Delete a row from a text file with Python

Question

I have a file where each line starts with a number. The user can delete a row by typing in the number of the row the user would like to delete.

The issue I\'m having

Answer 1

You could mmap the file... after haven read the suitable documentation...

Answer 2

def DeleteToDo():
    print ("Which Item Do You Want To Delete?")
    DeleteItem = raw_input(">") #select a line number to delete
    print ("Are You Sure You Want To Delete Number" + DeleteItem + "(y/n)")
    DeleteItem=int(DeleteItem) 
    VerifyDelete = str.lower(raw_input(">"))
    if VerifyDelete == "y":
        FILE = open('data.txt',"r") #open the file (tried w+ as well, entire file is deleted)
        lines=[x.strip() for x in FILE if int(x[:x.index('.')])!=DeleteItem] #read all the lines first except the line which matches the line number to be deleted
        FILE.close()
        FILE = open('data.txt',"w")#open the file again
        for x in lines:FILE.write(x+'\n')    #write the data to the file

    else:
        print ("Nothing Deleted")
DeleteToDo()

Answer 3

You don't need to check for the lines numbers in your file, you can do something like this:

def DeleteToDo(self):
    print "Which Item Do You Want To Delete?"
    DeleteItem = int(raw_input(">")) - 1
    print "Are You Sure You Want To Delete Number" + str(DeleteItem) + "(y/n)"
    VerifyDelete = str.lower(raw_input(">"))
    if VerifyDelete == "y":
        with open(ToDo.filename,"r") as f:
            lines = ''.join([a for i,a in enumerate(f) if i != DeleteItem])

        with open(ToDo.filename, "w") as f:
            f.write(lines)
    else:
        print "Nothing Deleted"

Answer 4

Instead of writing out all lines one by one to the file, delete the line from memory (to which you read the file using readlines()) and then write the memory back to disk in one shot. That way you will get the result you want, and you won't have to clog the I/O.

Answer 5

When you open a file for writing, you clobber the file (delete its current contents and start a new file). You can find this out by reading documentation for the open() command.

When you open a file for appending, you do not clobber the file. But how can you delete just one line? A file is a sequence of bytes stored on a storage device; there is no way for you to delete one line and have all the other lines automatically "slide down" into new positions on the storage device.

(If your data was stored in a database, you could actually delete just one "row" from the database; but a file is not a database.)

So, the traditional way to solve this: you read from the original file, and you copy it to a new output file. As you copy, you perform any desired edits; for example, you can delete a line simply by not copying that one line; or you can insert a line by writing it in the new file.

Then, once you have successfully written the new file, and successfully closed it, if there is no error, you go ahead and rename the new file back to the same name as the old file (which clobbers the old file).

In Python, your code should be something like this:

import os

# "num_to_delete" was specified by the user earlier.

# I'm assuming that the number to delete is set off from
# the rest of the line with a space.

s_to_delete = str(num_to_delete) + ' '
def want_input_line(line):
    return not line.startswith(s_to_delete)

in_fname = "original_input_filename.txt"
out_fname = "temporary_filename.txt"

with open(in_fname) as in_f, open(out_fname, "w") as out_f:
    for line in in_f:
        if want_input_line(line):
            out_f.write(line)

os.rename(out_fname, in_fname)

Note that if you happen to have a file called temporary_filename.txt it will be clobbered by this code. Really we don't care what the filename is, and we can ask Python to make up some unique filename for us, using the tempfile module.

Any recent version of Python will let you use multiple statements in a single with statement, but if you happen to be using Python 2.6 or something you can nest two with statements to get the same effect:

with open(in_fname) as in_f:
    with open(out_fname, "w") as out_f:
        for line in in_f:
            ... # do the rest of the code

Also, note that I did not use the .readlines() method to get the input lines, because .readlines() reads the entire contents of the file into memory, all at once, and if the file is very large this will be slow or might not even work. You can simply write a for loop using the "file object" you get back from open(); this will give you one line at a time, and your program will work with even really large files.

EDIT: Note that my answer is assuming that you just want to do one editing step. As @jdi noted in comments for another answer, if you want to allow for "interactive" editing where the user can delete multiple lines, or insert lines, or whatever, then the easiest way is in fact to read all the lines into memory using .readlines(), insert/delete/update/whatever on the resulting list, and then only write out the list to a file a single time when editing is all done.