trmd_b1_ok
# -*- coding:utf-8 -*-
'''
从11c开始提取
'''
import re
import numpy as np
import os
year = '17A'
ss="./data/edmd/"
# filename=ss+"/EDMDI1.17A"
def get_tag():
try:
os.rename(ss+"/EDMDI1.17A",ss+"/EDMDI1.txt")
except:
pass
f1=open(ss+"/EDMDI1.txt")
p1=re.compile(r"^(?:\s{3}|X\s{2}|\W\s{2})([A-Z]{6})\s.+\n")
list_tag=list()
for line in f1.readlines():
# print(line)
match1=re.findall(p1,line)
# print(match1)
if match1:
for j in match1:
list_tag.append(j)
# filename_w1= ss+'%s'%list_tag[MM]
print(list_tag)
return list_tag
def trmd_b1_nonote(list_tag):
if not os.path.exists('./data/edmd/new/'):
os.makedirs('./data/edmd/new/') for MM in range(len(list_tag)):
try:
os.rename(ss+'%s_D.17A'%list_tag[MM],ss+'%s.txt'%list_tag[MM])
except:
break filename_w= ss+'new/%s_w.txt'%list_tag[MM]
if os.path.exists(filename_w):
os.remove(filename_w)
# import os # os.rename('./data/CODECO_D.02A','./data/CODECO_D.txt')
filename_r = ss+'%s.txt'%list_tag[MM] # txt文件和当前脚本在同一目录下，所以不用写具体路径
#00010 UNH Message header M 1
pattern1 = re.compile(r"(^\d{4,5})\s{3}[A-Z]{3}.+[CM]\s{3}\d*\s{1,}\|{0,}\n")#
pattern1_2 = re.compile(r"^\d{4,5}\s{3}([A-Z]{3}).+[CM]\s{3}\d*\s{1,}\|{0,}\n")#UNH
#pattern1_3 = re.compile(r"^\d{5}\s{3}[A-Z]{3}(.+)[CM]\s{3}\d*\s{1,}\|{0,}\n")#Message header
pattern1_4 = re.compile(r"^\d{4,5}\s{3}[A-Z]{3}.+([CM])\s{3}\d*\s{1,}\|{0,}\n")#C
pattern1_5 = re.compile(r"^\d{4,5}\s{3}[A-Z]{3}.+[CM]\s{3}(\d*)\s{1,}\|{0,}\n")#
#pattern2 = re.compile(r"^\d{5}.+Segment\sgroup\s(\d)*.+[CM]\s{3}\d*\-+\+\n" )#+结尾
#00050 ---- Segment group 1 ------------------ C 9----------------+
pattern4_1 = re.compile(r"(^\d{4,5}).+Segment\sgroup\s\d*.+[CM]\s{3}\d*.+\n")
pattern4_2 = re.compile(r"^\d{4,5}.+Segment\sgroup\s(\d*).+[CM]\s{3}\d*.+\n")
pattern4_3 = re.compile(r"^\d{4,5}.+Segment\sgroup\s\d*.+([CM])\s{3}\d*.+\n")
pattern4_4 = re.compile(r"^\d{4,5}.+Segment\sgroup\s\d*.+[CM]\s{3}(\d*).+\n")
#匹配每组的单独结尾的一行即没有Segment group的以+、+|、+||、+|||……结尾的的每个字段
#如00280 RNG Range details C 1---------------+|
pattern5_1 = re.compile(r"(^\d{4,5})\s{3}[A-Z]{3}.+[CM]\s{3}\d*\-+\+{1,10}\|{0,20}\n" )
pattern5_2 = re.compile(r"^\d{4,5}\s{3}([A-Z]{3}).+[CM]\s{3}\d*\-+\+{1,10}\|{0,20}\n" )
pattern5_3 = re.compile(r"^\d{4,5}\s{3}[A-Z]{3}.+([CM])\s{3}\d*\-+\+{1,10}\|{0,20}\n" )
pattern5_4 = re.compile(r"^\d{4,5}\s{3}[A-Z]{3}.+[CM]\s{3}(\d*)\-+\+{1,10}\|{0,20}\n" )
#以下是确定层级关系
#匹配每组的单独结尾的一行即没有Segment group的以+、+|、+||、+|||……结尾的
pattern5 = re.compile(r"^\d{5}\s{3}[A-Z]{3}.+[CM]\s{3}\d*\-+\+\|{0,10}\n" )
#匹配每组的开头一行即有Segment group的以+、+|、+||、+|||……结尾的
pattern2_1 = re.compile(r"^\d{5}.+Segment\sgroup\s(\d*).+[CM]\s{3}\d*\-+\+\n" )#+结尾
pattern2_2 = re.compile(r"^\d{5}.+Segment\sgroup\s(\d*).+[CM]\s{3}\d*\-+\+\|\n" )#+|结尾
pattern2_3 = re.compile(r"^\d{5}.+Segment\sgroup\s(\d*).+[CM]\s{3}\d*\-+\+\|\|\n" )#+||结尾
pattern2_4 = re.compile(r"^\d{5}.+Segment\sgroup\s(\d*).+[CM]\s{3}\d*\-+\+\|\|\|\n" )
