在Big Query中使用正則表達式來提取URL

Question

我一直試圖提取Big Query中我的「文本」列中存在的任何URL。列包含文本和遍佈整個（一個細胞可能包含多個URL）我試圖用這個正則表達式的URL的混合物：

SELECT 

    REGEXP_EXTRACT (Text, r'(http(s)?:\/\/.)?(www\.)?[-a-zA-Z0-9:%._\+~#=]{2,256}\.[a-z]{2,6}\b([-a-zA-Z0-9%_:?\+.~#&//=]*') 
FROM 
Data.Text_Files 

我目前得到「無法解析正則表達式」當我嘗試運行查詢。我試過修改它，但無濟於事。

正則表達式在一個在線構建器中工作，但我不確定如何將它合併到Big Query中。

任何幫助將不勝感激 - 或者至少指出如何將正則表達式合併到Big Query中！

Answer 1

嘗試以下 - 這是BigQuery的標準SQL（見Enabling Standard SQL和Migrating from legacy SQL）

WITH YourTable AS (
    SELECT 1 AS id, 'What have you tried so far? Please edit your question to show a [Minimal, Complete, and Verifiable example](http://stackoverflow.com/help/mcve) of the code that you are having problems with, then we can try to help with the specific problem. You can also read [How to Ask](http://stackoverflow.com/help/how-to-ask). ' AS Text UNION ALL 
    SELECT 2 AS id, 'Important on SO, you can mark accepted answer by using the tick on the left of the posted answer, below the voting. see http://meta.stackexchange.com/questions/5234/how-does-accepting-an-answer-work#5235 for why it is important. There are more ... You can check about what to do when someone answers your question - http://stackoverflow.com/help/someone-answers.' AS Text UNION ALL 
    SELECT 3 AS id, 'If an answer has helped you solve your problem and you accept it you should also consider voting it up. See more at http://stackoverflow.com/help/someone-answers and Upvote section in http://meta.stackexchange.com/questions/5234/how-does-accepting-an-answer-work#5235' AS Text 
) 
SELECT 
id, 
REGEXP_EXTRACT_ALL(Text, r'(?i:(?:(?:(?:ftp|https?):\/\/)(?:www\.)?|www\.)(?:[\da-z-_\.]+)(?:[a-z\.]{2,7})(?:[\/\w\.-_\?\&]*)*\/?)') AS URL 
FROM YourTable 

這爲您提供了id字段的輸出，並多次實地與所有相應的網址

如果你需要扁平結果 - 你可以使用下面的變化

WITH YourTable AS (
    SELECT 1 AS id, 'What have you tried so far? Please edit your question to show a [Minimal, Complete, and Verifiable example](http://stackoverflow.com/help/mcve) of the code that you are having problems with, then we can try to help with the specific problem. You can also read [How to Ask](http://stackoverflow.com/help/how-to-ask). ' AS Text UNION ALL 
    SELECT 2 AS id, 'Important on SO, you can mark accepted answer by using the tick on the left of the posted answer, below the voting. see http://meta.stackexchange.com/questions/5234/how-does-accepting-an-answer-work#5235 for why it is important. There are more ... You can check about what to do when someone answers your question - http://stackoverflow.com/help/someone-answers.' AS Text UNION ALL 
    SELECT 3 AS id, 'If an answer has helped you solve your problem and you accept it you should also consider voting it up. See more at http://stackoverflow.com/help/someone-answers and Upvote section in http://meta.stackexchange.com/questions/5234/how-does-accepting-an-answer-work#5235' AS Text 
) 
SELECT 
    id, URL  
FROM (
    SELECT id, REGEXP_EXTRACT_ALL(Text, r'(?i:(?:(?:(?:ftp|https?):\/\/)(?:www\.)?|www\.)(?:[\da-z-_\.]+)(?:[a-z\.]{2,7})(?:[\/\w\.-_\?\&]*)*\/?)') AS URL 
    FROM YourTable 
), UNNEST(URL) as URL 

注意：你可以在這裏使用任何正則表達式，你可以在網上找到 - 但是什麼必須是 - 只有一個匹配組是允許的！所以所有內部匹配組都應該使用?:進行轉義，您可以在上面的示例中看到它。因此，您希望在輸出中看到的唯一組應該保持原樣 - 無?:

Answer 2

你的正則表達式有一個不完整的捕獲組，並且有2個非轉義字符。我不知道你使用的是哪種在線正則表達式構建器，但是也許你忘了將新的正則表達式放入其中？是

存在的問題如下：

(http(s)?:\/\/.)?(www\.)?[-a-zA-Z0-9:%._\+~#=]{2,256}\.[a-z]{2,6}\b([-a-zA-Z0-9%_:?\+.~#&//=]* 
POINTERS TO PROBLEMS ON THIS LINE --->        ^1     ^^2 

1. 這是一個沒有結束捕獲組的開始。您可能需要在*之前)。
2. 所有斜槓都需要轉義。這應該是\/或甚至\/\\。

這裏是我的兩個建議實施的示例：https://regex101.com/r/pt1hqS/1

好運修復它！