Reworked comment & progress bar

This commit is contained in:
Collin R
2020-02-01 00:23:15 -08:00
parent 74771bd24d
commit 3dee25bd80
11 changed files with 464 additions and 303 deletions

View File

@@ -1,136 +1,169 @@
import configparser
import re
import praw
from . import comment, database
### Globals ###
CONFIGPATH = 'pointsbot.ini'
from . import config, database, level, reply
### Main Function ###
def run():
config = configparser.ConfigParser()
config.read(CONFIGPATH)
# Get the user flair levels in ascending order by point value
# TODO Make levels a dict instead
# TODO Could also make a Level class or namedtuple that contains more info, e.g.
# flair css or template id
# levels = [(o, config.getint('Levels', o)) for o in config.options('Levels')]
# levels.sort(key=lambda pair: pair[1])
levels = []
for opt in config.options('Levels'):
levels.append((opt.title(), config.getint('Levels', opt)))
levels.sort(key=lambda pair: pair[1])
cfg = config.load()
levels = cfg.levels
# Connect to Reddit
reddit = praw.Reddit(site_name=config['Core']['praw_site_name'])
subreddit = reddit.subreddit(config['Core']['subreddit_name'])
reddit = praw.Reddit(site_name=cfg.praw_site_name)
subreddit = reddit.subreddit(cfg.subreddit_name)
print(f'Connected to Reddit as {reddit.user.me()}')
print(f'Read-only? {reddit.read_only}')
print(f'Watching subreddit {subreddit.title}')
print(f'Is mod? {bool(subreddit.moderator(redditor=reddit.user.me()))}')
print_level(0, f'Connected to Reddit as {reddit.user.me()}')
print_level(1, f'Read-only? {reddit.read_only}')
print_level(0, f'Watching subreddit {subreddit.title}')
is_mod = bool(subreddit.moderator(redditor=reddit.user.me()))
print_level(1, f'Is mod? {is_mod}')
# TODO pass database path instead of setting global variable
db = database.Database(config['Core']['database_name'])
testpoints = [1, 3, 5, 10, 15, 30, 45, 75] + list(range(100, 551, 50))
# The pattern to look for in comments when determining whether to award a point
# solved_pat = re.compile('!solved', re.IGNORECASE)
for sub in subreddit.new():
if sub.title == 'Testing comment scenarios':
redditor = sub.author
for points in testpoints:
body = f'Solver: {redditor}\n\nTotal points after solving: {points}'
print_level(0, body)
comm = sub.reply(body)
if comm:
level_info = level.user_level_info(points, levels)
body = reply.make(redditor, points, level_info)
comm.reply(body)
else:
print_level(1, 'ERROR: Unable to comment')
break
def real_run():
cfg = config.load()
levels = cfg.levels
# Connect to Reddit
reddit = praw.Reddit(site_name=cfg.praw_site_name)
subreddit = reddit.subreddit(cfg.subreddit_name)
print_level(0, f'Connected to Reddit as {reddit.user.me()}')
print_level(1, f'Read-only? {reddit.read_only}')
print_level(0, f'Watching subreddit {subreddit.title}')
is_mod = bool(subreddit.moderator(redditor=reddit.user.me()))
print_level(1, f'Is mod? {is_mod}')
db = database.Database(cfg.database_path)
# The pattern that determines whether a post is marked as solved
# Could also just use re.IGNORECASE flag
solved_pat = re.compile('![Ss]olved')
# Monitor new comments for confirmed solutions
for comm in subreddit.stream.comments(skip_existing=True):
print('Found comment')
print(f'Comment text: "{comm.body}"')
# Passing pause_after=0 will bypass the internal exponential delay, but have
# to check if any comments are returned with each query
for comm in subreddit.stream.comments(skip_existing=True, pause_after=0):
if comm is None:
continue
print_level(0, '\nFound comment')
print_level(1, f'Comment text: "{comm.body}"')
if marks_as_solved(comm, solved_pat):
# Ensure that this is the first "!solved" comment in the thread
submission = comm.submission
# Retrieve any comments hidden by "more comments"
submission.comments.replace_more(limit=0)
# Search the flattened comments tree
is_first_solution = True
for subcomm in submission.comments.list():
if (subcomm.id != comm.id
and marks_as_solved(subcomm, solved_pat)
and subcomm.created_utc < comm.created_utc):
# There is an earlier comment for the same submission
# already marked as a solution by the OP
is_first_solution = False
break
if not is_first_solution:
if not is_first_solution(comm, solved_pat):
# Skip this "!solved" comment and wait for the next
print('This is not the first solution')
print_level(1, 'Not the first solution')
continue
print('This is the first solution')
print_level(1, 'This is the first solution')
print_solution_info(comm)
solution = comm.parent()
solver = solution.author
print(f'Adding point for {solver.name}')
solver = comm.parent().author
print_level(1, f'Adding point for {solver.name}')
db.add_point(solver)
points = db.get_points(solver)
print(f'Points for {solver.name}: {points}')
print_level(1, f'Points for {solver.name}: {points}')
# Reply to the comment containing the solution
reply_body = comment.make(solver, points, levels)
print(f'Replying with: "{reply_body}"')
solution.reply(reply_body)
level_info = level.user_level_info(points, levels)
# TODO move comment to the end and use some things, e.g. whether
# flair is set, when building comment (to avoid duplicate logic)
# Reply to the comment marking the submission as solved
reply_body = reply.make(solver, points, level_info)
# reply_body = reply.make(solver, points, levels)
print_level(1, f'Replying with: "{reply_body}"')
comm.reply(reply_body)
# Check if (non-mod) user flair should be updated to new level
for levelname, levelpoints in levels:
# If the redditor's points total is equal to one of the levels,
# that means they just reached that level
if points == levelpoints:
print('User reached new level')
if not subreddit.moderator(redditor=solver):
# TODO can also use the keyword arg css_class *or*
# flair_template_id to style the flair
print(f'Setting flair text to {levelname}')
subreddit.flair.set(solver, text=levelname)
else:
print('Don\'t update flair b/c user is mod')
# Don't need to check the rest of the levels
break
if level_info.current and level_info.current.points == points:
print_level(1, f'User reached level: {level_info.current.name}')
if not subreddit.moderator(redditor=solver):
print_level(2, 'Setting flair')
print_level(3, f'Flair text: {level_info.current.name}')
# TODO
# print_level(3, f'Flair template ID: {}')
subreddit.flair.set(solver, text=levelname)
else:
print_level(2, 'Solver is mod; don\'t alter flair')
else:
print('Not a "!solved" comment')
print_level(1, 'Not a "!solved" comment')
### Auxiliary Functions ###
### Reddit Comment Functions ###
def marks_as_solved(comment, solved_pattern):
'''Return True if not top-level comment, from OP, contains "!Solved"; False
otherwise.
'''Return True if the comment meets the criteria for marking the submission
as solved, False otherwise.
'''
#comment.refresh() # probably not needed, but the docs are a tad unclear
return (not comment.is_root
and comment.is_submitter
and not comment.parent().is_submitter
and solved_pattern.search(comment.body))
### Debugging ###
def is_first_solution(solved_comment, solved_pattern):
# Retrieve any comments hidden by "more comments"
# Passing limit=0 will replace all "more comments"
submission = solved_comment.submission
submission.comments.replace_more(limit=0)
# Search the flattened comments tree
for comment in submission.comments.list():
if (comment.id != solved_comment.id
and marks_as_solved(comment, solved_pattern)
and comment.created_utc < solved_comment.created_utc):
# There is an earlier comment for the same submission
# already marked as a solution by the OP
return False
return True
def find_solver(comment):
# TODO
pass
def make_flair(points, levels):
# TODO
pass
### Debugging & Logging ###
def print_level(num_levels, string):
print('\t' * num_levels + string)
def print_solution_info(comm):
print('Submission solved')
print('\tSolution comment:')
print(f'\t\tAuthor: {comm.parent().author.name}')
print(f'\t\tBody: {comm.parent().body}')
print('\t"Solved" comment:')
print(f'\t\tAuthor: {comm.author.name}')
print(f'\t\tBody: {comm.body}')
print_level(1, 'Submission solved')
print_level(2, 'Solution comment:')
print_level(3, f'Author: {comm.parent().author.name}')
print_level(3, f'Body: {comm.parent().body}')
print_level(2, '"Solved" comment:')
print_level(3, f'Author: {comm.author.name}')
print_level(3, f'Body: {comm.body}')