pattern2_5 = re.compile(r"^\d{5}.+Segment\sgroup\s(\d*).+[CM]\s{3}\d*\-+\+\|\|\|\|\n" )
pattern2_6 = re.compile(r"^\d{5}.+Segment\sgroup\s(\d*).+[CM]\s{3}\d*\-+\+\|\|\|\|\|\n" )
pattern2_7 = re.compile(r"^\d{5}.+Segment\sgroup\s(\d*).+[CM]\s{3}\d*\-+\+\|\|\|\|\|\|\n" )
#匹配有同时多个组同时结束的情况，即以++、++|、++||……++、++|、++||……等结尾的
pattern3_1 = re.compile(r"^\d{5}.+[CM]\s{3}\d*\-+\+{2}\|{0,20}\n")# 匹配++、++|、++||……等结尾
pattern3_2 = re.compile(r"^\d{5}.+[CM]\s{3}\d*\-+\+{3}\|{0,20}\n")# 匹配+++、+++|、+++||……等结尾
pattern3_3 = re.compile(r"^\d{5}.+[CM]\s{3}\d*\-+\+{4}\|{0,20}\n")
pattern3_4 = re.compile(r"^\d{5}.+[CM]\s{3}\d*\-+\+{5}\|{0,20}\n")
pattern3_5 = re.compile(r"^\d{5}.+[CM]\s{3}\d*\-+\+{6}\|{0,20}\n")
pattern3_6 = re.compile(r"^\d{5}.+[CM]\s{3}\d*\-+\+{7}\|{0,20}\n") flag = 0
#listgr中第一个不为0的点
pos = -1
listgr =[0,0,0,0,0,0,0,0,0,0] fr = open(filename_r)
w2 = open(filename_w,'a')#a代表追加 w代表重写
# w2.write("code_pos,parent,TRSD_tag,year,list_tag[MM],S,R")
for line in fr.readlines():
matcher1 = re.findall(pattern1,line)
matcher1_2 = re.findall(pattern1_2,line)
#matcher1_3 = re.findall(pattern1_3,line)
matcher1_4 = re.findall(pattern1_4,line)
matcher1_5 = re.findall(pattern1_5,line)
matcher2_1 = re.findall(pattern2_1,line)
matcher2_2 = re.findall(pattern2_2,line)
matcher2_3 = re.findall(pattern2_3,line)
matcher2_4 = re.findall(pattern2_4,line)
matcher2_5 = re.findall(pattern2_5,line)
matcher2_6 = re.findall(pattern2_6,line)
matcher2_7 = re.findall(pattern2_7,line)
matcher3_1 = re.findall(pattern3_1,line)
matcher3_2 = re.findall(pattern3_2,line)
matcher3_3 = re.findall(pattern3_3,line)
matcher3_4 = re.findall(pattern3_4,line)
matcher3_5 = re.findall(pattern3_5,line)
matcher3_6 = re.findall(pattern3_6,line)
matcher4_1 = re.findall(pattern4_1,line)
matcher4_2 = re.findall(pattern4_2,line)
matcher4_3 = re.findall(pattern4_3,line)
matcher4_4 = re.findall(pattern4_4,line)
matcher5 = re.findall(pattern5,line)
matcher5_1 = re.findall(pattern5_1,line)
matcher5_2 = re.findall(pattern5_2,line)
matcher5_3 = re.findall(pattern5_3,line)
matcher5_4 = re.findall(pattern5_4,line) if matcher4_1!=[]:
w2.write("\n")
for j in matcher4_1:
for k in j:
w2.write(k)
if matcher4_2!=[]:
w2.write(",")
#写入parent列
if pos!= -1:
numgr =listgr[pos]
else:
numgr = 0
if numgr ==0:
w2.write("SG0,")
else:
w2.write("SG"+str(numgr)+",")
for j in matcher4_2:
for k in j:
w2.write("SG"+str(k))
if matcher4_3!=[]:
flag = 3
w2.write(",")
#默认写入year,list_tag[MM]两列
w2.write(year+","+list_tag[MM]+",")
for j in matcher4_3:
for k in j:
w2.write(k)
if matcher4_4!=[]:
w2.write(",")
for j in matcher4_4:
for k in j:
w2.write(k)
if matcher5_1!=[]:
w2.write("\n")
for j in matcher5_1:
for k in j:
w2.write(k)
if matcher5_2!=[]:
w2.write(",")
#写入parent列
if pos!= -1:
numgr =listgr[pos]
else:
numgr = 0
if numgr ==0:
w2.write("SG0,")
else:
