view data_manager/add_ctat_centrifuge_index.py @ 0:b4d4f0d76e94 draft default tip

Uploaded

#!/usr/bin/env python
# ref: https://galaxyproject.org/admin/tools/data-managers/how-to/define/

# Rewritten by H.E. Cicada Brokaw Dennis from a source downloaded from the toolshed and
# other example code on the web.
# This allows downloading of a centrifuge index, or specification of its disk location.
# This index is one of the input paramters needed by the ctat_metagenomics tool.
# At the moment only one index is supported by the ctat_metagenomics tool:
# ftp://ftp.ccb.jhu.edu/pub/infphilo/centrifuge/data/p_compressed+h+v.tar.gz

import argparse
import os
#import tarfile
#import urllib
import subprocess

# The following is used to generate a unique_id value
from datetime import *

# Remove the following line when testing without galaxy package:
from galaxy.util.json import to_json_string
# Am not using the following:
# from galaxy.util.json import from_json_string

# The FileListParser is used by get_ctat_genome_filenames(),
# which is called by the Data Manager interface (.xml file) to get
# the filenames that are available online at broadinstitute.org
# Not sure best way to do it. 
# This object uses HTMLParser to look through the html 
# searching for the filenames within anchor tags.
import urllib2
from HTMLParser import HTMLParser

_CTAT_CentrifugeIndexPage_URL = 'https://ccb.jhu.edu/software/centrifuge/'
_CTAT_CentrifugeDownload_URL = 'ftp://ftp.ccb.jhu.edu/pub/infphilo/centrifuge/data/p_compressed+h+v.tar.gz'
_CTAT_CentrifugeIndexTableName = 'ctat_centrifuge_indexes'
_CTAT_CentrifugeDir_Name = 'p_compressed+h+v'
_CTAT_Centrifuge_DisplayNamePrefix = 'CTAT_CentrifugeIndex_'
_CentrifugeIndexFileExtension = 'cf'
_NumBytesNeededForIndex = 7400130287 # 6.9 GB
#_DownloadFileSize = 5790678746 # 5.4 Gigabytes.
_Download_TestFile = 'write_testfile.txt'
_DownloadSuccessFile = 'download_succeeded.txt'

class FileListParser(HTMLParser):
    def __init__(self):
        # Have to use direct call to super class rather than using super():
        # super(FileListParser, self).__init__()
        # because HTMLParser is an "old style" class and its inheritance chain does not include object.
        HTMLParser.__init__(self)
        self.filenames = set()
    def handle_starttag(self, tag, attrs):
        # Look for filename references in anchor tags and add them to filenames.
        if tag == "a":
            # The tag is an anchor tag.
            for attribute in attrs:
                # print "Checking: {:s}".format(str(attribute))
                if attribute[0] == "href":
                    # Does the href have a tar.gz in it?
                    if ("tar.gz" in attribute[1]) and ("md5" not in attribute[1]):
                        # Add the value to filenames.
                        self.filenames.add(attribute[1])            
# End of class FileListParser

def get_ctat_centrifuge_index_locations():
    # For dynamic options need to return an interable with contents that are tuples with 3 items.
    # Item one is a string that is the display name put into the option list.
    # Item two is the value that is put into the parameter associated with the option list.
    # Item three is a True or False value, indicating whether the item is selected.
    options = []
    # open the url and retrieve the filenames of the files in the directory.
    resource = urllib2.urlopen(_CTAT_CentrifugeIndexPage_URL)
    theHTML = resource.read()
    filelist_parser = FileListParser()
    filelist_parser.feed(theHTML)
    # This is what was returned on 2018-04-23
    # ftp://ftp.ccb.jhu.edu/pub/infphilo/centrifuge/data/p_compressed_2018_4_15.tar.gz
    # ftp://ftp.ccb.jhu.edu/pub/infphilo/centrifuge/data/nt_2018_3_3.tar.gz
    # ftp://ftp.ccb.jhu.edu/pub/infphilo/centrifuge/data/p_compressed+h+v.tar.gz
    # ftp://ftp.ccb.jhu.edu/pub/infphilo/centrifuge/data/p+h+v.tar.gz
    # Which could be hard coded:
    # options.append(("p_compressed+h+v", "ftp://ftp.ccb.jhu.edu/pub/infphilo/centrifuge/data/p_compressed+h+v.tar.gz", True))
    # options.append(("p+h+v", "ftp://ftp.ccb.jhu.edu/pub/infphilo/centrifuge/data/p+h+v.tar.gz", False))
    # options.append(("nt_2018_3_3", "ftp://ftp.ccb.jhu.edu/pub/infphilo/centrifuge/data/nt_2018_3_3.tar.gz", False))
    # options.append(("p_compressed_2018_4_15", "ftp://ftp.ccb.jhu.edu/pub/infphilo/centrifuge/data/p_compressed_2018_4_15.tar.gz", False))
    # but only returning the one we want, which for now is assumed to be present.
    # For now, I am printing the list, just so I can see what was returned, 
    # print "FYI: The URL's that were found on Centrifuge's page are:"
    # print "\t" + "\n\t".join(filelist_parser.filenames)
    # For now instead of sending back the list of found URL's, send back the one URL we want.
    # Currently, only one of the options is supported.
    options.append((_CTAT_CentrifugeDir_Name, _CTAT_CentrifugeDownload_URL, True))
    print "The list of items being returned for the option menu is:"
    print str(options)
    return options 

