sql字段解析器的实现示例

时间:2022-05-19 00:11:47

用例:有一段sql语句,我们需要从中截取出所有字段部分,以便进行后续的类型推断或者别名字段抽取定义,请给出此解析方法。

想来很简单吧,因为 sql 中的字段列表,使用方式有限,比如 a as b, a, a b...

1. 解题思路

  如果不想做复杂处理,最容易想到的,就是直接用某个特征做分割即可。比如,先截取出 字段列表部分,然后再用逗号',' 分割,就可以得到一个个的字段了。然后再要细分,其实只需要用 as 进行分割就可以了。

  看起来好像可行,但是存在许多漏洞,首先,这里面有太多的假设:各种截取部分要求必须符合要求,必须没有多余的逗号,必须要有as 等等。这明显不符合要求了。

  其二,我们可以换一种转换方式。比如先截取到field部分,然后先以 as 分割,再以逗号分割,然后取最后一个词作为field。

  看起来好像更差了,截取到哪里已经完全不知道了。即原文已经被破坏殆尽,而且同样要求要有 as 转换标签,而且对于函数觊觎有 as 的场景,就完全错误了。

  其三,最好还是自行一个个单词地解析,field 字段无外乎几种情况,1. 普通字段如 select a; 2. 带as的普通字段如 select a as b; 3. 带函数的字段如 select coalesce(a, b); 4. 带函数且带as的字段如 select coalesce(a, b) ab; 5. 函数内带as的字段如 select cast(a as string) b; ...   我们只需依次枚举对应的情况,就可以将字段解析出来了。

  看起来是个不错的想法。但是具体实现如何?

2. 具体解析实现

  主要分两个部分,1. 需要定义一个解析后的结果数据结构,以便清晰描述字段信息; 2. 分词解析sql并以结构体返回;

  我们先来看看整个算法核心:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
/**
 * 功能描述: 简单sql字段解析器
 *
 *        样例如1:
 *          select COALESCE(t1.xno, t2.xno, t3.xno) as xno,
 *             case when t1.no is not null then 1 else null end as xxk001,
 *             case when t2.no is not null then 1 else null end as xxk200,
 *             case when t3.xno is not null then 1 else null end as xx3200
 *             from xxk001 t1
 *               full join xxkj100 t2 on t1.xno = t2.xno
 *               full join xxkj200 t3 on t1.xno = t3.xno;
 *
 *        样例如2:
 *          select cast(a as string) as b from ccc;
 *
 *        样例如3:
 *          with a as(select cus,x1 from b1), b as (select cus,x2 from b2)
 *              select a.cus as a_cus from a join b on a.cus=b.cus where xxx;
 *
 *        样例如4:
 *         select a.xno,b.xx from a_tb as a join b_tb as b on a.id = b.id
 *
 *        样例如5:
 *          select cast  \t(a as string) a_str, cc (a as double) a_double from x
 *
 */
public class SimpleSqlFieldParser {
 
    /**
     * 解析一段次标签sql 中的字段列表
     *
     * @param sql 原始sql, 需如 select xx from xxx join ... 格式
     * @return 字段列表
     */
    public static List<SelectFieldClauseDescriptor> parse(String sql) {
        String columnPart = adaptFieldPartSql(sql);
        int deep = 0;
        List<StringBuilder> fieldTokenSwap = new ArrayList<>();
        StringBuilder currentTokenBuilder = new StringBuilder();
        List<SelectFieldClauseDescriptor> fieldList = new ArrayList<>();
        fieldTokenSwap.add(currentTokenBuilder);
        int len = columnPart.length();
        char[] columnPartChars = columnPart.toCharArray();
        for(int i = 0; i < len; i++) {
            // 空格忽略,换行忽略,tab忽略
            // 字符串相接
            // 左(号入栈,++deep;
            // 右)号出栈,--deep;
            // deep>0 忽略所有其他直接拼接
            // as 则取下一个值为fieldName
            // case 则直接取到end为止;
            //,号则重置token,构建结果集
            char currentChar = columnPartChars[i];
            switch (currentChar) {
                case '(':
                    ++deep;
                    currentTokenBuilder.append(currentChar);
                    break;
                case ')':
                    --deep;
                    currentTokenBuilder.append(currentChar);
                    break;
                case ',':
                    if(deep == 0) {
                        addNewField(fieldList, fieldTokenSwap, true);
                        fieldTokenSwap = new ArrayList<>();
                        currentTokenBuilder = new StringBuilder();
                        fieldTokenSwap.add(currentTokenBuilder);
                        break;
                    }
                    currentTokenBuilder.append(currentChar);
                    break;
                case ' ':
                case '\t':
                case '\r':
                case '\n':
                    if(deep > 0) {
                        currentTokenBuilder.append(currentChar);
                        continue;
                    }
                    if(currentTokenBuilder.length() == 0) {
                        continue;
                    }
                    // original_name as   --> alias
                    if(i + 1 < len) {
                        int j = i + 1;
                        // 收集连续的空格
                        StringBuilder spaceHolder = new StringBuilder();
                        boolean isNextLeftBracket = false;
                        do {
                            char nextChar = columnPart.charAt(j++);
                            if(nextChar == ' ' || nextChar == '\t'
                                    || nextChar == '\r' || nextChar == '\n') {
                                spaceHolder.append(nextChar);
                                continue;
                            }
                            if(nextChar == '(') {
                                isNextLeftBracket = true;
                            }
                            break;
                        } while (j < len);
                        if(isNextLeftBracket) {
                            currentTokenBuilder.append(currentChar);
                        }
                        if(spaceHolder.length() > 0) {
                            currentTokenBuilder.append(spaceHolder);
                            i += spaceHolder.length();
                        }
                        if(isNextLeftBracket) {
                            // continue next for, function begin
                            continue;
                        }
                    }
                    if(fieldTokenSwap.size() == 1) {
                        if(fieldTokenSwap.get(0).toString().equalsIgnoreCase("case")) {
                            String caseWhenPart = CommonUtil.readSplitWord(
                                    columnPartChars, i, " ", "end");
                            currentTokenBuilder.append(caseWhenPart);
                            if(caseWhenPart.length() <= 0) {
                                throw new BizException("语法错误,未找到case..when的结束符");
                            }
                            i += caseWhenPart.length();
                        }
                    }
                    addNewField(fieldList, fieldTokenSwap, false);
                    currentTokenBuilder = new StringBuilder();
                    fieldTokenSwap.add(currentTokenBuilder);
                    break;
                    // 空格忽略
                default:
                    currentTokenBuilder.append(currentChar);
                    break;
            }
 