w2.write("SG"+str(numgr)+",")
for j in matcher5_2:
for k in j:
w2.write(k)
if matcher5_3!=[]:
flag = 3
w2.write(",")
#默认写入year,list_tag[MM]两列
w2.write(year+","+list_tag[MM]+",")
for j in matcher5_3:
for k in j:
w2.write(k)
if matcher5_4!=[]:
w2.write(",")
for j in matcher5_4:
for k in j:
w2.write(k)
#确定层级关系，也就是确定listgr
if(matcher5!=[]):
for i in listgr:
if i==0:
pos = listgr.index(i)-1
break
listgr[pos]=0
if (matcher2_1!=[]):
# print "2_1"
for j in matcher2_1:
# print j
if(listgr[0]==0):
listgr[0]=j
else:
listgr[0]=0
# print listgr
if (matcher2_2!=[]):
for j in matcher2_2:
#numgr_d = j
if(listgr[1]==0):
listgr[1]=j
else:
listgr[1]=0
if (matcher2_3!=[]):
for j in matcher2_3:
if(listgr[2]==0):
listgr[2]=j
else:
listgr[2]=0
if (matcher2_4!=[]):
for j in matcher2_4:
if(listgr[3]==0):
listgr[3]=j
else:
listgr[3]=0
if (matcher2_5!=[]):
for j in matcher2_5:
if(listgr[4]==0):
listgr[4]=j
else:
listgr[4]=0
if (matcher2_6!=[]):
for j in matcher2_6:
if(listgr[5]==0):
listgr[5]=j
else:
listgr[5]=0
if (matcher2_7!=[]):
for j in matcher2_7:
if(listgr[6]==0):
listgr[6]=j
else:
listgr[6]=0
if (matcher3_1!=[]):
for i in listgr:
if i==0:
pos = listgr.index(i)-1
break
listgr[pos]=0
listgr[pos-1]=0
if (matcher3_2!=[]):
for i in listgr:
if i==0:
pos = listgr.index(i)-1
break
for k in range((pos-2),(pos+1)):
listgr[k]=0
if (matcher3_3!=[]):
for i in listgr:
if i==0:
pos = listgr.index(i)-1
break
for k in range((pos-3),(pos+1)):
listgr[k]=0
if (matcher3_4!=[]):
for i in listgr:
if i==0:
pos = listgr.index(i)-1
break
for k in range(pos-4,pos+1):
listgr[k]=0
if (matcher3_5!=[]):
for i in listgr:
if i==0:
pos = listgr.index(i)-1
break
for k in range(pos-5,pos+1):
listgr[k]=0
if (matcher3_6!=[]):
for i in listgr:
if i==0:
pos = listgr.index(i)-1
break
for k in range(pos-6,pos+1):
listgr[k]=0
#确定层级关系结束
if (matcher1!=[]):
flag = 1
w2.write("\n")
for j in matcher1:
for k in j:
w2.write(k)
#print listgr
#判断当前lit不为0的位置
for i in listgr:
if i==0:
pos = listgr.index(i)-1
break
if matcher1_2!=[]:
flag = 2
w2.write(",")
#写入parent列
if pos!= -1:
numgr =listgr[pos]
else:
numgr = 0
if numgr ==0:
w2.write("SG0,")
else:
w2.write("SG"+str(numgr)+",")
for j in matcher1_2:
for k in j:
w2.write(k)
# if matcher1_3!=[]:
# flag = 3
# w2.write(",")
# for j in matcher1_3:
# for k in j:
# w2.write(k)
if matcher1_4!=[]:
flag = 4
w2.write(",")
#默认写入year,list_tag[MM]两列
w2.write(year+","+list_tag[MM]+",")
for j in matcher1_4:
for k in j:
w2.write(k)
if ((matcher1_5!=[])and(flag ==4)):
flag = 5
w2.write(",")
for j in matcher1_5:
for k in j:
w2.write(k)
w2.close()
fr.close()
def trmd_b1_note(list_tag):
for MM in range(len(list_tag)):
filename_r = ss+'%s.txt'%list_tag[MM]
filename_w = ss+'new/%s_wnote.txt'%list_tag[MM]
if os.path.exists(filename_w):
os.remove(filename_w) fr = open(filename_r)
w2 = open(filename_w,'a')
m=0
for line in fr.readlines():
list1 = [3,6,9,12,15,18,21,24,27,30]