# The following was used by the example program to get input parameters through the json.
# Just leaving here for reference.
# We are getting all of our parameter values through command line arguments.
#def get_reference_id_name(params):
#    genome_id = params['param_dict']['genome_id']
#    genome_name = params['param_dict']['genome_name']
#    return genome_id, genome_name
#
#def get_url(params):
#    trained_url = params['param_dict']['trained_url']
#    return trained_url

def download_index(src_location, destination, force_download):
    # We do not know if the index has been downloaded already.
    # This function returns whether or not the index actually gets downloaded.
    index_was_downloaded = False
    # Get the root filename of the Genome Directory. 
    # The part after the last '/' and before the first '.'
    root_index_dirname = src_location.split("/")[-1].split(".")[0]

    # We want to make sure that destination is absolute fully specified path.
    cannonical_destination = os.path.realpath(destination) 
    if cannonical_destination.split("/")[-1] != root_index_dirname:
        cannonical_destination += "/" + root_index_dirname
    if os.path.exists(cannonical_destination):
        if not os.path.isdir(cannonical_destination):
            raise ValueError("The destination is not a directory: " + \
                             "{:s}".format(cannonical_destination))
        # else all is good. It is a directory.
    else:
        # We need to create it.
        try:
            os.makedirs(cannonical_destination)
        except os.error:
            print "ERROR: Trying to create the following directory path:"
            print "\t{:s}".format(cannonical_destination)
            raise

    # Make sure the directory now exists and we can write to it.
    if not os.path.exists(cannonical_destination):
        # It should have been created, but if it doesn't exist at this point
        # in the code, something is wrong. Raise an error.
        raise OSError("The destination directory could not be created: " + \
                      "{:s}".format(cannonical_destination))
    test_writing_file = "{:s}/{:s}".format(cannonical_destination, _Download_TestFile)
    try:
        filehandle = open(test_writing_file, "w")
        filehandle.write("Testing writing to this file.")
        filehandle.close()
        os.remove(test_writing_file)
    except IOError:
        print "The destination directory could not be written into: " + \
                      "{:s}".format(cannonical_destination)
        raise
    
    # Get the list of files in the directory,
    # We use it to check for a previous download or extraction among other things.
    orig_files_in_destdir = set(os.listdir(cannonical_destination))
    # See whether the file has been downloaded already.
    download_success_file_path = "{:s}/{:s}".format(cannonical_destination, _DownloadSuccessFile)
    if (_DownloadSuccessFile not in orig_files_in_destdir) or force_download:
        # Check whether there is enough space on the device for the index.
        statvfs = os.statvfs(cannonical_destination)
        # fs_size = statvfs.f_frsize * statvfs.f_blocks          # Size of filesystem in bytes
        # num_free_bytes = statvfs.f_frsize * statvfs.f_bfree    # Actual number of free bytes
        num_avail_bytes = statvfs.f_frsize * statvfs.f_bavail    # Number of free bytes that ordinary users
                                                                 # are allowed to use (excl. reserved space)
        if (num_avail_bytes < _NumBytesNeededForIndex):
            raise OSError("There is insufficient space ({:s} bytes)".format(str(num_avail_bytes)) + \
                          " on the device of the destination directory: " + \
                          "{:s}".format(cannonical_destination))
    
