Fix #2524: ignore specific commits when running update-authors.js (#2543)

Co-authored-by: Glen Whitney <[email protected]>

Author: josdejong

AUTHORS

@@ -178,7 +178,6 @@ thetazero <[email protected]>
 Alfonso Valenciana <[email protected]>
 David <[email protected]>
 Glen Whitney <[email protected]>
-Divya Yeruva <[email protected]>
 Eternal-Rise <[email protected]>
 yifanwww <[email protected]>
 NattapongSiri <[email protected]>
@@ -187,6 +186,7 @@ Abraham <[email protected]>
 Sam Estep <[email protected]>
 Sinan <[email protected]>
 HanchaiN <[email protected]>
+Divya Yeruva <[email protected]>
 simlaticak <[email protected]>
 
 # Generated by tools/update-authors.js

HISTORY.md

@@ -7,6 +7,8 @@
   of those functions  (#2531, #2539). Thanks @simlaticak and @gwhitney.
 - Fix #2532: matrix index symbol `end` not working when used inside
   a sub-expression.
+- Fix #2524: In generating AUTHORS list, ignore a list of specific commits
+  (e.g., to avoid spurious duplicates in list). (#2543)
 
 
 # 2022-04-19, version 10.5.0

tools/update-authors.js

@@ -25,7 +25,8 @@ class CaseIndifferentMap {
 const log = spawn(
   'git',
   // Inspect author name/email and body.
-  ['log', '--reverse', '--format=Author: %aN <%aE>\n%b'], {
+  ['log', '--reverse'],
+  {
     stdio: ['inherit', 'pipe', 'inherit']
   })
 const rl = readline.createInterface({ input: log.stdout })
@@ -77,17 +78,40 @@ const mailmap = new CaseIndifferentMap()
 
 const seen = new Set()
 
+// Commits from which we do not want the author to pop up in the AUTHORS list,
+// for example because the commit was done with a wrong git user account
+const ignoreCommits = [
+  '43d4551e7c19f51d30e71b35009437c7ec6491f0'
+]
+let currentCommit
+
 // Support regular git author metadata, as well as `Author:` and
 // `Co-authored-by:` in the message body. Both have been used in the past
 // to indicate multiple authors per commit, with the latter standardized
 // by GitHub now.
 const authorRe =
-  /(^Author:|^Co-authored-by:)\s+(?<author>[^<]+)\s+(?<email><[^>]+>)/i
+  /(^Author:|^\s*Co-authored-by:)\s+(?<author>[^<]+)\s+(?<email><[^>]+>)/i
+
+// Commit line regex. Example: "commit 123456"
+const commitRe = /^commit (?<commit>.+)$/i
+
 rl.on('line', (line) => {
-  const match = line.match(authorRe)
-  if (!match) return
+  const commitMatch = line.match(commitRe)
+  if (commitMatch) {
+    currentCommit = commitMatch.groups.commit
+    return
+  }
+
+  if (ignoreCommits.includes(currentCommit)) {
+    return
+  }
+
+  const authorMatch = line.match(authorRe)
+  if (!authorMatch) {
+    return
+  }
 
-  const { author, email } = match.groups
+  const { author, email } = authorMatch.groups
 
   if (seen.has(email) ||
       /@chromium\.org/.test(email) ||