Copybara | 854996b | 2021-09-07 19:36:02 +0000 | [diff] [blame^] | 1 | # Copyright 2016 The Chromium Authors. All rights reserved. |
| 2 | # Use of this source code is governed by a BSD-style |
| 3 | # license that can be found in the LICENSE file or at |
| 4 | # https://developers.google.com/open-source/licenses/bsd |
| 5 | |
| 6 | """A set of functions that provide persistence for Monorail issue tracking. |
| 7 | |
| 8 | This module provides functions to get, update, create, and (in some |
| 9 | cases) delete each type of business object. It provides a logical |
| 10 | persistence layer on top of an SQL database. |
| 11 | |
| 12 | Business objects are described in tracker_pb2.py and tracker_bizobj.py. |
| 13 | """ |
| 14 | from __future__ import print_function |
| 15 | from __future__ import division |
| 16 | from __future__ import absolute_import |
| 17 | |
| 18 | import collections |
| 19 | import json |
| 20 | import logging |
| 21 | import os |
| 22 | import time |
| 23 | import uuid |
| 24 | |
| 25 | from google.appengine.api import app_identity |
| 26 | from google.appengine.api import images |
| 27 | from third_party import cloudstorage |
| 28 | |
| 29 | import settings |
| 30 | from features import filterrules_helpers |
| 31 | from framework import authdata |
| 32 | from framework import exceptions |
| 33 | from framework import framework_bizobj |
| 34 | from framework import framework_constants |
| 35 | from framework import framework_helpers |
| 36 | from framework import gcs_helpers |
| 37 | from framework import permissions |
| 38 | from framework import sql |
| 39 | from infra_libs import ts_mon |
| 40 | from proto import project_pb2 |
| 41 | from proto import tracker_pb2 |
| 42 | from services import caches |
| 43 | from services import tracker_fulltext |
| 44 | from tracker import tracker_bizobj |
| 45 | from tracker import tracker_helpers |
| 46 | |
| 47 | # TODO(jojwang): monorail:4693, remove this after all 'stable-full' |
| 48 | # gates have been renamed to 'stable'. |
| 49 | FLT_EQUIVALENT_GATES = {'stable-full': 'stable', |
| 50 | 'stable': 'stable-full'} |
| 51 | |
| 52 | ISSUE_TABLE_NAME = 'Issue' |
| 53 | ISSUESUMMARY_TABLE_NAME = 'IssueSummary' |
| 54 | ISSUE2LABEL_TABLE_NAME = 'Issue2Label' |
| 55 | ISSUE2COMPONENT_TABLE_NAME = 'Issue2Component' |
| 56 | ISSUE2CC_TABLE_NAME = 'Issue2Cc' |
| 57 | ISSUE2NOTIFY_TABLE_NAME = 'Issue2Notify' |
| 58 | ISSUE2FIELDVALUE_TABLE_NAME = 'Issue2FieldValue' |
| 59 | COMMENT_TABLE_NAME = 'Comment' |
| 60 | COMMENTCONTENT_TABLE_NAME = 'CommentContent' |
| 61 | COMMENTIMPORTER_TABLE_NAME = 'CommentImporter' |
| 62 | ATTACHMENT_TABLE_NAME = 'Attachment' |
| 63 | ISSUERELATION_TABLE_NAME = 'IssueRelation' |
| 64 | DANGLINGRELATION_TABLE_NAME = 'DanglingIssueRelation' |
| 65 | ISSUEUPDATE_TABLE_NAME = 'IssueUpdate' |
| 66 | ISSUEFORMERLOCATIONS_TABLE_NAME = 'IssueFormerLocations' |
| 67 | REINDEXQUEUE_TABLE_NAME = 'ReindexQueue' |
| 68 | LOCALIDCOUNTER_TABLE_NAME = 'LocalIDCounter' |
| 69 | ISSUESNAPSHOT_TABLE_NAME = 'IssueSnapshot' |
| 70 | ISSUESNAPSHOT2CC_TABLE_NAME = 'IssueSnapshot2Cc' |
| 71 | ISSUESNAPSHOT2COMPONENT_TABLE_NAME = 'IssueSnapshot2Component' |
| 72 | ISSUESNAPSHOT2LABEL_TABLE_NAME = 'IssueSnapshot2Label' |
| 73 | ISSUEPHASEDEF_TABLE_NAME = 'IssuePhaseDef' |
| 74 | ISSUE2APPROVALVALUE_TABLE_NAME = 'Issue2ApprovalValue' |
| 75 | ISSUEAPPROVAL2APPROVER_TABLE_NAME = 'IssueApproval2Approver' |
| 76 | ISSUEAPPROVAL2COMMENT_TABLE_NAME = 'IssueApproval2Comment' |
| 77 | |
| 78 | |
| 79 | ISSUE_COLS = [ |
| 80 | 'id', 'project_id', 'local_id', 'status_id', 'owner_id', 'reporter_id', |
| 81 | 'opened', 'closed', 'modified', |
| 82 | 'owner_modified', 'status_modified', 'component_modified', |
| 83 | 'derived_owner_id', 'derived_status_id', |
| 84 | 'deleted', 'star_count', 'attachment_count', 'is_spam'] |
| 85 | ISSUESUMMARY_COLS = ['issue_id', 'summary'] |
| 86 | ISSUE2LABEL_COLS = ['issue_id', 'label_id', 'derived'] |
| 87 | ISSUE2COMPONENT_COLS = ['issue_id', 'component_id', 'derived'] |
| 88 | ISSUE2CC_COLS = ['issue_id', 'cc_id', 'derived'] |
| 89 | ISSUE2NOTIFY_COLS = ['issue_id', 'email'] |
| 90 | ISSUE2FIELDVALUE_COLS = [ |
| 91 | 'issue_id', 'field_id', 'int_value', 'str_value', 'user_id', 'date_value', |
| 92 | 'url_value', 'derived', 'phase_id'] |
| 93 | # Explicitly specify column 'Comment.id' to allow joins on other tables that |
| 94 | # have an 'id' column. |
| 95 | COMMENT_COLS = [ |
| 96 | 'Comment.id', 'issue_id', 'created', 'Comment.project_id', 'commenter_id', |
| 97 | 'deleted_by', 'Comment.is_spam', 'is_description', |
| 98 | 'commentcontent_id'] # Note: commentcontent_id must be last. |
| 99 | COMMENTCONTENT_COLS = [ |
| 100 | 'CommentContent.id', 'content', 'inbound_message'] |
| 101 | COMMENTIMPORTER_COLS = ['comment_id', 'importer_id'] |
| 102 | ABBR_COMMENT_COLS = ['Comment.id', 'commenter_id', 'deleted_by', |
| 103 | 'is_description'] |
| 104 | ATTACHMENT_COLS = [ |
| 105 | 'id', 'issue_id', 'comment_id', 'filename', 'filesize', 'mimetype', |
| 106 | 'deleted', 'gcs_object_id'] |
| 107 | ISSUERELATION_COLS = ['issue_id', 'dst_issue_id', 'kind', 'rank'] |
| 108 | ABBR_ISSUERELATION_COLS = ['dst_issue_id', 'rank'] |
| 109 | DANGLINGRELATION_COLS = [ |
| 110 | 'issue_id', 'dst_issue_project', 'dst_issue_local_id', |
| 111 | 'ext_issue_identifier', 'kind'] |
| 112 | ISSUEUPDATE_COLS = [ |
| 113 | 'id', 'issue_id', 'comment_id', 'field', 'old_value', 'new_value', |
| 114 | 'added_user_id', 'removed_user_id', 'custom_field_name'] |
| 115 | ISSUEFORMERLOCATIONS_COLS = ['issue_id', 'project_id', 'local_id'] |
| 116 | REINDEXQUEUE_COLS = ['issue_id', 'created'] |
| 117 | ISSUESNAPSHOT_COLS = ['id', 'issue_id', 'shard', 'project_id', 'local_id', |
| 118 | 'reporter_id', 'owner_id', 'status_id', 'period_start', 'period_end', |
| 119 | 'is_open'] |
| 120 | ISSUESNAPSHOT2CC_COLS = ['issuesnapshot_id', 'cc_id'] |
| 121 | ISSUESNAPSHOT2COMPONENT_COLS = ['issuesnapshot_id', 'component_id'] |
| 122 | ISSUESNAPSHOT2LABEL_COLS = ['issuesnapshot_id', 'label_id'] |
| 123 | ISSUEPHASEDEF_COLS = ['id', 'name', 'rank'] |
| 124 | ISSUE2APPROVALVALUE_COLS = ['approval_id', 'issue_id', 'phase_id', |
| 125 | 'status', 'setter_id', 'set_on'] |
| 126 | ISSUEAPPROVAL2APPROVER_COLS = ['approval_id', 'approver_id', 'issue_id'] |
| 127 | ISSUEAPPROVAL2COMMENT_COLS = ['approval_id', 'comment_id'] |
| 128 | |
| 129 | CHUNK_SIZE = 1000 |
| 130 | |
| 131 | |
| 132 | class IssueIDTwoLevelCache(caches.AbstractTwoLevelCache): |
| 133 | """Class to manage RAM and memcache for Issue IDs.""" |
| 134 | |
| 135 | def __init__(self, cache_manager, issue_service): |
| 136 | super(IssueIDTwoLevelCache, self).__init__( |
| 137 | cache_manager, 'issue_id', 'issue_id:', int, |
| 138 | max_size=settings.issue_cache_max_size) |
| 139 | self.issue_service = issue_service |
| 140 | |
| 141 | def _MakeCache(self, cache_manager, kind, max_size=None): |
| 142 | """Override normal RamCache creation with ValueCentricRamCache.""" |
| 143 | return caches.ValueCentricRamCache(cache_manager, kind, max_size=max_size) |
| 144 | |
| 145 | def _DeserializeIssueIDs(self, project_local_issue_ids): |
| 146 | """Convert database rows into a dict {(project_id, local_id): issue_id}.""" |
| 147 | return {(project_id, local_id): issue_id |
| 148 | for (project_id, local_id, issue_id) in project_local_issue_ids} |
| 149 | |
| 150 | def FetchItems(self, cnxn, keys): |
| 151 | """On RAM and memcache miss, hit the database.""" |
| 152 | local_ids_by_pid = collections.defaultdict(list) |
| 153 | for project_id, local_id in keys: |
| 154 | local_ids_by_pid[project_id].append(local_id) |
| 155 | |
| 156 | where = [] # We OR per-project pairs of conditions together. |
| 157 | for project_id, local_ids_in_project in local_ids_by_pid.items(): |
| 158 | term_str = ('(Issue.project_id = %%s AND Issue.local_id IN (%s))' % |
| 159 | sql.PlaceHolders(local_ids_in_project)) |
| 160 | where.append((term_str, [project_id] + local_ids_in_project)) |
| 161 | |
| 162 | rows = self.issue_service.issue_tbl.Select( |
| 163 | cnxn, cols=['project_id', 'local_id', 'id'], |
| 164 | where=where, or_where_conds=True) |
| 165 | return self._DeserializeIssueIDs(rows) |
| 166 | |
| 167 | def _KeyToStr(self, key): |
| 168 | """This cache uses pairs of ints as keys. Convert them to strings.""" |
| 169 | return '%d,%d' % key |
| 170 | |
| 171 | def _StrToKey(self, key_str): |
| 172 | """This cache uses pairs of ints as keys. Convert them from strings.""" |
| 173 | project_id_str, local_id_str = key_str.split(',') |
| 174 | return int(project_id_str), int(local_id_str) |
| 175 | |
| 176 | |
| 177 | class IssueTwoLevelCache(caches.AbstractTwoLevelCache): |
| 178 | """Class to manage RAM and memcache for Issue PBs.""" |
| 179 | |
| 180 | def __init__( |
| 181 | self, cache_manager, issue_service, project_service, config_service): |
| 182 | super(IssueTwoLevelCache, self).__init__( |
| 183 | cache_manager, 'issue', 'issue:', tracker_pb2.Issue, |
| 184 | max_size=settings.issue_cache_max_size) |
| 185 | self.issue_service = issue_service |
| 186 | self.project_service = project_service |
| 187 | self.config_service = config_service |
| 188 | |
| 189 | def _UnpackIssue(self, cnxn, issue_row): |
| 190 | """Partially construct an issue object using info from a DB row.""" |
| 191 | (issue_id, project_id, local_id, status_id, owner_id, reporter_id, |
| 192 | opened, closed, modified, owner_modified, status_modified, |
| 193 | component_modified, derived_owner_id, derived_status_id, |
| 194 | deleted, star_count, attachment_count, is_spam) = issue_row |
| 195 | |
| 196 | issue = tracker_pb2.Issue() |
| 197 | project = self.project_service.GetProject(cnxn, project_id) |
| 198 | issue.project_name = project.project_name |
| 199 | issue.issue_id = issue_id |
| 200 | issue.project_id = project_id |
| 201 | issue.local_id = local_id |
| 202 | if status_id is not None: |
| 203 | status = self.config_service.LookupStatus(cnxn, project_id, status_id) |
| 204 | issue.status = status |
| 205 | issue.owner_id = owner_id or 0 |
| 206 | issue.reporter_id = reporter_id or 0 |
| 207 | issue.derived_owner_id = derived_owner_id or 0 |
| 208 | if derived_status_id is not None: |
| 209 | derived_status = self.config_service.LookupStatus( |
| 210 | cnxn, project_id, derived_status_id) |
| 211 | issue.derived_status = derived_status |
| 212 | issue.deleted = bool(deleted) |
| 213 | if opened: |
| 214 | issue.opened_timestamp = opened |
| 215 | if closed: |
| 216 | issue.closed_timestamp = closed |
| 217 | if modified: |
| 218 | issue.modified_timestamp = modified |
| 219 | if owner_modified: |
| 220 | issue.owner_modified_timestamp = owner_modified |
| 221 | if status_modified: |
| 222 | issue.status_modified_timestamp = status_modified |
| 223 | if component_modified: |
| 224 | issue.component_modified_timestamp = component_modified |
| 225 | issue.star_count = star_count |
| 226 | issue.attachment_count = attachment_count |
| 227 | issue.is_spam = bool(is_spam) |
| 228 | return issue |
| 229 | |
| 230 | def _UnpackFieldValue(self, fv_row): |
| 231 | """Construct a field value object from a DB row.""" |
| 232 | (issue_id, field_id, int_value, str_value, user_id, date_value, url_value, |
| 233 | derived, phase_id) = fv_row |
| 234 | fv = tracker_bizobj.MakeFieldValue( |
| 235 | field_id, int_value, str_value, user_id, date_value, url_value, |
| 236 | bool(derived), phase_id=phase_id) |
| 237 | return fv, issue_id |
| 238 | |
| 239 | def _UnpackApprovalValue(self, av_row): |
| 240 | """Contruct an ApprovalValue PB from a DB row.""" |
| 241 | (approval_id, issue_id, phase_id, status, setter_id, set_on) = av_row |
| 242 | if status: |
| 243 | status_enum = tracker_pb2.ApprovalStatus(status.upper()) |
| 244 | else: |
| 245 | status_enum = tracker_pb2.ApprovalStatus.NOT_SET |
| 246 | av = tracker_pb2.ApprovalValue( |
| 247 | approval_id=approval_id, setter_id=setter_id, set_on=set_on, |
| 248 | status=status_enum, phase_id=phase_id) |
| 249 | return av, issue_id |
| 250 | |
| 251 | def _UnpackPhase(self, phase_row): |
| 252 | """Construct a Phase PB from a DB row.""" |
| 253 | (phase_id, name, rank) = phase_row |
| 254 | phase = tracker_pb2.Phase( |
| 255 | phase_id=phase_id, name=name, rank=rank) |
| 256 | return phase |
| 257 | |
| 258 | def _DeserializeIssues( |
| 259 | self, cnxn, issue_rows, summary_rows, label_rows, component_rows, |
| 260 | cc_rows, notify_rows, fieldvalue_rows, relation_rows, |
| 261 | dangling_relation_rows, phase_rows, approvalvalue_rows, |
| 262 | av_approver_rows): |
| 263 | """Convert the given DB rows into a dict of Issue PBs.""" |
| 264 | results_dict = {} |
| 265 | for issue_row in issue_rows: |
| 266 | issue = self._UnpackIssue(cnxn, issue_row) |
| 267 | results_dict[issue.issue_id] = issue |
| 268 | |
| 269 | for issue_id, summary in summary_rows: |
| 270 | results_dict[issue_id].summary = summary |
| 271 | |
| 272 | # TODO(jrobbins): it would be nice to order labels by rank and name. |
| 273 | for issue_id, label_id, derived in label_rows: |
| 274 | issue = results_dict.get(issue_id) |
| 275 | if not issue: |
| 276 | logging.info('Got label for an unknown issue: %r %r', |
| 277 | label_rows, issue_rows) |
| 278 | continue |
| 279 | label = self.config_service.LookupLabel(cnxn, issue.project_id, label_id) |
| 280 | assert label, ('Label ID %r on IID %r not found in project %r' % |
| 281 | (label_id, issue_id, issue.project_id)) |
| 282 | if derived: |
| 283 | results_dict[issue_id].derived_labels.append(label) |
| 284 | else: |
| 285 | results_dict[issue_id].labels.append(label) |
| 286 | |
| 287 | for issue_id, component_id, derived in component_rows: |
| 288 | if derived: |
| 289 | results_dict[issue_id].derived_component_ids.append(component_id) |
| 290 | else: |
| 291 | results_dict[issue_id].component_ids.append(component_id) |
| 292 | |
| 293 | for issue_id, user_id, derived in cc_rows: |
| 294 | if derived: |
| 295 | results_dict[issue_id].derived_cc_ids.append(user_id) |
| 296 | else: |
| 297 | results_dict[issue_id].cc_ids.append(user_id) |
| 298 | |
| 299 | for issue_id, email in notify_rows: |
| 300 | results_dict[issue_id].derived_notify_addrs.append(email) |
| 301 | |
| 302 | for fv_row in fieldvalue_rows: |
| 303 | fv, issue_id = self._UnpackFieldValue(fv_row) |
| 304 | results_dict[issue_id].field_values.append(fv) |
| 305 | |
| 306 | phases_by_id = {} |
| 307 | for phase_row in phase_rows: |
| 308 | phase = self._UnpackPhase(phase_row) |
| 309 | phases_by_id[phase.phase_id] = phase |
| 310 | |
| 311 | approvers_dict = collections.defaultdict(list) |
| 312 | for approver_row in av_approver_rows: |
| 313 | approval_id, approver_id, issue_id = approver_row |
| 314 | approvers_dict[approval_id, issue_id].append(approver_id) |
| 315 | |
| 316 | for av_row in approvalvalue_rows: |
| 317 | av, issue_id = self._UnpackApprovalValue(av_row) |
| 318 | av.approver_ids = approvers_dict[av.approval_id, issue_id] |
| 319 | results_dict[issue_id].approval_values.append(av) |
| 320 | if av.phase_id: |
| 321 | phase = phases_by_id[av.phase_id] |
| 322 | issue_phases = results_dict[issue_id].phases |
| 323 | if phase not in issue_phases: |
| 324 | issue_phases.append(phase) |
| 325 | # Order issue phases |
| 326 | for issue in results_dict.values(): |
| 327 | if issue.phases: |
| 328 | issue.phases.sort(key=lambda phase: phase.rank) |
| 329 | |
| 330 | for issue_id, dst_issue_id, kind, rank in relation_rows: |
| 331 | src_issue = results_dict.get(issue_id) |
| 332 | dst_issue = results_dict.get(dst_issue_id) |
| 333 | assert src_issue or dst_issue, ( |
| 334 | 'Neither source issue %r nor dest issue %r was found' % |
| 335 | (issue_id, dst_issue_id)) |
| 336 | if src_issue: |
| 337 | if kind == 'blockedon': |
| 338 | src_issue.blocked_on_iids.append(dst_issue_id) |
| 339 | src_issue.blocked_on_ranks.append(rank) |
| 340 | elif kind == 'mergedinto': |
| 341 | src_issue.merged_into = dst_issue_id |
| 342 | else: |
| 343 | logging.info('unknown relation kind %r', kind) |
| 344 | continue |
| 345 | |
| 346 | if dst_issue: |
| 347 | if kind == 'blockedon': |
| 348 | dst_issue.blocking_iids.append(issue_id) |
| 349 | |
| 350 | for row in dangling_relation_rows: |
| 351 | issue_id, dst_issue_proj, dst_issue_id, ext_id, kind = row |
| 352 | src_issue = results_dict.get(issue_id) |
| 353 | if kind == 'blockedon': |
| 354 | src_issue.dangling_blocked_on_refs.append( |
| 355 | tracker_bizobj.MakeDanglingIssueRef(dst_issue_proj, |
| 356 | dst_issue_id, ext_id)) |
| 357 | elif kind == 'blocking': |
| 358 | src_issue.dangling_blocking_refs.append( |
| 359 | tracker_bizobj.MakeDanglingIssueRef(dst_issue_proj, dst_issue_id, |
| 360 | ext_id)) |
| 361 | elif kind == 'mergedinto': |
