Libconfig is a library for reading, manipulating, and writing structured configuration files. The library features a fully reentrant parser and includes bindings for both the C and C++ programming languages.
The library runs on modern POSIX-compilant systems, such as Linux, Solaris, and Mac OS X (Darwin), as well as on Microsoft Windows 2000/XP and later (with either Microsoft Visual Studio 2005 or later, or the GNU toolchain via the MinGW environment).
There are several open-source configuration file libraries available as of this writing. This library was written because each of those libraries falls short in one or more ways. The main features of libconfig that set it apart from the other libraries are:
To use the library from C code, include the following preprocessor directive in your source files:
#include <libconfig.h>
To link with the library, specify `-lconfig' as an argument to the linker.
To use the library from C++, include the following preprocessor directive in your source files:
#include <libconfig.h++>
Or, alternatively:
#include <libconfig.hh>
The C++ API classes are defined in the namespace `libconfig', hence the following statement may optionally be used:
using namespace libconfig;
To link with the library, specify `-lconfig++' as an argument to the linker.
Libconfig is fully reentrant; the functions in the library do not make use of global variables and do not maintain state between successive calls. Therefore two independent configurations may be safely manipulated concurrently by two distinct threads.
Libconfig is not thread-safe. The library is not aware of the presence of threads and knows nothing about the host system's threading model. Therefore, if an instance of a configuration is to be accessed from multiple threads, it must be suitably protected by synchronization mechanisms like read-write locks or mutexes; the standard rules for safe multithreaded access to shared data must be observed.
Libconfig is not async-safe. Calls should not be made into the library from signal handlers, because some of the C library routines that it uses may not be async-safe.
Libconfig is not guaranteed to be cancel-safe. Since it is not aware of the host system's threading model, the library does not contain any thread cancellation points. In most cases this will not be an issue for multithreaded programs. However, be aware that some of the routines in the library (namely those that read/write configurations from/to files or streams) perform I/O using C library routines which may potentially block; whether these C library routines are cancel-safe or not depends on the host system.
On UNIX systems you can use the pkg-config utility (version 0.20 or later) to automatically select the appropriate compiler and linker switches for libconfig. Ensure that the environment variable `PKG_CONFIG_PATH' contains the absolute path to the lib/pkgconfig subdirectory of the libconfig installation. Then, you can link C programs with libconfig as follows:
gcc `pkg-config --cflags libconfig` myprogram.c -o myprogram \
`pkg-config --libs libconfig`
And similarly, for C++ programs:
g++ `pkg-config --cflags libconfig++` myprogram.cpp -o myprogram \
`pkg-config --libs libconfig++`
Note the backticks in the above examples.
Libconfig supports structured, hierarchical configurations. These configurations can be read from and written to files and manipulated in memory.
A configuration consists of a group of settings, which associate names with values. A value can be one of the following:
Consider the following configuration file for a hypothetical GUI application, which illustrates all of the elements of the configuration file grammar.
# Example application configuration file
version = "1.0";
application:
{
window:
{
title = "My Application";
size = { w = 640; h = 480; };
pos = { x = 350; y = 250; };
};
list = ( ( "abc", 123, true ), 1.234, ( /* an empty list */) );
books = ( { title = "Treasure Island";
author = "Robert Louis Stevenson";
price = 29.95;
qty = 5; },
{ title = "Snow Crash";
author = "Neal Stephenson";
price = 9.99;
qty = 8; } );
misc:
{
pi = 3.141592654;
bigint = 9223372036854775807L;
columns = [ "Last Name", "First Name", "MI" ];
bitmask = 0x1FC3;
};
};
|
Settings can be uniquely identified within the configuration by a path. The path is a dot-separated sequence of names, beginning at a top-level group and ending at the setting itself. Each name in the path is either the name of a setting; if the setting has no name because it is an element in a list or array, an integer index in square brackets can be used as the name.
For example, in our hypothetical configuration file, the path to the
x setting is application.window.pos.x; the path to the
version setting is simply version; and the path to the
title setting of the second book in the books list is
application.books.[1].title.
The datatype of a value is determined from the format of the value
itself. If the value is enclosed in double quotes, it is treated as a
string. If it looks like an integer or floating point number, it is
treated as such. If it is one of the values TRUE, true,
FALSE, or false (or any other mixed-case version of
those tokens, e.g., True or FaLsE), it is treated as a
boolean. If it consists of a comma-separated list of values enclosed
in square brackets, it is treated as an array. And if it consists of a
comma-separated list of values enclosed in parentheses, it is treated
as a list. Any value which does not meet any of these conditions is
considered invalid and results in a parse error.
