from bioblend import galaxy
import bioblend as bb
import argparse
import os
import sys
import subprocess
import json
from workflow import Workflow
from filetransfer import FileTransfer
# TODO: script description
"""
"""
class Autoload:
def __init__(self, json_in):
self.json_in = json_in
def main(self):
return None
def main():
parser = argparse.ArgumentParser(description="Input genus, species, strain, version")
parser.add_argument("json", type=str, help="Input JSON file")
parser.add_argument("--just-load", help="Only load data into galaxy, does not create nor run analyses in galaxy")
# CLI stuff
# parser.add_argument("--name", help="Sample species name, format: genus-species",type=str)
# parser.add_argument("--strain", help="Strain of the sample species", type=str)
# parser.add_argument("--version", help="Data version (e.g 1.0, 1.2, ...)", type=str)
# parser.add_argument("--common-name", help="Vernacular/common name of the species",type=str)
user_arguments = parser.parse_args()
# List that will hold all dicts from the JSON input file, containing parameters for each species
sp_dict_list = []
# JSON parsing and loading
with open(user_arguments.json, 'r') as infile:
json_sp_dict = json.load(infile)
json_sp_dump = json.dumps(json_sp_dict, indent=4, sort_keys=True)
for json_sp in json_sp_dict:
sp_dict_list.append(json_sp)
# Get variables from the current species dict
for sp_dict in sp_dict_list:
sp_params_list = []
genus = sp_dict["genus"]
genus_lower = genus[0].lower() + genus[1:]
species = sp_dict["species"]
genus_species = genus_lower + "_" + species
common = sp_dict["common"]
strain = sp_dict["strain"]
if strain != "":
genus_species_strain = genus_species + "_" + strain
else:
genus_species_strain = genus_species
ogs_version = sp_dict["ogs version"]
genome_version = sp_dict["genome version"]
performed_by = sp_dict["performed by"]
# CLI stuff
# genus = user_arguments.name.split('-')[0]
# genus_lower = genus[0].lower().genus[1:]
# genus_upper = genus[0].upper() + genus[1:]
# species = user_arguments.name.split('-')[1]
# strain = user_arguments.strain
# vernacular = user_arguments.common_name
# TODO: prompt y/n asking for the validity of info
# Test adress, change to abims-gga.sb-roscoff.fr/sp/ in production
instance_url = "http://localhost/sp/" + genus_lower + "_" + species + "/galaxy/"
print("Species: " + genus + " " + species + " (" + common + ")"
"\nStrain: " + strain +
"\nAccessing instance " + instance_url)
# Connect to the galaxy instance of the current species TODO: API key connection issues
gi = galaxy.GalaxyInstance(url=instance_url,
key="0e993414b2f876515e74dd890f16ffc7",
email="alebars@sb-roscoff.fr",
password="pouet",
verify=False)
"""
This part creates the current species directory and go to it
If it already exists, just move to it
To be expanded when docker-swarm is implemented (or all config files are ready), not useful for now
"""
main_dir = os.getcwd() + "/"
sp_dir = os.path.join(main_dir, genus_species) + "/"
try:
os.mkdir(sp_dir)
except FileExistsError:
print("Directory " + sp_dir + " already exists")
try:
os.chdir(sp_dir)
wd = os.getcwd()
except OSError:
print("Cannot access " + sp_dir + ", run with higher privileges")
break
# Production instance example TODO: secure pswd and API key? + manage API keys (if needed)
# gi = galaxy.GalaxyInstance(url="http://abims-gga.sb-roscoff.fr/sp/ectocarpus_species1/galaxy/",
# key="84dfbee3c0efa9155518f01fbeff57c8",
# email="gga@sb-roscoff.fr",
# password="****")
# Check connection to the current instance
try:
hl = gi.histories.get_histories()
except bb.ConnectionError:
print("Cannot connect to GGA instance @ " + instance_url)
else:
print("Successfully connected to instance " + instance_url)
# TODO: FTP stuff to retrieve the datasets (used in testing, not needed for production)
# try:
# os.mkdir("./src_data")
# except FileExistsError:
# print("src_data folder already exists for " + genus_species_strain)
# else:
# print("src_data folder created for " + genus_species_strain)
#
# TODO: load the data into the current species directory and load it into galaxy instance
# setup_data_libraries_cl = \
# "docker-compose exec galaxy /tool_deps/_conda/bin/python /opt/setup_data_libraries.py"
#
# try:
# setup_data_libraries = subprocess.Popen(setup_data_libraries_cl.split(), stdout=subprocess.PIPE)
# # output message from the data loading script
# setup_data_libraries_output = setup_data_libraries.communicate()
# except Exception:
# print("Cannot load data into container for " + genus_species_strain)
# break
# else:
# print("Data successfully loaded into docker container for " + genus_species_strain)
# TODO: datamapping to run workflow (get datasets from instance)
datamap = {}