        #Previous code to download and untar. Not using anymore.
        #full_filepath = os.path.join(destination, src_filename)
        #
        #Download ref: https://dzone.com/articles/how-download-file-python
        #f = urllib2.urlopen(ctat_resource_lib_url)
        #data = f.read()
        #with open(full_filepath, 'wb') as code:
        #    code.write(data)
        #
        #Another way to download:
        #try: 
        #    urllib.urlretrieve(url=ctat_resource_lib_url, filename=full_filepath)
        #
        #Then untar the file.
        #try: 
        #    tarfile.open(full_filepath, mode='r:*').extractall()
        
        if (_DownloadSuccessFile in orig_files_in_destdir):
            # Since we are redoing the download, 
            # the success file needs to be removed
            # until the download has succeeded.
            os.remove(download_success_file_path)
        # We want to transfer and untar the file without storing the tar file, because that
        # adds all that much more space to the needed amount of free space on the disk.
        # Use subprocess to pipe the output of curl into tar.
        # Make curl silent so progress is not printed to stderr.
        command = "curl --silent {:s} | tar -xzf - -C {:s}".format(src_location, cannonical_destination)
        try: # to send the command that downloads and extracts the file.
            command_output = subprocess.check_output(command, shell=True)
            # FIX - not sure check_output is what we want to use. If we want to have an error raised on
            # any problem, maybe we should not be checking output.
        except subprocess.CalledProcessError:
            print "ERROR: Trying to run the following command:\n\t{:s}".format(command)
            raise
        else:
            index_was_downloaded = True

    # Some code to help us if errors occur.
    print "\n*******************************\nFinished download and extraction."
    if os.path.exists(cannonical_destination) and os.path.isdir(cannonical_destination):
        subprocess.check_call("ls -lad {:s}/* 2>&1".format(cannonical_destination), shell=True)
    
    files_in_destdir = set(os.listdir(cannonical_destination))
    found_filenames = set()
    for filename in files_in_destdir:
        # There should be three files, but some OS's might have created
        # other files in the directory, or maybe the user did.
        # Look for the index files.
        # The download files' names should start with the root_index_dirname
        # print "Is root: {:s} in file: {:s}".format(root_index_dirname, filename)
        if root_index_dirname in filename:
            found_filenames.add(filename)
    # print "The found_filenames are:\n\t{:s}".format(str(found_filenames))
    if (len(found_filenames) >= 3):
        # FIX - we could md5 the files to make sure they are correct.
        # Or at least check their sizes, to see if the download completed ok.
        # Also we could check the names of the files.
        try:
            # Create a file to indicate that the download succeeded.
            subprocess.check_call("touch {:s}".format(download_success_file_path), shell=True)
        except IOError:
            print "The download_success file could not be created: " + \
                      "{:s}".format(download_success_file_path)
            raise
    else:
        print "After download, the potential index files found are:\n\t{:s}".format(str(found_filenames))
        raise ValueError("ERROR: Could not find the extracted index files " + \
                         "in the destination directory:\n\t{:s}".format(cannonical_destination))

    return (cannonical_destination, root_index_dirname, index_was_downloaded)
        
def main():
    #Parse Command Line
    # print "At start before parsing arguments."
    parser = argparse.ArgumentParser()
    parser.add_argument('-d', '--download_location', default="", \
        help='This is the download location of the centrifuge index.')
    parser.add_argument('-n', '--display_name', default="", \
        help='Is used as the selector text for the entry of this Centrifuge Index in the data table.')
    parser.add_argument('-p', '--destination_path', \
        help='Full path of the Centrifuge Index location or destination, either where it is, or where it will be placed.')
    parser.add_argument('-o', '--output_filename', \
        help='Name of the output file, where the json dictionary will be written.')
    parser.add_argument('-f', '--force_download', 
        help='Forces download of the Centrifuge Index, even if previously downloaded. ' + \
             'Requires download_location to be set in order to work.', action="store_true")
    args = parser.parse_args()

    # All of the input parameters are written by default to the output file prior to
    # this program being called.
    # But I do not get input values from the json file, but rather from command line.
    # Just leaving the following code as a comment, in case it might be useful to someone later.
    # params = from_json_string(open(filename).read())
    # target_directory = params['output_data'][0]['extra_files_path']
    # os.mkdir(target_directory)

