CREATE DATABASE   The file names (db-filename-string, log-filename-string, and mirror-filename-string) are strings containing operating system file names. As literal strings, they must be enclosed in single quotes.

You cannot specify utility_db for db-filename-string. This name is reserved for the utility database.

ACCENT clause   This clause is used to specify accent sensitivity for the database. Support for this clause is deprecated. Use the collation tailoring options provided for the COLLATION and NCHAR COLLATION clauses to specify accent sensitivity.

The ACCENT clause applies only when using the UCA (Unicode Collation Algorithm) for the collation specified in the COLLATION or NCHAR COLLATION clause. ACCENT RESPECT causes the UCA string comparison to respect accent differences between letters. For example, e is less than é. ACCENT FRENCH is similar to ACCENT RESPECT, except that accents are compared from right to left, consistent with the rules of the French language. ACCENT IGNORE causes string comparisons to ignore accents. For example, e is equal to é.

If accent sensitivity is not specified when the database is created, the default accent sensitivity for comparisons and sorting is insensitive, with one exception; for Japanese databases created with a UCA collation, the default accent sensitivity is sensitive.

ASE COMPATIBLE clause   Do not create the SYS.SYSCOLUMNS and SYS.SYSINDEXES views. By default, these views are created for compatibility with system tables available in Watcom SQL (version 4 and earlier of this software). These views conflict with the Adaptive Server Enterprise compatibility views dbo.syscolumns and dbo.sysindexes.

BLANK PADDING clause   SQL Anywhere compares all strings as if they are varying length and stored using the VARCHAR domain. This includes string comparisons involving fixed length CHAR or NCHAR columns. In addition, SQL Anywhere never trims or pads values with trailing blanks when the values are stored in the database.

By default, SQL Anywhere treats blanks as significant characters. For example, the value 'a ' (the character 'a' followed by a blank) is not equivalent to the single-character string 'a'. Inequality comparisons also treat a blank as any other character in the collation.

If blank padding is enabled (specifying BLANK PADDING ON), the semantics of string comparisons more closely follow the ANSI/ISO SQL standard. With blank-padding enabled, SQL Anywhere ignores trailing blanks in any comparison.

In the example above, an equality comparison of 'a ' to 'a' in a blank-padded database returns TRUE. With a blank-padded database, fixed-length string values are padded with blanks when they are fetched by an application. Whether the application receives a string truncation warning on such an assignment is controlled by the ansi_blanks connection option.

CASE clause   This clause is used to specify case sensitivity for the database. Support for this clause is deprecated. Use the collation tailoring options provided for the COLLATION and NCHAR COLLATION clauses to specify case sensitivity.

CASE RESPECT causes case-sensitive string comparisons for all CHAR and NCHAR data types. Comparisons using UCA consider the case of a letter only if the base letters and accents are all equal. For all other collations, uppercase and lowercase letters are distinct; for example, a is less than A, which is less than b, and so on. CASE IGNORE causes case-insensitive string comparisons. Uppercase and lowercase letters are considered to be exactly equal.

If case sensitivity is not specified when the database is created, default case sensitivity for comparisons and sorting is insensitive, with one exception; for Japanese databases created with a UCA collation, default case sensitivity is sensitive.

CASE RESPECT is provided for compatibility with the ISO/ANSI SQL standard. Identifiers in the database are always case insensitive, even in case-sensitive databases.

CHECKSUM clause   Checksums are used to determine whether a database page has been modified on disk. When you create a database with global checksums enabled, a checksum is calculated for each page just before it is written to disk. The next time the page is read from disk, the page's checksum is recalculated and compared to the checksum stored on the page. If the checksums are different, then the page has been modified on disk and an error occurs. Databases created with global checksums enabled can also be validated using checksums. You can check whether a database was created with global checksums enabled by executing the following statement:

SELECT DB_PROPERTY ( 'Checksum' );

This query returns ON if global checksums are turned on, otherwise, it returns OFF. Global checksums are turned on by default, so if the CHECKSUM clause is omitted, ON is applied.

Regardless of the setting of this clause, the database server always enables write checksums for databases running on storage devices such as removable drives, and databases running on Windows Mobile to help provide early detection if the database file becomes corrupt. It also calculates checksums for critical pages during validation activities.

For databases that do not have global checksums enabled, you can enable write checksums by using the -wc options.

COLLATION clause   The collation specified by the COLLATION clause is used for sorting and comparison of character data types (CHAR, VARCHAR, and LONG VARCHAR). The collation provides character comparison and ordering information for the encoding (character set) being used. If the COLLATION clause is not specified, SQL Anywhere chooses a collation based on the operating system language and encoding.

The collation can be chosen from the list of collations that use the SQL Anywhere Collation Algorithm (SACA), or it can be the Unicode Collation Algorithm (UCA). If UCA is specified, you should also specify the ENCODING clause.

It is important to choose your collation carefully. It cannot be changed after the database has been created.