# generate workflow file and run it in the galaxy instance
workflow = Workflow(parameters_dict=sp_dict, instance=gi)
print("Generating custom workflow for " + genus_species_strain)
wf_dict = workflow.generate(working_directory=wd, main_directory=main_dir)
gi.histories.create_history(name=str(genus_species_strain + "_" + genome_version))
hi = gi.histories.get_histories(name=str(genus_species_strain + "_" + genome_version))
hi_id = hi[0]["id"]
# print(hi)
li = gi.libraries.get_libraries() # only one library
# print(li)
li_id = gi.libraries.get_libraries()[0]["id"] # project data folder/library
# print(li_id)
fo_gi = gi.libraries.get_folders(library_id=li_id) # data location (project data)
# print(fo_gi)
fo_id = {}
current_fo_name = ""
# folders ids: access to data to run the first tools
for i in fo_gi:
for k, v in i.items():
if k == "name":
fo_id[v] = 0
current_fo_name = v
if k == "id":
fo_id[current_fo_name] = v
print("IDs: ")
datasets = {}
for k, v in fo_id.items():
print("\t" + k + ": " + v)
if k == "/genome":
sub_folder_content = gi.folders.show_folder(folder_id=v, contents=True)
# print(sub_folder_content)
for k2, v2 in sub_folder_content.items():
# print(k2)
# print(v2)
for e in v2:
if type(e) == dict:
if e["name"].endswith(".fa"):
datasets["genome_file"] = e["ldda_id"]
print("\t\t" + e["name"] + ": " + e["ldda_id"])
elif k == "/annotation/" + genus_species:
sub_folder_content = gi.folders.show_folder(folder_id=v, contents=True)
# print(sub_folder_content)
for k2, v2 in sub_folder_content.items():
# print(k2)
# print(v2)
for e in v2:
if type(e) == dict:
# TODO: manage several files of the same type
if e["name"].endswith("transcripts-gff.fa"):
datasets["transcripts_file"] = e["ldda_id"]
print("\t\t" + e["name"] + ": " + e["ldda_id"])
elif e["name"].endswith("proteins.fasta"):
datasets["proteins_file"] = e["ldda_id"]
print("\t\t" + e["name"] + ": " + e["ldda_id"])
elif e["name"].endswith(".gff"):
datasets["gff_file"] = e["ldda_id"]
print("\t\t" + e["name"] + ": " + e["ldda_id"])
current_hi = gi.histories.get_current_history()["id"]
print("History ID:" + current_hi)
gi.histories.upload_dataset_from_library(history_id=hi_id, lib_dataset_id=datasets["genome_file"])
gi.histories.upload_dataset_from_library(history_id=hi_id, lib_dataset_id=datasets["gff_file"])
gi.histories.upload_dataset_from_library(history_id=hi_id, lib_dataset_id=datasets["transcripts_file"])
gi.histories.upload_dataset_from_library(history_id=hi_id, lib_dataset_id=datasets["proteins_file"])
# print(gi.tools.get_tool_panel())
# print(wf_dict)
wf_dict_2 = json.loads(wf_dict) # add to workflow class
gi.workflows.import_workflow_dict(workflow_dict=wf_dict_2) # doesn't work with eval()
# # PRE FILLED METHOD
# wf_id = workflow.store()
# hi_id = gi.histories.get_histories()[0]["id"]
# print("Workflow id: " + wf_id)
# print("History id: " + hi_id)
# wf_show = gi.workflows.show_workflow(workflow_id=wf_id)
# # print(wf_show["inputs"]) # ->no inputs
# # workflow.port()
# li_id = gi.libraries.get_libraries()[0]["id"]
# # gi.folders.update_folder() # TODO: add method to enable data updates
# tsi = gi.toolshed.get_repositories() # tool shed repo access point
# # print(gi.users.get_users()) # TODO: users management
# # print(gi.libraries.show_library(li_id))
# # TODO: create a NEW history, import data to galaxy from library (src_data)
# fo_gi = gi.libraries.get_folders(library_id=li_id) # data location
# fo_id = {}
# current_fo_name = ""
# # print(fo_gi)
# # folders ids: access to data to run the first tools
# for i in fo_gi:
# for k, v in i.items():
# if k == "name":
# fo_id[v] = 0
# current_fo_name = v
# if k == "id":
# fo_id[current_fo_name] = v
# print("Folders id: ")
# for k, v in fo_id.items():
# print("\t" + k + ": " + v)
# workflow.show()
# # gi.workflows.run_workflow(workflow_id=wf_id) # pre filled workflow, use the set on runtime approach instead
# for testing, purge configured @ ~/config/galaxy.yml.docker_sample
# gi.histories.delete_history(history_id=hi_id, purge=True)
os.chdir(main_dir)
print("\n")
if __name__ == "__main__":
main()