#! /usr/bin/env python3
# -.- coding: utf-8 -.-

# Zeitgeist
#
# Copyright © 2009-2010 Markus Korn <thekorn@gmx.de>
# Copyright © 2010 Mikkel Kamstrup Erlandsen <mikkel.kamstrup@gmail.com>
# Copyright © 2010 Canonical Ltd.
#             By Mikkel Kamstrup Erlandsen <mikkel.kamstrup@gmail.com>
# Copyright © 2011 Collabora Ltd.
#             By Siegfried-Angel Gevatter Pujals <siegfried@gevatter.com>
#             By Seif Lotfy <seif@lotfy.com>
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU Lesser General Public License as published by
# the Free Software Foundation, either version 2.1 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU Lesser General Public License for more details.
#
# You should have received a copy of the GNU Lesser General Public License
# along with this program.  If not, see <http://www.gnu.org/licenses/>.

import os
import re
import sys
import glob
import codecs
from io import StringIO
import collections
import argparse

import rdflib
from rdflib import RDF, RDFS
from rdflib.plugin import register

from rdflib.serializer import Serializer
from rdflib.parser import StringInputSource
from rdflib.namespace import Namespace

NIENS = Namespace("http://www.semanticdesktop.org/ontologies/2007/01/19/nie#")

class SymbolCollection(dict):

	closed = False

	_by_namespace = None

	def __init__(self):
		self._by_namespace = collections.defaultdict(lambda: [])

	def register(self, uri, parents, display_name, doc):
		assert not self.closed
		symbol = Symbol(self, uri, parents, display_name, doc)
		self[uri] = symbol

	def post_process(self):
		self.closed = True
		symbols = list(self.values())
		for symbol in symbols:
			for (i, parent) in enumerate(symbol.parents):
				symbol.parents[i] = self[parent]
			self._by_namespace[symbol.namespace].append(symbol)

	def iter_by_namespace(self):
		return list(self._by_namespace.items())

	def debug_print(self):
		symbols = list(self.registered_symbols.values())
		for symbol in symbols:
			symbol.debug_print()
			print()

class Symbol(object):

	name = None
	namespace = None
	uri = None
	parents = None
	display_name = None
	doc = None

	_collection = None
	_children = None
	_all_children = None

	def __init__(self, collection, uri, parents, display_name, doc):
		self._collection = collection
		self.uri = str(uri)
		self.namespace, self.name = self.uri[self.uri.rfind('/')+1:].split('#')
		self.name = Utils.camel2upper(self.name)
		self.namespace = self.namespace.upper()
		self.parents = [str(parent) for parent in parents]
		self.display_name = str(display_name) if display_name is not None \
			else self.name
		self.doc = str(doc)

	@property
	def children(self):
		""" Return all direct children of this Symbol. """
		if self._children is None:
			childs = set()
			values = list(self._collection.values())
			for symbol in values:
				if self in symbol.parents:
					childs.add(symbol)
			self._children = childs
		return self._children

	@property
	def all_children(self):
		""" Return all children of this Symbol, recursively. """
		if self._all_children is None:
			all_children = set()
			for symbol in self.children:
				all_children.update([symbol])
				all_children.update(symbol.all_children)
			self._all_children = all_children
		return self._all_children

	def debug_print(self):
		print("Name: %s" % self.name)
		print("  URI: %s" % self.uri)
		print("  Display Name: %s" % self.display_name)
		print("  Parents: %s" % ', '.join([str(p) for p in self.parents]))
		doc = self.doc if len(self.doc) <= 50 else "%s..." % self.doc[:47]
		print("  Description: %s" % doc)

	def __str__(self):
		return self.name

	def __doc__(self):
		return self.doc

	def __cmp__(self, other):
		return cmp(self.namespace, other.namespace) or \
			cmp(self.name, other.name)

	def __lt__(self, other):
		return self.name < other.name

	def __hash__(self):
		return self.uri.__hash__()

class Utils(object):

	@staticmethod
	def escape_chars(text, quotes='"', strip=True):
		assert len(quotes) == 1
		text = text.replace('%s' % quotes, '\\%s' % quotes)
		if strip:
			text = text.strip()
		return text

	@staticmethod
	def camel2upper(name):
		"""
		Convert CamelCase to CAMEL_CASE
		"""
		result = ""
		for i in range(len(name) - 1) :
			if name[i].islower() and name[i+1].isupper():
				result += name[i].upper() + "_"
			else:
				result += name[i].upper()
		result += name[-1].upper()
		return result

