def save_review(request):
# unpack lock secret, starting point, identity, text
lock_secret = request.POST.getone('lock_secret')
starting_point = int(request.POST.getone('starting_point'))
identity_name = request.POST.getone('identity_name')
identity_email = request.POST.getone('identity_email')
review_text_1 = request.POST.getone('review_text_1')
review_text_2 = request.POST.getone('review_text_2')
with repos.commit_lock:
# find and validate the lock
repo = repos.repo_from_request(request)
index = repo.index
if not repos.lock_is_valid(repo, index, 'review', starting_point, lock_secret):
raise ValueError('Invalid lock')
# save review texts
repos.save_snippet_text(repo, index, starting_point, review_text_1)
repos.save_snippet_text(repo, index, starting_point + _snippet_ms(), review_text_2)
# remove the lock
repos.remove_lock(repo, index, 'review', starting_point)
# remove the review from remaining reviews
repos.remove_review_from_remaining(repo, index, starting_point)
# commit with identity
os.environ['GIT_AUTHOR_NAME'] = identity_name
os.environ['GIT_AUTHOR_EMAIL'] = identity_email
index.commit('review: save')
# return empty indicating success
return Response('', content_type='text/plain')
def lock_review(request):
# unpack identity
identity_name = request.POST.getone('identity_name')
identity_email = request.POST.getone('identity_email')
with repos.commit_lock:
repo = repos.repo_from_request(request)
index = repo.index
# find and lock available review
starting_point, lock_secret_or_message = repos.lock_available_review(repo, index)
# if found,
if starting_point is not None:
# commit with identity
os.environ['GIT_AUTHOR_NAME'] = identity_name
os.environ['GIT_AUTHOR_EMAIL'] = identity_email
index.commit('review: lock')
# return review info and snippet texts
review_text_1 = repos.snippet_text(repo, index, starting_point)
review_text_2 = repos.snippet_text(repo, index, starting_point + _snippet_ms())
body = json.dumps({
'lock_acquired': True,
'starting_point': starting_point,
'ending_point': starting_point + (_snippet_ms() * 2),
'lock_secret': lock_secret_or_message,
'review_text_1': review_text_1,
'review_text_2': review_text_2,
})
return Response(body, content_type='application/json')
else:
# return message
body = json.dumps({
'lock_acquired': False,
'message': lock_secret_or_message,
})
return Response(body, content_type='application/json')
def snippet_info(request):
repo, commit = repos.repo_from_request(request)
starting_point = int(request.GET.getone('starting_point'))
# Return cached if found.
cache_key = 'snippet_info-{0}-{1}'.format(commit.hexsha, starting_point)
content, mtime = cache.get_cached_content(cache_key)
if content is None or request.GET.has_key('nocache'):
starting_point = int(request.GET.getone('starting_point'))
filename = '{0:016d}.txt'.format(starting_point)
contributor_list = []
mtime = None
for commit in repo.iter_commits(commit, paths=filename):
if mtime is None:
# Use the most recent modification time.
mtime = commit.authored_date
contributor = dict(author_name=commit.author.name)
if contributor not in contributor_list:
contributor_list.append(contributor)
contributor_list.reverse()
info = dict(
contributor_list=contributor_list,
)
content = json.dumps(info)
cache.cache_content(cache_key, content, mtime=mtime)
return Response(body=content, content_type='application/json', date=mtime)
def save_snippet(request):
# unpack lock secret, starting point, identity, text
lock_secret = request.POST.getone('lock_secret')
starting_point = int(request.POST.getone('starting_point'))
identity_name = request.POST.getone('identity_name')
identity_email = request.POST.getone('identity_email')
snippet_text = transcripts.normalized_text(
request.POST.getone('snippet_text').strip())
inline = request.POST.get('inline') == '1'
with repos.commit_lock:
# find and validate the lock
# Ignore commit; we want to save to master.
repo, commit = repos.repo_from_request(request, rev='master')
index = repo.index
if not repos.lock_is_valid(repo, index, 'snippet', starting_point, lock_secret):
raise ValueError('Invalid lock')
# save the snippet text
repos.save_snippet_text(repo, index, starting_point, snippet_text)
# remove the lock
repos.remove_lock(repo, index, 'snippet', starting_point)
# remove the snippet from remaining snippets
if snippet_text:
repos.remove_snippet_from_remaining(repo, index, starting_point)
# commit with identity
os.environ['GIT_AUTHOR_NAME'] = identity_name
os.environ['GIT_AUTHOR_EMAIL'] = identity_email
commit_message = 'snippet: %s, saved by %s' % (_label_from_ms(starting_point), identity_name)
if inline:
commit_message += ' (inline)'
index.commit(commit_message)
return Response('', content_type='text/plain')
def save_review(request):
# unpack lock secret, starting point, identity, text
lock_secret = request.POST.getone('lock_secret')
starting_point = int(request.POST.getone('starting_point'))
identity_name = request.POST.getone('identity_name')
identity_email = request.POST.getone('identity_email')
review_text_1 = transcripts.normalized_text(
request.POST.getone('review_text_1'))
review_text_2 = transcripts.normalized_text(
request.POST.getone('review_text_2'))
with repos.commit_lock:
# find and validate the lock
# Ignore commit; we want to save to master.
repo, commit = repos.repo_from_request(request, rev='master')
index = repo.index
if not repos.lock_is_valid(repo, index, 'review', starting_point, lock_secret):
raise ValueError('Invalid lock')
# save review texts
repos.save_snippet_text(repo, index, starting_point, review_text_1)
repos.save_snippet_text(repo, index, starting_point + _snippet_ms(), review_text_2)
# remove the lock
repos.remove_lock(repo, index, 'review', starting_point)
# remove the review from remaining reviews
repos.remove_review_from_remaining(repo, index, starting_point)
# commit with identity
os.environ['GIT_AUTHOR_NAME'] = identity_name
os.environ['GIT_AUTHOR_EMAIL'] = identity_email
commit_message = 'review: %s, saved by %s' % (_label_from_ms(starting_point), identity_name)
index.commit(commit_message)
# return empty indicating success
return Response('', content_type='text/plain')
def snippet_mp3(request):
repo, commit = repos.repo_from_request(request)
# Get information needed from settings and repository.
settings = app_settings()
full_mp3 = os.path.join(
settings['fanscribed.audio'],
'{0}.mp3'.format(request.host),
)
transcription_info, _ = repos.json_file_at_commit(
repo, 'transcription.json', commit, required=True)
duration = transcription_info['duration']
snippet_cache = settings['fanscribed.snippet_cache']
snippet_url_prefix = settings['fanscribed.snippet_url_prefix']
# Get information needed from GET params.
starting_point = int(request.GET.getone('starting_point'))
length = int(request.GET.getone('length'))
padding = int(request.GET.getone('padding'))
snippet_path = mp3.snippet_path(
full_mp3=full_mp3,
duration=duration,
output_path=snippet_cache,
starting_point=starting_point,
length=length,
padding=padding,
)
relative_path = os.path.relpath(snippet_path, snippet_cache)
snippet_url = urlparse.urljoin(snippet_url_prefix, relative_path)
raise HTTPFound(location=snippet_url)
def save_snippet(request):
# unpack lock secret, starting point, identity, text
lock_secret = request.POST.getone('lock_secret')
starting_point = int(request.POST.getone('starting_point'))
identity_name = request.POST.getone('identity_name')
identity_email = request.POST.getone('identity_email')
snippet_text = request.POST.getone('snippet_text')
inline = request.POST.get('inline') == '1'
with repos.commit_lock:
# find and validate the lock
repo = repos.repo_from_request(request)
index = repo.index
if not repos.lock_is_valid(repo, index, 'snippet', starting_point, lock_secret):
raise ValueError('Invalid lock')
# save the snippet text
repos.save_snippet_text(repo, index, starting_point, snippet_text)
# remove the lock
repos.remove_lock(repo, index, 'snippet', starting_point)
# remove the snippet from remaining snippets
repos.remove_snippet_from_remaining(repo, index, starting_point)
# commit with identity
os.environ['GIT_AUTHOR_NAME'] = identity_name
os.environ['GIT_AUTHOR_EMAIL'] = identity_email
commit_message = 'snippet: save'
if inline:
commit_message += ' (inline)'
index.commit(commit_message)
# return structure of snippet including resolved speaker names
# for potential rendering
master = repo.tree('master')
speakers_map = repos.speakers_map(master)
text = snippet_text.strip()
lines = _split_lines_and_expand_abbreviations(text, speakers_map)
return Response(json.dumps(lines), content_type='application/json')
def view(request):
repo = repos.repo_from_request(request)
master = repo.tree('master')
transcription_info = repos.transcription_info(master)
raw_snippets = {}
for obj in master:
if isinstance(obj, git.Blob):
name, ext = os.path.splitext(obj.name)
if ext == '.txt':
try:
starting_point = int(name)
except ValueError:
pass
else:
raw_snippets[starting_point] = obj.data_stream.read().decode('utf8')
