Python to extract data from a file

Question

I am trying to extract the text between that has specific text file:

----
data1
data1
data1
extractme
----
data2
data2
data2
----
data3
data3
extractme
---         


        

Answer 1

This works well enough for me. Your sample data is in a file called "data.txt" and the output goes to "result.txt"

inFile = open("data.txt")
outFile = open("result.txt", "w")
buffer = []
keepCurrentSet = True
for line in inFile:
    buffer.append(line)
    if line.startswith("----"):
        #---- starts a new data set
        if keepCurrentSet:
            outFile.write("".join(buffer))
        #now reset our state
        keepCurrentSet = False
        buffer = []
    elif line.startswith("extractme"):
        keepCurrentSet = True
inFile.close()
outFile.close()

Answer 2

For Python2

#!/usr/bin/env python

with open("infile.txt") as infile:
    with open("outfile.txt","w") as outfile:
        collector = []
        for line in infile:
            if line.startswith("----"):
                collector = []
            collector.append(line)
            if line.startswith("extractme"):
                for outline in collector:
                    outfile.write(outline)

For Python3

#!/usr/bin/env python3

with open("infile.txt") as infile, open("outfile.txt","w") as outfile:
    collector = []
    for line in infile:
        if line.startswith("----"):
            collector = []
        collector.append(line)
        if line.startswith("extractme"):
            for outline in collector:
                outfile.write(outline)

Answer 3

I imagine the change in number of dashes (4 in the input, sometimes 4 and sometimes 3 in the output) is an error and not actually desired (since no algorithm is even hinted at, to explain how many dashes are to be output on different occasions).

I would structure the task in terms of reading and yielding one block of lines at a time:

def readbyblock(f):
  while True:
      block = []
      for line in f:
          if line = '----\n': break
          block.append(line)
      if not block: break
      yield block

so that the (selective) output can be neatly separated from the input:

with open('infile.txt') as fin:
    with open('oufile.txt', 'w') as fou:
        for block in readbyblock(fin):
            if 'extractme\n' in block:
                fou.writelines(block)
                fou.write('----\n')

This is not optimal, performance-wise, if the blocks are large, since it has a separate loop on all lines in the block implied in the if clause. So, a good refactoring might be:

def selectivereadbyblock(f, marker='extractme\n'):
  while True:
      block = []
      extract = False
      for line in f:
          if line = '----\n': break
          block.append(line)
          if line==marker: extract = True
      if not block: break
      if extract: yield block

with open('infile.txt') as fin:
    with open('oufile.txt', 'w') as fou:
        for block in selectivereadbyblock(fin):
            fou.writelines(block)
            fou.write('----\n')

Parameterizing the separators (now hard-coded as '----\n' for both input and output) is another reasonable coding tweak.

Answer 4

data=open("file").read().split("----")
print '----'.join([ i for i in data if "extractme" in i ])