How to convert CString to string without loosing any characters

Question

How to convert CString to string without loosing any characters

sarath govind 1

I have a MFC CString variable , I want to convert it to a string . Actually this CString variable holds a value which have many junk characters like this (!(¾�ÊO€N„+ E^]«��3àâ;bÛ) in between . I have tried possibilities like CT2CA , LPCTSTR etc. These conversion results to string having length not same as CString length . There is large difference between the sizes. Looking forward a method which does the correct conversion without loosing any characters

David Lowndes 4,726 Reputation points

2021-07-28T13:18:30.647+00:00

Any conversion between a Unicode and ANSI/MBCS string will potentially loose characters (they'' usually be replaced with '?')that aren't in the code page you're converting to.
Your subject mentions null characters, which implies that your string has several null separated strings, though you fail to mention that in the question, so I'm wondering whether that is the case or not?
Clarify your question with some specific details so that someone can give you a more informed answer.
sarath govind 1 Reputation point

2021-07-28T13:29:23.61+00:00

modified the question. thank you
Viorel 122.5K Reputation points

2021-07-28T14:10:03.86+00:00

It is not clear what kind of CString do you have (CStringA or CStringW). If it is a CStringW, it can be converted to char* or CStringA strings using UTF-8 encoding and WideCharToMultiByte function (assuming that the source string is valid). The length in bytes will not be the same as CString length.

How did you get such an unnatural string? Maybe you should avoid this.
David Lowndes 4,726 Reputation points

2021-07-28T16:09:13.46+00:00

I'm afraid it's still not clear to me.
You say "junk" and "between", which implies characters you don't want intermixed with others you do want. How have you got into that situation - if that is the situation?
You say you have a CString. Are you building for Unicode, in which case it's a wide (Unicode) CStringW, or ANSI/MBCS, in which case it's a CStringA.
Fundamentally we need more details - perhaps a short example that fully illustrates the issue is needed.
sarath govind 1 Reputation point

2021-07-29T07:38:21.71+00:00

Actually the thing is that, I am having an AES encrypted text it consists of junk characters and '\0' character . This \0 character is there in between the CString value and at the end of the CString too. This is creating the issue when I convert CString to another string type.

while performing conversion, it reads up to the first '\0' characters. The remaining characters are not copying. Is there any other way to achieve this ?
sarath govind 1 Reputation point

2021-07-29T07:39:19.47+00:00

Actually the thing is that, I am having an AES encrypted text it consists of junk characters and '\0' character . This \0 character is there in between the CString value and at the end of the CString too. This is creating the issue when I convert CString to another string type.

while performing conversion, it reads up to the first '\0' characters. The remaining characters are not copying. Is there any other way to achieve this ?
Viorel 122.5K Reputation points

2021-07-29T08:19:53.633+00:00

For encrypted bytes you can use an array of bytes (CByteArray, or *char** and int to keep the length). It is not clear why do you need a string instead of array of bytes. You can convert a byte array to hex string (see AtlHexEncode) or to Base64 string.
David Lowndes 4,726 Reputation points

2021-07-29T08:52:27.827+00:00

As Viorel-1 says, if you've got output from encryption, you should have a byte array, not a string. It would only be a displayable string if it'd been converted to a specific string representation such as base64 or hex format (which your example clearly wasn't). See this site for an example.

What you may have, is the raw byte array stored in a CString (it ought to be a CStringA), and that's confusing matters for you.

We don't know why you need to convert this to another string form, but you may need to convert it to a base64 (or hex) string - you'll have to ask someone at your end who knows what they want you to give them.

1 answer

Your answer

David Lowndes 4,726 Reputation points

2021-07-28T13:18:30.647+00:00

Any conversion between a Unicode and ANSI/MBCS string will potentially loose characters (they'' usually be replaced with '?')that aren't in the code page you're converting to.
Your subject mentions null characters, which implies that your string has several null separated strings, though you fail to mention that in the question, so I'm wondering whether that is the case or not?
Clarify your question with some specific details so that someone can give you a more informed answer.
sarath govind 1 Reputation point

2021-07-28T13:29:23.61+00:00

modified the question. thank you
Viorel 122.5K Reputation points

2021-07-28T14:10:03.86+00:00

It is not clear what kind of CString do you have (CStringA or CStringW). If it is a CStringW, it can be converted to char* or CStringA strings using UTF-8 encoding and WideCharToMultiByte function (assuming that the source string is valid). The length in bytes will not be the same as CString length.

How did you get such an unnatural string? Maybe you should avoid this.
David Lowndes 4,726 Reputation points

2021-07-28T16:09:13.46+00:00

I'm afraid it's still not clear to me.
You say "junk" and "between", which implies characters you don't want intermixed with others you do want. How have you got into that situation - if that is the situation?
You say you have a CString. Are you building for Unicode, in which case it's a wide (Unicode) CStringW, or ANSI/MBCS, in which case it's a CStringA.
Fundamentally we need more details - perhaps a short example that fully illustrates the issue is needed.
sarath govind 1 Reputation point

2021-07-29T07:38:21.71+00:00

Actually the thing is that, I am having an AES encrypted text it consists of junk characters and '\0' character . This \0 character is there in between the CString value and at the end of the CString too. This is creating the issue when I convert CString to another string type.

while performing conversion, it reads up to the first '\0' characters. The remaining characters are not copying. Is there any other way to achieve this ?
sarath govind 1 Reputation point

2021-07-29T07:39:19.47+00:00

Actually the thing is that, I am having an AES encrypted text it consists of junk characters and '\0' character . This \0 character is there in between the CString value and at the end of the CString too. This is creating the issue when I convert CString to another string type.

while performing conversion, it reads up to the first '\0' characters. The remaining characters are not copying. Is there any other way to achieve this ?
Viorel 122.5K Reputation points

2021-07-29T08:19:53.633+00:00

For encrypted bytes you can use an array of bytes (CByteArray, or *char** and int to keep the length). It is not clear why do you need a string instead of array of bytes. You can convert a byte array to hex string (see AtlHexEncode) or to Base64 string.
David Lowndes 4,726 Reputation points

2021-07-29T08:52:27.827+00:00

As Viorel-1 says, if you've got output from encryption, you should have a byte array, not a string. It would only be a displayable string if it'd been converted to a specific string representation such as base64 or hex format (which your example clearly wasn't). See this site for an example.

What you may have, is the raw byte array stored in a CString (it ought to be a CStringA), and that's confusing matters for you.

We don't know why you need to convert this to another string form, but you may need to convert it to a base64 (or hex) string - you'll have to ask someone at your end who knows what they want you to give them.

Answer 1

Guido Franzke 2,191

Hello,
I assume, you build in Unicode (because ANSI does not make sense in my opinion).
So CString is a CStringW. std::string is always of type char. You should convert the CStringW to std::wstring. Read the documentation: basic_string

If this is not the case: as others have already mentioned before, give us a short example what you want to do.

Regards, Guido

sarath govind 1 Reputation point

2021-07-29T07:39:03.3+00:00

Actually the thing is that, I am having an AES encrypted text it consists of junk characters and '\0' character . This \0 character is there in between the CString value and at the end of the CString too. This is creating the issue when I convert CString to another string type.

while performing conversion, it reads up to the first '\0' characters. The remaining characters are not copying. Is there any other way to achieve this ?
Guido Franzke 2,191 Reputation points

2021-07-29T08:57:54.507+00:00

Accordingly, you must use a memory buffer. Do not use string classes like CString, std::string,... for a memory buffer. Strings will per definition always end at the (first) '\0' character.

Share via

How to convert CString to string without loosing any characters

1 answer

Your answer