在For循环之后清除Python中的元组

我正在努力弄清楚如何在从数据库读取后清除元组“数据”

工艺流程：

Update has started
((10938L, u"@anonymized @anonymized House of Cards will be nothing compared  to the Drumpf's!And worst of all is that Americans chose him as president?", u'New York, USA'),)
Update has started
((10938L, u"@anonymized @anonymized House of Cards will be nothing compared to the Drumpf's!And worst of all is that Americans chose him as president?", u'New York, USA'),)
Update has started
((10938L, u"@anonymized @anonymized House of Cards will be nothing compared to the Drumpf's!And worst of all is that Americans chose him as president?", u'New York, USA'),)
Update has started
((10938L, u"@anonymized @anonymized House of Cards will be nothing compared to the Drumpf's!And worst of all is that Americans chose him as president?", u'New York, USA'),)

class analyzeRecords():
    def batchUpdate(self):

     global data

     #Select up to 1 record

     b.execute(""" SELECT id, tweet,tweet_location_guess FROM rawTweets where compoundScore IS NULL LIMIT 0,1 """)

     #Put that data into a tuple

     data = b.fetchall()

     print(data)

     #print that update has started

     print("Update has started")

     for row in data:
          idMatch = row[0]
          cleanTweet = reduce(lambda x, y: x.replace(y, d[y]), sorted(d, key=lambda i: len(i), reverse=True), row[1].lower())
          sentimentScores = analyzer.polarity_scores(cleanTweet)
          overallScore = sentimentScores["compound"]

          u.execute("""UPDATE rawTweets SET tweet=%s, compoundScore=%s where id=%s""",
                    (cleanTweet, overallScore, idMatch))
          update.commit()

l = task.LoopingCall(analyzeRecords().batchUpdate)
l.start(timeout) # call every sixty seconds

reactor.run()       

每隔X分钟，我就会调用batchUpdate

batchUpdate拉入符合特定条件的记录

我们迭代这些记录以执行更新

进程结束并等待下一个调用

问题：

Update has started
((10938L, u"@anonymized @anonymized House of Cards will be nothing compared  to the Drumpf's!And worst of all is that Americans chose him as president?", u'New York, USA'),)
Update has started
((10938L, u"@anonymized @anonymized House of Cards will be nothing compared to the Drumpf's!And worst of all is that Americans chose him as president?", u'New York, USA'),)
Update has started
((10938L, u"@anonymized @anonymized House of Cards will be nothing compared to the Drumpf's!And worst of all is that Americans chose him as president?", u'New York, USA'),)
Update has started
((10938L, u"@anonymized @anonymized House of Cards will be nothing compared to the Drumpf's!And worst of all is that Americans chose him as president?", u'New York, USA'),)

class analyzeRecords():
    def batchUpdate(self):

     global data

     #Select up to 1 record

     b.execute(""" SELECT id, tweet,tweet_location_guess FROM rawTweets where compoundScore IS NULL LIMIT 0,1 """)

     #Put that data into a tuple

     data = b.fetchall()

     print(data)

     #print that update has started

     print("Update has started")

     for row in data:
          idMatch = row[0]
          cleanTweet = reduce(lambda x, y: x.replace(y, d[y]), sorted(d, key=lambda i: len(i), reverse=True), row[1].lower())
          sentimentScores = analyzer.polarity_scores(cleanTweet)
          overallScore = sentimentScores["compound"]

          u.execute("""UPDATE rawTweets SET tweet=%s, compoundScore=%s where id=%s""",
                    (cleanTweet, overallScore, idMatch))
          update.commit()

l = task.LoopingCall(analyzeRecords().batchUpdate)
l.start(timeout) # call every sixty seconds

reactor.run()       