| 362 | src_issue.merged_into_external = ext_id |
| 363 | else: |
| 364 | logging.warn('unhandled danging relation kind %r', kind) |
| 365 | continue |
| 366 | |
| 367 | return results_dict |
| 368 | |
| 369 | # Note: sharding is used to here to allow us to load issues from the replicas |
| 370 | # without placing load on the primary DB. Writes are not sharded. |
| 371 | # pylint: disable=arguments-differ |
| 372 | def FetchItems(self, cnxn, issue_ids, shard_id=None): |
| 373 | """Retrieve and deserialize issues.""" |
| 374 | issue_rows = self.issue_service.issue_tbl.Select( |
| 375 | cnxn, cols=ISSUE_COLS, id=issue_ids, shard_id=shard_id) |
| 376 | |
| 377 | summary_rows = self.issue_service.issuesummary_tbl.Select( |
| 378 | cnxn, cols=ISSUESUMMARY_COLS, shard_id=shard_id, issue_id=issue_ids) |
| 379 | label_rows = self.issue_service.issue2label_tbl.Select( |
| 380 | cnxn, cols=ISSUE2LABEL_COLS, shard_id=shard_id, issue_id=issue_ids) |
| 381 | component_rows = self.issue_service.issue2component_tbl.Select( |
| 382 | cnxn, cols=ISSUE2COMPONENT_COLS, shard_id=shard_id, issue_id=issue_ids) |
| 383 | cc_rows = self.issue_service.issue2cc_tbl.Select( |
| 384 | cnxn, cols=ISSUE2CC_COLS, shard_id=shard_id, issue_id=issue_ids) |
| 385 | notify_rows = self.issue_service.issue2notify_tbl.Select( |
| 386 | cnxn, cols=ISSUE2NOTIFY_COLS, shard_id=shard_id, issue_id=issue_ids) |
| 387 | fieldvalue_rows = self.issue_service.issue2fieldvalue_tbl.Select( |
| 388 | cnxn, cols=ISSUE2FIELDVALUE_COLS, shard_id=shard_id, |
| 389 | issue_id=issue_ids) |
| 390 | approvalvalue_rows = self.issue_service.issue2approvalvalue_tbl.Select( |
| 391 | cnxn, cols=ISSUE2APPROVALVALUE_COLS, issue_id=issue_ids) |
| 392 | phase_ids = [av_row[2] for av_row in approvalvalue_rows] |
| 393 | phase_rows = [] |
| 394 | if phase_ids: |
| 395 | phase_rows = self.issue_service.issuephasedef_tbl.Select( |
| 396 | cnxn, cols=ISSUEPHASEDEF_COLS, id=list(set(phase_ids))) |
| 397 | av_approver_rows = self.issue_service.issueapproval2approver_tbl.Select( |
| 398 | cnxn, cols=ISSUEAPPROVAL2APPROVER_COLS, issue_id=issue_ids) |
| 399 | if issue_ids: |
| 400 | ph = sql.PlaceHolders(issue_ids) |
| 401 | blocked_on_rows = self.issue_service.issuerelation_tbl.Select( |
| 402 | cnxn, cols=ISSUERELATION_COLS, issue_id=issue_ids, kind='blockedon', |
| 403 | order_by=[('issue_id', []), ('rank DESC', []), ('dst_issue_id', [])]) |
| 404 | blocking_rows = self.issue_service.issuerelation_tbl.Select( |
| 405 | cnxn, cols=ISSUERELATION_COLS, dst_issue_id=issue_ids, |
| 406 | kind='blockedon', order_by=[('issue_id', []), ('dst_issue_id', [])]) |
| 407 | unique_blocking = tuple( |
| 408 | row for row in blocking_rows if row not in blocked_on_rows) |
| 409 | merge_rows = self.issue_service.issuerelation_tbl.Select( |
| 410 | cnxn, cols=ISSUERELATION_COLS, |
| 411 | where=[('(issue_id IN (%s) OR dst_issue_id IN (%s))' % (ph, ph), |
| 412 | issue_ids + issue_ids), |
| 413 | ('kind != %s', ['blockedon'])]) |
| 414 | relation_rows = blocked_on_rows + unique_blocking + merge_rows |
| 415 | dangling_relation_rows = self.issue_service.danglingrelation_tbl.Select( |
| 416 | cnxn, cols=DANGLINGRELATION_COLS, issue_id=issue_ids) |
| 417 | else: |
| 418 | relation_rows = [] |
| 419 | dangling_relation_rows = [] |
| 420 | |
| 421 | issue_dict = self._DeserializeIssues( |
| 422 | cnxn, issue_rows, summary_rows, label_rows, component_rows, cc_rows, |
| 423 | notify_rows, fieldvalue_rows, relation_rows, dangling_relation_rows, |
| 424 | phase_rows, approvalvalue_rows, av_approver_rows) |
| 425 | logging.info('IssueTwoLevelCache.FetchItems returning: %r', issue_dict) |
| 426 | return issue_dict |
| 427 | |
| 428 | |
| 429 | class CommentTwoLevelCache(caches.AbstractTwoLevelCache): |
| 430 | """Class to manage RAM and memcache for IssueComment PBs.""" |
| 431 | |
| 432 | def __init__(self, cache_manager, issue_svc): |
| 433 | super(CommentTwoLevelCache, self).__init__( |
| 434 | cache_manager, 'comment', 'comment:', tracker_pb2.IssueComment, |
| 435 | max_size=settings.comment_cache_max_size) |
| 436 | self.issue_svc = issue_svc |
| 437 | |
| 438 | # pylint: disable=arguments-differ |
| 439 | def FetchItems(self, cnxn, keys, shard_id=None): |
| 440 | comment_rows = self.issue_svc.comment_tbl.Select(cnxn, |
| 441 | cols=COMMENT_COLS, id=keys, shard_id=shard_id) |
| 442 | |
| 443 | if len(comment_rows) < len(keys): |
| 444 | self.issue_svc.replication_lag_retries.increment() |
| 445 | logging.info('issue3755: expected %d, but got %d rows from shard %d', |
| 446 | len(keys), len(comment_rows), shard_id) |
| 447 | shard_id = None # Will use Primary DB. |
| 448 | comment_rows = self.issue_svc.comment_tbl.Select( |
| 449 | cnxn, cols=COMMENT_COLS, id=keys, shard_id=None) |
| 450 | logging.info( |
| 451 | 'Retry got %d comment rows from the primary DB', len(comment_rows)) |
| 452 | |
| 453 | cids = [row[0] for row in comment_rows] |
| 454 | commentcontent_ids = [row[-1] for row in comment_rows] |
| 455 | content_rows = self.issue_svc.commentcontent_tbl.Select( |
| 456 | cnxn, cols=COMMENTCONTENT_COLS, id=commentcontent_ids, |
| 457 | shard_id=shard_id) |
| 458 | approval_rows = self.issue_svc.issueapproval2comment_tbl.Select( |
| 459 | cnxn, cols=ISSUEAPPROVAL2COMMENT_COLS, comment_id=cids) |
| 460 | amendment_rows = self.issue_svc.issueupdate_tbl.Select( |
| 461 | cnxn, cols=ISSUEUPDATE_COLS, comment_id=cids, shard_id=shard_id) |
| 462 | attachment_rows = self.issue_svc.attachment_tbl.Select( |
| 463 | cnxn, cols=ATTACHMENT_COLS, comment_id=cids, shard_id=shard_id) |
| 464 | importer_rows = self.issue_svc.commentimporter_tbl.Select( |
| 465 | cnxn, cols=COMMENTIMPORTER_COLS, comment_id=cids, shard_id=shard_id) |
| 466 | |
| 467 | comments = self.issue_svc._DeserializeComments( |
| 468 | comment_rows, content_rows, amendment_rows, attachment_rows, |
| 469 | approval_rows, importer_rows) |
| 470 | |
| 471 | comments_dict = {} |
| 472 | for comment in comments: |
| 473 | comments_dict[comment.id] = comment |
| 474 | |
| 475 | return comments_dict |
| 476 | |
| 477 | |
| 478 | class IssueService(object): |
| 479 | """The persistence layer for Monorail's issues, comments, and attachments.""" |
| 480 | spam_labels = ts_mon.CounterMetric( |
| 481 | 'monorail/issue_svc/spam_label', |
| 482 | 'Issues created, broken down by spam label.', |
| 483 | [ts_mon.StringField('type')]) |
| 484 | replication_lag_retries = ts_mon.CounterMetric( |
| 485 | 'monorail/issue_svc/replication_lag_retries', |
| 486 | 'Counts times that loading comments from a replica failed', |
| 487 | []) |
| 488 | issue_creations = ts_mon.CounterMetric( |
| 489 | 'monorail/issue_svc/issue_creations', |
| 490 | 'Counts times that issues were created', |
| 491 | []) |
| 492 | comment_creations = ts_mon.CounterMetric( |
| 493 | 'monorail/issue_svc/comment_creations', |
| 494 | 'Counts times that comments were created', |
| 495 | []) |
| 496 | |
| 497 | def __init__(self, project_service, config_service, cache_manager, |
| 498 | chart_service): |
| 499 | """Initialize this object so that it is ready to use. |
| 500 | |
| 501 | Args: |
| 502 | project_service: services object for project info. |
| 503 | config_service: services object for tracker configuration info. |
| 504 | cache_manager: local cache with distributed invalidation. |
| 505 | chart_service (ChartService): An instance of ChartService. |
| 506 | """ |
| 507 | # Tables that represent issue data. |
| 508 | self.issue_tbl = sql.SQLTableManager(ISSUE_TABLE_NAME) |
| 509 | self.issuesummary_tbl = sql.SQLTableManager(ISSUESUMMARY_TABLE_NAME) |
| 510 | self.issue2label_tbl = sql.SQLTableManager(ISSUE2LABEL_TABLE_NAME) |
| 511 | self.issue2component_tbl = sql.SQLTableManager(ISSUE2COMPONENT_TABLE_NAME) |
| 512 | self.issue2cc_tbl = sql.SQLTableManager(ISSUE2CC_TABLE_NAME) |
| 513 | self.issue2notify_tbl = sql.SQLTableManager(ISSUE2NOTIFY_TABLE_NAME) |
| 514 | self.issue2fieldvalue_tbl = sql.SQLTableManager(ISSUE2FIELDVALUE_TABLE_NAME) |
| 515 | self.issuerelation_tbl = sql.SQLTableManager(ISSUERELATION_TABLE_NAME) |
| 516 | self.danglingrelation_tbl = sql.SQLTableManager(DANGLINGRELATION_TABLE_NAME) |
| 517 | self.issueformerlocations_tbl = sql.SQLTableManager( |
| 518 | ISSUEFORMERLOCATIONS_TABLE_NAME) |
| 519 | self.issuesnapshot_tbl = sql.SQLTableManager(ISSUESNAPSHOT_TABLE_NAME) |
| 520 | self.issuesnapshot2cc_tbl = sql.SQLTableManager( |
| 521 | ISSUESNAPSHOT2CC_TABLE_NAME) |
| 522 | self.issuesnapshot2component_tbl = sql.SQLTableManager( |
| 523 | ISSUESNAPSHOT2COMPONENT_TABLE_NAME) |
| 524 | self.issuesnapshot2label_tbl = sql.SQLTableManager( |
| 525 | ISSUESNAPSHOT2LABEL_TABLE_NAME) |
| 526 | self.issuephasedef_tbl = sql.SQLTableManager(ISSUEPHASEDEF_TABLE_NAME) |
| 527 | self.issue2approvalvalue_tbl = sql.SQLTableManager( |
| 528 | ISSUE2APPROVALVALUE_TABLE_NAME) |
| 529 | self.issueapproval2approver_tbl = sql.SQLTableManager( |
| 530 | ISSUEAPPROVAL2APPROVER_TABLE_NAME) |
| 531 | self.issueapproval2comment_tbl = sql.SQLTableManager( |
| 532 | ISSUEAPPROVAL2COMMENT_TABLE_NAME) |
| 533 | |
| 534 | # Tables that represent comments. |
| 535 | self.comment_tbl = sql.SQLTableManager(COMMENT_TABLE_NAME) |
| 536 | self.commentcontent_tbl = sql.SQLTableManager(COMMENTCONTENT_TABLE_NAME) |
| 537 | self.commentimporter_tbl = sql.SQLTableManager(COMMENTIMPORTER_TABLE_NAME) |
| 538 | self.issueupdate_tbl = sql.SQLTableManager(ISSUEUPDATE_TABLE_NAME) |
| 539 | self.attachment_tbl = sql.SQLTableManager(ATTACHMENT_TABLE_NAME) |
| 540 | |
| 541 | # Tables for cron tasks. |
| 542 | self.reindexqueue_tbl = sql.SQLTableManager(REINDEXQUEUE_TABLE_NAME) |
| 543 | |
| 544 | # Tables for generating sequences of local IDs. |
| 545 | self.localidcounter_tbl = sql.SQLTableManager(LOCALIDCOUNTER_TABLE_NAME) |
| 546 | |
| 547 | # Like a dictionary {(project_id, local_id): issue_id} |
| 548 | # Use value centric cache here because we cannot store a tuple in the |
| 549 | # Invalidate table. |
| 550 | self.issue_id_2lc = IssueIDTwoLevelCache(cache_manager, self) |
| 551 | # Like a dictionary {issue_id: issue} |
| 552 | self.issue_2lc = IssueTwoLevelCache( |
| 553 | cache_manager, self, project_service, config_service) |
| 554 | |
| 555 | # Like a dictionary {comment_id: comment) |
| 556 | self.comment_2lc = CommentTwoLevelCache( |
| 557 | cache_manager, self) |
| 558 | |
| 559 | self._config_service = config_service |
| 560 | self.chart_service = chart_service |
| 561 | |
| 562 | ### Issue ID lookups |
| 563 | |
| 564 | def LookupIssueIDsFollowMoves(self, cnxn, project_local_id_pairs): |
| 565 | # type: (MonorailConnection, Sequence[Tuple(int, int)]) -> |
| 566 | # (Sequence[int], Sequence[Tuple(int, int)]) |
| 567 | """Find the global issue IDs given the project ID and local ID of each. |
| 568 | |
| 569 | If any (project_id, local_id) pairs refer to an issue that has been moved, |
| 570 | the issue ID will still be returned. |
| 571 | |
| 572 | Args: |
| 573 | cnxn: Monorail connection. |
| 574 | project_local_id_pairs: (project_id, local_id) pairs to look up. |
| 575 | |
| 576 | Returns: |
| 577 | A tuple of two items. |
| 578 | 1. A sequence of global issue IDs in the `project_local_id_pairs` order. |
| 579 | 2. A sequence of (project_id, local_id) containing each pair provided |
| 580 | for which no matching issue is found. |
| 581 | """ |
| 582 | |
| 583 | issue_id_dict, misses = self.issue_id_2lc.GetAll( |
| 584 | cnxn, project_local_id_pairs) |
| 585 | for miss in misses: |
| 586 | project_id, local_id = miss |
| 587 | issue_id = int( |
| 588 | self.issueformerlocations_tbl.SelectValue( |
| 589 | cnxn, |
| 590 | 'issue_id', |
| 591 | default=0, |
| 592 | project_id=project_id, |
| 593 | local_id=local_id)) |
| 594 | if issue_id: |
| 595 | misses.remove(miss) |
| 596 | issue_id_dict[miss] = issue_id |
| 597 | # Put the Issue IDs in the order specified by project_local_id_pairs |
| 598 | issue_ids = [ |
| 599 | issue_id_dict[pair] |
| 600 | for pair in project_local_id_pairs |
| 601 | if pair in issue_id_dict |
| 602 | ] |
| 603 | |
| 604 | return issue_ids, misses |
| 605 | |
| 606 | def LookupIssueIDs(self, cnxn, project_local_id_pairs): |
| 607 | """Find the global issue IDs given the project ID and local ID of each.""" |
| 608 | issue_id_dict, misses = self.issue_id_2lc.GetAll( |
| 609 | cnxn, project_local_id_pairs) |
| 610 | |
| 611 | # Put the Issue IDs in the order specified by project_local_id_pairs |
| 612 | issue_ids = [issue_id_dict[pair] for pair in project_local_id_pairs |
| 613 | if pair in issue_id_dict] |
| 614 | |
| 615 | return issue_ids, misses |
| 616 | |
| 617 | def LookupIssueID(self, cnxn, project_id, local_id): |
| 618 | """Find the global issue ID given the project ID and local ID.""" |
| 619 | issue_ids, _misses = self.LookupIssueIDs(cnxn, [(project_id, local_id)]) |
| 620 | try: |
| 621 | return issue_ids[0] |
| 622 | except IndexError: |
| 623 | raise exceptions.NoSuchIssueException() |
| 624 | |
| 625 | def ResolveIssueRefs( |
| 626 | self, cnxn, ref_projects, default_project_name, refs): |
| 627 | """Look up all the referenced issues and return their issue_ids. |
| 628 | |
| 629 | Args: |
| 630 | cnxn: connection to SQL database. |
| 631 | ref_projects: pre-fetched dict {project_name: project} of all projects |
| 632 | mentioned in the refs as well as the default project. |
| 633 | default_project_name: string name of the current project, this is used |
| 634 | when the project_name in a ref is None. |
| 635 | refs: list of (project_name, local_id) pairs. These are parsed from |
| 636 | textual references in issue descriptions, comments, and the input |
| 637 | in the blocked-on field. |
| 638 | |
| 639 | Returns: |
| 640 | A list of issue_ids for all the referenced issues. References to issues |
| 641 | in deleted projects and any issues not found are simply ignored. |
| 642 | """ |
| 643 | if not refs: |
| 644 | return [], [] |
| 645 | |
| 646 | project_local_id_pairs = [] |
| 647 | for project_name, local_id in refs: |
| 648 | project = ref_projects.get(project_name or default_project_name) |
| 649 | if not project or project.state == project_pb2.ProjectState.DELETABLE: |
| 650 | continue # ignore any refs to issues in deleted projects |
| 651 | project_local_id_pairs.append((project.project_id, local_id)) |
| 652 | |
| 653 | return self.LookupIssueIDs(cnxn, project_local_id_pairs) # tuple |
| 654 | |
| 655 | def LookupIssueRefs(self, cnxn, issue_ids): |
| 656 | """Return {issue_id: (project_name, local_id)} for each issue_id.""" |
| 657 | issue_dict, _misses = self.GetIssuesDict(cnxn, issue_ids) |
| 658 | return { |
| 659 | issue_id: (issue.project_name, issue.local_id) |
| 660 | for issue_id, issue in issue_dict.items()} |
| 661 | |
| 662 | ### Issue objects |
| 663 | |
| 664 | def CreateIssue( |
| 665 | self, |
| 666 | cnxn, |
| 667 | services, |
| 668 | issue, |
| 669 | marked_description, |
| 670 | attachments=None, |
| 671 | index_now=False, |
| 672 | importer_id=None): |
| 673 | """Create and store a new issue with all the given information. |
| 674 | |
| 675 | Args: |
| 676 | cnxn: connection to SQL database. |
| 677 | services: persistence layer for users, issues, and projects. |
| 678 | issue: Issue PB to create. |
| 679 | marked_description: issue description with initial HTML markup. |
| 680 | attachments: [(filename, contents, mimetype),...] attachments uploaded at |
| 681 | the time the comment was made. |
| 682 | index_now: True if the issue should be updated in the full text index. |
| 683 | importer_id: optional user ID of API client importing issues for users. |
| 684 | |
| 685 | Returns: |
| 686 | A tuple (the newly created Issue PB and Comment PB for the |
| 687 | issue description). |
| 688 | """ |
| 689 | project_id = issue.project_id |
| 690 | reporter_id = issue.reporter_id |
| 691 | timestamp = issue.opened_timestamp |
| 692 | config = self._config_service.GetProjectConfig(cnxn, project_id) |
| 693 | |
| 694 | iids_to_invalidate = set() |
| 695 | if len(issue.blocked_on_iids) != 0: |
| 696 | iids_to_invalidate.update(issue.blocked_on_iids) |
| 697 | if len(issue.blocking_iids) != 0: |
| 698 | iids_to_invalidate.update(issue.blocking_iids) |
| 699 | |
| 700 | comment = self._MakeIssueComment( |
| 701 | project_id, reporter_id, marked_description, |
| 702 | attachments=attachments, timestamp=timestamp, |
| 703 | is_description=True, importer_id=importer_id) |
| 704 | |
| 705 | reporter = services.user.GetUser(cnxn, reporter_id) |
| 706 | project = services.project.GetProject(cnxn, project_id) |
| 707 | reporter_auth = authdata.AuthData.FromUserID(cnxn, reporter_id, services) |
| 708 | is_project_member = framework_bizobj.UserIsInProject( |
| 709 | project, reporter_auth.effective_ids) |
| 710 | classification = services.spam.ClassifyIssue( |
