#!/usr/bin/python
# Recursively list FTP directory contents
#
# Copyright (C) 2013 Peter Wu <lekensteyn@gmail.com>

# Prepend directory name to files and list files:
# awk '/^Directory/{if($0 != "Directory /")dir=substr($0, index($0, "/"))}/^-/{n=index($0,$9);print substr($0,1, n - 1) dir "/" substr($0,n)}'

from __future__ import print_function

import sys
from ftplib import FTP, all_errors
import re
from datetime import datetime, timezone, date

# None = auto-detect (FEAT), True = use LIST, False = use MLSD
use_list = None

user = "anonymous"
passwd = "anon"
host = None
port = 21
path = "/"

patt_url = re.compile(r"""
^
(?:ftp://)?
(?:
	(?P<user>.+?)
	(?:
		:(?P<passwd>.+?)
	)?
	@
)?
(?P<host>[a-zA-Z0-9.-]+)
(:(?P<port>\d+))?
(?P<path>/.*)?
$
""", re.VERBOSE)

if len(sys.argv) >= 2:
	m = patt_url.match(sys.argv[1])
	if m:
		if m.group("user") is not None:
			user = m.group("user")
		if m.group("passwd") is not None:
			passwd = m.group("passwd")
		host = m.group("host")
		if m.group("port") is not None:
			port = int(m.group("port"))
		if m.group("path") is not None:
			path = m.group("path")

if host is None:
	print("Usage: python", sys.argv[0],
		"[ftp://][user[:pass]@]ftp.example.com[:21][/path]",
		file=sys.stderr)
	sys.exit(1)

# directories for which the contents needs to be retrieved
dirs = []
dirs.append(path)

patt_list = re.compile(r"""
^
.[-rwxXst]{9}\ + # permissions
\d+\ +          # links count
\S+\ +          # owner
\S+\ +          # group
\d+\ +          # size
\S{3}\ +        # month
\d{2}\ +        # day
(?:\d{4,}|\d{2}:\d{2})  # year or time
\ (.+?)         # name
(?:\ ->\ (.+))? # symlink
[\r\n]*         # line contains
$
""", re.VERBOSE)

def queue_dir(cwd, name):
	if cwd == "/":
		dir = "/" + name
	else:
		dir = cwd + "/" + name
	print("Queuing", dir, file=sys.stderr)
	dirs.append(dir)

# LIST handler
def get_ls_processor(cwd):
	def process_dir(line):
		print(line)
		if line[0] == 'd':
			m = patt_list.match(line)
			name = m.group(1)
			queue_dir(cwd, name)
#		elif line[0] == 'l':
#			m = patt_list.match(line)
#			print("Link", m.group(2))
	return process_dir

# MLSD handler
def format_perms(mode):
	"""Turns a numeric UNIX mode into human-readable form
	"""
	str = ""
	for i in range(0, 3):
		o = 0o100 >> (3 * i)
		str += "r" if mode & (4 * o) else "-"
		str += "w" if mode & (2 * o) else "-"
		if mode & (0o4000 >> i): # setuid, setgid or sticky bit
			if o == 0o001: # "world"
				str += "t" if mode & o else "T"
			else:
				str += "s" if mode & o else "S"
		else:
			str += "x" if mode & (1 * o) else "-"
	return str

def format_type_fact(type):
	if type == "file":
		return "-"
	elif type in ("cdir", "pdir", "dir"):
		return "d"
	# TODO: handle OS.name=type
	else:
		return "?"

def dt_from_ftp(timeval):
	timeval = timeval.split(".")[0]
	return datetime.strptime(timeval, "%Y%m%d%H%M%S").replace(tzinfo=timezone.utc)

def format_mlsd(name, facts):

	if "type" in facts:
		mode_desc = format_type_fact(facts["type"])
	else:
		mode_desc = "?"

	if "unix.mode" in facts:
		perm = int(facts["unix.mode"], 8)
		mode_desc += format_perms(perm)
	else:
		mode_desc += "?" * 9

	user = "?" if not "unix.owner" in facts else facts["unix.owner"]
	group = "?" if not "unix.group" in facts else facts["unix.group"]
	size = "" if not "size" in facts else int(facts["size"])

	modtime = 0 if not "modify" in facts else dt_from_ftp(facts["modify"])
	if date.today().year == modtime.year:
		date_str = modtime.strftime("%b %d %H:%M")
	else:
		date_str = modtime.strftime("%b %d  %Y")

	line = mode_desc + " "
	#line += " {links:4s}".format(links=-1)
	line += " {user:8s} {group:8s} {size:8}".format(user=user, group=group, size=size)
	line += " " + date_str + " " + name
	return line

def get_features(ftp):
	"""Returns supported FTP features (FEAT) in a list"""
	# cut "211-Features" and "211 End"
	feats = ftp.sendcmd("FEAT").splitlines()[1:-1]
	return [ f.strip() for f in feats ]

with FTP() as ftp:
	ftp.connect(host, port)
	ftp.login(user, passwd)
	print(ftp.getwelcome(), file=sys.stderr)
	feats = get_features(ftp)

	if use_list is None:
		# use list only if supported
		use_list = not any(f.startswith("MLST ") for f in feats)

	while dirs:
		dir = dirs.pop()
		try:
			ftp.cwd(dir)
		except all_errors as e:
			print("Failed to chdir to " + dir + ":" + str(e), file=sys.stderr)
			continue
		print("Directory", ftp.pwd())
		if use_list:
			ftp.retrlines('LIST', get_ls_processor(dir))
		else:
			for name, facts in ftp.mlsd():
				if name not in (".", ".."):
					print(format_mlsd(name, facts))
					if facts["type"] == "dir":
						queue_dir(dir, name)
		print("Queue size:", len(dirs), file=sys.stderr)