#       Licensed to the Apache Software Foundation (ASF) under one
#       or more contributor license agreements.  See the NOTICE file
#       distributed with this work for additional information
#       regarding copyright ownership.  The ASF licenses this file
#       to you under the Apache License, Version 2.0 (the
#       "License"); you may not use this file except in compliance
#       with the License.  You may obtain a copy of the License at
#
#         http://www.apache.org/licenses/LICENSE-2.0
#
#       Unless required by applicable law or agreed to in writing,
#       software distributed under the License is distributed on an
#       "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
#       KIND, either express or implied.  See the License for the
#       specific language governing permissions and limitations
#       under the License.

import bson
import datetime
import json
import logging
import multiprocessing
import re
import string
import sys

import colander as col
from ming.orm import session, ThreadLocalORMSession
from tg import tmpl_context as c, app_globals as g

from allura import model as M
from allura.lib import helpers as h
from allura.lib.plugin import ProjectRegistrationProvider

log = logging.getLogger(__name__)


class TroveCategory():

    def __init__(self, root_type=''):
        self.root_type = root_type

    def deserialize(self, node, cstruct):
        if cstruct is col.null:
            return col.null
        cat = M.TroveCategory.query.get(fullpath=cstruct)
        if not cat:
            cat = M.TroveCategory.query.get(fullname=cstruct)
        if not cat:
            raise col.Invalid(node,
                              '"%s" is not a valid trove category.' % cstruct)
        if not cat.fullpath.startswith(self.root_type):
            raise col.Invalid(node,
                              '"%s" is not a valid "%s" trove category.' %
                              (cstruct, self.root_type))
        return cat


class User():

    def deserialize(self, node, cstruct):
        if cstruct is col.null:
            return col.null
        user = M.User.by_username(cstruct)
        if not user:
            raise col.Invalid(node,
                              'Invalid username "%s".' % cstruct)
        return user


class ProjectName(object):

    def __init__(self, name, shortname):
        self.name = name
        self.shortname = shortname


class ProjectNameType():

    def deserialize(self, node, cstruct):
        if cstruct is col.null:
            return col.null
        name = cstruct
        shortname = re.sub("[^A-Za-z0-9 ]", "", name).lower()
        shortname = re.sub(" ", "-", shortname)
        return ProjectName(name, shortname)


class ProjectShortnameType():

    def __init__(self, nbhd, update):
        self.nbhd = nbhd
        self.update = update

    def deserialize(self, node, cstruct):
        if cstruct is col.null:
            return col.null
        return ProjectRegistrationProvider.get().shortname_validator.to_python(cstruct,
                                                                               check_allowed=not self.update,
                                                                               neighborhood=self.nbhd)


class Award():

    def __init__(self, nbhd):
        self.nbhd = nbhd

    def deserialize(self, node, cstruct):
        if cstruct is col.null:
            return col.null
        award = M.Award.query.find(dict(short=cstruct,
                                        created_by_neighborhood_id=self.nbhd._id)).first()
        if not award:
            # try to look up the award by _id
            award = M.Award.query.find(dict(_id=bson.ObjectId(cstruct),
                                            created_by_neighborhood_id=self.nbhd._id)).first()
        if not award:
            raise col.Invalid(node,
                              'Invalid award "%s".' % cstruct)
        return award


class TroveTopics(col.SequenceSchema):
    trove_topics = col.SchemaNode(TroveCategory("Topic"))


class TroveLicenses(col.SequenceSchema):
    trove_license = col.SchemaNode(TroveCategory("License"))


class TroveDatabases(col.SequenceSchema):
    trove_databases = col.SchemaNode(TroveCategory("Database Environment"))


class TroveStatuses(col.SequenceSchema):
    trove_statuses = col.SchemaNode(TroveCategory("Development Status"))


class TroveAudiences(col.SequenceSchema):
    trove_audience = col.SchemaNode(TroveCategory("Intended Audience"))


class TroveOSes(col.SequenceSchema):
    trove_oses = col.SchemaNode(TroveCategory("Operating System"))


class TroveLanguages(col.SequenceSchema):
    trove_languages = col.SchemaNode(TroveCategory("Programming Language"))