| 711 | issue, comment, reporter, is_project_member) |
| 712 | |
| 713 | if classification['confidence_is_spam'] > settings.classifier_spam_thresh: |
| 714 | issue.is_spam = True |
| 715 | predicted_label = 'spam' |
| 716 | else: |
| 717 | predicted_label = 'ham' |
| 718 | |
| 719 | logging.info('classified new issue as %s' % predicted_label) |
| 720 | self.spam_labels.increment({'type': predicted_label}) |
| 721 | |
| 722 | # Create approval surveys |
| 723 | approval_comments = [] |
| 724 | if len(issue.approval_values) != 0: |
| 725 | approval_defs_by_id = {ad.approval_id: ad for ad in config.approval_defs} |
| 726 | for av in issue.approval_values: |
| 727 | ad = approval_defs_by_id.get(av.approval_id) |
| 728 | if ad: |
| 729 | survey = '' |
| 730 | if ad.survey: |
| 731 | questions = ad.survey.split('\n') |
| 732 | survey = '\n'.join(['<b>' + q + '</b>' for q in questions]) |
| 733 | approval_comments.append(self._MakeIssueComment( |
| 734 | project_id, reporter_id, survey, timestamp=timestamp, |
| 735 | is_description=True, approval_id=ad.approval_id)) |
| 736 | else: |
| 737 | logging.info('Could not find ApprovalDef with approval_id %r', |
| 738 | av.approval_id) |
| 739 | |
| 740 | issue.local_id = self.AllocateNextLocalID(cnxn, project_id) |
| 741 | self.issue_creations.increment() |
| 742 | issue_id = self.InsertIssue(cnxn, issue) |
| 743 | comment.issue_id = issue_id |
| 744 | self.InsertComment(cnxn, comment) |
| 745 | for approval_comment in approval_comments: |
| 746 | approval_comment.issue_id = issue_id |
| 747 | self.InsertComment(cnxn, approval_comment) |
| 748 | |
| 749 | issue.issue_id = issue_id |
| 750 | |
| 751 | # ClassifyIssue only returns confidence_is_spam, but |
| 752 | # RecordClassifierIssueVerdict records confidence of |
| 753 | # ham or spam. Therefore if ham, invert score. |
| 754 | confidence = classification['confidence_is_spam'] |
| 755 | if not issue.is_spam: |
| 756 | confidence = 1.0 - confidence |
| 757 | |
| 758 | services.spam.RecordClassifierIssueVerdict( |
| 759 | cnxn, issue, predicted_label=='spam', |
| 760 | confidence, classification['failed_open']) |
| 761 | |
| 762 | if permissions.HasRestrictions(issue, 'view'): |
| 763 | self._config_service.InvalidateMemcache( |
| 764 | [issue], key_prefix='nonviewable:') |
| 765 | |
| 766 | # Add a comment to existing issues saying they are now blocking or |
| 767 | # blocked on this issue. |
| 768 | blocked_add_issues = self.GetIssues(cnxn, issue.blocked_on_iids) |
| 769 | for add_issue in blocked_add_issues: |
| 770 | self.CreateIssueComment( |
| 771 | cnxn, add_issue, reporter_id, content='', |
| 772 | amendments=[tracker_bizobj.MakeBlockingAmendment( |
| 773 | [(issue.project_name, issue.local_id)], [], |
| 774 | default_project_name=add_issue.project_name)]) |
| 775 | blocking_add_issues = self.GetIssues(cnxn, issue.blocking_iids) |
| 776 | for add_issue in blocking_add_issues: |
| 777 | self.CreateIssueComment( |
| 778 | cnxn, add_issue, reporter_id, content='', |
| 779 | amendments=[tracker_bizobj.MakeBlockedOnAmendment( |
| 780 | [(issue.project_name, issue.local_id)], [], |
| 781 | default_project_name=add_issue.project_name)]) |
| 782 | |
| 783 | self._UpdateIssuesModified( |
| 784 | cnxn, iids_to_invalidate, modified_timestamp=timestamp) |
| 785 | |
| 786 | if index_now: |
| 787 | tracker_fulltext.IndexIssues( |
| 788 | cnxn, [issue], services.user, self, self._config_service) |
| 789 | else: |
| 790 | self.EnqueueIssuesForIndexing(cnxn, [issue.issue_id]) |
| 791 | |
| 792 | return issue, comment |
| 793 | |
| 794 | def AllocateNewLocalIDs(self, cnxn, issues): |
| 795 | # Filter to just the issues that need new local IDs. |
| 796 | issues = [issue for issue in issues if issue.local_id < 0] |
| 797 | |
| 798 | for issue in issues: |
| 799 | if issue.local_id < 0: |
| 800 | issue.local_id = self.AllocateNextLocalID(cnxn, issue.project_id) |
| 801 | |
| 802 | self.UpdateIssues(cnxn, issues) |
| 803 | |
| 804 | logging.info("AllocateNewLocalIDs") |
| 805 | |
| 806 | def GetAllIssuesInProject( |
| 807 | self, cnxn, project_id, min_local_id=None, use_cache=True): |
| 808 | """Special query to efficiently get ALL issues in a project. |
| 809 | |
| 810 | This is not done while the user is waiting, only by backround tasks. |
| 811 | |
| 812 | Args: |
| 813 | cnxn: connection to SQL database. |
| 814 | project_id: the ID of the project. |
| 815 | min_local_id: optional int to start at. |
| 816 | use_cache: optional boolean to turn off using the cache. |
| 817 | |
| 818 | Returns: |
| 819 | A list of Issue protocol buffers for all issues. |
| 820 | """ |
| 821 | all_local_ids = self.GetAllLocalIDsInProject( |
| 822 | cnxn, project_id, min_local_id=min_local_id) |
| 823 | return self.GetIssuesByLocalIDs( |
| 824 | cnxn, project_id, all_local_ids, use_cache=use_cache) |
| 825 | |
| 826 | def GetAnyOnHandIssue(self, issue_ids, start=None, end=None): |
| 827 | """Get any one issue from RAM or memcache, otherwise return None.""" |
| 828 | return self.issue_2lc.GetAnyOnHandItem(issue_ids, start=start, end=end) |
| 829 | |
| 830 | def GetIssuesDict(self, cnxn, issue_ids, use_cache=True, shard_id=None): |
| 831 | # type: (MonorailConnection, Collection[int], Optional[Boolean], |
| 832 | # Optional[int]) -> (Dict[int, Issue], Sequence[int]) |
| 833 | """Get a dict {iid: issue} from the DB or cache. |
| 834 | |
| 835 | Returns: |
| 836 | A dict {iid: issue} from the DB or cache. |
| 837 | A sequence of iid that could not be found. |
| 838 | """ |
| 839 | issue_dict, missed_iids = self.issue_2lc.GetAll( |
| 840 | cnxn, issue_ids, use_cache=use_cache, shard_id=shard_id) |
| 841 | if not use_cache: |
| 842 | for issue in issue_dict.values(): |
| 843 | issue.assume_stale = False |
| 844 | return issue_dict, missed_iids |
| 845 | |
| 846 | def GetIssues(self, cnxn, issue_ids, use_cache=True, shard_id=None): |
| 847 | # type: (MonorailConnection, Sequence[int], Optional[Boolean], |
| 848 | # Optional[int]) -> (Sequence[int]) |
| 849 | """Get a list of Issue PBs from the DB or cache. |
| 850 | |
| 851 | Args: |
| 852 | cnxn: connection to SQL database. |
| 853 | issue_ids: integer global issue IDs of the issues. |
| 854 | use_cache: optional boolean to turn off using the cache. |
| 855 | shard_id: optional int shard_id to limit retrieval. |
| 856 | |
| 857 | Returns: |
| 858 | A list of Issue PBs in the same order as the given issue_ids. |
| 859 | """ |
| 860 | issue_dict, _misses = self.GetIssuesDict( |
| 861 | cnxn, issue_ids, use_cache=use_cache, shard_id=shard_id) |
| 862 | |
| 863 | # Return a list that is ordered the same as the given issue_ids. |
| 864 | issue_list = [issue_dict[issue_id] for issue_id in issue_ids |
| 865 | if issue_id in issue_dict] |
| 866 | |
| 867 | return issue_list |
| 868 | |
| 869 | def GetIssue(self, cnxn, issue_id, use_cache=True): |
| 870 | """Get one Issue PB from the DB. |
| 871 | |
| 872 | Args: |
| 873 | cnxn: connection to SQL database. |
| 874 | issue_id: integer global issue ID of the issue. |
| 875 | use_cache: optional boolean to turn off using the cache. |
| 876 | |
| 877 | Returns: |
| 878 | The requested Issue protocol buffer. |
| 879 | |
| 880 | Raises: |
| 881 | NoSuchIssueException: the issue was not found. |
| 882 | """ |
| 883 | issues = self.GetIssues(cnxn, [issue_id], use_cache=use_cache) |
| 884 | try: |
| 885 | return issues[0] |
| 886 | except IndexError: |
| 887 | raise exceptions.NoSuchIssueException() |
| 888 | |
| 889 | def GetIssuesByLocalIDs( |
| 890 | self, cnxn, project_id, local_id_list, use_cache=True, shard_id=None): |
| 891 | """Get all the requested issues. |
| 892 | |
| 893 | Args: |
| 894 | cnxn: connection to SQL database. |
| 895 | project_id: int ID of the project to which the issues belong. |
| 896 | local_id_list: list of integer local IDs for the requested issues. |
| 897 | use_cache: optional boolean to turn off using the cache. |
| 898 | shard_id: optional int shard_id to choose a replica. |
| 899 | |
| 900 | Returns: |
| 901 | List of Issue PBs for the requested issues. The result Issues |
| 902 | will be ordered in the same order as local_id_list. |
| 903 | """ |
| 904 | issue_ids_to_fetch, _misses = self.LookupIssueIDs( |
| 905 | cnxn, [(project_id, local_id) for local_id in local_id_list]) |
| 906 | issues = self.GetIssues( |
| 907 | cnxn, issue_ids_to_fetch, use_cache=use_cache, shard_id=shard_id) |
| 908 | return issues |
| 909 | |
| 910 | def GetIssueByLocalID(self, cnxn, project_id, local_id, use_cache=True): |
| 911 | """Get one Issue PB from the DB. |
| 912 | |
| 913 | Args: |
| 914 | cnxn: connection to SQL database. |
| 915 | project_id: the ID of the project to which the issue belongs. |
| 916 | local_id: integer local ID of the issue. |
| 917 | use_cache: optional boolean to turn off using the cache. |
| 918 | |
| 919 | Returns: |
| 920 | The requested Issue protocol buffer. |
| 921 | """ |
| 922 | issues = self.GetIssuesByLocalIDs( |
| 923 | cnxn, project_id, [local_id], use_cache=use_cache) |
| 924 | try: |
| 925 | return issues[0] |
| 926 | except IndexError: |
| 927 | raise exceptions.NoSuchIssueException( |
| 928 | 'The issue %s:%d does not exist.' % (project_id, local_id)) |
| 929 | |
| 930 | def GetOpenAndClosedIssues(self, cnxn, issue_ids): |
| 931 | """Return the requested issues in separate open and closed lists. |
| 932 | |
| 933 | Args: |
| 934 | cnxn: connection to SQL database. |
| 935 | issue_ids: list of int issue issue_ids. |
| 936 | |
| 937 | Returns: |
| 938 | A pair of lists, the first with open issues, second with closed issues. |
| 939 | """ |
| 940 | if not issue_ids: |
| 941 | return [], [] # make one common case efficient |
| 942 | |
| 943 | issues = self.GetIssues(cnxn, issue_ids) |
| 944 | project_ids = {issue.project_id for issue in issues} |
| 945 | configs = self._config_service.GetProjectConfigs(cnxn, project_ids) |
| 946 | open_issues = [] |
| 947 | closed_issues = [] |
| 948 | for issue in issues: |
| 949 | config = configs[issue.project_id] |
| 950 | if tracker_helpers.MeansOpenInProject( |
| 951 | tracker_bizobj.GetStatus(issue), config): |
| 952 | open_issues.append(issue) |
| 953 | else: |
| 954 | closed_issues.append(issue) |
| 955 | |
| 956 | return open_issues, closed_issues |
| 957 | |
| 958 | # TODO(crbug.com/monorail/7822): Delete this method when V0 API retired. |
| 959 | def GetCurrentLocationOfMovedIssue(self, cnxn, project_id, local_id): |
| 960 | """Return the current location of a moved issue based on old location.""" |
| 961 | issue_id = int(self.issueformerlocations_tbl.SelectValue( |
| 962 | cnxn, 'issue_id', default=0, project_id=project_id, local_id=local_id)) |
| 963 | if not issue_id: |
| 964 | return None, None |
| 965 | project_id, local_id = self.issue_tbl.SelectRow( |
| 966 | cnxn, cols=['project_id', 'local_id'], id=issue_id) |
| 967 | return project_id, local_id |
| 968 | |
| 969 | def GetPreviousLocations(self, cnxn, issue): |
| 970 | """Get all the previous locations of an issue.""" |
| 971 | location_rows = self.issueformerlocations_tbl.Select( |
| 972 | cnxn, cols=['project_id', 'local_id'], issue_id=issue.issue_id) |
| 973 | locations = [(pid, local_id) for (pid, local_id) in location_rows |
| 974 | if pid != issue.project_id or local_id != issue.local_id] |
| 975 | return locations |
| 976 | |
| 977 | def GetCommentsByUser(self, cnxn, user_id): |
| 978 | """Get all comments created by a user""" |
| 979 | comments = self.GetComments(cnxn, commenter_id=user_id, |
| 980 | is_description=False, limit=10000) |
| 981 | return comments |
| 982 | |
| 983 | def GetIssueActivity(self, cnxn, num=50, before=None, after=None, |
| 984 | project_ids=None, user_ids=None, ascending=False): |
| 985 | |
| 986 | if project_ids: |
| 987 | use_clause = ( |
| 988 | 'USE INDEX (project_id) USE INDEX FOR ORDER BY (project_id)') |
| 989 | elif user_ids: |
| 990 | use_clause = ( |
| 991 | 'USE INDEX (commenter_id) USE INDEX FOR ORDER BY (commenter_id)') |
| 992 | else: |
| 993 | use_clause = '' |
| 994 | |
| 995 | # TODO(jrobbins): make this into a persist method. |
| 996 | # TODO(jrobbins): this really needs permission checking in SQL, which |
| 997 | # will be slow. |
| 998 | where_conds = [('Issue.id = Comment.issue_id', [])] |
| 999 | if project_ids is not None: |
| 1000 | cond_str = 'Comment.project_id IN (%s)' % sql.PlaceHolders(project_ids) |
| 1001 | where_conds.append((cond_str, project_ids)) |
| 1002 | if user_ids is not None: |
| 1003 | cond_str = 'Comment.commenter_id IN (%s)' % sql.PlaceHolders(user_ids) |
| 1004 | where_conds.append((cond_str, user_ids)) |
| 1005 | |
| 1006 | if before: |
| 1007 | where_conds.append(('created < %s', [before])) |
| 1008 | if after: |
| 1009 | where_conds.append(('created > %s', [after])) |
| 1010 | if ascending: |
| 1011 | order_by = [('created', [])] |
| 1012 | else: |
| 1013 | order_by = [('created DESC', [])] |
| 1014 | |
| 1015 | comments = self.GetComments( |
| 1016 | cnxn, joins=[('Issue', [])], deleted_by=None, where=where_conds, |
| 1017 | use_clause=use_clause, order_by=order_by, limit=num + 1) |
| 1018 | return comments |
| 1019 | |
| 1020 | def GetIssueIDsReportedByUser(self, cnxn, user_id): |
| 1021 | """Get all issue IDs created by a user""" |
| 1022 | rows = self.issue_tbl.Select(cnxn, cols=['id'], reporter_id=user_id, |
| 1023 | limit=10000) |
| 1024 | return [row[0] for row in rows] |
| 1025 | |
| 1026 | def InsertIssue(self, cnxn, issue): |
| 1027 | """Store the given issue in SQL. |
| 1028 | |
| 1029 | Args: |
| 1030 | cnxn: connection to SQL database. |
| 1031 | issue: Issue PB to insert into the database. |
| 1032 | |
| 1033 | Returns: |
| 1034 | The int issue_id of the newly created issue. |
| 1035 | """ |
| 1036 | status_id = self._config_service.LookupStatusID( |
| 1037 | cnxn, issue.project_id, issue.status) |
| 1038 | row = (issue.project_id, issue.local_id, status_id, |
| 1039 | issue.owner_id or None, |
| 1040 | issue.reporter_id, |
| 1041 | issue.opened_timestamp, |
| 1042 | issue.closed_timestamp, |
| 1043 | issue.modified_timestamp, |
| 1044 | issue.owner_modified_timestamp, |
| 1045 | issue.status_modified_timestamp, |
| 1046 | issue.component_modified_timestamp, |
| 1047 | issue.derived_owner_id or None, |
| 1048 | self._config_service.LookupStatusID( |
| 1049 | cnxn, issue.project_id, issue.derived_status), |
| 1050 | bool(issue.deleted), |
| 1051 | issue.star_count, issue.attachment_count, |
| 1052 | issue.is_spam) |
| 1053 | # ISSUE_COLs[1:] to skip setting the ID |
| 1054 | # Insert into the Primary DB. |
| 1055 | generated_ids = self.issue_tbl.InsertRows( |
| 1056 | cnxn, ISSUE_COLS[1:], [row], commit=False, return_generated_ids=True) |
| 1057 | issue_id = generated_ids[0] |
| 1058 | issue.issue_id = issue_id |
| 1059 | self.issue_tbl.Update( |
| 1060 | cnxn, {'shard': issue_id % settings.num_logical_shards}, |
| 1061 | id=issue.issue_id, commit=False) |
| 1062 | |
| 1063 | self._UpdateIssuesSummary(cnxn, [issue], commit=False) |
| 1064 | self._UpdateIssuesLabels(cnxn, [issue], commit=False) |
| 1065 | self._UpdateIssuesFields(cnxn, [issue], commit=False) |
| 1066 | self._UpdateIssuesComponents(cnxn, [issue], commit=False) |
| 1067 | self._UpdateIssuesCc(cnxn, [issue], commit=False) |
| 1068 | self._UpdateIssuesNotify(cnxn, [issue], commit=False) |
| 1069 | self._UpdateIssuesRelation(cnxn, [issue], commit=False) |
| 1070 | self._UpdateIssuesApprovals(cnxn, issue, commit=False) |
| 1071 | self.chart_service.StoreIssueSnapshots(cnxn, [issue], commit=False) |
| 1072 | cnxn.Commit() |
| 1073 | self._config_service.InvalidateMemcache([issue]) |
| 1074 | |
| 1075 | return issue_id |
| 1076 | |
| 1077 | def UpdateIssues( |
| 1078 | self, cnxn, issues, update_cols=None, just_derived=False, commit=True, |
| 1079 | invalidate=True): |
| 1080 | """Update the given issues in SQL. |
| 1081 | |
| 1082 | Args: |
| 1083 | cnxn: connection to SQL database. |
| 1084 | issues: list of issues to update, these must have been loaded with |
| 1085 | use_cache=False so that issue.assume_stale is False. |
| 1086 | update_cols: optional list of just the field names to update. |
| 1087 | just_derived: set to True when only updating derived fields. |
| 1088 | commit: set to False to skip the DB commit and do it in the caller. |
| 1089 | invalidate: set to False to leave cache invalidatation to the caller. |
| 1090 | """ |
| 1091 | if not issues: |
| 1092 | return |
| 1093 | |
| 1094 | for issue in issues: # slow, but mysql will not allow REPLACE rows. |
| 1095 | assert not issue.assume_stale, ( |
| 1096 | 'issue2514: Storing issue that might be stale: %r' % issue) |
| 1097 | delta = { |
| 1098 | 'project_id': issue.project_id, |
| 1099 | 'local_id': issue.local_id, |
| 1100 | 'owner_id': issue.owner_id or None, |
