#!/usr/bin/python
# -*- coding: utf8 -*-

"""
This Galaxy tool permits to prepare your files to be ready for
Assembly Hub visualization.
Program test arguments:
hubArchiveCreator.py -g test-data/augustusDbia3.gff3 -f test-data/dbia3.fa -d . -u ./tools -o output.html
"""

import argparse
import collections
import json
import logging
import os
import sys

# Internal dependencies
from Bam import Bam
from BedSimpleRepeats import BedSimpleRepeats
from BedSpliceJunctions import BedSpliceJunctions
from Bed import Bed
from BigWig import BigWig
from util.Fasta import Fasta
from util.Filters import TraceBackFormatter
from Gff3 import Gff3
from Gtf import Gtf
from Psl import Psl
from TrackHub import TrackHub
from bigPsl import bigPsl
from BedBlastAlignments import BedBlastAlignments

# TODO: Verify each subprocessed dependency is accessible [gff3ToGenePred, genePredToBed, twoBitInfo, faToTwoBit, bedToBigBed, sort


def main(argv):
    # Command Line parsing init
    parser = argparse.ArgumentParser(description='Create a foo.txt inside the given folder.')

    # Reference genome mandatory
    parser.add_argument('-f', '--fasta', help='Fasta file of the reference genome')

    # GFF3 Management
    parser.add_argument('--gff3', action='append', help='GFF3 format')

    # GTF Management
    parser.add_argument('--gtf', action='append', help='GTF format')

    # Bed4+12 (TrfBig)
    parser.add_argument('--bedSimpleRepeats', action='append', help='Bed4+12 format, using simpleRepeats.as')

    # Bed12+1 (regtools)
    parser.add_argument('--bedSpliceJunctions', action='append', help='Bed12+1 format, using spliceJunctions.as')

    # Generic Bed (Blastx transformed to bed)
    parser.add_argument('--bed', action='append', help='Bed generic format')

    # BigPsl (blat alignment)
    parser.add_argument('--bigpsl', action='append', help='bigPsl format, using bigPsl.as')

    # Bed12+12 (tblastn alignment)
    parser.add_argument('--bedBlastAlignments', action='append', help='Bed12+12 format, using bigPsl.as')

    # BigWig Management
    parser.add_argument('--bigwig', action='append', help='BigWig format')

    # Bam Management
    parser.add_argument('--bam', action='append', help='Bam format')

    # Psl Management
    parser.add_argument('--psl', action='append', help='Psl format')

    # TODO: Check if the running directory can have issues if we run the tool outside
    parser.add_argument('-d', '--directory',
                        help='Running tool directory, where to find the templates. Default is running directory')
    parser.add_argument('-u', '--ucsc_tools_path',
                        help='Directory where to find the executables needed to run this tool')
    parser.add_argument('-e', '--extra_files_path',
                        help='Name, in galaxy, of the output folder. Where you would want to build the Track Hub Archive')
    parser.add_argument('-o', '--output', help='Name of the HTML summarizing the content of the Track Hub Archive')

    parser.add_argument('-j', '--data_json', help='Json containing the metadata of the inputs')

    parser.add_argument('--user_email', help='Email of the user who launched the Hub Archive Creation')

    parser.add_argument('--genome_name', help='UCSC Genome Browser assembly ID')

    parser.add_argument('--debug_mode', action='store_true', help='Allow more details about the errors')

    # Begin init variables

    toolDirectory = '.'
    extra_files_path = '.'

    # Get the args passed in parameter
    args = parser.parse_args()

    extra_files_path = args.extra_files_path
    toolDirectory = args.directory

    #### Logging management ####
    # If we are in Debug mode, also print in stdout the debug dump

    configure_logger(extra_files_path=extra_files_path, debug=args.debug_mode)

    #### END Logging management ####

    array_inputs_reference_genome = json.loads(args.fasta)

    # TODO: Replace these with the object Fasta
    input_fasta_file = array_inputs_reference_genome["false_path"]
    input_fasta_file_name = sanitize_name_input(array_inputs_reference_genome["name"])
    genome_name = sanitize_name_input(args.genome_name)

    reference_genome = Fasta(input_fasta_file,
                             input_fasta_file_name, genome_name)

    user_email = args.user_email


    # TODO: Use a class to have a better management of the structure of these inputs
    # These inputs are populated in the Galaxy Wrapper xml and are in this format:
    # ARRAY[DICT{FILE_PATH: DICT{NAME: NAME_VALUE, EXTRA_DATA: EXTRA_DATA_VALUE}}]
    # EXTRA_DATA could be anything, for example the index of a BAM => {"index", FILE_PATH}
    array_inputs_bam = args.bam
    array_inputs_bed_generic = args.bed
    array_inputs_bed_simple_repeats = args.bedSimpleRepeats
    array_inputs_bed_splice_junctions = args.bedSpliceJunctions
    array_inputs_bigwig = args.bigwig
    array_inputs_gff3 = args.gff3
    array_inputs_gtf = args.gtf
    array_inputs_psl = args.psl
    array_inputs_bigpsl = args.bigpsl
    array_inputs_bed_blast_alignments = args.bedBlastAlignments

    outputFile = args.output

    json_inputs_data = args.data_json

    # TODO: Instead use a class to properly store the objects, with object_hook
    inputs_data = json.loads(json_inputs_data)
    # We remove the spaces in ["name"] of inputs_data
    sanitize_name_inputs(inputs_data)

