view nist_wrapper.py @ 1:964ec10d568c

new NIST wrapper demo tools
author pieter.lukasse@wur.nl
date Thu, 22 Jan 2015 16:15:30 +0100
parents cce6989ed423
children c3dc158717fc
line wrap: on
line source

#!/usr/bin/env python
# encoding: utf-8
'''
Module wrapping the NIST MSSEARCH application for matching 
spectra one or more spectra libraries.   
'''
import csv
import sys
import fileinput
import urllib2
import time
import utils
import uuid
import os
import subprocess
from report_generator import ReportGenerator

__author__ = "Pieter Lukasse"
__contact__ = "pieterlukasse@gmail.com,pieter.lukasse@wur.nl"
__copyright__ = "Copyright, 2015"
__license__ = "Apache v2"



def _prepare_NIST(uuid_value, nist_home_dir, nist_ini_file, spectrum_file, is_wine):
    '''
     executes the following steps:
     - copy  nist_home_dir folder to  nist_home_dir+_uuid
     - copy spectrum_file.msp to  ~/.wine/drive_c/NIST_uid
     - creates nist_home_dir+_uuid/MSSEARCH/AUTOIMP.MSD -> pointing to C:\NIST_uid\MSSEARCH\temp.msd (in case of is_wine) or to nist_home_dir+_uuidM\SSEARCH\temp.msd
     - creates nist_home_dir+_uuid/MSSEARCH/temp.msd -> pointing to  C:\NIST_uid\spectrum_file.msp (in case of is_wine) or to nist_home_dir+_uuid\spectrum_file.msp
       and the  text "10 724" in the second row
     - copy nistms.INI to  nist_home_dir+_uuid/MSSEARCH , overwriting the existing one
     - in case of is_wine:
        replace all occurrences of C:\NIST with C:\NIST_uid in this new nistms.INI
       else:
        replace all occurrences of C:\NIST with nist_home_dir+_uuid in this new nistms.INI
    '''
    if nist_home_dir.endswith("/") or nist_home_dir.endswith("\\"):
        nist_home_dir = nist_home_dir[:-1]
        
    # small validation for wine scenario
    if is_wine and not nist_home_dir.endswith("drive_c/NIST"):
        raise Exception('Error: invalid NIST home. For wine usage NIST home dir must be in the .wine folder and then in drive_c/NIST')
            
    new_nist_home = nist_home_dir+uuid_value
    utils.copy_dir(nist_home_dir, new_nist_home)
    
    utils.copy_file(spectrum_file, new_nist_home+"/spectrum_file.msp")
    
    # remove old file:
    os.remove(new_nist_home+"/MSSEARCH/AUTOIMP.MSD")
    with open(new_nist_home + "/MSSEARCH/AUTOIMP.MSD", "a") as text_file:
        if is_wine:
            text_file.write("C:\\NIST" + uuid_value + "\\MSSEARCH\\temp.msd")
        else:
            text_file.write(new_nist_home + "\\MSSEARCH\\temp.msd")
    
    with open(new_nist_home + "/MSSEARCH/temp.msd", "a") as text_file:
        if is_wine:
            text_file.write("C:\\NIST" + uuid_value + "\\spectrum_file.msp\n")
        else:
            text_file.write(new_nist_home + "\\spectrum_file.msp\n")
        text_file.write("10 724")
    
    replacement_text = new_nist_home
    if is_wine:
        replacement_text = "C:\\NIST" + uuid_value
    
    # remove old file
    os.remove(new_nist_home+"/MSSEARCH/nistms.INI")
    # make new one
    o = open(new_nist_home+"/MSSEARCH/nistms.INI","a") #open for append
    # TODO : this loop/replace below is a bit limited to specific variables...either test different NIST versions or make more generic (harder in case of wine, or we need extra "home in .INI file"  parameter):
    for line in open(nist_ini_file):
        if "Library Directory=" in line:
            line = "Library Directory="+ new_nist_home + "\\MSSEARCH\\\n"
        if "Dir=" in line:
            line = "Dir="+ replacement_text + "\\MSSEARCH\\\n"

        o.write(line) 
    o.close()
    
