全文搜索
执行全文搜索
全文搜索可在较大的文本中查找单词或短语。您可以在特定文本字段或所有文本字段中进行搜索。
本文很好地概述了最相关的全文搜索功能。请在参考文档中查找有关所有全文搜索功能的更多详细信息。
本文中的示例使用具有以下字段的架构：
| 字段名称 | 字段类型 |
|---|---|
brand |
TEXT |
model |
TEXT |
description |
TEXT |
单个单词
要在所有文本字段中搜索单词（或词干），您可以构造以下简单查询：
FT.SEARCH index "word"
您可能希望将搜索限制为特定文本字段，而不是在所有文本字段中搜索。
FT.SEARCH index "@field: word"
自然语言中经常出现的单词，例如the或a对于英语，则不会编入索引，并且不会返回搜索结果。您可以在停用词文章中找到更多详细信息。
以下示例搜索描述中包含单词 'kids' 的所有自行车：
Phrase
A phrase is a sentence, sentence fragment, or small group of words. You can find further details about how to find exact phrases in the exact match article.
Word prefix
You can also search for words that match a given prefix.
FT.SEARCH index "prefix*"
FT.SEARCH index "@field: prefix*"
Important:
The prefix needs to be at least two characters long.
Here is an example that shows you how to search for bicycles with a brand that starts with 'ka':
Word suffix
Similar to the prefix, it is also possible to search for words that share the same suffix.
FT.SEARCH index "*suffix"
You can also combine prefix- and suffix-based searches within a query expression.
FT.SEARCH index "*infix*"
Here is an example that finds all brands that end with 'bikes':
Fuzzy search
A fuzzy search allows you to find documents with words that approximately match your search term. To perform a fuzzy search, you wrap search terms with pairs of % characters. A single pair represents a (Levenshtein) distance of one, two pairs represent a distance of two, and three pairs, the maximum distance, represents a distance of three.
Here is the command that searches across all text fields with a distance of one:
FT.SEARCH index "%word%"
The following example finds all documents that contain a word that has a distance of one to the incorrectly spelled word 'optamized'. You can see that this matches the word 'optimized'.
If you want to increase the maximum word distance to two, you can use the following query:
Unicode considerations
Redis Query Engine only supports Unicode characters in the basic multilingual plane; U+0000 to U+FFFF. Unicode characters beyond U+FFFF, such as Emojis, are not supported and would not be retrieved by queries including such characters in the following use cases:
- Querying TEXT fields with Prefix/Suffix/Infix
- Querying TEXT fields with fuzzy
Examples:
redis> FT.CREATE idx SCHEMA tag TAG text TEXT
OK
redis> HSET doc:1 tag '😀😁🙂' text '😀😁🙂'
(integer) 2
redis> HSET doc:2 tag '😀😁🙂abc' text '😀😁🙂abc'
(integer) 2
redis> FT.SEARCH idx '@text:(*😀😁🙂)' NOCONTENT
1) (integer) 0
redis> FT.SEARCH idx '@text:(*😀😁🙂*)' NOCONTENT
1) (integer) 0
redis> FT.SEARCH idx '@text:(😀😁🙂*)' NOCONTENT
1) (integer) 0
redis> FT.SEARCH idx '@text:(%😀😁🙃%)' NOCONTENT
1) (integer) 0
On this page