    # TODO: Check here all the binaries / tools we need. Exception if missing

    # Create the Track Hub folder
    trackHub = TrackHub(reference_genome, user_email, outputFile, extra_files_path, toolDirectory)

    all_datatype_dictionary = {}

    for (inputs, datatype_class) in [
                        (array_inputs_bam, Bam),
                        (array_inputs_bed_generic, Bed),
                        (array_inputs_bigwig, BigWig),
                        (array_inputs_bed_simple_repeats, BedSimpleRepeats),
                        (array_inputs_bed_splice_junctions, BedSpliceJunctions),
                        (array_inputs_gff3, Gff3),
                        (array_inputs_gtf, Gtf),
                        (array_inputs_psl, Psl),
                        (array_inputs_bigpsl, bigPsl),
                        (array_inputs_bed_blast_alignments, BedBlastAlignments)]:
        if inputs:
            all_datatype_dictionary.update(create_ordered_datatype_objects(datatype_class, inputs, inputs_data))

    # Create Ordered Dictionary to add the tracks in the tool form order
    all_datatype_ordered_dictionary = collections.OrderedDict(all_datatype_dictionary)

    logging.debug("----- End of all_datatype_dictionary processing -----")
    logging.debug("all_datatype_ordered_dictionary keys are: {0}".format(all_datatype_ordered_dictionary.values()))

    logging.debug("----- Beginning of Track adding processing -----")
    for index, datatypeObject in all_datatype_ordered_dictionary.iteritems():
        trackHub.addTrack(datatypeObject.track.trackDb)
    logging.debug("----- End of Track adding processing -----")

    # We process all the modifications to create the zip file
    #trackHub.createZip()

    # We terminate le process and so create a HTML file summarizing all the files
    trackHub.terminate()

    logging.debug('#### End of HubArchiveCreator Debug Mode: Bye! ####')

    sys.exit(0)


def sanitize_name_input(string_to_sanitize):
    """
    Sanitize the string passed in parameter by replacing '/' and ' ' by '_'

    :param string_to_sanitize:
    :return :

    :Example:

    >>> sanitize_name_input('this/is an//example')
    this_is_an__example
    """
    return string_to_sanitize \
            .replace("/", "_") \
            .replace(" ", "_")


def sanitize_name_inputs(inputs_data):
    """
    Sanitize value of the keys "name" of the dictionary passed in parameter.

    Because sometimes output from Galaxy, or even just file name, from user inputs, have spaces.
    Also, it can contain '/' character and could break the use of os.path function.

    :param inputs_data: dict[string, dict[string, string]]
    """
    for key in inputs_data:
        inputs_data[key]["name"] = sanitize_name_input(inputs_data[key]["name"])


def create_ordered_datatype_objects(ExtensionClass, array_inputs, inputs_data):
    """
    Function which executes the creation all the necessary files / folders for a special Datatype, for TrackHub
    and update the dictionary of datatype

    :param ExtensionClass:
    :param array_inputs:
    :param inputs_data:
    :type ExtensionClass: Datatype
    :type array_inputs: list[string]
    :type inputs_data: dict
    :rtype: dict
    """

    datatype_dictionary = {}

    # TODO: Optimize this double loop
    for input_false_path in array_inputs:
        for key, data_value in inputs_data.items():
            if key == input_false_path:
                logging.debug("input_false_path: " + input_false_path)
                logging.debug("data_value: " + str(data_value))
                extensionObject = ExtensionClass(input_false_path, data_value)
                datatype_dictionary.update({data_value["order_index"]: extensionObject})
    return datatype_dictionary

def configure_logger(extra_files_path=None, debug=False):
    if not extra_files_path:
        raise Exception("Extra files path is not set. Stopping the application")


    # All case log: log everything in a .log file
    logger_file_name = ''.join([__name__, '.log'])
    logging_file_path = os.path.join(extra_files_path, logger_file_name)

    logging.basicConfig(filename=logging_file_path, level=logging.DEBUG)

    log_stdout = logging.StreamHandler(sys.stdout)
    if not debug:
        configure_logger_user(log_stdout)
    else:
        configure_logger_dev(log_stdout)

    # stderr configuration
    configure_logger_stderr()

    logging.debug('#### Welcome in HubArchiveCreator Debug Mode ####\n')

def configure_logger_user(log_stdout=None):
    """
    User Logger is defined as following:
        - User needs to have WARN, ERROR and CRITICAL but well formatted / without traceback
            in STDOUT
        - Still access to full, brute and traceback for errors
            in STDERR
        - And further access to debug if needed
            in .log

    """

    if not log_stdout:
        raise Exception("No log_stdout given. Stopping the application")

    # stdout for INFO / WARN / ERROR / CRITICAL
    log_stdout.setLevel(logging.INFO)

    formatter = TraceBackFormatter('%(message)s')

    log_stdout.setFormatter(formatter)

    logging.getLogger().addHandler(log_stdout)

def configure_logger_dev(log_stdout=None):
    """
    Dev Logger is defined as following:
        - Dev needs to have WARN, ERROR and CRITICAL but well formatted / without traceback, in stdout
        - Still access to full, brute and traceback in stderr for errors
        - And further access to debug if needed

    """
    if not log_stdout:
        raise Exception("No log_stdout given. Stopping the application")
    log_format = '%(message)s'

    # stdout and stderr and both identical for INFO / WARN / ERROR / CRITICAL
    log_stdout.setLevel(logging.DEBUG)

    formatter = logging.Formatter(log_format)

    log_stdout.setFormatter(formatter)

    logging.getLogger().addHandler(log_stdout)

def configure_logger_stderr():
    """
    Configure what should be logged in stderr
    """
    log_error = logging.StreamHandler(sys.stderr)
    log_error.setLevel(logging.ERROR)
    log_error_format = '%(message)s'

    formatter_error = logging.Formatter(log_error_format)

    log_error.setFormatter(formatter_error)

    logging.getLogger().addHandler(log_error)

if __name__ == "__main__":
    logging.getLogger(__name__)
    main(sys.argv)