diff options
Diffstat (limited to 'uploader/phenotypes')
| -rw-r--r-- | uploader/phenotypes/models.py | 30 | ||||
| -rw-r--r-- | uploader/phenotypes/views.py | 82 |
2 files changed, 101 insertions, 11 deletions
diff --git a/uploader/phenotypes/models.py b/uploader/phenotypes/models.py index 3946a0f..3d656d2 100644 --- a/uploader/phenotypes/models.py +++ b/uploader/phenotypes/models.py @@ -96,20 +96,34 @@ def dataset_phenotypes(# pylint: disable=[too-many-arguments, too-many-positiona xref_ids: tuple[int, ...] = tuple() ) -> tuple[dict, ...]: """Fetch the actual phenotypes.""" - _query = ( - "SELECT pheno.*, pxr.Id AS xref_id, pxr.InbredSetId, ist.InbredSetCode " + _narrow_by_ids = ( + f" AND pxr.Id IN ({', '.join(['%s'] * len(xref_ids))})" + if len(xref_ids) > 0 else "") + _narrow_by_limit = ( + f" LIMIT {limit} OFFSET {offset}" if bool(limit) else "") + _pub_query = ( + "SELECT pub.* " + "FROM PublishXRef AS pxr " + "INNER JOIN Publication AS pub ON pxr.PublicationId=pub.Id " + "WHERE pxr.InbredSetId=%s") + _narrow_by_ids + _pheno_query = (( + "SELECT pheno.*, pxr.Id AS xref_id, pxr.InbredSetId, pxr.PublicationId, " + "ist.InbredSetCode " "FROM Phenotype AS pheno " "INNER JOIN PublishXRef AS pxr ON pheno.Id=pxr.PhenotypeId " "INNER JOIN PublishFreeze AS pf ON pxr.InbredSetId=pf.InbredSetId " "INNER JOIN InbredSet AS ist ON pf.InbredSetId=ist.Id " - "WHERE pxr.InbredSetId=%s AND pf.Id=%s") + ( - f" AND pxr.Id IN ({', '.join(['%s'] * len(xref_ids))})" - if len(xref_ids) > 0 else "") + ( - f" LIMIT {limit} OFFSET {offset}" if bool(limit) else "") + "WHERE pxr.InbredSetId=%s AND pf.Id=%s") + + _narrow_by_ids + + _narrow_by_limit) with conn.cursor(cursorclass=DictCursor) as cursor: - cursor.execute(_query, (population_id, dataset_id) + xref_ids) + cursor.execute(_pub_query, (population_id,) + xref_ids) debug_query(cursor, logger) - return tuple(dict(row) for row in cursor.fetchall()) + _pubs = {row["Id"]: dict(row) for row in cursor.fetchall()} + cursor.execute(_pheno_query, (population_id, dataset_id) + xref_ids) + debug_query(cursor, logger) + return tuple({**dict(row), "publication": _pubs[row["PublicationId"]]} + for row in cursor.fetchall()) def __phenotype_se__(cursor: BaseCursor, xref_id, dataids_and_strainids): diff --git a/uploader/phenotypes/views.py b/uploader/phenotypes/views.py index ab3b3d2..c03f3f5 100644 --- a/uploader/phenotypes/views.py +++ b/uploader/phenotypes/views.py @@ -1,4 +1,6 @@ """Views handling ('classical') phenotypes."""# pylint: disable=[too-many-lines] +import io +import csv import sys import uuid import json @@ -21,12 +23,14 @@ from gn_libs import jobs as gnlibs_jobs from gn_libs.jobs.jobs import JobNotFound from gn_libs.mysqldb import database_connection +from werkzeug.datastructures import Headers from flask import (flash, request, jsonify, redirect, Blueprint, - current_app as app) + current_app as app, + Response as FlaskResponse) from r_qtl import r_qtl2_qc as rqc from r_qtl import exceptions as rqe @@ -313,6 +317,11 @@ def create_dataset(species: dict, population: dict, **kwargs):# pylint: disable= dataset_shortname = ( form["dataset-shortname"] or form["dataset-name"]).strip() _pheno_dataset = save_new_dataset( + # It's not necessary to update the authorisation server to register + # new phenotype resource here, since each phenotype trait