All names are case-sensitive. They may consist only of alphanumeric characters, dashes (`-'), underscores (`_'), and asterisks (`*'), and must begin with a letter or asterisk. No other characters are allowed.
In C and C++, integer, 64-bit integer, floating point, and string
values are mapped to the types long, long long,
double, and const char *, respectively. The boolean type
is mapped to int in C and bool in C++.
The following sections describe the elements of the configuration file grammar in additional detail.
A setting has the form:
name = value ;
or:
name : value ;
The trailing semicolon is required. Whitespace is not significant.
The value may be a scalar value, an array, a group, or a list.
A group has the form:
{ settings ... }
Groups can contain any number of settings, but each setting must have a unique name within the group.
An array has the form:
[ value, value ... ]
An array may have zero or more elements, but the elements must all be scalar values of the same type.
A list has the form:
( value, value ... )
A list may have zero or more elements, each of which can be a scalar value, an array, a group, or another list.
Integers can be represented in one of two ways: as a series of one or more decimal digits (`0' - `9'), with an optional leading sign character (`+' or `-'); or as a hexadecimal value consisting of the characters `0x' followed by a series of one or more hexadecimal digits (`0' - `9', `A' - `F', `a' - `f').
Long long (64-bit) integers are represented identically to integers, except that an 'L' character is appended to indicate a 64-bit value. For example, `0L' indicates a 64-bit integer value 0.
Floating point values consist of a series of one or more digits, one decimal point, an optional leading sign character (`+' or `-'), and an optional exponent. An exponent consists of the letter `E' or `e', an optional sign character, and a series of one or more digits.
Boolean values may have one of the following values: `true', `false', or any mixed-case variation thereof.
String values consist of arbitrary text delimited by double quotes. Literal double quotes can be escaped by preceding them with a backslash: `\"'. The escape sequences `\\', `\f', `\n', `\r', and `\t' are also recognized, and have the usual meaning. No other escape sequences are currently supported.
Adjacent strings are automatically concatenated, as in C/C++ source code. This is useful for formatting very long strings as sequences of shorter strings. For example, the following constructs are equivalent:
"The quick brown fox jumped over the lazy dog."
"The quick brown fox" " jumped over the lazy dog."
"The quick" /* comment */ " brown fox " // another comment "jumped over the lazy dog."
Three types of comments are allowed within a configuration:
As expected, comment delimiters appearing within quoted strings are treated as literal text.
Comments are ignored when the configuration is read in, so they are not treated as part of the configuration. Therefore if the configuration is written back out to a stream, any comments that were present in the original configuration will be lost.
This chapter describes the C library API. The type config_t represents a configuration, and the type config_setting_t represents a configuration setting.
The boolean values CONFIG_TRUE and CONFIG_FALSE are
macros defined as (1) and (0), respectively.
These functions initialize and destroy the configuration object config.
config_init()initializes config as a new, empty configuration.
config_destroy()destroys the configuration config, deallocating all memory associated with the configuration, but not including the config_t structure itself.
This function reads and parses a configuration from the given stream into the configuration object config. It returns
CONFIG_TRUEon success, orCONFIG_FALSEon failure; theconfig_error_text()andconfig_error_line()functions, described below, can be used to obtain information about the error.
This function reads and parses a configuration from the file named filename into the configuration object config. It returns
CONFIG_TRUEon success, orCONFIG_FALSEon failure; theconfig_error_text()andconfig_error_line()functions, described below, can be used to obtain information about the error.
This function writes the configuration config to the given stream.
This function writes the configuration config to the file named filename. It returns
CONFIG_TRUEon success, orCONFIG_FALSEon failure.
These functions, which are implemented as macros, return the text and line number of the parse error, if one occurred during a call to
config_read()orconfig_read_file(). Storage for the string returned byconfig_error_text()is managed by the library and released automatically when the configuration is destroyed; the string must not be freed by the caller.
config_set_auto_convert()enables number auto-conversion for the configuration config if flag is non-zero, and disables it otherwise. When this feature is enabled, an attempt to retrieve a floating point setting's value into an integer (or vice versa), or store an integer to a floating point setting's value (or vice versa) will cause the library to silently perform the necessary conversion (possibly leading to loss of data), rather than reporting failure. By default this feature is disabled.
config_get_auto_convert()returnsCONFIG_TRUEif number auto-conversion is currently enabled for config; otherwise it returnsCONFIG_FALSE.
