1) By "Scrapers don't save metadata properly" in my last commit, I meant only URL and accessDate, though on second thought they probably will work.

1b) However, I also did, in fact, break scraping completely, so my previous statement was actually correct. Fix for that coming right up.

2) Fixed problem with translators table getting wiped out completely whenever system.sql was updated (from r671, I believe). Right. Moved the DROP and CREATE statements for translators into translators.sql.

scrapers.sql

@@ -1,4 +1,20 @@
--- 92
+-- 93
+
+DROP TABLE IF EXISTS translators;
+CREATE TABLE translators (
+    translatorID TEXT PRIMARY KEY,
+    lastUpdated DATETIME,
+    inRepository INT,
+    priority INT,
+    translatorType INT,
+    label TEXT,
+    creator TEXT,
+    target TEXT,
+    detectCode TEXT,
+    code TEXT
+);
+DROP INDEX IF EXISTS translators_type;
+CREATE INDEX translators_type ON translators(translatorType);
 
 -- Set the following timestamp to the most recent scraper update date
 REPLACE INTO "version" VALUES ('repository', STRFTIME('%s', '2006-10-01 17:00:00'));

system.sql

@@ -76,22 +76,6 @@
         creatorType TEXT
     );
     
-    DROP TABLE IF EXISTS translators;
-    CREATE TABLE translators (
-        translatorID TEXT PRIMARY KEY,
-        lastUpdated DATETIME,
-        inRepository INT,
-        priority INT,
-        translatorType INT,
-        label TEXT,
-        creator TEXT,
-        target TEXT,
-        detectCode TEXT,
-        code TEXT
-    );
-    DROP INDEX IF EXISTS translators_type;
-    CREATE INDEX translators_type ON translators(translatorType);
-    
     DROP TABLE IF EXISTS transactionSets;
     CREATE TABLE transactionSets (
         transactionSetID INTEGER PRIMARY KEY,