def write_issue(ticket, outfile):
"""Dumps a csv line *row* from the issue query to *outfile*.
"""
# Issue text body
body = ticket[3]['description']
body = trac_to_gh(body)
# Default state: open (no known resolution)
state = STATES.get(ticket[3]['status'], 'open')
# Trac will have stored some kind of username.
reporter = ticket[3]['reporter']
# Not sure whether we have a related github account for that user.
if USERNAMES.get(reporter):
userdata = USERNAMES[reporter]
else: # If we do not, at least mention the user in our issue body
userdata = DEFAULT_USER
body = ('This issue was reported by **%s**\r\n\r\n' % reporter) + body
# Whether this is stored in 'milestone' or '__group__' depends on the
# query type. Try to find the data or assign the default milestone 0.
milestone_info = ticket[3]['milestone']
milestone = MILESTONES.get(milestone_info, 3)
labels = [] # Collect random tags that might serve as labels
for tag in ('type', 'component', 'priority'):
if ticket[3].get(tag) and LABELS.get(ticket[3][tag]):
label = LABELS[ticket[3][tag]]
labels.append({'name': github_label(label)})
# Dates
updated_at = DateTime(str(ticket[2])).ISO8601()
created_at = DateTime(str(ticket[1])).ISO8601()
# Now prepare writing all data into the json files
dct = {
'title': ticket[3]['summary'],
'body': body,
'state': state,
'user': userdata,
'milestone': int(milestone),
'labels': labels,
'updated_at': updated_at,
'created_at': created_at,
}
# Assigned user in trac and github account of that assignee
# assigned_trac = ticket[3]['owner']
# assigned = USERNAMES.get(assigned_trac)
# Assigning really does not make sense without github account
# if state == 'open' and assigned and assigned['login'] != 'fifengine':
# print assigned
# dct['assignee'] = assigned
# Everything collected, write the json file
json.dump(dct, outfile, indent=5)
def main():
#######################################################################
# Gather information about our tickets (mainly assembles comment list)
#######################################################################
# Stores a list of comments for each ticket by ID
#prock - load pickle file here and generate a row
f = open(TICKET_FILE, 'rb')
tickets = pickle.load(f)
f.close()
comment_coll = collections.defaultdict(list)
for ticketid in tickets:
comments = tickets[ticketid][4]
for comment in comments:
if ("comment" in comment):
dct = massage_comment(ticketid, comment[0],comment[1],comment[4])
comment_coll[ticketid].append(dct)
#######################################################################
# Write the ticket comments to json files indicating their parent issue
#######################################################################
for ticket, data in comment_coll.iteritems():
if (comment_coll[ticket][0]['body'] != '.'):
count = 0
for row in data:
with open(COMMENTS_PATH % (ticket,count), 'w') as f:
json.dump(row, f, indent=5)
count = count + 1
#######################################################################
# Write the actual ticket data to separate json files (GitHub API v3)
#######################################################################
for ticketid in tickets:
with open(ISSUES_PATH % str(ticketid), 'w') as f:
write_issue(tickets[ticketid], f)
#######################################################################
# Finally, dump all milestones and the related data. This script is not
# attempting to extract due dates or other data. We just manually mined
# the milestone names once and stored that in MILESTONES for reference.
#######################################################################
for name, id in MILESTONES.iteritems():
with open(MILESTONES_PATH % id, 'w') as f:
dct = {
'number': id,
'creator': DEFAULT_USER,
'title': name,
}
json.dump(dct, f, indent=5)
def main():
#######################################################################
# Gather information about our tickets (mainly assembles comment list)
#######################################################################
# Stores a list of comments for each ticket by ID
comments = collections.defaultdict(list)
comment_rows = csv.reader(open(CSVFILE, 'rb'),
delimiter=CSVDELIM, quotechar=CSVESCAPE)
for row in comment_rows:
try:
ticket, date, author, body = row
except: # malformed ticket query, fix in trac or the csv file!
print '/!\\ Please at check this csv row: /!\\\n', row
continue
if not ticket: # nothing we can do if there is no ticket to assign
continue
dct = massage_comment(ticket, date, author, body)
comments[ticket].append(dct) # defaultdict, append always works
#######################################################################
# Write the ticket comments to json files indicating their parent issue
#######################################################################
for ticket, data in comments.iteritems():
with open(ISSUES_PATH % ticket, 'w') as f:
json.dump(data, f, indent=5)
#######################################################################
# Write the actual ticket data to separate json files (GitHub API v3)
#######################################################################
csv_data = urllib.urlopen(TRAC_REPORT_URL)
ticket_data = csv.DictReader(csv_data)
for row in ticket_data:
if not (row.get('summary') and row.get('ticket')):
continue
with open(ISSUES_PATH % row['ticket'], 'w') as f:
write_issue(row, f)
#######################################################################
# Finally, dump all milestones and the related data. This script is not
# attempting to extract due dates or other data. We just manually mined
# the milestone names once and stored that in MILESTONES for reference.
#######################################################################
for name, id in MILESTONES.iteritems():
with open(MILESTONES_PATH % id, 'w') as f:
dct = {
'number': id,
'creator': DEFAULT_USER,
'title': name,
}
json.dump(dct, f, indent=5)
def write_issue(row, outfile):
"""Dumps a csv line *row* from the issue query to *outfile*.
"""
for key, value in row.items():
row[key] = row[key].decode('utf-8')
# Issue text body
body = row.get('_description', u'')
body = trac_to_gh(body) + '\r\n\r\n' \
'[> Link to originally reported Trac ticket <] ({url})'.format(
url=TRAC_TICKET_URL % row['ticket'])
# Default state: open (no known resolution)
state = STATES.get(row.get('status'), 'open')
# Trac will have stored some kind of username.
reporter = row['_reporter']
# Not sure whether we have a related github account for that user.
if USERNAMES.get(reporter):
userdata = USERNAMES[reporter]
else: # If we do not, at least mention the user in our issue body
userdata = DEFAULT_USER
body = ('This issue was reported by **%s**\r\n\r\n' % reporter) + body
# Whether this is stored in 'milestone' or '__group__' depends on the
# query type. Try to find the data or assign the default milestone 0.
milestone_info = row.get(('milestone'), row.get('__group__'))
milestone = MILESTONES.get(milestone_info, 0)
labels = [] # Collect random tags that might serve as labels
for tag in ('type', 'component', 'priority'):
if row.get(tag) and LABELS.get(row[tag]):
label = LABELS[row[tag]]
labels.append({'name': github_label(label)})
# Also attach a special label to our starter tasks.
# Again, please ignore this.
#if row['ticket'] in easy_tickets:
# labels.append({'name': unicode(LABELS.get('start').lower())})
# Dates
updated_at = row.get('modified') or row.get('_changetime')
created_at = row.get('created') or updated_at
# Now prepare writing all data into the json files
dct = {
'title': row['summary'],
'body': body,
'state': state,
'user': userdata,
'milestone': {'number': milestone},
'labels': labels,
'updated_at': updated_at,
'created_at': created_at,
}
# Assigned user in trac and github account of that assignee
assigned_trac = row.get('owner')
assigned = USERNAMES.get(assigned_trac)
# Assigning really does not make sense without github account
if state == 'open' and assigned:
dct['assignee'] = assigned
# Everything collected, write the json file
json.dump(dct, outfile, indent=5)
