Add TweetTrove#FillMissingUserIDs helper method

scraper/tweet_trove.go

@@ -1,5 +1,9 @@
 package scraper
 
+import (
+	"fmt"
+)
+
 type TweetTrove struct {
 	Tweets    map[TweetID]Tweet
 	Users     map[UserID]User
@@ -30,7 +34,6 @@ func (trove TweetTrove) Transform() (tweets []Tweet, retweets []Retweet, users [
 	for _, val := range trove.Retweets {
 		retweets = append(retweets, val)
 	}
-
 	return
 }  // TODO: refactor until this function isn't needed anymore
 
@@ -51,3 +54,40 @@ func (t1 *TweetTrove) MergeWith(t2 TweetTrove) {
 
 	t1.TombstoneUsers = append(t1.TombstoneUsers, t2.TombstoneUsers...)
 }
+
+/**
+ * Checks for tombstoned tweets and fills in their UserIDs based on the collected tombstoned users.
+
+ * To be called after calling "scraper.GetUser" on all the tombstoned users.
+ *
+ * At this point, those users should have been added to this trove's Users collection, and the
+ * Tweets have a field `UserHandle` which can be used to pair them with newly fetched Users.
+ *
+ * This will still fail if the user deleted their account (instead of getting banned, blocking the
+ * quote-tweeter, etc), because then that user won't show up .
+ */
+func (trove *TweetTrove) FillMissingUserIDs() {
+	for i := range trove.Tweets {
+		tweet := trove.Tweets[i]
+		if tweet.UserID != 0 {
+			// No need to fill this tweet's user_id, it's already filled
+			continue
+		}
+
+		handle := tweet.UserHandle
+		is_user_found := false
+		for _, u := range trove.Users {
+			if u.Handle == handle {
+				tweet.UserID = u.ID
+				is_user_found = true
+				break
+			}
+		}
+		if !is_user_found {
+			// The user probably deleted deleted their account, and thus `scraper.GetUser` failed.  So
+			// they're not in this trove's Users.
+			panic(fmt.Sprintf("Couldn't fill out this Tweet's UserID: %d, %s", tweet.ID, tweet.UserHandle))
+		}
+		trove.Tweets[i] = tweet
+	}
+}

scraper/tweet_trove_test.go

@@ -47,3 +47,22 @@ func TestMergeTweetTroves(t *testing.T) {
 	assert.Equal(3, len(trove1.Retweets))
 	assert.Equal(3, len(trove1.TombstoneUsers))
 }
+
+func TestFillMissingUserIDs(t *testing.T) {
+	assert := assert.New(t)
+	u1 := User{ID: 1, Handle: "a"}
+
+	t1 := Tweet{ID: 1, UserID: 1}
+	t2 := Tweet{ID: 2, UserHandle: "a"}
+
+	trove := NewTweetTrove()
+	trove.Users[u1.ID] = u1
+	trove.Tweets[t1.ID] = t1
+	trove.Tweets[t2.ID] = t2
+
+	assert.NotEqual(trove.Tweets[2].UserID, UserID(1))
+
+	trove.FillMissingUserIDs()
+
+	assert.Equal(trove.Tweets[2].UserID, UserID(1))
+}