对batchUpdate函数的每次后续调用都不会产生新数据。元组“data”包含与以前相同的值

简化示例（仅拉取一条记录，每1秒安排一次）：

Update has started
((10938L, u"@anonymized @anonymized House of Cards will be nothing compared  to the Drumpf's!And worst of all is that Americans chose him as president?", u'New York, USA'),)
Update has started
((10938L, u"@anonymized @anonymized House of Cards will be nothing compared to the Drumpf's!And worst of all is that Americans chose him as president?", u'New York, USA'),)
Update has started
((10938L, u"@anonymized @anonymized House of Cards will be nothing compared to the Drumpf's!And worst of all is that Americans chose him as president?", u'New York, USA'),)
Update has started
((10938L, u"@anonymized @anonymized House of Cards will be nothing compared to the Drumpf's!And worst of all is that Americans chose him as president?", u'New York, USA'),)

class analyzeRecords():
    def batchUpdate(self):

     global data

     #Select up to 1 record

     b.execute(""" SELECT id, tweet,tweet_location_guess FROM rawTweets where compoundScore IS NULL LIMIT 0,1 """)

     #Put that data into a tuple

     data = b.fetchall()

     print(data)

     #print that update has started

     print("Update has started")

     for row in data:
          idMatch = row[0]
          cleanTweet = reduce(lambda x, y: x.replace(y, d[y]), sorted(d, key=lambda i: len(i), reverse=True), row[1].lower())
          sentimentScores = analyzer.polarity_scores(cleanTweet)
          overallScore = sentimentScores["compound"]

          u.execute("""UPDATE rawTweets SET tweet=%s, compoundScore=%s where id=%s""",
                    (cleanTweet, overallScore, idMatch))
          update.commit()

l = task.LoopingCall(analyzeRecords().batchUpdate)
l.start(timeout) # call every sixty seconds

reactor.run()       

代码：

Update has started
((10938L, u"@anonymized @anonymized House of Cards will be nothing compared  to the Drumpf's!And worst of all is that Americans chose him as president?", u'New York, USA'),)
Update has started
((10938L, u"@anonymized @anonymized House of Cards will be nothing compared to the Drumpf's!And worst of all is that Americans chose him as president?", u'New York, USA'),)
Update has started
((10938L, u"@anonymized @anonymized House of Cards will be nothing compared to the Drumpf's!And worst of all is that Americans chose him as president?", u'New York, USA'),)
Update has started
((10938L, u"@anonymized @anonymized House of Cards will be nothing compared to the Drumpf's!And worst of all is that Americans chose him as president?", u'New York, USA'),)

class analyzeRecords():
    def batchUpdate(self):

     global data

     #Select up to 1 record

     b.execute(""" SELECT id, tweet,tweet_location_guess FROM rawTweets where compoundScore IS NULL LIMIT 0,1 """)

     #Put that data into a tuple

     data = b.fetchall()

     print(data)

     #print that update has started

     print("Update has started")

     for row in data:
          idMatch = row[0]
          cleanTweet = reduce(lambda x, y: x.replace(y, d[y]), sorted(d, key=lambda i: len(i), reverse=True), row[1].lower())
          sentimentScores = analyzer.polarity_scores(cleanTweet)
          overallScore = sentimentScores["compound"]

          u.execute("""UPDATE rawTweets SET tweet=%s, compoundScore=%s where id=%s""",
                    (cleanTweet, overallScore, idMatch))
          update.commit()

l = task.LoopingCall(analyzeRecords().batchUpdate)
l.start(timeout) # call every sixty seconds

reactor.run()       

---

在代码中，执行以下两行：

 global data

 data = b.fetchall()

综上所述，这两条语句应该覆盖先前

数据
变量中的任何内容

我要指出的是，这个函数似乎不需要全局变量——您可以而且可能应该使用局部变量

无论如何，我不认为问题在于存在一些神秘的遗留数据，除非定义了
b.fetchall（）
对象来实现这一点。（例如，如果查询中有错误，或者查询没有返回匹配项，它是如何进行通信的？如果它引发或返回一个您忽略的值，则可能
fetchall
可能会返回过时的数据，因为您应该检查该值，而不是调用fetchall…）

我建议您看看
execute
和
fetchall
是如何协同工作的，还可以看看for循环。我看到
b.execute
和
u.execute
和
update.commit
。看起来您有许多不同的数据库连接。也许你喜欢。或者你复制/粘贴代码，你真的应该这样做：

u.execute(...)
u.commit()

我可能是误会了。您能在函数末尾简单地设置
data=None
吗？但是
b.execute
返回不同的数据吗？有什么理由将数据定义为全局的吗？我对数据库连接/游标很笨。问题解决了。谢谢