Forum Discussion
Solved
Removing HTML tags and reordering text
Howdy folks, I have a situation where a text string is being imported that could contain hyperlinks (it might not contain any, or it could contain several). What I want to do is extract the text and...
Hi afhealey ,
You can splilt column many times by specific string and filter the specfic string to achieve this in power query. The whole query would be like this:
let Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45Wiik1MDBOLgBTqSEZmcUKQFScn5uqUJJaUQKR1YdKx+ShqIZwEhUyilLTbGOUMkpKCqz09cvLy/XS8/PTc1L1kvNzY5QUilJzgLJ5+Wn5OTn55TEQG1PdwUoUcjLzsqGWJCKbisNKBTIdlJSZlw5yjj5O9zgBVRB0jVJsLAA=", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [Data = _t]), #"Changed Type" = Table.TransformColumnTypes(Source,{{"Data", type text}}), #"Split Column by Delimiter" = Table.ExpandListColumn(Table.TransformColumns(#"Changed Type", {{"Data", Splitter.SplitTextByDelimiter("<p>", QuoteStyle.Csv), let itemType = (type nullable text) meta [Serialized.Text = true] in type {itemType}}}), "Data"), #"Split Column by Delimiter1" = Table.ExpandListColumn(Table.TransformColumns(#"Split Column by Delimiter", {{"Data", Splitter.SplitTextByDelimiter(" ", QuoteStyle.Csv), let itemType = (type nullable text) meta [Serialized.Text = true] in type {itemType}}}), "Data"), #"Changed Type1" = Table.TransformColumnTypes(#"Split Column by Delimiter1",{{"Data", type text}}), #"Filtered Rows" = Table.SelectRows(#"Changed Type1", each Text.Contains([Data], "href")), #"Split Column by Delimiter2" = Table.ExpandListColumn(Table.TransformColumns(#"Filtered Rows", {{"Data", Splitter.SplitTextByDelimiter("=", QuoteStyle.Csv), let itemType = (type nullable text) meta [Serialized.Text = true] in type {itemType}}}), "Data"), #"Changed Type2" = Table.TransformColumnTypes(#"Split Column by Delimiter2",{{"Data", type text}}), #"Filtered Rows1" = Table.SelectRows(#"Changed Type2", each not Text.Contains([Data], "href")) in #"Filtered Rows1"Best Regards,
Community Support Team _ Yingjie Li
If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.
I really doubt Power Query would cope with it with ease considering following cases in practice, CAPITAL letters, arbitary blank spaces, etc,
<A HREF = "http://www.google.com" REL="nofollow">Google link</A>
<a href="http://www.google.com" rel="nofollow">Google link</a>
<a href= 'http://www.yahoo.com' rel="nofollow">Yahoo link</a>Resort to regular expression with embedded R or Python script using such a pattern,
(?<=href)(?:[^"\']*)(?P<DELIM>["\'])(?P<LINK>.*?)(?P=DELIM)