Problem with curvy quotations marks when doing a merge.

[Diego Malo]

Hello,

I am using Stata18. I am attempting to merge with string variables. However, some of the string variables in my master data have curly quotation marks like ” or like ’. I cannot replace them directly in my do file as when I tried to replace them, it appears directly " instead of ” even I copy the ” directly.

Then, I am trying different solutions but none of them have worked. Here are some of the attempts I made:

First solution:

replace language_name = subinstr(language_name, "\u201D", `"\""', .)

Second solution:

local quote "\""
replace language_name = subinstr(language_name, "\u201D", "`quote'", .)

These attempts resulted in errors of invalid syntax or did not have any effect. Any alternative to deal with this issue?

Thank you.

[Nick Cox]

I would try

replace language_name = subinstr(language_name, XXX, YYY , .) where XXX and YYY are replaced by calls to uchar() or ustrunescape() chartab from SSC can help mightily here.

[Diego Malo]

Thanks a lot Nick Cox

I could solve the issue of the ” with the following code:

replace language_name = subinstr(language_name, uchar(8221), uchar(34), .)

However, I am still finding issues with ’. I am using the followind code:

replace language_name = subinstr(language_name, uchar(2019), uchar(27), .)

However, It does not have any effect. Do you know why?

[Nick Cox]

That looks like uchar(39) to me. If it's something else, then as said chartab from SSC will tell you what you have.

[Diego Malo]

Thanks a lot for the help. It works! I write the code I did in case it can be helpful to anyone else:

ssc install chartab

chartab language_name, noascii


   decimal  hexadecimal   character |     frequency    unique name
------------------------------------+---------------------------------------------
     8,217       \u2019       ’     |           207    RIGHT SINGLE QUOTATION MARK


replace language_name = subinstr(language_name, uchar(8217), uchar(27), .)