blocks|key|948719|text|您必须编写一个解析器来解析这样的查询字符串。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|948720|正则表达式对于在查询字符串中查找“verbs”和“nouns”很有用，但您可能还需要一个描述查询语言的非上下文语法，例如：|948721|<QUERY>+:=+<TARGET_SPEC>
<TARGET_SPEC>+:=+<OBJECT>+'that+can'+<ABILITY>
<TARGET_SPEC>+:=+<OBJECT>++
<OBJECT>+:=+<COLOR>+<WHAT>+
<OBJECT>+:=+<WHAT>
<COLOR>+:=+'blue'+%7C+'red'+%7C+'purple'+%7C+'green'+
<WHAT>+:=+<ITEM>+%7C+<HERO>
<ITEM>+:=+<ADJECTIVE>+<ITEM>
<ADJECTIVE>+:=+'brigade'+%7C+'hero'+%7C+'magic'+%7C+'enhanced'+%7C+'rustproof'
<ITEM>+:=+'enhancements'+%7C+'sword'+%7C+'potion'
<HERO>+:=+<HERO>+'of'+<COUNTRY>
<HERO>+:=+'kings'+%7C+'knights'+%7C+'thiefs'
<COUNTRY>+:=+'israel'+%7C+'palestine'+%7C+'jordan'+%7C+'egypt'
<ABILITY>+:=+<ABILITY>+'and'+<ABILITY>
<ABILITY>+:=+'swim'+%7C+'dance'+%7C+discard+<DISCARDABLE>+%7C+'kill'+<HERO>+%7C+'use'+<ITEM>
<DISCARDABLE>+:=+'ec's'+%7C+'et's'+%7C+'etc'|code-block|syntax|javascript|948722|基于这样的语法构建的解析器将能够确定查询的哪个部分是对象，这是一种能力，颜色，国家等。例如，给定输入字符串‘红色的约旦骑士，可以游泳’，解析器将选择正确的规则并应用它们：|948723|<QUERY>+:=+'red+knights+of+jordan+that+can+swim'
<TARGET_SPEC>+:=+'red+knights+of+jordan+that+can+swim'
<TARGET_SPEC>+:=+'red+knights+of+jordan'+'that+can'+'swim'
<OBJECT>+:=+'red+knights+of+jordan'
<ABILITY>+:=+'swim'
<COLOR>+:=+'red'
<WHAT>+:=+'knights+of+jordan'
<HERO>+:=+'knights'+'of'+'jordan'
<HERO>+:=+'knights'
<COUNTRY>+:=+'jordan'|948724|基于提取的信息，您将能够创建搜索条件。|948725|使用语法还有一个额外的好处，那就是解决了一些其他方法很难解决的歧义--例如，如果用户要求“可以杀死白衣骑士的红色国王”，那么通过将每个单词与可用的颜色列表进行匹配来查找颜色的简单算法将会失败。|948726|我推荐阅读一本关于编译器设计的书--+Dragon+Book是一个经典的选择(你不必全部阅读，只需要阅读关于词法分析器和解析器的部分)。|offset|length|948727|如果您不想自己编写整个解析器(因为这可能非常耗时且容易出错)，那么您将需要一个解析器生成器(即，一个为给定语法创建解析器源代码的程序)；here是一个有一些建议的问题。|948728|您还应该考虑阅读有关自然语言处理技术的书籍。斯坦福大学here有一门在线课程，我现在正在“参加”它，我可以全心全意地推荐它。|948729|entityMap|0|LINK|mutability|MUTABLE|url|http://en.wikipedia.org/wiki/Compilers:_Principles,_Techniques,_and_Tools|1|https://stackoverflow.com/questions/3720362/what-is-a-good-parser-generator-for-php|2|http://www.coursera.org/nlp^0|0|0|0|0|0|0|0|J|B|0|0|1W|4|1|0|R|4|2|0^^$0|@$1|2|3|4|5|6|7|1A|8|@]|9|@]|A|$]]|$1|B|3|C|5|6|7|1B|8|@]|9|@]|A|$]]|$1|D|3|E|5|F|7|1C|8|@]|9|@]|A|$G|H]]|$1|I|3|J|5|6|7|1D|8|@]|9|@]|A|$]]|$1|K|3|L|5|F|7|1E|8|@]|9|@]|A|$G|H]]|$1|M|3|N|5|6|7|1F|8|@]|9|@]|A|$]]|$1|O|3|P|5|6|7|1G|8|@]|9|@]|A|$]]|$1|Q|3|R|5|6|7|1H|8|@]|9|@$S|1I|T|1J|1|1K]]|A|$]]|$1|U|3|V|5|6|7|1L|8|@]|9|@$S|1M|T|1N|1|1O]]|A|$]]|$1|W|3|X|5|6|7|1P|8|@]|9|@$S|1Q|T|1R|1|1S]]|A|$]]|$1|Y|3|-4|5|6|7|1T|8|@]|9|@]|A|$]]]|Z|$10|$5|11|12|13|A|$14|15]]|16|$5|11|12|13|A|$14|17]]|18|$5|11|12|13|A|$14|19]]]]

