From 2594693a46e701180c2bd8505d8b84986fad927d Mon Sep 17 00:00:00 2001 From: guillep2k Date: Tue, 6 Aug 2019 01:38:16 -0300 Subject: [PATCH] Remove camel case tokenization from repo indexer (#7733) * Make repository indexer tokenize by camel case selectable * Revert "Make repository indexer tokenize by camel case selectable" This reverts commit a8d4e40af56ffeab9c040a13e01b9e0d5487bd3b. * Remove app.ini setting, permanently remove camel case tokenizer * Increase repoIndexerLatestVersion to force indexes to be rebuilt * Changes introduced by the commmand make vendor --- modules/indexer/repo.go | 5 +- .../analysis/token/camelcase/camelcase.go | 78 ------------- .../bleve/analysis/token/camelcase/parser.go | 109 ------------------ .../bleve/analysis/token/camelcase/states.go | 87 -------------- vendor/modules.txt | 1 - 5 files changed, 2 insertions(+), 278 deletions(-) delete mode 100644 vendor/github.com/blevesearch/bleve/analysis/token/camelcase/camelcase.go delete mode 100644 vendor/github.com/blevesearch/bleve/analysis/token/camelcase/parser.go delete mode 100644 vendor/github.com/blevesearch/bleve/analysis/token/camelcase/states.go diff --git a/modules/indexer/repo.go b/modules/indexer/repo.go index 287d23854..787f9a946 100644 --- a/modules/indexer/repo.go +++ b/modules/indexer/repo.go @@ -12,7 +12,6 @@ import ( "github.com/blevesearch/bleve" "github.com/blevesearch/bleve/analysis/analyzer/custom" - "github.com/blevesearch/bleve/analysis/token/camelcase" "github.com/blevesearch/bleve/analysis/token/lowercase" "github.com/blevesearch/bleve/analysis/token/unique" "github.com/blevesearch/bleve/analysis/tokenizer/unicode" @@ -24,7 +23,7 @@ const ( repoIndexerAnalyzer = "repoIndexerAnalyzer" repoIndexerDocType = "repoIndexerDocType" - repoIndexerLatestVersion = 1 + repoIndexerLatestVersion = 2 ) // repoIndexer (thread-safe) index for repository contents @@ -111,7 +110,7 @@ func createRepoIndexer(path string, latestVersion int) error { "type": custom.Name, "char_filters": []string{}, "tokenizer": unicode.Name, - "token_filters": []string{unicodeNormalizeName, camelcase.Name, lowercase.Name, unique.Name}, + "token_filters": []string{unicodeNormalizeName, lowercase.Name, unique.Name}, }); err != nil { return err } diff --git a/vendor/github.com/blevesearch/bleve/analysis/token/camelcase/camelcase.go b/vendor/github.com/blevesearch/bleve/analysis/token/camelcase/camelcase.go deleted file mode 100644 index 42b7483c6..000000000 --- a/vendor/github.com/blevesearch/bleve/analysis/token/camelcase/camelcase.go +++ /dev/null @@ -1,78 +0,0 @@ -// Copyright (c) 2016 Couchbase, Inc. -// -// Licensed 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. - -package camelcase - -import ( - "bytes" - "unicode/utf8" - - "github.com/blevesearch/bleve/analysis" - "github.com/blevesearch/bleve/registry" -) - -const Name = "camelCase" - -// CamelCaseFilter splits a given token into a set of tokens where each resulting token -// falls into one the following classes: -// 1) Upper case followed by lower case letters. -// Terminated by a number, an upper case letter, and a non alpha-numeric symbol. -// 2) Upper case followed by upper case letters. -// Terminated by a number, an upper case followed by a lower case letter, and a non alpha-numeric symbol. -// 3) Lower case followed by lower case letters. -// Terminated by a number, an upper case letter, and a non alpha-numeric symbol. -// 4) Number followed by numbers. -// Terminated by a letter, and a non alpha-numeric symbol. -// 5) Non alpha-numeric symbol followed by non alpha-numeric symbols. -// Terminated by a number, and a letter. -// -// It does a one-time sequential pass over an input token, from left to right. -// The scan is greedy and generates the longest substring that fits into one of the classes. -// -// See the test file for examples of classes and their parsings. -type CamelCaseFilter struct{} - -func NewCamelCaseFilter() *CamelCaseFilter { - return &CamelCaseFilter{} -} - -func (f *CamelCaseFilter) Filter(input analysis.TokenStream) analysis.TokenStream { - rv := make(analysis.TokenStream, 0, len(input)) - - nextPosition := 1 - for _, token := range input { - runeCount := utf8.RuneCount(token.Term) - runes := bytes.Runes(token.Term) - - p := NewParser(runeCount, nextPosition, token.Start) - for i := 0; i < runeCount; i++ { - if i+1 >= runeCount { - p.Push(runes[i], nil) - } else { - p.Push(runes[i], &runes[i+1]) - } - } - rv = append(rv, p.FlushTokens()...) - nextPosition = p.NextPosition() - } - return rv -} - -func CamelCaseFilterConstructor(config map[string]interface{}, cache *registry.Cache) (analysis.TokenFilter, error) { - return NewCamelCaseFilter(), nil -} - -func init() { - registry.RegisterTokenFilter(Name, CamelCaseFilterConstructor) -} diff --git a/vendor/github.com/blevesearch/bleve/analysis/token/camelcase/parser.go b/vendor/github.com/blevesearch/bleve/analysis/token/camelcase/parser.go deleted file mode 100644 index ff4ce2fea..000000000 --- a/vendor/github.com/blevesearch/bleve/analysis/token/camelcase/parser.go +++ /dev/null @@ -1,109 +0,0 @@ -// Copyright (c) 2016 