From 7d3a08128bcadd562e7eaa48644c830ed166c98c Mon Sep 17 00:00:00 2001 From: emkael Date: Tue, 31 Dec 2019 13:19:48 +0100 Subject: Migrating JFR match results to result info client model --- jfr_playoff/data/info.py | 298 ++----------------------------------- jfr_playoff/data/match/__init__.py | 9 ++ jfr_playoff/data/match/jfrdb.py | 75 ++++++++++ jfr_playoff/data/match/jfrhtml.py | 192 ++++++++++++++++++++++++ jfr_playoff/sql.py | 5 +- 5 files changed, 295 insertions(+), 284 deletions(-) (limited to 'jfr_playoff') diff --git a/jfr_playoff/data/info.py b/jfr_playoff/data/info.py index e245f82..342a878 100644 --- a/jfr_playoff/data/info.py +++ b/jfr_playoff/data/info.py @@ -1,12 +1,10 @@ +import copy import glob import inspect -import re from os.path import dirname, basename, isfile, join from urlparse import urljoin -import jfr_playoff.sql as p_sql from jfr_playoff.dto import Match, Team -from jfr_playoff.remote import RemoteUrl as p_remote from jfr_playoff.logger import PlayoffLogger @@ -176,90 +174,7 @@ class MatchInfo(ResultInfo): self.info.id, teams) return scores_fetched, teams_fetched, teams - def __get_db_teams(self, teams, fetch_scores): - row = self.database.fetch( - self.config['database'], p_sql.MATCH_RESULTS, - (self.config['table'], self.config['round'])) - for i in range(0, 2): - teams[i].name = [row[i]] - teams[i].known_teams = 1 - if fetch_scores: - teams[0].score = row[3] + row[5] - teams[1].score = row[4] + row[6] - if row[2] > 0: - teams[0].score += row[2] - else: - teams[1].score -= row[2] - PlayoffLogger.get('matchinfo').info( - 'db scores for match #%d: %s', self.info.id, teams) - return teams - - def __find_table_row(self, url): - html_content = p_remote.fetch(url) - for row in html_content.select('tr tr'): - for cell in row.select('td.t1'): - if cell.text.strip() == str(self.config['table']): - PlayoffLogger.get('matchinfo.html').debug( - 'HTML row for table %d found: %s', - self.config['table'], row) - return row - PlayoffLogger.get('matchinfo.html').debug( - 'HTML row for table %d not found', - self.config['table']) - return None - - def __get_html_teams(self, teams, fetch_score): - if self.info.link is None: - raise ValueError('link not set') - row = self.__find_table_row(self.info.link) - if row is None: - raise ValueError('table row not found') - try: - scores = [ - float(text) for text - in row.select('td.bdc')[-1].contents - if isinstance(text, unicode)] - except ValueError: - # single-segment match - try: - # running single-segment - scores = [ - float(text.strip()) for text - in row.select('td.bdcg a')[-1].contents - if isinstance(text, unicode)] - except IndexError: - try: - # static single-segment - scores = [ - float(text.strip()) for text - in row.select('td.bdc a')[-1].contents - if isinstance(text, unicode)] - except IndexError: - # toweled single-segment - scores = [0.0, 0.0] - # carry-over - carry_over = [ - float(text.strip()) if len(text.strip()) > 0 else 0.0 for text - in row.select('td.bdc')[0].contents - if isinstance(text, unicode)] - if len(carry_over) < 2: - # no carry-over, possibly no carry-over cells or empty - carry_over = [0.0, 0.0] - for i in range(0, 2): - scores[i] += carry_over[i] - team_names = [[text for text in link.contents - if isinstance(text, unicode)][0].strip(u'\xa0') - for link in row.select('a[onmouseover]')] - for i in range(0, 2): - teams[i].name = [team_names[i]] - teams[i].known_teams = 1 - teams[i].score = scores[i] - PlayoffLogger.get('matchinfo').info( - 'HTML scores for match #%d: %s', - self.info.id, teams) - return teams - - def __get_config_teams(self, teams): + def _get_config_teams(self, teams): for i in range(0, 2): match_teams = [] possible_teams = [] @@ -310,25 +225,16 @@ class MatchInfo(ResultInfo): if scores_fetched: self.info.running = int(self.config.get('running', -1)) if not teams_fetched: - try: - try: - if self.database is None: - raise KeyError('database not configured') - if 'database' not in self.config: - raise KeyError('database not configured') - self.info.teams = self.