You'll have to write a parser to parse such query strings. 

Regular expressions will be useful to find 'verbs' and 'nouns' in query strings, but you probably will also need a non-context grammar describing the language of your queries, for example something like this:

<pre><code>&lt;QUERY&gt; := &lt;TARGET_SPEC&gt;
&lt;TARGET_SPEC&gt; := &lt;OBJECT&gt; 'that can' &lt;ABILITY&gt;
&lt;TARGET_SPEC&gt; := &lt;OBJECT&gt; 
&lt;OBJECT&gt; := &lt;COLOR&gt; &lt;WHAT&gt; 
&lt;OBJECT&gt; := &lt;WHAT&gt;
&lt;COLOR&gt; := 'blue' | 'red' | 'purple' | 'green' 
&lt;WHAT&gt; := &lt;ITEM&gt; | &lt;HERO&gt;
&lt;ITEM&gt; := &lt;ADJECTIVE&gt; &lt;ITEM&gt;
&lt;ADJECTIVE&gt; := 'brigade' | 'hero' | 'magic' | 'enhanced' | 'rustproof'
&lt;ITEM&gt; := 'enhancements' | 'sword' | 'potion'
&lt;HERO&gt; := &lt;HERO&gt; 'of' &lt;COUNTRY&gt;
&lt;HERO&gt; := 'kings' | 'knights' | 'thiefs'
&lt;COUNTRY&gt; := 'israel' | 'palestine' | 'jordan' | 'egypt'
&lt;ABILITY&gt; := &lt;ABILITY&gt; 'and' &lt;ABILITY&gt;
&lt;ABILITY&gt; := 'swim' | 'dance' | discard &lt;DISCARDABLE&gt; | 'kill' &lt;HERO&gt; | 'use' &lt;ITEM&gt;
&lt;DISCARDABLE&gt; := 'ec's' | 'et's' | 'etc'
</code></pre>

Parser built around such a grammar will be able to determine which part of your query is an object, which is an ability, color, country etc. For example, given input string 'red knights of jordan that can swim', parser will choose right rules and apply them:

<pre><code>&lt;QUERY&gt; := 'red knights of jordan that can swim'
&lt;TARGET_SPEC&gt; := 'red knights of jordan that can swim'
&lt;TARGET_SPEC&gt; := 'red knights of jordan' 'that can' 'swim'
&lt;OBJECT&gt; := 'red knights of jordan'
&lt;ABILITY&gt; := 'swim'
&lt;COLOR&gt; := 'red'
&lt;WHAT&gt; := 'knights of jordan'
&lt;HERO&gt; := 'knights' 'of' 'jordan'
&lt;HERO&gt; := 'knights'
&lt;COUNTRY&gt; := 'jordan'
</code></pre>

Basing on extracted information you'll be able to create search criteria. 

Using a grammar has an additional benefit of resolving some ambiguities that would be hard to resolve any other way - for example, if user asks for 'red kings that can kill white knights', simple algorithms that just look for color by matching every word with list of colors available will fail.

