Skip to content

A Q&A platform software for teams at any scales. Whether it's a community forum, help center, or knowledge management platform, you can always count on Apache Answer.

feat(search): highlight search results

Octobug5e477482982c
diff --git a/internal/repo/search_common/search_repo.go b/internal/repo/search_common/search_repo.go
index 0443bfe9..cd7de696 100644
--- a/internal/repo/search_common/search_repo.go
+++ b/internal/repo/search_common/search_repo.go
@@ -22,12 +22,13 @@ package search_common
import (
"context"
"fmt"
- tagcommon "github.com/apache/incubator-answer/internal/service/tag_common"
- "github.com/apache/incubator-answer/plugin"
"strconv"
"strings"
"time"
+ tagcommon "github.com/apache/incubator-answer/internal/service/tag_common"
+ "github.com/apache/incubator-answer/plugin"
+
"github.com/apache/incubator-answer/pkg/htmltext"
"github.com/apache/incubator-answer/internal/base/data"
@@ -234,7 +235,7 @@ func (sr *searchRepo) SearchContents(ctx context.Context, words []string, tagIDs
err = errors.InternalServer(reason.DatabaseError).WithError(err).WithStack()
return
} else {
- resp, err = sr.parseResult(ctx, res)
+ resp, err = sr.parseResult(ctx, res, words)
return
}
}
@@ -342,7 +343,7 @@ func (sr *searchRepo) SearchQuestions(ctx context.Context, words []string, tagID
if len(tr) != 0 {
total = converter.StringToInt64(string(tr[0]["total"]))
}
- resp, err = sr.parseResult(ctx, res)
+ resp, err = sr.parseResult(ctx, res, words)
if err != nil {
err = errors.InternalServer(reason.DatabaseError).WithError(err).WithStack()
}
@@ -436,7 +437,7 @@ func (sr *searchRepo) SearchAnswers(ctx context.Context, words []string, tagIDs
}
total = converter.StringToInt64(string(tr[0]["total"]))
- resp, err = sr.parseResult(ctx, res)
+ resp, err = sr.parseResult(ctx, res, words)
if err != nil {
err = errors.InternalServer(reason.DatabaseError).WithError(err).WithStack()
}
@@ -460,7 +461,7 @@ func (sr *searchRepo) parseOrder(ctx context.Context, order string) (res string)
}
// ParseSearchPluginResult parse search plugin result
-func (sr *searchRepo) ParseSearchPluginResult(ctx context.Context, sres []plugin.SearchResult) (resp []*schema.SearchResult, err error) {
+func (sr *searchRepo) ParseSearchPluginResult(ctx context.Context, sres []plugin.SearchResult, words []string) (resp []*schema.SearchResult, err error) {
var (
qres []map[string][]byte
res = make([]map[string][]byte, 0)
@@ -483,11 +484,11 @@ func (sr *searchRepo) ParseSearchPluginResult(ctx context.Context, sres []plugin
}
res = append(res, qres[0])
}
- return sr.parseResult(ctx, res)
+ return sr.parseResult(ctx, res, words)
}
// parseResult parse search result, return the data structure
-func (sr *searchRepo) parseResult(ctx context.Context, res []map[string][]byte) (resp []*schema.SearchResult, err error) {
+func (sr *searchRepo) parseResult(ctx context.Context, res []map[string][]byte, words []string) (resp []*schema.SearchResult, err error) {
questionIDs := make([]string, 0)
userIDs := make([]string, 0)
resultList := make([]*schema.SearchResult, 0)
@@ -508,7 +509,7 @@ func (sr *searchRepo) parseResult(ctx context.Context, res []map[string][]byte)
QuestionID: QuestionID,
Title: string(r["title"]),
UrlTitle: htmltext.UrlTitle(string(r["title"])),
- Excerpt: htmltext.FetchExcerpt(string(r["parsed_text"]), "...", 240),
+ Excerpt: htmltext.FetchMatchedExcerpt(string(r["parsed_text"]), words, "...", 100),
CreatedAtParsed: tp.Unix(),
UserInfo: &schema.SearchObjectUser{
ID: string(r["user_id"]),
diff --git a/internal/service/search_common/search.go b/internal/service/search_common/search.go
index c19fdc49..ba6f6295 100644
--- a/internal/service/search_common/search.go
+++ b/internal/service/search_common/search.go
