model: comments: allow selective retrieval of inline comments

Allow retrieving inline comments of only a specified file and possibly line
number, by filtering at database level.

This will be useful when adding more code context in comment notification
emails.

kallithea/model/comment.py

@@ -1,283 +1,304 b''
 # -*- coding: utf-8 -*-
 # This program is free software: you can redistribute it and/or modify
 # it under the terms of the GNU General Public License as published by
 # the Free Software Foundation, either version 3 of the License, or
 # (at your option) any later version.
 #
 # This program is distributed in the hope that it will be useful,
 # but WITHOUT ANY WARRANTY; without even the implied warranty of
 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 # GNU General Public License for more details.
 #
 # You should have received a copy of the GNU General Public License
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 """
 kallithea.model.comment
 ~~~~~~~~~~~~~~~~~~~~~~~
 
 comments model for Kallithea
 
 This file was forked by the Kallithea project in July 2014.
 Original author and date, and relevant copyright and licensing information is below:
 :created_on: Nov 11, 2011
 :author: marcink
 :copyright: (c) 2013 RhodeCode GmbH, and others.
 :license: GPLv3, see LICENSE.md for more details.
 """
 
 import logging
 
 from tg.i18n import ugettext as _
 from collections import defaultdict
 
 from kallithea.lib.utils2 import extract_mentioned_users, safe_unicode
 from kallithea.lib import helpers as h
 from kallithea.model.db import ChangesetComment, User, \
     PullRequest, Repository
 from kallithea.model.notification import NotificationModel
 from kallithea.model.meta import Session
 
 log = logging.getLogger(__name__)
 
 
 def _list_changeset_commenters(revision):
     return (Session().query(User)
         .join(ChangesetComment.author)
         .filter(ChangesetComment.revision == revision)
         .all())
 
 def _list_pull_request_commenters(pull_request):
     return (Session().query(User)
         .join(ChangesetComment.author)
         .filter(ChangesetComment.pull_request_id == pull_request.pull_request_id)
         .all())
 
 
 class ChangesetCommentsModel(object):
 
     def _get_notification_data(self, repo, comment, author, comment_text,
                                line_no=None, revision=None, pull_request=None,
                                status_change=None, closing_pr=False):
         """
         :returns: tuple (subj,body,recipients,notification_type,email_kwargs)
         """
         # make notification
         body = comment_text  # text of the comment
         line = ''
         if line_no:
             line = _('on line %s') % line_no
 
         # changeset
         if revision:
             notification_type = NotificationModel.TYPE_CHANGESET_COMMENT
             cs = repo.scm_instance.get_changeset(revision)
             desc = cs.short_id
 
             threading = ['%s-rev-%s@%s' % (repo.repo_name, revision, h.canonical_hostname())]
             if line_no: # TODO: url to file _and_ line number
                 threading.append('%s-rev-%s-line-%s@%s' % (repo.repo_name, revision, line_no,
                                                            h.canonical_hostname()))
             comment_url = h.canonical_url('changeset_home',
                 repo_name=repo.repo_name,
                 revision=revision,
                 anchor='comment-%s' % comment.comment_id)
             subj = safe_unicode(
                 h.link_to('Re changeset: %(desc)s %(line)s' % \
                           {'desc': desc, 'line': line},
                           comment_url)
             )
             # get the current participants of this changeset
             recipients = _list_changeset_commenters(revision)
             # add changeset author if it's known locally
             cs_author = User.get_from_cs_author(cs.author)
             if not cs_author:
                 # use repo owner if we cannot extract the author correctly
                 # FIXME: just use committer name even if not a user
                 cs_author = repo.owner
             recipients.append(cs_author)
 
             email_kwargs = {
                 'status_change': status_change,
                 'cs_comment_user': author.full_name_and_username,
                 'cs_target_repo': h.canonical_url('summary_home', repo_name=repo.repo_name),
                 'cs_comment_url': comment_url,
                 'cs_url': h.canonical_url('changeset_home', repo_name=repo.repo_name, revision=revision),
                 'raw_id': revision,
                 'message': cs.message,
                 'message_short': h.shorter(cs.message, 50, firstline=True),
                 'cs_author': cs_author,
                 'repo_name': repo.repo_name,
                 'short_id': h.short_id(revision),
                 'branch': cs.branch,
                 'comment_username': author.username,
                 'threading': threading,
             }
         # pull request
         elif pull_request:
             notification_type = NotificationModel.TYPE_PULL_REQUEST_COMMENT
             desc = comment.pull_request.title
             _org_ref_type, org_ref_name, _org_rev = comment.pull_request.org_ref.split(':')
             _other_ref_type, other_ref_name, _other_rev = comment.pull_request.other_ref.split(':')
             threading = ['%s-pr-%s@%s' % (pull_request.other_repo.repo_name,
