After noticing an application tended to discard random emails due to incorrect string value errors, I went though and switched many text columns to use the utf8 column charset and the default column collate (utf8_general_ci) so that it would accept them. This fixed most of the errors, and made the application stop getting sql errors when it hit non-latin emails, too.
Despite this, some of the emails are still causing the program to hit incorrect string value errrors: (Incorrect string value: '\xE4\xC5\xCC\xC9\xD3\xD8...' for column 'contents' at row 1)
The contents column is a MEDIUMTEXT datatybe which uses the utf8 column charset and the utf8_general_ci column collate. There are no flags that I can toggle in this column.
Keeping in mind that I don't want to touch or even look at the application source code unless absolutely necessary:
What is causing that error? (yes, I know the emails are full of random garbage, but I thought utf8 would be pretty permissive)
How can I fix it?
What are the likely effects of such a fix?
One thing I considered was switching to a utf8 varchar([some large number]) with the binary flag turned on, but I'm rather unfamiliar with MySQL, and have no idea if such a fix makes sense.
UPDATE to the below answer:
The time the question was asked, "UTF8" in MySQL meant utf8mb3. In the meantime, utf8mb4 was added, but to my knowledge MySQLs "UTF8" was not switched to mean utf8mb4.
That means, you'd need to specifically put "utf8mb4", if you mean it (and you should use utf8mb4)
I'll keep this here instead of just editing the answer, to make clear there is still a difference when saying "UTF8"
Original
I would not suggest Richies answer, because you are screwing up the data inside the database. You would not fix your problem but try to "hide" it and not being able to perform essential database operations with the crapped data.
If you encounter this error either the data you are sending is not UTF-8 encoded, or your connection is not UTF-8. First, verify, that the data source (a file, ...) really is UTF-8.
Then, check your database connection, you should do this after connecting:
SET NAMES 'utf8mb4';
SET CHARACTER SET utf8mb4;
Next, verify that the tables where the data is stored have the utf8mb4 character set:
SELECT
`tables`.`TABLE_NAME`,
`collations`.`character_set_name`
FROM
`information_schema`.`TABLES` AS `tables`,
`information_schema`.`COLLATION_CHARACTER_SET_APPLICABILITY` AS `collations`
WHERE
`tables`.`table_schema` = DATABASE()
AND `collations`.`collation_name` = `tables`.`table_collation`
;
Last, check your database settings:
mysql> show variables like '%colla%';
mysql> show variables like '%charac%';
If source, transport and destination are utf8mb4, your problem is gone;)
MySQL’s utf-8 types are not actually proper utf-8 – it only uses up to three bytes per character and supports only the Basic Multilingual Plane (i.e. no Emoji, no astral plane, etc.).
If you need to store values from higher Unicode planes, you need the utf8mb4 encodings.
The table and fields have the wrong encoding; however, you can convert them to UTF-8.
ALTER TABLE logtest CONVERT TO CHARACTER SET utf8 COLLATE utf8_general_ci;
ALTER TABLE logtest DEFAULT CHARACTER SET utf8 COLLATE utf8_general_ci;
ALTER TABLE logtest CHANGE title title VARCHAR(100) CHARACTER SET utf8 COLLATE utf8_general_ci;
"\xE4\xC5\xCC\xC9\xD3\xD8" isn't valid UTF-8. Tested using Python:
>>> "\xE4\xC5\xCC\xC9\xD3\xD8".decode("utf-8")
...
UnicodeDecodeError: 'utf8' codec can't decode bytes in position 0-2: invalid data
If you're looking for a way to avoid decoding errors within the database, the cp1252 encoding (aka "Windows-1252" aka "Windows Western European") is the most permissive encoding there is - every byte value is a valid code point.
Of course it's not going to understand genuine UTF-8 any more, nor any other non-cp1252 encoding, but it sounds like you're not too concerned about that?
I solved this problem today by altering the column to 'LONGBLOB' type which stores raw bytes instead of UTF-8 characters.
The only disadvantage of doing this is that you have to take care of the encoding yourself. If one client of your application uses UTF-8 encoding and another uses CP1252, you may have your emails sent with incorrect characters. To avoid this, always use the same encoding (e.g. UTF-8) across all your applications.
