SQL string comparison, greater than and less than operators

The comparison operators (including < and >) "work" with string values as well as numbers.

For MySQL

By default, string comparisons are not case sensitive and use the current character set. The default is latin1 (cp1252 West European), which also works well for English.

String comparisons will be case sensitive when the characterset collation of the strings being compared is case sensitive, i.e. the name of the character set ends in _cs rather than _ci. There's really no point in repeating all of the information that's available in MySQL Reference Manual here.

MySQL Comparison Operators Reference:
http://dev.mysql.com/doc/refman/5.5/en/comparison-operators.html

More information about MySQL charactersets/collations:
http://dev.mysql.com/doc/refman/5.5/en/charset.html

To answer the specific questions you asked:

Q: is this a possible way to compare strings in SQL?

A: Yes, in both MySQL and SQL Server

Q: and how does it act?

A: A comparison operator returns a boolean, either TRUE, FALSE or NULL.

Q: a string less than another one comes before in dictionary order? For example, ball is less than water?

A: Yes, because 'b' comes before 'w' in the characteset collation, the expression

  'ball' < 'water'

will return TRUE. (This depends on the characterset and on the collation.

Q: and this comparison is case sensitive?

A: Whether a particular comparison is case sensitive or not depends on the database server; by default, both SQL Server and MySQL are case insensitive.

In MySQL it is possible to make string comparisons by specifying a characterset collation that is case sensitive (the characterset name will end in _cs rather than _ci)

Q: For example BALL < water, the upper case character does affect these comparison?

A: By default, in both SQL Server and MySQL, the expression

  'BALL' < 'water'

would return TRUE.

In Microsoft SQL Server, collation determines to dictionary rules for comparing and sorting character data with regards to:

• case sensitivity
• accent sensitivity
• width sensitivity
• kana sensitivity

SQL Server also includes binary collations where comparison and sorting is done by binary code point rather than dictionary rules. Once can choose from many collations according to the desired sensitivity behavior. The default collation selected for Latin-based language locales during SQL installation is case insensitive and accent sensitive.

Collation is specified at the instance (during installation), database, and column level. Instance collation determines the collation of Instance-level objects like logins and database names as well as identifiers for variables, GOTO labels and temporary tables. Database collation (same as instance collation by default), determines the collation of database identifiers like table and column names as well as literal expressions. Column collation (same as database collation by default) determines the collation of that column.

It is certainly possible compare strings using '<', '>', '<>', ,LIKE, BETWEEN, etc.

if you are using Mybatis or XML based technique to execute SQL query, you have to use <![CDATA[your_symbol-here]]> to avoid that issue.

'ball' <![CDATA[<]]> 'water'