Wiki copymirror

This is a mirror of the wiki at pascal.chat. Editing, history and sign-in are there.

FPC Unicode support

English · Español · Français · Русский

Introduction

Up to and including FPC 2.6.x, the RTL was based on the ones of Turbo Pascal and Delphi 7. This means it was primarily based around the shortstring, ansistring and pchar types. None of these types had any encoding information associated with them, but were implicitly assumed to be encoded in the "default system encoding" and were passed on to OS API calls without any conversion.

In Delphi 2009, Embarcadero switched the entire RTL over to the UnicodeString type, which represents strings using UTF-16. Additionally, they also made the AnsiString type "code page-aware". This means that AnsiStrings from then on contain the code page according to which their data should be interpreted.

FPC's language-level support for these string types is already available in current stable versions of the compiler (FPC 3.0.0 and up). The RTL level support is not yet complete. This page gives an overview of the code page-related behaviour of these string types, the current level of support in the RTL, and possible future ways of how this support may be improved.

Backward compatibility

If you have existing code that works in a defined way () with a previous version of FPC and make no changes to it, it should continue to work unmodified with the new FPC version. Guaranteeing this is the main purpose of the multitude of DefaultCodePage variables and their default values as described below.

(*) this primarily means: you do not store data in an ansistring that has been encoded using something else than the system's default code page, and subsequently pass this string as-is to an FPC RTL routine. E.g., current Lazarus code is generally fine, as you are supposed to call UTF8ToAnsi() before passing its strings to FPC RTL routines.

If your existing code did use ansistrings in an unsupported way, namely by storing data in it that is not encoded in the system's default code page and not taking care when interfacing with other code (such as RTL routines), you still may be able to work around most of the issues if this data always uses the same encoding. In that case, you can call SetMultiByteConversionCodePage() when starting your program, with as argument the code page of the data that your ansistrings contain. Note that this will also affect the interpretation of all ShortString, AnsiChar and PAnsiChar data.

Code pages

A code page defines how the individual bytes of a string should be interpreted, i.e., which letter, symbol or other graphic character corresponds to every byte or sequence of bytes.

Code page identifiers

A code page identifier is always stored as a TSystemCodePage, which is an alias for Word. The value represents the corresponding code page as defined by [http://msdn.microsoft.com/en-us/library/windows/desktop/dd317756(v=vs.85).aspx Microsoft Windows]. Additionally, there are 3 special code page values:

Note: code page identifiers are different from codepage names as used in the {$codepage xxx} directives (which is available in current stable FPC already). Codepage names are the names of individual codepage units exposed by the charset unit, which have names such as cp866 and cp1251 and utf8.

Code page settings

The system unit contains several global variables that indicate the default code page used for certain operations.

DefaultSystemCodePage

DefaultFileSystemCodePage

DefaultRTLFileSystemCodePage

Source file codepage

The source file codepage determines how string constants are interpreted, and where the compiler will insert codepage conversion operations when assigning one string type to another.

The source file codepage is determined as follows:

String/character types

Shortstring

The code page of a shortstring is implicitly CP_ACP and hence will always be equal to the current value of DefaultSystemCodePage.

PAnsiChar/AnsiChar

These types are the same as the old PChar/Char types. In all compiler modes except for {$mode delphiunicode}, PChar/Char are also still aliases for PAnsiChar/AnsiChar. Their code page is implicitly CP_ACP and hence will always be equal to the current value of DefaultSystemCodePage.

PWideChar/PUnicodeChar and WideChar/UnicodeChar

These types remain unchanged. WideChar/UnicodeChar can contain a single UTF-16 code unit, while PWideChar/PUnicodeChar point to a single or an array of UTF-16 code units.

In {$mode delphiunicode}, PChar becomes an alias for PWideChar/PUnicodeChar and Char becomes an alias for WideChar/UnicodeChar.

UnicodeString/WideString

These types behave the same as in previous versions:

Ansistring

AnsiStrings are reference-counted types with a maximum length of high(SizeInt) bytes. Additionally, they now also have code page information associated with them.

The most important thing to understand about the new AnsiString type is that it both has a declared/static/preferred/default code page (called declared code page from now on), and a dynamic code page. The declared code page tells the compiler that when assigning something to that AnsiString, it should first convert the data to that declared code page (except if it is CP_NONE, see RawByteString below). The dynamic code page is a property of the AnsiString which, similar to the length and the reference count, defines the actual code page of the data currently held by that AnsiString.

Declared code page

The declared code page of an AnsiString can only be defined by declaring a new type as follows:

pascal
type
  CP866String = type AnsiString(866); // note the extra "type"

The declared code page of a variable declared as plain AnsiString is CP_ACP. In effect, the AnsiString type is now semantically defined in the System unit as:

pascal
type
  AnsiString = type AnsiString(CP_ACP);

Another predefined AnsiString(X) type in the System unit is UTF8String:

pascal
type
  UTF8String = type AnsiString(CP_UTF8);

Once you have defined such a custom AnsiString(X) type, you can use it to declare variables, parameters, fields etc as usual.

Note that CP_UTF16 and CP_UTF16BE are not valid as code pages for AnsiStrings. The result of defining an AnsiString with such a code page is undefined.

Dynamic code page

If a string with a declared code page SOURCE_CP is assigned to a string with declared code page DEST_CP , then

These rules mean that it is perfectly possible for an AnsiString variable to get a dynamic code page that differs from its declared code page. E.g. in the third case SOURCE_CP could be CP_ACP, while after the assignment it may have a dynamic code page equal to DefaultSystemCodePage.

