&#8221;" ><!-- right double quotation mark, U+201D ISOnum -->
<!ENTITY bdquo "&#8222;" ><!-- double low-9 quotation mark, U+201E NEW -->
<!ENTITY dagger "&#8224;" ><!-- dagger, U+2020 ISOpub -->
<!ENTITY Dagger "&#8225;" ><!-- double dagger, U+2021 ISOpub -->
<!ENTITY permil "&#8240;" ><!-- per mille sign, U+2030 ISOtech -->
<!-- lsaquo is proposed but not yet ISO standardized -->
<!ENTITY lsaquo "&#8249;" ><!-- single left-pointing angle quotation mark, U+2039 ISO proposed -->
<!-- rsaquo is proposed but not yet ISO standardized -->
<!ENTITY rsaquo "&#8250;" ><!-- single right-pointing angle quotation mark, U+203A ISO proposed -->
<!ENTITY euro "&#8364;" ><!-- euro sign, U+20AC NEW -->
<!-- end of xhtml-special.ent -->
F.1.3. XHTML Mathematical, Greek, and Symbolic Characters
You can download this version of this file from http://www.w3.org/TR/2010/REC-xhtml-modularization/DTD/xhtml-symbol.ent. The latest version is available at http://www.w3.org/MarkUp/DTD/xhtml-symbol.ent.
<!-- ...................................................................... -->
<!-- ISO Math, Greek and Symbolic Character Entity Set for XHTML .......... -->
<!-- file: xhtml-symbol.ent
Typical invocation:
<!ENTITY % xhtml-symbol
PUBLIC "-//W3C//ENTITIES Symbols for XHTML//EN"
"xhtml-symbol.ent" >
%xhtml-symbol;
This DTD module is identified by the PUBLIC and SYSTEM identifiers:
PUBLIC "-//W3C//ENTITIES Symbols for XHTML//EN"
SYSTEM "http://www.w3.org/MarkUp/DTD/xhtml-symbol.ent"
Revision: $Id: xhtml-symbol.ent,v 4.1 2001/04/10 09:34:14 altheim Exp $ SMI
Portions (C) International Organization for Standardization 1986:
Permission to copy in any form is granted for use with conforming
SGML systems and applications as defined in ISO 8879, provided
this notice is included in all copies.
-->
<!-- Relevant ISO entity set is given unless names are newly introduced.
New names (i.e., not in ISO 8879 [SGML] list) do not clash with
any existing ISO 8879 entity names. ISO 10646 [ISO10646] character
numbers are given for each character, in hex. Entity values are
decimal conversions of the ISO 10646 values and refer to the
document character set. Names are Unicode [UNICODE] names.
-->
<!-- Latin Extended-B -->
<!ENTITY fnof "&#402;" ><!-- latin small f with hook = function
= florin, U+0192 ISOtech -->
<!-- Greek -->
<!ENTITY Alpha "&#913;" ><!-- greek capital letter alpha, U+0391 -->
<!ENTITY Beta "&#914;" ><!-- greek capital letter beta, U+0392 -->
<!ENTITY Gamma "&#915;" ><!-- greek capital letter gamma, U+0393 ISOgrk3 -->
<!ENTITY Delta "&#916;" ><!-- greek capital letter delta, U+0394 ISOgrk3 -->
<!ENTITY Epsilon "&#917;" ><!-- greek capital letter epsilon, U+0395 -->
<!ENTITY Zeta "&#918;" ><!-- greek capital letter zeta, U+0396 -->
<!ENTITY Eta "&#919;" ><!-- greek capital letter eta, U+0397 -->
<!ENTITY Theta "&#920;" ><!-- greek capital letter theta, U+0398 ISOgrk3 -->
<!ENTITY Iota "&#921;" ><!-- greek capital letter iota, U+0399 -->
<!ENTITY Kappa "&#922;" ><!-- greek capital letter kappa, U+039A -->
<!ENTITY Lambda "&#923;" ><!-- greek capital letter lambda, U+039B ISOgrk3 -->
<!ENTITY Mu "&#924;" ><!-- greek capital letter mu, U+039C -->
<!ENTITY Nu "&#925;" ><!-- greek capital letter nu, U+039D -->
<!ENTITY Xi "&#926;" ><!-- greek capital letter xi, U+039E ISOgrk3 -->
<!ENTITY Omicron "&#927;" ><!-- greek capital letter omicron, U+039F -->
<!ENTITY Pi "&#928;" ><!-- greek capital letter pi, U+03A0 ISOgrk3 -->
<!ENTITY Rho "&#929;" ><!-- greek capital letter rho, U+03A1 -->
<!-- there is no Sigmaf, and no U+03A2 character either -->
<!ENTITY Sigma "&#931;" ><!-- greek capital letter sigma, U+03A3 ISOgrk3 -->
<!ENTITY Tau "&#932;" ><!-- greek capital letter tau, U+03A4 -->
<!ENTITY Upsilon "&#933;" ><!-- greek capital letter upsilon,
U+03A5 ISOgrk3 -->
<!ENTITY Phi "&#934;" ><!-- greek capital letter phi, U+03A6 ISOgrk3 -->
<!ENTITY Chi "&#935;" ><!-- greek capital letter chi, U+03A7 -->
<!ENTITY Psi "&#936;" ><!-- greek capital letter psi, U+03A8 ISOgrk3 -->
<!ENTITY Omega "&#937;" ><!-- greek capital letter omega, U+03A9 ISOgrk3 -->
<!ENTITY alpha "&#945;" ><!-- greek small letter alpha, U+03B1 ISOgrk3 -->
<!ENTITY beta "&#946;" ><!-- greek small letter beta, U+03B2 ISOgrk3 -->
<!ENTITY gamma "&#947;" ><!-- greek small letter gamma, U+03B3 ISOgrk3 -->
<!ENTITY delta "&#948;" ><!-- greek small letter delta, U+03B4 ISOgrk3 -->
<!ENTITY epsilon "&#949;" ><!-- greek small letter epsilon, U+03B5 ISOgrk3 -->
<!ENTITY zeta "&#950;" ><!-- greek small letter zeta, U+03B6 ISOgrk3 -->
<!ENTITY eta "&#951;" ><!-- greek small letter eta, U+03B7 ISOgrk3 -->
<!ENTITY theta "&#952;" ><!-- greek small letter theta, U+03B8 ISOgrk3 -->
<!ENTITY iota "&#953;" ><!-- greek small letter iota, U+03B9 ISOgrk3 -->
<!ENTITY kappa "&#954;" ><!-- greek small letter kappa, U+03BA ISOgrk3 -->
<!ENTITY lambda "&#955;" ><!-- greek small letter lambda, U+03BB ISOgrk3 -->
<!ENTITY mu "&#956;" ><!-- greek small letter mu, U+03BC ISOgrk3 -->
<!ENTITY nu "&#957;" ><!-- greek small letter nu, U+03BD ISOgrk3 -->
<!ENTITY xi "&#958;" ><!-- greek small letter xi, U+03BE ISOgrk3 -->
<!ENTITY omicron "&#959;" ><!-- greek small letter omicron, U+03BF NEW -->
<!ENTITY pi "&#960;" ><!-- greek small letter pi, U+03C0 ISOgrk3 -->
<!ENTITY rho "&#961;" ><!-- greek small letter rho, U+03C1 ISOgrk3 -->
<!ENTITY sigmaf "&#962;" ><!-- greek small letter final sigma, U+03C2 ISOgrk3 -->
<!ENTITY sigma "&#963;" ><!-- greek small letter sigma, U+03C3 ISOgrk3 -->
<!ENTITY tau "&#964;" ><!-- greek small letter tau, U+03C4 ISOgrk3 -->
<!ENTITY upsilon "&#965;" ><!-- greek small letter upsilon, U+03C5 ISOgrk3 -->
<!ENTITY phi "&#966;" ><!-- greek small letter phi, U+03C6 ISOgrk3 -->
<!ENTITY chi "&#967;" ><!-- greek small letter chi, U+03C7 ISOgrk3 -->
<!ENTITY psi "&#968;" ><!-- greek small letter psi, U+03C8 ISOgrk3 -->
<!ENTITY omega "&#969;" ><!-- greek small letter omega, U+03C9 ISOgrk3 -->
<!ENTITY thetasym "&#977;" ><!-- greek small letter theta symbol, U+03D1 NEW -->
<!ENTITY upsih "&#978;" ><!-- greek upsilon with hook symbol, U+03D2 NEW -->
<!ENTITY piv "&#982;" ><!-- greek pi symbol, U+03D6 ISOgrk3 -->
<!-- General Punctuation -->
<!ENTITY bull "&#8226;" ><!-- bullet = black small circle, U+2022 ISOpub -->
<!-- bullet is NOT the same as bullet operator, U+2219 -->
<!ENTITY hellip "&#8230;" ><!-- horizontal ellipsis = three dot leader, U+2026 ISOpub -->
<!ENTITY prime "&#8242;" ><!-- prime = minutes = feet, U+2032 ISOtech -->
<!ENTITY Prime "&#8243;" ><!-- double prime = seconds = inches, U+2033 ISOtech -->
<!ENTITY oline "&#8254;" ><!-- overline = spacing overscore, U+203E NEW -->
<!ENTITY frasl "&#8260;" ><!-- fraction slash, U+2044 NEW -->
<!-- Letterlike Symbols -->
<!ENTITY weierp "&#8472;" ><!-- script capital P = power set = Weierstrass p, U+2118 ISOamso -->
<!ENTITY image "&#8465;" ><!-- blackletter capital I = imaginary part, U+2111 ISOamso -->
<!ENTITY real "&#8476;" ><!-- blackletter capital R = real part symbol, U+211C ISOamso -->
<!ENTITY trade "&#8482;" ><!-- trade mark sign, U+2122 ISOnum -->
<!ENTITY alefsym "&#8501;" ><!-- alef symbol = first transfinite cardinal, U+2135 NEW -->
<!-- alef symbol is NOT the same as hebrew letter alef, U+05D0 although
the same glyph could be used to depict both characters -->
<!-- Arrows -->
<!ENTITY larr "&#8592;" ><!-- leftwards arrow, U+2190 ISOnum -->
<!ENTITY uarr "&#8593;" ><!-- upwards arrow, U+2191 ISOnum-->
<!ENTITY rarr "&#8594;" ><!-- rightwards arrow, U+2192 ISOnum -->
<!ENTITY darr "&#8595;" ><!-- downwards arrow, U+2193 ISOnum -->
<!ENTITY harr "&#8596;" ><!-- left right arrow, U+2194 ISOamsa -->
<!ENTITY crarr "&#8629;" ><!-- downwards arrow with corner leftwards
= carriage return, U+21B5 NEW -->
<!ENTITY lArr "&#8656;" ><!-- leftwards double arrow, U+21D0 ISOtech -->
<!-- Unicode does not say that lArr is the same as the 'is implied by' arrow
but also does not have any other character for that function. So ? lArr can
be used for 'is implied by' as ISOtech suggests -->
<!ENTITY uArr "&#8657;" ><!-- upwards double arrow, U+21D1 ISOamsa -->
<!ENTITY rArr "&#8658;" ><!-- rightwards double arrow, U+21D2 ISOtech -->
<!-- Unicode does not say this is the 'implies' character but does not have
another character with this function so ?
