SQLFluff自定义规则Rule_LL99未加载，Lint无预期结果求助

自定义规则代码 #

规则文件custom_rules.py与.sqlfluff同级，代码如下：

from sqlfluff.core.rules import BaseRule, LintResult
from sqlfluff.core.parser import BaseSegment

class Rule_LL99(BaseRule):
    """Ensure columns with functions are aliased"""

    def _eval(self, segment: BaseSegment, **kwargs):
        print(f"Evaluating segment: {segment}")

        if segment.is_type('function'):
            parent_segment = segment.parent
            if parent_segment and not any(child.is_type('alias_expression') for child in parent_segment.segments):
                return LintResult(anchor=segment, description="Columns where Function has been used must be aliased")
        return None

测试SQL #

SELECT
  id,
  status,
  LOWER("text"),
  CAST("test" AS STRING)
FROM test

规则加载配置 #

在.sqlfluff中配置：

[sqlfluff:rules]
custom_rules = custom_rules::Rule_LL99

运行Lint后的输出 #

=== [dbt templater] Sorting Nodes...
=== [dbt templater] Compiling dbt project...
=== [dbt templater] Project Compiled.
All Finished 📜 🎉!

问题描述 #

单独写Python脚本测试时发现规则名称相关问题，但Lint过程中没有任何报错提示，排查困难。本人非Python开发者，求指出配置或代码中的明显问题。完整SQLFluff配置如下：

[sqlfluff]
templater = dbt
dialect = bigquery
exclude_rules = RF02, AL01, AL05, ST07, AL03

[sqlfluff:templater:dbt]
project_dir = ./

[sqlfluff:indentation]
tab_space_size = 2
indented_joins = true
indented_using_on = false

[sqlfluff:rules]
custom_rules = custom_rules::Rule_LL99

[sqlfluff:rules:capitalisation.keywords]
capitalisation_policy = upper

[sqlfluff:rules:capitalisation.identifiers]
extended_capitalisation_policy = lower

[sqlfluff:rules:capitalisation.functions]
extended_capitalisation_policy = upper

[sqlfluff:rules:capitalisation.literals]
capitalisation_policy = upper

[sqlfluff:rules:capitalisation.types]
extended_capitalisation_policy = upper

[sqlfluff:rules:aliasing.length]
min_alias_length = 3

问题分析与修复方案 #

1. 规则命名不符合SQLFluff规范 #

SQLFluff对自定义规则的命名有强制要求：必须遵循Rule_{X}{XX}格式，其中X是单个字母（代表规则类别），后面跟三位数字。你的规则名Rule_LL99用了两个字母，不符合规范，导致SQLFluff无法正确识别加载规则，因此Lint过程中既不会触发规则，也不会输出加载失败的报错。

修复：将规则名改为符合规范的格式，比如Rule_L099，同时同步修改配置中的规则引用：

[sqlfluff:rules]
custom_rules = custom_rules::Rule_L099

2. 规则逻辑存在错误 #

当前代码中，function的父节点是expression，而alias_expression属于select_target节点的子元素，所以通过parent_segment无法找到别名，导致规则逻辑失效。

修复：调整规则逻辑，直接遍历select_target节点，检查其中是否包含函数且未设置别名：

from sqlfluff.core.rules import BaseRule, LintResult
from sqlfluff.core.parser import BaseSegment

class Rule_L099(BaseRule):
    """Ensure columns with functions are aliased"""

    def _eval(self, segment: BaseSegment, **kwargs):
        # 只处理SELECT语句中的目标列节点
        if segment.is_type('select_target'):
            # 检查当前列是否包含函数
            has_function = any(child.is_type('function') for child in segment.recursive_crawl('function'))
            # 检查当前列是否有别名
            has_alias = any(child.is_type('alias_expression') for child in segment.segments)
            
            if has_function and not has_alias:
                # 找到第一个函数节点作为报错锚点
                func_segment = next(segment.recursive_crawl('function'))
                return LintResult(anchor=func_segment, description="使用函数的列必须指定别名")
        return None

3. 开启调试日志排查加载问题 #

默认SQLFluff日志级别较低，看不到规则加载失败的细节。运行Lint时加上--verbose参数，或者设置环境变量LOG_LEVEL=DEBUG，就能看到规则加载的详细日志，方便排查问题：

sqlfluff lint --verbose

内容的提问来源于stack exchange，提问作者Moseleyi