# Go through all snippets, whether they've been transcribed or not.
snippets = []
speakers_map = repos.speakers_map(master)
for starting_point in range(0, transcription_info['duration'], _snippet_ms()):
text = raw_snippets.get(starting_point, '').strip()
lines = _split_lines_and_expand_abbreviations(text, speakers_map)
snippets.append((starting_point, lines))
return dict(
_standard_response(repo, master),
snippets=sorted(snippets),
)
def rss_completion(request, percentage_gap=5):
repo, commit = repos.repo_from_request(request)
# Return cached if found.
cache_key = 'rss_completion commit={0} percentage_gap={1}'.format(commit.hexsha, percentage_gap)
content, mtime = cache.get_cached_content(cache_key)
if content is None or request.GET.has_key('nocache'):
mtime = commit.authored_date
tree = commit.tree
percent_reviewed_to_report = set(xrange(percentage_gap, 100, percentage_gap))
percent_reviewed_to_report.add(100)
percent_transcribed_to_report = set(xrange(percentage_gap, 100, percentage_gap))
percent_transcribed_to_report.add(100)
transcription_info, _ = repos.json_file_at_commit(
repo, 'transcription.json', commit, required=True)
duration = transcription_info['duration']
snippet_ms = _snippet_ms()
snippets_total = duration / snippet_ms
if duration % snippet_ms:
snippets_total += 1
reviews_total = snippets_total - 1
completions = []
now_url = 'http://{host}/?src=rc'.format(**dict(
host=request.host,
))
for c in reversed(list(repo.iter_commits(commit, ['remaining_reviews.json', 'remaining_snippets.json']))):
this_url = 'http://{host}/?rev={rev}&src=rc'.format(**dict(
host=request.host,
rev=c.hexsha,
))
snippets_remaining = len(repos.get_remaining_snippets(c.tree))
snippets_completed = snippets_total - snippets_remaining
snippets_percent = snippets_completed * 100 / snippets_total
reviews_remaining = len(repos.get_remaining_reviews(c.tree))
reviews_completed = reviews_total - reviews_remaining
reviews_percent = reviews_completed * 100 / reviews_total
if snippets_percent in percent_transcribed_to_report:
completions.append((
snippets_percent, None, this_url, c.authored_date,
))
percent_transcribed_to_report.remove(snippets_percent)
elif reviews_percent in percent_reviewed_to_report:
completions.append((
None, reviews_percent, this_url, c.authored_date,
))
percent_reviewed_to_report.remove(reviews_percent)
# Report completion in chrono order.
completions.reverse()
pub_date = commit.authored_date
data = dict(
_standard_response(repo, commit),
completions=completions,
now_url=now_url,
pub_date=pub_date,
request=request,
rfc822_from_time=rfc822_from_time,
)
content = render('fanscribed:templates/rss_completion.xml.mako', data, request=request)
cache.cache_content(cache_key, content, mtime)
return Response(content, content_type='application/rss+xml', date=mtime)
def rss_basic(request, max_actions=50):
repo, commit = repos.repo_from_request(request)