I recommend reading a book on compiler design - <a href="http://en.wikipedia.org/wiki/Compilers:_Principles,_Techniques,_and_Tools" rel="nofollow noreferrer">Dragon Book</a> is a classic choice (you don't have to read all of it, just the part about lexers and parsers). 

If you don't want to code the entire parser yourself (as this can be pretty time-consuming and error-prone), you'll need a parser generator (that is, a program that creates parser source code for given grammar); <a href="https://stackoverflow.com/questions/3720362/what-is-a-good-parser-generator-for-php">here</a> is a question with some suggestions for PHP.

You should also consider reading up on Natural Language Processing techniques. There is an online course from Stanford University <a href="http://www.coursera.org/nlp" rel="nofollow noreferrer">here</a>, I'm 'attending' it now and can wholeheartedly recommend it.

blocks|key|1970306|text|我真的很喜欢socha's+suggestion，但我也会考虑一个简单得多的。|type|unstyled|depth|inlineStyleRanges|entityRanges|offset|length|data|1970307|如果您有一个已知搜索词的字典，并且能够纠正它们的语法和语法(提示:使用您的数据库和OED作为缓存层，在Google上抛出任何缓存未命中)，您可以通过将每个搜索词binary+bucket+sorting到已知类型的集合中来执行搜索。使用您的示例，每个存储桶将是:+brigade_purple，brigade_blue，type_hero，type_evil，您的每个特殊能力，以及您的每个特殊类型标识符。|1970308|对于每张卡，构造一个符合您的存储桶的位域。对于每个用户查询，构造相同的查询。然后，通过执行数据库的逐位遍历返回符合您的位掩码的结果，对于这个玩具示例，我假设它的形状类似于B%2B+tree，按照最接近掩码的结果按主位顺序排序。这样做的好处是可以扩展到支持位字段的最大长度，这在许多数据库实现中实际上是无界的。|1970309|好的，这有点技术性。这就是我在任何情况下构建搜索数据库的方式。|1970310|entityMap|0|LINK|mutability|MUTABLE|url|https://stackoverflow.com/a/9921031/517815|1|http://www.oed.com/|2|http://en.wikipedia.org/wiki/Bucket_sort|3|http://en.wikipedia.org/wiki/B%2B_tree^0|6|I|0|0|15|3|1|28|L|2|0|2D|7|3|0|0^^$0|@$1|2|3|4|5|6|7|X|8|@]|9|@$A|Y|B|Z|1|10]]|C|$]]|$1|D|3|E|5|6|7|11|8|@]|9|@$A|12|B|13|1|14]|$A|15|B|16|1|17]]|C|$]]|$1|F|3|G|5|6|7|18|8|@]|9|@$A|19|B|1A|1|1B]]|C|$]]|$1|H|3|I|5|6|7|1C|8|@]|9|@]|C|$]]|$1|J|3|-4|5|6|7|1D|8|@]|9|@]|C|$]]]|K|$L|$5|M|N|O|C|$P|Q]]|R|$5|M|N|O|C|$P|S]]|T|$5|M|N|O|C|$P|U]]|V|$5|M|N|O|C|$P|W]]]]

I really like <a href="https://stackoverflow.com/a/9921031/517815">socha's suggestion</a>, but I would consider a much simpler one as well. 

If you have a dictionary of known search terms and the ability to correct them for syntax and grammar (hint: use your database and <a href="http://www.oed.com/" rel="nofollow noreferrer">OED</a> as a cache layer, throwing any cache misses at Google), you can perform the search by <a href="http://en.wikipedia.org/wiki/Bucket_sort" rel="nofollow noreferrer">binary bucket sorting</a> each term into sets of known types. Using your example, each bucket would be: brigade_purple, brigade_blue, type_hero, type_evil, each of your special abilities, and each of your special type identifiers.

