Show More
@@ -1,109 +1,149 b'' | |||
|
1 | 1 | #!/usr/bin/env python |
|
2 | 2 | """Simple tools to query github.com and gather stats about issues. |
|
3 | 3 | """ |
|
4 | 4 | #----------------------------------------------------------------------------- |
|
5 | 5 | # Imports |
|
6 | 6 | #----------------------------------------------------------------------------- |
|
7 | 7 | |
|
8 | 8 | from __future__ import print_function |
|
9 | 9 | |
|
10 | 10 | import json |
|
11 | import re | |
|
11 | 12 | import sys |
|
12 | 13 | |
|
13 | 14 | from datetime import datetime, timedelta |
|
14 | 15 | from urllib import urlopen |
|
15 | 16 | |
|
16 | 17 | #----------------------------------------------------------------------------- |
|
18 | # Globals | |
|
19 | #----------------------------------------------------------------------------- | |
|
20 | ||
|
21 | ISO8601 = "%Y-%m-%dT%H:%M:%SZ" | |
|
22 | PER_PAGE = 100 | |
|
23 | ||
|
24 | element_pat = re.compile(r'<(.+?)>') | |
|
25 | rel_pat = re.compile(r'rel=[\'"](\w+)[\'"]') | |
|
26 | ||
|
27 | #----------------------------------------------------------------------------- | |
|
17 | 28 | # Functions |
|
18 | 29 | #----------------------------------------------------------------------------- |
|
19 | 30 | |
|
20 | def get_issues(project="ipython/ipython/", state="open"): | |
|
31 | def parse_link_header(headers): | |
|
32 | link_s = headers.get('link', '') | |
|
33 | urls = element_pat.findall(link_s) | |
|
34 | rels = rel_pat.findall(link_s) | |
|
35 | d = {} | |
|
36 | for rel,url in zip(rels, urls): | |
|
37 | d[rel] = url | |
|
38 | return d | |
|
39 | ||
|
40 | def get_paged_request(url): | |
|
41 | """get a full list, handling APIv3's paging""" | |
|
42 | results = [] | |
|
43 | while url: | |
|
44 | print("fetching %s" % url, file=sys.stderr) | |
|
45 | f = urlopen(url) | |
|
46 | results.extend(json.load(f)) | |
|
47 | links = parse_link_header(f.headers) | |
|
48 | url = links.get('next') | |
|
49 | return results | |
|
50 | ||
|
51 | def get_issues(project="ipython/ipython", state="closed", pulls=False): | |
|
21 | 52 | """Get a list of the issues from the Github API.""" |
|
22 | f = urlopen("http://github.com/api/v2/json/issues/list/%s%s" % (project, | |
|
23 | state)) | |
|
24 | return json.load(f)['issues'] | |
|
53 | which = 'pulls' if pulls else 'issues' | |
|
54 | url = "https://api.github.com/repos/%s/%s?state=%s&per_page=%i" % (project, which, state, PER_PAGE) | |
|
55 | return get_paged_request(url) | |
|
25 | 56 | |
|
26 | 57 | |
|
27 | 58 | def _parse_datetime(s): |
|
28 | 59 | """Parse dates in the format returned by the Github API.""" |
|
29 | return datetime.strptime(s.rpartition(" ")[0], "%Y/%m/%d %H:%M:%S") | |
|
60 | if s: | |
|
61 | return datetime.strptime(s, ISO8601) | |
|
62 | else: | |
|
63 | return datetime.fromtimestamp(0) | |
|
30 | 64 | |
|
31 | 65 | |
|
32 | 66 | def issues2dict(issues): |
|
33 | 67 | """Convert a list of issues to a dict, keyed by issue number.""" |
|
34 | 68 | idict = {} |
|
35 | 69 | for i in issues: |
|
36 | 70 | idict[i['number']] = i |
|
37 | 71 | return idict |
|
38 | 72 | |
|
39 | 73 | |
|
40 | 74 | def is_pull_request(issue): |
|
41 | 75 | """Return True if the given issue is a pull request.""" |
|
42 | 76 | return 'pull_request_url' in issue |
|
43 | 77 | |
|
44 | 78 | |
|
45 |
def issues_closed_since(period=timedelta(days=365), project="ipython/ipython |
|
|
79 | def issues_closed_since(period=timedelta(days=365), project="ipython/ipython", pulls=False): | |
|
46 | 80 | """Get all issues closed since a particular point in time. period |
|
47 | 81 | can either be a datetime object, or a timedelta object. In the |
|
48 | 82 | latter case, it is used as a time before the present.""" |
|
49 | allclosed = get_issues(project=project, state='closed') | |
|
83 | ||
|
84 | which = 'pulls' if pulls else 'issues' | |
|
85 | ||
|