	@staticmethod
	def replace_items(item_set, item_map):
		if not item_set:
			return
		for item, value in list(item_map.items()):
			try:
				item_set.remove(item)
			except KeyError:
				# item is not in set
				continue
			else:
				# item was in set, replace it with value
				item_set.add(value)

	@staticmethod
	def indent(text, indentation):
		return re.sub(r'(?m)^(.+)$', r'%s\1' % indentation, text)

class OntologyParser(object):

	symbols = None

	def __init__(self, directory):
		if not os.path.isdir(directory):
			raise SystemExit('Directory doesn\'t exist: %s' % directory)
		self.symbols = self._parse(glob.glob(os.path.join(directory, '*.trig')))

	def _parse(self, trig_files):
		"""
		Parse an RDFXML stream into a SymbolCollection.
		"""
		
		ontology = rdflib.ConjunctiveGraph()
		for trig_path in trig_files:
			with open(trig_path, "r") as trig_file:
				ontology.parse(trig_file, format='trig')
		
		def _get_all_classes(*super_classes):
			for cls in super_classes:
				for subclass in ontology.subjects(RDFS.subClassOf, cls):
					yield subclass
					for x in _get_all_classes(subclass):
						yield x
		
		parent_classes = [NIENS['InformationElement'], NIENS['DataObject']]
		symbol_classes = set(_get_all_classes(*parent_classes))
		all_symbols = symbol_classes.union(parent_classes)
		
		symbols = SymbolCollection()
		for symbol in sorted(all_symbols):
			# URI
			uri = str(symbol)

			# Description
			comments = list(ontology.objects(symbol, RDFS.comment))
			doc = comments[0] if comments else ''

			# Display name
			labels = list(ontology.objects(symbol, RDFS.label))
			display_name = (labels[0]) if labels else None

			# Parents
			parents = set(ontology.objects(symbol, RDFS.subClassOf)
				).intersection(all_symbols)

			if symbol in symbol_classes:
				assert parents

			# And we have a new Symbol!
			symbols.register(uri, parents, display_name, doc)

		symbols.post_process()
		return symbols

class GenericSerializer(object):

	parser = None
	symbols = None

	def __init__(self, parser):
		self.parser = parser
		self.symbols = parser.symbols

class PythonSerializer(GenericSerializer):

	def dump(self):
		for symbol in sorted(self.symbols.values()):
			parents = set((symbol.uri for symbol in symbol.parents))
			Utils.replace_items(parents, {
				str(NIENS['InformationElement']): 'Interpretation',
				str(NIENS['DataObject']): 'Manifestation' })
			print("Symbol('%s', parent=%r, uri='%s', display_name='%s', " \
				"doc='%s', auto_resolve=False)" % (symbol.name, parents,
				symbol.uri, Utils.escape_chars(symbol.display_name, '\''),
				Utils.escape_chars(symbol.doc, '\'')))

class ValaSerializer(GenericSerializer):

	@staticmethod
	def symbol_link(symbol):
		return '%s.%s' % (symbol.namespace, symbol.name)

	@classmethod
	def build_doc(cls, symbol, doc_prefix=""):
		"""
		Build a C-style docstring for gtk-doc processing.
		"""
		uri_link = '[[%s]]' % (symbol.uri)
		doc = symbol.doc

		# List children
		children = [cls.symbol_link(child) for child in symbol.children]
		if children:
			doc += '\n\n Children: %s' % ', '.join('{@link %s}' % child
				for child in children)
		else:
			doc += '\n\n Children: None'

		# List parents
		parents = [cls.symbol_link(parent) for parent in symbol.parents]
		if parents:
			doc += '\n\n Parents: %s' % ', '.join('{@link %s}' % parent
				for parent in parents)
		else:
			doc += '\n\n Parents: None'
	
		# Convert docstring to gtk-doc style C comment
		doc = doc.replace('\n', '\n *')
		doc = '/**\n * %s:\n *\n * %s%s\n *\n * %s\n */' % (
			symbol.name, doc_prefix, uri_link, doc)
		return doc

	def dump_uris(self, dest):
		dest.write('namespace Zeitgeist\n{\n')
		for namespace, symbols in sorted(self.symbols.iter_by_namespace()):
			dest.write('\n    namespace %s\n    {\n\n' % namespace)
			for symbol in sorted(symbols):
				# FIXME: (event/subject) interpretation/manifestation ??
				doc = self.build_doc(symbol,
					doc_prefix='Macro defining the interpretation type ')
				dest.write('	    %s\n' % doc.replace('\n', '\n	    '
					).strip())
				dest.write('	    public const string %s = "%s";\n\n' % (
					symbol.name, symbol.uri))
			dest.write('    }\n')
		dest.write('}\n')