Refer to this page http://dev.mysql.com/doc/refman/5.0/en/blob.html for more details of the differences between TEXT/LONGTEXT and BLOB/LONGBLOB. There are also many other arguments on the web discussing these two.
First check if your default_character_set_name is utf8.
SELECT default_character_set_name FROM information_schema.SCHEMATA S WHERE schema_name = "DBNAME";
If the result is not utf8 you must convert your database. At first you must save a dump.
To change the character set encoding to UTF-8 for all of the tables in the specified database, type the following command at the command line. Replace DBNAME with the database name:
mysql --database=DBNAME -B -N -e "SHOW TABLES" | awk '{print "SET foreign_key_checks = 0; ALTER TABLE", $1, "CONVERT TO CHARACTER SET utf8 COLLATE utf8_general_ci; SET foreign_key_checks = 1; "}' | mysql --database=DBNAME
To change the character set encoding to UTF-8 for the database itself, type the following command at the mysql> prompt. Replace DBNAME with the database name:
ALTER DATABASE DBNAME CHARACTER SET utf8 COLLATE utf8_general_ci;
You can now retry to to write utf8 character into your database. This solution help me when i try to upload 200000 row of csv file into my database.
Although your collation is set to utf8_general_ci, I suspect that the character encoding of the database, table or even column may be different.
ALTER TABLE tabale_name MODIFY COLUMN column_name VARCHAR(255)
CHARACTER SET utf8 COLLATE utf8_general_ci NOT NULL;
In general, this happens when you insert strings to columns with incompatible encoding/collation.
I got this error when I had TRIGGERs, which inherit server's collation for some reason.
And mysql's default is (at least on Ubuntu) latin-1 with swedish collation.
Even though I had database and all tables set to UTF-8, I had yet to set my.cnf:
/etc/mysql/my.cnf :
[mysqld]
character-set-server=utf8
default-character-set=utf8
And this must list all triggers with utf8-*:
select TRIGGER_SCHEMA, TRIGGER_NAME, CHARACTER_SET_CLIENT, COLLATION_CONNECTION, DATABASE_COLLATION from information_schema.TRIGGERS
And some of variables listed by this should also have utf-8-* (no latin-1 or other encoding):
show variables like 'char%';
I got a similar error (Incorrect string value: '\xD0\xBE\xDO\xB2. ...' for 'content' at row 1). I have tried to change character set of column to utf8mb4 and after that the error has changed to 'Data too long for column 'content' at row 1'.
It turned out that mysql shows me wrong error. I turned back character set of column to utf8 and changed type of the column to MEDIUMTEXT. After that the error disappeared.
I hope it helps someone.
By the way MariaDB in same case (I have tested the same INSERT there) just cut a text without error.
That error means that either you have the string with incorrect encoding (e.g. you're trying to enter ISO-8859-1 encoded string into UTF-8 encoded column), or the column does not support the data you're trying to enter.
In practice, the latter problem is caused by MySQL UTF-8 implementation that only supports UNICODE characters that need 1-3 bytes when represented in UTF-8. See "Incorrect string value" when trying to insert UTF-8 into MySQL via JDBC? for details. The trick is to use column type utf8mb4 instead of type utf8 which doesn't actually support all of UTF-8 despite the name. The former type is the correct type to use for all UTF-8 strings.
In my case, Incorrect string value: '\xCC\x88'..., the problem was that an o-umlaut was in its decomposed state. This question-and-answer helped me understand the difference between o¨ and ö. In PHP, the fix for me was to use PHP's Normalizer library. E.g., Normalizer::normalize('o¨', Normalizer::FORM_C).
The solution for me when running into this Incorrect string value: '\xF8' for column error using scriptcase was to be sure that my database is set up for utf8 general ci and so are my field collations. Then when I do my data import of a csv file I load the csv into UE Studio then save it formatted as utf8 and Voila! It works like a charm, 29000 records in there no errors. Previously I was trying to import an excel created csv.
I have tried all of the above solutions (which all bring valid points), but nothing was working for me.
Until I found that my MySQL table field mappings in C# was using an incorrect type: MySqlDbType.Blob . I changed it to MySqlDbType.Text and now I can write all the UTF8 symbols I want!
p.s. My MySQL table field is of the "LongText" type. However, when I autogenerated the field mappings using MyGeneration software, it automatically set the field type as MySqlDbType.Blob in C#.
