|
| 1 | +#!/usr/bin/env python3 |
| 2 | +""" |
| 3 | +Glific — Weekly Release Tracker → Google Sheets |
| 4 | +
|
| 5 | +Runs every Saturday. Fetches all GitHub releases published in the org during |
| 6 | +the current week (Monday–Saturday) and appends one row per week to a dedicated |
| 7 | +'Releases' tab in the Google Sheet. Weeks are grouped under their month so |
| 8 | +the sheet stays easy to read at a glance. |
| 9 | +
|
| 10 | +Required env vars: |
| 11 | + GITHUB_TOKEN - GitHub PAT with repo/read:org scope |
| 12 | + ORG - GitHub organisation name |
| 13 | + GOOGLE_CREDENTIALS_JSON - Service account JSON key (full JSON string) |
| 14 | + SPREADSHEET_ID - Target Google Sheet ID |
| 15 | +""" |
| 16 | + |
| 17 | +import json |
| 18 | +import os |
| 19 | +import sys |
| 20 | +from datetime import date, datetime, timedelta, timezone |
| 21 | + |
| 22 | +import requests |
| 23 | +from dotenv import load_dotenv |
| 24 | +from google.oauth2 import service_account |
| 25 | +from googleapiclient.discovery import build |
| 26 | + |
| 27 | +load_dotenv() |
| 28 | + |
| 29 | +SCOPES = ["https://www.googleapis.com/auth/spreadsheets"] |
| 30 | +SHEET_NAME = "Releases" |
| 31 | +HEADER = ["Month", "Week Start", "Week End", "Release Count", "Releases", "Last Updated"] |
| 32 | +EXCLUDED_REPOS = {"support-process"} |
| 33 | + |
| 34 | + |
| 35 | +# ── Date helpers ────────────────────────────────────────────────────────────── |
| 36 | + |
| 37 | + |
| 38 | +def _week_bounds(reference: date) -> tuple[date, date]: |
| 39 | + """Return (monday, saturday) for the week containing *reference*.""" |
| 40 | + monday = reference - timedelta(days=reference.weekday()) # weekday() 0=Mon |
| 41 | + saturday = monday + timedelta(days=5) |
| 42 | + return monday, saturday |
| 43 | + |
| 44 | + |
| 45 | +def _month_label(d: date) -> str: |
| 46 | + return d.strftime("%B %Y") |
| 47 | + |
| 48 | + |
| 49 | +# ── GitHub REST API ─────────────────────────────────────────────────────────── |
| 50 | + |
| 51 | + |
| 52 | +def _gh_headers(token: str) -> dict: |
| 53 | + return { |
| 54 | + "Authorization": f"Bearer {token}", |
| 55 | + "Accept": "application/vnd.github+json", |
| 56 | + "X-GitHub-Api-Version": "2022-11-28", |
| 57 | + } |
| 58 | + |
| 59 | + |
| 60 | +def _list_repos(org: str, token: str) -> list[str]: |
| 61 | + """Return all non-archived repo names in the org.""" |
| 62 | + repos = [] |
| 63 | + page = 1 |
| 64 | + while True: |
| 65 | + resp = requests.get( |
| 66 | + f"https://api.github.com/orgs/{org}/repos", |
| 67 | + params={"per_page": 100, "page": page, "type": "all"}, |
| 68 | + headers=_gh_headers(token), |
| 69 | + timeout=30, |
| 70 | + ) |
| 71 | + resp.raise_for_status() |
| 72 | + data = resp.json() |
| 73 | + for r in data: |
| 74 | + if not r.get("archived") and r["name"] not in EXCLUDED_REPOS: |
| 75 | + repos.append(r["name"]) |
| 76 | + if len(data) < 100: |
| 77 | + break |
| 78 | + page += 1 |
| 79 | + return repos |
| 80 | + |
| 81 | + |
| 82 | +def _fetch_releases_in_range( |
| 83 | + org: str, repos: list[str], start: date, end: date, token: str |
| 84 | +) -> list[dict]: |
| 85 | + """ |
| 86 | + Return releases published between *start* (inclusive) and *end* (inclusive). |
| 87 | + Each entry: {"repo": str, "tag": str, "name": str, "published_at": str} |
| 88 | + """ |
| 89 | + start_dt = datetime(start.year, start.month, start.day, tzinfo=timezone.utc) |
| 90 | + end_dt = datetime(end.year, end.month, end.day, 23, 59, 59, tzinfo=timezone.utc) |
| 91 | + |
| 92 | + releases = [] |
| 93 | + for repo in repos: |
| 94 | + page = 1 |
| 95 | + while True: |
| 96 | + resp = requests.get( |
| 97 | + f"https://api.github.com/repos/{org}/{repo}/releases", |
| 98 | + params={"per_page": 100, "page": page}, |
| 99 | + headers=_gh_headers(token), |