These functions locate the setting in the configuration config specified by the path path. They return the value of the setting on success, or a 0 or
NULLvalue if the setting was not found or if the type of the value did not match the type requested.Storage for the string returned by
config_lookup_string()is managed by the library and released automatically when the setting is destroyed or when the setting's value is changed; the string must not be freed by the caller.
This function locates the setting in the configuration config specified by the path path. It returns a pointer to the
config_setting_tstructure on success, orNULLif the setting was not found.
These functions return the value of the given setting. If the type of the setting does not match the type requested, a 0 or
NULLvalue is returned. Storage for the string returned byconfig_setting_get_string()is managed by the library and released automatically when the setting is destroyed or when the setting's value is changed; the string must not be freed by the caller.
These functions set the value of the given setting to value. On success, they return
CONFIG_TRUE. If the setting does not match the type of the value, they returnCONFIG_FALSE.config_setting_set_string()makes a copy of the passed string value, so it may be subsequently freed or modified by the caller without affecting the value of the setting.
These functions get and set the external format for the setting setting.
The format must be one of the constants
CONFIG_FORMAT_DEFAULTorCONFIG_FORMAT_HEX. All settings support theCONFIG_FORMAT_DEFAULTformat. TheCONFIG_FORMAT_HEXformat specifies hexadecimal formatting for integer values, and hence only applies to settings of typeCONFIG_TYPE_INTandCONFIG_TYPE_INT64. If format is invalid for the given setting, it is ignored.
config_setting_set_format()returnsCONFIG_TRUEon success andCONFIG_FALSEon failure.
This function fetches the child setting named name from the group setting. It returns the requested setting on success, or
NULLif the setting was not found or if setting is not a group.
This function fetches the element at the given index idx in the setting setting, which must be an array, list, or group. It returns the requested setting on success, or
NULLif idx is out of range or if setting is not an array, list, or group.
These functions return the value at the specified index idx in the setting setting. If the setting is not an array or list, or if the type of the element does not match the type requested, or if idx is out of range, they return 0 or
NULL. Storage for the string returned byconfig_setting_get_string_elem()is managed by the library and released automatically when the setting is destroyed or when its value is changed; the string must not be freed by the caller.
These functions set the value at the specified index idx in the setting setting to value. If idx is negative, a new element is added to the end of the array or list. On success, these functions return a pointer to the setting representing the element. If the setting is not an array or list, or if the setting is an array and the type of the array does not match the type of the value, or if idx is out of range, they return
NULL.config_setting_set_string_elem()makes a copy of the passed string value, so it may be subsequently freed or modified by the caller without affecting the value of the setting.
This function adds a new child setting or element to the setting parent, which must be a group, array, or list. If parent is an array or list, the name parameter is ignored and may be
NULL.The function returns the new setting on success, or
NULLif parent is not a group, array, or list; or if there is already a child setting of parent named name; or if type is invalid.
This function removes and destroys the setting named name from the parent setting parent, which must be a group. Any child settings of the setting are recursively destroyed as well.
The function returns
CONFIG_TRUEon success. If parent is not a group, or if it has no setting with the given name, it returnsCONFIG_FALSE.
This function removes the child setting at the given index idx from the setting parent, which must be a group, list, or array. Any child settings of the removed setting are recursively destroyed as well.
The function returns
CONFIG_TRUEon success. If parent is not a group, list, or array, or if idx is out of range, it returnsCONFIG_FALSE.
This function returns the root setting for the configuration config. The root setting is a group.
This function returns the name of the given setting, or
NULLif the setting has no name. Storage for the returned string is managed by the library and released automatically when the setting is destroyed; the string must not be freed by the caller.
This function returns the parent setting of the given setting, or
NULLif setting is the root setting.
This function returns
CONFIG_TRUEif the given setting is the root setting, andCONFIG_FALSEotherwise.
This function returns the index of the given setting within its parent setting. If setting is the root setting, this function returns -1.
This function returns the number of settings in a group, or the number of elements in a list or array. For other types of settings, it returns 0.
This function returns the type of the given setting. The return value is one of the constants
CONFIG_TYPE_INT,CONFIG_TYPE_INT64,CONFIG_TYPE_FLOAT,CONFIG_TYPE_STRING,CONFIG_TYPE_BOOL,CONFIG_TYPE_ARRAY,CONFIG_TYPE_LIST, orCONFIG_TYPE_GROUP.
These convenience functions, which are implemented as macros, test if the setting setting is of a given type. They return
CONFIG_TRUEorCONFIG_FALSE.