__get_db_teams( - self.info.teams, not scores_fetched) - except (IOError, TypeError, IndexError, KeyError) as e: - PlayoffLogger.get('matchinfo').warning( - 'fetching DB scores for match #%d failed: %s(%s)', - self.info.id, type(e).__name__, str(e)) - self.info.teams = self.__get_html_teams( - self.info.teams, not scores_fetched) - except (TypeError, IndexError, KeyError, IOError, ValueError) as e: + teams = self.call_client( + 'fetch_teams', None, + copy.deepcopy(self.info.teams)) + if teams is None: PlayoffLogger.get('matchinfo').warning( - 'fetching HTML scores for match #%d failed: %s(%s)', - self.info.id, type(e).__name__, str(e)) - self.info.teams = self.__get_config_teams(self.info.teams) + 'fetching teams for match #%d failed, reverting to config', + self.info.id) + self.info.teams = self._get_config_teams(self.info.teams) + else: + self.info.teams = teams for team in range(0, len(self.info.teams)): if isinstance(self.config['teams'][team], dict): self.info.teams[team].place = self.config['teams'][team].get( @@ -344,128 +250,9 @@ class MatchInfo(ResultInfo): 'predicted team list after resolving aliases: %s', self.info.teams[team].possible_name) - def __get_db_board_count(self): - towels = self.database.fetch( - self.config['database'], p_sql.TOWEL_COUNT, - (self.config['table'], self.config['round'])) - row = [0 if r is None - else r for r in - self.database.fetch( - self.config['database'], p_sql.BOARD_COUNT, - (self.config['table'], self.config['round']))] - boards_to_play = int(row[0]) - boards_played = max(int(row[1]), 0) - if boards_to_play > 0: - boards_played += int(towels[0]) - PlayoffLogger.get('matchinfo').info( - 'DB board count for match #%d: %d/%d', - self.info.id, boards_played, boards_to_play) - return boards_played, boards_to_play - - def __has_segment_link(self, cell): - links = [link for link in cell.select('a[href]') - if re.match(r'^.*\d+t\d+-\d+\.htm$', link['href'])] - return len(links) > 0 - - def __has_towel_image(self, cell): - return len(cell.select('img[alt="towel"]')) > 0 - - def __get_html_running_boards(self, cell): - return int(cell.contents[-1].strip()) - - def __get_html_segment_board_count(self, segment_url): - segment_content = p_remote.fetch(segment_url) - board_rows = [row for row in segment_content.find_all('tr') if len(row.select('td.bdcc a.zb')) > 0] - board_count = len(board_rows) - played_boards = len([ - row for row in board_rows if len( - ''.join([cell.text.strip() for cell in row.select('td.bdc')])) > 0]) - return played_boards, board_count - - def __get_finished_info(self, cell): - segment_link = cell.select('a[href]') - if len(segment_link) > 0: - segment_url = re.sub( - r'\.htm$', '.html', - urljoin(self.info.link, segment_link[0]['href'])) - try: - played_boards, board_count = self.__get_html_segment_board_count(segment_url) - PlayoffLogger.get('matchinfo').info( - 'HTML played boards count for segment: %d/%d', - played_boards, board_count) - return board_count, played_boards >= board_count - except IOError as e: - PlayoffLogger.get('matchinfo').info( - 'cannot fetch HTML played boards count for segment: %s(%s)', - self.info.id, type(e).__name__, str(e)) - return 0, False - return 0, False - - def __get_html_board_count(self): - if self.info.link is None: - raise ValueError('link not set') - row = self.__find_table_row(self.info.link) - if row is None: - raise ValueError('table row not found') - for selector in ['td.bdc', 'td.bdcg']: - cells = row.select(selector) - segments = [cell for cell in cells if self.__has_segment_link(cell)] - towels = [cell for cell in cells if self.__has_towel_image(cell)] - if len(segments) == 0: - # in single-segment match, there are no td.bdc cells with segment links - # but maybe it's a multi-segment match with towels - if len(towels) > 0: - PlayoffLogger.get('matchinfo').info( - 'HTML board count for match #%d: all towels', self.info.id) - return 1, 1 # entire match is toweled, so mark as finished - else: - # not a single-segment match, no need to look for td.bdcg cells - break - if len(segments) == 0: - raise ValueError('segments not found') - running_segments = row.select('td.bdca') - running_boards = sum([self.