change "simple" analyzer to use "letter" tokenizer

this change improves compatibility with the simple analyzer
defined by Lucene.  this has important implications for
some perf tests as well as they often use the simple
analyzer.

analysis/analyzers/simple_analyzer/simple_analyzer.go

@@ -12,14 +12,14 @@ package simple_analyzer
 import (
 	"github.com/blevesearch/bleve/analysis"
 	"github.com/blevesearch/bleve/analysis/token_filters/lower_case_filter"
-	"github.com/blevesearch/bleve/analysis/tokenizers/unicode"
+	"github.com/blevesearch/bleve/analysis/tokenizers/letter"
 	"github.com/blevesearch/bleve/registry"
 )
 
 const Name = "simple"
 
 func AnalyzerConstructor(config map[string]interface{}, cache *registry.Cache) (*analysis.Analyzer, error) {
-	tokenizer, err := cache.TokenizerNamed(unicode.Name)
+	tokenizer, err := cache.TokenizerNamed(letter.Name)
 	if err != nil {
 		return nil, err
 	}

analysis/tokenizers/character/character.go

@@ -0,0 +1,71 @@
+//  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 character
+
+import (
+	"unicode/utf8"
+
+	"github.com/blevesearch/bleve/analysis"
+)
+
+type IsTokenRune func(r rune) bool
+
+type CharacterTokenizer struct {
+	isTokenRun IsTokenRune
+}
+
+func NewCharacterTokenizer(f IsTokenRune) *CharacterTokenizer {
+	return &CharacterTokenizer{
+		isTokenRun: f,
+	}
+}
+
+func (c *CharacterTokenizer) Tokenize(input []byte) analysis.TokenStream {
+
+	rv := make(analysis.TokenStream, 0, 1024)
+
+	offset := 0
+	start := 0
+	end := 0
+	count := 0
+	for currRune, size := utf8.DecodeRune(input[offset:]); currRune != utf8.RuneError; currRune, size = utf8.DecodeRune(input[offset:]) {
+		isToken := c.isTokenRun(currRune)
+		if isToken {
+			end = offset + size
+		} else {
+			if end-start > 0 {
+				// build token
+				rv = append(rv, &analysis.Token{
+					Term:     input[start:end],
+					Start:    start,
+					End:      end,
+					Position: count + 1,
+					Type:     analysis.AlphaNumeric,
+				})
+				count++
+			}
+			start = offset + size
+			end = start
+		}
+		offset += size
+	}
+	// if we ended in the middle of a token, finish it
+	if end-start > 0 {
+		// build token
+		rv = append(rv, &analysis.Token{
+			Term:     input[start:end],
+			Start:    start,
+			End:      end,
+			Position: count + 1,
+			Type:     analysis.AlphaNumeric,
+		})
+	}
+	return rv
+}

analysis/tokenizers/character/character_test.go

@@ -0,0 +1,79 @@
+//  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 character
+
+import (
+	"reflect"
+	"testing"
+	"unicode"
+
+	"github.com/blevesearch/bleve/analysis"
+)
+
+func TestCharacterTokenizer(t *testing.T) {
+	tests := []struct {
+		input  []byte
+		output analysis.TokenStream
+	}{
+		{
+			[]byte("Hello World."),
+			analysis.TokenStream{
+				{
+					Start:    0,
+					End:      5,
+					Term:     []byte("Hello"),
+					Position: 1,
+					Type:     analysis.AlphaNumeric,
+				},
+				{
+					Start:    6,
+					End:      11,
+					Term:     []byte("World"),
+					Position: 2,
+					Type:     analysis.AlphaNumeric,
+				},
+			},
+		},
+		{
+			[]byte("dominique@mcdiabetes.com"),
+			analysis.TokenStream{
+				{
+					Start:    0,
+					End:      9,
+					Term:     []byte("dominique"),
+					Position: 1,
+					Type:     analysis.AlphaNumeric,
+				},
+				{
+					Start:    10,
+					End:      20,
+					Term:     []byte("mcdiabetes"),
+					Position: 2,
+					Type:     analysis.AlphaNumeric,
+				},
+				{
+					Start:    21,
+					End:      24,
+					Term:     []byte("com"),
+					Position: 3,
+					Type:     analysis.AlphaNumeric,
+				},
+			},
+		},
+	}
+
+	tokenizer := NewCharacterTokenizer(unicode.IsLetter)
+	for _, test := range tests {
+		actual := tokenizer.Tokenize(test.input)
+		if !reflect.DeepEqual(actual, test.output) {
+			t.Errorf("Expected %v, got %v for %s", test.output, actual, string(test.input))
+		}
+	}
+}

analysis/tokenizers/letter/letter.go

@@ -0,0 +1,28 @@
+//  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 letter
+
+import (
+	"unicode"
+
+	"github.com/blevesearch/bleve/analysis"
+	"github.com/blevesearch/bleve/analysis/tokenizers/character"
+	"github.com/blevesearch/bleve/registry"
+)
+
+const Name = "letter"
+
+func TokenizerConstructor(config map[string]interface{}, cache *registry.Cache) (analysis.Tokenizer, error) {
+	return character.NewCharacterTokenizer(unicode.IsLetter), nil
+}
+
+func init() {
+	registry.RegisterTokenizer(Name, TokenizerConstructor)
+}