Interestingly, I have been using the MySqlDbType.Blob type with UTF8 characters for many months with no trouble, until one day I tried writing a string with some specific characters in it.
Hope this helps someone who is struggling to find a reason for the error.
If you happen to process the value with some string function before saving, make sure the function can properly handle multibyte characters. String functions that cannot do that and are, say, attempting to truncate might split one of the single multibyte characters in the middle, and that can cause such string error situations.
In PHP for instance, you would need to switch from substr to mb_substr.
I added binary before the column name and solve the charset error.
insert into tableA values(binary stringcolname1);
Hi i also got this error when i use my online databases from godaddy server
i think it has the mysql version of 5.1 or more. but when i do from my localhost server (version 5.7) it was fine after that i created the table from local server and copied to the online server using mysql yog i think the problem is with character set
Screenshot Here
To fix this error I upgraded my MySQL database to utf8mb4 which supports the full Unicode character set by following this detailed tutorial. I suggest going through it carefully, because there are quite a few gotchas (e.g. the index keys can become too large due to the new encodings after which you have to modify field types).
There's good answers in here. I'm just adding mine since I ran into the same error but it turned out to be a completely different problem. (Maybe on the surface the same, but a different root cause.)
For me the error happened for the following field:
#Column(nullable = false, columnDefinition = "VARCHAR(255)")
private URI consulUri;
This ends up being stored in the database as a binary serialization of the URI class. This didn't raise any flags with unit testing (using H2) or CI/integration testing (using MariaDB4j), it blew up in our production-like setup. (Though, once the problem was understood, it was easy enough to see the wrong value in the MariaDB4j instance; it just didn't blow up the test.) The solution was to build a custom type mapper:
package redacted;
import javax.persistence.AttributeConverter;
import java.net.URI;
import java.net.URISyntaxException;
import static java.lang.String.format;
public class UriConverter implements AttributeConverter<URI, String> {
#Override
public String convertToDatabaseColumn(URI attribute) {
return attribute.toString();
}
#Override
public URI convertToEntityAttribute(String field) {
try {
return new URI(field);
}
catch (URISyntaxException e) {
throw new RuntimeException(format("could not convert database field to URI: %s", field));
}
}
}
Used as follows:
#Column(nullable = false, columnDefinition = "VARCHAR(255)")
#Convert(converter = UriConverter.class)
private URI consulUri;
As far as Hibernate is involved, it seems it has a bunch of provided type mappers, including for java.net.URL, but not for java.net.URI (which is what we needed here).
In my case that problem was solved by changing Mysql column encoding to 'binary' (data type will be changed automatically to VARBINARY). Probably I will not be able to filter or search with that column, but I'm no need for that.
In my case ,first i've meet a '???' in my website, then i check Mysql's character set which is latin now ,so i change it into utf-8,then i restart my project ,then i got the same error with you , then i found that i forget to change the database's charset and change into utf-8, boom,it worked.
I tried almost every steps mentioned here. None worked. Downloaded mariadb. It worked. I know this is not a solution yet this might help somebody to identify the problem quickly or give a temporary solution.
Server version: 10.2.10-MariaDB - MariaDB Server
Protocol version: 10
Server charset: UTF-8 Unicode (utf8)
I had a table with a varbinary column that I wanted to convert to utf8mb4 varchar. Unfortunately some of the existing data was invalid UTF-8 and the ALTER query returned Incorrect string value for various rows.
I tried every suggestion I could find regarding cast / convert / char_length = length etc. but nothing in SQL detected the erroneous values, other than the ALTER query returning bad rows one by one. I would love a pure SQL solution to remove the bad values. Sadly this solution is not pretty
I ended up select *'ing the entire table into PHP, where the erroneous rows could be detected en-masse by:
if (empty(htmlspecialchars($row['whatever'])))
The problem can also be caused by the client if the charset is not set to utf8mb4. so even if every Database, Table and Column is set to utf8mb4 you will still get an error, for instance in PyCharm.
For Python, set the charset of the connection in the MySQL Connector connect method:
mydb = mysql.connector.connect(
host="IP or Host",
user="<user>",
passwd="<password>",
database="<yourDB>",
# set charset to utf8mb4 to support emojis
charset='utf8mb4'
)
I know i`m late to the ball but someone else might come accross the problem i had with this and be happy to read my workaround.
