Firebird Character Sets and Collations

Every CHAR or VARCHAR field can (or, better: must) have a character set assigned. Firebird uses this information to correctly store the bytes that make up the character string.

In order to be able to sort or compare strings, you also need to define a collation. A collation defines the sort ordering and uppercase conversions for a string.

Firebird is unable to transliterate between character sets. So you must set the correct values on the server and on the client if everything is to work smoothely.

An Example

In the German language there are the "Umlauts", special vowels with a double-dot (diaeresis) over them. A common last name in Germany is Müller. (If you don't have umlauts on your keyboard, you could also write "Mueller", but that's not what we want to discuss here ;-)

When you convert Müller to uppercase you get MÜLLER, so there is an uppercase Ü and a lowercase ü.

When you want to perform a lexicographic compare on the name, you have several options:

  • You can treat the Ü like a U (German "Duden" dictionary)
  • You can treat the Ü like UE (German telephone book)
  • You can treat the Ü like a special character, sorted in after Z (a common practice in Scandinavia)

Creating a Database

You can define the default character set for a new database in the CREATE DATABASE statement:

CREATE DATABASE <database>
USER <username>
PASSWORD <password>
PAGE_SIZE <pagesize>
DEFAULT CHARACTER SET <charset>

For example:

CREATE DATABASE localhost:meter
USER SYSDBA
PASSWORD masterkey
PAGE_SIZE 4096
DEFAULT CHARACTER SET ISO8859_1;

From now on, any VARCHAR or CHAR field will default to the ISO8859_1 character set. You can, however, specify a special character set for each column:

CREATE TABLE users (
CZECH_NAME VARCHAR(50) CHARACTER SET ISO8859_2,
...

Collations

There is no default collation. So you should define a collation for every field that is to be used for sorting (ORDER BY) or comparing (UPPER):

CREATE TABLE users (
NAME VARCHAR(50) COLLATE DE_DE,
...

COLLATE DE_DE means: "use a collation for the German language (the first DE), applying the rules from Germany (the second DE)"

You can also specify the collation with the ORDER BY clause:

ORDER BY LASTNAME COLLATE FR_CA, FIRSTNAME COLLATE FR_CA

or with the WHERE clause:

WHERE LASTNAME COLLATE FR_CA = :lastnametosearch

or when searching:

WHERE UPPER (LAST_NAME COLLATE SV_SV) = 'PAULSEN';

The UPPER() function

UPPER() only works correctly if there is a collation defined for the parameter field:

WHERE UPPER (NAME COLLATE DE_DE) = 'MÜLLER';

Specifying the client character set

ISQL

SET NAMES ISO8859_1;

InterBase Objects (Ibo) by Jason Wharton

The TIb_Connection class has a string property named CharSet. Assign it the name of the character set to use:

Ib_Connection1.CharSet := 'ISO8859_1';

InterBase Express (IBX), built into Delphi

The TIbDatabase class has a TStrings property named Params. Add a field with the name lc_ctype and specify the character set:

IbDatabase1.Params.Add ('lc_ctype=ISO8859_1');

PHP

In PHP you define the Client Character Set when you connect (or pconnect) to the database.

$db = ibase_connect ($Name, $Usr, $Pwd, "ISO8859_1");

Conversions

Conversions between character sets are always done as: CHARSET1 -> UNICODE -> CHARSET2

With NONE or OCTETS as the connection character set, the bytes are just copied: NONE/OCTETS -> CHARSET2 and CHARSET1 -> NONE/OCTETS.

Case insensitive searching

I have written a separate article about this.


Character Sets and Collations

Character Set Languages Collation Comments
Generic Character Sets
NONE All NONE No character set applied. With this character set setting, Firebird is unable to perform conversion operations like UPPER() correctly on anything other than the standard 26 latin letters.
OCTETS | BINARY All OCTETS Same as NONE. Cannot be used as client connection character set. Space character is #x00. Will be displayed as hex in ISQL 2.0.
ASCII English ASCII English
Unicode based Character Sets
UNICODE_FSS All UNICODE_FSS Unicode UTF-8. An old implementation that accepts malformed strings and does not enforce correct max. string length. All characters 3 bytes, no case mapping.