@@ -21,6 +21,7 @@ package search_common
import (
"context"
+
"github.com/apache/incubator-answer/internal/schema"
"github.com/apache/incubator-answer/plugin"
)
@@ -29,5 +30,5 @@ type SearchRepo interface {
SearchContents(ctx context.Context, words []string, tagIDs [][]string, userID string, votes, page, size int, order string) (resp []*schema.SearchResult, total int64, err error)
SearchQuestions(ctx context.Context, words []string, tagIDs [][]string, notAccepted bool, views, answers int, page, size int, order string) (resp []*schema.SearchResult, total int64, err error)
SearchAnswers(ctx context.Context, words []string, tagIDs [][]string, accepted bool, questionID string, page, size int, order string) (resp []*schema.SearchResult, total int64, err error)
- ParseSearchPluginResult(ctx context.Context, sres []plugin.SearchResult) (resp []*schema.SearchResult, err error)
+ ParseSearchPluginResult(ctx context.Context, sres []plugin.SearchResult, words []string) (resp []*schema.SearchResult, err error)
}
diff --git a/internal/service/search_service.go b/internal/service/search_service.go
index 46d71374..430fa6cf 100644
--- a/internal/service/search_service.go
+++ b/internal/service/search_service.go
@@ -21,6 +21,7 @@ package service
import (
"context"
+
"github.com/apache/incubator-answer/internal/schema"
"github.com/apache/incubator-answer/internal/service/search_common"
"github.com/apache/incubator-answer/internal/service/search_parser"
@@ -93,6 +94,6 @@ func (ss *SearchService) searchByPlugin(ctx context.Context, finder plugin.Searc
res, resp.Total, err = finder.SearchAnswers(ctx, cond.Convert2PluginSearchCond(dto.Page, dto.Size, dto.Order))
}
- resp.SearchResults, err = ss.searchRepo.ParseSearchPluginResult(ctx, res)
+ resp.SearchResults, err = ss.searchRepo.ParseSearchPluginResult(ctx, res, cond.Words)
return resp, err
}
diff --git a/pkg/htmltext/htmltext.go b/pkg/htmltext/htmltext.go
index 792063f4..df8866ca 100644
--- a/pkg/htmltext/htmltext.go
+++ b/pkg/htmltext/htmltext.go
@@ -20,18 +20,36 @@
package htmltext
import (
- "github.com/Chain-Zhang/pinyin"
- "github.com/apache/incubator-answer/pkg/checker"
"io"
"net/http"
"net/url"
"regexp"
"strings"
+ "unicode/utf8"
+ "github.com/Chain-Zhang/pinyin"
"github.com/Machiel/slugify"
strip "github.com/grokify/html-strip-tags-go"
+
+ "github.com/apache/incubator-answer/pkg/checker"
+ "github.com/apache/incubator-answer/pkg/converter"
)
+// min() and max() can be removed starting from Go1.21
+func max(a, b int) int {
+ if a > b {
+ return a
+ }
+ return b
+}
+
+func min(a, b int) int {
+ if a < b {
+ return a
+ }
+ return b
+}
+
// ClearText clear HTML, get the clear text
func ClearText(html string) (text string) {
if len(html) == 0 {
@@ -123,6 +141,82 @@ func FetchExcerpt(html, trimMarker string, limit int) (text string) {
return
}
+// findFirstMatchedWord returns the first matched word and its index
+func findFirstMatchedWord(text string, words []string) (string, int) {
+ if len(text) == 0 || len(words) == 0 {
+ return "", 0
+ }
+
+ words = converter.UniqueArray(words)
+ firstWord := ""
+ firstIndex := len(text)
+
+ for _, word := range words {
+ if idx := strings.Index(text, word); idx != -1 && idx < firstIndex {
+ firstIndex = idx
+ firstWord = word
+ }
+ }
+
+ if firstIndex != len(text) {
+ return firstWord, firstIndex
+ }
+
+ return "", 0
+}
+
+// getRuneRange returns the valid begin and end indexes of the runeText
+func getRuneRange(runeText []rune, offset, limit int) (begin, end int) {
+ runeLen := len(runeText)
+
+ limit = min(runeLen, max(0, limit))
+ begin = min(runeLen, max(0, offset))
+ end = min(runeLen, begin+limit)
+
+ return
+}
+
+// FetchRangedExcerpt returns a ranged excerpt from the HTML string.
