#!/usr/bin/python
# -*- coding: utf-8 -*-
# Setup VIM: ex: noet ts=2 sw=2 :
#
# Main Script doing the conversion.
# See README for details.
#
# Author: Elan Ruusamäe <glen@pld-linux.org>
# Version: 1.0

import sys, os, os.path, re, codecs
import getopt
from MoinMoin import user, wikiutil
from MoinMoin.web.contexts import ScriptContext as RequestCLI
from MoinMoin.logfile import editlog
from MoinMoin.Page import Page
from shutil import copyfile, copystat
from os import listdir, mkdir
from os.path import isdir, basename
from doku import DokuWiki
from moinformat import moin2doku
import random

# sys.setdefaultencoding() does not exist, here!
reload(sys)  # Reload does the trick!
sys.setdefaultencoding('cp1252')

USEC = 1000000

def init_dirs(output_dir):
	if not isdir(output_dir):
		print >> sys.stderr, "Output directory doesn't exist!"
		sys.exit(1)

	pagedir = os.path.join(output_dir, 'pages')
	if not isdir(pagedir):
		mkdir(pagedir)

	mediadir = os.path.join(output_dir, 'media')
	if not isdir(mediadir):
		mkdir(mediadir)

	metadir = os.path.join(output_dir, 'meta')
	if not isdir(metadir):
		mkdir(metadir)

def readfile(filename):
	f = open(filename, 'r')
	text = f.read()
	return unicode(text.decode('utf-8'))

def writefile(filename, content, overwrite=False):
	dir = os.path.dirname(os.path.abspath(filename))
	if not isdir(dir):
		os.makedirs(dir);

	if os.path.exists(filename) and overwrite == False:
		raise (OSError, 'File already exists: %s' % filename)

	# ensure it's a list
	if not isinstance(content, (list, tuple)):
		content = [content]

	f = codecs.open(filename, 'w', 'utf-8')
	f.writelines([line + u'\n' for line in content])
	f.close()


def encode_relaxed(text):
	return text.encode("ascii", errors="ignore")


# page = MoinMoin Page oject
# ns = DokuWiki namespace where attachments to copy
def copy_attachments(page, ns,randomID):
	srcdir = page.getPagePath('attachments', check_create = 0)
	if not isdir(srcdir):
		return

	attachment_dir = os.path.join(output_dir, 'media', dw.mediaFN(ns))
	if not isdir(attachment_dir):
		os.makedirs(attachment_dir);

	attachments = listdir(srcdir)
	for attachment in attachments:
		try:
			src = os.path.join(srcdir, attachment)
			dst = os.path.join(output_dir, 'media', dw.mediaFN(dw.cleanID(u"%s/%s" % (ns, str(randomID)+attachment))))
			copyfile(src, dst)
			copystat(src, dst)
		except UnicodeDecodeError:
			print 'ERROR: unable to convert attachment "%s"' % encode_relaxed(attachment)

def print_help():
	program = sys.argv[0]
	print "Usage: %s OPTIONS" % program
	print "Convert MoinMoin pages to DokuWiki."
	print "Options:"
	print "-d DIR  - output directory"
	print "-a      - Convert Attic pages (history)"
	print "-f      - overwrite output files"
	print "-F FILE - convert single file"
	print "-r FILE - write config for redirect plugin"
	print ""
	print "%s -a -d /var/lib/dokuwiki" % program
	print "%s -F moinmoin/data/pages/frontpage -d out" % program
	sys.exit(0)

# return unicode encoded wikiname
# input is a dir from moinmoin pages/ dir
def wikiname(filename):
	return wikiutil.unquoteWikiname(basename(filename))

def convert_editlog(page, output = None, overwrite = False):
	pagedir = page.getPagePath()
	pagename = wikiname(pagedir)
	if not output:
		output = pagename
	edit_log = editlog.EditLog(request, page.getPagePath('edit-log'))

	changes = {}
	for log in edit_log:
		# not supported. perhaps add anyway?
		if log.action in ('ATTNEW', 'ATTDEL', 'ATTDRW'):
			continue

		# 1201095949  192.168.2.23    E   start   glen@delfi.ee
		author = log.hostname
		if log.userid:
			userdata = user.User(request, log.userid)
			if userdata.name:
				author = userdata.name

		try:
			action = {
				'SAVE' : 'E',
				'SAVENEW' : 'C',
				'SAVE/REVERT' : 'R',
			}[log.action]
		except KeyError:
			action = log.action

		mtime = str(log.ed_time_usecs / USEC)
		changes[mtime] = u"\t".join([mtime, log.addr, action, dw.cleanID(log.pagename), author, log.comment])

	# see if we have missing entries, try to recover
	page = Page(request, pagename)
	if len(page.getRevList()) != len(changes):
		print "RECOVERING edit-log, missing %d entries" % (len(page.getRevList()) - len(changes))
		for rev in page.getRevList():
			page = Page(request, pagename, rev = rev)
			mtime = page.mtime_usecs() / USEC

			if not mtime:
				pagefile, realrev, exists = page.get_rev(rev = rev);
				if os.path.exists(pagefile):
					mtime = int(os.path.getmtime(pagefile))
					print "Recovered %s: %s" % (rev, mtime)

			mtime = str(mtime)
			if not changes.has_key(mtime):
				changes[mtime] = u"\t".join([mtime, '127.0.0.1', '?', dw.cleanID(pagename), 'root', 'recovered entry'])
				print "ADDING %s" % mtime

	changes = sorted(changes.values())
	out_file = os.path.join(output_dir, 'meta', dw.metaFN(output, '.changes'))
	writefile(out_file, changes, overwrite = overwrite)

def convertfile(page, output = None, overwrite = False):
	pagedir = page.getPagePath()
	pagename = wikiname(pagedir)

	if not output:
		output = pagename

	print "Converting %s" % encode_relaxed(pagename)

	if page.isUnderlayPage():
		print "underlay: %s" % page.request.cfg.data_underlay_dir
		print "underlay: %s" % request.cfg.data_underlay_dir
		print "SKIP UNDERLAY: %s" % encode_relaxed(pagename)
		return False

	current_exists = page.exists()
	current_rev = page.current_rev()

	if convert_attic:
		revs = page.getRevList()
	else:
		revs = [current_rev]

