Update stage3 status

README.md

@@ -1 +1,49 @@
-GOOOOOOO
+# cnblogs archiver
+
+## How can I help? 
+
+### Binary
+
+Go to [release](https://git.saveweb.org/saveweb/cnblogs/releases) page, downlaod `cnblogs_posts_list` and run it.
+
+WARNING: DO NOT run `cnblogs_posts_list` concurrently (on same IP), you may be banned by cnblogs.
+
+### With Docker
+
+```bash
+export ARCHIVIST=<your_node_name> # a string that can uniquely identify your node (for example: bob-gcloud-514). (Legal characters: letters, numbers, -, _)
+```
+
+```bash
+if [[ -z "$ARCHIVIST" ]]; then
+    echo "WARN: ARCHIVIST must be set"
+    exit 1
+fi
+_image="icecodexi/saveweb:cnblogs"
+docker pull "${_image}" \
+    && docker stop cnblogs
+docker rm -f cnblogs \
+    && docker run --env ARCHIVIST="$ARCHIVIST" --restart always \
+        --volume /etc/localtime:/etc/localtime:ro \
+        --cpu-shares 512 --memory 512M --memory-swap 512M \
+        --label=com.centurylinklabs.watchtower.enable=true \
+        --detach  --name cnblogs \
+        "${_image}"
+```
+
+
+## Archiving stages
+
+### stage1：detect all blogids (~~finished~~)
+
+run `cnblogs_rss_detect`
+
+### stage2：iterate all blogids and collect all posts' URLs (~~finished~~)
+
+run `cnblogs_posts_list`
+
+<!-- ### stage3：导出文章 urls.txt 并发送给 ArchiveTeam -->
+
+### stage3：export all posts' URLs and send to ArchiveTeam (~~finished~~)
+
+### stage4：also download all posts' HTMLs by ourselves (TODO)

cmd/cnblogs_posts_list/cnblogs_posts_list.go

@@ -69,6 +69,30 @@ func claimWorker(i int, tracker *savewebtracker.Tracker) {
 	}
 }
 
+// Refactored deduplication function
+func dedup_append(allPostMetas, newPostMetas []cnblogs_api.PostMeta) ([]cnblogs_api.PostMeta, int) {
+	founds := 0
+	for _, newPostMeta := range newPostMetas {
+		_found := false
+		for all_idx, allPostMeta := range allPostMetas {
+			if newPostMeta.URL == allPostMeta.URL {
+				_found = true
+				founds++
+				// replace the old
+				Logger.Println("dedup_append: replace the old", allPostMeta, "with", newPostMeta)
+				allPostMetas[all_idx] = newPostMeta
+				break
+			}
+		}
+		if !_found {
+			allPostMetas = append(allPostMetas, newPostMeta)
+		}
+	}
+	return allPostMetas, founds
+}
+
+var DUP_THRESHOLD_RATE = 3
+
 func ProcesserWorker(i int, tracker *savewebtracker.Tracker) {
 	Logger.Println("[START] ProcesserWorker", i)
 	defer Logger.Println("[STOP] ProcesserWorker", i, " exited...")
@@ -83,7 +107,13 @@ func ProcesserWorker(i int, tracker *savewebtracker.Tracker) {
 			Logger.Panicln(head, err)
 		}
 		all_postMetas := []cnblogs_api.PostMeta{}
+		dups_found := 0
 		for page := 1; ; page++ {
+			if dups_found > len(all_postMetas)*DUP_THRESHOLD_RATE {
+				Logger.Println(head, "Dups found", dups_found, "exceeds the threshold", DUP_THRESHOLD_RATE, ", break")
+				break
+			}
+
 			Logger.Println(head, "Processing", blogURI, "page:", page, "Got:", len(all_postMetas))
 			htmlBody, statusCode, err := cnblogs_api.GetBlogHomepage(tracker.HTTP_client, blogURI, page)
 			if err != nil {
@@ -104,7 +134,9 @@ func ProcesserWorker(i int, tracker *savewebtracker.Tracker) {
 				break
 			}
 			Logger.Println(head, "Got", postMetas)
-			all_postMetas = append(all_postMetas, postMetas...)
+			_all_postMetas, _founds_in_the_page := dedup_append(all_postMetas, postMetas)
+			dups_found += _founds_in_the_page
+			all_postMetas = _all_postMetas
 		}
 
 		items := []savewebtracker.Item{}
@@ -170,7 +202,7 @@ func ShowStatus(t *savewebtracker.Tracker) {
 }
 
 func main() {
-	tracker := savewebtracker.GetTracker(project_id, "0.2", savewebtracker.Archivist())
+	tracker := savewebtracker.GetTracker(project_id, "0.3.0", savewebtracker.Archivist())
 	tracker.PING_client = GetRetryableHttpClient(10*time.Second, DEBUG)
 	// tracker.HTTP_client = GetRetryableHttpClient(10*time.Second, DEBUG)
 	tracker.SelectBestTracker().StartSelectTrackerBackground().StartFetchProjectBackground()

pkg/cnblogs_api.go

@@ -145,7 +145,6 @@ func EnsureHomepageOK(htmldata string) bool {
 		"currentBlogApp",
 		"application/rss+xml",
 		"antiforgery_token",
-		"poweredby",
 	}
 	for _, flag := range Flags {
 		if !strings.Contains(htmldata, flag) {