Superseded in Firebird 2.0 with the UTF8 character set. Deprecated.

UTF8 All UCS_BASIC UCS_BASIC sorts in Unicode code-point order (Firebird 2.0)
UNICODE Sorts using the Unicode Collation Algorithm (UCA) (Firebird 2.0)
UTF-8 Case insensitive collation (Firebird 2.1)
UNICODE_CI_AI Case insentive, Accent insensitive collation for Unicode (Firebird 2.5)
Current Character Sets
ISO8859_1 Western Europe ISO8859_1 Latin-1
DA_DA Danish/Danmark
DE_DE German/Germany
DU_NL Dutch/The Netherlands
EN_UK English/United Kingdom
EN_US English/USA
ES_ES Spanish/Spain
ES_ES_CI_AI Spanish/Spain, case insensitive, accent insensitive (Firebird 2.0)
FI_FI Finnish/Finnland
FR_CA French/Canada
FR_FR French/France
FR_FR_CI_AI French/France, case insensitive, accent insensitive (Firebird 2.1)
IS_IS Icelandic/Iceland
IT_IT Italian/Italy
NO_NO Norwegian/Norway
PT_PT Portuguese/Portugal
PT_BR Portuguese/Brasil (Firebird 2.0). Case+Accent insensitive
SV_SV Swedish/Sweden
ISO8859_2 Central Europe ISO8859_2 Central Europe
CS_CZ Czech
ISO_HUN Hungarian
ISO_PLK Polish (Firebird 2.0)
ISO8859_3 Southern Europe ISO8859_3 Maltese, Esperanto
ISO8859_4 North European ISO8859_4 Estonian, Latvian, Lithuanian, Greenlandic, Lappish
ISO8859_5 Cyrillic ISO8859_5 Russian, Ukrainian
ISO8859_6 Arabic ISO8859_6  
ISO8859_7 Modern Greek ISO8859_7  
ISO8859_8 Hebrew ISO8859_8  
ISO8859_9 Turkish ISO8859_9  
ISO8859_13 Baltic ISO8859_13 Baltic
LT_LT Lithuanian
WIN1250 Central Europe WIN1250 Central Europe
BS_BA Bosnian (Firebird 2.0)
WIN_CZ Czech, case-insensitive (Firebird 2.0)
WIN_CZ_CI_AI Czech, case-insensitive, accent-insensitive (Firebird 2.0)
PXW_CSY Czech
PXW_HUN Hungarian
PXW_HUNDC Hungarian, Dictionary sort
PXW_PLK Polish
PXW_SLOV Slovanian
WIN1251 Cyrillic WIN1251 Cyrillic
WIN1251_UA Ukrainian
PXW_CYRL Cyrillic, Paradox compatibility
WIN1252 Western Europe, America WIN1252 Latin-1 with Windows extensions
WIN_PTBR Brasilian Portuguese (Firebird 2.0). Case+Accent insensitive
PXW_INTL Paradox ANSI International
PXW_INTL850 Paradox Multi-Lingual Latin-1
PXW_NORDAN4 Paradox Norwegian and Danish
PXW_SPAN Paradox Spanish
PXW_SWEDFIN Paradox Swedish, Finnish
WIN1253 Modern Greek WIN1253  
PXW_GREEK Paradox Greek
WIN1254 Turkish WIN1254  
PXW_TURK Paradox Turkish
WIN1255 Hebrew WIN1255  
WIN1256 Arabic WIN1256  
WIN1257 Baltic WIN1257 Baltic
WIN1257_LV Latvian dictionary collation (Firebird 2.0)
WIN1257_LT Lithuanian dictionary collation (Firebird 2.0)
WIN1257_EE Estonian dictionary collation (Firebird 2.0)
