我有以下python代码可检索某些关键字后的第一个单词:
file_tokens = ('DATABASE', 'EXTERNAL_FILE', 'FILE', 'FILENAME', 'INCLUDE')
# match newline, only spaces, then exact token, then spaces, then everything but whitespace
search_pattern = r'\n\s*({})\s+([^\s]*)'.format('|'.join(file_tokens))
matches = re.findall(search_pattern, file_content_string, flags=re.IGNORECASE) # find matches
它在类似于以下字符串(包括换行符和回车符)中的运行很漂亮:
# originaly spe1 but with grd ecl file meddled with for nesting
include tests
SIMULATION
SIMULATION_TYPE SUBSURFACE
PROCESS_MODELS
SUBSURFACE_FLOW Flow
MODE BLACK_OIL
OPTIONS
ANALYTICAL_JACOBIAN
ISOTHERMAL
/
/ ! end of subsurface_flow
/ ! end of process models
CHECKPOINT
/
END !! end simulation block
SUBSURFACE
external_file example1.dat
include example2.dat
包含以下内容的匹配项:
matches = [example1.dat,example2.dat]
但是它无法生成一个仅包含关键字和其他文本的简单字符串,如下所示:
external_file example3.dat
include example4.dat
返回一个空数组或仅返回最后一个项目(有点随机):
matches = [example4.dat]或matchs = []
有什么主意吗?谢谢。
更新
确定,修改导入文本后:
external_file example3.dat
include example4.dat
database example5.dat
我意识到我的matchs数组仅缺少第一项:
matches = [example4.dat,example5.dat]
如何修改正则表达式以包含example3.dat?
答案 0 :(得分:0)
我会解决一些不同的问题。
import re
test1 = """include tests
SIMULATION
SIMULATION_TYPE SUBSURFACE
PROCESS_MODELS
SUBSURFACE_FLOW Flow
MODE BLACK_OIL
OPTIONS
ANALYTICAL_JACOBIAN
ISOTHERMAL
/
/ ! end of subsurface_flow
/ ! end of process models
CHECKPOINT
/A
END !! end simulation block
SUBSURFACE
external_file example1.dat
include example2.dat"""
test2 = """external_file example3.dat
include example4.dat"""
token = re.findall(r'\S+', test1)
token
>>>['include',
'tests',
'SIMULATION',
'SIMULATION_TYPE',
'SUBSURFACE',
'PROCESS_MODELS',
'SUBSURFACE_FLOW',
'Flow',
'MODE',
'BLACK_OIL',
'OPTIONS',
'ANALYTICAL_JACOBIAN',
'ISOTHERMAL',
'/',
'/',
'!',
'end',
'of',
'subsurface_flow',
'/',
'!',
'end',
'of',
'process',
'models',
'CHECKPOINT',
'/',
'END',
'!!',
'end',
'simulation',
'block',
'SUBSURFACE',
'external_file',
'example1.dat',
'include',
'example2.dat']
当您对您的单词进行标记后,我会建立二元语法
bi_grams = [(a,b) for a,b in zip(token[:-1], token[1:]) ]
然后过滤那些包含文件令牌作为第一个条目的二元组合
file_tokens = ('DATABASE', 'EXTERNAL_FILE', 'FILE', 'FILENAME', 'INCLUDE')
bi_grams_of_interest = [bi_gram for bi_gram in bi_grams if bi_gram[0].upper() in file_tokens]
bi_grams_of_interest
>>>[('include', 'tests'),
('external_file', 'example1.dat'),
('include', 'example2.dat')]
如果为test2运行此命令,则会得到以下输出
>>>[('external_file', 'example3.dat'), ('include', 'example4.dat')]
答案 1 :(得分:0)
您需要将\n
替换为^
,并将re.M
添加到标志:
r'(?mi)^\s*(?:{})\s+(\S+)'.format('|'.join(file_tokens))
现在,^\s*
将匹配行的开头,然后匹配0个或多个空格。
请参见Python demo:
import re
file_content_string="""external_file example3.dat
include example4.dat
database example5.dat"""
file_tokens = ('DATABASE', 'EXTERNAL_FILE', 'FILE', 'FILENAME', 'INCLUDE')
search_pattern = r'^\s*(?:{})\s+(\S+)'.format('|'.join(file_tokens))
matches = re.findall(search_pattern, file_content_string, flags=re.I|re.M)
print(matches)
输出:['example3.dat', 'example4.dat', 'example5.dat']