I have come accross this problem with french characters. turns out i the text I was copying had encoding the accents on some charaatcers as 2 chars and others as single chars...
i couldn`t find how to set my table to accept the strings so i ended up changing the diacritics in my text import.
here is a list of them as double characters to search for them in your texts.
ùòìàè
áéíóú
ûôêâî
ç
1 - You have to declare in your connection the propertie of enconding UTF8. http://php.net/manual/en/mysqli.set-charset.php.
2 - If you are using mysql commando line to execute a script, you have to use the flag, like:
Cmd: C:\wamp64\bin\mysql\mysql5.7.14\bin\mysql.exe -h localhost -u root -P 3306 --default-character-set=utf8 omega_empresa_parametros_336 < C:\wamp64\www\PontoEletronico\PE10002Corporacao\BancoDeDadosModelo\omega_empresa_parametros.sql
I have a database table with a column where I categorized Persian alphabetic letters to select with MySQL WHERE later. everything works fine for all letters, but I have a problem while selecting letter (چ) which is stored as (Ù†) in database and (ن) which is stored as (Ú†).
first I thought the problem could be from inserting same letters, but when I checked in database , letters where stored with different encoding I mean (Ù†) and (Ú†).
when I zoom in these letters the tick over U is different. both letters are echoed correctly when I echo them on webpage, but when I choose to select letters WHERE letter = 'چ' it shows letters with (ن) too!!!
all of the webpages that insert and read data from DB are in UTF-8 and database collation is utf_persian-ci.
I cant find where the problem is with this? any help is appreciated,
Mojibake. (or not; see below) Probably:
The bytes you have in the client are correctly encoded in utf8 (good).
You connected with SET NAMES latin1 (or set_charset('latin1') or ...), probably by default. (It should have been utf8.)
The column in the tables may or may not have been CHARACTER SET utf8, but it should have been that.
For PHP:
⚈ mysqli interface: mysqli_set_charset('utf8') function.
⚈ PDO interface: set the charset attribute of the PDO dsn or via SET NAMES utf8.
The COLLATION (eg, utf8_persion_ci) is not relevant to Mojibake. It is relevant to how characters are ordered.
Edit
You say "is stored as (Ù†)" -- How do you know? Most attempts to see what is stored are subject to the client fiddling with the bytes. This is a sure way to see what is there:
SELECT col, HEX(col) FROM tbl ...
For چ, the HEX should be DA86 for proper utf8 (or utf8mb4) encoding. If you get C39AE280A0, then you have "double encoding". In general, Arabic/Persian/Farsi should be of the form Dxyy.
If you read چ while connected with latin1, you will get Ù†, which is DA86 in latin1 encoding (Ù = DA and † = 86).
ن encodes as D986.
Double Encoding
I used hex(col) to send query and got C399E280A0 for ن and C39AE280A0 for چ .
So, you have "double encoding", not "Mojibake".
C399 is utf8 for Ù; E280A0 is utf8 for †. Your character was changed from latin1 to utf8 twice. Usually the end result is invisible to the outside world, but messed up in the table. That is because the SELECT decodes twice. However, since you are seeing only one decode, things are not that simple.
Caveat: You have a situation where I have not experimented; the advice I give you could be wrong.
Here's what probably happened.
The client had characters encoded as utf8 (good) hex: D986;
When inserting, the application lied by claiming that the client had latin1 encoding. (This is the old default.); D9 converted to Ù and 86 converted to †;
The column in the table declared CHARACTER SET utf8 (good). But now the Ù is stored as C399 and the † is stored as E280A0, for a total of 5 bytes;
When reading the connection claimed utf8 (good) for the client, so those 5 bytes were turned back into Ù†;
The client dutifully said the utf8 data was Ù†.
Notice the imbalance between the INSERT and the SELECT. You tagged this PHP; did PHP both write and read the data? Did it have a different setting for the charset for writing and reading?
The problem seems to be only in setting the charset for writing. It needed to be explicitly utf8, not defaulting to latin1.
But what about the data? If everything I said (about double encoding) matches what you have, then an UPDATE can fix the data. See my blog for the details.