# Return cached if found.
cache_key = 'rss_basic commit={0} max_actions={1}'.format(commit.hexsha, max_actions)
content, mtime = cache.get_cached_content(cache_key)
if content is None or request.GET.has_key('nocache'):
mtime = commit.authored_date
tree = commit.tree
actions = [
# dict(author=AUTHOR, date=DATE, position=POSITION, this_url=URL, now_url=URL),
]
# Starting from the request's commit, iterate backwards.
for c in repo.iter_commits(commit):
snippets_affected = set()
for filename in c.stats.files:
ms = _ms_from_snippet_filename(filename)
if ms is not None:
snippets_affected.add(ms)
if snippets_affected:
earliest_ms = min(snippets_affected)
anchor = _anchor_from_ms(earliest_ms)
position = _label_from_ms(earliest_ms)
author = c.author
date = c.authored_date
kwargs = dict(
host=request.host,
rev=c.hexsha,
anchor=anchor,
)
now_url = 'http://{host}/?src=rb#{anchor}'.format(**kwargs)
this_url = 'http://{host}/?rev={rev}&src=rb#{anchor}'.format(**kwargs)
actions.append(dict(
author=author,
date=date,
position=position,
this_url=this_url,
now_url=now_url,
))
if len(actions) >= max_actions:
break
# Report actions in chrono order.
actions.reverse()
if not actions:
pub_date = time.time()
else:
pub_date = actions[-1]['date']
data = dict(
_standard_response(repo, commit),
actions=actions,
pub_date=pub_date,
request=request,
rfc822_from_time=rfc822_from_time,
)
content = render('fanscribed:templates/rss_basic.xml.mako', data, request=request)
cache.cache_content(cache_key, content, mtime)
return Response(content, content_type='application/rss+xml', date=mtime)
def transcription_json(request):
# No rendering or processing, no need to cache.
repo, commit = repos.repo_from_request(request)
info, mtime = repos.json_file_at_commit(
repo, 'transcription.json', commit, required=True)
# Inject additional information into the info dict.
settings = app_settings()
info['snippet_ms'] = int(settings['fanscribed.snippet_seconds']) * 1000
info['snippet_padding_ms'] = int(float(settings['fanscribed.snippet_padding_seconds']) * 1000)
return Response(body=json.dumps(info), content_type='application/json')
def custom_js(request):
# No rendering or processing, no need to cache.
repo, commit = repos.repo_from_request(request)
tree = commit.tree
if 'custom.js' in tree:
mtime = repo.iter_commits(commit, 'custom.js').next().authored_date
blob = tree['custom.js']
content = blob.data_stream.read().decode('utf8')
return Response(content, date=mtime, content_type='text/javascript')
else:
# Not yet created.
raise HTTPNotFound()
def progress(request):
repo, commit = repos.repo_from_request(request)
# Return cached if found.
cache_key = 'progress-{0}'.format(commit.hexsha)
content, mtime = cache.get_cached_content(cache_key)
if content is None or request.GET.has_key('nocache'):
tree = commit.tree
info, _ = repos.json_file_at_commit(
repo, 'transcription.json', commit, required=True)
content = json.dumps(_progress_dicts(tree, info))
mtime = commit.authored_date
cache.cache_content(cache_key, content, mtime=mtime)
return Response(body=content, content_type='application/json', date=mtime)
def snippet_info(request):
starting_point = int(request.GET.getone('starting_point'))
filename = '{0:016d}.txt'.format(starting_point)
repo = repos.repo_from_request(request)
contributor_list = []
for commit in repo.iter_commits('master', paths=filename):
contributor = dict(author_name=commit.author.name)
if contributor not in contributor_list:
contributor_list.append(contributor)
contributor_list.reverse()
info = dict(
contributor_list=contributor_list,
)
return Response(body=json.dumps(info), content_type='application/json')
def save_duration(request):
registry = get_current_registry()
settings = registry.settings
init_password = settings['fanscribed.init_password'].strip()
given_password = request.POST.getone('init_password').strip()
if given_password != init_password:
response = 'Wrong password'
else:
duration = int(request.POST.getone('duration'))
bytes_total = int(request.POST.getone('bytes_total'))
identity_name = request.POST.getone('identity_name')
identity_email = request.POST.getone('identity_email')