| 100 | + timeout=30, |
| 101 | + ) |
| 102 | + if resp.status_code == 404: |
| 103 | + break |
| 104 | + resp.raise_for_status() |
| 105 | + data = resp.json() |
| 106 | + found_any = False |
| 107 | + for r in data: |
| 108 | + pub = r.get("published_at") |
| 109 | + if not pub: |
| 110 | + continue |
| 111 | + pub_dt = datetime.fromisoformat(pub.rstrip("Z")).replace(tzinfo=timezone.utc) |
| 112 | + if pub_dt < start_dt: |
| 113 | + # Releases are newest-first; once we go past the window, stop. |
| 114 | + found_any = False |
| 115 | + break |
| 116 | + found_any = True |
| 117 | + if pub_dt <= end_dt: |
| 118 | + releases.append({ |
| 119 | + "repo": repo, |
| 120 | + "tag": r.get("tag_name", ""), |
| 121 | + "name": r.get("name") or r.get("tag_name", ""), |
| 122 | + "published_at": pub, |
| 123 | + }) |
| 124 | + if len(data) < 100 or not found_any: |
| 125 | + break |
| 126 | + page += 1 |
| 127 | + return releases |
| 128 | + |
| 129 | + |
| 130 | +# ── Google Sheets ───────────────────────────────────────────────────────────── |
| 131 | + |
| 132 | + |
| 133 | +def _sheets_client(): |
| 134 | + creds_json = os.environ.get("GOOGLE_CREDENTIALS_JSON") |
| 135 | + if not creds_json: |
| 136 | + sys.exit("GOOGLE_CREDENTIALS_JSON not set") |
| 137 | + info = json.loads(creds_json) |
| 138 | + creds = service_account.Credentials.from_service_account_info(info, scopes=SCOPES) |
| 139 | + return build("sheets", "v4", credentials=creds, cache_discovery=False).spreadsheets() |
| 140 | + |
| 141 | + |
| 142 | +def _ensure_sheet_tab(sheets, spreadsheet_id: str): |
| 143 | + meta = sheets.get(spreadsheetId=spreadsheet_id).execute() |
| 144 | + existing = [s["properties"]["title"] for s in meta.get("sheets", [])] |
| 145 | + if SHEET_NAME not in existing: |
| 146 | + sheets.batchUpdate( |
| 147 | + spreadsheetId=spreadsheet_id, |
| 148 | + body={"requests": [{"addSheet": {"properties": {"title": SHEET_NAME}}}]}, |
| 149 | + ).execute() |
| 150 | + print(f"Created sheet tab '{SHEET_NAME}'.") |
| 151 | + |
| 152 | + |
| 153 | +def _ensure_header(sheets, spreadsheet_id: str): |
| 154 | + result = sheets.values().get( |
| 155 | + spreadsheetId=spreadsheet_id, |
| 156 | + range=f"'{SHEET_NAME}'!A1:{chr(64 + len(HEADER))}1", |
| 157 | + ).execute() |
| 158 | + if not result.get("values"): |
| 159 | + sheets.values().update( |
| 160 | + spreadsheetId=spreadsheet_id, |
| 161 | + range=f"'{SHEET_NAME}'!A1", |
| 162 | + valueInputOption="RAW", |
| 163 | + body={"values": [HEADER]}, |
| 164 | + ).execute() |
| 165 | + print("Header row written.") |
| 166 | + |
| 167 | + |
| 168 | +def _all_rows(sheets, spreadsheet_id: str) -> list[list[str]]: |
| 169 | + result = sheets.values().get( |
| 170 | + spreadsheetId=spreadsheet_id, |
| 171 | + range=f"'{SHEET_NAME}'!A:F", |
| 172 | + ).execute() |
| 173 | + return result.get("values", []) |
| 174 | + |
| 175 | + |
| 176 | +def _find_week_row(rows: list[list[str]], week_start_iso: str) -> int | None: |
| 177 | + """Return 1-based sheet row index where column B matches week_start_iso, or None.""" |
| 178 | + for idx, row in enumerate(rows, start=1): |
| 179 | + if len(row) >= 2 and row[1] == week_start_iso: |
| 180 | + return idx |
| 181 | + return None |
| 182 | + |
| 183 | + |
| 184 | +def _write_row(sheets, spreadsheet_id: str, row_index: int, values: list): |
| 185 | + sheets.values().update( |
| 186 | + spreadsheetId=spreadsheet_id, |
| 187 | + range=f"'{SHEET_NAME}'!A{row_index}:{chr(64 + len(HEADER))}{row_index}", |
| 188 | + valueInputOption="RAW", |
| 189 | + body={"values": [values]}, |
| 190 | + ).execute() |
| 191 | + |
| 192 | + |
| 193 | +def _month_already_in_sheet(rows: list[list[str]], month_label: str) -> bool: |
| 194 | + """True if any existing row already carries this month label in column A.""" |