rArr can be used for 'implies' as ISOtech suggests -->
<!ENTITY dArr "&#8659;" ><!-- downwards double arrow, U+21D3 ISOamsa -->
<!ENTITY hArr "&#8660;" ><!-- left right double arrow, U+21D4 ISOamsa -->
<!-- Mathematical Operators -->
<!ENTITY forall "&#8704;" ><!-- for all, U+2200 ISOtech -->
<!ENTITY part "&#8706;" ><!-- partial differential, U+2202 ISOtech -->
<!ENTITY exist "&#8707;" ><!-- there exists, U+2203 ISOtech -->
<!ENTITY empty "&#8709;" ><!-- empty set = null set, U+2205 ISOamso -->
<!ENTITY nabla "&#8711;" ><!-- nabla = backward difference, U+2207 ISOtech -->
<!ENTITY isin "&#8712;" ><!-- element of, U+2208 ISOtech -->
<!ENTITY notin "&#8713;" ><!-- not an element of, U+2209 ISOtech -->
<!ENTITY ni "&#8715;" ><!-- contains as member, U+220B ISOtech -->
<!-- should there be a more memorable name than 'ni'? -->
<!ENTITY prod "&#8719;" ><!-- n-ary product = product sign, U+220F ISOamsb -->
<!-- prod is NOT the same character as U+03A0 'greek capital letter pi' though
the same glyph might be used for both -->
<!ENTITY sum "&#8721;" ><!-- n-ary sumation, U+2211 ISOamsb -->
<!-- sum is NOT the same character as U+03A3 'greek capital letter sigma'
though the same glyph might be used for both -->
<!ENTITY minus "&#8722;" ><!-- minus sign, U+2212 ISOtech -->
<!ENTITY lowast "&#8727;" ><!-- asterisk operator, U+2217 ISOtech -->
<!ENTITY radic "&#8730;" ><!-- square root = radical sign, U+221A ISOtech -->
<!ENTITY prop "&#8733;" ><!-- proportional to, U+221D ISOtech -->
<!ENTITY infin "&#8734;" ><!-- infinity, U+221E ISOtech -->
<!ENTITY ang "&#8736;" ><!-- angle, U+2220 ISOamso -->
<!ENTITY and "&#8743;" ><!-- logical and = wedge, U+2227 ISOtech -->
<!ENTITY or "&#8744;" ><!-- logical or = vee, U+2228 ISOtech -->
<!ENTITY cap "&#8745;" ><!-- intersection = cap, U+2229 ISOtech -->
<!ENTITY cup "&#8746;" ><!-- union = cup, U+222A ISOtech -->
<!ENTITY int "&#8747;" ><!-- integral, U+222B ISOtech -->
<!ENTITY there4 "&#8756;" ><!-- therefore, U+2234 ISOtech -->
<!ENTITY sim "&#8764;" ><!-- tilde operator = varies with = similar to, U+223C ISOtech -->
<!-- tilde operator is NOT the same character as the tilde, U+007E,
although the same glyph might be used to represent both -->
<!ENTITY cong "&#8773;" ><!-- approximately equal to, U+2245 ISOtech -->
<!ENTITY asymp "&#8776;" ><!-- almost equal to = asymptotic to, U+2248 ISOamsr -->
<!ENTITY ne "&#8800;" ><!-- not equal to, U+2260 ISOtech -->
<!ENTITY equiv "&#8801;" ><!-- identical to, U+2261 ISOtech -->
<!ENTITY le "&#8804;" ><!-- less-than or equal to, U+2264 ISOtech -->
<!ENTITY ge "&#8805;" ><!-- greater-than or equal to, U+2265 ISOtech -->
<!ENTITY sub "&#8834;" ><!-- subset of, U+2282 ISOtech -->
<!ENTITY sup "&#8835;" ><!-- superset of, U+2283 ISOtech -->
<!-- note that nsup, 'not a superset of, U+2283' is not covered by the Symbol
font encoding and is not included. Should it be, for symmetry?
It is in ISOamsn -->
<!ENTITY nsub "&#8836;" ><!-- not a subset of, U+2284 ISOamsn -->
<!ENTITY sube "&#8838;" ><!-- subset of or equal to, U+2286 ISOtech -->
<!ENTITY supe "&#8839;" ><!-- superset of or equal to, U+2287 ISOtech -->
<!ENTITY oplus "&#8853;" ><!-- circled plus = direct sum, U+2295 ISOamsb -->
<!ENTITY otimes "&#8855;" ><!-- circled times = vector product, U+2297 ISOamsb -->
<!ENTITY perp "&#8869;" ><!-- up tack = orthogonal to = perpendicular, U+22A5 ISOtech -->
<!ENTITY sdot "&#8901;" ><!-- dot operator, U+22C5 ISOamsb -->
<!-- dot operator is NOT the same character as U+00B7 middle dot -->
<!-- Miscellaneous Technical -->
<!ENTITY lceil "&#8968;" ><!-- left ceiling = apl upstile, U+2308 ISOamsc -->
<!ENTITY rceil "&#8969;" ><!-- right ceiling, U+2309 ISOamsc -->
<!ENTITY lfloor "&#8970;" ><!-- left floor = apl downstile, U+230A ISOamsc -->
<!ENTITY rfloor "&#8971;" ><!-- right floor, U+230B ISOamsc -->
<!ENTITY lang "&#9001;" ><!-- left-pointing angle bracket = bra, U+2329 ISOtech -->
<!-- lang is NOT the same character as U+003C 'less than'
or U+2039 'single left-pointing angle quotation mark' -->
<!ENTITY rang "&#9002;" ><!-- right-pointing angle bracket = ket, U+232A ISOtech -->
<!-- rang is NOT the same character as U+003E 'greater than'
or U+203A 'single right-pointing angle quotation mark' -->
<!-- Geometric Shapes -->
<!ENTITY loz "&#9674;" ><!-- lozenge, U+25CA ISOpub -->
<!-- Miscellaneous Symbols -->
<!ENTITY spades "&#9824;" ><!-- black spade suit, U+2660 ISOpub -->
<!-- black here seems to mean filled as opposed to hollow -->
<!ENTITY clubs "&#9827;" ><!-- black club suit = shamrock, U+2663 ISOpub -->
<!ENTITY hearts "&#9829;" ><!-- black heart suit = valentine, U+2665 ISOpub -->
<!ENTITY diams "&#9830;" ><!-- black diamond suit, U+2666 ISOpub -->
<!-- end of xhtml-symbol.ent -->
"""
return text
def get_apache_license():
license = r"""/**
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
"""
return license
main()

366
fine-lucene/src/main/java/com/fr/third/org/apache/lucene/util/automaton/UTF32ToUTF8.py
Unescape View File

@ -1,366 +0,0 @@
# Licensed to the Apache Software Foundation (ASF) under one or more
# contributor license agreements. See the NOTICE file distributed with
# this work for additional information regarding copyright ownership.
# The ASF licenses this file to You under the Apache License, Version 2.0
# (the "License"); you may not use this file except in compliance with
# the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import types
import os
import sys
import random
MAX_UNICODE = 0x10FFFF
# TODO
# - could be more minimal
# - eg when bracket lands on a utf8 boundary, like 3 - 2047 -- they can share the two * edges
# - also 3 2048 or 3 65536 -- it should not have an * down the red path, but it does
# MASKS[0] is bottom 1-bit
# MASKS[1] is bottom 2-bits
# ...
utf8Ranges = [(0, 127),
(128, 2047),
(2048, 65535),
(65536, 1114111)]
typeToColor = {'startend': 'purple',
'start': 'blue',
'end': 'red'}
class FSA:
def __init__(self):
# maps fromNode -> (startUTF8, endUTF8, endNode)
self.states = {}
self.nodeUpto = 0
def run(self, bytes):
state = self.start
for b in bytes:
found = False
oldState = state
for label, s, e, n in self.states[state][1:]:
if b >= s and b <= e:
if found:
raise RuntimeError('state %s has ambiguous output for byte %s' % (oldState, b))
state = n
found = True
if not found:
return -1
return state
def addEdge(self, n1, n2, v1, v2, label):
"""
Adds edge from n1-n2, utf8 byte range v1-v2.
