Fix using cached WACZ filename if already set ahead of time. (#783)

- if <uid>:nextWacz filename already exists, actually get it and use
that!
- don't merge cdx if not generating wacz yet, use same condition for
both bump version to 1.5.8
- fix follow-up to #748, fix #747

package.json

@@ -1,6 +1,6 @@
 {
   "name": "browsertrix-crawler",
-  "version": "1.5.7",
+  "version": "1.5.8",
   "main": "browsertrix-crawler",
   "type": "module",
   "repository": "https://github.com/webrecorder/browsertrix-crawler",

src/crawler.ts

@@ -1730,9 +1730,13 @@ self.__bx_behaviors.selectMainBehavior();
       await this.combineWARC();
     }
 
+    const generateFiles =
+      !this.params.dryRun &&
+      (!this.interruptReason || this.finalExit || this.uploadAndDeleteLocal);
+
     if (
       (this.params.generateCDX || this.params.generateWACZ) &&
-      !this.params.dryRun
+      generateFiles
     ) {
       logger.info("Merging CDX");
       await this.crawlState.setStatus(
@@ -1746,11 +1750,7 @@ self.__bx_behaviors.selectMainBehavior();
       );
     }
 
-    if (
-      this.params.generateWACZ &&
-      !this.params.dryRun &&
-      (!this.interruptReason || this.finalExit || this.uploadAndDeleteLocal)
-    ) {
+    if (this.params.generateWACZ && generateFiles) {
       const uploaded = await this.generateWACZ();
 
       if (uploaded && this.uploadAndDeleteLocal) {

src/util/state.ts

@@ -420,12 +420,30 @@ return inx;
   async setWACZFilename(): Promise<string> {
     const filename = process.env.STORE_FILENAME || "@ts-@id.wacz";
     this.waczFilename = interpolateFilename(filename, this.key);
-    await this.redis.hsetnx(
-      `${this.key}:nextWacz`,
-      this.uid,
-      this.waczFilename,
-    );
-    return this.waczFilename;
+    if (
+      !(await this.redis.hsetnx(
+        `${this.key}:nextWacz`,
+        this.uid,
+        this.waczFilename,
+      ))
+    ) {
+      this.waczFilename = await this.redis.hget(
+        `${this.key}:nextWacz`,
+        this.uid,
+      );
+      logger.debug(
+        "Keeping WACZ Filename",
+        { filename: this.waczFilename },
+        "state",
+      );
+    } else {
+      logger.debug(
+        "Using New WACZ Filename",
+        { filename: this.waczFilename },
+        "state",
+      );
+    }
+    return this.waczFilename!;
   }
 
   async getWACZFilename(): Promise<string> {