| 1101 | 'status_id': self._config_service.LookupStatusID( |
| 1102 | cnxn, issue.project_id, issue.status) or None, |
| 1103 | 'opened': issue.opened_timestamp, |
| 1104 | 'closed': issue.closed_timestamp, |
| 1105 | 'modified': issue.modified_timestamp, |
| 1106 | 'owner_modified': issue.owner_modified_timestamp, |
| 1107 | 'status_modified': issue.status_modified_timestamp, |
| 1108 | 'component_modified': issue.component_modified_timestamp, |
| 1109 | 'derived_owner_id': issue.derived_owner_id or None, |
| 1110 | 'derived_status_id': self._config_service.LookupStatusID( |
| 1111 | cnxn, issue.project_id, issue.derived_status) or None, |
| 1112 | 'deleted': bool(issue.deleted), |
| 1113 | 'star_count': issue.star_count, |
| 1114 | 'attachment_count': issue.attachment_count, |
| 1115 | 'is_spam': issue.is_spam, |
| 1116 | } |
| 1117 | if update_cols is not None: |
| 1118 | delta = {key: val for key, val in delta.items() |
| 1119 | if key in update_cols} |
| 1120 | self.issue_tbl.Update(cnxn, delta, id=issue.issue_id, commit=False) |
| 1121 | |
| 1122 | if not update_cols: |
| 1123 | self._UpdateIssuesLabels(cnxn, issues, commit=False) |
| 1124 | self._UpdateIssuesCc(cnxn, issues, commit=False) |
| 1125 | self._UpdateIssuesFields(cnxn, issues, commit=False) |
| 1126 | self._UpdateIssuesComponents(cnxn, issues, commit=False) |
| 1127 | self._UpdateIssuesNotify(cnxn, issues, commit=False) |
| 1128 | if not just_derived: |
| 1129 | self._UpdateIssuesSummary(cnxn, issues, commit=False) |
| 1130 | self._UpdateIssuesRelation(cnxn, issues, commit=False) |
| 1131 | |
| 1132 | self.chart_service.StoreIssueSnapshots(cnxn, issues, commit=False) |
| 1133 | |
| 1134 | iids_to_invalidate = [issue.issue_id for issue in issues] |
| 1135 | if just_derived and invalidate: |
| 1136 | self.issue_2lc.InvalidateAllKeys(cnxn, iids_to_invalidate) |
| 1137 | elif invalidate: |
| 1138 | self.issue_2lc.InvalidateKeys(cnxn, iids_to_invalidate) |
| 1139 | if commit: |
| 1140 | cnxn.Commit() |
| 1141 | if invalidate: |
| 1142 | self._config_service.InvalidateMemcache(issues) |
| 1143 | |
| 1144 | def UpdateIssue( |
| 1145 | self, cnxn, issue, update_cols=None, just_derived=False, commit=True, |
| 1146 | invalidate=True): |
| 1147 | """Update the given issue in SQL. |
| 1148 | |
| 1149 | Args: |
| 1150 | cnxn: connection to SQL database. |
| 1151 | issue: the issue to update. |
| 1152 | update_cols: optional list of just the field names to update. |
| 1153 | just_derived: set to True when only updating derived fields. |
| 1154 | commit: set to False to skip the DB commit and do it in the caller. |
| 1155 | invalidate: set to False to leave cache invalidatation to the caller. |
| 1156 | """ |
| 1157 | self.UpdateIssues( |
| 1158 | cnxn, [issue], update_cols=update_cols, just_derived=just_derived, |
| 1159 | commit=commit, invalidate=invalidate) |
| 1160 | |
| 1161 | def _UpdateIssuesSummary(self, cnxn, issues, commit=True): |
| 1162 | """Update the IssueSummary table rows for the given issues.""" |
| 1163 | self.issuesummary_tbl.InsertRows( |
| 1164 | cnxn, ISSUESUMMARY_COLS, |
| 1165 | [(issue.issue_id, issue.summary) for issue in issues], |
| 1166 | replace=True, commit=commit) |
| 1167 | |
| 1168 | def _UpdateIssuesLabels(self, cnxn, issues, commit=True): |
| 1169 | """Update the Issue2Label table rows for the given issues.""" |
| 1170 | label_rows = [] |
| 1171 | for issue in issues: |
| 1172 | issue_shard = issue.issue_id % settings.num_logical_shards |
| 1173 | # TODO(jrobbins): If the user adds many novel labels in one issue update, |
| 1174 | # that could be slow. Solution is to add all new labels in a batch first. |
| 1175 | label_rows.extend( |
| 1176 | (issue.issue_id, |
| 1177 | self._config_service.LookupLabelID(cnxn, issue.project_id, label), |
| 1178 | False, |
| 1179 | issue_shard) |
| 1180 | for label in issue.labels) |
| 1181 | label_rows.extend( |
| 1182 | (issue.issue_id, |
| 1183 | self._config_service.LookupLabelID(cnxn, issue.project_id, label), |
| 1184 | True, |
| 1185 | issue_shard) |
| 1186 | for label in issue.derived_labels) |
| 1187 | |
| 1188 | self.issue2label_tbl.Delete( |
| 1189 | cnxn, issue_id=[issue.issue_id for issue in issues], |
| 1190 | commit=False) |
| 1191 | self.issue2label_tbl.InsertRows( |
| 1192 | cnxn, ISSUE2LABEL_COLS + ['issue_shard'], |
| 1193 | label_rows, ignore=True, commit=commit) |
| 1194 | |
| 1195 | def _UpdateIssuesFields(self, cnxn, issues, commit=True): |
| 1196 | """Update the Issue2FieldValue table rows for the given issues.""" |
| 1197 | fieldvalue_rows = [] |
| 1198 | for issue in issues: |
| 1199 | issue_shard = issue.issue_id % settings.num_logical_shards |
| 1200 | for fv in issue.field_values: |
| 1201 | fieldvalue_rows.append( |
| 1202 | (issue.issue_id, fv.field_id, fv.int_value, fv.str_value, |
| 1203 | fv.user_id or None, fv.date_value, fv.url_value, fv.derived, |
| 1204 | fv.phase_id or None, issue_shard)) |
| 1205 | |
| 1206 | self.issue2fieldvalue_tbl.Delete( |
| 1207 | cnxn, issue_id=[issue.issue_id for issue in issues], commit=False) |
| 1208 | self.issue2fieldvalue_tbl.InsertRows( |
| 1209 | cnxn, ISSUE2FIELDVALUE_COLS + ['issue_shard'], |
| 1210 | fieldvalue_rows, commit=commit) |
| 1211 | |
| 1212 | def _UpdateIssuesComponents(self, cnxn, issues, commit=True): |
| 1213 | """Update the Issue2Component table rows for the given issues.""" |
| 1214 | issue2component_rows = [] |
| 1215 | for issue in issues: |
| 1216 | issue_shard = issue.issue_id % settings.num_logical_shards |
| 1217 | issue2component_rows.extend( |
| 1218 | (issue.issue_id, component_id, False, issue_shard) |
| 1219 | for component_id in issue.component_ids) |
| 1220 | issue2component_rows.extend( |
| 1221 | (issue.issue_id, component_id, True, issue_shard) |
| 1222 | for component_id in issue.derived_component_ids) |
| 1223 | |
| 1224 | self.issue2component_tbl.Delete( |
| 1225 | cnxn, issue_id=[issue.issue_id for issue in issues], commit=False) |
| 1226 | self.issue2component_tbl.InsertRows( |
| 1227 | cnxn, ISSUE2COMPONENT_COLS + ['issue_shard'], |
| 1228 | issue2component_rows, ignore=True, commit=commit) |
| 1229 | |
| 1230 | def _UpdateIssuesCc(self, cnxn, issues, commit=True): |
| 1231 | """Update the Issue2Cc table rows for the given issues.""" |
| 1232 | cc_rows = [] |
| 1233 | for issue in issues: |
| 1234 | issue_shard = issue.issue_id % settings.num_logical_shards |
| 1235 | cc_rows.extend( |
| 1236 | (issue.issue_id, cc_id, False, issue_shard) |
| 1237 | for cc_id in issue.cc_ids) |
| 1238 | cc_rows.extend( |
| 1239 | (issue.issue_id, cc_id, True, issue_shard) |
| 1240 | for cc_id in issue.derived_cc_ids) |
| 1241 | |
| 1242 | self.issue2cc_tbl.Delete( |
| 1243 | cnxn, issue_id=[issue.issue_id for issue in issues], commit=False) |
| 1244 | self.issue2cc_tbl.InsertRows( |
| 1245 | cnxn, ISSUE2CC_COLS + ['issue_shard'], |
| 1246 | cc_rows, ignore=True, commit=commit) |
| 1247 | |
| 1248 | def _UpdateIssuesNotify(self, cnxn, issues, commit=True): |
| 1249 | """Update the Issue2Notify table rows for the given issues.""" |
| 1250 | notify_rows = [] |
| 1251 | for issue in issues: |
| 1252 | derived_rows = [[issue.issue_id, email] |
| 1253 | for email in issue.derived_notify_addrs] |
| 1254 | notify_rows.extend(derived_rows) |
| 1255 | |
| 1256 | self.issue2notify_tbl.Delete( |
| 1257 | cnxn, issue_id=[issue.issue_id for issue in issues], commit=False) |
| 1258 | self.issue2notify_tbl.InsertRows( |
| 1259 | cnxn, ISSUE2NOTIFY_COLS, notify_rows, ignore=True, commit=commit) |
| 1260 | |
| 1261 | def _UpdateIssuesRelation(self, cnxn, issues, commit=True): |
| 1262 | """Update the IssueRelation table rows for the given issues.""" |
| 1263 | relation_rows = [] |
| 1264 | blocking_rows = [] |
| 1265 | dangling_relation_rows = [] |
| 1266 | for issue in issues: |
| 1267 | for i, dst_issue_id in enumerate(issue.blocked_on_iids): |
| 1268 | rank = issue.blocked_on_ranks[i] |
| 1269 | relation_rows.append((issue.issue_id, dst_issue_id, 'blockedon', rank)) |
| 1270 | for dst_issue_id in issue.blocking_iids: |
| 1271 | blocking_rows.append((dst_issue_id, issue.issue_id, 'blockedon')) |
| 1272 | for dst_ref in issue.dangling_blocked_on_refs: |
| 1273 | if dst_ref.ext_issue_identifier: |
| 1274 | dangling_relation_rows.append(( |
| 1275 | issue.issue_id, None, None, |
| 1276 | dst_ref.ext_issue_identifier, 'blockedon')) |
| 1277 | else: |
| 1278 | dangling_relation_rows.append(( |
| 1279 | issue.issue_id, dst_ref.project, dst_ref.issue_id, |
| 1280 | None, 'blockedon')) |
| 1281 | for dst_ref in issue.dangling_blocking_refs: |
| 1282 | if dst_ref.ext_issue_identifier: |
| 1283 | dangling_relation_rows.append(( |
| 1284 | issue.issue_id, None, None, |
| 1285 | dst_ref.ext_issue_identifier, 'blocking')) |
| 1286 | else: |
| 1287 | dangling_relation_rows.append(( |
| 1288 | issue.issue_id, dst_ref.project, dst_ref.issue_id, |
| 1289 | dst_ref.ext_issue_identifier, 'blocking')) |
| 1290 | if issue.merged_into: |
| 1291 | relation_rows.append(( |
| 1292 | issue.issue_id, issue.merged_into, 'mergedinto', None)) |
| 1293 | if issue.merged_into_external: |
| 1294 | dangling_relation_rows.append(( |
| 1295 | issue.issue_id, None, None, |
| 1296 | issue.merged_into_external, 'mergedinto')) |
| 1297 | |
| 1298 | old_blocking = self.issuerelation_tbl.Select( |
| 1299 | cnxn, cols=ISSUERELATION_COLS[:-1], |
| 1300 | dst_issue_id=[issue.issue_id for issue in issues], kind='blockedon') |
| 1301 | relation_rows.extend([ |
| 1302 | (row + (0,)) for row in blocking_rows if row not in old_blocking]) |
| 1303 | delete_rows = [row for row in old_blocking if row not in blocking_rows] |
| 1304 | |
| 1305 | for issue_id, dst_issue_id, kind in delete_rows: |
| 1306 | self.issuerelation_tbl.Delete(cnxn, issue_id=issue_id, |
| 1307 | dst_issue_id=dst_issue_id, kind=kind, commit=False) |
| 1308 | self.issuerelation_tbl.Delete( |
| 1309 | cnxn, issue_id=[issue.issue_id for issue in issues], commit=False) |
| 1310 | self.issuerelation_tbl.InsertRows( |
| 1311 | cnxn, ISSUERELATION_COLS, relation_rows, ignore=True, commit=commit) |
| 1312 | self.danglingrelation_tbl.Delete( |
| 1313 | cnxn, issue_id=[issue.issue_id for issue in issues], commit=False) |
| 1314 | self.danglingrelation_tbl.InsertRows( |
| 1315 | cnxn, DANGLINGRELATION_COLS, dangling_relation_rows, ignore=True, |
| 1316 | commit=commit) |
| 1317 | |
| 1318 | def _UpdateIssuesModified( |
| 1319 | self, cnxn, iids, modified_timestamp=None, invalidate=True): |
| 1320 | """Store a modified timestamp for each of the specified issues.""" |
| 1321 | if not iids: |
| 1322 | return |
| 1323 | delta = {'modified': modified_timestamp or int(time.time())} |
| 1324 | self.issue_tbl.Update(cnxn, delta, id=iids, commit=False) |
| 1325 | if invalidate: |
| 1326 | self.InvalidateIIDs(cnxn, iids) |
| 1327 | |
| 1328 | def _UpdateIssuesApprovals(self, cnxn, issue, commit=True): |
| 1329 | """Update the Issue2ApprovalValue table rows for the given issue.""" |
| 1330 | self.issue2approvalvalue_tbl.Delete( |
| 1331 | cnxn, issue_id=issue.issue_id, commit=commit) |
| 1332 | av_rows = [(av.approval_id, issue.issue_id, av.phase_id, |
| 1333 | av.status.name.lower(), av.setter_id, av.set_on) for |
| 1334 | av in issue.approval_values] |
| 1335 | self.issue2approvalvalue_tbl.InsertRows( |
| 1336 | cnxn, ISSUE2APPROVALVALUE_COLS, av_rows, commit=commit) |
| 1337 | |
| 1338 | approver_rows = [] |
| 1339 | for av in issue.approval_values: |
| 1340 | approver_rows.extend([(av.approval_id, approver_id, issue.issue_id) |
| 1341 | for approver_id in av.approver_ids]) |
| 1342 | self.issueapproval2approver_tbl.Delete( |
| 1343 | cnxn, issue_id=issue.issue_id, commit=commit) |
| 1344 | self.issueapproval2approver_tbl.InsertRows( |
| 1345 | cnxn, ISSUEAPPROVAL2APPROVER_COLS, approver_rows, commit=commit) |
| 1346 | |
| 1347 | def UpdateIssueStructure(self, cnxn, config, issue, template, reporter_id, |
| 1348 | comment_content, commit=True, invalidate=True): |
| 1349 | """Converts the phases and approvals structure of the issue into the |
| 1350 | structure of the given template.""" |
| 1351 | # TODO(jojwang): Remove Field defs that belong to any removed approvals. |
| 1352 | approval_defs_by_id = {ad.approval_id: ad for ad in config.approval_defs} |
| 1353 | issue_avs_by_id = {av.approval_id: av for av in issue.approval_values} |
| 1354 | |
| 1355 | new_approval_surveys = [] |
| 1356 | new_issue_approvals = [] |
| 1357 | |
| 1358 | for template_av in template.approval_values: |
| 1359 | existing_issue_av = issue_avs_by_id.get(template_av.approval_id) |
| 1360 | |
| 1361 | # Update all approval surveys so latest ApprovalDef survey changes |
| 1362 | # appear in the converted issue's approval values. |
| 1363 | ad = approval_defs_by_id.get(template_av.approval_id) |
| 1364 | new_av_approver_ids = [] |
| 1365 | if ad: |
| 1366 | new_av_approver_ids = ad.approver_ids |
| 1367 | new_approval_surveys.append( |
| 1368 | self._MakeIssueComment( |
| 1369 | issue.project_id, reporter_id, ad.survey, |
| 1370 | is_description=True, approval_id=ad.approval_id)) |
| 1371 | else: |
| 1372 | logging.info('ApprovalDef not found for approval %r', template_av) |
| 1373 | |
| 1374 | # Keep approval values as-is if it exists in issue and template |
| 1375 | if existing_issue_av: |
| 1376 | new_av = tracker_bizobj.MakeApprovalValue( |
| 1377 | existing_issue_av.approval_id, |
| 1378 | approver_ids=existing_issue_av.approver_ids, |
| 1379 | status=existing_issue_av.status, |
| 1380 | setter_id=existing_issue_av.setter_id, |
| 1381 | set_on=existing_issue_av.set_on, |
| 1382 | phase_id=template_av.phase_id) |
| 1383 | new_issue_approvals.append(new_av) |
| 1384 | else: |
| 1385 | new_av = tracker_bizobj.MakeApprovalValue( |
| 1386 | template_av.approval_id, approver_ids=new_av_approver_ids, |
| 1387 | status=template_av.status, phase_id=template_av.phase_id) |
| 1388 | new_issue_approvals.append(new_av) |
| 1389 | |
| 1390 | template_phase_by_name = { |
| 1391 | phase.name.lower(): phase for phase in template.phases} |
| 1392 | issue_phase_by_id = {phase.phase_id: phase for phase in issue.phases} |
| 1393 | updated_fvs = [] |
| 1394 | # Trim issue FieldValues or update FieldValue phase_ids |
| 1395 | for fv in issue.field_values: |
| 1396 | # If a fv's phase has the same name as a template's phase, update |
| 1397 | # the fv's phase_id to that of the template phase's. Otherwise, |
| 1398 | # remove the fv. |
| 1399 | if fv.phase_id: |
| 1400 | issue_phase = issue_phase_by_id.get(fv.phase_id) |
| 1401 | if issue_phase and issue_phase.name: |
| 1402 | template_phase = template_phase_by_name.get(issue_phase.name.lower()) |
| 1403 | # TODO(jojwang): monorail:4693, remove this after all 'stable-full' |
| 1404 | # gates have been renamed to 'stable'. |
| 1405 | if not template_phase: |
| 1406 | template_phase = template_phase_by_name.get( |
| 1407 | FLT_EQUIVALENT_GATES.get(issue_phase.name.lower())) |
| 1408 | if template_phase: |
| 1409 | fv.phase_id = template_phase.phase_id |
| 1410 | updated_fvs.append(fv) |
| 1411 | # keep all fvs that do not belong to phases. |
| 1412 | else: |
| 1413 | updated_fvs.append(fv) |
| 1414 | |
| 1415 | fd_names_by_id = {fd.field_id: fd.field_name for fd in config.field_defs} |
| 1416 | amendment = tracker_bizobj.MakeApprovalStructureAmendment( |
| 1417 | [fd_names_by_id.get(av.approval_id) for av in new_issue_approvals], |
| 1418 | [fd_names_by_id.get(av.approval_id) for av in issue.approval_values]) |
| 1419 | |
| 1420 | # Update issue structure in RAM. |
| 1421 | issue.approval_values = new_issue_approvals |
| 1422 | issue.phases = template.phases |
| 1423 | issue.field_values = updated_fvs |
| 1424 | |
| 1425 | # Update issue structure in DB. |
| 1426 | for survey in new_approval_surveys: |
| 1427 | survey.issue_id = issue.issue_id |
| 1428 | self.InsertComment(cnxn, survey, commit=False) |
| 1429 | self._UpdateIssuesApprovals(cnxn, issue, commit=False) |
| 1430 | self._UpdateIssuesFields(cnxn, [issue], commit=False) |
| 1431 | comment_pb = self.CreateIssueComment( |
| 1432 | cnxn, issue, reporter_id, comment_content, |
| 1433 | amendments=[amendment], commit=False) |
| 1434 | |
| 1435 | if commit: |
| 1436 | cnxn.Commit() |
| 1437 | |
| 1438 | if invalidate: |
| 1439 | self.InvalidateIIDs(cnxn, [issue.issue_id]) |
| 1440 | |
| 1441 | return comment_pb |
| 1442 | |
| 1443 | def DeltaUpdateIssue( |
| 1444 | self, cnxn, services, reporter_id, project_id, |
| 1445 | config, issue, delta, index_now=False, comment=None, attachments=None, |
| 1446 | iids_to_invalidate=None, rules=None, predicate_asts=None, |
| 1447 | is_description=False, timestamp=None, kept_attachments=None, |
| 1448 | importer_id=None, inbound_message=None): |