For each card, construct a bitfield conforming to your buckets. For each user query, construct the same. Then, return the results that conform to your bitmask, by performing a bitwise traversal of the database, which I assume for this toy example to be shaped like a <a href="http://en.wikipedia.org/wiki/B+_tree" rel="nofollow noreferrer">B+ tree</a>, sorting by results closest to the mask in major-bit order. This has the benefit of being extensible up to the maximum length of your backing bitfield, which can be practically unbounded in many database implementations.

Okay, so that's a bit technical. It's how I would construct the search database, in any event.

blocks|key|1965233|text|使用TierTempCur作为|type|unstyled|depth|inlineStyleRanges|entityRanges|data|1965234|--/*Use+Rela+table+to+get+the+offspring+of+the+parent*/+

++++++++(
++++++++++++SELECT+Rela.ID_RSSD_PARENT
++++++++++++++++,+Rela.ID_RSSD_OFFSPRING
++++++++++++++++,+'12/31/2011'+AS+REPORT_DATE
++++++++++++++++,+1+As+TREE_LVL+
++++++++++++++++,+CHECKSUM(ID_RSSD_PARENT,+ID_RSSD_OFFSPRING)+As+CHKSUM+
++++++++++++++++,+RIGHT('000000000'%2B+CONVERT(VARCHAR(MAX),ID_RSSD_OFFSPRING),9)+AS+RSSD_PATH+
++++++++++++FROM+CUV_RELATIONSHIPS+As+Rela
++++++++++++WHERE+ID_RSSD_PARENT+=+451965+AND+'12/31/2011'+BETWEEN+D_DT_START+AND+D_DT_END+
++++++++++++++++AND+Rela.CTRL_IND+=+1++++++--/*+indicates+subsidiary+*/
++++++++++++++++AND+Rela.OTHER_BASIS_IND+not+in+(3,8)++--/*+Per+DM's+job+*/

++++++++++++UNION+ALL

++++++++++++SELECT+Rela.ID_RSSD_PARENT
++++++++++++++++,+Rela.ID_RSSD_OFFSPRING
++++++++++++++++,+REPORT_DATE
++++++++++++++++,+TREE_LVL+%2B+1+As+TREE_LVL
++++++++++++++++,+CHECKSUM(Rela.ID_RSSD_PARENT,+Rela.ID_RSSD_OFFSPRING)+As+CHKSUM+
++++++++++++++++,+Tmp.RSSD_PATH+%2B+'\'+%2B+RIGHT('000000000'%2B+CONVERT(VARCHAR(MAX),Rela.ID_RSSD_OFFSPRING),9)+AS+RSSD_PATH+
++++++++++++FROM+CUV_RELATIONSHIPS+As+Rela
++++++++++++INNER+JOIN+TierTempCur+As+Tmp+
++++++++++++++ON+Rela.ID_RSSD_PARENT+=+Tmp.ID_RSSD_OFFSPRING
++++++++++++++AND+REPORT_DATE+BETWEEN+Rela.D_DT_START+AND+Rela.D_DT_END
++++++++++++WHERE+TREE_LVL+<+20++++++++++--/*max+depth+for+the+tier+is+20+--+to+end+self+referencing+parent/child+relationships+*/
++++++++++++++AND+Rela.CTRL_IND+=+1++++++--/*+indicates+subsidiary+*/
++++++++++++++AND+Rela.OTHER_BASIS_IND+not+in+(3,8)
++++++++),|code-block|syntax|javascript|1965235|entityMap^0|0|0^^$0|@$1|2|3|4|5|6|7|I|8|@]|9|@]|A|$]]|$1|B|3|C|5|D|7|J|8|@]|9|@]|A|$E|F]]|$1|G|3|-4|5|6|7|K|8|@]|9|@]|A|$]]]|H|$]]