+// Note: offset is a rune index, not a byte index
+func FetchRangedExcerpt(html, trimMarker string, offset int, limit int) (text string) {
+ if len(html) == 0 {
+ text = html
+ return
+ }
+
+ runeText := []rune(ClearText(html))
+ begin, end := getRuneRange(runeText, offset, limit)
+ text = string(runeText[begin:end])
+
+ if begin > 0 {
+ text = trimMarker + text
+ }
+ if end < len(runeText) {
+ text = text + trimMarker
+ }
+
+ return
+}
+
+// FetchMatchedExcerpt returns the matched excerpt according to the words
+func FetchMatchedExcerpt(html string, words []string, trimMarker string, trimLength int) string {
+ text := ClearText(html)
+ matchedWord, matchedIndex := findFirstMatchedWord(text, words)
+ runeIndex := utf8.RuneCountInString(text[0:matchedIndex])
+
+ trimLength = max(0, trimLength)
+ runeOffset := runeIndex - trimLength
+ runeLimit := trimLength + trimLength + utf8.RuneCountInString(matchedWord)
+
+ textRuneCount := utf8.RuneCountInString(text)
+ if runeOffset+runeLimit > textRuneCount {
+ // Reserved extra chars before the matched word
+ runeOffset = textRuneCount - runeLimit
+ }
+
+ return FetchRangedExcerpt(html, trimMarker, runeOffset, runeLimit)
+}
+
func GetPicByUrl(Url string) string {
res, err := http.Get(Url)
if err != nil {
diff --git a/pkg/htmltext/htmltext_test.go b/pkg/htmltext/htmltext_test.go
index 4c520d7a..d549d887 100644
--- a/pkg/htmltext/htmltext_test.go
+++ b/pkg/htmltext/htmltext_test.go
@@ -85,3 +85,122 @@ func TestUrlTitle(t *testing.T) {
fmt.Println(formatTitle)
}
}
+
+func TestFindFirstMatchedWord(t *testing.T) {
+ var (
+ expectedWord,
+ actualWord string
+ expectedIndex,
+ actualIndex int
+ )
+
+ text := "Hello, I have 中文 and 😂 and I am supposed to work fine."
+
+ // test find nothing
+ expectedWord, expectedIndex = "", 0
+ actualWord, actualIndex = findFirstMatchedWord(text, []string{"youcantfindme"})
+ assert.Equal(t, expectedWord, actualWord)
+ assert.Equal(t, expectedIndex, actualIndex)
+
+ // test find one word
+ expectedWord, expectedIndex = "文", 17
+ actualWord, actualIndex = findFirstMatchedWord(text, []string{"文"})
+ assert.Equal(t, expectedWord, actualWord)
+ assert.Equal(t, expectedIndex, actualIndex)
+
+ // test find multiple matched words
+ expectedWord, expectedIndex = "Hello", 0
+ actualWord, actualIndex = findFirstMatchedWord(text, []string{"Hello", "文"})
+ assert.Equal(t, expectedWord, actualWord)
+ assert.Equal(t, expectedIndex, actualIndex)
+}
+
+func TestGetRuneRange(t *testing.T) {
+ var (
+ expectedBegin,
+ expectedEnd,
+ actualBegin,
+ actualEnd int
+ )
+
+ runeText := []rune("Hello, I have 中文 and 😂.")