These convenience functions, which are implemented as macros, test if the setting setting is of an aggregate type (a group, array, or list), of a scalar type (integer, 64-bit integer, floating point, boolean, or string), and of a number (integer, 64-bit integer, or floating point), respectively. They return
CONFIG_TRUEorCONFIG_FALSE.
This function returns the line number of the configuration file or stream at which the setting setting was parsed. This information is useful for reporting application-level errors. If the setting was not read from a file or stream, or if the line number is otherwise unavailable, the function returns 0.
These functions make it possible to attach arbitrary data to each setting structure, for instance a “wrapper” or “peer” object written in another programming language. The destructor function, if one has been supplied via a call to
config_set_destructor(), will be called by the library to dispose of this data when the setting itself is destroyed. There is no default destructor.
This function assigns the destructor function destructor for the configuration config. This function accepts a single
void *argument and has no return value. Seeconfig_setting_set_hook()above for more information.
This chapter describes the C++ library API. The class Config
represents a configuration, and the class Setting represents a
configuration setting. Note that by design, neither of these classes
provides a public copy constructor or assignment operator. Therefore,
instances of these classes may only be passed between functions via
references or pointers.
The library defines a group of exceptions, all of which extend the
common base exception ConfigException.
A SettingTypeException is thrown when the type of a setting's
value does not match the type requested.
A SettingNotFoundException is thrown when a setting is not found.
A SettingNameException is thrown when an attempt is made to add
a new setting with a non-unique or invalid name.
A ParseException is thrown when a parse error occurs while
reading a configuration from a stream.
A FileIOException is thrown when an I/O error occurs while
reading/writing a configuration from/to a file.
The remainder of this chapter describes the methods for manipulating configurations and configuration settings.
These methods create and destroy
Configobjects.
The
read()method reads and parses a configuration from the given stream. AParseExceptionis thrown if a parse error occurs.The
write()method writes the configuration to the given stream.
The
readFile()method reads and parses a configuration from the file named filename. AParseExceptionis thrown if a parse error occurs. AFileIOExceptionis thrown if the file cannot be read.The
writeFile()method writes the configuration to the file named filename. AFileIOExceptionis thrown if the file cannot be written.
If a call to
readFile()orread()resulted in aParseException, these methods can be called on the exception object to obtain the text and line number of the parse error. Storage for the string returned bygetError()is managed by the library; the string must not be freed by the caller.
setAutoConvert()enables number auto-conversion for the configuration if flag istrue, and disables it otherwise. When this feature is enabled, an attempt to assign a floating point setting to an integer (or vice versa), or assign an integer to a floating point setting (or vice versa) will cause the library to silently perform the necessary conversion (possibly leading to loss of data), rather than throwing aSettingTypeException. By default this feature is disabled.
getAutoConvert()returnstrueif number auto-conversion is currently enabled for the configuration; otherwise it returnsfalse.
This method returns the root setting for the configuration, which is a group.
These methods locate the setting specified by the path path. If the requested setting is not found, a
SettingNotFoundExceptionis thrown.
These methods test if a setting with the given path exists in the configuration. They return
trueif the setting exists, and false otherwise. These methods do not throw exceptions.
These are convenience methods for looking up the value of a setting with the given path. If the setting is found and is of an appropriate type, the value is stored in value and the method returns
true. Otherwise, value is left unmodified and the method returns false. These methods do not throw exceptions.Storage for const char * values is managed by the library and released automatically when the setting is destroyed or when its value is changed; the string must not be freed by the caller. For safety and convenience, always assigning string values to a
std::stringis suggested.Since these methods have boolean return values and do not throw exceptions, they can be used within boolean logic expressions. The following example presents a concise way to look up three values at once and perform error handling if any of them are not found or are of the wrong type:
int var1; double var2; const char *var3; if(config.lookupValue("values.var1", var1) && config.lookupValue("values.var2", var2) && config.lookupValue("values.var3", var3)) { // use var1, var2, var3 } else { // error handling here }This approach also takes advantage of the short-circuit evaluation rules of C++, i.e., if the first lookup fails (returning
false), the remaining lookups are skipped entirely.
These cast operators allow a
Settingobject to be assigned to a variable of type bool if it is of typeTypeBoolean; int, unsigned int, long, or unsigned long if it is of typeTypeInt;long longorunsigned long longif it is of typeTypeInt64, float or double if it is of typeTypeFloat; or const char * or std::string if it is of typeTypeString.Storage for const char * return values is managed by the library and released automatically when the setting is destroyed or when its value is changed; the string must not be freed by the caller. For safety and convenience, always assigning string return values to a
std::stringis suggested.The following examples demonstrate this usage:
long width = config.lookup("application.window.size.w"); bool splashScreen = config.lookup("application.splash_screen"); std::string title = config.lookup("application.window.title");Note that certain conversions can lead to loss of precision or clipping of values, e.g., assigning a negative value to an unsigned int (in which case the value will be treated as 0), or a double-precision value to a float. The library does not treat these lossy conversions as errors.
