pfurio

#!/usr/bin/env python

from pymongo import MongoClient
from bson import ObjectId
import datetime
import argparse

parser = argparse.ArgumentParser(description='Get the number of times each user have logged in during a period of time.')
parser.add_argument('host', metavar='host', help='Mongo host and port. Example: localhost:27017.')
# parser.add_argument('database', metavar='database', help='Mongo database name. Example: opencga_catalog.')

pfurio

// 1. Load the function from https://gist.github.com/pfurio/f7cd90af08e0073699f0beeeef1958ba

// 2. Load all the scripts present in the gist in the CATALOG database

pfurio

// Extracted from https://stackoverflow.com/questions/3066586/get-string-in-yyyymmdd-format-from-js-date-object
Date.prototype.opencgaDate = function() {
  // We will get the current day minus 7 days
  this.setDate(this.getDate() - 7);
  var mm = this.getMonth() + 1; // getMonth() is zero-based
  var dd = this.getDate();

  return [this.getFullYear(),
          (mm>9 ? '' : '0') + mm,
          (dd>9 ? '' : '0') + dd,

pfurio

// Use migration script https://gist.github.com/pfurio/f7cd90af08e0073699f0beeeef1958ba

/*
Migrate NUMERIC variables to DOUBLE https://github.com/opencb/opencga/issues/545
*/

// This function receives an array of variables (not variableSet),
// and will modify NUMERIC variables to DOUBLE recursively
function modifyNumeric(variables) {
    var modified = false;

pfurio

migrateCollection("metadata", {}, {config: 1, acl: 1}, function (bulk, doc) {
    var id = doc._id;
 
    var permissions = ["VIEW_FILE_HEADERS", "VIEW_FILE_CONTENTS", "VIEW_FILES", "WRITE_FILES", "VIEW_JOBS", "WRITE_JOBS"];
    doc.acl[0].permissions = permissions;
    doc.config.acl[0].permissions = permissions;
 
    bulk.find({_id: id}).update({$set: {"acl": doc.acl, "config.acl": doc.config.acl}});
});
 

pfurio

function migrateCollection(collection, query, projection, migrateFunc) {
    var bulk = db.getCollection(collection).initializeOrderedBulkOp();
    var count = 0;
    var bulkSize = 500;
    db.getCollection(collection).find(query,projection).forEach(function(doc) {
        migrateFunc(bulk, doc);
        if ( bulk.nUpdateOps + bulk.nInsertOps + bulk.nRemoveOps >= bulkSize ) {
            count += bulk.nUpdateOps + bulk.nInsertOps + bulk.nRemoveOps;
            print("Execute bulk! " + count);
            bulk.execute();

pfurio

# -*- coding: utf-8 -*-

# Author: Pedro Furió Tarí
# Data: 06/05/2013
#
# This script will take a sorted and indexed bam file and return different text files with the counts
# Useful when doing a trimming step for pair-end. After trimming, we have to make sure that we have the pairs correctly paired

import getopt, sys, os.path
import pysam