+ runeLen := len(runeText)
+
+ // test get range of negative offset and negative limit
+ expectedBegin, expectedEnd = 0, 0
+ actualBegin, actualEnd = getRuneRange(runeText, -1, -1)
+ assert.Equal(t, expectedBegin, actualBegin)
+ assert.Equal(t, expectedEnd, actualEnd)
+
+ // test get range of exceeding offset and exceeding limit
+ expectedBegin, expectedEnd = runeLen, runeLen
+ actualBegin, actualEnd = getRuneRange(runeText, runeLen+1, runeLen+1)
+ assert.Equal(t, expectedBegin, actualBegin)
+ assert.Equal(t, expectedEnd, actualEnd)
+
+ // test get range of normal offset and exceeding limit
+ expectedBegin, expectedEnd = 3, runeLen
+ actualBegin, actualEnd = getRuneRange(runeText, 3, runeLen)
+ assert.Equal(t, expectedBegin, actualBegin)
+ assert.Equal(t, expectedEnd, actualEnd)
+
+ // test get range of normal offset and normal limit
+ expectedBegin, expectedEnd = 3, 10
+ actualBegin, actualEnd = getRuneRange(runeText, 3, 7)
+ assert.Equal(t, expectedBegin, actualBegin)
+ assert.Equal(t, expectedEnd, actualEnd)
+}
+
+func TestFetchRangedExcerpt(t *testing.T) {
+ var (
+ expected,
+ actual string
+ )
+
+ // test english string
+ expected = "hello..."
+ actual = FetchRangedExcerpt("<p>hello world</p>", "...", 0, 5)
+ assert.Equal(t, expected, actual)
+
+ // test string with offset
+ expected = "...llo你好..."
+ actual = FetchRangedExcerpt("<p>hello你好world</p>", "...", 2, 5)
+ assert.Equal(t, expected, actual)
+
+ // test mixed string with emoticon with offset
+ expected = "...你好😂..."
+ actual = FetchRangedExcerpt("<p>hello你好😂world</p>", "...", 5, 3)
+ assert.Equal(t, expected, actual)
+
+ // test mixed string with offset and exceeding limit
+ expected = "...你好😂world"
+ actual = FetchRangedExcerpt("<p>hello你好😂world</p>", "...", 5, 100)
+ assert.Equal(t, expected, actual)
+}
+
+func TestFetchMatchedExcerpt(t *testing.T) {
+ var (
+ expected,
+ actual string
+ )
+
+ html := "<p>Hello, I have 中文 and 😂 and I am supposed to work fine</p>"
+
+ // test find nothing
+ // it should return from the begin with double trimLength text
+ expected = "Hello, I h..."
+ actual = FetchMatchedExcerpt(html, []string{"youcantfindme"}, "...", 5)
+ assert.Equal(t, expected, actual)
+
+ // test find the word at the end
+ // it should return the word beginning with double trimLenth plus len(word)
+ expected = "... work fine"
+ actual = FetchMatchedExcerpt(html, []string{"youcant", "fine"}, "...", 3)
+ assert.Equal(t, expected, actual)
+
+ // test find multiple words
+ // it should return the first matched word with trimmedText
+ expected = "... have 中文 and 😂..."
+ actual = FetchMatchedExcerpt(html, []string{"中文", "😂"}, "...", 6)
+ assert.Equal(t, expected, actual)
+}
diff --git a/ui/src/common/pattern.ts b/ui/src/common/pattern.ts
index 847c6d76..133228b1 100644
--- a/ui/src/common/pattern.ts
+++ b/ui/src/common/pattern.ts
@@ -20,6 +20,8 @@
const pattern = {
email:
/^(([^<>()[\]\\.,;:\s@"]+(\.[^<>()[\]\\.,;:\s@"]+)*)|(".+"))@((\[[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}])|(([a-zA-Z\-0-9\u00A0-\uD7FF\uF900-\uFDCF\uFDF0-\uFFEF]+\.)+[a-zA-Z\u00A0-\uD7FF\uF900-\uFDCF\uFDF0-\uFFEF]{2,}))$/,
+ search:
+ /(\[.*\])|(is:answer)|(is:question)|(score:\d*)|(user:\S*)|(answers:\d*)/g,
uaWeChat: /micromessenger/i,
uaWeCom: /wxwork/i,
uaDingTalk: /dingtalk/i,
diff --git a/ui/src/components/HighlightText/index.scss b/ui/src/components/HighlightText/index.scss
new file mode 100644
index 00000000..da0468e6
--- /dev/null
+++ b/ui/src/components/HighlightText/index.scss
@@ -0,0 +1,22 @@
+/*
+ * 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.