| 195 | + return any(row and row[0] == month_label for row in rows) |
| 196 | + |
| 197 | + |
| 198 | +# ── Main ────────────────────────────────────────────────────────────────────── |
| 199 | + |
| 200 | + |
| 201 | +def main(): |
| 202 | + token = os.environ.get("GITHUB_TOKEN") |
| 203 | + org = os.environ.get("ORG") |
| 204 | + spreadsheet_id = os.environ.get("SPREADSHEET_ID") |
| 205 | + |
| 206 | + if not all([token, org, spreadsheet_id]): |
| 207 | + sys.exit("Missing required env vars: GITHUB_TOKEN, ORG, SPREADSHEET_ID") |
| 208 | + |
| 209 | + saturday_input = os.environ.get("SATURDAY_DATE", "").strip() |
| 210 | + if saturday_input: |
| 211 | + try: |
| 212 | + reference = date.fromisoformat(saturday_input) |
| 213 | + except ValueError: |
| 214 | + sys.exit(f"Invalid SATURDAY_DATE '{saturday_input}' — expected YYYY-MM-DD.") |
| 215 | + if reference.weekday() != 5: # 5 = Saturday |
| 216 | + sys.exit( |
| 217 | + f"SATURDAY_DATE '{saturday_input}' is not a Saturday " |
| 218 | + f"(it's a {reference.strftime('%A')})." |
| 219 | + ) |
| 220 | + today = reference |
| 221 | + else: |
| 222 | + today = date.today() |
| 223 | + |
| 224 | + week_start, week_end = _week_bounds(today) |
| 225 | + month_label = _month_label(week_end) # week belongs to the month it ends in |
| 226 | + |
| 227 | + print(f"Week : {week_start} → {week_end}") |
| 228 | + print(f"Month : {month_label}") |
| 229 | + print(f"Fetching repos for org '{org}'…") |
| 230 | + |
| 231 | + repos = _list_repos(org, token) |
| 232 | + print(f"Repos found: {len(repos)}") |
| 233 | + |
| 234 | + print(f"Fetching releases published {week_start} → {week_end}…") |
| 235 | + releases = _fetch_releases_in_range(org, repos, week_start, week_end, token) |
| 236 | + print(f"Releases found: {len(releases)}") |
| 237 | + for r in releases: |
| 238 | + print(f" {r['repo']} {r['tag']} ({r['published_at']})") |
| 239 | + |
| 240 | + # Deduplicate by date: multiple repos releasing on the same day count as 1. |
| 241 | + release_dates = { |
| 242 | + datetime.fromisoformat(r["published_at"].rstrip("Z")).date() |
| 243 | + for r in releases |
| 244 | + } |
| 245 | + release_count = len(release_dates) |
| 246 | + print(f"Unique release days: {release_count}") |
| 247 | + |
| 248 | + # Build a compact summary string: "repo@tag, repo@tag, …" |
| 249 | + release_summary = ", ".join(f"{r['repo']}@{r['tag']}" for r in releases) or "—" |
| 250 | + |
| 251 | + sheets = _sheets_client() |
| 252 | + _ensure_sheet_tab(sheets, spreadsheet_id) |
| 253 | + _ensure_header(sheets, spreadsheet_id) |
| 254 | + |
| 255 | + rows = _all_rows(sheets, spreadsheet_id) |
| 256 | + |
| 257 | + # Decide whether to show the month label (only on the first week of that month) |
| 258 | + month_cell = month_label if not _month_already_in_sheet(rows[1:], month_label) else "" |
| 259 | + |
| 260 | + row_values = [ |
| 261 | + month_cell, |
| 262 | + week_start.isoformat(), |
| 263 | + week_end.isoformat(), |
| 264 | + release_count, |
| 265 | + release_summary, |
| 266 | + today.isoformat(), |
| 267 | + ] |
| 268 | + |
| 269 | + existing_row = _find_week_row(rows, week_start.isoformat()) |
| 270 | + if existing_row: |
| 271 | + print(f"Updating existing row {existing_row}…") |
| 272 | + # Preserve the month cell if it was already set (don't blank it on re-run) |
| 273 | + if rows[existing_row - 1] and rows[existing_row - 1][0]: |
| 274 | + row_values[0] = rows[existing_row - 1][0] |
| 275 | + _write_row(sheets, spreadsheet_id, existing_row, row_values) |
| 276 | + else: |
| 277 | + next_row = len(rows) + 1 |
| 278 | + print(f"Appending new row {next_row}…") |
| 279 | + _write_row(sheets, spreadsheet_id, next_row, row_values) |
| 280 | + |
| 281 | + print("Done.") |
| 282 | + |
| 283 | + |
| 284 | +if __name__ == "__main__": |
| 285 | + main() |
0 commit comments