50 | 86 | if isinstance(period, timedelta): |
|
51 | 87 | period = datetime.now() - period |
|
52 | return [i for i in allclosed if _parse_datetime(i['closed_at']) > period] | |
|
88 | url = "https://api.github.com/repos/%s/%s?state=closed&sort=updated&since=%s&per_page=%i" % (project, which, period.strftime(ISO8601), PER_PAGE) | |
|
89 | allclosed = get_paged_request(url) | |
|
90 | # allclosed = get_issues(project=project, state='closed', pulls=pulls, since=period) | |
|
91 | filtered = [i for i in allclosed if _parse_datetime(i['closed_at']) > period] | |
|
92 | return filtered | |
|
53 | 93 | |
|
54 | 94 | |
|
55 | 95 | def sorted_by_field(issues, field='closed_at', reverse=False): |
|
56 | 96 | """Return a list of issues sorted by closing date date.""" |
|
57 | 97 | return sorted(issues, key = lambda i:i[field], reverse=reverse) |
|
58 | 98 | |
|
59 | 99 | |
|
60 | 100 | def report(issues, show_urls=False): |
|
61 | 101 | """Summary report about a list of issues, printing number and title. |
|
62 | 102 | """ |
|
63 | 103 | # titles may have unicode in them, so we must encode everything below |
|
64 | 104 | if show_urls: |
|
65 | 105 | for i in issues: |
|
66 | print('* `%d <%s>`_: %s' % (i['number'], | |
|
67 | i['html_url'].encode('utf-8'), | |
|
106 | role = 'ghpull' if 'merged' in i else 'ghissue' | |
|
107 | print('* :%s:`%d`: %s' % (role, i['number'], | |
|
68 | 108 | i['title'].encode('utf-8'))) |
|
69 | 109 | else: |
|
70 | 110 | for i in issues: |
|
71 | 111 | print('* %d: %s' % (i['number'], i['title'].encode('utf-8'))) |
|
72 | 112 | |
|
73 | 113 | #----------------------------------------------------------------------------- |
|
74 | 114 | # Main script |
|
75 | 115 | #----------------------------------------------------------------------------- |
|
76 | 116 | |
|
77 | 117 | if __name__ == "__main__": |
|
78 | 118 | # Whether to add reST urls for all issues in printout. |
|
79 | 119 | show_urls = True |
|
80 | 120 | |
|
81 | 121 | # By default, search one month back |
|
82 | 122 | if len(sys.argv) > 1: |
|
83 | 123 | days = int(sys.argv[1]) |
|
84 | 124 | else: |
|
85 | 125 | days = 30 |
|
86 | 126 | |
|
87 | 127 | # turn off to play interactively without redownloading, use %run -i |
|
88 | 128 | if 1: |
|
89 | issues = issues_closed_since(timedelta(days=days)) | |
|
129 | issues = issues_closed_since(timedelta(days=days), pulls=False) | |
|
130 | pulls = issues_closed_since(timedelta(days=days), pulls=True) | |
|
90 | 131 | |
|
91 | 132 | # For regular reports, it's nice to show them in reverse chronological order |
|
92 | 133 | issues = sorted_by_field(issues, reverse=True) |
|
93 | ||
|
94 | # Break up into pull requests and regular issues | |
|
95 | pulls = filter(is_pull_request, issues) | |
|
96 | regular = filter(lambda i: not is_pull_request(i), issues) | |
|
97 | n_issues, n_pulls, n_regular = map(len, (issues, pulls, regular)) | |
|
134 | pulls = sorted_by_field(pulls, reverse=True) | |
|
135 | ||
|
136 | n_issues, n_pulls = map(len, (issues, pulls)) | |
|
137 | n_total = n_issues + n_pulls | |
|
98 | 138 | |
|
99 | 139 | # Print summary report we can directly include into release notes. |
|
100 |
print("Git |
|
|
140 | print("GitHub stats for the last %d days." % days) | |
|
101 | 141 | print("We closed a total of %d issues, %d pull requests and %d regular \n" |
|
102 | 142 | "issues; this is the full list (generated with the script \n" |
|
103 |
"`tools/github_stats.py`):" % (n_ |
|
|
143 | "`tools/github_stats.py`):" % (n_total, n_pulls, n_issues)) | |
|
104 | 144 | print() |
|
105 |
print('Pull |
|
|
145 | print('Pull Requests (%d):\n' % n_pulls) | |
|
106 | 146 | report(pulls, show_urls) |
|
107 | 147 | print() |
|
108 |
print(' |
|
|
109 |
report( |
|
|
148 | print('Issues (%d):\n' % n_issues) | |
|
149 | report(issues, show_urls) |
General Comments 0
You need to be logged in to leave comments.
Login now