| 1449 | """Update the issue in the database and return a set of update tuples. |
| 1450 | |
| 1451 | Args: |
| 1452 | cnxn: connection to SQL database. |
| 1453 | services: connections to persistence layer. |
| 1454 | reporter_id: user ID of the user making this change. |
| 1455 | project_id: int ID for the current project. |
| 1456 | config: ProjectIssueConfig PB for this project. |
| 1457 | issue: Issue PB of issue to update. |
| 1458 | delta: IssueDelta object of fields to update. |
| 1459 | index_now: True if the issue should be updated in the full text index. |
| 1460 | comment: This should be the content of the comment |
| 1461 | corresponding to this change. |
| 1462 | attachments: List [(filename, contents, mimetype),...] of attachments. |
| 1463 | iids_to_invalidate: optional set of issue IDs that need to be invalidated. |
| 1464 | If provided, affected issues will be accumulated here and, the caller |
| 1465 | must call InvalidateIIDs() afterwards. |
| 1466 | rules: optional list of preloaded FilterRule PBs for this project. |
| 1467 | predicate_asts: optional list of QueryASTs for the rules. If rules are |
| 1468 | provided, then predicate_asts should also be provided. |
| 1469 | is_description: True if the comment is a new description for the issue. |
| 1470 | timestamp: int timestamp set during testing, otherwise defaults to |
| 1471 | int(time.time()). |
| 1472 | kept_attachments: This should be a list of int attachment ids for |
| 1473 | attachments kept from previous descriptions, if the comment is |
| 1474 | a change to the issue description |
| 1475 | importer_id: optional ID of user ID for an API client that is importing |
| 1476 | issues and attributing them to other users. |
| 1477 | inbound_message: optional string full text of an email that caused |
| 1478 | this comment to be added. |
| 1479 | |
| 1480 | Returns: |
| 1481 | A tuple (amendments, comment_pb) with a list of Amendment PBs that |
| 1482 | describe the set of metadata updates that the user made, and the |
| 1483 | resulting IssueComment (or None if no comment was created). |
| 1484 | """ |
| 1485 | timestamp = timestamp or int(time.time()) |
| 1486 | old_effective_owner = tracker_bizobj.GetOwnerId(issue) |
| 1487 | old_effective_status = tracker_bizobj.GetStatus(issue) |
| 1488 | old_components = set(issue.component_ids) |
| 1489 | |
| 1490 | logging.info( |
| 1491 | 'Bulk edit to project_id %s issue.local_id %s, comment %r', |
| 1492 | project_id, issue.local_id, comment) |
| 1493 | if iids_to_invalidate is None: |
| 1494 | iids_to_invalidate = set([issue.issue_id]) |
| 1495 | invalidate = True |
| 1496 | else: |
| 1497 | iids_to_invalidate.add(issue.issue_id) |
| 1498 | invalidate = False # Caller will do it. |
| 1499 | |
| 1500 | # Store each updated value in the issue PB, and compute Update PBs |
| 1501 | amendments, impacted_iids = tracker_bizobj.ApplyIssueDelta( |
| 1502 | cnxn, self, issue, delta, config) |
| 1503 | iids_to_invalidate.update(impacted_iids) |
| 1504 | |
| 1505 | # If this was a no-op with no comment, bail out and don't save, |
| 1506 | # invalidate, or re-index anything. |
| 1507 | if (not amendments and (not comment or not comment.strip()) and |
| 1508 | not attachments): |
| 1509 | logging.info('No amendments, comment, attachments: this is a no-op.') |
| 1510 | return [], None |
| 1511 | |
| 1512 | # Note: no need to check for collisions when the user is doing a delta. |
| 1513 | |
| 1514 | # update the modified_timestamp for any comment added, even if it was |
| 1515 | # just a text comment with no issue fields changed. |
| 1516 | issue.modified_timestamp = timestamp |
| 1517 | |
| 1518 | # Update the closed timestamp before filter rules so that rules |
| 1519 | # can test for closed_timestamp, and also after filter rules |
| 1520 | # so that closed_timestamp will be set if the issue is closed by the rule. |
| 1521 | tracker_helpers.UpdateClosedTimestamp(config, issue, old_effective_status) |
| 1522 | if rules is None: |
| 1523 | logging.info('Rules were not given') |
| 1524 | rules = services.features.GetFilterRules(cnxn, config.project_id) |
| 1525 | predicate_asts = filterrules_helpers.ParsePredicateASTs( |
| 1526 | rules, config, []) |
| 1527 | |
| 1528 | filterrules_helpers.ApplyGivenRules( |
| 1529 | cnxn, services, issue, config, rules, predicate_asts) |
| 1530 | tracker_helpers.UpdateClosedTimestamp(config, issue, old_effective_status) |
| 1531 | if old_effective_owner != tracker_bizobj.GetOwnerId(issue): |
| 1532 | issue.owner_modified_timestamp = timestamp |
| 1533 | if old_effective_status != tracker_bizobj.GetStatus(issue): |
| 1534 | issue.status_modified_timestamp = timestamp |
| 1535 | if old_components != set(issue.component_ids): |
| 1536 | issue.component_modified_timestamp = timestamp |
| 1537 | |
| 1538 | # Store the issue in SQL. |
| 1539 | self.UpdateIssue(cnxn, issue, commit=False, invalidate=False) |
| 1540 | |
| 1541 | comment_pb = self.CreateIssueComment( |
| 1542 | cnxn, issue, reporter_id, comment, amendments=amendments, |
| 1543 | is_description=is_description, attachments=attachments, commit=False, |
| 1544 | kept_attachments=kept_attachments, timestamp=timestamp, |
| 1545 | importer_id=importer_id, inbound_message=inbound_message) |
| 1546 | self._UpdateIssuesModified( |
| 1547 | cnxn, iids_to_invalidate, modified_timestamp=issue.modified_timestamp, |
| 1548 | invalidate=invalidate) |
| 1549 | |
| 1550 | # Add a comment to the newly added issues saying they are now blocking |
| 1551 | # this issue. |
| 1552 | for add_issue in self.GetIssues(cnxn, delta.blocked_on_add): |
| 1553 | self.CreateIssueComment( |
| 1554 | cnxn, add_issue, reporter_id, content='', |
| 1555 | amendments=[tracker_bizobj.MakeBlockingAmendment( |
| 1556 | [(issue.project_name, issue.local_id)], [], |
| 1557 | default_project_name=add_issue.project_name)], |
| 1558 | timestamp=timestamp, importer_id=importer_id) |
| 1559 | # Add a comment to the newly removed issues saying they are no longer |
| 1560 | # blocking this issue. |
| 1561 | for remove_issue in self.GetIssues(cnxn, delta.blocked_on_remove): |
| 1562 | self.CreateIssueComment( |
| 1563 | cnxn, remove_issue, reporter_id, content='', |
| 1564 | amendments=[tracker_bizobj.MakeBlockingAmendment( |
| 1565 | [], [(issue.project_name, issue.local_id)], |
| 1566 | default_project_name=remove_issue.project_name)], |
| 1567 | timestamp=timestamp, importer_id=importer_id) |
| 1568 | |
| 1569 | # Add a comment to the newly added issues saying they are now blocked on |
| 1570 | # this issue. |
| 1571 | for add_issue in self.GetIssues(cnxn, delta.blocking_add): |
| 1572 | self.CreateIssueComment( |
| 1573 | cnxn, add_issue, reporter_id, content='', |
| 1574 | amendments=[tracker_bizobj.MakeBlockedOnAmendment( |
| 1575 | [(issue.project_name, issue.local_id)], [], |
| 1576 | default_project_name=add_issue.project_name)], |
| 1577 | timestamp=timestamp, importer_id=importer_id) |
| 1578 | # Add a comment to the newly removed issues saying they are no longer |
| 1579 | # blocked on this issue. |
| 1580 | for remove_issue in self.GetIssues(cnxn, delta.blocking_remove): |
| 1581 | self.CreateIssueComment( |
| 1582 | cnxn, remove_issue, reporter_id, content='', |
| 1583 | amendments=[tracker_bizobj.MakeBlockedOnAmendment( |
| 1584 | [], [(issue.project_name, issue.local_id)], |
| 1585 | default_project_name=remove_issue.project_name)], |
| 1586 | timestamp=timestamp, importer_id=importer_id) |
| 1587 | |
| 1588 | if not invalidate: |
| 1589 | cnxn.Commit() |
| 1590 | |
| 1591 | if index_now: |
| 1592 | tracker_fulltext.IndexIssues( |
| 1593 | cnxn, [issue], services.user_service, self, self._config_service) |
| 1594 | else: |
| 1595 | self.EnqueueIssuesForIndexing(cnxn, [issue.issue_id]) |
| 1596 | |
| 1597 | return amendments, comment_pb |
| 1598 | |
| 1599 | def InvalidateIIDs(self, cnxn, iids_to_invalidate): |
| 1600 | """Invalidate the specified issues in the Invalidate table and memcache.""" |
| 1601 | issues_to_invalidate = self.GetIssues(cnxn, iids_to_invalidate) |
| 1602 | self.InvalidateIssues(cnxn, issues_to_invalidate) |
| 1603 | |
| 1604 | def InvalidateIssues(self, cnxn, issues): |
| 1605 | """Invalidate the specified issues in the Invalidate table and memcache.""" |
| 1606 | iids = [issue.issue_id for issue in issues] |
| 1607 | self.issue_2lc.InvalidateKeys(cnxn, iids) |
| 1608 | self._config_service.InvalidateMemcache(issues) |
| 1609 | |
| 1610 | def RelateIssues(self, cnxn, issue_relation_dict, commit=True): |
| 1611 | """Update the IssueRelation table rows for the given relationships. |
| 1612 | |
| 1613 | issue_relation_dict is a mapping of 'source' issues to 'destination' issues, |
| 1614 | paired with the kind of relationship connecting the two. |
| 1615 | """ |
| 1616 | relation_rows = [] |
| 1617 | for src_iid, dests in issue_relation_dict.items(): |
| 1618 | for dst_iid, kind in dests: |
| 1619 | if kind == 'blocking': |
| 1620 | relation_rows.append((dst_iid, src_iid, 'blockedon', 0)) |
| 1621 | elif kind == 'blockedon': |
| 1622 | relation_rows.append((src_iid, dst_iid, 'blockedon', 0)) |
| 1623 | elif kind == 'mergedinto': |
| 1624 | relation_rows.append((src_iid, dst_iid, 'mergedinto', None)) |
| 1625 | |
| 1626 | self.issuerelation_tbl.InsertRows( |
| 1627 | cnxn, ISSUERELATION_COLS, relation_rows, ignore=True, commit=commit) |
| 1628 | |
| 1629 | def CopyIssues(self, cnxn, dest_project, issues, user_service, copier_id): |
| 1630 | """Copy the given issues into the destination project.""" |
| 1631 | created_issues = [] |
| 1632 | iids_to_invalidate = set() |
| 1633 | |
| 1634 | for target_issue in issues: |
| 1635 | assert not target_issue.assume_stale, ( |
| 1636 | 'issue2514: Copying issue that might be stale: %r' % target_issue) |
| 1637 | new_issue = tracker_pb2.Issue() |
| 1638 | new_issue.project_id = dest_project.project_id |
| 1639 | new_issue.project_name = dest_project.project_name |
| 1640 | new_issue.summary = target_issue.summary |
| 1641 | new_issue.labels.extend(target_issue.labels) |
| 1642 | new_issue.field_values.extend(target_issue.field_values) |
| 1643 | new_issue.reporter_id = copier_id |
| 1644 | |
| 1645 | timestamp = int(time.time()) |
| 1646 | new_issue.opened_timestamp = timestamp |
| 1647 | new_issue.modified_timestamp = timestamp |
| 1648 | |
| 1649 | target_comments = self.GetCommentsForIssue(cnxn, target_issue.issue_id) |
| 1650 | initial_summary_comment = target_comments[0] |
| 1651 | |
| 1652 | # Note that blocking and merge_into are not copied. |
| 1653 | if target_issue.blocked_on_iids: |
| 1654 | blocked_on = target_issue.blocked_on_iids |
| 1655 | iids_to_invalidate.update(blocked_on) |
| 1656 | new_issue.blocked_on_iids = blocked_on |
| 1657 | |
| 1658 | # Gather list of attachments from the target issue's summary comment. |
| 1659 | # MakeIssueComments expects a list of [(filename, contents, mimetype),...] |
| 1660 | attachments = [] |
| 1661 | for attachment in initial_summary_comment.attachments: |
| 1662 | object_path = ('/' + app_identity.get_default_gcs_bucket_name() + |
| 1663 | attachment.gcs_object_id) |
| 1664 | with cloudstorage.open(object_path, 'r') as f: |
| 1665 | content = f.read() |
| 1666 | attachments.append( |
| 1667 | [attachment.filename, content, attachment.mimetype]) |
| 1668 | |
| 1669 | if attachments: |
| 1670 | new_issue.attachment_count = len(attachments) |
| 1671 | |
| 1672 | # Create the same summary comment as the target issue. |
| 1673 | comment = self._MakeIssueComment( |
| 1674 | dest_project.project_id, copier_id, initial_summary_comment.content, |
| 1675 | attachments=attachments, timestamp=timestamp, is_description=True) |
| 1676 | |
| 1677 | new_issue.local_id = self.AllocateNextLocalID( |
| 1678 | cnxn, dest_project.project_id) |
| 1679 | issue_id = self.InsertIssue(cnxn, new_issue) |
| 1680 | comment.issue_id = issue_id |
| 1681 | self.InsertComment(cnxn, comment) |
| 1682 | |
| 1683 | if permissions.HasRestrictions(new_issue, 'view'): |
| 1684 | self._config_service.InvalidateMemcache( |
| 1685 | [new_issue], key_prefix='nonviewable:') |
| 1686 | |
| 1687 | tracker_fulltext.IndexIssues( |
| 1688 | cnxn, [new_issue], user_service, self, self._config_service) |
| 1689 | created_issues.append(new_issue) |
| 1690 | |
| 1691 | # The referenced issues are all modified when the relationship is added. |
| 1692 | self._UpdateIssuesModified( |
| 1693 | cnxn, iids_to_invalidate, modified_timestamp=timestamp) |
| 1694 | |
| 1695 | return created_issues |
| 1696 | |
| 1697 | def MoveIssues(self, cnxn, dest_project, issues, user_service): |
| 1698 | """Move the given issues into the destination project.""" |
| 1699 | old_location_rows = [ |
| 1700 | (issue.issue_id, issue.project_id, issue.local_id) |
| 1701 | for issue in issues] |
| 1702 | moved_back_iids = set() |
| 1703 | |
| 1704 | former_locations_in_project = self.issueformerlocations_tbl.Select( |
| 1705 | cnxn, cols=ISSUEFORMERLOCATIONS_COLS, |
| 1706 | project_id=dest_project.project_id, |
| 1707 | issue_id=[issue.issue_id for issue in issues]) |
| 1708 | former_locations = { |
| 1709 | issue_id: local_id |
| 1710 | for issue_id, project_id, local_id in former_locations_in_project} |
| 1711 | |
| 1712 | # Remove the issue id from issue_id_2lc so that it does not stay |
| 1713 | # around in cache and memcache. |
| 1714 | # The Key of IssueIDTwoLevelCache is (project_id, local_id). |
| 1715 | self.issue_id_2lc.InvalidateKeys( |
| 1716 | cnxn, [(issue.project_id, issue.local_id) for issue in issues]) |
| 1717 | self.InvalidateIssues(cnxn, issues) |
| 1718 | |
| 1719 | for issue in issues: |
| 1720 | if issue.issue_id in former_locations: |
| 1721 | dest_id = former_locations[issue.issue_id] |
| 1722 | moved_back_iids.add(issue.issue_id) |
| 1723 | else: |
| 1724 | dest_id = self.AllocateNextLocalID(cnxn, dest_project.project_id) |
| 1725 | |
| 1726 | issue.local_id = dest_id |
| 1727 | issue.project_id = dest_project.project_id |
| 1728 | issue.project_name = dest_project.project_name |
| 1729 | |
| 1730 | # Rewrite each whole issue so that status and label IDs are looked up |
| 1731 | # in the context of the destination project. |
| 1732 | self.UpdateIssues(cnxn, issues) |
| 1733 | |
| 1734 | # Comments also have the project_id because it is needed for an index. |
| 1735 | self.comment_tbl.Update( |
| 1736 | cnxn, {'project_id': dest_project.project_id}, |
| 1737 | issue_id=[issue.issue_id for issue in issues], commit=False) |
| 1738 | |
| 1739 | # Record old locations so that we can offer links if the user looks there. |
| 1740 | self.issueformerlocations_tbl.InsertRows( |
| 1741 | cnxn, ISSUEFORMERLOCATIONS_COLS, old_location_rows, ignore=True, |
| 1742 | commit=False) |
| 1743 | cnxn.Commit() |
| 1744 | |
| 1745 | tracker_fulltext.IndexIssues( |
| 1746 | cnxn, issues, user_service, self, self._config_service) |
| 1747 | |
| 1748 | return moved_back_iids |
| 1749 | |
| 1750 | def ExpungeFormerLocations(self, cnxn, project_id): |
| 1751 | """Delete history of issues that were in this project but moved out.""" |
| 1752 | self.issueformerlocations_tbl.Delete(cnxn, project_id=project_id) |
| 1753 | |
| 1754 | def ExpungeIssues(self, cnxn, issue_ids): |
| 1755 | """Completely delete the specified issues from the database.""" |
| 1756 | logging.info('expunging the issues %r', issue_ids) |
| 1757 | tracker_fulltext.UnindexIssues(issue_ids) |
| 1758 | |
| 1759 | remaining_iids = issue_ids[:] |
| 1760 | |
| 1761 | # Note: these are purposely not done in a transaction to allow |
| 1762 | # incremental progress in what might be a very large change. |
| 1763 | # We are not concerned about non-atomic deletes because all |
| 1764 | # this data will be gone eventually anyway. |
| 1765 | while remaining_iids: |
| 1766 | iids_in_chunk = remaining_iids[:CHUNK_SIZE] |
| 1767 | remaining_iids = remaining_iids[CHUNK_SIZE:] |
| 1768 | self.issuesummary_tbl.Delete(cnxn, issue_id=iids_in_chunk) |
| 1769 | self.issue2label_tbl.Delete(cnxn, issue_id=iids_in_chunk) |
| 1770 | self.issue2component_tbl.Delete(cnxn, issue_id=iids_in_chunk) |
| 1771 | self.issue2cc_tbl.Delete(cnxn, issue_id=iids_in_chunk) |
| 1772 | self.issue2notify_tbl.Delete(cnxn, issue_id=iids_in_chunk) |
| 1773 | self.issueupdate_tbl.Delete(cnxn, issue_id=iids_in_chunk) |
| 1774 | self.attachment_tbl.Delete(cnxn, issue_id=iids_in_chunk) |
| 1775 | self.comment_tbl.Delete(cnxn, issue_id=iids_in_chunk) |
| 1776 | self.issuerelation_tbl.Delete(cnxn, issue_id=iids_in_chunk) |
| 1777 | self.issuerelation_tbl.Delete(cnxn, dst_issue_id=iids_in_chunk) |
| 1778 | self.danglingrelation_tbl.Delete(cnxn, issue_id=iids_in_chunk) |
| 1779 | self.issueformerlocations_tbl.Delete(cnxn, issue_id=iids_in_chunk) |
| 1780 | self.reindexqueue_tbl.Delete(cnxn, issue_id=iids_in_chunk) |
