"""
python字符串切片处理DB2 DEL文件断行问题实现不以引号开头的行与上一行连接.py
题目来源 http://www.bathome.net/thread-38164-1-1.html
依山居  17:47 2015/11/24
这个解决思路在脑子里绕了好几天~找到\n也就是换行符，同时切出它的下一位不为"号的字符。
也就是不以引号开头行的第一个字。然后再替换掉\n
感觉这个处理方法,不需要正则表达式可靠性也很高。
"""
import time
start=time.time()
print("运行中..."*3)

with open("a.txt") as f:
    txt=f.read()
#增加字符长度,取\n的下一位字符时就不用担心和考虑越界的问题
txt=txt+"###"
rn=len(txt)
#找到\n也就是换行符，同时切取出它的下一位不为"号的字符。
rs=[txt[r]+txt[r+1] for r in range(rn-3) if ("\n" in txt[r]) and ('\"' not in txt[r+1])]
#rs=['\n2', '\n5', '\n2', '\n3'...]
for r in rs:
    #设r为\n3,替换为r[1]也就是换为3，相当于删掉了\n。这个不以引开头的行就能与上行连接了。
    txt=txt.replace(r,r[1])
txt=txt.rstrip("###")+"\n"
with open("aa.txt","w+") as f:
    f.write(txt)

end=time.time()
pt=end-start
print("运行耗时：",pt)
try:
    input("按回车退出")
except SyntaxError:
    pass
复制代码

"""
生成测试数据
依山居 20:01 2015/11/24
"""
import time
start=time.time()
print("运行中..."*3)

a=""""1","1
2","123","","201
5-10-31",
"1","1
2","12
3","","",
"1","12","123","123
4"
"A","AB","ABC","ABCD"
"R","RS","RS
T","RSTU"
"E","EF","EFG","EFGH"
"c","EF","EFG","EFGH"
"d","EF","EFG","EFGH
xxx"
"f","EF","EFG","EFGH
13:05 2015/11/17"
"c","EF","EFG","EFGH"
"c","EF","EFG","EFGH"
"c","EF","EFG","EFGH"
"c","EF","EFG","EFGH"
"""

b=a*1000000
with open("a.txt","w") as f:
    f.write(b)

end=time.time()
pt=end-start
print("运行耗时：",pt)
try:
    input("按回车退出")
except SyntaxError:
    pass
复制代码

"""
python列表切片处理DEL文件断行.py
这版的代码是之前列表切片版的改进，目的是把DEL文件中不以引号开头的行与上一行连接
下一行不以引号开头则清除行后的空白字符，包括\n，在写入的时候就自动与上一行连接了。
http://www.bathome.net/thread-38164-1-1.html
"""

import time
start=time.time()
print("运行中..."*3)

txt=[]
with open("a.txt") as f:
    txt=f.readlines()

    end=time.time()
    pt=end-start
    print("readlines运行耗时：",pt)

    ln=len(txt)
    print("需要处理的文件行数：",ln)

    end=time.time()
    pt=end-start
    print("len(txt)运行耗时：",pt)
    #为了防指针越界，所以rn需要减1
    #flines=[txt[r] if ('\"' in txt[r+1][0]) else txt[r].rstrip() for r in range(ln-1)]
    #与上一行等效写法
    flines=[txt[r] if (txt[r+1].startswith('\"')) else txt[r].rstrip() for r in range(ln-1)]
	
end=time.time()
pt=end-start
print("flines运行耗时：",pt)

with open("out.txt","w+") as f:
    f.writelines(flines)
    #补上最后一行
    f.write(txt[-1])
    f.close()
    
txt[:]=[]
flines[:]=[]

end=time.time()
pt=end-start
print("运行耗时：",pt)
try:
    input("按回车退出")
except SyntaxError:
    pass

"""
测试结果，列表切片和模式匹配处理速度相当，670M的文本，大约60-70秒左右，
包括写入时间,算出处理速度大约是每秒处理10M左右。
实际发现内存占用峰值会高出好几倍，python 进程会短时占用好几个G内存的情况。

"""
复制代码