正则表达式（R&Python)

regular expression

1.R,strongly recommend this blog

The table_info examples are following:

du_mtime_cinema_showtime20190606

du_amap_shoppingmall_indoor_201903_d4

du_amap_shopping_mall_info_2017

du_amap_ship_201909

I want tables which start with"du_amap_" and end with year/month, so in the tables above,

I only want the fourth one. Below, in R, character escape (the backlash character \) should be \\.

"^"means the start and "$" means the end.

^can delete but $ can't because we use "grep" function.

keyword_all <-'^du_amap_.+2\\d{5}$'
keyword_table <- grep(keyword_all, table_info$Tables_in_risingdata, value =T)

str_extract_all is a function which only filters out the characters that fits the pattern.

The below codes extract the last six numbers:year and month

table_name_body <- 'amap_cvs_citycount'
month<-str_extract_all(string=keyword_table[p],pattern='\\d.+')%>% as.character()

* means the pattern in front of it will appear one or more times, | means or, and . means any characters.

Below codes means deleting "du_amap_" and "_201..".

keyword <- gsub('.*amap_|_201.*', '', table_name_body)
shoppingmall_amap$name<-gsub('(\\(.*\\))',"",shoppingmall_amap$name)

latitude and longtitude

\\d{2}[.]\\d+

find Chinese

[\u4E00-\u9FA5\\s]+ #many characters，including space
[\u4E00-\u9FA5]+ #many characters，not including space
[\u4E00-\u9FA5] #one character

2.Python

包

import re

查找数字，注意这里python转义只有一个\，但R里转义要两个：\\

pattern1 = re.compile(r'\d+')

这里是找表格里每行以(080)开头的数字

pattern1 = re.compile(r'\(080\)\d+')
fixed_line_all=pd.DataFrame()
for i in range(len(calls_pd[0])):
        fixed_line=pattern1.findall(calls_pd[0][i])
        fixed_line_all=set(fixed_line_all).union(fixed_line)
fixed_line_all=pd.DataFrame(fixed_line_all)

这里提取以7、8、9开头的前四位数

pattern2=re.compile(r'^(7\d{3}|8\d{3}|9\d{3})')
for i in range(len(fixed_line_bind[1])):
        mobile_line=pattern2.findall(fixed_line_bind[1][i])
        mobile_bang=set(mobile_bang).union(mobile_line)
mobile_bang=pd.DataFrame(mobile_bang)