Couchbase, Inc. -// -// Licensed 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. - -package camelcase - -import ( - "github.com/blevesearch/bleve/analysis" -) - -func (p *Parser) buildTokenFromTerm(buffer []rune) *analysis.Token { - term := analysis.BuildTermFromRunes(buffer) - token := &analysis.Token{ - Term: term, - Position: p.position, - Start: p.index, - End: p.index + len(term), - } - p.position++ - p.index += len(term) - return token -} - -// Parser accepts a symbol and passes it to the current state (representing a class). -// The state can accept it (and accumulate it). Otherwise, the parser creates a new state that -// starts with the pushed symbol. -// -// Parser accumulates a new resulting token every time it switches state. -// Use FlushTokens() to get the results after the last symbol was pushed. -type Parser struct { - bufferLen int - buffer []rune - current State - tokens []*analysis.Token - position int - index int -} - -func NewParser(length, position, index int) *Parser { - return &Parser{ - bufferLen: length, - buffer: make([]rune, 0, length), - tokens: make([]*analysis.Token, 0, length), - position: position, - index: index, - } -} - -func (p *Parser) Push(sym rune, peek *rune) { - if p.current == nil { - // the start of parsing - p.current = p.NewState(sym) - p.buffer = append(p.buffer, sym) - - } else if p.current.Member(sym, peek) { - // same state, just accumulate - p.buffer = append(p.buffer, sym) - - } else { - // the old state is no more, thus convert the buffer - p.tokens = append(p.tokens, p.buildTokenFromTerm(p.buffer)) - - // let the new state begin - p.current = p.NewState(sym) - p.buffer = make([]rune, 0, p.bufferLen) - p.buffer = append(p.buffer, sym) - } -} - -// Note. States have to have different starting symbols. -func (p *Parser) NewState(sym rune) State { - var found State - - found = &LowerCaseState{} - if found.StartSym(sym) { - return found - } - - found = &UpperCaseState{} - if found.StartSym(sym) { - return found - } - - found = &NumberCaseState{} - if found.StartSym(sym) { - return found - } - - return &NonAlphaNumericCaseState{} -} - -func (p *Parser) FlushTokens() []*analysis.Token { - p.tokens = append(p.tokens, p.buildTokenFromTerm(p.buffer)) - return p.tokens -} - -func (p *Parser) NextPosition() int { - return p.position -} diff --git a/vendor/github.com/blevesearch/bleve/analysis/token/camelcase/states.go b/vendor/github.com/blevesearch/bleve/analysis/token/camelcase/states.go deleted file mode 100644 index 97e08d044..000000000 --- a/vendor/github.com/blevesearch/bleve/analysis/token/camelcase/states.go +++ /dev/null @@ -1,87 +0,0 @@ -// Copyright (c) 2016 Couchbase, Inc. -// -// Licensed 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. - -package camelcase - -import ( - "unicode" -) - -// States codify the classes that the parser recognizes. -type State interface { - // is _sym_ the start character - StartSym(sym rune) bool - - // is _sym_ a member of a class. - // peek, the next sym on the tape, can also be used to determine a class. - Member(sym rune, peek *rune) bool -} - -type LowerCaseState struct{} - -func (s *LowerCaseState) Member(sym rune, peek *rune) bool { - return unicode.IsLower(sym) -} - -func (s *LowerCaseState) StartSym(sym rune) bool { - return s.Member(sym, nil) -} - -type UpperCaseState struct { - startedCollecting bool // denotes that the start character has been read - collectingUpper bool // denotes if this is a class of all upper case letters -} - -func (s *UpperCaseState) Member(sym rune, peek *rune) bool { - if !(unicode.IsLower(sym) || unicode.IsUpper(sym)) { - return false - } - - if peek != nil && unicode.IsUpper(sym) && unicode.IsLower(*peek) { - return false - } - - if !s.startedCollecting { - // now we have to determine if upper-case letters are collected. - s.startedCollecting = true - s.collectingUpper = unicode.IsUpper(sym) - return true - } - - return s.collectingUpper == unicode.IsUpper(sym) -} - -func (s *UpperCaseState) StartSym(sym rune) bool { - return unicode.IsUpper(sym) -} - -type NumberCaseState struct{} - -func (s *NumberCaseState) Member(sym rune, peek *rune) bool { - return unicode.IsNumber(sym) -} - -func (s *NumberCaseState) StartSym(sym rune) bool { - return s.Member(sym, nil) -} - -type NonAlphaNumericCaseState struct{} - -func (s *NonAlphaNumericCaseState) Member(sym rune, peek *rune) bool { - return !unicode.IsLower(sym) && !unicode.IsUpper(sym) && !unicode.IsNumber(sym) -} - -func (s *NonAlphaNumericCaseState) StartSym(sym rune) bool { - return s.Member(sym, nil) -} diff --git a/vendor/modules.txt b/vendor/modules.txt index 187d128d7..703f161a8 100644 --- a/vendor/modules.txt +++ b/vendor/modules.txt @@ -24,7 +24,6 @@ github.com/beorn7/perks/quantile # github.com/blevesearch/bleve v0.0.0-20190214220507-05d86ea8f6e3 github.com/blevesearch/bleve github.com/blevesearch/bleve/analysis/analyzer/custom -github.com/blevesearch/bleve/analysis/token/camelcase github.com/blevesearch/bleve/analysis/token/lowercase github.com/blevesearch/bleve/analysis/token/unicodenorm github.com/blevesearch/bleve/analysis/token/unique