Optionally, you can specify collation tailoring options (collation-tailoring-string) for additional control over the sorting and comparing of characters. These options take the form of keyword=value pairs, assembled in parentheses, following the collation name. For example, ... CHAR COLLATION 'UCA(locale=es;case=respect;accent=respect)'.

DATABASE SIZE clause   Use this optional clause to set the initial size of the database file. You can use KB, MB, GB, or PAGES to specify units of kilobytes, megabytes, gigabytes, or pages respectively.

Specifying the file size at creation time is a way of pre-allocating space for the file. This helps reduce the risk of running out of space on the drive the database is located on. As well, it can help improve performance by increasing the amount of data that can be stored in the database before the database server needs to grow the database, which can be a time-consuming operation.

DBA USER clause   Use this clause to specify a DBA user for the database. When you use this clause, you can no longer connect to the database as the default DBA user. If you do not specify this clause, the default DBA user ID is created.

DBA PASSWORD clause   You can specify a different password for the DBA database user. If you do not specify this clause, the default password (sql) is used for the DBA user.

ENCODING clause   Most collations specified in the COLLATION clause dictate both the encoding (character set) and ordering. For those collations, the ENCODING clause should not be specified. However, if the value specified in the COLLATION clause is UCA (Unicode Collation Algorithm), use the ENCODING clause to specify a locale-specific encoding and get the benefits of the UCA for comparison and ordering. The ENCODING clause may specify UTF-8 or any single-byte encoding for CHAR data types. ENCODING may not specify a multibyte encoding other than UTF-8.

If you choose the UCA collation, you can optionally specify collation tailoring options.

If COLLATION is set to UCA and ENCODING is not specified, then SQL Anywhere uses UTF-8.

ENCRYPTED or ENCRYPTED TABLE clause   Encryption makes stored data unreadable. Use the ENCRYPTED keyword (without TABLE) when you want to encrypt the entire database. Use the ENCRYPTED TABLE clause when you only want to enable table encryption. Enabling table encryption means that the tables that are subsequently created or altered using the ENCRYPTED clause are encrypted using the settings you specified at database creation.

There are two levels of database and table encryption: simple and strong. Simple encryption is equivalent to obfuscation. The data is unreadable, but someone with cryptographic expertise could decipher the data. Strong encryption makes the data is unreadable and virtually undecipherable.

For simple encryption, specify ENCRYPTED ON ALGORITHM SIMPLE, or ENCRYPTED ALGORITHM SIMPLE, or specify the ENCRYPTED ON clause without specifying an algorithm or key.

For strong encryption, specify ENCRYPTED ON ALGORITHM with a 128-bit or 256-bit AES algorithm, and the KEY clause to specify an encryption key. It is recommended that you choose a value for your key that is at least 16 characters long, contains a mix of uppercase and lowercase, and includes numbers, letters, and special characters.

On Windows Mobile, the AES_FIPS and AES256_FIPS algorithms are only supported with ARM processors.

You can also create an encrypted copy of an existing database using the CREATE ENCRYPTED DATABASE statement.

JCONNECT clause   To allow the jConnect JDBC driver access to system catalog information, specify JCONNECT ON. This clause installs the system objects that provide jConnect support. Specify JCONNECT OFF if you want to exclude the jConnect system objects. You can still use JDBC, as long as you do not access system information. JCONNECT is ON by default.

PAGE SIZE clause   The page size for a database can be 2048, 4096, 8192, 16384, or 32768 bytes. The default page size is 4096 bytes. Large databases generally obtain performance benefits from a larger page size, but there can be additional overhead associated with large page sizes.

For example:

CREATE DATABASE 'c:\\databases\\my_db.db'
PAGE SIZE 4096;

NCHAR COLLATION clause   The collation specified by the NCHAR COLLATION clause is used for sorting and comparison of national character data types (NCHAR, NVARCHAR, and LONG NVARCHAR). The collation provides character ordering information for the UTF-8 encoding (character set) used for national characters. If the NCHAR COLLATION clause is not specified, SQL Anywhere uses the Unicode Collation Algorithm (UCA). The only other allowed collation is UTF8BIN, which provides a binary ordering of all characters whose encoding is greater than 0x7E.

Optionally, you can specify collation tailoring options (collation-tailoring-string) for additional control over the sorting and comparing of characters. These options take the form of keyword=value pairs, assembled in a quoted string, following the collation name. For example, ... NCHAR COLLATION 'UCA(locale=es;case=respect;accent=respect)'. If you specify the ACCENT or CASE clause and a collation tailoring string that contains settings for case and accent, the values of the ACCENT and CASE clauses are used as defaults only.

TRANSACTION LOG clause   The transaction log is a file where the database server logs all changes made to the database. The transaction log plays a key role in backup and recovery, and in data replication.

The MIRROR clause of the TRANSACTION clause allows you to provide a file name if you want to use a transaction log mirror. A transaction log mirror is an identical copy of a transaction log, usually maintained on a separate device, for greater protection of your data. By default, SQL Anywhere does not use a transaction log mirror.