        }
        // 处理剩余尚未存储的字段信息
        addNewField(fieldList, fieldTokenSwap, true);
        return fieldList;
    }
 
    /**
     * 新增一个字段描述
     *
     * @param fieldList 字段容器
     * @param fieldTokenSwap 候选词
     */
    private static void addNewField(List<SelectFieldClauseDescriptor> fieldList,
                                    List<StringBuilder> fieldTokenSwap,
                                    boolean forceAdd) {
        int ts = fieldTokenSwap.size();
        if(ts == 1 && forceAdd) {
            // db.original_name,
            String fieldName = fieldTokenSwap.get(0).toString();
            String alias = fieldName;
            if(fieldName.contains(".")) {
                alias = fieldName.substring(fieldName.lastIndexOf('.') + 1);
            }
            fieldList.add(new SelectFieldClauseDescriptor(fieldName, alias));
            return;
        }
        if(ts < 2) {
            return;
        }
        if(ts == 2) {
            // original_name alias,
            if(fieldTokenSwap.get(1).toString().equalsIgnoreCase("as")) {
                return;
            }
            fieldList.add(new SelectFieldClauseDescriptor(
                    fieldTokenSwap.get(0).toString(),
                    fieldTokenSwap.get(1).toString()));
        }
        else if(ts == 3) {
            // original_name as alias,
            fieldList.add(new SelectFieldClauseDescriptor(
                    fieldTokenSwap.get(0).toString(),
                    fieldTokenSwap.get(2).toString()));
        }
        else {
            throw new BizException("字段语法解析错误,超过3个以字段描述信息:" + ts);
        }
    }
 
    // 截取适配 field 字段信息部分
    private static String adaptFieldPartSql(String fullSql) {
        int start = fullSql.lastIndexOf("select ");
        int end = fullSql.lastIndexOf(" from");
        String columnPart = fullSql.substring(start + "select ".length(), end);
        return columnPart.trim();
    }
 
}

  应该说是比较简单的,一个for, 一个 switch ,就搞定了。其他的,更多的是逻辑判定。

  下面我们来看看字段描述类的写法,其实就是两个字段,源字段和别名。

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
/**
 * 功能描述: sql字段描述 select 字段描述类
 *
 */
public class SelectFieldClauseDescriptor {
    private String fieldName;
    private String alias;
 
    public SelectFieldClauseDescriptor(String fieldName, String alias) {
        this.fieldName = fieldName;
        this.alias = alias;
    }
 
    public String getFieldName() {
        return fieldName;
    }
 
    public String getAlias() {
        return alias;
    }
 
 
    @Override
    public boolean equals(Object o) {
        if (this == o) return true;
        if (o == null || getClass() != o.getClass()) return false;
        SelectFieldClauseDescriptor that = (SelectFieldClauseDescriptor) o;
        return Objects.equals(fieldName, that.fieldName) &&
                Objects.equals(alias, that.alias);
    }
 
    @Override
    public int hashCode() {
        return Objects.hash(fieldName, alias);
    }
 