__get_html_running_boards(segment) for segment in running_segments]) - finished_segments = [] - boards_in_segment = None - for segment in segments: - if segment not in running_segments: - boards, is_finished = self.__get_finished_info(segment) - if is_finished: - finished_segments.append(segment) - if boards_in_segment is None and boards > 0: - boards_in_segment = boards - if 'bdcg' in segments[0]['class']: - # only a single-segment match will yield td.bdcg cells with segment scores - total_boards = boards_in_segment - else: - PlayoffLogger.get('matchinfo').info( - 'HTML board count for match #%d, found: %d finished segments, %d towels, %d boards per segment and %d boards in running segment', - self.info.id, len(finished_segments), len(towels), boards_in_segment, running_boards) - total_boards = (len(segments) + len(towels) + len(running_segments)) * boards_in_segment - played_boards = (len(towels) + len(finished_segments)) * boards_in_segment + running_boards - PlayoffLogger.get('matchinfo').info( - 'HTML board count for match #%d: %d/%d', - self.info.id, played_boards, total_boards) - return played_boards, total_boards - def _fetch_board_count(self): - boards_played = 0 - boards_to_play = 0 - try: - if self.database is None: - raise KeyError('database not configured') - boards_played, boards_to_play = self.__get_db_board_count() - except (IOError, TypeError, IndexError, KeyError) as e: - PlayoffLogger.get('matchinfo').warning( - 'fetching board count from DB for match #%d failed: %s(%s)', - self.info.id, type(e).__name__, str(e)) - try: - boards_played, boards_to_play = self.__get_html_board_count() - except (TypeError, IndexError, KeyError, IOError, ValueError) as e: - PlayoffLogger.get('matchinfo').warning( - 'fetching board count from HTML for match #%d failed: %s(%s)', - self.info.id, type(e).__name__, str(e)) - pass + boards_played, boards_to_play = self.call_client( + 'board_count', (0, 0)) if boards_played > 0: self.info.running = -1 \ if boards_played >= boards_to_play \ @@ -489,66 +276,13 @@ class MatchInfo(ResultInfo): self.info.possible_loser = self.info.teams[0].name[0] self.info.possible_winner = self.info.teams[1].name[0] - def __get_db_running_link(self, prefix, round_no): - current_segment = int( - self.database.fetch( - self.config['database'], p_sql.CURRENT_SEGMENT, ())[0]) - PlayoffLogger.get('matchinfo').info( - 'fetched running segment from DB for match #%d: %d', - self.info.id, current_segment) - return '%s%st%d-%d.html' % ( - prefix, round_no, self.config['table'], current_segment) - - def __get_html_running_link(self): - if self.info.link is None: - raise ValueError('link not set') - row = self.__find_table_row(self.info.link) - running_link = row.select('td.bdcg a[href]') - if len(running_link) == 0: - raise ValueError('running link not found') - PlayoffLogger.get('matchinfo').info( - 'fetched running link from HTML for match #%d: %s', - self.info.id, running_link) - return urljoin(self.info.link, running_link[0]['href']) - def _determine_running_link(self): if self.info.link is None: return - match_link = self.info.link - link_match = re.match(r'^(.*)runda(\d+)\.html$', self.info.link) - if link_match: - try: - if self.database is None: - raise KeyError('database not configured') - self.info.link = self.__get_db_running_link( - link_match.group(1), link_match.group(2)) - except (IOError, TypeError, IndexError, KeyError) as e: - PlayoffLogger.get('matchinfo').warning( - 'cannot determine running link from DB for match #%d: %s(%s)', - self.info.id, type(e).__name__, str(e)) - try: - self.info.link = self.__get_html_running_link() - except (TypeError, IndexError, KeyError, IOError, ValueError) as e: - PlayoffLogger.get('matchinfo').warning( - 'cannot determine running link from HTML for match #%d: %s(%s)', - self.info.id, type(e).