# Update transcription info.
repo = repos.repo_from_request(request)
master = repo.tree('master')
blob = master['transcription.json']
transcription_info = json.load(blob.data_stream)
if ('duration' not in transcription_info
and 'bytes_total' not in transcription_info
):
transcription_info['duration'] = int(duration)
transcription_info['bytes_total'] = int(bytes_total)
# Save transcription info.
with repos.commit_lock:
repo.heads['master'].checkout()
index = repo.index
filename = os.path.join(repo.working_dir, 'transcription.json')
with open(filename, 'wb') as f:
json.dump(transcription_info, f, indent=4)
index.add(['transcription.json'])
# Also save remaining snippets and reviews.
locks = repos.get_locks(master)
repos.save_locks(repo, index, locks)
snippets = range(0, transcription_info['duration'], _snippet_ms())
reviews = snippets[:-1]
repos.save_remaining_snippets(repo, index, snippets)
repos.save_remaining_reviews(repo, index, reviews)
# Perform commit.
os.environ['GIT_AUTHOR_NAME'] = identity_name
os.environ['GIT_AUTHOR_EMAIL'] = identity_email
index.commit('initialize')
# Respond to client.
response = 'Committed - Duration: {0}, Bytes: {1}'.format(duration, bytes_total)
else:
response = 'Already initialized'.format(duration, bytes_total)
return Response(response, content_type='text/plain')
def lock_snippet(request):
# First make sure IP address is not banned.
message = _banned_message(request)
if message is not None:
body = json.dumps({
'lock_acquired': False,
'message': message,
})
return Response(body, content_type='application_json')
# unpack identity
identity_name = request.POST.getone('identity_name')
identity_email = request.POST.getone('identity_email')
# get snippet if specified
desired_starting_point = request.POST.get('starting_point', None)
if desired_starting_point is not None:
desired_starting_point = int(desired_starting_point)
with repos.commit_lock:
repo, commit = repos.repo_from_request(request, rev='master')
index = repo.index
# find and lock available snippet
starting_point, lock_secret_or_message = repos.lock_available_snippet(repo, index, desired_starting_point)
# if found,
if starting_point is not None:
# commit with identity
os.environ['GIT_AUTHOR_NAME'] = identity_name
os.environ['GIT_AUTHOR_EMAIL'] = identity_email
commit_message = 'snippet: %s, locked by %s' % (_label_from_ms(starting_point), identity_name)
index.commit(commit_message)
# return snippet info and text
snippet_text = repos.snippet_text(repo, index, starting_point)
body = json.dumps({
'lock_acquired': True,
'starting_point': starting_point,
'ending_point': starting_point + _snippet_ms(),
'lock_secret': lock_secret_or_message,
'snippet_text': snippet_text,
})
return Response(body, content_type='application/json')
else:
# return message
body = json.dumps({
'lock_acquired': False,
'message': lock_secret_or_message,
})
return Response(body, content_type='application/json')
def lock_review(request):
# First make sure IP address is not banned.
message = _banned_message(request)
if message is not None:
body = json.dumps({
'lock_acquired': False,
'message': message,
})
return Response(body, content_type='application_json')
# unpack identity
identity_name = request.POST.getone('identity_name')
identity_email = request.POST.getone('identity_email')
with repos.commit_lock:
# Ignore commit; we want to save to master.
repo, commit = repos.repo_from_request(request, rev='master')
index = repo.index
# find and lock available review
starting_point, lock_secret_or_message = repos.lock_available_review(repo, index)
# if found,
if starting_point is not None:
# commit with identity
os.environ['GIT_AUTHOR_NAME'] = identity_name
os.environ['GIT_AUTHOR_EMAIL'] = identity_email
commit_message = 'review: %s, locked by %s' % (_label_from_ms(starting_point), identity_name)
index.commit(commit_message)
# return review info and snippet texts
review_text_1 = repos.snippet_text(repo, index, starting_point)
review_text_2 = repos.snippet_text(repo, index, starting_point + _snippet_ms())
body = json.dumps({
'lock_acquired': True,
'starting_point': starting_point,
'ending_point': starting_point + (_snippet_ms() * 2),
'lock_secret': lock_secret_or_message,
'review_text_1': review_text_1,
'review_text_2': review_text_2,
})
return Response(body, content_type='application/json')
else:
# return message
body = json.dumps({
'lock_acquired': False,
'message': lock_secret_or_message,
})
return Response(body, content_type='application/json')
def cancel_review(request):
# unpack lock secret, starting point, identity
lock_secret = request.POST.getone('lock_secret')
starting_point = int(request.POST.getone('starting_point'))
identity_name = request.POST.getone('identity_name')
identity_email = request.POST.getone('identity_email')
with repos.commit_lock:
# find and validate the lock
repo = repos.repo_from_request(request)
index = repo.index
if not repos.lock_is_valid(repo, index, 'review', starting_point, lock_secret):
raise ValueError('Invalid lock')
# remove the lock
repos.remove_lock(repo, index, 'review', starting_point)
# commit with identity
os.environ['GIT_AUTHOR_NAME'] = identity_name
os.environ['GIT_AUTHOR_EMAIL'] = identity_email
index.commit('review: cancel')
# return empty indicating success
return Response('', content_type='text/plain')
def post_speakers_txt(request):
text = request.POST.getone('text')
identity_name = request.POST.getone('identity_name')
identity_email = request.POST.getone('identity_email')
# Save transcription info.
repo, commit = repos.repo_from_request(request, rev='master')
with repos.commit_lock:
repo.heads['master'].checkout()
index = repo.index
filename = os.path.join(repo.working_dir, 'speakers.txt')
with open(filename, 'wb') as f:
f.write(text.encode('utf8'))
index.add(['speakers.txt'])
os.environ['GIT_AUTHOR_NAME'] = identity_name
os.environ['GIT_AUTHOR_EMAIL'] = identity_email
index.commit('speakers: save')
# Reload from repo and serve it up.
commit = repo.commit('master') # Refresh commit to match latest master.
text, mtime = repos.file_at_commit(repo, 'speakers.txt', commit)
return Response(text, content_type='text/plain', date=mtime)
def read(request):
repo, commit = repos.repo_from_request(request)
# Return cached if found.
cache_key = 'view-{0}'.format(commit.hexsha)
content, mtime = cache.get_cached_content(cache_key)
if content is None or request.GET.has_key('nocache'):
mtime = commit.authored_date
tree = commit.tree
transcription_info, _ = repos.json_file_at_commit(
repo, 'transcription.json', commit, required=True)
raw_snippets = {}
for obj in tree:
if isinstance(obj, git.Blob):
name, ext = os.path.splitext(obj.name)
if ext == '.txt':
try:
starting_point = int(name)
except ValueError:
pass
else:
raw_snippets[starting_point] = obj.data_stream.read().decode('utf8')
# Go through all snippets, whether they've been transcribed or not.
snippets = []
speakers_map = repos.speakers_map(repo, commit)
for starting_point in range(0, transcription_info['duration'], _snippet_ms()):
text = raw_snippets.get(starting_point, '').strip()
lines = _split_lines_and_expand_abbreviations(text, speakers_map)
snippets.append((starting_point, lines))
data = dict(
_standard_response(repo, commit),
snippets=sorted(snippets),
preamble_incomplete=repos.file_at_commit(
repo, 'preamble_incomplete.html', commit,
)[0],
preamble_completed=repos.file_at_commit(
repo, 'preamble_completed.html', commit,
)[0],
)
content = render('fanscribed:templates/view.mako', data, request=request)
cache.cache_content(cache_key, content, mtime)
return Response(content, date=mtime)
def all_contributors(request):
repo = repos.repo_from_request(request)
master = repo.tree('master')