    return new_nist_home

def _run_NIST(new_nist_home, output_file, is_wine):
    '''
      - run : (wine) new_nist_home/MSSEARCH/nistms$.exe /INSTRUMENT /PAR=2
      - monitor : new_nist_home/MSSEARCH/SRCREADY.TXT for content = "1"
         - when ready:
           > copy SRCRESLT.TXT to output_file
           > kill nist process
           > (optional)remove ~/.wine/drive_c/NIST_uid/
           > finish
    '''
    # to avoid conflicts in the orphan process killing (see end of this method), we will 
    # only run NIST again after previous nistms.exe process has been killed:
    # TODO : solution is currently only for wine (in the windows context the solution is not there yet, but parallel calls are not expected as in windows we only run tests one by one for now)
#     if is_wine:
#         while True:
#             # check if process exists. If not, break loop and continue
#             pid = utils.get_process_pid("nistms.exe")
#             if pid == -1:
#                 break
#             time.sleep(2)
    
    
    # remove old file, if it is there:
    file_to_monitor = new_nist_home+"/MSSEARCH/SRCREADY.TXT"
    if os.path.exists(file_to_monitor):
        os.remove(file_to_monitor)
    
    exec_path = new_nist_home + "/MSSEARCH/nistms$.exe" 
    
    pro = ""
    if is_wine:
        print "calling wine with " + exec_path
        cmd = ["wine "+ exec_path + " /INSTRUMENT /PAR=2"]
        # The os.setsid() is passed in the argument preexec_fn so
        # it's run after the fork() and before  exec() to run the shell.
        pro = subprocess.Popen(cmd, stdout=subprocess.PIPE, 
                           shell=True, preexec_fn=os.setsid)
    else:
        cmd = [
            exec_path,
            "/INSTRUMENT",
            "/PAR=2"]
        subprocess.call(cmd)
     
    
    # monitor process by checking state file:
    while True:
        # check if SRCREADY.TXT is there already:
        if os.path.exists(file_to_monitor):
            break
        time.sleep(2)
        
    # kill process:
    #p.terminate() - not needed, nistm$ will terminate...nistms.exe is the one that 
    #stays open...and orphan..killing it:
    
    if is_wine:
#         pid = utils.get_process_pid("nistms.exe")
#         os.kill(pid, 9)
        os.killpg(pro.pid, 9)    
    else:
        # windows case:
        proc_name = "nistms.exe"
        os.system("taskkill /f /im " + proc_name)
  
    # copy SRCRESLT.TXT to output_file
    result_file = new_nist_home+"/MSSEARCH/SRCRESLT.TXT"
    utils.copy_file(result_file, output_file)
    

def _create_html_report(output_html_report, output_html_report_files_path, hits_dict, spectra_dict):
    '''
    This report will contain a page that displays essentially the same list as found in the 
    tabular output file (rendered with datatables jquery plugin), with some extra features:
     - when user clicks on an entry, it should display the query spectrum and the hit spectrum
       in "head to tail" and "difference" mode (see galaxy/report_example.png)
        -> the query spectrum can be generated from the data in the input MSP file
        -> the library "online representative" spectrum can be generated from data returned by http://webbook.nist.gov/cgi/cbook.cgi?JCAMP=C537268&Index=0&Type=Mass ,
           where C537268 in this case is the CAS ID without the '-' separators
    '''
    # step 1 : generate HTML via the jinja template engine
    # step 1.1: make sure to link the query spectrum data to the corresponding html object for quick rendering when needed
    
    html_file = open(output_html_report,'w')
    html_render = ReportGenerator(os.path.dirname(__file__), 'templates/main_template.html',hits_dict, spectra_dict)
    html_render.render(html_file)
    # copy necessary .js files as well:
    templates_folder = os.path.dirname(__file__) + '/templates/'
    utils.copy_file(templates_folder + 'spectrum_gen.js', output_html_report_files_path+"/spectrum_gen.js")
    utils.copy_dir(templates_folder + 'lib', output_html_report_files_path+'/lib' )
    utils.copy_dir(templates_folder + 'images', output_html_report_files_path+'/images' )
    
    html_file.close()