+ */
+
+.highlight-text > mark {
+ padding: 0;
+}
diff --git a/ui/src/components/HighlightText/index.tsx b/ui/src/components/HighlightText/index.tsx
new file mode 100644
index 00000000..b5acacc1
--- /dev/null
+++ b/ui/src/components/HighlightText/index.tsx
@@ -0,0 +1,46 @@
+/*
+ * 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 { memo, FC } from 'react';
+
+import './index.scss';
+
+interface IProps {
+ text: string;
+ keywords: string[];
+}
+
+const Index: FC<IProps> = ({ text = '', keywords = [] }) => {
+ const regex = new RegExp(`(${keywords.join('|')})`, 'gi');
+
+ return (
+ <span className="highlight-text">
+ {text.split(regex).map((piece: string, index: number) => {
+ const key = `${piece.substring(0, 6)}_${index}`;
+ return keywords.find((kw: string) => kw === piece) ? (
+ <mark key={key}>{piece}</mark>
+ ) : (
+ piece
+ );
+ })}
+ </span>
+ );
+};
+
+export default memo(Index);
diff --git a/ui/src/components/index.ts b/ui/src/components/index.ts
index 7725cf43..fdf052d3 100644
--- a/ui/src/components/index.ts
+++ b/ui/src/components/index.ts
@@ -59,6 +59,7 @@ import CustomSidebar from './CustomSidebar';
import ImgViewer from './ImgViewer';
import SideNav from './SideNav';
import PluginRender from './PluginRender';
+import HighlightText from './HighlightText';
export {
Avatar,
@@ -105,5 +106,6 @@ export {
ImgViewer,
SideNav,
PluginRender,
+ HighlightText,
};
export type { EditorRef, JSONSchema, UISchema };
diff --git a/ui/src/pages/Search/components/SearchItem/index.tsx b/ui/src/pages/Search/components/SearchItem/index.tsx
index b1ebca56..a31dab56 100644
--- a/ui/src/pages/Search/components/SearchItem/index.tsx
+++ b/ui/src/pages/Search/components/SearchItem/index.tsx
@@ -18,12 +18,19 @@
*/
import { memo, FC } from 'react';
-import { Link } from 'react-router-dom';
+import { Link, useSearchParams } from 'react-router-dom';
import { ListGroupItem } from 'react-bootstrap';
import { useTranslation } from 'react-i18next';
import { pathFactory } from '@/router/pathFactory';
-import { Tag, FormatTime, BaseUserCard, Counts } from '@/components';
+import {
+ Tag,
+ FormatTime,
+ BaseUserCard,
+ Counts,
+ HighlightText,
+} from '@/components';
+import Pattern from '@/common/pattern';
import type { SearchResItem } from '@/common/interface';
import { escapeRemove } from '@/utils';
@@ -47,6 +54,10 @@ const Index: FC<Props> = ({ data }) => {
});
}
+ const [searchParams] = useSearchParams();
+ const q = searchParams.get('q');
+ const keywords = q?.replace(Pattern.search, '').split(' ') || [];
+
return (
<ListGroupItem className="py-3 px-0 border-start-0 border-end-0 bg-transparent">
<div className="mb-2 clearfix">
@@ -56,7 +67,7 @@ const Index: FC<Props> = ({ data }) => {
{data.object_type === 'question' ? 'Q' : 'A'}
</span>
<Link className="h5 mb-0 link-dark text-break" to={itemUrl}>
- {data.object.title}
+ <HighlightText text={data.object.title} keywords={keywords} />
{data.object.status === 'closed'
? ` [${t('closed', { keyPrefix: 'question' })}]`
: null}
@@ -88,7 +99,10 @@ const Index: FC<Props> = ({ data }) => {
{data.object?.excerpt && (
<p className="small text-truncate-2 mb-2 last-p text-break">
- {escapeRemove(data.object.excerpt)}
+ <HighlightText
+ text={escapeRemove(data.object.excerpt) || ''}
+ keywords={keywords}
+ />
</p>
)}