WITH TierTempCur As

<pre><code>--/*Use Rela table to get the offspring of the parent*/ 

 (
 SELECT Rela.ID_RSSD_PARENT
 , Rela.ID_RSSD_OFFSPRING
 , '12/31/2011' AS REPORT_DATE
 , 1 As TREE_LVL 
 , CHECKSUM(ID_RSSD_PARENT, ID_RSSD_OFFSPRING) As CHKSUM 
 , RIGHT('000000000'+ CONVERT(VARCHAR(MAX),ID_RSSD_OFFSPRING),9) AS RSSD_PATH 
 FROM CUV_RELATIONSHIPS As Rela
 WHERE ID_RSSD_PARENT = 451965 AND '12/31/2011' BETWEEN D_DT_START AND D_DT_END 
 AND Rela.CTRL_IND = 1 --/* indicates subsidiary */
 AND Rela.OTHER_BASIS_IND not in (3,8) --/* Per DM's job */

 UNION ALL

 SELECT Rela.ID_RSSD_PARENT
 , Rela.ID_RSSD_OFFSPRING
 , REPORT_DATE
 , TREE_LVL + 1 As TREE_LVL
 , CHECKSUM(Rela.ID_RSSD_PARENT, Rela.ID_RSSD_OFFSPRING) As CHKSUM 
 , Tmp.RSSD_PATH + '\' + RIGHT('000000000'+ CONVERT(VARCHAR(MAX),Rela.ID_RSSD_OFFSPRING),9) AS RSSD_PATH 
 FROM CUV_RELATIONSHIPS As Rela
 INNER JOIN TierTempCur As Tmp 
 ON Rela.ID_RSSD_PARENT = Tmp.ID_RSSD_OFFSPRING
 AND REPORT_DATE BETWEEN Rela.D_DT_START AND Rela.D_DT_END
 WHERE TREE_LVL &lt; 20 --/*max depth for the tier is 20 -- to end self referencing parent/child relationships */
 AND Rela.CTRL_IND = 1 --/* indicates subsidiary */
 AND Rela.OTHER_BASIS_IND not in (3,8)
 ),
</code></pre>

I'm developing a search engine for a CCG. I want the user to be able to find cards based on a query like, <code>"blue brigade hero enhancements that can discard ec's"</code> or <code>"purple kings of israel"</code>. There are many variables to search for: brigades (purple, blue), types (heroes, evil characters [ec's]), special abilities (discard), and identifiers (kings of israel). I'm thinking about regexing to find common search parameters. I know this won't be easy, and it will take a long time to fine tune, but can someone point me in the right direction? Is regex even a recommend solution? I don't know if it's important, but I'm using php and mysql.

how to break apart a search query

翻译质量差，导致语言生硬或混乱。

没有提供实际的解决方法或示例。

解答不清晰，无法理解或解决问题。

页面排版不美观，阅读体验差。

文章

问答

视频

学习中心

腾讯云实验室

直播

竞赛

腾讯云代码分析专区

腾讯iOA零信任安全管理系统专区

腾讯云架构师技术同盟交流圈

腾讯云数据库专区

腾讯云顾问专区

腾讯云原生专区

腾讯混元专区

腾讯云TCE专区

腾讯云Lighthouse专区

腾讯云HAI专区

腾讯云Edgeone专区

腾讯云存储专区

腾讯云智能专区

腾讯轻联专区 

腾讯云开发专区

TAPD专区

腾讯轻量云游戏服专区

腾讯云最具价值专家

腾讯云架构师技术同盟

腾讯云创作之星

腾讯云开发者先锋

腾讯云代码助手

云原生构建

TAPD 敏捷项目管理

Cloud Studio

SDK中心

API中心

命令行工具

涵盖代码开发、场景应用、自动测试全流程，助你从零构建专属AI助手

一站式MCP教程库，解锁AI应用新玩法

我正在为CCG开发一个搜索引擎。我希望用户能够找到基于查询的卡，如"blue brigade hero enhancements that can discard ec's"或"purple kings of israel"。有许多变量可以搜索:旅(紫色，蓝色)，类型(英雄，邪恶人物ec)，特殊能力(丢弃)和身份(以色...

问如何分解搜索查询
EN

Stack Overflow用户

社区

活动

圈层

关于

腾讯云开发者

热门产品

热门推荐

更多推荐

问如何分解搜索查询EN