                                           pull_request.pull_request_id,
                                           h.canonical_hostname())]
             if line_no: # TODO: url to file _and_ line number
                 threading.append('%s-pr-%s-line-%s@%s' % (pull_request.other_repo.repo_name,
                                                           pull_request.pull_request_id, line_no,
                                                           h.canonical_hostname()))
             comment_url = pull_request.url(canonical=True,
                 anchor='comment-%s' % comment.comment_id)
             subj = safe_unicode(
                 h.link_to('Re pull request %(pr_nice_id)s: %(desc)s %(line)s' %
                           {'desc': desc,
                            'pr_nice_id': comment.pull_request.nice_id(),
                            'line': line},
                           comment_url)
             )
             # get the current participants of this pull request
             recipients = _list_pull_request_commenters(pull_request)
             recipients.append(pull_request.owner)
             recipients += pull_request.get_reviewer_users()
 
             # set some variables for email notification
             email_kwargs = {
                 'pr_title': pull_request.title,
                 'pr_title_short': h.shorter(pull_request.title, 50),
                 'pr_nice_id': pull_request.nice_id(),
                 'status_change': status_change,
                 'closing_pr': closing_pr,
                 'pr_comment_url': comment_url,
                 'pr_url': pull_request.url(canonical=True),
                 'pr_comment_user': author.full_name_and_username,
                 'pr_target_repo': h.canonical_url('summary_home',
                                    repo_name=pull_request.other_repo.repo_name),
                 'pr_target_branch': other_ref_name,
                 'pr_source_repo': h.canonical_url('summary_home',
                                    repo_name=pull_request.org_repo.repo_name),
                 'pr_source_branch': org_ref_name,
                 'pr_owner': pull_request.owner,
                 'pr_owner_username': pull_request.owner.username,
                 'repo_name': pull_request.other_repo.repo_name,
                 'comment_username': author.username,
                 'threading': threading,
             }
 
         return subj, body, recipients, notification_type, email_kwargs
 
     def create(self, text, repo, author, revision=None, pull_request=None,
                f_path=None, line_no=None, status_change=None, closing_pr=False,
                send_email=True):
         """
         Creates a new comment for either a changeset or a pull request.
         status_change and closing_pr is only for the optional email.
 
         Returns the created comment.
         """
         if not status_change and not text:
             log.warning('Missing text for comment, skipping...')
             return None
 
         repo = Repository.guess_instance(repo)
         author = User.guess_instance(author)
         comment = ChangesetComment()
         comment.repo = repo
         comment.author = author
         comment.text = text
         comment.f_path = f_path
         comment.line_no = line_no
 
         if revision is not None:
             comment.revision = revision
         elif pull_request is not None:
             pull_request = PullRequest.guess_instance(pull_request)
             comment.pull_request = pull_request
         else:
             raise Exception('Please specify revision or pull_request_id')
 
         Session().add(comment)
         Session().flush()
 
         if send_email:
             (subj, body, recipients, notification_type,
              email_kwargs) = self._get_notification_data(
                                 repo, comment, author,
                                 comment_text=text,
                                 line_no=line_no,
                                 revision=revision,
                                 pull_request=pull_request,
                                 status_change=status_change,
                                 closing_pr=closing_pr)
             email_kwargs['is_mention'] = False
             # create notification objects, and emails
             NotificationModel().create(
                 created_by=author, subject=subj, body=body,
                 recipients=recipients, type_=notification_type,
                 email_kwargs=email_kwargs,
             )
 
             mention_recipients = extract_mentioned_users(body).difference(recipients)
             if mention_recipients:
                 email_kwargs['is_mention'] = True
                 subj = _('[Mention]') + ' ' + subj
                 # FIXME: this subject is wrong and unused!
                 NotificationModel().create(
                     created_by=author, subject=subj, body=body,
                     recipients=mention_recipients,
                     type_=notification_type,
                     email_kwargs=email_kwargs
                 )
 
         return comment
 
     def delete(self, comment):
         comment = ChangesetComment.guess_instance(comment)
         Session().delete(comment)
 
         return comment
 
     def get_comments(self, repo_id, revision=None, pull_request=None):
         """
         Gets general comments for either revision or pull_request.