Note: as mentioned above, whether or not a potential code page conversion happens only depends on the declared code pages of the involved strings. This means that if you assign one AnsiString(X) to another AnsiString(X) and the former's dynamic code was different from X, the string data will not be converted to code page X by the assignment.

RawByteString

The RawByteString type is defined as:

pascal
type
  RawByteString = type AnsiString(CP_NONE);

As mentioned earlier, the results of operations on strings with the CP_NONE code page are undefined. As it does not make sense to define a type in the RTL whose behaviour is undefined, the behaviour of RawByteString is somewhat different than that of other AnsiString(X) types.

As a first approximation, RawByteString can be thought of as an "untyped AnsiString": assigning an AnsiString(X) to a RawByteString has exactly the same behaviour as assigning that AnsiString(X) to another AnsiString(X) variable with the same value of X: no code page conversion or copying occurs, just the reference count is increased.

Less intuitive is probably that when a RawByteString is assigned to an AnsiString(X), the same happens: no code page conversion or copying, just the reference count is increased. Note that this means that results from functions returning a RawByteString will never be converted to the destination's declared code page. This is another way in which the dynamic code page of an AnsiString(X) can become different from its declared code page.

This type is mainly used to declare const, constref and value parameters that accept any AnsiString(X) value without converting it to a predefined declared code page. Note that if you do this, the routine accepting those parameters should be able to handle strings with any possible dynamic code page.

var and out parameters can also be declared as RawByteString, but in this case the compiler will give an error if an AnsiString(X) whose declared code page is different from CP_NONE is passed in. This is consistent with var and out parameters in general: they require an exactly matching type to be passed in. You can add an explicit RawByteString() typecast around an argument to remove this error, but then you must be prepared to deal with the fact that the returned string can have any dynamic code page.

String concatenations

Normally, in Pascal the result type of an expression is independent of how its result is used afterwards. E.g. multiplying two longints on a 32 bit platform and assigning the result to an int64 will still perform the multiplication using 32 bit arithmetic, and only afterwards the result is converted to 64 bit.

Code page-aware strings are the only exception to this rule: concatenating two or more strings always occurs without data loss, although afterwards the resulting string will of course still be converted to the declared code page of the destination (which may result in data loss).

Assigning the result of a concatenation to a RawByteString is again special:

String constants

The compiler has to know the code page according to which it should interpret string constants, as it may have to convert them at compile time. Normally, a string constant is interpreted according to the source file codepage. If the source file codepage is CP_ACP, a default is used instead: in that case, during conversions the constant strings are assumed to have code page 28591 (ISO 8859-1 Latin 1; Western European).

When a string constant is assigned to an AnsiString(X) either in code or as part of a typed constant or variable initialisation, then

Similarly, if a string constant is assigned to a UnicodeString, the string constant is converted, at compile time, from the source file code page to UTF-16.

For ShortString and PChar, the same rule as for AnsiString(CP_ACP) is followed.

Note that symbolic string constants will be converted at compile time to the appropriate string type and code page whenever they are used. This means that there is no speed overhead when using a single string constant in multiple code page and string type contexts, only some data size overhead.

From the above it follows that to ensure predictable interpretation of string constants in your source code, it is best to either include an explicit {$codepage xxx} directive (or use the equivalent -Fc command line option), or to save the source code in UTF-8 with a BOM.

String indexing

Nothing changes to string indexing. Every string element of a UnicodeString/WideString is two bytes and every string element of all other strings is one byte. The string indexing mechanism completely ignores code pages and composite code points.

RTL changes

In order to fully guarantee data integrity in the presence of codepage-aware strings, all routines in the RTL and packages that accept AnsiString parameters must be adapted. The reason is that if their parameters remain plain AnsiString, then any string with a different declared code page will be converted to DefaultSystemCodePage when it is passed in. This can result in data loss.

Until now, primarily routines dealing with file system access have been updated to preserve all character data. Below is an exhaustive list of all routines that preserve the string encoding in FPC 3.0. Unless where explicitly noted otherwise, these routines also all have overloads that accept UnicodeString parameters.

RTL todos

As the above list is exhaustive, no other RTL routines support arbitrary code pages yet. This section contains a list of gotchas that some people have identified and, if possible, workarounds. Note that routines not mentioned here nor above are equally unsafe as the ones that are explicitly mentioned.

TFormatSettings and DefaultFormatSettings

The type of ThousandSeparator and DecimalSeparator is AnsiChar type. This means that if DefaultSystemCodePage is UTF-8 and the locale's separator is more than one byte long in that encoding, these fields are not large enough. Examples are the French and Russian non-breaking white space character used to represent the ThousandSeparator.

Old/obsolete sections

Warning: These sections are kept for historical reference - please update the sections above with this information if it is still applicable. Since FPC 2.7 (development version before the release of 3.0.0), extensive Unicode support has been implemented.

User visible changes

Full support of code page aware strings is not possible without breaking some existing code. The following list tries to summarize the most important user visible changes.

See Also

From the Free Pascal wiki, page "FPC Unicode support", as it stood on 2023-11-28 after a change by Mattias2, out of the wiki's dump of 2024-05-26. Copyright stays with the author who wrote it, according to the old wiki's Licensing page. If you have concerns about this material, message the admin.

Revision 1, last changed by WikiArchive .