diff options
Diffstat (limited to 'uploader/phenotypes/views.py')
| -rw-r--r-- | uploader/phenotypes/views.py | 77 |
1 files changed, 72 insertions, 5 deletions
diff --git a/uploader/phenotypes/views.py b/uploader/phenotypes/views.py index 3d2ff76..9c737fc 100644 --- a/uploader/phenotypes/views.py +++ b/uploader/phenotypes/views.py @@ -3,6 +3,7 @@ import sys import csv import uuid import json +import logging import datetime import tempfile from typing import Any @@ -900,9 +901,9 @@ def edit_download_phenotype_data(# pylint: disable=[unused-argument] filename = Path(tmpdir).joinpath("tempfile.tsv") with open(filename, mode="w") as outfile: outfile.write( - "# **DO NOT** delete the 'UniqueIdentifier' field. It is used " - "by the system to identify and edit the correct row(s) in the " - "database.\n") + "# **DO NOT** delete the 'UniqueIdentifier' row. It is used " + "by the system to identify and edit the correct rows and " + "columns in the database.\n") outfile.write( "# The '…_description' fields are useful for you to figure out " "what row you are working on. Changing any of this fields will " @@ -920,7 +921,8 @@ def edit_download_phenotype_data(# pylint: disable=[unused-argument] "Pre_publication_abbreviation", "Pre_publication_description", "Original_description", - "Post_publication_abbreviation" + "Post_publication_abbreviation", + "PubMed_ID" ] + samples_list, dialect="excel-tab") writer.writeheader() @@ -957,4 +959,69 @@ def edit_upload_phenotype_data(# pylint: disable=[unused-argument] dataset=dataset, activelink="edit-phenotype") - return "NOT Implemented: Would do actual edit." + edit_file = save_file(request.files["file-upload-bulk-edit-upload"], + Path(app.config["UPLOAD_FOLDER"])) + + from gn_libs import jobs as gnlibs_jobs + from gn_libs import sqlite3 + jobs_db = app.config["ASYNCHRONOUS_JOBS_SQLITE_DB"] + with sqlite3.connection(jobs_db) as conn: + job_id = uuid.uuid4() + _job = gnlibs_jobs.launch_job( + gnlibs_jobs.initialise_job(conn, + job_id, + [ + sys.executable, "-u", + "-m", "scripts.phenotypes_bulk_edit", + app.config["SQL_URI"], + jobs_db, + str(job_id), + "--log-level", + logging.getLevelName( + app.logger.getEffectiveLevel() + ).lower() + ], + "phenotype-bulk-edit", + extra_meta = { + "edit-file": str(edit_file) + }), + jobs_db, + f"{app.config['UPLOAD_FOLDER']}/job_errors", + worker_manager="gn_libs.jobs.launcher") + + + return """ + <p>The following steps need to be performed: + <ol> + <li>Check that all IDs exist</li> + <li>Check for mandatory values</li> + <li>Update descriptions in the database (where changed)</li> + <li>Update publications in the database (where changed): + <ol> + <li>If <strong>PubMed_ID</strong> exists in our database, simply update the + 'PublicationId' value in the 'PublishXRef' table.</li> + <li>If <strong>PubMed_ID</strong> does not exists in our database: + <ol> + <li>fetch the publication's details from PubMed using the new + <strong>PubMed_ID</strong> value.</li> + <li>create a new publication in our database using the fetched data</li> + <li>Update 'PublicationId' value in 'PublishXRef' with ID of newly created + publication</li> + </ol> + </ol> + </li> + <li>Update values in the database (where changed)</li> + </ol> + </p> + + <p><strong>Note:</strong> + <ul> + <li>If a strain that did not have a value is given a value, then we need to + add a new cross-reference for the new DataId created.</li> + <li>If a strain that had a value has its value deleted and left blank, we + need to remove the cross-reference for the existing DataId — or, should we + enter the NULL value instead? Removing the cross-reference might be more + trouble than it is worth.</li> + </ul> + </p> + """ |