WIN1258 Vietnamese   Vietnamese (Firebird 2.0)
MS-DOS, dBASE and Paradox compatibility
DOS437 Western Europe, America DOS437 English/USA
DB_DEU437 dBASE German
DB_ESP437 dBASE Spanish
DB_FIN437 dBASE Finnish
DB_FRA437 dBASE French
DB_ITA437 dBASE Italian
DB_NLD437 dBASE Dutch
DB_SVE437 dBASE Swedisch
DB_UK437 dBASE English/UK
DB_US437 dBASE English/US
PDOX_ASCII Paradox ASCII code page
PDOX_INTL Paradox International English code page
PDOX_SWEDFIN Paradox Swedish/Finnish code page
DOS737 Greek DOS737 Greek
DOS775 Baltic DOS775 Baltic
DOS850 Western Europe, America DOS850 Latin-1 (without Euro € symbol)
DB_DEU850 dBASE German
DB_ESP850 dBASE Spanish
DB_FRA850 dBASE French/France
DB_FRC850 dBASE French/Canada
DB_ITA850 dBASE Italian
DB_NLD850 dBASE Dutch
DB_PTB850 dBASE Portuguese/Brasil
DB_SVE850 dBASE Swedish
DB_UK850 dBASE English/UK
DB_US850 dBASE English/USA
DOS852 Central Europe DOS852 Latin-2 (Central Europe)
DB_CSY dBASE Czech
DB_PLK dBASE Polish
DB_SLO dBASE Slovakian
PDOX_CSY Paradox Czech
PDOX_HUN Paradox Hungarian
PDOX_PLK Paradox Polish
PDOX_SLO Paradox Slovakian
DOS857 Turkish DOS857 Turkish
DB_TRK dBASE Turkish
DOS858   DOS858 Latin-1 plus Euro symbol €
DOS860 Portuguese DOS860 Portuguese
DB_PTG860 dBASE Portuguese
DOS861 Icelandic DOS861 Icelandic
PDOX_ISL Paradox Icelandic
DOS862 Hebrew DOS862 Hebrew
DOS863 Canadian French DOS863 French/Canada
DB_FRC863 dBASE French/Canada
DOS864 Arabic DOS864 Arabic
DOS865 Scandinavian DOS865 Nordic
DB_NOR865 dBASE Norwegian
DB_DAN865 dBASE Danish
PDOX_NORDAN4 Paradox Norwegian
DOS866 Russian DOS866 Russian
DOS869 Greek DOS869 Modern Greek
Others
BIG_5 Chinese BIG_5 Chinese
KOI8R Russian   Russian character set and dictionary collation (Firebird 2.0)
KOI8U Ukrainian   Ukrainian character set and dictionary collation (Firebird 2.0)
CYRL Russian/Ukrainian CYRL Cyrillic
DB_RUS dBASE Russian
PDOX_CYRL Paradox Cyrillic
KSC_5601 Korean KSC_5601 Unified Korean Hangeul, also known as windows-949
KSC_DICTIONARY Korean dictionary ordering
NEXT NeXT Computers NEXT NeXTSTEP encoding
NXT_DEU German
NXT_ESP Spanish
NXT_FRA French
NXT_ITA Italian
NXT_US US-English
SJIS_0208 Japanese SJIS_0208 Shift-JIS
EUCJ_0208 Japanese EUCJ_0208 EUC Japanese
GB_2312 Chinese GB_2312 Simplified Chinese (HongKong, PRC), a subset of GBK/windows-936
CP943C Japanese CP943C_UNICODE Japanese character set (Firebird 2.1)
TIS620 Thai TIS620_UNICODE Thai character set, single byte (Firebird 2.1)