# Get list of all snippet files.
snippet_filenames = []
for blob in master:
if isinstance(blob, git.Blob):
if len(blob.name) == 20 and blob.name.endswith('.txt'):
# Found a snippet.
snippet_filenames.append(blob.name)
# Find contributors to snippets.
contributor_list = []
for commit in repo.iter_commits('master', snippet_filenames):
contributor = dict(author_name=commit.author.name)
if contributor not in contributor_list:
contributor_list.append(contributor)
contributor_list.sort()
info = dict(
contributor_list=contributor_list,
)
return Response(body=json.dumps(info), content_type='application/json')
def cancel_review(request):
# unpack lock secret, starting point, identity
lock_secret = request.POST.getone('lock_secret')
starting_point = int(request.POST.getone('starting_point'))
identity_name = request.POST.getone('identity_name')
identity_email = request.POST.getone('identity_email')
with repos.commit_lock:
# find and validate the lock
# Ignore commit; we want to save to master.
repo, commit = repos.repo_from_request(request, rev='master')
index = repo.index
if not repos.lock_is_valid(repo, index, 'review', starting_point, lock_secret):
raise ValueError('Invalid lock')
# remove the lock
repos.remove_lock(repo, index, 'review', starting_point)
# commit with identity
os.environ['GIT_AUTHOR_NAME'] = identity_name
os.environ['GIT_AUTHOR_EMAIL'] = identity_email
commit_message = 'review: %s, cancel by %s' % (_label_from_ms(starting_point), identity_name)
index.commit(commit_message)
# return empty indicating success
return Response('', content_type='text/plain')
def snippets_updated(request):
"""Return formatted snippets that have been updated since the given revision."""
repo, request_commit = repos.repo_from_request(request)
since_rev = request.GET.getone('since')
# Return cached if found.
cache_key = 'updated-{0}-{1}'.format(request_commit.hexsha, since_rev)
content, mtime = cache.get_cached_content(cache_key)
if content is None or request.GET.has_key('nocache'):
since_commit = repo.commit(since_rev)
files_updated = set()
for commit in repo.iter_commits(request_commit):
# Have we reached the end?
if commit == since_commit:
break
# Look for snippet changes.
for filename in commit.stats.files:
if len(filename) == 20 and filename.endswith('.txt'):
files_updated.add(filename)
tree = request_commit.tree
speakers_map = repos.speakers_map(repo, request_commit)
snippets = []
for filename in files_updated:
starting_point = int(filename[:16])
snippet = dict(
starting_point=starting_point,
)
text = tree[filename].data_stream.read().strip()
snippet['lines'] = _split_lines_and_expand_abbreviations(text, speakers_map)
snippets.append(snippet)
data = dict(
latest_revision=repos.latest_revision(repo),
snippets=snippets,
)
content = json.dumps(data)
mtime = request_commit.authored_date
cache.cache_content(cache_key, content, mtime)
return Response(content, content_type='application/json', date=mtime)
def speakers_txt(request):
repo = repos.repo_from_request(request)
master = repo.tree('master')
text = repos.speakers_text(master)
return Response(text, content_type='text/plain')
def transcription_json(request):
repo = repos.repo_from_request(request)
master = repo.tree('master')
info = repos.transcription_info(master)
return Response(body=json.dumps(info), content_type='application/json')
def progress(request):
repo = repos.repo_from_request(request)
master = repo.tree('master')
info = repos.transcription_info(master)
return Response(body=json.dumps(_progress_dicts(master, info)), content_type='application/json')
def edit(request):
repo, commit = repos.repo_from_request(request, rev='master')
return dict(
_standard_response(repo, commit),
)
def custom_css(request):
repo = repos.repo_from_request(request)
master = repo.tree('master')
text = repos.custom_css(master)
return Response(text, content_type='text/css')
def rss_kudos(request, max_hours=24, default_minutes=60):
repo, commit = repos.repo_from_request(request)
