Downloading a file into memory

I am writing a python script and I just need the second line of a series of very small text files. I would like to extract this without saving the file to my harddrive as I currently do.

I have 开发者_JAVA技巧found a few threads that reference the TempFile and StringIO modules but I was unable to make much sense of them.

Currently I download all of the files and name them sequentially like 1.txt, 2.txt, etc, then go through all of them and extract the second line. I would like to open the file grab the line then move on to finding and opening and reading the next file.

Here is what I do currently with writing it to my HDD:

while (count4 <= num_files):
    file_p = [directory,str(count4),'.txt']
    file_path = ''.join(file_p)        
    cand_summary = string.strip(linecache.getline(file_path, 2))
    linkFile = open('Summary.txt', 'a')
    linkFile.write(cand_summary)
    linkFile.write("\n")
    count4 = count4 + 1
    linkFile.close()

---

Just replace the file writing with a call to append() on a list. For example:

summary = []
while (count4 <= num_files):
    file_p = [directory,str(count4),'.txt']
    file_path = ''.join(file_p)        
    cand_summary = string.strip(linecache.getline(file_path, 2))
    summary.append(cand_summary)
    count4 = count4 + 1

As an aside you would normally write count += 1. Also it looks like count4 uses 1-based indexing. That seems pretty unusual for Python.

---

You open and close the output file in every iteration.

Why not simply do

with open("Summary.txt", "w") as linkfile:
    while (count4 <= num_files):
        file_p = [directory,str(count4),'.txt']
        file_path = ''.join(file_p)        
        cand_summary = linecache.getline(file_path, 2).strip() # string module is deprecated
        linkFile.write(cand_summary)
        linkFile.write("\n")
        count4 = count4 + 1

Also, linecache is probably not the right tool here since it's optimized for reading multiple lines from the same file, not the same line from multiple files.

Instead, better do

with open(file_path, "r") as infile:
    dummy = infile.readline()
    cand_summary = infile.readline.strip()

Also, if you drop the strip() method, you don't have to re-add the \n, but who knows why you have that in there. Perhaps .lstrip() would be better?

Finally, what's with the manual while loop? Why not use a for loop?

Lastly, after your comment, I understand you want to put the result in a list instead of a file. OK.

All in all:

summary = []
for count in xrange(num_files):
    file_p = [directory,str(count),'.txt'] # or count+1, if you start at 1
    file_path = ''.join(file_p)        
    with open(file_path, "r") as infile:
        dummy = infile.readline()
        cand_summary = infile.readline().strip()
        summary.append(cand_summary)