Which one to choose?

The question now is: which character set do I choose for my database?

Note: I don't have any experience with Asian scripts (Chinese, Korean, Japanese) so I can't give you any hint on these

  • You should chose the DOS, dBASE and Paradox character sets only if you have legacy applications to support
  • The WINxxx character sets are extensions of the corresponding ISOxxx character sets, however you will have problems on non-Windows systems. So if you have a cross-platform application, stay with the ISOxxx character sets.
  • The ISOxxx character sets are missing a few characters of the WINxxx character sets (like typographic dash signs) or they can have some different characters, so your application must be prepared to handle this.

Unicode?

The Unicode situation dramatically improved with Firebird 2.0. Now there is the new UTF8 character set that correctly handles Unicode strings in UTF-8 format. The Unicode collation algorithm has been implemented so now you can use UPPER() and the new LOWER() function without the need to specify a collation.


See also: www.firebirdsql.org/index.php?op=doc&id=fb_1_5_charsets

See also: www.collation-charts.org

Firebird Character Sets and Collations的更多相关文章

  1. 10.1.5 Connection Character Sets and Collations

    10.1.5 Connection Character Sets and Collations Several character set and collation system variables ...

  2. MySQL: Connection Character Sets and Collations

    character_set_server collation_servercharacter_set_databasecollation_database character_set_clientch ...

  3. 02:PostgreSQL Character Sets

    在利用postGIS导入shapefile文件到postgresql数据库的时候,老是提示字符串的问题,或者是乱码,试了好几种都不行,于是度娘之.... 使用默认的UTF8,提示信息是:建议使用LAT ...

  4. Character Sets: Migrating to utf8mb4 with pt_online_schema_change

    David Berube  | June 12, 2018 |  Posted In: MySQL Modern applications often feature the use of data ...

  5. Character Sets, Collation, Unicode :: utf8_unicode_ci vs utf8_general_ci

    w Hi, You can check and compare sort orders provided by these two collations here: http://www.collat ...

  6. information_schema系列之字符集校验(CHARACTER_SETS,COLLATIONS,COLLATION_CHARACTER_SET_APPLICABILITY)

    1:CHARACTER_SETS 首先看一下查询前十条的结果: root@localhost [information_schema]>select * from CHARACTER_SETS ...

  7. MySQL设置字符集CHARACTER SET

    本文地址:http://www.cnblogs.com/yhLinux/p/4036506.html 在 my.cnf 配置文件中设置相关选项,改变为相应的character set. 设置数据库编码 ...

  8. MySQL基础知识:Character Set和Collation

    A character set is a set of symbols and encodings. A collation is a set of rules for comparing chara ...

  9. mysql set names 命令和 mysql 字符编码问题

    先看下面的执行结果: (root@localhost)[(none)]mysql>show variables like 'character%'; +--------------------- ...

随机推荐

  1. 【udacity】机器学习-神经网络

    Evernote Export 1.神经网络 神经元 细胞的主体称为细胞体,然后有轴突.突触 他们构建的方式是可以调整的 我们会有一些输入的放电信号视为放电频率或输入的强度 X1​w1​X2​w2​X ...

  2. 【剑指Offer】43、左旋转字符串

      题目描述:   汇编语言中有一种移位指令叫做循环左移(ROL),现在有个简单的任务,就是用字符串模拟这个指令的运算结果.对于一个给定的字符序列S,请你把其循环左移K位后的序列输出.例如,字符序列S ...

  3. DOM学习之充实文档内容

    HTML代码 <!DOCTYPE html> <html> <head> <meta charset="utf-8"> <me ...

  4. 59.bouncing results

        一.bouncing results成因及解决方案 bouncing results问题,两个document排序,field值相同:不同的shard上,可能排序不同:每次请求轮询路由到不同的 ...

  5. 【ABCD组】Scrum meeting 4

    前言 第4次会议在6月16日由组长在教9 405召开. 主要对下一步的工作进行说明安排,时长90min. 主要内容 分配下阶段任务,争取在这阶段完成软件的设计阶段 任务分配 姓名 当前阶段任务 贡献时 ...

  6. python 图片滑动窗口

    METHOD #1: No smooth, just scaling. def pyramid(image, scale=1.5, minSize=(30, 30)): # yield the ori ...

  7. mysql中使用order 出现错误

  8. UEditor使用总结(与SpringMVC整合)

    最近再弄富文本框,选择了UEditor,原因是:界面漂亮,百度开源的然而, 开启整合之路(怎么做) 1.下载插件 下载只有将插件放在Webapp下,如图 2.修改 导入之后我们就需要修改一些参数已满足 ...

  9. 国产芯片任重道远 国科微SSD主控芯片的“追赶之路”(不能只提供一颗芯片,而是要将芯片、国密算法、固件Firmware、BIOS和操作系统紧密联系在一起,变成完整解决方案交给行业用户,才能真正体现自身的价值)

    集微网消息,“中国芯”战略之路道阻且长,踏入这个赛道的攻坚者们需要十年如一日的技术突破,需要集合产业势能,共同协作,方能建立中国核心技术真正的竞争力. 国产化之路任重道远,SSD芯片初见成效 信息时代 ...

  10. 简单list转化为tree结构

    1 @Override public List<Module> getTree() { List<Module> list = moduleMapper.queryList() ...