	# Generate random ID Number for collision avoidance when attachments in Namespace have the same name
	randomID = random.randint(101,999)

	for rev in revs:
		page = Page(request, pagename, rev = rev)
		pagefile, realrev, exists = page.get_rev(rev = rev);

		mtime = page.mtime_usecs() / USEC

		if not mtime:
			if os.path.exists(pagefile) != exists:
				raise Exception, "IT SHOULD NOT HAPPEN"

			if os.path.exists(pagefile):
				mtime = int(os.path.getmtime(pagefile))
				print "recovered %s: %s" % (rev, mtime)

			if not mtime:
				print "NO REVISION: for %s" % encode_relaxed(pagefile)
				continue

		if rev == current_rev:
			out_file = os.path.join(output_dir, 'pages', dw.wikiFN(output))
			if not convert_attic and not exists:
				# if not converting attic, allow current version may not exist anymore
				continue
		else:
			out_file = os.path.join(output_dir, 'attic', dw.wikiFN(output, str(mtime)))

		content = moin2doku(pagename, page.get_raw_body(),randomID)
		if len(content) == 0:
#			raise Exception, "No content"
			print "NO CONTENT: exists: %s,%s" % (exists, os.path.exists(pagefile))

		writefile(out_file, content, overwrite = overwrite)
		copystat(pagefile, out_file)

	ID = dw.cleanID(output)
	copy_attachments(page, dw.getNS(ID),randomID)

	# convert edit-log, it's always present even if current page is not
	convert_editlog(page, output = output, overwrite = overwrite)

	# add to redirect.conf if filenames differ
	# and page must exist (no redirect for deleted pages)
	if redirect_conf and current_exists:
		# redirect dokuwiki plugin is quite picky
		# - it doesn't understand if entries are not lowercase
		# - it doesn't understand if paths are separated by forward slash
		old_page = pagename.lower().replace('/', ':').replace(' ', '_')
		if old_page != ID:
			redirect_map[old_page] = ID

	print "Converted %s as %s" % (encode_relaxed(pagename), dw.wikiFN(output))

	return True

#
# "main" starts here
#

# setup utf8 output
if sys.stdout.isatty():
	default_encoding = sys.stdout.encoding
else:
	import locale
	default_encoding = locale.getpreferredencoding()
sys.stdout = codecs.getwriter(default_encoding)(sys.stdout);

try:
	opts, args = getopt.getopt(sys.argv[1:], 'hfad:p:r:i:I:', [ "help" ])
except getopt.GetoptError, e:
	print >> sys.stderr, 'Incorrect parameters! Use --help switch to learn more.: %s' % e
	sys.exit(1)

overwrite = False
convert_page = None
output_dir = None
convert_attic = False
redirect_conf = False
redirect_map = {}
page_filter = []
for o, a in opts:
	if o == "--help" or o == "-h":
		print_help()
	if o == "-f":
		overwrite = True
	if o == "-a":
		convert_attic = True
	if o == "-r":
		redirect_conf = a
	if o == "-i":
		page_filter.append(a)
	if o == "-I":
		page_filter.extend(readfile(a).split("\n"))
	if o == "-d":
		output_dir = a
	if o == "-p":
		convert_page = a

if not output_dir:
	print_help()
	sys.exit(1)

print "Output dir is: '%s'" % output_dir
init_dirs(output_dir)

dw = DokuWiki()
request = RequestCLI()
pages = {}

if convert_page != None:
	pagename = wikiname(convert_page)
	pages[pagename] = pagename
else:
	filter = None
	if page_filter:
		def name_filter(name):
			return name not in page_filter
		filter = name_filter

	# get list of all pages in wiki
	# hide underlay dir temporarily
	underlay_dir = request.rootpage.cfg.data_underlay_dir
	print(underlay_dir)
	request.rootpage.cfg.data_underlay_dir = None
	pages = request.rootpage.getPageList(user = '', exists = not convert_attic, filter = filter)
	pages = dict(zip(pages, pages))
	# restore
	request.rootpage.cfg.data_underlay_dir = underlay_dir

	# insert frontpage,
	# so that MoinMoin frontpage gets saved as DokuWiki frontpage based on their configs
	frontpage = wikiutil.getFrontPage(request)
	if pages.has_key(frontpage.page_name):
		del pages[frontpage.page_name]
	pages[dw.getId()] = frontpage.page_name

print "--------------------------------------------------"
for output, pagename in pages.items():
	print " - %s" % encode_relaxed(pagename)
print "--------------------------------------------------"

converted = 0
for output, pagename in pages.items():
	page = Page(request, pagename)
	res = convertfile(page, output = output, overwrite = overwrite)
	if res != None:
		converted += 1
print "Processed %d files, converted %d" % (len(pages), converted)

if redirect_conf:
	print "Writing %s: %d items" % (redirect_conf, len(redirect_map))
	content = [u"\t".join(pair) for pair in redirect_map.items()]
	writefile(redirect_conf, content, overwrite = overwrite)