"""
assert n1 in self.states
assert type(v1) is types.IntType
assert type(v2) is types.IntType
self.states[n1].append((label, v1, v2, n2))
def addNode(self, label=None):
try:
self.states[self.nodeUpto] = [label]
return self.nodeUpto
finally:
self.nodeUpto += 1
def toDOT(self, label):
__l = []
w = __l.append
endNode = startNode = None
for id, details in self.states.items():
name = details[0]
if name == 'end':
endNode = id
elif name == 'start':
startNode = id
w('digraph %s {' % label)
w(' rankdir=LR;')
w(' size="8,5";')
w(' node [color=white label=""]; Ns;')
w(' node [color=black];')
w(' node [shape=doublecircle, label=""]; N%s [label="%s"];' % (endNode, endNode))
w(' node [shape=circle];')
w(' N%s [label="%s"];' % (startNode, startNode))
w(' Ns -> N%s;' % startNode)
for id, details in self.states.items():
edges = details[1:]
w(' N%s [label="%s"];' % (id, id))
for type, s, e, dest in edges:
c = typeToColor.get(type, 'black')
if type == 'all*':
# special case -- matches any utf8 byte at this point
label = '*'
elif s == e:
label = '%s' % binary(s)
else:
label = '%s-%s' % (binary(s), binary(e))
w(' N%s -> N%s [label="%s" color="%s"];' % (id, dest, label, c))
if name == 'end':
endNode = id
elif name == 'start':
startNode = id
w('}')
return '\n'.join(__l)
def toPNG(self, label, pngOut):
open('tmp.dot', 'wb').write(self.toDOT(label))
if os.system('dot -Tpng tmp.dot -o %s' % pngOut):
raise RuntimeException('dot failed')
MASKS = []
v = 2
for i in range(32):
MASKS.append(v-1)
v *= 2
def binary(x):
if x == 0:
return '00000000'
l = []
while x > 0:
if x & 1 == 1:
l.append('1')
else:
l.append('0')
x = x >> 1
# big endian!
l.reverse()
l2 = []
while len(l) > 0:
s = ''.join(l[-8:])
if len(s) < 8:
s = '0'*(8-len(s)) + s
l2.append(s)
del l[-8:]
return ' '.join(l2)
def getUTF8Rest(code, numBytes):
l = []
for i in range(numBytes):
l.append((128 | (code & MASKS[5]), 6))
code = code >> 6
l.reverse()
return tuple(l)
def toUTF8(code):
# code = Unicode code point
assert code >= 0
assert code <= MAX_UNICODE
if code < 128:
# 0xxxxxxx
bytes = ((code, 7),)
elif code < 2048:
# 110yyyxx 10xxxxxx
byte1 = (6 << 5) | (code >> 6)
bytes = ((byte1, 5),) + getUTF8Rest(code, 1)
elif code < 65536:
# 1110yyyy 10yyyyxx 10xxxxxx
len = 3
byte1 = (14 << 4) | (code >> 12)
bytes = ((byte1, 4),) + getUTF8Rest(code, 2)
else:
# 11110zzz 10zzyyyy 10yyyyxx 10xxxxxx
len = 4
byte1 = (30 << 3) | (code >> 18)
bytes = ((byte1, 3),) + getUTF8Rest(code, 3)
return bytes
def all(fsa, startNode, endNode, startCode, endCode, left):
if len(left) == 0:
fsa.addEdge(startNode, endNode, startCode, endCode, 'all')
else:
lastN = fsa.addNode()
fsa.addEdge(startNode, lastN, startCode, endCode, 'all')
while len(left) > 1:
n = fsa.addNode()
fsa.addEdge(lastN, n, 128, 191, 'all*')
left = left[1:]
lastN = n
fsa.addEdge(lastN, endNode, 128, 191, 'all*')
def start(fsa, startNode, endNode, utf8, doAll):
if len(utf8) == 1:
fsa.addEdge(startNode, endNode, utf8[0][0], utf8[0][0] | MASKS[utf8[0][1]-1], 'start')
else:
n = fsa.addNode()
fsa.addEdge(startNode, n, utf8[0][0], utf8[0][0], 'start')
start(fsa, n, endNode, utf8[1:], True)
end = utf8[0][0] | MASKS[utf8[0][1]-1]
if doAll and utf8[0][0] != end:
all(fsa, startNode, endNode, utf8[0][0]+1, end, utf8[1:])
def end(fsa, startNode, endNode, utf8, doAll):
if len(utf8) == 1:
fsa.addEdge(startNode, endNode, utf8[0][0] & ~MASKS[utf8[0][1]-1], utf8[0][0], 'end')
else:
if utf8[0][1] == 5:
# special case -- avoid created unused edges (utf8 doesn't accept certain byte sequences):
start = 194
else:
start = utf8[0][0] & (~MASKS[utf8[0][1]-1])
if doAll and utf8[0][0] != start:
all(fsa, startNode, endNode, start, utf8[0][0]-1, utf8[1:])
n = fsa.addNode()
fsa.addEdge(startNode, n, utf8[0][0], utf8[0][0], 'end')
end(fsa, n, endNode, utf8[1:], True)
def build(fsa,
startNode, endNode,
startUTF8, endUTF8):
# Break into start, middle, end:
if startUTF8[0][0] == endUTF8[0][0]:
# Degen case: lead with the same byte:
if len(startUTF8) == 1 and len(endUTF8) == 1:
fsa.addEdge(startNode, endNode, startUTF8[0][0], endUTF8[0][0], 'startend')
return
else:
assert len(startUTF8) != 1
assert len(endUTF8) != 1
n = fsa.addNode()
# single value edge
fsa.addEdge(startNode, n, startUTF8[0][0], startUTF8[0][0], 'single')
build(fsa, n, endNode, startUTF8[1:], endUTF8[1:])
elif len(startUTF8) == len(endUTF8):
if len(startUTF8) == 1:
fsa.addEdge(startNode, endNode, startUTF8[0][0], endUTF8[0][0], 'startend')
else:
start(fsa, startNode, endNode, startUTF8, False)
if endUTF8[0][0] - startUTF8[0][0] > 1:
all(fsa, startNode, endNode, startUTF8[0][0]+1, endUTF8[0][0]-1, startUTF8[1:])
end(fsa, startNode, endNode, endUTF8, False)
else:
# start
start(fsa, startNode, endNode, startUTF8, True)
# possibly middle
byteCount = 1+len(startUTF8)
while byteCount < len(endUTF8):
s = toUTF8(utf8Ranges[byteCount-1][0])
e = toUTF8(utf8Ranges[byteCount-1][1])
all(fsa, startNode, endNode,
s[0][0],
e[0][0],
s[1:])
byteCount += 1
# end
end(fsa, startNode, endNode, endUTF8, True)
def main():
if len(sys.argv) not in (3, 4):
print
print 'Usage: python %s startUTF32 endUTF32 [testCode]' % sys.argv[0]
print
sys.exit(1)
utf32Start = int(sys.argv[1])
utf32End = int(sys.argv[2])
if utf32Start > utf32End:
print 'ERROR: start must be <= end'
sys.exit(1)
fsa = FSA()
fsa.start = fsa.addNode('start')
fsa.end = fsa.addNode('end')
print 's=%s' % ' '.join([binary(x[0]) for x in toUTF8(utf32Start)])
print 'e=%s' % ' '.join([binary(x[0]) for x in toUTF8(utf32End)])
if len(sys.argv) == 4:
print 't=%s [%s]' % \
(' '.join([binary(x[0]) for x in toUTF8(int(sys.argv[3]))]),
' '.join(['%2x' % x[0] for x in toUTF8(int(sys.argv[3]))]))
build(fsa, fsa.start, fsa.end,
toUTF8(utf32Start),
toUTF8(utf32End))
fsa.toPNG('test', '/tmp/outpy.png')
print 'Saved to /tmp/outpy.png...'
test(fsa, utf32Start, utf32End, 100000);
def test(fsa, utf32Start, utf32End, count):
# verify correct ints are accepted
for i in range(count):
r = random.randint(utf32Start, utf32End)
dest = fsa.run([tup[0] for tup in toUTF8(r)])
if dest != fsa.end:
print 'FAILED: valid %s (%s) is not accepted' % (r, ' '.join([binary(x[0]) for x in toUTF8(r)]))
return False
invalidRange = MAX_UNICODE - (utf32End - utf32Start + 1)
if invalidRange >= 0:
# verify invalid ints are not accepted
for i in range(count):
r = random.randint(0, invalidRange-1)
if r >= utf32Start:
r = utf32End + 1 + r - utf32Start
dest = fsa.run([tup[0] for tup in toUTF8(r)])
if dest != -1:
print 'FAILED: invalid %s (%s) is accepted' % (r, ' '.join([binary(x[0]) for x in toUTF8(r)]))
return False
return True
def stress():
print 'Testing...'
