diff options
| author | Aleksandr Nogikh <nogikh@google.com> | 2024-10-25 14:29:36 +0200 |
|---|---|---|
| committer | Aleksandr Nogikh <nogikh@google.com> | 2024-10-25 12:42:32 +0000 |
| commit | 65e8686b0e9e909b6ea5629f95a9b14e81927872 (patch) | |
| tree | 72d5dfa978f09fca4df8a6265a51a9f92f747af8 /pkg | |
| parent | 42a1ab121510d4ace36fddeb7396bdb7f28bd489 (diff) | |
dashboard: adapt to the new Googe Groups footer
It changed recently and it broke our email discussion link extraction.
Closes #5431.
Diffstat (limited to 'pkg')
| -rw-r--r-- | pkg/email/parser.go | 8 | ||||
| -rw-r--r-- | pkg/email/parser_test.go | 29 |
2 files changed, 35 insertions, 2 deletions
diff --git a/pkg/email/parser.go b/pkg/email/parser.go index e8aa1a8c0..b748d02c3 100644 --- a/pkg/email/parser.go +++ b/pkg/email/parser.go @@ -11,6 +11,7 @@ import ( "mime/multipart" "mime/quotedprintable" "net/mail" + "net/url" "regexp" "sort" "strings" @@ -59,8 +60,8 @@ const ( cmdTest5 ) -var groupsLinkRe = regexp.MustCompile("\nTo view this discussion on the web visit" + - " (https://groups\\.google\\.com/.*?)\\.(?:\r)?\n") +var groupsLinkRe = regexp.MustCompile(`(?m)\nTo view this discussion (?:on the web )?visit` + + ` (https://groups\.google\.com/.*?)\.(:?$|\n|\r)`) func prepareEmails(list []string) map[string]bool { ret := make(map[string]bool) @@ -153,6 +154,9 @@ func Parse(r io.Reader, ownEmails, goodLists, domains []string) (*Email, error) link := "" if match := groupsLinkRe.FindStringSubmatchIndex(bodyStr); match != nil { link = bodyStr[match[2]:match[3]] + if unescaped, err := url.QueryUnescape(link); err == nil { + link = unescaped + } } author := CanonicalEmail(from[0].Address) diff --git a/pkg/email/parser_test.go b/pkg/email/parser_test.go index 239b795a3..4c042bcfc 100644 --- a/pkg/email/parser_test.go +++ b/pkg/email/parser_test.go @@ -461,6 +461,35 @@ For more options, visit https://groups.google.com/d/optout.`, }, }, }}, + {`Date: Sun, 7 May 2017 19:54:00 -0700 +Message-ID: <123> +Subject: new footer +From: Bob <bob@example.com> +To: syzbot <foo+4564456@bar.com> +Content-Type: text/plain; charset="UTF-8" + +some title + +-- +You received this message because you are subscribed to the Google Groups "syzkaller-bugs" group. +To unsubscribe from this group and stop receiving emails from it, send an email to syzkaller-bugs+unsubscribe@googlegroups.com. +To view this discussion visit https://groups.google.com/d/msgid/syzkaller-bugs/671b7fb2.050a0220.2e773.0000.GAE%40google.com.`, + Email{ + BugIDs: []string{"4564456"}, + MessageID: "<123>", + Date: time.Date(2017, time.May, 7, 19, 54, 0, 0, parseTestZone), + Link: "https://groups.google.com/d/msgid/syzkaller-bugs/671b7fb2.050a0220.2e773.0000.GAE@google.com", + Subject: "new footer", + Author: "bob@example.com", + Cc: []string{"bob@example.com"}, + Body: `some title + +-- +You received this message because you are subscribed to the Google Groups "syzkaller-bugs" group. +To unsubscribe from this group and stop receiving emails from it, send an email to syzkaller-bugs+unsubscribe@googlegroups.com. +To view this discussion visit https://groups.google.com/d/msgid/syzkaller-bugs/671b7fb2.050a0220.2e773.0000.GAE%40google.com.`, + Patch: "", + }}, {`Date: Sun, 7 May 2017 19:54:00 -0700 Message-ID: <123> |