| 1781 | self.issue_tbl.Delete(cnxn, id=iids_in_chunk) |
| 1782 | |
| 1783 | def SoftDeleteIssue(self, cnxn, project_id, local_id, deleted, user_service): |
| 1784 | """Set the deleted boolean on the indicated issue and store it. |
| 1785 | |
| 1786 | Args: |
| 1787 | cnxn: connection to SQL database. |
| 1788 | project_id: int project ID for the current project. |
| 1789 | local_id: int local ID of the issue to freeze/unfreeze. |
| 1790 | deleted: boolean, True to soft-delete, False to undelete. |
| 1791 | user_service: persistence layer for users, used to lookup user IDs. |
| 1792 | """ |
| 1793 | issue = self.GetIssueByLocalID(cnxn, project_id, local_id, use_cache=False) |
| 1794 | issue.deleted = deleted |
| 1795 | self.UpdateIssue(cnxn, issue, update_cols=['deleted']) |
| 1796 | tracker_fulltext.IndexIssues( |
| 1797 | cnxn, [issue], user_service, self, self._config_service) |
| 1798 | |
| 1799 | def DeleteComponentReferences(self, cnxn, component_id): |
| 1800 | """Delete any references to the specified component.""" |
| 1801 | # TODO(jrobbins): add tasks to re-index any affected issues. |
| 1802 | # Note: if this call fails, some data could be left |
| 1803 | # behind, but it would not be displayed, and it could always be |
| 1804 | # GC'd from the DB later. |
| 1805 | self.issue2component_tbl.Delete(cnxn, component_id=component_id) |
| 1806 | |
| 1807 | ### Local ID generation |
| 1808 | |
| 1809 | def InitializeLocalID(self, cnxn, project_id): |
| 1810 | """Initialize the local ID counter for the specified project to zero. |
| 1811 | |
| 1812 | Args: |
| 1813 | cnxn: connection to SQL database. |
| 1814 | project_id: int ID of the project. |
| 1815 | """ |
| 1816 | self.localidcounter_tbl.InsertRow( |
| 1817 | cnxn, project_id=project_id, used_local_id=0, used_spam_id=0) |
| 1818 | |
| 1819 | def SetUsedLocalID(self, cnxn, project_id): |
| 1820 | """Set the local ID counter based on existing issues. |
| 1821 | |
| 1822 | Args: |
| 1823 | cnxn: connection to SQL database. |
| 1824 | project_id: int ID of the project. |
| 1825 | """ |
| 1826 | highest_id = self.GetHighestLocalID(cnxn, project_id) |
| 1827 | self.localidcounter_tbl.InsertRow( |
| 1828 | cnxn, replace=True, used_local_id=highest_id, project_id=project_id) |
| 1829 | return highest_id |
| 1830 | |
| 1831 | def AllocateNextLocalID(self, cnxn, project_id): |
| 1832 | """Return the next available issue ID in the specified project. |
| 1833 | |
| 1834 | Args: |
| 1835 | cnxn: connection to SQL database. |
| 1836 | project_id: int ID of the project. |
| 1837 | |
| 1838 | Returns: |
| 1839 | The next local ID. |
| 1840 | """ |
| 1841 | try: |
| 1842 | next_local_id = self.localidcounter_tbl.IncrementCounterValue( |
| 1843 | cnxn, 'used_local_id', project_id=project_id) |
| 1844 | except AssertionError as e: |
| 1845 | logging.info('exception incrementing local_id counter: %s', e) |
| 1846 | next_local_id = self.SetUsedLocalID(cnxn, project_id) + 1 |
| 1847 | return next_local_id |
| 1848 | |
| 1849 | def GetHighestLocalID(self, cnxn, project_id): |
| 1850 | """Return the highest used issue ID in the specified project. |
| 1851 | |
| 1852 | Args: |
| 1853 | cnxn: connection to SQL database. |
| 1854 | project_id: int ID of the project. |
| 1855 | |
| 1856 | Returns: |
| 1857 | The highest local ID for an active or moved issues. |
| 1858 | """ |
| 1859 | highest = self.issue_tbl.SelectValue( |
| 1860 | cnxn, 'MAX(local_id)', project_id=project_id) |
| 1861 | highest = highest or 0 # It will be None if the project has no issues. |
| 1862 | highest_former = self.issueformerlocations_tbl.SelectValue( |
| 1863 | cnxn, 'MAX(local_id)', project_id=project_id) |
| 1864 | highest_former = highest_former or 0 |
| 1865 | return max(highest, highest_former) |
| 1866 | |
| 1867 | def GetAllLocalIDsInProject(self, cnxn, project_id, min_local_id=None): |
| 1868 | """Return the list of local IDs only, not the actual issues. |
| 1869 | |
| 1870 | Args: |
| 1871 | cnxn: connection to SQL database. |
| 1872 | project_id: the ID of the project to which the issue belongs. |
| 1873 | min_local_id: point to start at. |
| 1874 | |
| 1875 | Returns: |
| 1876 | A range object of local IDs from 1 to N, or from min_local_id to N. It |
| 1877 | may be the case that some of those local IDs are no longer used, e.g., |
| 1878 | if some issues were moved out of this project. |
| 1879 | """ |
| 1880 | if not min_local_id: |
| 1881 | min_local_id = 1 |
| 1882 | highest_local_id = self.GetHighestLocalID(cnxn, project_id) |
| 1883 | return list(range(min_local_id, highest_local_id + 1)) |
| 1884 | |
| 1885 | def ExpungeLocalIDCounters(self, cnxn, project_id): |
| 1886 | """Delete history of local ids that were in this project.""" |
| 1887 | self.localidcounter_tbl.Delete(cnxn, project_id=project_id) |
| 1888 | |
| 1889 | ### Comments |
| 1890 | |
| 1891 | def _UnpackComment( |
| 1892 | self, comment_row, content_dict, inbound_message_dict, approval_dict, |
| 1893 | importer_dict): |
| 1894 | """Partially construct a Comment PB from a DB row.""" |
| 1895 | (comment_id, issue_id, created, project_id, commenter_id, |
| 1896 | deleted_by, is_spam, is_description, commentcontent_id) = comment_row |
| 1897 | comment = tracker_pb2.IssueComment() |
| 1898 | comment.id = comment_id |
| 1899 | comment.issue_id = issue_id |
| 1900 | comment.timestamp = created |
| 1901 | comment.project_id = project_id |
| 1902 | comment.user_id = commenter_id |
| 1903 | comment.content = content_dict.get(commentcontent_id, '') |
| 1904 | comment.inbound_message = inbound_message_dict.get(commentcontent_id, '') |
| 1905 | comment.deleted_by = deleted_by or 0 |
| 1906 | comment.is_spam = bool(is_spam) |
| 1907 | comment.is_description = bool(is_description) |
| 1908 | comment.approval_id = approval_dict.get(comment_id) |
| 1909 | comment.importer_id = importer_dict.get(comment_id) |
| 1910 | return comment |
| 1911 | |
| 1912 | def _UnpackAmendment(self, amendment_row): |
| 1913 | """Construct an Amendment PB from a DB row.""" |
| 1914 | (_id, _issue_id, comment_id, field_name, |
| 1915 | old_value, new_value, added_user_id, removed_user_id, |
| 1916 | custom_field_name) = amendment_row |
| 1917 | amendment = tracker_pb2.Amendment() |
| 1918 | field_enum = tracker_pb2.FieldID(field_name.upper()) |
| 1919 | amendment.field = field_enum |
| 1920 | |
| 1921 | # TODO(jrobbins): display old values in more cases. |
| 1922 | if new_value is not None: |
| 1923 | amendment.newvalue = new_value |
| 1924 | if old_value is not None: |
| 1925 | amendment.oldvalue = old_value |
| 1926 | if added_user_id: |
| 1927 | amendment.added_user_ids.append(added_user_id) |
| 1928 | if removed_user_id: |
| 1929 | amendment.removed_user_ids.append(removed_user_id) |
| 1930 | if custom_field_name: |
| 1931 | amendment.custom_field_name = custom_field_name |
| 1932 | return amendment, comment_id |
| 1933 | |
| 1934 | def _ConsolidateAmendments(self, amendments): |
| 1935 | """Consoliodate amendments of the same field in one comment into one |
| 1936 | amendment PB.""" |
| 1937 | |
| 1938 | fields_dict = {} |
| 1939 | result = [] |
| 1940 | |
| 1941 | for amendment in amendments: |
| 1942 | key = amendment.field, amendment.custom_field_name |
| 1943 | fields_dict.setdefault(key, []).append(amendment) |
| 1944 | for (field, _custom_name), sorted_amendments in sorted(fields_dict.items()): |
| 1945 | new_amendment = tracker_pb2.Amendment() |
| 1946 | new_amendment.field = field |
| 1947 | for amendment in sorted_amendments: |
| 1948 | if amendment.newvalue is not None: |
| 1949 | if new_amendment.newvalue is not None: |
| 1950 | # NOTE: see crbug/monorail/8272. BLOCKEDON and BLOCKING changes |
| 1951 | # are all stored in newvalue e.g. (newvalue = -b/123 b/124) and |
| 1952 | # external bugs and monorail bugs are stored in separate amendments. |
| 1953 | # Without this, the values of external bug amendments and monorail |
| 1954 | # blocker bug amendments may overwrite each other. |
| 1955 | new_amendment.newvalue += (' ' + amendment.newvalue) |
| 1956 | else: |
| 1957 | new_amendment.newvalue = amendment.newvalue |
| 1958 | if amendment.oldvalue is not None: |
| 1959 | new_amendment.oldvalue = amendment.oldvalue |
| 1960 | if amendment.added_user_ids: |
| 1961 | new_amendment.added_user_ids.extend(amendment.added_user_ids) |
| 1962 | if amendment.removed_user_ids: |
| 1963 | new_amendment.removed_user_ids.extend(amendment.removed_user_ids) |
| 1964 | if amendment.custom_field_name: |
| 1965 | new_amendment.custom_field_name = amendment.custom_field_name |
| 1966 | result.append(new_amendment) |
| 1967 | return result |
| 1968 | |
| 1969 | def _UnpackAttachment(self, attachment_row): |
| 1970 | """Construct an Attachment PB from a DB row.""" |
| 1971 | (attachment_id, _issue_id, comment_id, filename, filesize, mimetype, |
| 1972 | deleted, gcs_object_id) = attachment_row |
| 1973 | attach = tracker_pb2.Attachment() |
| 1974 | attach.attachment_id = attachment_id |
| 1975 | attach.filename = filename |
| 1976 | attach.filesize = filesize |
| 1977 | attach.mimetype = mimetype |
| 1978 | attach.deleted = bool(deleted) |
| 1979 | attach.gcs_object_id = gcs_object_id |
| 1980 | return attach, comment_id |
| 1981 | |
| 1982 | def _DeserializeComments( |
| 1983 | self, comment_rows, commentcontent_rows, amendment_rows, attachment_rows, |
| 1984 | approval_rows, importer_rows): |
| 1985 | """Turn rows into IssueComment PBs.""" |
| 1986 | results = [] # keep objects in the same order as the rows |
| 1987 | results_dict = {} # for fast access when joining. |
| 1988 | |
| 1989 | content_dict = dict( |
| 1990 | (commentcontent_id, content) for |
| 1991 | commentcontent_id, content, _ in commentcontent_rows) |
| 1992 | inbound_message_dict = dict( |
| 1993 | (commentcontent_id, inbound_message) for |
| 1994 | commentcontent_id, _, inbound_message in commentcontent_rows) |
| 1995 | approval_dict = dict( |
| 1996 | (comment_id, approval_id) for approval_id, comment_id in |
| 1997 | approval_rows) |
| 1998 | importer_dict = dict(importer_rows) |
| 1999 | |
| 2000 | for comment_row in comment_rows: |
| 2001 | comment = self._UnpackComment( |
| 2002 | comment_row, content_dict, inbound_message_dict, approval_dict, |
| 2003 | importer_dict) |
| 2004 | results.append(comment) |
| 2005 | results_dict[comment.id] = comment |
| 2006 | |
| 2007 | for amendment_row in amendment_rows: |
| 2008 | amendment, comment_id = self._UnpackAmendment(amendment_row) |
| 2009 | try: |
| 2010 | results_dict[comment_id].amendments.extend([amendment]) |
| 2011 | except KeyError: |
| 2012 | logging.error('Found amendment for missing comment: %r', comment_id) |
| 2013 | |
| 2014 | for attachment_row in attachment_rows: |
| 2015 | attach, comment_id = self._UnpackAttachment(attachment_row) |
| 2016 | try: |
| 2017 | results_dict[comment_id].attachments.append(attach) |
| 2018 | except KeyError: |
| 2019 | logging.error('Found attachment for missing comment: %r', comment_id) |
| 2020 | |
| 2021 | for c in results: |
| 2022 | c.amendments = self._ConsolidateAmendments(c.amendments) |
| 2023 | |
| 2024 | return results |
| 2025 | |
| 2026 | # TODO(jrobbins): make this a private method and expose just the interface |
| 2027 | # needed by activities.py. |
| 2028 | def GetComments( |
| 2029 | self, cnxn, where=None, order_by=None, content_only=False, **kwargs): |
| 2030 | """Retrieve comments from SQL.""" |
| 2031 | shard_id = sql.RandomShardID() |
| 2032 | order_by = order_by or [('created', [])] |
| 2033 | comment_rows = self.comment_tbl.Select( |
| 2034 | cnxn, cols=COMMENT_COLS, where=where, |
| 2035 | order_by=order_by, shard_id=shard_id, **kwargs) |
| 2036 | cids = [row[0] for row in comment_rows] |
| 2037 | commentcontent_ids = [row[-1] for row in comment_rows] |
| 2038 | content_rows = self.commentcontent_tbl.Select( |
| 2039 | cnxn, cols=COMMENTCONTENT_COLS, id=commentcontent_ids, |
| 2040 | shard_id=shard_id) |
| 2041 | approval_rows = self.issueapproval2comment_tbl.Select( |
| 2042 | cnxn, cols=ISSUEAPPROVAL2COMMENT_COLS, comment_id=cids) |
| 2043 | amendment_rows = [] |
| 2044 | attachment_rows = [] |
| 2045 | importer_rows = [] |
| 2046 | if not content_only: |
| 2047 | amendment_rows = self.issueupdate_tbl.Select( |
| 2048 | cnxn, cols=ISSUEUPDATE_COLS, comment_id=cids, shard_id=shard_id) |
| 2049 | attachment_rows = self.attachment_tbl.Select( |
| 2050 | cnxn, cols=ATTACHMENT_COLS, comment_id=cids, shard_id=shard_id) |
| 2051 | importer_rows = self.commentimporter_tbl.Select( |
| 2052 | cnxn, cols=COMMENTIMPORTER_COLS, comment_id=cids, shard_id=shard_id) |
| 2053 | |
| 2054 | comments = self._DeserializeComments( |
| 2055 | comment_rows, content_rows, amendment_rows, attachment_rows, |
| 2056 | approval_rows, importer_rows) |
| 2057 | return comments |
| 2058 | |
| 2059 | def GetComment(self, cnxn, comment_id): |
| 2060 | """Get the requested comment, or raise an exception.""" |
| 2061 | comments = self.GetComments(cnxn, id=comment_id) |
| 2062 | try: |
| 2063 | return comments[0] |
| 2064 | except IndexError: |
| 2065 | raise exceptions.NoSuchCommentException() |
| 2066 | |
| 2067 | def GetCommentsForIssue(self, cnxn, issue_id): |
| 2068 | """Return all IssueComment PBs for the specified issue. |
| 2069 | |
| 2070 | Args: |
| 2071 | cnxn: connection to SQL database. |
| 2072 | issue_id: int global ID of the issue. |
| 2073 | |
| 2074 | Returns: |
| 2075 | A list of the IssueComment protocol buffers for the description |
| 2076 | and comments on this issue. |
| 2077 | """ |
| 2078 | comments = self.GetComments(cnxn, issue_id=[issue_id]) |
| 2079 | for i, comment in enumerate(comments): |
| 2080 | comment.sequence = i |
| 2081 | |
| 2082 | return comments |
| 2083 | |
| 2084 | |
| 2085 | def GetCommentsByID(self, cnxn, comment_ids, sequences, use_cache=True, |
| 2086 | shard_id=None): |
| 2087 | """Return all IssueComment PBs by comment ids. |
| 2088 | |
| 2089 | Args: |
| 2090 | cnxn: connection to SQL database. |
| 2091 | comment_ids: a list of comment ids. |
| 2092 | sequences: sequence of the comments. |
| 2093 | use_cache: optional boolean to enable the cache. |
| 2094 | shard_id: optional int shard_id to limit retrieval. |
| 2095 | |
| 2096 | Returns: |
| 2097 | A list of the IssueComment protocol buffers for comment_ids. |
| 2098 | """ |
| 2099 | # Try loading issue comments from a random shard to reduce load on |
| 2100 | # primary DB. |
| 2101 | if shard_id is None: |
| 2102 | shard_id = sql.RandomShardID() |
| 2103 | |
| 2104 | comment_dict, _missed_comments = self.comment_2lc.GetAll(cnxn, comment_ids, |
| 2105 | use_cache=use_cache, shard_id=shard_id) |
| 2106 | |
| 2107 | comments = sorted(list(comment_dict.values()), key=lambda x: x.timestamp) |
| 2108 | |
| 2109 | for i in range(len(comment_ids)): |
| 2110 | comments[i].sequence = sequences[i] |
| 2111 | |
| 2112 | return comments |
| 2113 | |
| 2114 | # TODO(jrobbins): remove this method because it is too slow when an issue |
| 2115 | # has a huge number of comments. |
| 2116 | def GetCommentsForIssues(self, cnxn, issue_ids, content_only=False): |
| 2117 | """Return all IssueComment PBs for each issue ID in the given list. |
| 2118 | |
| 2119 | Args: |
| 2120 | cnxn: connection to SQL database. |
| 2121 | issue_ids: list of integer global issue IDs. |
| 2122 | content_only: optional boolean, set true for faster loading of |
| 2123 | comment content without attachments and amendments. |
| 2124 | |
| 2125 | Returns: |
| 2126 | Dict {issue_id: [IssueComment, ...]} with IssueComment protocol |
| 2127 | buffers for the description and comments on each issue. |
| 2128 | """ |
| 2129 | comments = self.GetComments( |
| 2130 | cnxn, issue_id=issue_ids, content_only=content_only) |
| 2131 | |
| 2132 | comments_dict = collections.defaultdict(list) |
| 2133 | for comment in comments: |
| 2134 | comment.sequence = len(comments_dict[comment.issue_id]) |
| 2135 | comments_dict[comment.issue_id].append(comment) |
| 2136 | |
| 2137 | return comments_dict |
| 2138 | |
| 2139 | def InsertComment(self, cnxn, comment, commit=True): |
| 2140 | """Store the given issue comment in SQL. |
| 2141 | |
| 2142 | Args: |
| 2143 | cnxn: connection to SQL database. |
| 2144 | comment: IssueComment PB to insert into the database. |
| 2145 | commit: set to False to avoid doing the commit for now. |
| 2146 | """ |
| 2147 | commentcontent_id = self.commentcontent_tbl.InsertRow( |
| 2148 | cnxn, content=comment.content, |
| 2149 | inbound_message=comment.inbound_message, commit=False) |
| 2150 | comment_id = self.comment_tbl.InsertRow( |
| 2151 | cnxn, issue_id=comment.issue_id, created=comment.timestamp, |
| 2152 | project_id=comment.project_id, |
| 2153 | commenter_id=comment.user_id, |