iter = 0
while True:
if iter % 10 == 0:
print '%s...' % iter
iter += 1
v1 = random.randint(0, MAX_UNICODE)
v2 = random.randint(0, MAX_UNICODE)
if v2 < v1:
v1, v2 = v2, v1
utf32Start = v1
utf32End = v2
fsa = FSA()
fsa.start = fsa.addNode('start')
fsa.end = fsa.addNode('end')
build(fsa, fsa.start, fsa.end,
toUTF8(utf32Start),
toUTF8(utf32End))
if not test(fsa, utf32Start, utf32End, 10000):
print 'FAILED on utf32Start=%s utf32End=%s' % (utf32Start, utf32End)
if __name__ == '__main__':
if len(sys.argv) > 1:
main()
else:
stress()

500
fine-lucene/src/main/java/com/fr/third/org/apache/lucene/util/automaton/createLevAutomata.py
Unescape View File

@ -1,500 +0,0 @@
# Licensed to the Apache Software Foundation (ASF) under one or more
# contributor license agreements. See the NOTICE file distributed with
# this work for additional information regarding copyright ownership.
# The ASF licenses this file to You under the Apache License, Version 2.0
# (the "License"); you may not use this file except in compliance with
# the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# Note, this file is known to work with rev 120 of the moman
# repository (http://bitbucket.org/jpbarrette/moman/overview)
#
# See also: http://sites.google.com/site/rrettesite/moman
import math
import os
import sys
#sys.path.insert(0, 'moman/finenight/python')
sys.path.insert(0, '../../../../../../../../build/core/moman/finenight/python')
try:
from possibleStates import genTransitions
except ImportError:
from finenight.possibleStates import genTransitions
MODE = 'array'
PACKED = True
WORD = 64
LOG2_WORD = int(math.log(WORD)/math.log(2))
#MODE = 'switch'
class LineOutput:
def __init__(self, indent=''):
self.l = []
self._indent = self.startIndent = indent
self.inComment = False
def __call__(self, s, indent=0):
if s.find('}') != -1:
assert self._indent != self.startIndent
self._indent = self._indent[:-2]
if indent != 0:
indent0 = ' ' * (len(self._indent)/2+indent)
else:
indent0 = self._indent
if s.find('/*') != -1:
if s.find('*/') == -1:
self.inComment = True
elif s.find('*/') != -1:
self.inComment = True
if self.inComment:
self.l.append(indent0 + s)
else:
self.l.append(indent0 + s.lstrip())
self.inComment = self.inComment and s.find('*/') == -1
if s.find('{') != -1:
self._indent += ' '
def __str__(self):
if True:
assert self._indent == self.startIndent, 'indent %d vs start indent %d' % \
(len(self._indent), len(self.startIndent))
return '\n'.join(self.l)
def indent(self):
self._indent += ' '
def outdent(self):
assert self._indent != self.startIndent
self._indent = self._indent[:-2]
def charVarNumber(charVar):
"""
Maps binary number (eg [1, 0, 1]) to its decimal value (5).
"""
p = 1
sum = 0
downTo = len(charVar)-1
while downTo >= 0:
sum += p * int(charVar[downTo])
p *= 2
downTo -= 1
return sum
def main():
if len(sys.argv) != 3:
print
print 'Usage: python -u %s N <True/False>' % sys.argv[0]
print
print 'NOTE: the resulting .java file is created in the current working dir!'
print
sys.exit(1)
n = int(sys.argv[1])
transpose = (sys.argv[2] == "True")
tables = genTransitions(n, transpose)
stateMap = {}
# init null state
stateMap['[]'] = -1
# init start state
stateMap['[(0, 0)]'] = 0
w = LineOutput()
w('package com.fr.third.org.apache.lucene.util.automaton;')
w('')
w('/*')
w(' * Licensed to the Apache Software Foundation (ASF) under one or more')
w(' * contributor license agreements. See the NOTICE file distributed with')
w(' * this work for additional information regarding copyright ownership.')
w(' * The ASF licenses this file to You under the Apache License, Version 2.0')
w(' * (the "License"); you may not use this file except in compliance with')
w(' * the License. You may obtain a copy of the License at')
w(' *')
w(' * http://www.apache.org/licenses/LICENSE-2.0')
w(' *')
w(' * Unless required by applicable law or agreed to in writing, software')
w(' * distributed under the License is distributed on an "AS IS" BASIS,')
w(' * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.')
w(' * See the License for the specific language governing permissions and')
w(' * limitations under the License.')
w(' */')
w('')
w('// The following code was generated with the moman/finenight pkg')
w('// This package is available under the MIT License, see NOTICE.txt')
w('// for more details.')
w('')
w('import com.fr.third.org.apache.lucene.util.automaton.LevenshteinAutomata.ParametricDescription;')
w('')
if transpose:
w('/** Parametric description for generating a Levenshtein automaton of degree %s, ' % n)
w(' with transpositions as primitive edits */')
className = 'Lev%dTParametricDescription' % n
else:
w('/** Parametric description for generating a Levenshtein automaton of degree %s */' % n)
className = 'Lev%dParametricDescription' % n
w('class %s extends ParametricDescription {' % className)
w('')
w('@Override')
w('int transition(int absState, int position, int vector) {')
w(' // null absState should never be passed in')
w(' assert absState != -1;')
w('')
w(' // decode absState -> state, offset')
w(' int state = absState/(w+1);')
w(' int offset = absState%(w+1);')
w(' assert offset >= 0;')
w('')
machines = []
for i, map in enumerate(tables):
if i == 0:
w('if (position == w) {')
elif i == len(tables)-1:
w('} else {')
else:
w('} else if (position == w-%d) {' % i)
if i != 0 and MODE == 'switch':
w('switch(vector) {')
l = map.items()
l.sort()
numCasesPerVector = None
numVectors = len(l)
if MODE == 'array':
toStateArray = []
toOffsetIncrArray = []
for charVar, states in l:
# somehow it's a string:
charVar = eval(charVar)
if i != 0 and MODE == 'switch':
w('case %s: // <%s>' % (charVarNumber(charVar), ','.join([str(x) for x in charVar])))
w.indent()
l = states.items()
byFromState = {}
# first pass to assign states
byAction = {}
for s, (toS, offset) in l:
state = str(s)
toState = str(toS)
if state not in stateMap:
stateMap[state] = len(stateMap)-1
if toState not in stateMap:
stateMap[toState] = len(stateMap)-1
byFromState[stateMap[state]] = (1+stateMap[toState], offset)
fromStateDesc = s[1:len(s)-1]
toStateDesc = ', '.join([str(x) for x in toS])
tup = (stateMap[toState], toStateDesc, offset)
if tup not in byAction:
byAction[tup] = []
byAction[tup].append((fromStateDesc, stateMap[state]))
if numCasesPerVector is None:
numCasesPerVector = len(l)
else:
# we require this to be uniform... empirically it seems to be!
assert numCasesPerVector == len(l)
if MODE == 'array':
for s in range(numCasesPerVector):
toState, offsetIncr = byFromState[s]
toStateArray.append(toState)
toOffsetIncrArray.append(offsetIncr)
else:
# render switches
w('switch(state) { // %s cases' % len(l))
for (toState, toStateDesc, offset), lx in byAction.items():
for fromStateDesc, fromState in lx:
w('case %s: // %s' % (fromState, fromStateDesc))
w.indent()
w(' state = %s; // %s' % (toState, toStateDesc))
if offset > 0:
w(' offset += %s;' % offset)
w('break;')
w.outdent()
w('}')
if i != 0:
w('break;')
w.outdent()
if MODE == 'array':
# strangely state can come in wildly out of bounds....
w(' if (state < %d) {' % numCasesPerVector)
w(' final int loc = vector * %d + state;' % numCasesPerVector)
if PACKED:
w(' offset += unpack(offsetIncrs%d, loc, NBITSOFFSET%d);' % (i, i))
w(' state = unpack(toStates%d, loc, NBITSSTATES%d)-1;' % (i, i))
else:
w(' offset += offsetIncrs%d[loc];' % i)
w(' state = toStates%d[loc]-1;' % i)
w(' }')
elif i != 0:
w('}')
machines.append((toStateArray, toOffsetIncrArray, numCasesPerVector, numVectors))
# ends switch statement for machine
w('}')
w('')
w(' if (state == -1) {')
w(' // null state')
w(' return -1;')
w(' } else {')
w(' // translate back to abs')
w(' return state*(w+1)+offset;')
w(' }')
# ends transition method
w('}')
subs = []
if MODE == 'array':
w.indent()
for i, (toStateArray, toOffsetIncrsArray, numCasesPerVector, numVectors) in enumerate(machines):
w('')
w.outdent()
w('// %d vectors; %d states per vector; array length = %d' % \
(numVectors, numCasesPerVector, numVectors*numCasesPerVector))
w.indent()
if PACKED:
# pack in python
l, nbits = pack(toStateArray)
subs.append(('NBITSSTATES%d' % i, str(nbits)))
w(' private final static long[] toStates%d = new long[] /*%d bits per value */ %s;' % \
(i, nbits, renderList([hex(long(x)) for x in l])))
l, nbits = pack(toOffsetIncrsArray)
subs.append(('NBITSOFFSET%d' % i, str(nbits)))
w(' private final static long[] offsetIncrs%d = new long[] /*%d bits per value */ %s;' % \
(i, nbits, renderList([hex(long(x)) for x in l])))
else:
w(' private final static int[] toStates%d = new int[] %s;' % \
(i, renderList([str(x) for x in toStateArray])))
w(' private final static int[] offsetIncrs%d = new int[] %s;' % \
(i, renderList([str(x) for x in toStateArray])))
w.outdent()
stateMap2 = dict([[v,k] for k,v in stateMap.items()])
w('')
w('// state map')
sum = 0
minErrors = []
for i in xrange(len(stateMap2)-1):
w('// %s -> %s' % (i, stateMap2[i]))
# we replace t-notation as its not relevant here
st = stateMap2[i].replace('t', '')
v = eval(st)
minError = min([-i+e for i, e in v])
c = len(v)
sum += c
minErrors.append(minError)
w('')
w.indent()
#w('private final static int[] minErrors = new int[] {%s};' % ','.join([str(x) for x in minErrors]))
w.outdent()
w('')
w(' public %s(int w) {' % className)
w(' super(w, %d, new int[] {%s});' % (n, ','.join([str(x) for x in minErrors])), indent=1)
w(' }')
if 0:
w('')
w('@Override')
w('public int size() { // this can now move up?')
