读取一行将其存储在一个变量中，然后读取另一行并返回第一行。蟒蛇2

Question

这是一个棘手的问题，我已经阅读了很多关于它的帖子，但我无法让它发挥作用。

我有一个大文件。我需要逐行阅读它，一旦我到达"Total is: (any decimal number)" 形式的行，就取这个字符串并将数字保存在一个变量中。如果数字大于 40.0，那么我需要找到 Total 行上方的第四行（例如，如果 Total 行是第 39 行，则此行将是第 35 行）。该行的格式为"(number).(space)(substring)"。最后，我需要将这个子字符串解析出来并对其进行进一步处理。

这是一个输入文件的示例：

many lines that we don't care about
many lines that we don't care about
...
1. Hi45
People: bla bla bla bla bla bla
whitespace
bla bla bla bla bla
Total is: (*here there will be a decimal number*)
bla bla
white space
...
more lines we don't care about
and then more lines and then
again we get
2. How144
People: bla bla bla bla bla bla
whitespace
bla bla bla bla bla
Total is: (*here there will be a decimal number*)
bla bla
white space

我尝试了很多东西，包括使用re.search() 方法从我需要关注的每一行中捕获我需要的内容。

这是我从另一个 stackoverflow 问答中修改的代码：

import re
import linecache
number = ""
higher_line = ""
found_line = ""

with open("filename_with_many_lines.txt") as aFile:
    for num, line in enumerate(aFile, 1):
        searchObj = re.search(r'(\bTotal\b)(\s)(\w+)(\:)(\s)(\d+.\d+)', line)
        if searchObj:
            print "this is on line", line
            print "this is the line number:", num
            var1 = searchObj.group(6)
            print var1
            if float(var1) > 40.0:
                number = num
                higher_line = number - 4
                print number
                print higher_line

                found_line = linecache.getline("filename_with_many_lines.txt", higher_line)
                print "found the", found_line

预期的输出是：

this is on line Total is: 45.5
this is the line number: 14857
14857
14853
found the 1. Hi145
this is on line Total is: 62.1
this is the line number: 14985
14985
14981
found the 2.How144

Answer 1

如果您需要的行总是在Total is: 行上方四行，您可以将前面的行保留在有界deque 中。

from collections import deque

with open(filename, 'r') as file:
    previous_lines = deque(maxlen=4)
    for line in file:
        if line.startswith('Total is: '):
            try:
                higher_line = previous_lines[-4]
                # store higher_line, do calculations, whatever
                break  # if you only want to do this once; else let it keep going
            except IndexError:
                # we don't have four previous lines yet
                # I've elected to simply skip this total line in that case
                pass
        previous_lines.append(line)

有界deque（具有最大长度的）如果添加新项目会导致其超过其最大长度，则会丢弃另一侧的项目。在这种情况下，我们将字符串附加到deque 的右侧，因此一旦deque 的长度达到4，我们附加到右侧的每个新字符串都会导致它丢弃一个字符串左侧。因此，在for 循环的开头，deque 将包含当前行之前的四行，最旧的行位于最左侧（索引0）。

事实上，the documentation on collections.deque 提到的用例与我们的非常相似：

有限长度双端队列提供类似于 Unix 中的 tail 过滤器的功能。它们还可用于跟踪仅关注最近活动的交易和其他数据池。

Answer 2

这会将以数字和点开头的行存储到名为prevline 的变量中。仅当re.search 返回匹配对象时，我们才打印prevline。

import re
with open("file") as aFile:
    prevline = ""
    for num, line in enumerate(aFile,1):
        m = re.match(r'\d+\.\s*.*', line)                                # stores the match object of the line which starts with a number and a dot
        if m:                                              
            prevline += re.match(r'\d+\.\s*(.*)', line).group()         # If there is any match found then this would append the whole line to the variable prevline. You could also write this line as prevline += m.group()

        searchObj = re.search(r'(\bTotal\b\s+\w+:\s+(\d+\.\d+))', line)  # Search for the line which contains the string Total plus a word plus a colon and a float number
        if searchObj:                                                   # if there is any
            score = float(searchObj.group(2))                           # then the float number is assigned to the variable called score
            if score > 40.0:                                            # Do all the below operations only if the float number we fetched was greater than 40.0
                print "this is the line number: ", num
                print "this is the line", searchObj.group(1)
                print num
                print num-4
                print "found the", prevline
                prevline = ""

输出：

this is on line Total is: 45.5
this is the line number:  8
8
4
found the 1. Hi45
this is on line Total is: 62.1
this is the line number:  20
20
16
found the 2. How144

Answer 3

我建议对 Blacklight Shining 的帖子进行编辑，该帖子基于其 deque 解决方案，但被拒绝，建议改为将其作为答案。下面，我将展示 Blacklight 的解决方案如何解决您的问题，如果您只是盯着它看一会儿。

with open(filename, 'r') as file:
    # Clear: we don't care about checking the first 4 lines for totals.
    # Instead, we just store them for later.
    previousLines = []
    previousLines.append(file.readline())
    previousLines.append(file.readline())
    previousLines.append(file.readline())
    previousLines.append(file.readline())

    # The earliest we should expect a total is at line 5.
    for lineNum, line in enumerate(file, 5):
        if line.startswith('Total is: '):
            prevLine = previousLines[0]
            high_num = prevLine.split()[1] # A
            score = float(line.strip("Total_is: ").strip("\n").strip()) # B

            if score > 40.0:
                # That's all! We've now got everything we need.
                # Display results as shown in example code.
                print "this is the line number : ", lineNum
                print "this is the line ", line.strip('\n')
                print lineNum
                print (lineNum - 4)
                print "found the ", prevLine

        # Critical - remove old line & push current line onto deque.
        previousLines = previousLines[1:] + [line]

我没有利用deque，但我的代码命令式地完成了同样的事情。我认为这不一定是比其他任何一个更好的答案。我发布它是为了展示如何使用非常简单的算法和简单的工具来解决您尝试解决的问题。 （将 Avinash 巧妙的 17 线解决方案与我简化的 18 线解决方案进行比较。）

这种简化的方法不会让任何阅读您的代码的人看起来像个向导，但它也不会意外匹配中间行中的任何内容。如果您对使用正则表达式不满意，那么只需修改 A 和 B 行。通用解决方案仍然有效。

重点是，记住第 4 行后面的内容的一种简单方法是将最后四行存储在内存中。