| 2154 | deleted_by=comment.deleted_by or None, |
| 2155 | is_spam=comment.is_spam, is_description=comment.is_description, |
| 2156 | commentcontent_id=commentcontent_id, |
| 2157 | commit=False) |
| 2158 | comment.id = comment_id |
| 2159 | if comment.importer_id: |
| 2160 | self.commentimporter_tbl.InsertRow( |
| 2161 | cnxn, comment_id=comment_id, importer_id=comment.importer_id) |
| 2162 | |
| 2163 | amendment_rows = [] |
| 2164 | for amendment in comment.amendments: |
| 2165 | field_enum = str(amendment.field).lower() |
| 2166 | if (amendment.get_assigned_value('newvalue') is not None and |
| 2167 | not amendment.added_user_ids and not amendment.removed_user_ids): |
| 2168 | amendment_rows.append(( |
| 2169 | comment.issue_id, comment_id, field_enum, |
| 2170 | amendment.oldvalue, amendment.newvalue, |
| 2171 | None, None, amendment.custom_field_name)) |
| 2172 | for added_user_id in amendment.added_user_ids: |
| 2173 | amendment_rows.append(( |
| 2174 | comment.issue_id, comment_id, field_enum, None, None, |
| 2175 | added_user_id, None, amendment.custom_field_name)) |
| 2176 | for removed_user_id in amendment.removed_user_ids: |
| 2177 | amendment_rows.append(( |
| 2178 | comment.issue_id, comment_id, field_enum, None, None, |
| 2179 | None, removed_user_id, amendment.custom_field_name)) |
| 2180 | # ISSUEUPDATE_COLS[1:] to skip id column. |
| 2181 | self.issueupdate_tbl.InsertRows( |
| 2182 | cnxn, ISSUEUPDATE_COLS[1:], amendment_rows, commit=False) |
| 2183 | |
| 2184 | attachment_rows = [] |
| 2185 | for attach in comment.attachments: |
| 2186 | attachment_rows.append([ |
| 2187 | comment.issue_id, comment.id, attach.filename, attach.filesize, |
| 2188 | attach.mimetype, attach.deleted, attach.gcs_object_id]) |
| 2189 | self.attachment_tbl.InsertRows( |
| 2190 | cnxn, ATTACHMENT_COLS[1:], attachment_rows, commit=False) |
| 2191 | |
| 2192 | if comment.approval_id: |
| 2193 | self.issueapproval2comment_tbl.InsertRows( |
| 2194 | cnxn, ISSUEAPPROVAL2COMMENT_COLS, |
| 2195 | [(comment.approval_id, comment_id)], commit=False) |
| 2196 | |
| 2197 | if commit: |
| 2198 | cnxn.Commit() |
| 2199 | |
| 2200 | def _UpdateComment(self, cnxn, comment, update_cols=None): |
| 2201 | """Update the given issue comment in SQL. |
| 2202 | |
| 2203 | Args: |
| 2204 | cnxn: connection to SQL database. |
| 2205 | comment: IssueComment PB to update in the database. |
| 2206 | update_cols: optional list of just the field names to update. |
| 2207 | """ |
| 2208 | delta = { |
| 2209 | 'commenter_id': comment.user_id, |
| 2210 | 'deleted_by': comment.deleted_by or None, |
| 2211 | 'is_spam': comment.is_spam, |
| 2212 | } |
| 2213 | if update_cols is not None: |
| 2214 | delta = {key: val for key, val in delta.items() |
| 2215 | if key in update_cols} |
| 2216 | |
| 2217 | self.comment_tbl.Update(cnxn, delta, id=comment.id) |
| 2218 | self.comment_2lc.InvalidateKeys(cnxn, [comment.id]) |
| 2219 | |
| 2220 | def _MakeIssueComment( |
| 2221 | self, project_id, user_id, content, inbound_message=None, |
| 2222 | amendments=None, attachments=None, kept_attachments=None, timestamp=None, |
| 2223 | is_spam=False, is_description=False, approval_id=None, importer_id=None): |
| 2224 | """Create in IssueComment protocol buffer in RAM. |
| 2225 | |
| 2226 | Args: |
| 2227 | project_id: Project with the issue. |
| 2228 | user_id: the user ID of the user who entered the comment. |
| 2229 | content: string body of the comment. |
| 2230 | inbound_message: optional string full text of an email that |
| 2231 | caused this comment to be added. |
| 2232 | amendments: list of Amendment PBs describing the |
| 2233 | metadata changes that the user made along w/ comment. |
| 2234 | attachments: [(filename, contents, mimetype),...] attachments uploaded at |
| 2235 | the time the comment was made. |
| 2236 | kept_attachments: list of Attachment PBs for attachments kept from |
| 2237 | previous descriptions, if the comment is a description |
| 2238 | timestamp: time at which the comment was made, defaults to now. |
| 2239 | is_spam: True if the comment was classified as spam. |
| 2240 | is_description: True if the comment is a description for the issue. |
| 2241 | approval_id: id, if any, of the APPROVAL_TYPE FieldDef this comment |
| 2242 | belongs to. |
| 2243 | importer_id: optional User ID of script that imported the comment on |
| 2244 | behalf of a user. |
| 2245 | |
| 2246 | Returns: |
| 2247 | The new IssueComment protocol buffer. |
| 2248 | |
| 2249 | The content may have some markup done during input processing. |
| 2250 | |
| 2251 | Any attachments are immediately stored. |
| 2252 | """ |
| 2253 | comment = tracker_pb2.IssueComment() |
| 2254 | comment.project_id = project_id |
| 2255 | comment.user_id = user_id |
| 2256 | comment.content = content or '' |
| 2257 | comment.is_spam = is_spam |
| 2258 | comment.is_description = is_description |
| 2259 | if not timestamp: |
| 2260 | timestamp = int(time.time()) |
| 2261 | comment.timestamp = int(timestamp) |
| 2262 | if inbound_message: |
| 2263 | comment.inbound_message = inbound_message |
| 2264 | if amendments: |
| 2265 | logging.info('amendments is %r', amendments) |
| 2266 | comment.amendments.extend(amendments) |
| 2267 | if approval_id: |
| 2268 | comment.approval_id = approval_id |
| 2269 | |
| 2270 | if attachments: |
| 2271 | for filename, body, mimetype in attachments: |
| 2272 | gcs_object_id = gcs_helpers.StoreObjectInGCS( |
| 2273 | body, mimetype, project_id, filename=filename) |
| 2274 | attach = tracker_pb2.Attachment() |
| 2275 | # attachment id is determined later by the SQL DB. |
| 2276 | attach.filename = filename |
| 2277 | attach.filesize = len(body) |
| 2278 | attach.mimetype = mimetype |
| 2279 | attach.gcs_object_id = gcs_object_id |
| 2280 | comment.attachments.extend([attach]) |
| 2281 | logging.info("Save attachment with object_id: %s" % gcs_object_id) |
| 2282 | |
| 2283 | if kept_attachments: |
| 2284 | for kept_attach in kept_attachments: |
| 2285 | (filename, filesize, mimetype, deleted, |
| 2286 | gcs_object_id) = kept_attach[3:] |
| 2287 | new_attach = tracker_pb2.Attachment( |
| 2288 | filename=filename, filesize=filesize, mimetype=mimetype, |
| 2289 | deleted=bool(deleted), gcs_object_id=gcs_object_id) |
| 2290 | comment.attachments.append(new_attach) |
| 2291 | logging.info("Copy attachment with object_id: %s" % gcs_object_id) |
| 2292 | |
| 2293 | if importer_id: |
| 2294 | comment.importer_id = importer_id |
| 2295 | |
| 2296 | return comment |
| 2297 | |
| 2298 | def CreateIssueComment( |
| 2299 | self, cnxn, issue, user_id, content, inbound_message=None, |
| 2300 | amendments=None, attachments=None, kept_attachments=None, timestamp=None, |
| 2301 | is_spam=False, is_description=False, approval_id=None, commit=True, |
| 2302 | importer_id=None): |
| 2303 | """Create and store a new comment on the specified issue. |
| 2304 | |
| 2305 | Args: |
| 2306 | cnxn: connection to SQL database. |
| 2307 | issue: the issue on which to add the comment, must be loaded from |
| 2308 | database with use_cache=False so that assume_stale == False. |
| 2309 | user_id: the user ID of the user who entered the comment. |
| 2310 | content: string body of the comment. |
| 2311 | inbound_message: optional string full text of an email that caused |
| 2312 | this comment to be added. |
| 2313 | amendments: list of Amendment PBs describing the |
| 2314 | metadata changes that the user made along w/ comment. |
| 2315 | attachments: [(filename, contents, mimetype),...] attachments uploaded at |
| 2316 | the time the comment was made. |
| 2317 | kept_attachments: list of attachment ids for attachments kept from |
| 2318 | previous descriptions, if the comment is an update to the description |
| 2319 | timestamp: time at which the comment was made, defaults to now. |
| 2320 | is_spam: True if the comment is classified as spam. |
| 2321 | is_description: True if the comment is a description for the issue. |
| 2322 | approval_id: id, if any, of the APPROVAL_TYPE FieldDef this comment |
| 2323 | belongs to. |
| 2324 | commit: set to False to not commit to DB yet. |
| 2325 | importer_id: user ID of an API client that is importing issues. |
| 2326 | |
| 2327 | Returns: |
| 2328 | The new IssueComment protocol buffer. |
| 2329 | |
| 2330 | Note that we assume that the content is safe to echo out |
| 2331 | again. The content may have some markup done during input |
| 2332 | processing. |
| 2333 | """ |
| 2334 | if is_description: |
| 2335 | kept_attachments = self.GetAttachmentsByID(cnxn, kept_attachments) |
| 2336 | else: |
| 2337 | kept_attachments = [] |
| 2338 | |
| 2339 | comment = self._MakeIssueComment( |
| 2340 | issue.project_id, user_id, content, amendments=amendments, |
| 2341 | inbound_message=inbound_message, attachments=attachments, |
| 2342 | timestamp=timestamp, is_spam=is_spam, is_description=is_description, |
| 2343 | kept_attachments=kept_attachments, approval_id=approval_id, |
| 2344 | importer_id=importer_id) |
| 2345 | comment.issue_id = issue.issue_id |
| 2346 | |
| 2347 | if attachments or kept_attachments: |
| 2348 | issue.attachment_count = ( |
| 2349 | issue.attachment_count + len(attachments) + len(kept_attachments)) |
| 2350 | self.UpdateIssue(cnxn, issue, update_cols=['attachment_count']) |
| 2351 | |
| 2352 | self.comment_creations.increment() |
| 2353 | self.InsertComment(cnxn, comment, commit=commit) |
| 2354 | |
| 2355 | return comment |
| 2356 | |
| 2357 | def SoftDeleteComment( |
| 2358 | self, cnxn, issue, issue_comment, deleted_by_user_id, |
| 2359 | user_service, delete=True, reindex=False, is_spam=False): |
| 2360 | """Mark comment as un/deleted, which shows/hides it from average users.""" |
| 2361 | # Update number of attachments |
| 2362 | attachments = 0 |
| 2363 | if issue_comment.attachments: |
| 2364 | for attachment in issue_comment.attachments: |
| 2365 | if not attachment.deleted: |
| 2366 | attachments += 1 |
| 2367 | |
| 2368 | # Delete only if it's not in deleted state |
| 2369 | if delete: |
| 2370 | if not issue_comment.deleted_by: |
| 2371 | issue_comment.deleted_by = deleted_by_user_id |
| 2372 | issue.attachment_count = issue.attachment_count - attachments |
| 2373 | |
| 2374 | # Undelete only if it's in deleted state |
| 2375 | elif issue_comment.deleted_by: |
| 2376 | issue_comment.deleted_by = 0 |
| 2377 | issue.attachment_count = issue.attachment_count + attachments |
| 2378 | |
| 2379 | issue_comment.is_spam = is_spam |
| 2380 | self._UpdateComment( |
| 2381 | cnxn, issue_comment, update_cols=['deleted_by', 'is_spam']) |
| 2382 | self.UpdateIssue(cnxn, issue, update_cols=['attachment_count']) |
| 2383 | |
| 2384 | # Reindex the issue to take the comment deletion/undeletion into account. |
| 2385 | if reindex: |
| 2386 | tracker_fulltext.IndexIssues( |
| 2387 | cnxn, [issue], user_service, self, self._config_service) |
| 2388 | else: |
| 2389 | self.EnqueueIssuesForIndexing(cnxn, [issue.issue_id]) |
| 2390 | |
| 2391 | ### Approvals |
| 2392 | |
| 2393 | def GetIssueApproval(self, cnxn, issue_id, approval_id, use_cache=True): |
| 2394 | """Retrieve the specified approval for the specified issue.""" |
| 2395 | issue = self.GetIssue(cnxn, issue_id, use_cache=use_cache) |
| 2396 | approval = tracker_bizobj.FindApprovalValueByID( |
| 2397 | approval_id, issue.approval_values) |
| 2398 | if approval: |
| 2399 | return issue, approval |
| 2400 | raise exceptions.NoSuchIssueApprovalException() |
| 2401 | |
| 2402 | def DeltaUpdateIssueApproval( |
| 2403 | self, cnxn, modifier_id, config, issue, approval, approval_delta, |
| 2404 | comment_content=None, is_description=False, attachments=None, |
| 2405 | commit=True, kept_attachments=None): |
| 2406 | """Update the issue's approval in the database.""" |
| 2407 | amendments = [] |
| 2408 | |
| 2409 | # Update status in RAM and DB and create status amendment. |
| 2410 | if approval_delta.status: |
| 2411 | approval.status = approval_delta.status |
| 2412 | approval.set_on = approval_delta.set_on or int(time.time()) |
| 2413 | approval.setter_id = modifier_id |
| 2414 | status_amendment = tracker_bizobj.MakeApprovalStatusAmendment( |
| 2415 | approval_delta.status) |
| 2416 | amendments.append(status_amendment) |
| 2417 | |
| 2418 | self._UpdateIssueApprovalStatus( |
| 2419 | cnxn, issue.issue_id, approval.approval_id, approval.status, |
| 2420 | approval.setter_id, approval.set_on) |
| 2421 | |
| 2422 | # Update approver_ids in RAM and DB and create approver amendment. |
| 2423 | approvers_add = [approver for approver in approval_delta.approver_ids_add |
| 2424 | if approver not in approval.approver_ids] |
| 2425 | approvers_remove = [approver for approver in |
| 2426 | approval_delta.approver_ids_remove |
| 2427 | if approver in approval.approver_ids] |
| 2428 | if approvers_add or approvers_remove: |
| 2429 | approver_ids = [approver for approver in |
| 2430 | list(approval.approver_ids) + approvers_add |
| 2431 | if approver not in approvers_remove] |
| 2432 | approval.approver_ids = approver_ids |
| 2433 | approvers_amendment = tracker_bizobj.MakeApprovalApproversAmendment( |
| 2434 | approvers_add, approvers_remove) |
| 2435 | amendments.append(approvers_amendment) |
| 2436 | |
| 2437 | self._UpdateIssueApprovalApprovers( |
| 2438 | cnxn, issue.issue_id, approval.approval_id, approver_ids) |
| 2439 | |
| 2440 | fv_amendments = tracker_bizobj.ApplyFieldValueChanges( |
| 2441 | issue, config, approval_delta.subfield_vals_add, |
| 2442 | approval_delta.subfield_vals_remove, approval_delta.subfields_clear) |
| 2443 | amendments.extend(fv_amendments) |
| 2444 | if fv_amendments: |
| 2445 | self._UpdateIssuesFields(cnxn, [issue], commit=False) |
| 2446 | |
| 2447 | label_amendment = tracker_bizobj.ApplyLabelChanges( |
| 2448 | issue, config, approval_delta.labels_add, approval_delta.labels_remove) |
| 2449 | if label_amendment: |
| 2450 | amendments.append(label_amendment) |
| 2451 | self._UpdateIssuesLabels(cnxn, [issue], commit=False) |
| 2452 | |
| 2453 | comment_pb = self.CreateIssueComment( |
| 2454 | cnxn, issue, modifier_id, comment_content, amendments=amendments, |
| 2455 | approval_id=approval.approval_id, is_description=is_description, |
| 2456 | attachments=attachments, commit=False, |
| 2457 | kept_attachments=kept_attachments) |
| 2458 | |
| 2459 | if commit: |
| 2460 | cnxn.Commit() |
| 2461 | self.issue_2lc.InvalidateKeys(cnxn, [issue.issue_id]) |
| 2462 | |
| 2463 | return comment_pb |
| 2464 | |
| 2465 | def _UpdateIssueApprovalStatus( |
| 2466 | self, cnxn, issue_id, approval_id, status, setter_id, set_on): |
| 2467 | """Update the approvalvalue for the given issue_id's issue.""" |
| 2468 | set_on = set_on or int(time.time()) |
| 2469 | delta = { |
| 2470 | 'status': status.name.lower(), |
| 2471 | 'setter_id': setter_id, |
| 2472 | 'set_on': set_on, |
| 2473 | } |
| 2474 | self.issue2approvalvalue_tbl.Update( |
| 2475 | cnxn, delta, approval_id=approval_id, issue_id=issue_id, |
| 2476 | commit=False) |
| 2477 | |
| 2478 | def _UpdateIssueApprovalApprovers( |
| 2479 | self, cnxn, issue_id, approval_id, approver_ids): |
| 2480 | """Update the list of approvers allowed to approve an issue's approval.""" |
| 2481 | self.issueapproval2approver_tbl.Delete( |
| 2482 | cnxn, issue_id=issue_id, approval_id=approval_id, commit=False) |
| 2483 | self.issueapproval2approver_tbl.InsertRows( |
| 2484 | cnxn, ISSUEAPPROVAL2APPROVER_COLS, [(approval_id, approver_id, issue_id) |
| 2485 | for approver_id in approver_ids], |
| 2486 | commit=False) |
| 2487 | |
| 2488 | ### Attachments |
| 2489 | |
| 2490 | def GetAttachmentAndContext(self, cnxn, attachment_id): |
| 2491 | """Load a IssueAttachment from database, and its comment ID and IID. |
| 2492 | |
| 2493 | Args: |
| 2494 | cnxn: connection to SQL database. |
| 2495 | attachment_id: long integer unique ID of desired issue attachment. |
| 2496 | |
| 2497 | Returns: |
| 2498 | An Attachment protocol buffer that contains metadata about the attached |
| 2499 | file, or None if it doesn't exist. Also, the comment ID and issue IID |
| 2500 | of the comment and issue that contain this attachment. |
| 2501 | |
| 2502 | Raises: |
| 2503 | NoSuchAttachmentException: the attachment was not found. |
| 2504 | """ |
| 2505 | if attachment_id is None: |
| 2506 | raise exceptions.NoSuchAttachmentException() |
| 2507 | |
| 2508 | attachment_row = self.attachment_tbl.SelectRow( |
| 2509 | cnxn, cols=ATTACHMENT_COLS, id=attachment_id) |
| 2510 | if attachment_row: |
| 2511 | (attach_id, issue_id, comment_id, filename, filesize, mimetype, |
| 2512 | deleted, gcs_object_id) = attachment_row |
| 2513 | if not deleted: |
| 2514 | attachment = tracker_pb2.Attachment( |
| 2515 | attachment_id=attach_id, filename=filename, filesize=filesize, |