    @Override
    public String toString() {
        return "SelectFieldClauseDescriptor{" +
                "fieldName='" + fieldName + '\'' +
                ", alias='" + alias + '\'' +
                '}';
    }
}

它存在的意义,仅仅是为了使用方更方便取值,以为更进一步的解析提供了依据。

3. 单元测试

  其实像写这种工具类,单元测试最是方便简单。因为最初的结果,我们早已预料,以测试驱动开发最合适不过了。而且,基本上一出现不符合预期的值时,很快速就定位问题了。

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
/**
 * 功能描述: sql字段解析器测试
 **/
public class SimpleSqlFieldParserTest {
 
    @Test
    public void testParse() {
        String sql;
        List<SelectFieldClauseDescriptor> parsedFieldList;
        sql = "select COALESCE(t1.xno, t2.xno, t3.xno) as xno,\n" +
                "   case when t1.xno is not null then 1 else null end as xxk001,\n" +
                "   case when t2.xno is not null then 1 else null end as xxk200,\n" +
                "   case when t3.xno is not null then 1 else null end as xx3200\n" +
                "   from xxk001 t1\n" +
                "     full join xxkj100 t2 on t1.xno = t2.xno\n" +
                "     full join xxkj200 t3 on t1.xno = t3.xno;";
        parsedFieldList = SimpleSqlFieldParser.parse(sql);
        System.out.println("result:");
        parsedFieldList.forEach(System.out::println);
        Assert.assertEquals("字段个数解析不正确",
                4, parsedFieldList.size());
        Assert.assertEquals("字段别名解析不正确",
                "xno", parsedFieldList.get(0).getAlias());
        Assert.assertEquals("字段别名解析不正确",
                "xx3200", parsedFieldList.get(3).getAlias());
 
        sql = "select cast(a as string) as b from ccc;";
        parsedFieldList = SimpleSqlFieldParser.parse(sql);
        System.out.println("result:");
        parsedFieldList.forEach(System.out::println);
        Assert.assertEquals("字段个数解析不正确",
                1, parsedFieldList.size());
        Assert.assertEquals("字段别名解析不正确",
                "b", parsedFieldList.get(0).getAlias());
 
        sql = "with a as(select cus,x1 from b1), b as (select cus,x2 from b2)\n" +
                "    select a.cus as a_cus, cast(a \nas string) as a_cus2, " +
                "b.x2 b2 from a join b on a.cus=b.cus where xxx;";
        parsedFieldList = SimpleSqlFieldParser.parse(sql);
        System.out.println("result:");
        parsedFieldList.forEach(System.out::println);
        Assert.assertEquals("字段个数解析不正确",
                3, parsedFieldList.size());
        Assert.assertEquals("字段别名解析不正确",
                "a_cus", parsedFieldList.get(0).getAlias());
        Assert.assertEquals("字段别名解析不正确",
                "b2", parsedFieldList.get(2).getAlias());
 
        sql = "select a.xno,b.xx,qqq from a_tb as a join b_tb as b on a.id = b.id";
        parsedFieldList = SimpleSqlFieldParser.parse(sql);
        System.out.println("result:");
        parsedFieldList.forEach(System.out::println);
        Assert.assertEquals("字段个数解析不正确",
                3, parsedFieldList.size());
        Assert.assertEquals("字段别名解析不正确",
                "xno", parsedFieldList.get(0).getAlias());
        Assert.assertEquals("字段别名解析不正确",
                "qqq", parsedFieldList.get(2).getAlias());
 
        sql = "select cast (a.a_int as string) a_str, b.xx, coalesce  \n( a, b, c) qqq from a_tb as a join b_tb as b on a.id = b.id";
        parsedFieldList = SimpleSqlFieldParser.parse(sql);
        System.out.println("result:");
        parsedFieldList.forEach(System.out::println);
        Assert.assertEquals("字段个数解析不正确",
                3, parsedFieldList.size());
        Assert.assertEquals("字段别名解析不正确",
                "a_str", parsedFieldList.get(0).getAlias());
        Assert.assertEquals("字段原始名解析不正确",
                "cast (a.a_int as string)", parsedFieldList.get(0).getFieldName());
        Assert.assertEquals("字段别名解析不正确",
                "qqq", parsedFieldList.get(2).getAlias());
        Assert.assertEquals("字段原始名解析不正确",
                "coalesce  \n( a, b, c)", parsedFieldList.get(2).getFieldName());
    }
}

至此,一个简单的字段解析器完成。小工具,供参考!

到此这篇关于sql字段解析器的实现示例的文章就介绍到这了,更多相关sql字段解析器内容请搜索服务器之家以前的文章或继续浏览下面的相关文章希望大家以后多多支持服务器之家!

原文链接:https://www.cnblogs.com/yougewe/p/14911443.html