w(' return %d*(w+1);' % (len(stateMap2)-1))
w('}')
w('')
w('@Override')
w('public int getPosition(int absState) { // this can now move up?')
w(' return absState % (w+1);')
w('}')
w('')
w('@Override')
w('public boolean isAccept(int absState) { // this can now move up?')
w(' // decode absState -> state, offset')
w(' int state = absState/(w+1);')
w(' if (true || state < minErrors.length) {')
w(' int offset = absState%(w+1);')
w(' assert offset >= 0;')
w(' return w - offset + minErrors[state] <= %d;' % n)
w(' } else {')
w(' return false;')
w(' }')
w('}')
if MODE == 'array' and PACKED:
# we moved into super class
if False:
w('')
v = 2
l = []
for i in range(63):
l.append(hex(v-1))
v *= 2
w('private final static long[] MASKS = new long[] {%s};' % ','.join(l), indent=1)
w('')
# unpack in java
w('private int unpack(long[] data, int index, int bitsPerValue) {')
w(' final long bitLoc = bitsPerValue * index;')
w(' final int dataLoc = (int) (bitLoc >> %d);' % LOG2_WORD)
w(' final int bitStart = (int) (bitLoc & %d);' % (WORD-1))
w(' //System.out.println("index=" + index + " dataLoc=" + dataLoc + " bitStart=" + bitStart + " bitsPerV=" + bitsPerValue);')
w(' if (bitStart + bitsPerValue <= %d) {' % WORD)
w(' // not split')
w(' return (int) ((data[dataLoc] >> bitStart) & MASKS[bitsPerValue-1]);')
w(' } else {')
w(' // split')
w(' final int part = %d-bitStart;' % WORD)
w(' return (int) (((data[dataLoc] >> bitStart) & MASKS[part-1]) +')
w(' ((data[1+dataLoc] & MASKS[bitsPerValue-part-1]) << part));', indent=1)
w(' }')
w('}')
# class
w('}')
w('')
fileOut = '%s.java' % className
s = str(w)
for sub, repl in subs:
s = s.replace(sub, repl)
open(fileOut, 'wb').write(s)
print 'Wrote %s [%d lines; %.1f KB]' % \
(fileOut, len(w.l), os.path.getsize(fileOut)/1024.)
def renderList(l):
lx = [' ']
for i in xrange(len(l)):
if i > 0:
lx.append(',')
if i % 4 == 0:
lx.append('\n ')
lx.append(l[i])
return '{\n%s\n }' % ''.join(lx)
MASKS = []
v = 2
for i in xrange(63):
MASKS.append(v-1)
v *= 2
# packs into longs; returns long[], numBits
def pack(l):
maxV = max(l)
bitsPerValue = max(1, int(math.ceil(math.log(maxV+1)/math.log(2.0))))
bitsLeft = WORD
pendingValue = 0
packed = []
for i in xrange(len(l)):
v = l[i]
if pendingValue > 0:
bitsUsed = math.ceil(math.log(pendingValue)/math.log(2.0))
assert bitsUsed <= (WORD-bitsLeft), 'bitsLeft=%s (%s-%s=%s) bitsUsed=%s' % (bitsLeft, WORD, bitsLeft, WORD-bitsLeft, bitsUsed)
if bitsLeft >= bitsPerValue:
pendingValue += v << (WORD-bitsLeft)
bitsLeft -= bitsPerValue
if bitsLeft == 0:
packed.append(pendingValue)
bitsLeft = WORD
pendingValue = 0
else:
# split
# bottom bitsLeft go in current word:
pendingValue += (v & MASKS[bitsLeft-1]) << (WORD-bitsLeft)
packed.append(pendingValue)
pendingValue = v >> bitsLeft
bitsLeft = WORD - (bitsPerValue-bitsLeft)
if bitsLeft < WORD:
packed.append(pendingValue)
# verify(l, packed, bitsPerValue)
return packed, bitsPerValue
def verify(data, packedData, bitsPerValue):
for i in range(len(data)):
assert data[i] == unpack(packedData, i, bitsPerValue)
def unpack(data, index, bitsPerValue):
bitLoc = bitsPerValue * index
dataLoc = int(bitLoc >> LOG2_WORD)
bitStart = int(bitLoc & (WORD-1))
if bitStart + bitsPerValue <= WORD:
# not split
return int(((data[dataLoc] >> bitStart) & MASKS[bitsPerValue-1]))
else:
# split
part = WORD-bitStart;
return int((((data[dataLoc] >> bitStart) & MASKS[part-1]) +
((data[1+dataLoc] & MASKS[bitsPerValue-part-1]) << part)))
if __name__ == '__main__':
if not __debug__:
print
print 'ERROR: please run without -O'
print
sys.exit(1)
main()

335
fine-lucene/src/main/java/com/fr/third/org/apache/lucene/util/packed/gen_BulkOperation.py
Unescape View File

@ -1,335 +0,0 @@
#! /usr/bin/env python
# Licensed to the Apache Software Foundation (ASF) under one or more
# contributor license agreements. See the NOTICE file distributed with
# this work for additional information regarding copyright ownership.
# The ASF licenses this file to You under the Apache License, Version 2.0
# (the "License"); you may not use this file except in compliance with
# the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from fractions import gcd
"""Code generation for bulk operations"""
MAX_SPECIALIZED_BITS_PER_VALUE = 24;
PACKED_64_SINGLE_BLOCK_BPV = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 16, 21, 32]
OUTPUT_FILE = "BulkOperation.java"
HEADER = """// This file has been automatically generated, DO NOT EDIT
package com.fr.third.org.apache.lucene.util.packed;
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
"""
FOOTER="""
protected int writeLong(long block, byte[] blocks, int blocksOffset) {
for (int j = 1; j <= 8; ++j) {
blocks[blocksOffset++] = (byte) (block >>> (64 - (j << 3)));
}
return blocksOffset;
}
/**
* For every number of bits per value, there is a minimum number of
* blocks (b) / values (v) you need to write in order to reach the next block
* boundary:
* - 16 bits per value -> b=1, v=4
* - 24 bits per value -> b=3, v=8
* - 50 bits per value -> b=25, v=32
* - 63 bits per value -> b=63, v=64
* - ...
*
* A bulk read consists in copying <code>iterations*v</code> values that are
* contained in <code>iterations*b</code> blocks into a <code>long[]</code>
* (higher values of <code>iterations</code> are likely to yield a better
* throughput) => this requires n * (b + v) longs in memory.
*
* This method computes <code>iterations</code> as
* <code>ramBudget / (8 * (b + v))</code> (since a long is 8 bytes).
