Hyppää sisältöön

Ei vielä käännetty

Tätä sivua ei ole vielä käännetty suomeksi, joten se näytetään englanniksi. Auta kääntämään

Fileinput Module Complexity

The fileinput module provides an iterator for processing lines from multiple input files in a uniform way.

Common Operations

Operation Time Space Notes
input(files) O(B) O(L) B = total bytes read; L = longest line held at a time
input() with backup O(B) O(L) In-place edit with backup; same streaming bounds as above
filename() O(1) O(1) Get current file name
filelineno() O(1) O(1) Get current line in file
lineno() O(1) O(1) Get total line count
fileno() O(1) O(1) Get current file descriptor
isfirstline() O(1) O(1) Check first line of file
isstdin() O(1) O(1) Check if current file is stdin
nextfile() O(1) O(1) Skip to next file
close() O(1) O(1) Close file iterator
FileInput(...) O(1) O(1) Create iterator instance
hook_encoded() O(1) O(1) Encoding open hook factory
hook_compressed() O(1) O(1) Compressed-file open hook factory

Basic Line Iteration

input()

Time Complexity: O(n)

Where n = total lines across all files.

import fileinput

# Iterate lines from multiple files
for line in fileinput.input(['file1.txt', 'file2.txt']):
    print(line.rstrip())  # O(n) to process all lines

# With stdin
for line in fileinput.input():
    print(line)  # O(n) for all input

Space Complexity: O(1)

import fileinput

# Lazy iteration - no buffering
for line in fileinput.input(['huge_file.txt']):
    # Process one line at a time
    process(line)  # O(1) memory per iteration

In-Place Editing

Backup Creation

Time Complexity: O(n)

import fileinput

# Modify files in-place with backup
for line in fileinput.input(['file1.txt', 'file2.txt'], 
                           backup='.bak', inplace=True):
    # Original backed up as file1.txt.bak
    print(line.upper().rstrip())  # Write to stdout -> redirected to file

Space Complexity: O(1)

import fileinput

# Streaming in-place edit
for line in fileinput.input(inplace=True):
    print(line.rstrip())  # O(1) memory

In-Place Without Backup

Time Complexity: O(n)

import fileinput

# Edit without backup
for line in fileinput.input(['file.txt'], inplace=True, 
                           backup=''):
    print(line.rstrip())  # No backup created

# Original file modified, no .bak

Space Complexity: O(1)

import fileinput

# Direct modification
for line in fileinput.input(['large_file.txt'], inplace=True, 
                           backup=''):
    process(line)  # O(1) memory

Line Number Tracking

linenum() and fileno()

Time Complexity: O(1)

import fileinput

# Track line numbers
for line in fileinput.input(['file1.txt', 'file2.txt']):
    total_line_num = fileinput.lineno()      # O(1)
    file_line_num = fileinput.filelineno()   # O(1)
    filename = fileinput.filename()          # O(1)

    print(f'{filename}:{file_line_num}: {line.rstrip()}')

Space Complexity: O(1)

import fileinput

# Metadata stored during iteration
for line in fileinput.input():
    num = fileinput.lineno()  # O(1) space - counter

Context Manager Usage

Time Complexity: O(n)

import fileinput

# Use as context manager (Python 3.10+)
with fileinput.input(['file1.txt', 'file2.txt']) as f:
    for line in f:
        print(line.rstrip())  # O(n) for all lines
# Auto-cleanup

Space Complexity: O(1)

import fileinput

with fileinput.input(['file.txt']) as f:
    for line in f:
        process(line)  # O(1) streaming

Common Patterns

Simple Line Processing

import fileinput

# Process lines from multiple files - O(B) in total bytes, O(L) space.
# Lines are the unit you iterate, not the unit of work: every byte is still
# scanned to find the line breaks. Files are opened one at a time and lines
# yielded on demand, so memory holds one line - which can be arbitrarily long
for line in fileinput.input(['input1.txt', 'input2.txt']):
    # Line already has newline
    if line.startswith('#'):
        continue  # Skip comments

    print(line.rstrip())  # Remove trailing newline, print

In-Place Text Replacement

import fileinput
import sys

# Replace pattern in-place
def replace_in_files(files, old, new):
    for line in fileinput.input(files, inplace=True, backup='.bak'):
        print(line.replace(old, new).rstrip())