class TroveTranslations(col.SequenceSchema):
    trove_translations = col.SchemaNode(TroveCategory("Translations"))


class TroveUIs(col.SequenceSchema):
    trove_uis = col.SchemaNode(TroveCategory("User Interface"))


class Labels(col.SequenceSchema):
    label = col.SchemaNode(col.Str())


class Project(col.MappingSchema):
    name = col.SchemaNode(ProjectNameType())
    summary = col.SchemaNode(col.Str(), missing='')
    description = col.SchemaNode(col.Str(), missing='')
    admin = col.SchemaNode(User())
    private = col.SchemaNode(col.Bool(), missing=False)
    labels = Labels(missing=[])
    external_homepage = col.SchemaNode(col.Str(), missing='')
    video_url = col.SchemaNode(col.Str(), missing='')
    trove_root_databases = TroveDatabases(missing=None)
    trove_developmentstatuses = TroveStatuses(validator=col.Length(max=6), missing=None)
    trove_audiences = TroveAudiences(validator=col.Length(max=6), missing=None)
    trove_licenses = TroveLicenses(validator=col.Length(max=6), missing=None)
    trove_oses = TroveOSes(missing=None)
    trove_languages = TroveLanguages(validator=col.Length(max=6), missing=None)
    trove_topics = TroveTopics(validator=col.Length(max=3), missing=None)
    trove_natlanguages = TroveTranslations(missing=None)
    trove_environments = TroveUIs(missing=None)
    tool_data = col.SchemaNode(col.Mapping(unknown='preserve'), missing={})
    icon = col.SchemaNode(col.Str(), missing=None)
    # more fields are added dynamically to the schema in main()


def valid_shortname(project):
    if project.shortname:
        # already validated in ProjectShortnameType validator
        return True
    elif 3 <= len(project.name.shortname) <= 15:
        return True
    else:
        return 'Project shortname "%s" must be between 3 and 15 characters' \
            % project.name.shortname


class Projects(col.SequenceSchema):
    project = Project(validator=col.Function(valid_shortname))


class Object(object):

    def __init__(self, d):
        self.__dict__.update(d)


def trove_ids(orig, new_):
    if new_ is None:
        return orig
    return list(set(t._id for t in list(new_)))


def create_project(p, nbhd, options):
    worker_name = multiprocessing.current_process().name
    M.session.artifact_orm_session._get().skip_mod_date = True
    shortname = p.shortname or p.name.shortname
    project = M.Project.query.get(shortname=shortname,
                                  neighborhood_id=nbhd._id)
    project_template = nbhd.get_project_template()

    if project and not (options.update and p.shortname):
        log.warning('[%s] Skipping existing project "%s". To update an existing '
                    'project you must provide the project shortname and run '
                    'this script with --update.' % (worker_name, shortname))
        return 0

    if not project:
        log.info('[%s] Creating project "%s".' % (worker_name, shortname))
        try:
            project = nbhd.register_project(shortname,
                                            p.admin,
                                            project_name=p.name.name,
                                            private_project=p.private)
        except Exception, e:
            log.exception('[%s] %s' % (worker_name, str(e)))
            return 0
    else:
        log.info('[%s] Updating project "%s".' % (worker_name, shortname))

    project.notifications_disabled = True

    if options.ensure_tools and 'tools' in project_template:
        for i, tool in enumerate(project_template['tools'].iterkeys()):
            tool_config = project_template['tools'][tool]
            if project.app_instance(tool_config['mount_point']):
                continue
            tool_options = tool_config.get('options', {})
            for k, v in tool_options.iteritems():
                if isinstance(v, basestring):
                    tool_options[k] = string.Template(v).safe_substitute(
                        project.root_project.__dict__.get('root_project', {}))
            project.install_app(tool,
                                mount_label=tool_config['label'],
                                mount_point=tool_config['mount_point'],
                                **tool_options)