*/
public final int computeIterations(int valueCount, int ramBudget) {
final int iterations = (ramBudget >>> 3) / (blockCount() + valueCount());
if (iterations == 0) {
// at least 1
return 1;
} else if ((iterations - 1) * blockCount() >= valueCount) {
// don't allocate for more than the size of the reader
return (int) Math.ceil((double) valueCount / valueCount());
} else {
return iterations;
}
}
}
"""
def is_power_of_two(n):
return n & (n - 1) == 0
def casts(typ):
cast_start = "(%s) (" %typ
cast_end = ")"
if typ == "long":
cast_start = ""
cast_end = ""
return cast_start, cast_end
def hexNoLSuffix(n):
# On 32 bit Python values > (1 << 31)-1 will have L appended by hex function:
s = hex(n)
if s.endswith('L'):
s = s[:-1]
return s
def masks(bits):
if bits == 64:
return "", ""
return "(", " & %sL)" %(hexNoLSuffix((1 << bits) - 1))
def get_type(bits):
if bits == 8:
return "byte"
elif bits == 16:
return "short"
elif bits == 32:
return "int"
elif bits == 64:
return "long"
else:
assert False
def block_value_count(bpv, bits=64):
blocks = bpv
values = blocks * bits / bpv
while blocks % 2 == 0 and values % 2 == 0:
blocks /= 2
values /= 2
assert values * bpv == bits * blocks, "%d values, %d blocks, %d bits per value" %(values, blocks, bpv)
return (blocks, values)
def packed64(bpv, f):
blocks, values = block_value_count(bpv)
mask = (1 << bpv) - 1
f.write("\n")
f.write(" public BulkOperationPacked%d() {\n" %bpv)
f.write(" super(%d);\n" %bpv)
f.write(" assert blockCount() == %d;\n" %blocks)
f.write(" assert valueCount() == %d;\n" %values)
f.write(" }\n\n")
if bpv == 64:
f.write(""" @Override
public void decode(long[] blocks, int blocksOffset, long[] values, int valuesOffset, int iterations) {
System.arraycopy(blocks, blocksOffset, values, valuesOffset, valueCount() * iterations);
}
@Override
public void decode(long[] blocks, int blocksOffset, int[] values, int valuesOffset, int iterations) {
throw new UnsupportedOperationException();
}
@Override
public void decode(byte[] blocks, int blocksOffset, int[] values, int valuesOffset, int iterations) {
throw new UnsupportedOperationException();
}
@Override
public void decode(byte[] blocks, int blocksOffset, long[] values, int valuesOffset, int iterations) {
LongBuffer.wrap(values, valuesOffset, iterations * valueCount()).put(ByteBuffer.wrap(blocks, blocksOffset, 8 * iterations * blockCount()).asLongBuffer());
}
""")
else:
p64_decode(bpv, f, 32)
p64_decode(bpv, f, 64)
def p64_decode(bpv, f, bits):
blocks, values = block_value_count(bpv)
typ = get_type(bits)
cast_start, cast_end = casts(typ)
f.write(" @Override\n")
f.write(" public void decode(long[] blocks, int blocksOffset, %s[] values, int valuesOffset, int iterations) {\n" %typ)
if bits < bpv:
f.write(" throw new UnsupportedOperationException();\n")
else:
f.write(" for (int i = 0; i < iterations; ++i) {\n")
mask = (1 << bpv) - 1
if is_power_of_two(bpv):
f.write(" final long block = blocks[blocksOffset++];\n")
f.write(" for (int shift = %d; shift >= 0; shift -= %d) {\n" %(64 - bpv, bpv))
f.write(" values[valuesOffset++] = %s(block >>> shift) & %d%s;\n" %(cast_start, mask, cast_end))
f.write(" }\n")
else:
for i in xrange(0, values):
block_offset = i * bpv / 64
bit_offset = (i * bpv) % 64
if bit_offset == 0:
# start of block
f.write(" final long block%d = blocks[blocksOffset++];\n" %block_offset);
f.write(" values[valuesOffset++] = %sblock%d >>> %d%s;\n" %(cast_start, block_offset, 64 - bpv, cast_end))
elif bit_offset + bpv == 64:
# end of block
f.write(" values[valuesOffset++] = %sblock%d & %dL%s;\n" %(cast_start, block_offset, mask, cast_end))
elif bit_offset + bpv < 64:
# middle of block
f.write(" values[valuesOffset++] = %s(block%d >>> %d) & %dL%s;\n" %(cast_start, block_offset, 64 - bit_offset - bpv, mask, cast_end))
else:
# value spans across 2 blocks
mask1 = (1 << (64 - bit_offset)) -1
shift1 = bit_offset + bpv - 64
shift2 = 64 - shift1
f.write(" final long block%d = blocks[blocksOffset++];\n" %(block_offset + 1));
f.write(" values[valuesOffset++] = %s((block%d & %dL) << %d) | (block%d >>> %d)%s;\n" %(cast_start, block_offset, mask1, shift1, block_offset + 1, shift2, cast_end))
f.write(" }\n")
f.write(" }\n\n")
byte_blocks, byte_values = block_value_count(bpv, 8)
f.write(" @Override\n")
f.write(" public void decode(byte[] blocks, int blocksOffset, %s[] values, int valuesOffset, int iterations) {\n" %typ)
if bits < bpv:
f.write(" throw new UnsupportedOperationException();\n")
else:
if is_power_of_two(bpv) and bpv < 8:
f.write(" for (int j = 0; j < 8 * iterations; ++j) {\n")
f.write(" final byte block = blocks[blocksOffset++];\n")
for shift in xrange(8 - bpv, 0, -bpv):
f.write(" values[valuesOffset++] = (block >>> %d) & %d;\n" %(shift, mask))
f.write(" values[valuesOffset++] = block & %d;\n" %mask)
f.write(" }\n")
elif bpv == 8:
f.write(" for (int j = 0; j < 8 * iterations; ++j) {\n")
f.write(" values[valuesOffset++] = blocks[blocksOffset++] & 0xFF;\n")
f.write(" }\n")
elif is_power_of_two(bpv) and bpv > 8:
f.write(" for (int j = 0; j < %d * iterations; ++j) {\n" %(64 / bpv))
m = bits <= 32 and "0xFF" or "0xFFL"
f.write(" values[valuesOffset++] =")
for i in xrange(bpv / 8 - 1):
f.write(" ((blocks[blocksOffset++] & %s) << %d) |" %(m, bpv - 8))
f.write(" (blocks[blocksOffset++] & %s);\n" %m)
f.write(" }\n")
else:
f.write(" for (int i = 0; i < 8 * iterations; ++i) {\n")
for i in xrange(0, byte_values):
byte_start = i * bpv / 8
bit_start = (i * bpv) % 8
byte_end = ((i + 1) * bpv - 1) / 8
bit_end = ((i + 1) * bpv - 1) % 8
shift = lambda b: 8 * (byte_end - b - 1) + 1 + bit_end
if bit_start == 0:
f.write(" final %s byte%d = blocks[blocksOffset++] & 0xFF;\n" %(typ, byte_start))
for b in xrange(byte_start + 1, byte_end + 1):
f.write(" final %s byte%d = blocks[blocksOffset++] & 0xFF;\n" %(typ, b))
f.write(" values[valuesOffset++] =")
if byte_start == byte_end:
if bit_start == 0:
if bit_end == 7:
f.write(" byte%d" %byte_start)
else:
f.write(" byte%d >>> %d" %(byte_start, 7 - bit_end))
else:
if bit_end == 7:
f.write(" byte%d & %d" %(byte_start, 2 ** (8 - bit_start) - 1))
else:
f.write(" (byte%d >>> %d) & %d" %(byte_start, 7 - bit_end, 2 ** (bit_end - bit_start + 1) - 1))
else:
if bit_start == 0:
f.write(" (byte%d << %d)" %(byte_start, shift(byte_start)))
else:
f.write(" ((byte%d & %d) << %d)" %(byte_start, 2 ** (8 - bit_start) - 1, shift(byte_start)))
for b in xrange(byte_start + 1, byte_end):
f.write(" | (byte%d << %d)" %(b, shift(b)))
if bit_end == 7:
f.write(" | byte%d" %byte_end)
else:
f.write(" | (byte%d >>> %d)" %(byte_end, 7 - bit_end))
f.write(";\n")
f.write(" }\n")
f.write(" }\n\n")
if __name__ == '__main__':
f = open(OUTPUT_FILE, 'w')
f.write(HEADER)
f.write('\n')
f.write('''/**
* Efficient sequential read/write of packed integers.
*/\n''')
f.write('abstract class BulkOperation implements PackedInts.Decoder, PackedInts.Encoder {\n')
f.write(' private static final BulkOperation[] packedBulkOps = new BulkOperation[] {\n')
for bpv in xrange(1, 65):
if bpv > MAX_SPECIALIZED_BITS_PER_VALUE:
f.write(' new BulkOperationPacked(%d),\n' % bpv)
continue
f2 = open('BulkOperationPacked%d.java' % bpv, 'w')
f2.write(HEADER)
if bpv == 64:
f2.write('import java.nio.LongBuffer;\n')
f2.write('import java.nio.ByteBuffer;\n')
f2.write('\n')
f2.write('''/**
* Efficient sequential read/write of packed integers.
*/\n''')
f2.write('final class BulkOperationPacked%d extends BulkOperationPacked {\n' % bpv)
packed64(bpv, f2)
f2.write('}\n')
f2.close()
f.write(' new BulkOperationPacked%d(),\n' % bpv)
f.write(' };\n')
f.write('\n')
f.write(' // NOTE: this is sparse (some entries are null):\n')
f.write(' private static final BulkOperation[] packedSingleBlockBulkOps = new BulkOperation[] {\n')
for bpv in xrange(1, max(PACKED_64_SINGLE_BLOCK_BPV)+1):
if bpv in PACKED_64_SINGLE_BLOCK_BPV:
f.write(' new BulkOperationPackedSingleBlock(%d),\n' % bpv)
else:
f.write(' null,\n')
f.write(' };\n')
f.write('\n')
f.write("\n")
f.write(" public static BulkOperation of(PackedInts.Format format, int bitsPerValue) {\n")
f.write(" switch (format) {\n")
f.write(" case PACKED:\n")
f.write(" assert packedBulkOps[bitsPerValue - 1] != null;\n")
f.write(" return packedBulkOps[bitsPerValue - 1];\n")
f.write(" case PACKED_SINGLE_BLOCK:\n")
f.write(" assert packedSingleBlockBulkOps[bitsPerValue - 1] != null;\n")
f.write(" return packedSingleBlockBulkOps[bitsPerValue - 1];\n")
f.write(" default:\n")
f.write(" throw new AssertionError();\n")
f.write(" }\n")
f.write(" }\n")
f.write(FOOTER)
f.close()

175
fine-lucene/src/main/java/com/fr/third/org/apache/lucene/util/packed/gen_Direct.py
Unescape View File

@ -1,175 +0,0 @@
#! /usr/bin/env python
# Licensed to the Apache Software Foundation (ASF) under one or more
# contributor license agreements. See the NOTICE file distributed with
# this work for additional information regarding copyright ownership.
# The ASF licenses this file to You under the Apache License, Version 2.0
# (the "License"); you may not use this file except in compliance with
# the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
HEADER="""// This file has been automatically generated, DO NOT EDIT
package com.fr.third.org.apache.lucene.util.packed;
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
import com.fr.third.org.apache.lucene.store.DataInput;
import com.fr.third.org.apache.lucene.util.RamUsageEstimator;
import java.io.IOException;
import java.util.Arrays;
"""
TYPES = {8: "byte", 16: "short", 32: "int", 64: "long"}
MASKS = {8: " & 0xFFL", 16: " & 0xFFFFL", 32: " & 0xFFFFFFFFL", 64: ""}
CASTS = {8: "(byte) ", 16: "(short) ", 32: "(int) ", 64: ""}
if __name__ == '__main__':
for bpv in TYPES.keys():
type
f = open("Direct%d.java" %bpv, 'w')
f.write(HEADER)
f.write("""/**
* Direct wrapping of %d-bits values to a backing array.