# Usage
replace_in_files(['config.txt', 'data.txt'], 'old_value', 'new_value')
# Creates .bak files as backups

Count Lines in Multiple Files

import fileinput

# Count all lines
count = 0
for line in fileinput.input(['file1.txt', 'file2.txt']):
    count += 1
print(f'Total lines: {count}')  # O(n)

# Group by file
lines_per_file = {}
for line in fileinput.input(['file1.txt', 'file2.txt']):
    filename = fileinput.filename()
    lines_per_file[filename] = fileinput.filelineno()

Filter Lines Across Files

import fileinput

# Extract specific lines - O(B) over all bytes, O(L) for the current line
for line in fileinput.input(['file1.txt', 'file2.txt']):
    if 'ERROR' in line:
        filename = fileinput.filename()  # O(1)
        lineno = fileinput.lineno()      # O(1) - a running counter
        print(f'{filename}:{lineno}: {line.rstrip()}')

Add Line Numbers to Files

import fileinput

# O(B) in total bytes; in-place editing rewrites each file once
for line in fileinput.input(inplace=True):
    lineno = fileinput.lineno()  # O(1)
    print(f'{lineno:5d}: {line.rstrip()}')

# Reads from stdin, writes to stdout with line numbers
# Usage: python script.py < input.txt

Input Modes

Text Mode (Default)

import fileinput

# Read as text (default) - O(B) in total bytes, O(L) for the current line
# Decoding is part of the per-line cost; it does not change the complexity
for line in fileinput.input(['file.txt']):
    print(type(line))  # str
    print(line.rstrip())

Binary Mode (Not Directly Supported)

import fileinput

# fileinput works with text by default
# For binary, use regular file operations
with open('file.bin', 'rb') as f:
    for line in f:
        process(line)

Encoding Support

import fileinput

# Specify encoding (Python 3.10+)
# hook_encoded() is O(1) - it builds the opener, called once per file
for line in fileinput.input(['utf8_file.txt'], 
                           openhook=fileinput.hook_encoded("utf-8")):
    print(line.rstrip())  # Still O(B) in total bytes, O(L) per line

Performance Characteristics

Best Practices

import fileinput

# Good: Stream processing
for line in fileinput.input(['large_file.txt']):
    process(line)  # O(1) memory per line

# Good: Use context manager
with fileinput.input(['file.txt']) as f:
    for line in f:
        print(line.rstrip())

# Avoid: Loading all lines first
lines = []
for line in fileinput.input(['file.txt']):
    lines.append(line)  # O(n) memory

In-Place Editing Safety

import fileinput

# Good: Create backup
for line in fileinput.input(['config.txt'], 
                           inplace=True, backup='.bak'):
    print(line.rstrip())

# Risky: No backup
for line in fileinput.input(['config.txt'], 
                           inplace=True, backup=''):
    print(line.rstrip())  # Lost on error

Error Handling

import fileinput
import sys

def process_files_safely(files):
    try:
        # O(B) in total bytes; the try/except costs nothing until it raises
        for line in fileinput.input(files):
            try:
                process(line)
            except ValueError as e:
                filename = fileinput.filename()  # O(1)
                lineno = fileinput.lineno()      # O(1)
                print(f'Error {filename}:{lineno}: {e}', 
                      file=sys.stderr)
    finally:
        fileinput.close()

# Handle file not found
try:
    for line in fileinput.input(['nonexistent.txt']):
        pass
except FileNotFoundError as e:
    print(f'File error: {e}')

Comparison with Alternatives

import fileinput
from pathlib import Path

# fileinput (good for multiple files)
for line in fileinput.input(['file1.txt', 'file2.txt']):
    print(line.rstrip())  # O(n), O(1) memory

# pathlib (good for single operations)
for line in Path('file.txt').read_text().splitlines():
    print(line)  # O(n) time, O(n) memory

# direct open (good for single file)
with open('file.txt') as f:
    for line in f:
        print(line.rstrip())  # O(n) time, O(1) memory

# All similar time complexity, fileinput best for multiple files

Sys.argv Integration

import fileinput
import sys

# Automatically use command-line files or stdin
# O(B) in total bytes, O(L) per line - stdin streams the same way a file does
for line in fileinput.input():
    # Reads from files in sys.argv[1:] or stdin
    print(line.rstrip())

# Usage: python script.py file1.txt file2.txt
# Or:    python script.py < input.txt

Version Notes

  • Python 3.10+: Context manager support (with statement)
  • Python 3.10+: openhook parameter for encoding