     
     
     
    
    return None

def _get_extra_info_and_link_cols(data_found, data_type_found, query_link):
    '''
    This method will go over the data found and will return a 
    list with the following items:
    - Experiment details where hits have been found :
        'organism', 'tissue','experiment_name','user_name','column_type'
    - Link that executes same query
        
    '''
    # set() makes a unique list:
    organism_set = []
    tissue_set = []
    experiment_name_set = []
    user_name_set = []
    column_type_set = []
    cas_nr_set = []
    
    if 'organism' in data_found:
        organism_set = set(data_found['organism'])
    if 'tissue' in data_found:
        tissue_set = set(data_found['tissue'])
    if 'experiment_name' in data_found:
        experiment_name_set = set(data_found['experiment_name'])
    if 'user_name' in data_found:
        user_name_set = set(data_found['user_name'])
    if 'column_type' in data_found:
        column_type_set = set(data_found['column_type'])
    if 'CAS' in data_found:
        cas_nr_set = set(data_found['CAS'])        
    
    
    result = [data_type_found,
                            
              #To let Excel interpret as link, use e.g. =HYPERLINK("http://stackoverflow.com", "friendly name"): 
              "=HYPERLINK(\""+ query_link + "\", \"Link to entries found in DB \")"]
    return result



    
    
# alternative: ?    
#     s = requests.Session()
#     s.verify = False
#     #s.auth = (token01, token02)
#     resp = s.get(url, params={'name': 'anonymous'}, stream=True)
#     content = resp.content
#     # transform to dictionary:
    


def _save_data(data_rows, headers, out_csv):
    '''
    Writes tab-separated data to file
    @param data_rows: dictionary containing merged/enriched dataset
    @param out_csv: output csv file
    '''

    # Open output file for writing
    outfile_single_handle = open(out_csv, 'wb')
    output_single_handle = csv.writer(outfile_single_handle, delimiter="\t")

    # Write headers
    output_single_handle.writerow(headers)

    # Write one line for each row
    for data_row in data_rows:
        output_single_handle.writerow(data_row)

def _get_metexp_URL(metexp_dblink_file):
    '''
    Read out and return the URL stored in the given file.
    '''
    file_input = fileinput.input(metexp_dblink_file)
    try:
        for line in file_input:
            if line[0] != '#':
                # just return the first line that is not a comment line:
                return line
    finally:
        file_input.close()
    

def main():
    '''
    Wrapper main function
    
    The input expected is:
     NIST_HOME dir
     nistms.INI
     spectrum_file.msp 
     outputfileName 
     (optional) htmlReportFile
     (optional) htmlReportFile.files_path
    '''
    seconds_start = int(round(time.time()))
    
    nist_home_dir = sys.argv[1]
    nist_ini_file = sys.argv[2]
    spectrum_file = sys.argv[3]
    nist_output_file = sys.argv[4]
    final_output_file = sys.argv[5]
    # html report pars:
    output_html_report = None
    output_html_report_files_path = None
    if len(sys.argv) > 6:
        output_html_report = sys.argv[6]
        output_html_report_files_path = sys.argv[7]
    
    is_wine = False
    if "wine" in nist_home_dir:
        is_wine = True
    
    uuid_value = str(uuid.uuid4()) 

    # prepare NIST environment for running:
    new_nist_home = _prepare_NIST(uuid_value, nist_home_dir, nist_ini_file, spectrum_file, is_wine)
    
    # run NIST search command:
    _run_NIST(new_nist_home, nist_output_file, is_wine)

    # write output tabular:
    hits_dict = utils.get_nist_out_as_dict(nist_output_file) 
    utils.save_dict_as_tsv(hits_dict, final_output_file)
    
    # create report:
    if len(sys.argv) > 6:
        spectra_dict = utils.get_spectra_file_as_dict(spectrum_file)
        _create_html_report(output_html_report, output_html_report_files_path, hits_dict, spectra_dict)


    #_save_data(enriched_data, headers, output_result)
    
    seconds_end = int(round(time.time()))
    print "Took " + str(seconds_end - seconds_start) + " seconds"
                      
                      

if __name__ == '__main__':
    main()