* @lucene.internal
*/\n""" %bpv)
f.write("final class Direct%d extends PackedInts.MutableImpl {\n" %bpv)
f.write(" final %s[] values;\n\n" %TYPES[bpv])
f.write(" Direct%d(int valueCount) {\n" %bpv)
f.write(" super(valueCount, %d);\n" %bpv)
f.write(" values = new %s[valueCount];\n" %TYPES[bpv])
f.write(" }\n\n")
f.write(" Direct%d(DataInput in, int valueCount) throws IOException {\n" %bpv)
f.write(" this(valueCount);\n")
f.write(" for (int i = 0; i < valueCount; ++i) {\n")
f.write(" values[i] = in.read%s();\n" %TYPES[bpv].title())
f.write(" }\n")
if bpv != 64:
f.write(" final int mod = valueCount %% %d;\n" %(64 / bpv))
f.write(" if (mod != 0) {\n")
f.write(" for (int i = mod; i < %d; ++i) {\n" %(64 / bpv))
f.write(" in.read%s();\n" %TYPES[bpv].title())
f.write(" }\n")
f.write(" }\n")
f.write(" }\n")
f.write("""
@Override
public long get(final int index) {
return values[index]%s;
}
public void set(final int index, final long value) {
values[index] = %s(value);
}
public long ramBytesUsed() {
return RamUsageEstimator.sizeOf(values);
}
public void clear() {
Arrays.fill(values, %s0L);
}
@Override
public Object getArray() {
return values;
}
@Override
public boolean hasArray() {
return true;
}
""" %(MASKS[bpv], CASTS[bpv], CASTS[bpv]))
if bpv == 64:
f.write("""
@Override
public int get(int index, long[] arr, int off, int len) {
assert len > 0 : "len must be > 0 (got " + len + ")";
assert index >= 0 && index < valueCount;
assert off + len <= arr.length;
final int gets = Math.min(valueCount - index, len);
System.arraycopy(values, index, arr, off, gets);
return gets;
}
public int set(int index, long[] arr, int off, int len) {
assert len > 0 : "len must be > 0 (got " + len + ")";
assert index >= 0 && index < valueCount;
assert off + len <= arr.length;
final int sets = Math.min(valueCount - index, len);
System.arraycopy(arr, off, values, index, sets);
return sets;
}
@Override
public void fill(int fromIndex, int toIndex, long val) {
Arrays.fill(values, fromIndex, toIndex, val);
}
""")
else:
f.write("""
@Override
public int get(int index, long[] arr, int off, int len) {
assert len > 0 : "len must be > 0 (got " + len + ")";
assert index >= 0 && index < valueCount;
assert off + len <= arr.length;
final int gets = Math.min(valueCount - index, len);
for (int i = index, o = off, end = index + gets; i < end; ++i, ++o) {
arr[o] = values[i]%s;
}
return gets;
}
public int set(int index, long[] arr, int off, int len) {
assert len > 0 : "len must be > 0 (got " + len + ")";
assert index >= 0 && index < valueCount;
assert off + len <= arr.length;
final int sets = Math.min(valueCount - index, len);
for (int i = index, o = off, end = index + sets; i < end; ++i, ++o) {
values[i] = %sarr[o];
}
return sets;
}
@Override
public void fill(int fromIndex, int toIndex, long val) {
assert val == (val%s);
Arrays.fill(values, fromIndex, toIndex, %sval);
}
""" %(MASKS[bpv], CASTS[bpv], MASKS[bpv], CASTS[bpv]))
f.write("}\n")
f.close()

291
fine-lucene/src/main/java/com/fr/third/org/apache/lucene/util/packed/gen_Packed64SingleBlock.py
Unescape View File

@ -1,291 +0,0 @@
#! /usr/bin/env python
# Licensed to the Apache Software Foundation (ASF) under one or more
# contributor license agreements. See the NOTICE file distributed with
# this work for additional information regarding copyright ownership.
# The ASF licenses this file to You under the Apache License, Version 2.0
# (the "License"); you may not use this file except in compliance with
# the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
SUPPORTED_BITS_PER_VALUE = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 16, 21, 32]
HEADER="""// This file has been automatically generated, DO NOT EDIT
package com.fr.third.org.apache.lucene.util.packed;
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with this
* work for additional information regarding copyright ownership. The ASF
* licenses this file to You under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
* License for the specific language governing permissions and limitations under
* the License.
*/
import java.io.IOException;
import java.util.Arrays;
import com.fr.third.org.apache.lucene.store.DataInput;
import com.fr.third.org.apache.lucene.util.RamUsageEstimator;
/**
* This class is similar to {@link Packed64} except that it trades space for
* speed by ensuring that a single block needs to be read/written in order to
* read/write a value.
*/
abstract class Packed64SingleBlock extends PackedInts.MutableImpl {
public static final int MAX_SUPPORTED_BITS_PER_VALUE = %d;
private static final int[] SUPPORTED_BITS_PER_VALUE = new int[] {%s};
public static boolean isSupported(int bitsPerValue) {
return Arrays.binarySearch(SUPPORTED_BITS_PER_VALUE, bitsPerValue) >= 0;
}
private static int requiredCapacity(int valueCount, int valuesPerBlock) {
return valueCount / valuesPerBlock
+ (valueCount %% valuesPerBlock == 0 ? 0 : 1);
}
final long[] blocks;
Packed64SingleBlock(int valueCount, int bitsPerValue) {
super(valueCount, bitsPerValue);
assert isSupported(bitsPerValue);
final int valuesPerBlock = 64 / bitsPerValue;
blocks = new long[requiredCapacity(valueCount, valuesPerBlock)];
}
@Override
public void clear() {
Arrays.fill(blocks, 0L);
}
public long ramBytesUsed() {
return RamUsageEstimator.sizeOf(blocks);
}
@Override
public int get(int index, long[] arr, int off, int len) {
assert len > 0 : "len must be > 0 (got " + len + ")";
assert index >= 0 && index < valueCount;
len = Math.min(len, valueCount - index);
assert off + len <= arr.length;
final int originalIndex = index;
// go to the next block boundary
final int valuesPerBlock = 64 / bitsPerValue;
final int offsetInBlock = index %% valuesPerBlock;
if (offsetInBlock != 0) {
for (int i = offsetInBlock; i < valuesPerBlock && len > 0; ++i) {
arr[off++] = get(index++);
--len;
}
if (len == 0) {
return index - originalIndex;
}
}
// bulk get
assert index %% valuesPerBlock == 0;
final PackedInts.Decoder decoder = BulkOperation.of(PackedInts.Format.PACKED_SINGLE_BLOCK, bitsPerValue);
assert decoder.blockCount() == 1;
assert decoder.valueCount() == valuesPerBlock;
final int blockIndex = index / valuesPerBlock;
final int nblocks = (index + len) / valuesPerBlock - blockIndex;
decoder.decode(blocks, blockIndex, arr, off, nblocks);
final int diff = nblocks * valuesPerBlock;
index += diff; len -= diff;
if (index > originalIndex) {
// stay at the block boundary
return index - originalIndex;
} else {
// no progress so far => already at a block boundary but no full block to
// get
assert index == originalIndex;
return super.get(index, arr, off, len);
}
}
@Override
public int set(int index, long[] arr, int off, int len) {
assert len > 0 : "len must be > 0 (got " + len + ")";
assert index >= 0 && index < valueCount;
len = Math.min(len, valueCount - index);
assert off + len <= arr.length;
final int originalIndex = index;
// go to the next block boundary
final int valuesPerBlock = 64 / bitsPerValue;
final int offsetInBlock = index %% valuesPerBlock;
if (offsetInBlock != 0) {
for (int i = offsetInBlock; i < valuesPerBlock && len > 0; ++i) {
set(index++, arr[off++]);
--len;
}
if (len == 0) {
return index - originalIndex;
}
}
// bulk set
assert index %% valuesPerBlock == 0;
final BulkOperation op = BulkOperation.of(PackedInts.Format.PACKED_SINGLE_BLOCK, bitsPerValue);
assert op.blockCount() == 1;
assert op.valueCount() == valuesPerBlock;
final int blockIndex = index / valuesPerBlock;
final int nblocks = (index + len) / valuesPerBlock - blockIndex;
op.encode(arr, off, blocks, blockIndex, nblocks);
final int diff = nblocks * valuesPerBlock;
index += diff; len -= diff;
if (index > originalIndex) {
// stay at the block boundary
return index - originalIndex;
} else {
// no progress so far => already at a block boundary but no full block to
// set
assert index == originalIndex;
return super.set(index, arr, off, len);
}
}
@Override
public void fill(int fromIndex, int toIndex, long val) {
assert fromIndex >= 0;
assert fromIndex <= toIndex;