can, in + # theory, have its own access control allowing/disallowing access to + # it. In practice, however, we tend to gather multiple traits into a + # single resource for access control. cursor, population["Id"], form["dataset-name"].strip(), @@ -522,6 +531,65 @@ def job_status( @phenotypesbp.route( "<int:species_id>/populations/<int:population_id>/phenotypes/datasets" + "/<int:dataset_id>/job/<uuid:job_id>/download-errors", + methods=["GET"]) +@require_login +@with_dataset( + species_redirect_uri="species.populations.phenotypes.index", + population_redirect_uri="species.populations.phenotypes.select_population", + redirect_uri="species.populations.phenotypes.list_datasets") +def download_errors( + species: dict, + population: dict, + dataset: dict, + job_id: uuid.UUID, + **kwargs):# pylint: disable=[unused-argument] + """Download the list of errors as a CSV file.""" + with Redis.from_url(app.config["REDIS_URL"], decode_responses=True) as rconn: + try: + job = jobs.job(rconn, jobs.jobsnamespace(), str(job_id)) + _prefix_ = jobs.jobsnamespace() + _jobid_ = job['jobid'] + def __generate_chunks__(): + _errors_ = ( + json.loads(error) + for key in rconn.keys( + f"{_prefix_}:{str(_jobid_)}:*:errors:*") + for error in rconn.lrange(key, 0, -1)) + _chunk_no_ = 0 + _all_errors_printed_ = False + while not _all_errors_printed_: + _chunk_ = [] + try: + for _ in range(0, 1000): + _chunk_.append(next(_errors_)) + except StopIteration: + _all_errors_printed_ = True + if len(_chunk_) <= 0: + raise + + _out_ = io.StringIO() + _writer_ = csv.DictWriter(_out_, fieldnames=tuple(_chunk_[0].keys())) + if _chunk_no_ == 0: + _writer_.writeheader() + _writer_.writerows(_chunk_) + _chunk_no_ += 1 + yield _out_.getvalue() + if _all_errors_printed_: + return + + headers = Headers() + headers.set("Content-Disposition", + "attachment", + filename=f"{job['job-type']}_{job['jobid']}.csv") + return FlaskResponse( + __generate_chunks__(), mimetype="text/csv", headers=headers) + except jobs.JobNotFound as _jnf: + return render_template("jobs/job-not-found.html", job_id=job_id) + + +@phenotypesbp.route( + "<int:species_id>/populations/<int:population_id>/phenotypes/datasets" "/<int:dataset_id>/job/<uuid:job_id>/review", methods=["GET"]) @require_login @@ -599,6 +667,8 @@ def review_job_data( conn, int(_job_metadata["publicationid"])) if _job_metadata.get("publicationid") else None), + user=session.user_details(), + timestamp=datetime.datetime.now().isoformat(), activelink="add-phenotypes") @@ -678,9 +748,15 @@ def load_data_to_database( "publication_id": _meta["publicationid"], "authserver": oauth2client.authserver_uri(), "token": token["access_token"], + "dataname": request.form["data_name"].strip(), "success_handler": ( "uploader.phenotypes.views" - ".load_phenotypes_success_handler") + ".load_phenotypes_success_handler"), + **{ + key: request.form[key] + for key in ("data_description",) + if key in request.form.keys() + } }, external_id=session.logged_in_user_id()) ).then( @@ -1262,7 +1338,7 @@ def delete_phenotypes(# pylint: disable=[unused-argument, too-many-locals] loglevel=_loglevel) ).either(__handle_error__, proceed_to_job_status) case _: - _phenos = tuple() + _phenos: tuple[dict, ...] = tuple() if len(xref_ids) > 0: _phenos = dataset_phenotypes( conn, population["Id"], dataset["Id"], xref_ids=xref_ids) |