    # print "Arguments are parsed."
    print "\ndownload_location is {:s}".format(str(args.download_location))
    print "display_name is {:s}".format(str(args.display_name))
    print "destination_path is {:s}\n".format(str(args.destination_path))
    root_index_dirname = None
    # FIX - Prob don't need index_was_downloaded. Not doing anything with it.
    # But it indicates success downloading the index, so maybe should be checking it.
    index_was_downloaded = False
    if (args.download_location != ""):
        index_directory, root_index_dirname, index_was_downloaded = \
            download_index(src_location=args.download_location, \
                           destination=args.destination_path, \
                           force_download=args.force_download)
    else:
        cannonical_destination = os.path.realpath(args.destination_path)
        if not os.path.exists(cannonical_destination):
            raise ValueError("Cannot find the Centrifuge Index.\n" + \
                "The directory does not exist:\n\t{:s}".format(index_directory))
        # If args.destination_path is a directory containing 
        # a subdirectory that contains the index files,
        # then we need to set the index_directory to be that subdirectory.
        files_in_destination_path = os.listdir(cannonical_destination)
        if (len(files_in_destination_path) == 1):
            path_to_file = "{:s}/{:s}".format(cannonical_destination, files_in_destination_path[0])
            if os.path.isdir(path_to_file):
                index_directory = path_to_file
            else:
                index_directory = cannonical_destination
        else:
            index_directory = cannonical_destination
        # Get the root_index_dirname of the index from the index_directory name.
        root_index_dirname = index_directory.split("/")[-1].split(".")[0]

    # Check if there is an actual Centrifuge Index file in the index_directory.
    print "\nThe location of the Centrifuge Index is {:s}.\n".format(index_directory)
    files_in_index_directory = set(os.listdir(index_directory))
    index_file_found = False
    index_file_path = index_directory
    for filename in files_in_index_directory:
        # The current index is split into 3 files.
        # filenames are in the form: index_root_name.#.cf,
        # where # is a numeral (1, 2, or 3)
        # indicating the order of the files.
        if filename.split(".")[-1] == _CentrifugeIndexFileExtension:
            index_file_found = True
            # The centrifuge program wants the root name of the files to be final part of the path.
            index_file_path = "{:s}/{:s}".format(index_directory, filename.split(".")[0])
    if not index_file_found:
        raise ValueError("Cannot find any Centrifuge Index files.\n" + \
            "The contents of the directory {:s} are:\n\t".format(index_directory) + \
            "\n\t".join(files_in_index_directory))

    # Set the display_name
    if (args.display_name is None) or (args.display_name == ""):
        # Use the root_index_dirname.
        if (root_index_dirname != None) and (root_index_dirname != ""):
            display_name = _CTAT_Centrifuge_DisplayNamePrefix + root_index_dirname
        else:
            display_name = _CTAT_Centrifuge_DisplayNamePrefix + _CTAT_CentrifugeDir_Name
            print "WARNING: Did not set the display name. Using the default: {:s}".format(display_name_value)
    else:
        display_name = _CTAT_Centrifuge_DisplayNamePrefix + args.display_name
    display_name = display_name.replace(" ","_")

    # Set the unique_id
    datetime_stamp = datetime.now().strftime("_%Y_%m_%d_%H_%M_%S_%f")
    if (root_index_dirname != None) and (root_index_dirname != ""):
        unique_id = root_index_dirname + datetime_stamp
    else:
        unique_id = _CTAT_CentrifugeDir_Name + datetime_stamp

    print "The Index's display_name will be set to: {:s}\n".format(display_name)
    print "Its unique_id will be set to: {:s}\n".format(unique_id)
    print "Its dir_path will be set to: {:s}\n".format(index_file_path)

    data_manager_dict = {}
    data_manager_dict['data_tables'] = {}
    data_manager_dict['data_tables'][_CTAT_CentrifugeIndexTableName] = []
    data_table_entry = dict(value=unique_id, name=display_name, path=index_file_path)
    data_manager_dict['data_tables'][_CTAT_CentrifugeIndexTableName].append(data_table_entry)

    # Temporarily the output file's dictionary is written for debugging:
    print "The dictionary for the output file is:\n\t{:s}".format(str(data_manager_dict))
    # Save info to json file. This is used to transfer data from the DataManager tool, to the data manager,
    # which then puts it into the correct .loc file (I think).
    # Remove the following line when testing without galaxy package.
    open(args.output_filename, 'wb').write(to_json_string(data_manager_dict))

if __name__ == "__main__":
    main()