assert PackedInts.bitsRequired(val) <= bitsPerValue;
final int valuesPerBlock = 64 / bitsPerValue;
if (toIndex - fromIndex <= valuesPerBlock << 1) {
// there needs to be at least one full block to set for the block
// approach to be worth trying
super.fill(fromIndex, toIndex, val);
return;
}
// set values naively until the next block start
int fromOffsetInBlock = fromIndex %% valuesPerBlock;
if (fromOffsetInBlock != 0) {
for (int i = fromOffsetInBlock; i < valuesPerBlock; ++i) {
set(fromIndex++, val);
}
assert fromIndex %% valuesPerBlock == 0;
}
// bulk set of the inner blocks
final int fromBlock = fromIndex / valuesPerBlock;
final int toBlock = toIndex / valuesPerBlock;
assert fromBlock * valuesPerBlock == fromIndex;
long blockValue = 0L;
for (int i = 0; i < valuesPerBlock; ++i) {
blockValue = blockValue | (val << (i * bitsPerValue));
}
Arrays.fill(blocks, fromBlock, toBlock, blockValue);
// fill the gap
for (int i = valuesPerBlock * toBlock; i < toIndex; ++i) {
set(i, val);
}
}
@Override
protected PackedInts.Format getFormat() {
return PackedInts.Format.PACKED_SINGLE_BLOCK;
}
@Override
public String toString() {
return getClass().getSimpleName() + "(bitsPerValue=" + bitsPerValue
+ ", size=" + size() + ", elements.length=" + blocks.length + ")";
}
public static Packed64SingleBlock create(DataInput in,
int valueCount, int bitsPerValue) throws IOException {
Packed64SingleBlock reader = create(valueCount, bitsPerValue);
for (int i = 0; i < reader.blocks.length; ++i) {
reader.blocks[i] = in.readLong();
}
return reader;
}
""" %(SUPPORTED_BITS_PER_VALUE[-1], ", ".join(map(str, SUPPORTED_BITS_PER_VALUE)))
FOOTER = "}"
if __name__ == '__main__':
f = open("Packed64SingleBlock.java", 'w')
f.write(HEADER)
f.write(" public static Packed64SingleBlock create(int valueCount, int bitsPerValue) {\n")
f.write(" switch (bitsPerValue) {\n")
for bpv in SUPPORTED_BITS_PER_VALUE:
f.write(" case %d:\n" %bpv)
f.write(" return new Packed64SingleBlock%d(valueCount);\n" %bpv)
f.write(" default:\n")
f.write(" throw new IllegalArgumentException(\"Unsupported number of bits per value: \" + %d);\n" %bpv)
f.write(" }\n")
f.write(" }\n\n")
for bpv in SUPPORTED_BITS_PER_VALUE:
log_2 = 0
while (1 << log_2) < bpv:
log_2 = log_2 + 1
if (1 << log_2) != bpv:
log_2 = None
f.write(" static class Packed64SingleBlock%d extends Packed64SingleBlock {\n\n" %bpv)
f.write(" Packed64SingleBlock%d(int valueCount) {\n" %bpv)
f.write(" super(valueCount, %d);\n" %bpv)
f.write(" }\n\n")
f.write(" @Override\n")
f.write(" public long get(int index) {\n")
if log_2 is not None:
f.write(" final int o = index >>> %d;\n" %(6 - log_2))
f.write(" final int b = index & %d;\n" %((1 << (6 - log_2)) - 1))
f.write(" final int shift = b << %d;\n" %log_2)
else:
f.write(" final int o = index / %d;\n" %(64 / bpv))
f.write(" final int b = index %% %d;\n" %(64 / bpv))
f.write(" final int shift = b * %d;\n" %bpv)
f.write(" return (blocks[o] >>> shift) & %dL;\n" %((1 << bpv) - 1))
f.write(" }\n\n")
f.write(" @Override\n")
f.write(" public void set(int index, long value) {\n")
if log_2 is not None:
f.write(" final int o = index >>> %d;\n" %(6 - log_2))
f.write(" final int b = index & %d;\n" %((1 << (6 - log_2)) - 1))
f.write(" final int shift = b << %d;\n" %log_2)
else:
f.write(" final int o = index / %d;\n" %(64 / bpv))
f.write(" final int b = index %% %d;\n" %(64 / bpv))
f.write(" final int shift = b * %d;\n" %bpv)
f.write(" blocks[o] = (blocks[o] & ~(%dL << shift)) | (value << shift);\n" % ((1 << bpv) - 1))
f.write(" }\n\n")
f.write(" }\n\n")
f.write(FOOTER)
f.close()

161
fine-lucene/src/main/java/com/fr/third/org/apache/lucene/util/packed/gen_PackedThreeBlocks.py
Unescape View File

@ -1,161 +0,0 @@
#! /usr/bin/env python
# Licensed to the Apache Software Foundation (ASF) under one or more
# contributor license agreements. See the NOTICE file distributed with
# this work for additional information regarding copyright ownership.
# The ASF licenses this file to You under the Apache License, Version 2.0
# (the "License"); you may not use this file except in compliance with
# the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
HEADER="""// This file has been automatically generated, DO NOT EDIT
package com.fr.third.org.apache.lucene.util.packed;
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
import com.fr.third.org.apache.lucene.store.DataInput;
import com.fr.third.org.apache.lucene.util.RamUsageEstimator;
import java.io.IOException;
import java.util.Arrays;
"""
TYPES = {8: "byte", 16: "short"}
MASKS = {8: " & 0xFFL", 16: " & 0xFFFFL", 32: " & 0xFFFFFFFFL", 64: ""}
CASTS = {8: "(byte) ", 16: "(short) ", 32: "(int) ", 64: ""}
if __name__ == '__main__':
for bpv in TYPES.keys():
type
f = open("Packed%dThreeBlocks.java" %bpv, 'w')
f.write(HEADER)
f.write("""/**
* Packs integers into 3 %ss (%d bits per value).
* @lucene.internal
*/\n""" %(TYPES[bpv], bpv*3))
f.write("final class Packed%dThreeBlocks extends PackedInts.MutableImpl {\n" %bpv)
f.write(" final %s[] blocks;\n\n" %TYPES[bpv])
f.write(" public static final int MAX_SIZE = Integer.MAX_VALUE / 3;\n\n")
f.write(" Packed%dThreeBlocks(int valueCount) {\n" %bpv)
f.write(" super(valueCount, %d);\n" %(bpv*3))
f.write(" if (valueCount > MAX_SIZE) {\n")
f.write(" throw new ArrayIndexOutOfBoundsException(\"MAX_SIZE exceeded\");\n")
f.write(" }\n")
f.write(" blocks = new %s[valueCount * 3];\n" %TYPES[bpv])
f.write(" }\n\n")
f.write(" Packed%dThreeBlocks(DataInput in, int valueCount) throws IOException {\n" %bpv)
f.write(" this(valueCount);\n")
f.write(" for (int i = 0; i < 3 * valueCount; ++i) {\n")
f.write(" blocks[i] = in.read%s();\n" %TYPES[bpv].title())
f.write(" }\n")
f.write(" final int mod = blocks.length %% %d;\n" %(64 / bpv))
f.write(" if (mod != 0) {\n")
f.write(" for (int i = mod; i < %d; ++i) {\n" %(64 / bpv))
f.write(" in.read%s();\n" %TYPES[bpv].title())
f.write(" }\n")
f.write(" }\n")
f.write(" }\n")
f.write("""
@Override
public long get(int index) {
final int o = index * 3;
return (blocks[o]%s) << %d | (blocks[o+1]%s) << %d | (blocks[o+2]%s);
}
@Override
public int get(int index, long[] arr, int off, int len) {
assert len > 0 : "len must be > 0 (got " + len + ")";
assert index >= 0 && index < valueCount;
assert off + len <= arr.length;
final int gets = Math.min(valueCount - index, len);
for (int i = index * 3, end = (index + gets) * 3; i < end; i+=3) {
arr[off++] = (blocks[i]%s) << %d | (blocks[i+1]%s) << %d | (blocks[i+2]%s);
}
return gets;
}
@Override
public void set(int index, long value) {
final int o = index * 3;
blocks[o] = %s(value >>> %d);
blocks[o+1] = %s(value >>> %d);
blocks[o+2] = %svalue;
}
@Override
public int set(int index, long[] arr, int off, int len) {
assert len > 0 : "len must be > 0 (got " + len + ")";
assert index >= 0 && index < valueCount;
assert off + len <= arr.length;
final int sets = Math.min(valueCount - index, len);
for (int i = off, o = index * 3, end = off + sets; i < end; ++i) {
final long value = arr[i];
blocks[o++] = %s(value >>> %d);
blocks[o++] = %s(value >>> %d);
blocks[o++] = %svalue;
}
return sets;
}
@Override
public void fill(int fromIndex, int toIndex, long val) {
final %s block1 = %s(val >>> %d);
final %s block2 = %s(val >>> %d);
final %s block3 = %sval;
for (int i = fromIndex * 3, end = toIndex * 3; i < end; i += 3) {
blocks[i] = block1;
blocks[i+1] = block2;
blocks[i+2] = block3;
}
}
@Override
public void clear() {
Arrays.fill(blocks, %s0);
}
public long ramBytesUsed() {
return RamUsageEstimator.sizeOf(blocks);
}
@Override
public String toString() {
return getClass().getSimpleName() + "(bitsPerValue=" + bitsPerValue
+ ", size=" + size() + ", elements.length=" + blocks.length + ")";
}
}
""" %(MASKS[bpv], 2*bpv, MASKS[bpv], bpv, MASKS[bpv], MASKS[bpv], 2*bpv, MASKS[bpv], bpv, MASKS[bpv], CASTS[bpv], 2*bpv, CASTS[bpv], bpv, CASTS[bpv], CASTS[bpv],
2*bpv, CASTS[bpv], bpv, CASTS[bpv], TYPES[bpv], CASTS[bpv], 2*bpv, TYPES[bpv],
CASTS[bpv], bpv, TYPES[bpv], CASTS[bpv], CASTS[bpv]))
f.close()
Write Preview
Loading…
Cancel
Save