From c446679adbfc913488f4166615907062d4f17e27 Mon Sep 17 00:00:00 2001
From: Sergey Matveev <stargrave@stargrave.org>
Date: Fri, 18 Feb 2022 10:00:19 +0300
Subject: [PATCH] Use title+content as more reliable item identifier
---
cmd/feed2mdir/main.go | 6 +++++-
doc/mail.texi | 2 +-
2 files changed, 6 insertions(+), 2 deletions(-)
diff --git a/cmd/feed2mdir/main.go b/cmd/feed2mdir/main.go
index 9074cef..5783d1f 100644
--- a/cmd/feed2mdir/main.go
+++ b/cmd/feed2mdir/main.go
@@ -58,7 +58,6 @@ func main() {
} else {
when = &now
}
- fn := hex.EncodeToString(h.Sum(nil)[:sha512.Size/2])
var what string
if len(item.Content) == 0 {
what = item.Description
@@ -67,6 +66,11 @@ func main() {
}
what = strings.TrimPrefix(what, "<![CDATA[")
what = strings.TrimSuffix(what, "]]>")
+ h := sha512.New()
+ h.Write([]byte(item.Title))
+ h.Write([]byte{0})
+ h.Write([]byte(what))
+ fn := hex.EncodeToString(h.Sum(nil)[:sha512.Size/2])
exists := false
for _, d := range []string{"cur", "new"} {
entries, err := os.ReadDir(path.Join(mdir, d))
diff --git a/doc/mail.texi b/doc/mail.texi
index 346bc76..6e42273 100644
--- a/doc/mail.texi
+++ b/doc/mail.texi
@@ -15,7 +15,7 @@ Content-Type: text/html; charset=utf-8
Base64-encoded item's content. That is HTML as a rule.
@end verbatim
-Each message is placed under @code{HEX(SHA512/2(title+date))} filename.
+Each message is placed under @code{HEX(SHA512/2(title+content))} filename.
Unfortunately many feeds does not provide any date information at all.
Often multiple entries contain the same date. So you can not solely
depend on update/publish date.
--
2.51.0