	def dump_symbols(self, dest):
		dest.write('string uri, display_name, description;\n')
		dest.write('string[] parents, children, all_children;\n\n')
		for namespace, symbols in sorted(self.symbols.iter_by_namespace()):
			for symbol in sorted(symbols):
				parent_uris = ', '.join('%s.%s' % (s.namespace, s.name) for
					s in symbol.parents)
				children_uris = ', '.join('%s.%s' % (s.namespace, s.name)
					for s in symbol.children)
				all_children_uris = ', '.join('%s.%s' % (s.namespace,
					s.name) for s in symbol.all_children)
				dest.write('uri = Zeitgeist.%s.%s;\n' % (symbol.namespace,
					symbol.name))
				dest.write('description = "%s";\n' % Utils.escape_chars(
					symbol.doc, '"'));
				dest.write('display_name = "%s";\n' % Utils.escape_chars(
					symbol.display_name, '"'))
				dest.write('parents = { %s };\n' % parent_uris)
				dest.write('children = { %s };\n' % children_uris)
				dest.write('all_children = { %s };\n' % all_children_uris)
				dest.write('Symbol.Info.register (uri, display_name, description, ' \
					'parents, children, all_children);\n\n')

class OntologyCodeGenerator(object):

	_INSERTION_MARK = '// *insert-auto-generated-code*'

	_selfpath = None
	_parser = None
	_python_serializer = None
	_vala_serializer = None

	def __init__(self):
		self._selfpath = os.path.dirname(os.path.abspath(__file__))
		self._parser = OntologyParser(os.path.join(self._selfpath, 'ontology'))
		self._python_serializer = PythonSerializer(self._parser)
		self._vala_serializer = ValaSerializer(self._parser)

	def generate_python(self):
		self._python_serializer.dump()

	def generate_vala(self, uris_tpl, symbols_tpl, uris_out, symbols_out):
		self._write_file(uris_tpl, uris_out,
			self._vala_serializer.dump_uris, 'vala')
		self._write_file(symbols_tpl, symbols_out,
			self._vala_serializer.dump_symbols, 'vala')

	def _write_file(self, tplfilename, outfilename, content_generator, _type):
		sys.stderr.write("Generating %s..." % os.path.basename(outfilename))

		# Read template file
		if not os.path.isabs (tplfilename):
			tplfilename = os.path.join(os.getcwd(), tplfilename)

		template = open(tplfilename).read()

		# Generate output
		content = StringIO()
		content_generator(content)
		content = content.getvalue().strip('\n')

		# Write header
		output = StringIO()
		self._write_header(output, _type)

		# Write template, insert the generated output into the correct
		# position (marked by "// *insert-auto-generated-code*").
		insertion_pos = template.find(self._INSERTION_MARK)
		indentation = insertion_pos - template.rfind('\n', 0, insertion_pos) - 1
		start_pos = template.rfind('\n', 0, insertion_pos) + 1
		continue_pos = insertion_pos
		output.write(template[:start_pos])
		output.write(Utils.indent(content, ' ' * indentation))
		output.write(template[continue_pos+len(self._INSERTION_MARK):])

		# Write everything to the result file
		if os.path.isabs (outfilename):
			outpath = outfilename
		else:
			outpath = os.path.join(os.getcwd(), outfilename)

		open(outpath, 'w').write(output.getvalue())

	def _write_header(self, dest, _type):
		if _type == 'vala':
			dest.write('// This file has been auto-generated by the ' \
				'ontology2code script.\n')
			dest.write('// Do not modify it directly.\n\n')
		else:
			raise NotImplementedError

	def _generate_vala_uris(self, dest):
		pass

if __name__ == "__main__":
	parser = argparse.ArgumentParser()
	parser.add_argument('--vala', nargs=4, metavar=('URI_TEMPLATE', 'SYMBOLS_TEMPLATE', 'URI_DESTINATION', 'SYMBOLS_DESTINATION'))
	parser.add_argument('--dump-python', action='store_true')

	args = parser.parse_args()

	generator = OntologyCodeGenerator()
	if args.vala:
		generator.generate_vala(args.vala[0], args.vala[1], args.vala[2], args.vala[3])
	elif args.dump_python:
		generator.generate_python()

# vim:noexpandtab:ts=4:sw=4