This is a typical result of using a 'locale specific unicode encoding', in your case utf8_persian_ci. I expect that if you switch your collation to utf8_unicode_ci, it will work as expected.
If by any change you want to get rid of the case-insensitivity, you could switch to utf8_bin.
For further reference see the MySQL documentation.
I am working on a turkish website, which has stored many malformed turkish characters in a MySQL database, like:
- ş as þ
- ı as ý
- ğ as ð
- Ý as İ
i can not change the data in the database, because the database are updated daily and the new data will contain the malformed characters again. So my idea was to change the data in PHP instead of changing the data in the database. I have tried some steps:
Turkish characters are not displayed correctly
Fix Turkish Charset Issue Html / PHP (iconv?)
PHP Turkish Language displaying issue
PHP MYSQL encoding issue ( Turkish Characters )
I am using the PHP-MySQLi-Database-Class available on GitHub with utf8 as charset.
I have even tried to replace the malformed characters with str_replace, like:
$newString = str_replace ( chr ( 253 ), "ı", $newString );
My question is, how can i solve the issue without changing the characters in the database? Are there any best practices? Is it a good option just to replace the characters?
EDIT:
solved it by using
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-9" />
2022 update. I made a wide research and I found this solution and it's working.
let's say your db_connection is $mysqli:
$mysqli = mysqli_connect($hostname, $username, $password, $database) OR DIE ("Baglanti saglanamadi!");
just add this line after. it works like magic with all languages even Arabic:
mysqli_set_charset($mysqli, 'utf8');
Two solutions are good
PHP MYSQL encoding issue ( Turkish Characters )
PHP Turkish Language displaying issue
Also you can set configuration on phpMyAdmin
Operations > Table options > Collation > select utf8_general_ci
if you create the tables already edit the collation structures also
SELECT CONVERT(CONVERT(UNHEX('d0dddef0fdfe') USING ...) USING utf8);
latin5 / iso-8859-1 shows ĞİŞğış
latin1 / iso-8859-9 shows ÐÝÞðýþ
You are confusing two similar encodings; see the first paragraph in https://en.wikipedia.org/wiki/ISO/IEC_8859-9 .
"Collation" is only for sorting. But first you need to change the CHARACTER SET to latin5. Then change the collation to latin5_turkish_ci. (Since that is the default for latin5, no action need be taken.)
This may suffice to make the change in MySQL: EDIT 3
NO, this is probably wring -- ALTER TABLE tbl CONVERT TO CHARACTER SET latin5;
After seeing more of the issue, this "2-step ALTER" is probably correct:
ALTER TABLE Tbl MODIFY COLUMN col VARBINARY(...) ...;
ALTER TABLE Tbl MODIFY COLUMN col VARCHAR(...) ... CHARACTER SET latin5 ...;
Do that for each table. Be sure to test this on a copy of your data first.
The 2-step ALTER is useful for when the bytes are correct, but the CHARACTER SET is not.
CONVERT TO should be used when the characters are correct, but you want a different encoding (and CHARACTER SET). See Case 5.
Edit 1
E7 and FD and cp1250, dec8, latin1 and latin2 for ç and ý. FD in latin5 is ı. I conclude that your encoding is latin1, not latin5.
You say you cannot change the "scripts". Let's look at your limitations. Are you restricted on the INSERT side? Or the SELECT side? Or both? What is rendering the text; html? MySQL is willing to change from latin1 to/from latin5 and you insert/select (based on a few settings). And/or you could lie to HTML (via a meta tag) to get it to interpret the bytes differently. Please spell out the details of the data flow.
Edit 2
Given that the HEX in the table is E7FD6B6172FD6C6D6173FD6E61, and it should be rendered as çıkarılmasına, ... Note especially the second letter needs to show as ı (Turkish dotless small I), not ý (small Y with acute), correct?
Start by trying
<meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-9"/>
That should give you the `latin5 rendering, as you already found out. IANA Reference.
As for "Best practice", that would involve changing the way text is inserted. You have stated this as off-limits.
Apparently you have latin5 characters stored in a latin1 column. Since latin1 does not involve any checking, you can insert and retrieve latin5 characters without any trouble.
This does not address the desire to have Turkish collation. If necessary, I can probably concoct a way to specify Turkish ordering on particular statements; please provide a sample statement.