__name__, str(e)) - if self.info.link != match_link: - # we've detected a running segment link - # we should check if the segment's uploaded live - try: - boards_played, board_count = self.__get_html_segment_board_count(re.sub('\.htm$', '.html', self.info.link)) - except IOError as e: - PlayoffLogger.get('matchinfo').warning( - 'cannot determine running link (%s) board count for match #%d: %s(%s)', - self.info.link, self.info.id, type(e).__name__, str(e)) - boards_played = 0 - if not boards_played: - PlayoffLogger.get('matchinfo').warning( - 'running link (%s) for match #%d is not live, reverting to match link (%s)', - self.info.link, self.info.id, match_link) - self.info.link = match_link + self.info.link = self.call_client('running_link', self.info.link) def set_phase_link(self, phase_link): + prev_link = self.info.link if self.info.link is None: self.info.link = phase_link else: @@ -558,7 +292,7 @@ class MatchInfo(ResultInfo): 'applying phase link %s to match #%d: %s', phase_link, self.info.id, self.info.link) # re-init result info clients - if not len(self.clients) and (self.info.link is not None): + if (prev_link != self.info.link) and (self.info.link is not None): PlayoffLogger.get('matchinfo').info( 'config link changed, re-initializing result info client list') self.config['link'] = self.info.link diff --git a/jfr_playoff/data/match/__init__.py b/jfr_playoff/data/match/__init__.py index 7db05c2..6137078 100644 --- a/jfr_playoff/data/match/__init__.py +++ b/jfr_playoff/data/match/__init__.py @@ -4,3 +4,12 @@ from jfr_playoff.data.info import ResultInfoClient class MatchInfoClient(ResultInfoClient): def get_match_link(self): raise NotImplementedError + + def fetch_teams(self, teams): + raise NotImplementedError + + def board_count(self): + raise NotImplementedError + + def running_link(self): + raise NotImplementedError diff --git a/jfr_playoff/data/match/jfrdb.py b/jfr_playoff/data/match/jfrdb.py index 7d9e067..8dc57ab 100644 --- a/jfr_playoff/data/match/jfrdb.py +++ b/jfr_playoff/data/match/jfrdb.py @@ -1,3 +1,6 @@ +import re + +import jfr_playoff.sql as p_sql from jfr_playoff.data import TournamentInfo from jfr_playoff.data.match import MatchInfoClient from jfr_playoff.logger import PlayoffLogger @@ -26,3 +29,75 @@ class JFRDbMatchInfo(MatchInfoClient): PlayoffLogger.get('match.jfrdb').info( 'match #%d link fetched: %s', self.settings['id'], link) return link + + def fetch_teams(self, teams): + row = self.database.fetch( + self.settings['database'], p_sql.MATCH_RESULTS, + (self.settings['table'], self.settings['round'])) + for i in range(0, 2): + teams[i].name = [row[i]] + teams[i].known_teams = 1 + teams[0].score = row[3] + row[5] + teams[1].score = row[4] + row[6] + if row[2] > 0: + teams[0].score += row[2] + else: + teams[1].score -= row[2] + PlayoffLogger.get('match.jfrdb').info( + 'scores fetched: %s', teams) + return teams + + def board_count(self): + towels = self.database.fetch( + self.settings['database'], p_sql.TOWEL_COUNT, + (self.settings['table'], self.settings['round'])) + row = [0 if r is None + else r for r in + self.database.fetch( + self.settings['database'], p_sql.BOARD_COUNT, + (self.settings['table'], self.settings['round']))] + boards_to_play = int(row[0]) + boards_played = max(int(row[1]), 0) + if boards_to_play > 0: + boards_played += int(towels[0]) + PlayoffLogger.get('match.jfrdb').info( + 'board count: %d/%d', boards_played, boards_to_play) + return boards_played, boards_to_play + + def running_link(self): + match_link = self.settings['link'] + link_match = re.match(r'^(.