Python中如何基于指定表列列表筛选字符串内IF语句块？正则可行吗？

问题描述 #

我有一个结构重复的超长字符串，内容是针对数据库模式中各表列的IF语句块，示例如下：

IF ( 
ISFILTERED ( Table1[Column_1] ), 
VAR ___f = FILTERS ( Table1[Column_1] ) 
VAR ___r = COUNTROWS ( ___f ) 
VAR ___t = TOPN ( MaxFilters, ___f, Table1[Column_1])
VAR ___d = CONCATENATEX ( ___t, Table1[Column_1], ", " )
VAR ___x = "Table1[Column_1] = " & ___d & IF(___r > MaxFilters, ", ... [" & ___r & " items selected]" ) & " " 
RETURN ___x & UNICHAR(13) & UNICHAR(10)
)
& IF (
ISFILTERED ( Table1[Column_2] ), 
VAR ___f = FILTERS ( Table1[Column_2] ) 
VAR ___r = COUNTROWS ( ___f ) 
VAR ___t = TOPN ( MaxFilters, ___f, Table1[Column_2])
VAR ___d = CONCATENATEX ( ___t, Table1[Columnw_1], ", " )
VAR ___x = "Table1[Column_2] = " & ___d & IF(___r > MaxFilters, ", ... [" & ___r & " items selected]" ) & " " 
RETURN ___x & UNICHAR(13) & UNICHAR(10)
)
& IF (
...

同时持有需保留的表列列表：

['Table1[Column_1]', 'Table2[Column_4]', 'Table6[Column_22]']

要求仅保留列表中表列对应的完整IF语句块，删除其余块，优先使用Python实现。

可行方案 #

方案1：正则表达式匹配筛选 #

通过解析原始字符串，提取所有IF语句块，再筛选出包含目标列的块。这种方法适合原始字符串存在格式差异（如换行、空格不一致）的场景。

Python代码实现

import re

# 原始超长字符串（替换为你的实际内容）
original_str = """IF ( 
ISFILTERED ( Table1[Column_1] ), 
VAR ___f = FILTERS ( Table1[Column_1] ) 
VAR ___r = COUNTROWS ( ___f ) 
VAR ___t = TOPN ( MaxFilters, ___f, Table1[Column_1])
VAR ___d = CONCATENATEX ( ___t, Table1[Column_1], ", " )
VAR ___x = "Table1[Column_1] = " & ___d & IF(___r > MaxFilters, ", ... [" & ___r & " items selected]" ) & " " 
RETURN ___x & UNICHAR(13) & UNICHAR(10)
)
& IF (
ISFILTERED ( Table1[Column_2] ), 
VAR ___f = FILTERS ( Table1[Column_2] ) 
VAR ___r = COUNTROWS ( ___f ) 
VAR ___t = TOPN ( MaxFilters, ___f, Table1[Column_2])
VAR ___d = CONCATENATEX ( ___t, Table1[Columnw_1], ", " )
VAR ___x = "Table1[Column_2] = " & ___d & IF(___r > MaxFilters, ", ... [" & ___r & " items selected]" ) & " " 
RETURN ___x & UNICHAR(13) & UNICHAR(10)
)
& IF (
ISFILTERED ( Table2[Column_4] ), 
VAR ___f = FILTERS ( Table2[Column_4] ) 
VAR ___r = COUNTROWS ( ___f ) 
VAR ___t = TOPN ( MaxFilters, ___f, Table2[Column_4])
VAR ___d = CONCATENATEX ( ___t, Table2[Column_4], ", " )
VAR ___x = "Table2[Column_4] = " & ___d & IF(___r > MaxFilters, ", ... [" & ___r & " items selected]" ) & " " 
RETURN ___x & UNICHAR(13) & UNICHAR(10)
)
"""

# 需要保留的表列列表
keep_columns = ['Table1[Column_1]', 'Table2[Column_4]', 'Table6[Column_22]']

def extract_if_blocks(s):
    """提取所有完整的IF语句块"""
    blocks = []
    start_idx = None
    bracket_count = 0
    in_quotes = False

    for idx, char in enumerate(s):
        # 处理引号内的内容，避免误判括号
        if char == '"':
            in_quotes = not in_quotes
        if in_quotes:
            continue

        # 识别IF块的起始
        if char == '(' and idx >=3 and s[idx-3:idx] == 'IF ':
            start_idx = idx -3
            bracket_count = 1
        elif char == '(' and start_idx is not None:
            bracket_count +=1
        elif char == ')' and start_idx is not None:
            bracket_count -=1
            # 括号计数归0时，说明当前IF块结束
            if bracket_count == 0:
                block = s[start_idx:idx+1].strip()
                blocks.append(block)
                start_idx = None
                # 跳过后续的&和空白符
                while idx+1 < len(s) and s[idx+1] in '& \n\t':
                    idx +=1
    return blocks

# 提取所有IF块并筛选
all_blocks = extract_if_blocks(original_str)
filtered_blocks = [block for block in all_blocks if any(f'ISFILTERED ( {col} )' in block for col in keep_columns)]

# 拼接结果
final_result = ' & '.join(filtered_blocks)
print(final_result)

方案2：模板生成新字符串 #

如果所有IF语句块的结构完全固定，直接遍历目标表列，用模板生成对应块。这种方法更简洁可靠，无需处理复杂的正则匹配逻辑。

Python代码实现

# 需要保留的表列列表
keep_columns = ['Table1[Column_1]', 'Table2[Column_4]', 'Table6[Column_22]']

# 定义IF块模板（与原始结构完全一致）
if_block_template = """IF ( 
ISFILTERED ( {column} ), 
VAR ___f = FILTERS ( {column} ) 
VAR ___r = COUNTROWS ( ___f ) 
VAR ___t = TOPN ( MaxFilters, ___f, {column})
VAR ___d = CONCATENATEX ( ___t, {column}, ", " )
VAR ___x = "{column} = " & ___d & IF(___r > MaxFilters, ", ... [" & ___r & " items selected]" ) & " " 
RETURN ___x & UNICHAR(13) & UNICHAR(10)
)"""

# 生成所有目标列对应的IF块
filtered_blocks = [if_block_template.format(column=col) for col in keep_columns]

# 拼接结果
final_result = ' & '.join(filtered_blocks)
print(final_result)

方案选择建议 #

• 若原始字符串存在格式差异（如换行、空格不统一），优先使用正则表达式方案。
• 若所有IF块结构完全一致，优先使用模板生成方案，代码更简洁，出错概率更低。

内容的提问来源于stack exchange，提问作者Tom