 
         Returns a list, ordered by creation date.
         """
         return self._get_comments(repo_id, revision=revision, pull_request=pull_request,
                                   inline=False)
 
-    def get_inline_comments(self, repo_id, revision=None, pull_request=None):
+    def get_inline_comments(self, repo_id, revision=None, pull_request=None,
+                f_path=None, line_no=None):
         """
         Gets inline comments for either revision or pull_request.
 
         Returns a list of tuples with file path and list of comments per line number.
         """
         comments = self._get_comments(repo_id, revision=revision, pull_request=pull_request,
-                                      inline=True)
+                                      inline=True, f_path=f_path, line_no=line_no)
 
         paths = defaultdict(lambda: defaultdict(list))
         for co in comments:
             paths[co.f_path][co.line_no].append(co)
         return paths.items()
 
-    def _get_comments(self, repo_id, revision=None, pull_request=None, inline=False):
+    def _get_comments(self, repo_id, revision=None, pull_request=None,
+                inline=False, f_path=None, line_no=None):
         """
         Gets comments for either revision or pull_request_id, either inline or general.
+        If a file path and optionally line number are given, return only the matching inline comments.
         """
+        if f_path is None and line_no is not None:
+            raise Exception("line_no only makes sense if f_path is given.")
+
+        if inline is None and f_path is not None:
+            raise Exception("f_path only makes sense for inline comments.")
+
         q = Session().query(ChangesetComment)
 
         if inline:
-            q = q.filter(ChangesetComment.line_no != None) \
-                .filter(ChangesetComment.f_path != None)
+            if f_path is not None:
+                # inline comments for a given file...
+                q = q.filter(ChangesetComment.f_path == f_path)
+                if line_no is None:
+                    # ... on any line
+                    q = q.filter(ChangesetComment.line_no != None)
+                else:
+                    # ... on specific line
+                    q = q.filter(ChangesetComment.line_no == line_no)
+            else:
+                # all inline comments
+                q = q.filter(ChangesetComment.line_no != None) \
+                    .filter(ChangesetComment.f_path != None)
         else:
+            # all general comments
             q = q.filter(ChangesetComment.line_no == None) \
                 .filter(ChangesetComment.f_path == None)
 
         if revision is not None:
             q = q.filter(ChangesetComment.revision == revision) \
                 .filter(ChangesetComment.repo_id == repo_id)
         elif pull_request is not None:
             pull_request = PullRequest.guess_instance(pull_request)
             q = q.filter(ChangesetComment.pull_request == pull_request)
         else:
             raise Exception('Please specify either revision or pull_request')
 
         return q.order_by(ChangesetComment.created_on).all()

kallithea/tests/models/test_comments.py

@@ -1,155 +1,230 b''
 from kallithea.tests.base import *
 from kallithea.model.comment import ChangesetCommentsModel
 from kallithea.model.db import Repository
 
+import pytest
 from tg.util.webtest import test_context
 
 class TestComments(TestController):
 
-    def _check_comment_count(self, repo_id, revision, expected_len_comments, expected_len_inline_comments):
+    def _check_comment_count(self, repo_id, revision,
+            expected_len_comments, expected_len_inline_comments,
+            f_path=None, line_no=None
+    ):
         comments = ChangesetCommentsModel().get_comments(repo_id,
                 revision=revision)
         assert len(comments) == expected_len_comments
         inline_comments = ChangesetCommentsModel().get_inline_comments(repo_id,
-                revision=revision)
+                revision=revision, f_path=f_path, line_no=line_no)
         assert len(inline_comments) == expected_len_inline_comments
 
         return comments, inline_comments
 
     def test_create_delete_general_comment(self):
         with test_context(self.app):
             repo_id = Repository.get_by_repo_name(HG_REPO).repo_id
             revision = '9a7b4ff9e8b40bbda72fc75f162325b9baa45cda'
 
             self._check_comment_count(repo_id, revision,
                     expected_len_comments=0, expected_len_inline_comments=0)
 
             text = u'a comment'