# Get grouping parameters.
end_timestamp = int(request.GET.get('end', time.time()))
minutes_per_item = int(request.GET.get('minutes', default_minutes))
# Return cached if found.
cache_key = 'rss_kudos commit={0} minutes={1} max_hours={2} start={3}'.format(
commit.hexsha,
minutes_per_item,
max_hours,
end_timestamp,
)
content, mtime = cache.get_cached_content(cache_key)
if content is None or request.GET.has_key('nocache'):
# Get the list of kudos to give, or use the default.
kudos_txt, mtime = repos.file_at_commit(repo, 'kudos.txt', commit)
kudos_txt = kudos_txt or DEFAULT_KUDOS
kudos_lines = kudos_txt.strip().splitlines()
# Kudos templates might want transcription info.
transcription_info, _ = repos.json_file_at_commit(
repo, 'transcription.json', commit, required=True)
# Process the range of time needed for this RSS feed.
mtime = commit.authored_date
tree = commit.tree
timegroup_author_actions = {
# timegroup: {
# AUTHOR_NAME: dict(actions=[ACTION, ...], kudos=KUDOS),
# ACTION = dict(author=AUTHOR, date=DATE, position=POSITION, this_url=URL, now_url=URL)
# }
}
# Find the ending timestamp for the period of time that comes
# just before the current "partial" period of time.
max_timestamp = end_timestamp - (end_timestamp % (minutes_per_item * 60))
min_timestamp = max_timestamp - (max_hours * 60 * 60)
# Starting from the request's commit, iterate backwards.
for c in repo.iter_commits(commit):
if c.authored_date < min_timestamp:
break
snippets_affected = set()
for filename in c.stats.files:
ms = _ms_from_snippet_filename(filename)
if ms is not None:
snippets_affected.add(ms)
if snippets_affected:
earliest_ms = min(snippets_affected)
date = c.authored_date
timegroup = date - (date % (minutes_per_item * 60))
timegroup_authors = timegroup_author_actions.setdefault(timegroup, {})
author = c.author
author_actions = timegroup_authors.setdefault(author.name, dict(actions=[]))['actions']
anchor = _anchor_from_ms(earliest_ms)
position = _label_from_ms(earliest_ms)
kwargs = dict(
host=request.host,
rev=c.hexsha,
anchor=anchor,
)
now_url = 'http://{host}/?src=rk#{anchor}'.format(**kwargs)
this_url = 'http://{host}/?rev={rev}&src=rk#{anchor}'.format(**kwargs)
action = dict(
author=author,
author_name=author.name,
date=date,
position=position,
this_url=this_url,
now_url=now_url,
)
author_actions.append(action)
# Now create the kudos for each author.
for timegroup, authors in timegroup_author_actions.iteritems():
for author_name, author_info in authors.iteritems():
# Deterministically choose a pseudo-random kudos based on what we
# know about this author's contributions.
actions = author_info['actions']
latest_action = actions[0]
random.seed(latest_action['date'])
kudos_line = random.choice(kudos_lines).strip()
kudos_template = Template(text=kudos_line)
# Render it.
kudos = kudos_template.render(
author_name=author_name,
contributions=len(actions),
latest_action=latest_action,
transcription_info=transcription_info,
request=request,
)
# Keep it with the author info.
author_info['kudos'] = kudos
author_info['latest_action'] = latest_action
pub_date = commit.authored_date
data = dict(
_standard_response(repo, commit),
timegroup_author_actions=timegroup_author_actions,
pub_date=pub_date,
request=request,
rfc822_from_time=rfc822_from_time,
end_timestamp=end_timestamp,
minutes_per_item=minutes_per_item,
)
content = render('fanscribed:templates/rss_kudos.xml.mako', data, request=request)
cache.cache_content(cache_key, content, mtime)
return Response(content, content_type='application/rss+xml', date=mtime)
def speakers_txt(request):
# No rendering or processing, no need to cache.
repo, commit = repos.repo_from_request(request)
text, mtime = repos.file_at_commit(repo, 'speakers.txt', commit)
return Response(text, content_type='text/plain', date=mtime)