    project.summary = p.summary
    project.short_description = p.description
    project.external_homepage = p.external_homepage
    project.video_url = p.video_url
    project.last_updated = datetime.datetime.utcnow()
    # These properties may have been populated by nbhd template defaults in
    # register_project(). Overwrite if we have data, otherwise keep defaults.
    project.labels = p.labels or project.labels
    project.trove_root_database = trove_ids(project.trove_root_database, p.trove_root_databases)
    project.trove_developmentstatus = trove_ids(project.trove_developmentstatus, p.trove_developmentstatuses)
    project.trove_audience = trove_ids(project.trove_audience, p.trove_audiences)
    project.trove_license = trove_ids(project.trove_license, p.trove_licenses)
    project.trove_os = trove_ids(project.trove_os, p.trove_oses)
    project.trove_language = trove_ids(project.trove_language, p.trove_languages)
    project.trove_topic = trove_ids(project.trove_topic, p.trove_topics)
    project.trove_natlanguage = trove_ids(project.trove_natlanguage, p.trove_natlanguages)
    project.trove_environment = trove_ids(project.trove_environment, p.trove_environments)

    project.tool_data.update(p.tool_data)

    for a in p.awards:
        M.AwardGrant(app_config_id=bson.ObjectId(),
                     award_id=a._id,
                     granted_to_project_id=project._id,
                     granted_by_neighborhood_id=nbhd._id)

    if p.icon:
        with open(p.icon) as icon_file:
            project.save_icon(p.icon, icon_file)

    project.notifications_disabled = False
    with h.push_config(c, project=project, user=p.admin):
        ThreadLocalORMSession.flush_all()
        g.post_event('project_updated')
    session(project).clear()
    return 0


def create_projects(projects, nbhd, options):
    for p in projects:
        r = create_project(Object(p), nbhd, options)
        if r != 0:
            sys.exit(r)


def main(options):
    log.addHandler(logging.StreamHandler(sys.stdout))
    log.setLevel(getattr(logging, options.log_level.upper()))
    log.debug(options)

    nbhd = M.Neighborhood.query.get(name=options.neighborhood)
    if not nbhd:
        return 'Invalid neighborhood "%s".' % options.neighborhood

    data = json.load(open(options.file, 'r'))

    # dynamically add to the schema (e.g. if needs nbhd)
    project = Project()
    project.add(col.SchemaNode(col.Sequence(),
                               col.SchemaNode(Award(nbhd)),
                               name='awards', missing=[]))
    project.add(col.SchemaNode(ProjectShortnameType(nbhd, options.update),
                               name='shortname', missing=None))

    projects = []
    for datum in data:
        try:
            projects.append(project.deserialize(datum))
        except Exception:
            keep_going = options.validate_only
            log.error('Error on %s\n%s', datum['shortname'], datum, exc_info=keep_going)
            if not keep_going:
                raise

    log.debug(projects)

    if options.validate_only:
        return

    chunks = [projects[i::options.nprocs] for i in range(options.nprocs)]
    jobs = []
    for i in range(options.nprocs):
        p = multiprocessing.Process(target=create_projects,
                                    args=(chunks[i], nbhd, options), name='worker-' + str(i + 1))
        jobs.append(p)
        p.start()

    for j in jobs:
        j.join()
        if j.exitcode != 0:
            return j.exitcode
    return 0


def parse_options():
    import argparse
    parser = argparse.ArgumentParser(
        description='Import Allura project(s) from JSON file')
    parser.add_argument('file', metavar='JSON_FILE', type=str,
                        help='Path to JSON file containing project data.')
    parser.add_argument('neighborhood', metavar='NEIGHBORHOOD', type=str,
                        help='Destination Neighborhood shortname.')
    parser.add_argument('--log', dest='log_level', default='INFO',
                        help='Log level (DEBUG, INFO, WARNING, ERROR, CRITICAL).')
    parser.add_argument('--update', dest='update', default=False,
                        action='store_true',
                        help='Update existing projects. Without this option, existing '
                        'projects will be skipped.')
    parser.add_argument('--ensure-tools', dest='ensure_tools', default=False,
                        action='store_true',
                        help='Check nbhd project template for default tools, and install '
                        'them on the project(s) if not already installed.')
    parser.add_argument(
        '--nprocs', '-n', action='store', dest='nprocs', type=int,
        help='Number of processes to divide the work among.',
        default=multiprocessing.cpu_count())
    parser.add_argument('--validate-only', '-v', action='store_true', dest='validate_only',
                        help='Validate ALL records, make no changes')
    return parser.parse_args()


if __name__ == '__main__':
    sys.exit(main(parse_options()))