Files
游夏 36f4e3720e 🎨 Add Simplified/Traditional Chinese sensitivity option (#17846)
*  搜索支持区分繁简选项 Add Simplified/Traditional Chinese sensitivity option for search

新增搜索配置 hanSensitive(区分繁简,默认开启 = 现状行为,命名与语义
向 caseSensitive 对齐)。关闭后简体与繁体可互相检索。

- FTS 层:关闭区分繁简时,blocks_fts / blocks_fts_case_insensitive 建表
  追加 han_insensitive 分词器参数(依赖 88250/go-sqlite3 配套补丁),
  繁体在分词时按单字折叠为简体,token 偏移仍指向原文;
- 切换选项复用区分大小写现有的 FullReindex 流程;
- 高亮:EncloseHighlighting 在关闭区分繁简时将关键字逐字符展开为
  繁简等价字符类(如 诗经 -> [诗詩][经經]),与分词器使用同一份
  OpenCC TSCharacters 映射数据;
- 兼容:旧版本 conf.json 缺 hanSensitive 字段时按既往行为置为开启;
  setSearch 请求未携带该字段(现行前端/第三方调用)时保持当前值,
  避免被零值意外翻转并触发重建索引;
- 遵照配置界面重写计划(#17601),本次不改动设置界面与 i18n,仅添加
  配置数据结构字段与前端类型定义(config.d.ts)。

范围说明:本次仅覆盖块全文搜索(关键字/查询语法)的命中与高亮;
历史/资源文件搜索、数据库属性过滤、虚拟引用、查找替换的替换动作
不在本次范围内(替换保持字节精确以避免误改原文)。

Fixes #13304

* 🐛 add missing util.SearchHanSensitive referenced by highlighting and sql.SetHanSensitive

Copilot 评审指出 util.SearchHanSensitive 未定义导致无法编译:补上该全局
变量(默认 true,与既往行为一致,由 sql.SetHanSensitive 维护),并修正
conf/search.go 的 gofmt 对齐。已在本地以 dev 当前 go.mod(已合并的
88250/go-sqlite3 分词器)完成 go build -tags fts5 与 kernel/search 单测验证。
2026-06-12 15:59:50 +08:00

152 lines
4.0 KiB
Go

// SiYuan - Refactor your thinking
// Copyright (c) 2020-present, b3log.org
//
// This program is free software: you can redistribute it and/or modify
// it under the terms of the GNU Affero General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU Affero General Public License for more details.
//
// You should have received a copy of the GNU Affero General Public License
// along with this program. If not, see <https://www.gnu.org/licenses/>.
package search
import (
"fmt"
"regexp"
"strings"
"unicode/utf8"
"github.com/88250/gulu"
"github.com/88250/lute/lex"
"github.com/siyuan-note/siyuan/kernel/util"
)
func MarkText(text string, keyword string, beforeLen int, caseSensitive bool) (pos int, marked string) {
if "" == keyword {
return -1, text
}
keywords := SplitKeyword(keyword)
marked = EncloseHighlighting(text, keywords, "<mark>", "</mark>", caseSensitive, false)
pos = strings.Index(marked, "<mark>")
if 0 > pos {
return
}
var before []rune
var count int
for i := pos; 0 < i; { // 关键字前面太长的话缩短一些
r, size := utf8.DecodeLastRuneInString(marked[:i])
i -= size
before = append([]rune{r}, before...)
count++
if beforeLen < count {
before = append([]rune("..."), before...)
break
}
}
marked = string(before) + marked[pos:]
return
}
const (
TermSep = "__term@sep__"
SearchMarkLeft = "__@mark__"
SearchMarkRight = "__mark@__"
)
func SplitKeyword(keyword string) (keywords []string) {
keyword = strings.TrimSpace(keyword)
if "" == keyword {
return
}
words := strings.Split(keyword, TermSep)
if 1 < len(words) {
for _, word := range words {
if "" == word {
continue
}
keywords = append(keywords, word)
}
} else {
keywords = append(keywords, keyword)
}
return
}
func EncloseHighlighting(text string, keywords []string, openMark, closeMark string, caseSensitive, splitWords bool) (ret string) {
ic := "(?i)"
if caseSensitive {
ic = "(?)"
}
var re strings.Builder
re.WriteString(ic + "(")
for i, k := range keywords {
if "" == k {
continue
}
wordBoundary := false
if splitWords {
wordBoundary = lex.IsASCIILetterNums(gulu.Str.ToBytes(k)) // Improve virtual reference split words https://github.com/siyuan-note/siyuan/issues/7833
}
if !util.SearchHanSensitive {
// 不区分繁简:将关键字逐字符展开为繁简等价字符类,如 "诗经" -> "[诗詩][经經]"
k = hanInsensitiveRegexp(util.EscapeHTML(k))
} else {
k = regexp.QuoteMeta(util.EscapeHTML(k))
}
re.WriteString("(")
if wordBoundary {
re.WriteString("\\b")
}
re.WriteString(k)
if wordBoundary {
re.WriteString("\\b")
}
re.WriteString(")")
if i < len(keywords)-1 {
re.WriteString("|")
}
}
re.WriteString(")")
ret = util.EscapeHTML(text)
ret = strings.ReplaceAll(ret, "&#34;", "\ue000")
ret = strings.ReplaceAll(ret, "&lt;", "\ue001")
ret = strings.ReplaceAll(ret, "&gt;", "\ue002")
ret = strings.ReplaceAll(ret, "&#39;", "\ue003")
if reg, err := regexp.Compile(re.String()); err == nil {
ret = reg.ReplaceAllStringFunc(ret, func(s string) string { return openMark + s + closeMark })
}
ret = strings.ReplaceAll(ret, "\ue000", "&#34;")
ret = strings.ReplaceAll(ret, "\ue001", "&lt;")
ret = strings.ReplaceAll(ret, "\ue002", "&gt;")
ret = strings.ReplaceAll(ret, "\ue003", "&#39;")
// 搜索结果预览包含转义符问题 Search results preview contains escape character issue https://github.com/siyuan-note/siyuan/issues/9790
ret = strings.ReplaceAll(ret, "\\<span", "\\\\<span")
return
}
const (
MarkDataType = "search-mark"
VirtualBlockRefDataType = "virtual-block-ref"
)
func GetMarkSpanStart(dataType string) string {
return fmt.Sprintf("<span data-type=\"%s\">", dataType)
}
func GetMarkSpanEnd() string {
return "</span>"
}