You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@lucene.apache.org by "Michael Gibney (Jira)" <ji...@apache.org> on 2019/09/20 20:04:00 UTC
[jira] [Comment Edited] (LUCENE-8972) CharFilter version of
ICUTransformFilter, to better support dictionary-based tokenization
[ https://issues.apache.org/jira/browse/LUCENE-8972?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16934721#comment-16934721 ]
Michael Gibney edited comment on LUCENE-8972 at 9/20/19 8:03 PM:
-----------------------------------------------------------------
I have pushed [PR #892|https://github.com/apache/lucene-solr/pull/892], with the proposed new classes, tests, and docs. Initially I've mostly just used modified versions of the tests for {{ICUTransformFilter*}} ... (btw, {{testRandomStrings()}} is great!).
Most of the code complexity is due to the need to incrementally process one input character at a time in order to get offset correction as accurate as possible, and implement "rollback" (following the same approach as ICU Transliterator code does internally, but not exposed via public API).
The following discusses "rollback" in a little more depth, including some of the performance implications and an idea for future performance improvement:
Regarding "rollback", see comments "To understand the need for rollback" [in source code|https://github.com/unicode-org/icu/blob/a075ac9c/icu4j/main/classes/translit/src/com/ibm/icu/text/Transliterator.java#L1137] for private method {{Transliterator#filteredTransliterate(Replaceable, Position, boolean, boolean)}}. [CompoundTransliterator|https://github.com/unicode-org/icu/blob/a075ac9c/icu4j/main/classes/translit/src/com/ibm/icu/text/CompoundTransliterator.java]'s compliance with the extant top-level Transliterator abstraction here induces some serious performance hits (for some not-uncommon cases, like trailing NFC in the "Cyrillic-Latin" transliteration, shifting character blocks around on _every_ incremental character insertion. FWIW, "incremental character insertion and rollback" is essentially how ICU handles this situation in the source code referenced above).
For future consideration (absent a change in the ICU API) I'm thinking that it might be possible to reimplement the essence of CompoundTransliterator in external (Lucene) application code, with separately tracked "position" for each "leaf" Transliterator in the Transliterator tree. This would allow positions that were blocked partway through depth-first traveral of the Transliterator tree to avoid:
# being double-processing by (potentially not idempotent) leading Transliterators, and/or
# bypassing trailing Transliterators on account of higher-level filters that block the partially-processed character
My sense is that the performance gain could be significant.
was (Author: mgibney):
I have pushed [PR #892|https://github.com/apache/lucene-solr/pull/892], with the proposed new classes, tests, and docs. Initially I've mostly just used modified versions of the tests for {{ICUTransformFilter*}} ... (btw, {{testRandomStrings()}} is great!).
Most of the code complexity is due to the need to incrementally process one input character at a time in order get offset correction as accurate as possible, and implement "rollback" (following the same approach as ICU Transliterator code does internally, but not exposed via public API).
The following discusses "rollback" in a little more depth, including some of the performance implications and an idea for future performance improvement:
Regarding "rollback", see comments "To understand the need for rollback" [in source code|https://github.com/unicode-org/icu/blob/a075ac9c/icu4j/main/classes/translit/src/com/ibm/icu/text/Transliterator.java#L1137] for private method {{Transliterator#filteredTransliterate(Replaceable, Position, boolean, boolean)}}. [CompoundTransliterator|https://github.com/unicode-org/icu/blob/a075ac9c/icu4j/main/classes/translit/src/com/ibm/icu/text/CompoundTransliterator.java]'s compliance with the extant top-level Transliterator abstraction here induces some serious performance hits (for some not-uncommon cases, like trailing NFC in the "Cyrillic-Latin" transliteration, shifting character blocks around on _every_ incremental character insertion. FWIW, "incremental character insertion and rollback" is essentially how ICU handles this situation in the source code referenced above).
For future consideration (absent a change in the ICU API) I'm thinking that it might be possible to reimplement the essence of CompoundTransliterator in external (Lucene) application code, with separately tracked "position" for each "leaf" Transliterator in the Transliterator tree. This would allow positions that were blocked partway through depth-first traveral of the Transliterator tree to avoid:
# being double-processing by (potentially not idempotent) leading Transliterators, and/or
# bypassing trailing Transliterators on account of higher-level filters that block the partially-processed character
My sense is that the performance gain could be significant.
> CharFilter version of ICUTransformFilter, to better support dictionary-based tokenization
> -----------------------------------------------------------------------------------------
>
> Key: LUCENE-8972
> URL: https://issues.apache.org/jira/browse/LUCENE-8972
> Project: Lucene - Core
> Issue Type: New Feature
> Components: modules/analysis
> Affects Versions: master (9.0), 8.2
> Reporter: Michael Gibney
> Priority: Minor
> Time Spent: 10m
> Remaining Estimate: 0h
>
> The ICU Transliteration API is currently exposed through Lucene only post-tokinzer, via ICUTransformFilter. Some tokenizers (particularly dictionary-based) may assume pre-normalized input (e.g., for Chinese characters, there may be an assumption of traditional-only or simplified-only input characters, at the level of either all input, or per-dictionary-defined-token).
> The potential usefulness of a CharFilter that exposes the ICU Transliteration API was suggested in a [thread on the Solr mailing list|https://mail-archives.apache.org/mod_mbox/lucene-solr-user/201807.mbox/%3C4DAB7BA7-42A8-4009-8B49-60822B00DE7D%40wunderwood.org%3E], and my hope is that this issue can facilitate more detailed discussion of the proposed addition.
> A concrete example of mixed traditional/simplified characters that are currently tokenized differently by the ICUTokenizer are:
> * 红楼梦 (SSS)
> * 紅樓夢 (TTT)
> * 紅楼夢 (TST)
> The first two tokens (simplified-only and traditional-only, respectively) are included in the [CJ dictionary that backs ICUTokenizer|https://raw.githubusercontent.com/unicode-org/icu/release-62-1/icu4c/source/data/brkitr/dictionaries/cjdict.txt], but the last (a mixture of traditional and simplified characters) is not, and is not recognized as a token. Even _if_ we assume this to be an intentional omission from the dictionary that results in behavior that could be desirable for some use cases, there are surely some use cases that would benefit from a more permissive dictionary-based tokenization strategy (such as could be supported by pre-tokenizer transliteration).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org