*)runda(\d+)\.html$', match_link) + if link_match: + current_segment = int( + self.database.fetch( + self.settings['database'], p_sql.CURRENT_SEGMENT, + (self.settings['round'],))[0]) + PlayoffLogger.get('match.jfrdb').info( + 'fetched running segment: %d', current_segment) + match_link = '%s%st%d-%d.html' % ( + link_match.group(1), link_match.group(2), + self.settings['table'], current_segment) + if match_link != self.settings['link']: + PlayoffLogger.get('match.jfrdb').info( + 'checking if running link %s is live', match_link) + from jfr_playoff.data.match.jfrhtml import JFRHtmlMatchInfo + client = JFRHtmlMatchInfo(self.settings) + try: + boards_played, board_count = client.segment_board_count( + re.sub('\.htm$', '.html', match_link)) + except Exception as e: + PlayoffLogger.get('match.jfrdb').info( + 'unable to determine confirm link: %s(%s)', + type(e).__name__, str(e)) + boards_played = 0 + if not boards_played: + PlayoffLogger.get('match.jfrdb').info( + 'running link is not live - reverting to match link (%s)', + self.settings['link']) + match_link = self.settings['link'] + elif boards_played == board_count: + PlayoffLogger.get('match.jfrdb').info( + 'running link is finished - reverting to match link (%s)', + self.settings['link']) + match_link = self.settings['link'] + return match_link diff --git a/jfr_playoff/data/match/jfrhtml.py b/jfr_playoff/data/match/jfrhtml.py index 40f1395..b4a8808 100644 --- a/jfr_playoff/data/match/jfrhtml.py +++ b/jfr_playoff/data/match/jfrhtml.py @@ -1,4 +1,9 @@ +import copy +import re +from urlparse import urljoin + from jfr_playoff.data.match import MatchInfoClient +from jfr_playoff.remote import RemoteUrl as p_remote from jfr_playoff.logger import PlayoffLogger @@ -18,3 +23,190 @@ class JFRHtmlMatchInfo(MatchInfoClient): 'match #%d link pre-defined: %s', self.settings['id'], self.settings['link']) return self.settings['link'] + + def _find_table_row(self, url): + html_content = p_remote.fetch(url) + for row in html_content.select('tr tr'): + for cell in row.select('td.t1'): + if cell.text.strip() == str(self.settings['table']): + PlayoffLogger.get('match.jfrhtml').debug( + 'HTML row for table %d found: %s', + self.settings['table'], row) + return row + PlayoffLogger.get('match.jfrhtml').debug( + 'HTML row for table %d not found', + self.settings['table']) + return None + + def fetch_teams(self, teams): + if self.settings['link'] is None: + raise ValueError('link not set') + row = self._find_table_row(self.settings['link']) + if row is None: + raise ValueError('table row not found') + try: + scores = [ + float(text) for text + in row.select('td.bdc')[-1].contents + if isinstance(text, unicode)] + except ValueError: + # single-segment match + try: + # running single-segment + scores = [ + float(text.strip()) for text + in row.select('td.bdcg a')[-1].contents + if isinstance(text, unicode)] + except IndexError: + try: + # static single-segment + scores = [ + float(text.strip()) for text + in row.select('td.bdc a')[-1].contents + if isinstance(text, unicode)] + except IndexError: + # toweled single-segment + scores = [0.0, 0.0] + # carry-over + carry_over = [ + float(text.strip()) if len(text.strip()) > 0 else 0.0 for text + in row.select('td.bdc')[0].contents + if isinstance(text, unicode)] + if len(carry_over) < 2: + # no carry-over, possibly no carry-over cells or empty + carry_over = [0.0, 0.0] + for i in range(0, 2): + scores[i] += carry_over[i] + team_names = [[text for text in link.contents + if isinstance(text, unicode)][0].strip(u'\xa0') + for link in row.select('a[onmouseover]')] + for i in range(0, 2): + teams[i].name = [team_names[i]] + teams[i].known_teams = 1 + teams[i].score = scores[i] + PlayoffLogger.get('match.jfrhtml').info( + 'scores fetched: %s', teams) + return teams + + + def _has_segment_link(self, cell): + links = [link for link in cell.select('a[href]') + if re.match(r'^.