for i in range(10):
k = list1[i]
# print k
pattern1 = re.compile(r"^(\d{4,5})\s{"+str(k)+"}[^ ].+\n")
matcher1 = re.findall(pattern1,line)
if matcher1!=[]:
flag = 1
m = k
# print m
w2.write("\"\n")
# for j in matcher1:
# w2.write(j)
flag = 1
w2.write("\"")
break
v = m+5
#print v
pattern2 = re.compile(r"^\s{"+str(v)+"}([^ ].+)\n")
matcher2 = re.findall(pattern2,line)
if (matcher2!=[]):
for j in matcher2:
w2.write(j)
w2.write(" ")
#防止匹配到下面结构中的行
pattern3 = re.compile(r"(:?4.3\s{4}Message\sstructure)|(:?Pos\s+Tag\sName\s+S\s+R)")
matcher3 = re.findall(pattern3,line)
if (matcher3!=[]):
break
w2.write("\"")
w2.close( )
#把第一行的“修改为note
old_file=filename_w
fopen=open(old_file,'r')
w_str=""
i =0
for line in fopen:
i =i+1
if ((re.search("\"",line)) and (i ==1)):
line=re.sub('\"','code_pos,note',line)
w_str+=line
else:
w_str+=line
# print w_str
wopen=open(old_file,'w')
wopen.write(w_str)
fopen.close()
wopen.close()
def join(list_tag):
for MM in range(len(list_tag)):
f1 = open(ss+'new/%s_w.txt'%list_tag[MM])
f2 = open(ss+'new/%s_wnote.txt'%list_tag[MM]) list_note=[]
for line1 in f1:
# print(line1) list_note.append(line1) f1.close() # print(list_note)
f2_w= open(ss+'new/b1%s.csv'%year,'a')
# for i in range(len(list_note)):
j=0
# f2_r = open(ss+'/new/%s_w.txt'%list_tag[MM])
for line2 in f2: str11="%s,%s\n"%(list_note[j].strip('\n'),line2.strip('\n'))
j=j+1
# print(i)
# print(str11)
f2_w.write(str11) f2.close()
f2_w.close() if __name__ == '__main__':
list_tag=get_tag()
trmd_b1_nonote(list_tag)
trmd_b1_note(list_tag)
join(list_tag) """
特殊情况 """
trmd_b1_ok的更多相关文章
随机推荐
- （Python）numpy的argmax用法
解释 还是从一维数组出发.看下面的例子. import numpy as np a = np.array([3, 1, 2, 4, 6, 1]) print(np.argmax(a))4 argm ...
- Delphi编程实现是否开启“平滑屏幕字体边缘“
在Windows高级设置中的视觉效果中可以设置是否开启“平滑屏幕字休边缘”,可以通过编程的方式来实现: if SystemParametersInfo(SPI_SETFONTSMOOTHING, 1, ...
- Hibernate迫切左外连接和迫切内连接
•迫切左外连接: •LEFT JOIN FETCH 关键字表示迫切左外连接检索策略. –list() 方法返回的集合中存放实体对象的引用, 每个 Department 对象关联的 Employee ...
- 吴裕雄 数据挖掘与分析案例实战（4）——python数据处理工具：Pandas
# 导入模块import pandas as pdimport numpy as np # 构造序列gdp1 = pd.Series([2.8,3.01,8.99,8.59,5.18])print(g ...
- HTML 练习 做简历表
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/ ...
- linux下面得小数计算
可以通过命令行向awk中传递参数. 这样子传递进去的参数,在awk命令中可以访问.每一项都必须作为单一的一个参数来解释.所以,等号之间不能有空格. 比如说我们传递进去一个学生名字,想查这个学生的得分. ...
- LIS问题（DP解法）---poj1631
题目链接:http://poj.org/problem?id=1631 这个题题目有些难看懂hhh,但实质就是求LIS--longest increasing sequence. 以下介绍LIS的解法 ...
- ECMAScript5新特性之获取对象特有的属性
'use strict'; // 父类 function Fruit(){ } Fruit.prototype.name = '水果'; // 子类 function Apple(desc){ thi ...
- TZOJ 1545 Hurdles of 110m(01背包dp)
描述 In the year 2008, the 29th Olympic Games will be held in Beijing. This will signify the prosperit ...
- google thumbnailator
Thumbnailator 是一个优秀的图片处理的Google开源Java类库.处理效果远比Java API的好. 从API提供现有的图像文件和图像对象的类中简化了处理过程,两三行代码就能够从现有图片 ...