| 2516 | mimetype=mimetype, deleted=bool(deleted), |
| 2517 | gcs_object_id=gcs_object_id) |
| 2518 | return attachment, comment_id, issue_id |
| 2519 | |
| 2520 | raise exceptions.NoSuchAttachmentException() |
| 2521 | |
| 2522 | def GetAttachmentsByID(self, cnxn, attachment_ids): |
| 2523 | """Return all Attachment PBs by attachment ids. |
| 2524 | |
| 2525 | Args: |
| 2526 | cnxn: connection to SQL database. |
| 2527 | attachment_ids: a list of comment ids. |
| 2528 | |
| 2529 | Returns: |
| 2530 | A list of the Attachment protocol buffers for the attachments with |
| 2531 | these ids. |
| 2532 | """ |
| 2533 | attachment_rows = self.attachment_tbl.Select( |
| 2534 | cnxn, cols=ATTACHMENT_COLS, id=attachment_ids) |
| 2535 | |
| 2536 | return attachment_rows |
| 2537 | |
| 2538 | def _UpdateAttachment(self, cnxn, comment, attach, update_cols=None): |
| 2539 | """Update attachment metadata in the DB. |
| 2540 | |
| 2541 | Args: |
| 2542 | cnxn: connection to SQL database. |
| 2543 | comment: IssueComment PB to invalidate in the cache. |
| 2544 | attach: IssueAttachment PB to update in the DB. |
| 2545 | update_cols: optional list of just the field names to update. |
| 2546 | """ |
| 2547 | delta = { |
| 2548 | 'filename': attach.filename, |
| 2549 | 'filesize': attach.filesize, |
| 2550 | 'mimetype': attach.mimetype, |
| 2551 | 'deleted': bool(attach.deleted), |
| 2552 | } |
| 2553 | if update_cols is not None: |
| 2554 | delta = {key: val for key, val in delta.items() |
| 2555 | if key in update_cols} |
| 2556 | |
| 2557 | self.attachment_tbl.Update(cnxn, delta, id=attach.attachment_id) |
| 2558 | self.comment_2lc.InvalidateKeys(cnxn, [comment.id]) |
| 2559 | |
| 2560 | def SoftDeleteAttachment( |
| 2561 | self, cnxn, issue, issue_comment, attach_id, user_service, delete=True, |
| 2562 | index_now=False): |
| 2563 | """Mark attachment as un/deleted, which shows/hides it from avg users.""" |
| 2564 | attachment = None |
| 2565 | for attach in issue_comment.attachments: |
| 2566 | if attach.attachment_id == attach_id: |
| 2567 | attachment = attach |
| 2568 | |
| 2569 | if not attachment: |
| 2570 | logging.warning( |
| 2571 | 'Tried to (un)delete non-existent attachment #%s in project ' |
| 2572 | '%s issue %s', attach_id, issue.project_id, issue.local_id) |
| 2573 | return |
| 2574 | |
| 2575 | if not issue_comment.deleted_by: |
| 2576 | # Decrement attachment count only if it's not in deleted state |
| 2577 | if delete: |
| 2578 | if not attachment.deleted: |
| 2579 | issue.attachment_count = issue.attachment_count - 1 |
| 2580 | |
| 2581 | # Increment attachment count only if it's in deleted state |
| 2582 | elif attachment.deleted: |
| 2583 | issue.attachment_count = issue.attachment_count + 1 |
| 2584 | |
| 2585 | logging.info('attachment.deleted was %s', attachment.deleted) |
| 2586 | |
| 2587 | attachment.deleted = delete |
| 2588 | |
| 2589 | logging.info('attachment.deleted is %s', attachment.deleted) |
| 2590 | |
| 2591 | self._UpdateAttachment( |
| 2592 | cnxn, issue_comment, attachment, update_cols=['deleted']) |
| 2593 | self.UpdateIssue(cnxn, issue, update_cols=['attachment_count']) |
| 2594 | |
| 2595 | if index_now: |
| 2596 | tracker_fulltext.IndexIssues( |
| 2597 | cnxn, [issue], user_service, self, self._config_service) |
| 2598 | else: |
| 2599 | self.EnqueueIssuesForIndexing(cnxn, [issue.issue_id]) |
| 2600 | |
| 2601 | ### Reindex queue |
| 2602 | |
| 2603 | def EnqueueIssuesForIndexing(self, cnxn, issue_ids, commit=True): |
| 2604 | # type: (MonorailConnection, Collection[int], Optional[bool]) -> None |
| 2605 | """Add the given issue IDs to the ReindexQueue table.""" |
| 2606 | reindex_rows = [(issue_id,) for issue_id in issue_ids] |
| 2607 | self.reindexqueue_tbl.InsertRows( |
| 2608 | cnxn, ['issue_id'], reindex_rows, ignore=True, commit=commit) |
| 2609 | |
| 2610 | def ReindexIssues(self, cnxn, num_to_reindex, user_service): |
| 2611 | """Reindex some issues specified in the IndexQueue table.""" |
| 2612 | rows = self.reindexqueue_tbl.Select( |
| 2613 | cnxn, order_by=[('created', [])], limit=num_to_reindex) |
| 2614 | issue_ids = [row[0] for row in rows] |
| 2615 | |
| 2616 | if issue_ids: |
| 2617 | issues = self.GetIssues(cnxn, issue_ids) |
| 2618 | tracker_fulltext.IndexIssues( |
| 2619 | cnxn, issues, user_service, self, self._config_service) |
| 2620 | self.reindexqueue_tbl.Delete(cnxn, issue_id=issue_ids) |
| 2621 | |
| 2622 | return len(issue_ids) |
| 2623 | |
| 2624 | ### Search functions |
| 2625 | |
| 2626 | def RunIssueQuery( |
| 2627 | self, cnxn, left_joins, where, order_by, shard_id=None, limit=None): |
| 2628 | """Run a SQL query to find matching issue IDs. |
| 2629 | |
| 2630 | Args: |
| 2631 | cnxn: connection to SQL database. |
| 2632 | left_joins: list of SQL LEFT JOIN clauses. |
| 2633 | where: list of SQL WHERE clauses. |
| 2634 | order_by: list of SQL ORDER BY clauses. |
| 2635 | shard_id: int shard ID to focus the search. |
| 2636 | limit: int maximum number of results, defaults to |
| 2637 | settings.search_limit_per_shard. |
| 2638 | |
| 2639 | Returns: |
| 2640 | (issue_ids, capped) where issue_ids is a list of the result issue IDs, |
| 2641 | and capped is True if the number of results reached the limit. |
| 2642 | """ |
| 2643 | limit = limit or settings.search_limit_per_shard |
| 2644 | where = where + [('Issue.deleted = %s', [False])] |
| 2645 | rows = self.issue_tbl.Select( |
| 2646 | cnxn, shard_id=shard_id, distinct=True, cols=['Issue.id'], |
| 2647 | left_joins=left_joins, where=where, order_by=order_by, |
| 2648 | limit=limit) |
| 2649 | issue_ids = [row[0] for row in rows] |
| 2650 | capped = len(issue_ids) >= limit |
| 2651 | return issue_ids, capped |
| 2652 | |
| 2653 | def GetIIDsByLabelIDs(self, cnxn, label_ids, project_id, shard_id): |
| 2654 | """Return a list of IIDs for issues with any of the given label IDs.""" |
| 2655 | if not label_ids: |
| 2656 | return [] |
| 2657 | where = [] |
| 2658 | if shard_id is not None: |
| 2659 | slice_term = ('shard = %s', [shard_id]) |
| 2660 | where.append(slice_term) |
| 2661 | |
| 2662 | rows = self.issue_tbl.Select( |
| 2663 | cnxn, shard_id=shard_id, cols=['id'], |
| 2664 | left_joins=[('Issue2Label ON Issue.id = Issue2Label.issue_id', [])], |
| 2665 | label_id=label_ids, project_id=project_id, where=where) |
| 2666 | return [row[0] for row in rows] |
| 2667 | |
| 2668 | def GetIIDsByParticipant(self, cnxn, user_ids, project_ids, shard_id): |
| 2669 | """Return IIDs for issues where any of the given users participate.""" |
| 2670 | iids = [] |
| 2671 | where = [] |
| 2672 | if shard_id is not None: |
| 2673 | where.append(('shard = %s', [shard_id])) |
| 2674 | if project_ids: |
| 2675 | cond_str = 'Issue.project_id IN (%s)' % sql.PlaceHolders(project_ids) |
| 2676 | where.append((cond_str, project_ids)) |
| 2677 | |
| 2678 | # TODO(jrobbins): Combine these 3 queries into one with ORs. It currently |
| 2679 | # is not the bottleneck. |
| 2680 | rows = self.issue_tbl.Select( |
| 2681 | cnxn, cols=['id'], reporter_id=user_ids, |
| 2682 | where=where, shard_id=shard_id) |
| 2683 | for row in rows: |
| 2684 | iids.append(row[0]) |
| 2685 | |
| 2686 | rows = self.issue_tbl.Select( |
| 2687 | cnxn, cols=['id'], owner_id=user_ids, |
| 2688 | where=where, shard_id=shard_id) |
| 2689 | for row in rows: |
| 2690 | iids.append(row[0]) |
| 2691 | |
| 2692 | rows = self.issue_tbl.Select( |
| 2693 | cnxn, cols=['id'], derived_owner_id=user_ids, |
| 2694 | where=where, shard_id=shard_id) |
| 2695 | for row in rows: |
| 2696 | iids.append(row[0]) |
| 2697 | |
| 2698 | rows = self.issue_tbl.Select( |
| 2699 | cnxn, cols=['id'], |
| 2700 | left_joins=[('Issue2Cc ON Issue2Cc.issue_id = Issue.id', [])], |
| 2701 | cc_id=user_ids, |
| 2702 | where=where + [('cc_id IS NOT NULL', [])], |
| 2703 | shard_id=shard_id) |
| 2704 | for row in rows: |
| 2705 | iids.append(row[0]) |
| 2706 | |
| 2707 | rows = self.issue_tbl.Select( |
| 2708 | cnxn, cols=['Issue.id'], |
| 2709 | left_joins=[ |
| 2710 | ('Issue2FieldValue ON Issue.id = Issue2FieldValue.issue_id', []), |
| 2711 | ('FieldDef ON Issue2FieldValue.field_id = FieldDef.id', [])], |
| 2712 | user_id=user_ids, grants_perm='View', |
| 2713 | where=where + [('user_id IS NOT NULL', [])], |
| 2714 | shard_id=shard_id) |
| 2715 | for row in rows: |
| 2716 | iids.append(row[0]) |
| 2717 | |
| 2718 | return iids |
| 2719 | |
| 2720 | ### Issue Dependency Rankings |
| 2721 | |
| 2722 | def SortBlockedOn(self, cnxn, issue, blocked_on_iids): |
| 2723 | """Sort blocked_on dependencies by rank and dst_issue_id. |
| 2724 | |
| 2725 | Args: |
| 2726 | cnxn: connection to SQL database. |
| 2727 | issue: the issue being blocked. |
| 2728 | blocked_on_iids: the iids of all the issue's blockers |
| 2729 | |
| 2730 | Returns: |
| 2731 | a tuple (ids, ranks), where ids is the sorted list of |
| 2732 | blocked_on_iids and ranks is the list of corresponding ranks |
| 2733 | """ |
| 2734 | rows = self.issuerelation_tbl.Select( |
| 2735 | cnxn, cols=ISSUERELATION_COLS, issue_id=issue.issue_id, |
| 2736 | dst_issue_id=blocked_on_iids, kind='blockedon', |
| 2737 | order_by=[('rank DESC', []), ('dst_issue_id', [])]) |
| 2738 | ids = [row[1] for row in rows] |
| 2739 | ids.extend([iid for iid in blocked_on_iids if iid not in ids]) |
| 2740 | ranks = [row[3] for row in rows] |
| 2741 | ranks.extend([0] * (len(blocked_on_iids) - len(ranks))) |
| 2742 | return ids, ranks |
| 2743 | |
| 2744 | def ApplyIssueRerank( |
| 2745 | self, cnxn, parent_id, relations_to_change, commit=True, invalidate=True): |
| 2746 | """Updates rankings of blocked on issue relations to new values |
| 2747 | |
| 2748 | Args: |
| 2749 | cnxn: connection to SQL database. |
| 2750 | parent_id: the global ID of the blocked issue to update |
| 2751 | relations_to_change: This should be a list of |
| 2752 | [(blocker_id, new_rank),...] of relations that need to be changed |
| 2753 | commit: set to False to skip the DB commit and do it in the caller. |
| 2754 | invalidate: set to False to leave cache invalidatation to the caller. |
| 2755 | """ |
| 2756 | blocker_ids = [blocker for (blocker, rank) in relations_to_change] |
| 2757 | self.issuerelation_tbl.Delete( |
| 2758 | cnxn, issue_id=parent_id, dst_issue_id=blocker_ids, commit=False) |
| 2759 | insert_rows = [(parent_id, blocker, 'blockedon', rank) |
| 2760 | for (blocker, rank) in relations_to_change] |
| 2761 | self.issuerelation_tbl.InsertRows( |
| 2762 | cnxn, cols=ISSUERELATION_COLS, row_values=insert_rows, commit=commit) |
| 2763 | if invalidate: |
| 2764 | self.InvalidateIIDs(cnxn, [parent_id]) |
| 2765 | |
| 2766 | # Expunge Users from Issues system. |
| 2767 | def ExpungeUsersInIssues(self, cnxn, user_ids_by_email, limit=None): |
| 2768 | """Removes all references to given users from issue DB tables. |
| 2769 | |
| 2770 | This method will not commit the operations. This method will |
| 2771 | not make changes to in-memory data. |
| 2772 | |
| 2773 | Args: |
| 2774 | cnxn: connection to SQL database. |
| 2775 | user_ids_by_email: dict of {email: user_id} of all users we want |
| 2776 | to expunge. |
| 2777 | limit: Optional, the limit for each operation. |
| 2778 | |
| 2779 | Returns: |
| 2780 | A list of issue_ids that need to be reindexed. |
| 2781 | """ |
| 2782 | commit = False |
| 2783 | user_ids = list(user_ids_by_email.values()) |
| 2784 | user_emails = list(user_ids_by_email.keys()) |
| 2785 | # Track issue_ids for issues that will have different search documents |
| 2786 | # as a result of removing users. |
| 2787 | affected_issue_ids = [] |
| 2788 | |
| 2789 | # Reassign commenter_id and delete inbound_messages. |
| 2790 | shard_id = sql.RandomShardID() |
| 2791 | comment_content_id_rows = self.comment_tbl.Select( |
| 2792 | cnxn, cols=['Comment.id', 'Comment.issue_id', 'commentcontent_id'], |
| 2793 | commenter_id=user_ids, shard_id=shard_id, limit=limit) |
| 2794 | comment_ids = [row[0] for row in comment_content_id_rows] |
| 2795 | commentcontent_ids = [row[2] for row in comment_content_id_rows] |
| 2796 | if commentcontent_ids: |
| 2797 | self.commentcontent_tbl.Update( |
| 2798 | cnxn, {'inbound_message': None}, id=commentcontent_ids, commit=commit) |
| 2799 | if comment_ids: |
| 2800 | self.comment_tbl.Update( |
| 2801 | cnxn, {'commenter_id': framework_constants.DELETED_USER_ID}, |
| 2802 | id=comment_ids, |
| 2803 | commit=commit) |
| 2804 | affected_issue_ids.extend([row[1] for row in comment_content_id_rows]) |
| 2805 | |
| 2806 | # Reassign deleted_by comments deleted_by. |
| 2807 | self.comment_tbl.Update( |
| 2808 | cnxn, |
| 2809 | {'deleted_by': framework_constants.DELETED_USER_ID}, |
| 2810 | deleted_by=user_ids, |
| 2811 | commit=commit, limit=limit) |
| 2812 | |
| 2813 | # Remove users in field values. |
| 2814 | fv_issue_id_rows = self.issue2fieldvalue_tbl.Select( |
| 2815 | cnxn, cols=['issue_id'], user_id=user_ids, limit=limit) |
| 2816 | fv_issue_ids = [row[0] for row in fv_issue_id_rows] |
| 2817 | self.issue2fieldvalue_tbl.Delete( |
| 2818 | cnxn, user_id=user_ids, limit=limit, commit=commit) |
| 2819 | affected_issue_ids.extend(fv_issue_ids) |
| 2820 | |
| 2821 | # Remove users in approval values. |
| 2822 | self.issueapproval2approver_tbl.Delete( |
| 2823 | cnxn, approver_id=user_ids, commit=commit, limit=limit) |
| 2824 | self.issue2approvalvalue_tbl.Update( |
| 2825 | cnxn, |
| 2826 | {'setter_id': framework_constants.DELETED_USER_ID}, |
| 2827 | setter_id=user_ids, |
| 2828 | commit=commit, limit=limit) |
| 2829 | |
| 2830 | # Remove users in issue Ccs. |
| 2831 | cc_issue_id_rows = self.issue2cc_tbl.Select( |
| 2832 | cnxn, cols=['issue_id'], cc_id=user_ids, limit=limit) |
| 2833 | cc_issue_ids = [row[0] for row in cc_issue_id_rows] |
| 2834 | self.issue2cc_tbl.Delete( |
| 2835 | cnxn, cc_id=user_ids, limit=limit, commit=commit) |
| 2836 | affected_issue_ids.extend(cc_issue_ids) |
| 2837 | |
| 2838 | # Remove users in issue owners. |
| 2839 | owner_issue_id_rows = self.issue_tbl.Select( |
| 2840 | cnxn, cols=['id'], owner_id=user_ids, limit=limit) |
| 2841 | owner_issue_ids = [row[0] for row in owner_issue_id_rows] |
| 2842 | if owner_issue_ids: |
| 2843 | self.issue_tbl.Update( |
| 2844 | cnxn, {'owner_id': None}, id=owner_issue_ids, commit=commit) |
| 2845 | affected_issue_ids.extend(owner_issue_ids) |
| 2846 | derived_owner_issue_id_rows = self.issue_tbl.Select( |
| 2847 | cnxn, cols=['id'], derived_owner_id=user_ids, limit=limit) |
| 2848 | derived_owner_issue_ids = [row[0] for row in derived_owner_issue_id_rows] |
| 2849 | if derived_owner_issue_ids: |
| 2850 | self.issue_tbl.Update( |
| 2851 | cnxn, {'derived_owner_id': None}, |
| 2852 | id=derived_owner_issue_ids, |
| 2853 | commit=commit) |
| 2854 | affected_issue_ids.extend(derived_owner_issue_ids) |
| 2855 | |
| 2856 | # Remove users in issue reporters. |
| 2857 | reporter_issue_id_rows = self.issue_tbl.Select( |
| 2858 | cnxn, cols=['id'], reporter_id=user_ids, limit=limit) |
| 2859 | reporter_issue_ids = [row[0] for row in reporter_issue_id_rows] |
| 2860 | if reporter_issue_ids: |
| 2861 | self.issue_tbl.Update( |
| 2862 | cnxn, {'reporter_id': framework_constants.DELETED_USER_ID}, |
| 2863 | id=reporter_issue_ids, |
| 2864 | commit=commit) |
| 2865 | affected_issue_ids.extend(reporter_issue_ids) |
| 2866 | |
| 2867 | # Note: issueupdate_tbl's and issue2notify's user_id columns do not |
| 2868 | # reference the User table. So all values need to updated here before |
| 2869 | # User rows can be deleted safely. No limit will be applied. |
| 2870 | |
| 2871 | # Remove users in issue updates. |
| 2872 | self.issueupdate_tbl.Update( |
| 2873 | cnxn, |
| 2874 | {'added_user_id': framework_constants.DELETED_USER_ID}, |
| 2875 | added_user_id=user_ids, |
| 2876 | commit=commit) |
| 2877 | self.issueupdate_tbl.Update( |
| 2878 | cnxn, |
| 2879 | {'removed_user_id': framework_constants.DELETED_USER_ID}, |
| 2880 | removed_user_id=user_ids, |
| 2881 | commit=commit) |
| 2882 | |
| 2883 | # Remove users in issue notify. |
| 2884 | self.issue2notify_tbl.Delete( |
| 2885 | cnxn, email=user_emails, commit=commit) |
| 2886 | |
| 2887 | # Remove users in issue snapshots. |
| 2888 | self.issuesnapshot_tbl.Update( |
| 2889 | cnxn, |
| 2890 | {'owner_id': framework_constants.DELETED_USER_ID}, |
| 2891 | owner_id=user_ids, |
| 2892 | commit=commit, limit=limit) |
| 2893 | self.issuesnapshot_tbl.Update( |
| 2894 | cnxn, |
| 2895 | {'reporter_id': framework_constants.DELETED_USER_ID}, |
| 2896 | reporter_id=user_ids, |
| 2897 | commit=commit, limit=limit) |
| 2898 | self.issuesnapshot2cc_tbl.Delete( |
| 2899 | cnxn, cc_id=user_ids, commit=commit, limit=limit) |
| 2900 | |
| 2901 | return list(set(affected_issue_ids)) |