             new_comment = ChangesetCommentsModel().create(
                     text=text,
                     repo=HG_REPO,
                     author=TEST_USER_REGULAR_LOGIN,
                     revision=revision,
                     send_email=False)
 
             self._check_comment_count(repo_id, revision,
                     expected_len_comments=1, expected_len_inline_comments=0)
 
             ChangesetCommentsModel().delete(new_comment)
 
             self._check_comment_count(repo_id, revision,
                     expected_len_comments=0, expected_len_inline_comments=0)
 
     def test_create_delete_inline_comment(self):
         with test_context(self.app):
             repo_id = Repository.get_by_repo_name(HG_REPO).repo_id
             revision = '9a7b4ff9e8b40bbda72fc75f162325b9baa45cda'
 
             self._check_comment_count(repo_id, revision,
                     expected_len_comments=0, expected_len_inline_comments=0)
 
             text = u'an inline comment'
             f_path = u'vcs/tests/base.py'
             line_no = u'n50'
             new_comment = ChangesetCommentsModel().create(
                     text=text,
                     repo=HG_REPO,
                     author=TEST_USER_REGULAR_LOGIN,
                     revision=revision,
                     f_path=f_path,
                     line_no=line_no,
                     send_email=False)
 
             comments, inline_comments = self._check_comment_count(repo_id, revision,
                     expected_len_comments=0, expected_len_inline_comments=1)
             # inline_comments is a list of tuples (file_path, dict)
             # where the dict keys are line numbers and values are lists of comments
             assert inline_comments[0][0] == f_path
             assert len(inline_comments[0][1]) == 1
             assert line_no in inline_comments[0][1]
             assert inline_comments[0][1][line_no][0].text == text
 
             ChangesetCommentsModel().delete(new_comment)
 
             self._check_comment_count(repo_id, revision,
                     expected_len_comments=0, expected_len_inline_comments=0)
 
     def test_create_delete_multiple_inline_comments(self):
         with test_context(self.app):
             repo_id = Repository.get_by_repo_name(HG_REPO).repo_id
             revision = '9a7b4ff9e8b40bbda72fc75f162325b9baa45cda'
 
             self._check_comment_count(repo_id, revision,
                     expected_len_comments=0, expected_len_inline_comments=0)
 
             text = u'an inline comment'
             f_path = u'vcs/tests/base.py'
             line_no = u'n50'
             new_comment = ChangesetCommentsModel().create(
                     text=text,
                     repo=HG_REPO,
                     author=TEST_USER_REGULAR_LOGIN,
                     revision=revision,
                     f_path=f_path,
                     line_no=line_no,
                     send_email=False)
 
             text2 = u'another inline comment, same file'
             line_no2 = u'o41'
             new_comment2 = ChangesetCommentsModel().create(
                     text=text2,
                     repo=HG_REPO,
                     author=TEST_USER_REGULAR_LOGIN,
                     revision=revision,
                     f_path=f_path,
                     line_no=line_no2,
                     send_email=False)
 
             text3 = u'another inline comment, same file'
             f_path3 = u'vcs/tests/test_hg.py'
             line_no3 = u'n159'
             new_comment3 = ChangesetCommentsModel().create(
                     text=text3,
                     repo=HG_REPO,
                     author=TEST_USER_REGULAR_LOGIN,
                     revision=revision,
                     f_path=f_path3,
                     line_no=line_no3,
                     send_email=False)
 
             comments, inline_comments = self._check_comment_count(repo_id, revision,
                     expected_len_comments=0, expected_len_inline_comments=2)
             # inline_comments is a list of tuples (file_path, dict)
             # where the dict keys are line numbers and values are lists of comments
             assert inline_comments[1][0] == f_path
             assert len(inline_comments[1][1]) == 2
             assert inline_comments[1][1][line_no][0].text == text
             assert inline_comments[1][1][line_no2][0].text == text2
 
             assert inline_comments[0][0] == f_path3
             assert len(inline_comments[0][1]) == 1
             assert line_no3 in inline_comments[0][1]