*\d+t\d+-\d+\.htm$', link['href'])] + return len(links) > 0 + + def _has_towel_image(self, cell): + return len(cell.select('img[alt="towel"]')) > 0 + + def _get_html_running_boards(self, cell): + return int(cell.contents[-1].strip()) + + def segment_board_count(self, segment_url): + segment_content = p_remote.fetch(segment_url) + board_rows = [ + row for row + in segment_content.find_all('tr') + if len(row.select('td.bdcc a.zb')) > 0] + board_count = len(board_rows) + played_boards = len([ + row for row + in board_rows + if len(''.join([ + cell.text.strip() + for cell in row.select('td.bdc')])) > 0]) + return played_boards, board_count + + def _get_finished_info(self, cell): + segment_link = cell.select('a[href]') + if len(segment_link) > 0: + segment_url = re.sub( + r'\.htm$', '.html', + urljoin(self.settings['link'], segment_link[0]['href'])) + try: + played_boards, board_count = \ + self.segment_board_count(segment_url) + PlayoffLogger.get('match.jfrhtml').info( + 'HTML played boards count for segment: %d/%d', + played_boards, board_count) + return board_count, played_boards >= board_count + except IOError as e: + PlayoffLogger.get('match.jfrhtml').info( + 'cannot fetch played boards count for segment: %s(%s)', + type(e).__name__, str(e)) + return 0, False + return 0, False + + def board_count(self): + row = self._find_table_row(self.settings['link']) + if row is None: + raise ValueError('table row not found') + for selector in ['td.bdc', 'td.bdcg']: + cells = row.select(selector) + segments = [cell for cell in cells if self._has_segment_link(cell)] + towels = [cell for cell in cells if self._has_towel_image(cell)] + if len(segments) == 0: + # in single-segment match, + # there are no td.bdc cells with segment links, + # but maybe it's a multi-segment match with towels + if len(towels) > 0: + PlayoffLogger.get('match.jfrhtml').info( + 'board count: all towels') + return 1, 1 # entire match is toweled, so mark as finished + else: + # not a single-segment match + # no need to look for td.bdcg cells + break + if len(segments) == 0: + raise ValueError('segments not found') + running_segments = row.select('td.bdca') + running_boards = sum([ + self._get_html_running_boards(segment) + for segment + in running_segments]) + finished_segments = [] + boards_in_segment = None + for segment in segments: + if segment not in running_segments: + boards, is_finished = self._get_finished_info(segment) + if is_finished: + finished_segments.append(segment) + if boards_in_segment is None and boards > 0: + boards_in_segment = boards + if 'bdcg' in segments[0]['class']: + # only a single-segment match will yield + # td.bdcg cells with segment scores + total_boards = boards_in_segment + else: + PlayoffLogger.get('match.jfrhtml').info( + 'board count, found: %d finished segments, %d towels, ' \ + + '%d boards per segment and %d boards in running segment', + len(finished_segments), len(towels), + boards_in_segment, running_boards) + total_boards = ( + len(segments) + len(towels) + len(running_segments)) \ + * boards_in_segment + played_boards = (len(towels) + len(finished_segments)) \ + * boards_in_segment \ + + running_boards + PlayoffLogger.get('match.jfrhtml').info( + 'board count: %d/%d', played_boards, total_boards) + return played_boards, total_boards + + def running_link(self): + row = self._find_table_row(self.settings['link']) + running_link = row.select('td.bdcg a[href]') + if len(running_link) == 0: + raise ValueError('running link not found') + PlayoffLogger.get('match.jfrhtml').info( + 'fetched running link from HTML: %s', running_link) + match_link = urljoin(self.settings['link'], running_link[0]['href']) + try: + boards_played, board_count = self.segment_board_count( + re.sub('\.htm$', '.html', match_link)) + except Exception as e: + boards_played = 0 + if not boards_played: + PlayoffLogger.get('match.jfrhtml').info( + 'running link is not live - reverting to match link (%s)', + self.settings['link']) + match_link = self.settings['link'] + return match_link diff --git a/jfr_playoff/sql.py b/jfr_playoff/sql.py index 1eac414..6903ab1 100644 --- a/jfr_playoff/sql.py +++ b/jfr_playoff/sql.py @@ -34,8 +34,9 @@ GROUP BY #db#.segments.tabl, #db#.segments.rnd ''' CURRENT_SEGMENT = ''' -SELECT LEAST(segmentsperround, segm+1) -FROM #db#.admin +SELECT MAX(segment) +FROM #db#.scores +WHERE #db#.scores.rnd = %s AND #db#.scores.contract IS NOT NULL ''' PREFIX = ''' -- cgit v1.2.3