             assert inline_comments[0][1][line_no3][0].text == text3
 
             # now delete only one comment
             ChangesetCommentsModel().delete(new_comment2)
 
             comments, inline_comments = self._check_comment_count(repo_id, revision,
                     expected_len_comments=0, expected_len_inline_comments=2)
             # inline_comments is a list of tuples (file_path, dict)
             # where the dict keys are line numbers and values are lists of comments
             assert inline_comments[1][0] == f_path
             assert len(inline_comments[1][1]) == 1
             assert inline_comments[1][1][line_no][0].text == text
 
             assert inline_comments[0][0] == f_path3
             assert len(inline_comments[0][1]) == 1
             assert line_no3 in inline_comments[0][1]
             assert inline_comments[0][1][line_no3][0].text == text3
 
             # now delete all others
             ChangesetCommentsModel().delete(new_comment)
             ChangesetCommentsModel().delete(new_comment3)
 
             self._check_comment_count(repo_id, revision,
                     expected_len_comments=0, expected_len_inline_comments=0)
+
+    def test_selective_retrieval_of_inline_comments(self):
+        with test_context(self.app):
+            repo_id = Repository.get_by_repo_name(HG_REPO).repo_id
+            revision = '9a7b4ff9e8b40bbda72fc75f162325b9baa45cda'
+
+            self._check_comment_count(repo_id, revision,
+                    expected_len_comments=0, expected_len_inline_comments=0)
+
+            text = u'an inline comment'
+            f_path = u'vcs/tests/base.py'
+            line_no = u'n50'
+            new_comment = ChangesetCommentsModel().create(
+                    text=text,
+                    repo=HG_REPO,
+                    author=TEST_USER_REGULAR_LOGIN,
+                    revision=revision,
+                    f_path=f_path,
+                    line_no=line_no,
+                    send_email=False)
+
+            text2 = u'another inline comment, same file'
+            line_no2 = u'o41'
+            new_comment2 = ChangesetCommentsModel().create(
+                    text=text2,
+                    repo=HG_REPO,
+                    author=TEST_USER_REGULAR_LOGIN,
+                    revision=revision,
+                    f_path=f_path,
+                    line_no=line_no2,
+                    send_email=False)
+
+            text3 = u'another inline comment, same file'
+            f_path3 = u'vcs/tests/test_hg.py'
+            line_no3 = u'n159'
+            new_comment3 = ChangesetCommentsModel().create(
+                    text=text3,
+                    repo=HG_REPO,
+                    author=TEST_USER_REGULAR_LOGIN,
+                    revision=revision,
+                    f_path=f_path3,
+                    line_no=line_no3,
+                    send_email=False)
+
+            # now selectively retrieve comments of one file
+            comments, inline_comments = self._check_comment_count(repo_id, revision,
+                    f_path=f_path,
+                    expected_len_comments=0, expected_len_inline_comments=1)
+            # inline_comments is a list of tuples (file_path, dict)
+            # where the dict keys are line numbers and values are lists of comments
+            assert inline_comments[0][0] == f_path
+            assert len(inline_comments[0][1]) == 2
+            assert inline_comments[0][1][line_no][0].text == text
+            assert inline_comments[0][1][line_no2][0].text == text2
+
+            # now selectively retrieve comments of one file, one line
+            comments, inline_comments = self._check_comment_count(repo_id, revision,
+                    f_path=f_path, line_no=line_no2,
+                    expected_len_comments=0, expected_len_inline_comments=1)
+            # inline_comments is a list of tuples (file_path, dict)
+            # where the dict keys are line numbers and values are lists of comments
+            assert inline_comments[0][0] == f_path
+            assert len(inline_comments[0][1]) == 1
+            assert inline_comments[0][1][line_no2][0].text == text2
+
+            # verify that retrieval based on line_no but no f_path fails
+            with pytest.raises(Exception) as excinfo:
+                self._check_comment_count(repo_id, revision,
+                        f_path=None, line_no=line_no2,
+                        expected_len_comments=0, expected_len_inline_comments=0)
+            assert 'line_no only makes sense if f_path is given' in str(excinfo.value)