Perhaps surprisingly, the following code in particular will cause a compiler error:
std::string title; . . . title = config.lookup("application.window.title");This is because the assignment operator of
std::stringis being invoked with aSetting &as an argument. The compiler is unable to make an implicit conversion because both theconst char *and thestd::stringcast operators ofSettingare equally appropriate. This is not a bug in libconfig; providing only theconst char *cast operator would resolve this particular ambiguity, but would cause assignments tostd::stringlike the one in the previous example to produce a compiler error. (To understand why, see section 11.4.1 of The C++ Programming Language.)The solution to this problem is to use an explicit conversion that avoids the construction of an intermediate
std::stringobject, as follows:
std::string title; . . . title = (const char *)config.lookup("application.window.title");If the assignment is invalid due to a type mismatch, a
SettingTypeExceptionis thrown.
These assignment operators allow values of type bool, int, long, long long, float, double, const char *, and std::string to be assigned to a setting. In the case of strings, the library makes a copy of the passed string value, so it may be subsequently freed or modified by the caller without affecting the value of the setting.
If the assignment is invalid due to a type mismatch, a
SettingTypeExceptionis thrown.
A
Settingobject may be subscripted with an integer index idx if it is an array or list, or with either a string name or an integer index idx if it is a group. For example, the following code would produce the string `Last Name' when applied to the example configuration in Configuration Files.
Setting& setting = config.lookup("application.misc"); const char *s = setting["columns"][0];If the setting is not an array, list, or group, a
SettingTypeExceptionis thrown. If the subscript (idx or name) does not refer to a valid element, aSettingNotFoundExceptionis thrown.Iterating over a group's child settings with an integer index will return the settings in the same order that they appear in the configuration.
These are convenience methods for looking up the value of a child setting with the given name. If the setting is found and is of an appropriate type, the value is stored in value and the method returns
true. Otherwise, value is left unmodified and the method returns false. These methods do not throw exceptions.Storage for const char * values is managed by the library and released automatically when the setting is destroyed or when its value is changed; the string must not be freed by the caller. For safety and convenience, always assigning string values to a
std::stringis suggested.Since these methods have boolean return values and do not throw exceptions, they can be used within boolean logic expressions. The following example presents a concise way to look up three values at once and perform error handling if any of them are not found or are of the wrong type:
int var1; double var2; const char *var3; if(setting.lookupValue("var1", var1) && setting.lookupValue("var2", var2) && setting.lookupValue("var3", var3)) { // use var1, var2, var3 } else { // error handling here }This approach also takes advantage of the short-circuit evaluation rules of C++, e.g., if the first lookup fails (returning
false), the remaining lookups are skipped entirely.
These methods add a new child setting with the given name and type to the setting, which must be a group. They return a reference to the new setting. If the setting already has a child setting with the given name, or if the name is invalid, a
SettingNameExceptionis thrown. If the setting is not a group, aSettingTypeExceptionis thrown.Once a setting has been created, neither its name nor type can be changed.
This method adds a new element to the setting, which must be of type
TypeArrayorTypeList. If the setting is an array which currently has zero elements, the type parameter (which must beTypeInt,TypeInt64,TypeFloat,TypeBool, orTypeString) determines the type for the array; otherwise it must match the type of the existing elements in the array.The method returns the new setting on success. If type is a scalar type, the new setting will have a default value of 0, 0.0,
false, orNULL, depending on the type.The method throws a
SettingTypeExceptionif the setting is not an array or list, or if type is invalid.
These methods remove the child setting with the given name from the setting, which must be a group. Any child settings of the removed setting are recursively destroyed as well.
If the setting is not a group, a
SettingTypeExceptionis thrown. If the setting does not have a child setting with the given name, aSettingNotFoundExceptionis thrown.
This method removes the child setting at the given index idx from the setting, which must be a group, list, or array. Any child settings of the removed setting are recursively destroyed as well.
If the setting is not a group, list, or array, a
SettingTypeExceptionis thrown. If idx is out of range, aSettingNotFoundExceptionis thrown.
This method returns the name of the setting, or
NULLif the setting has no name. Storage for the returned string is managed by the library and released automatically when the setting is destroyed; the string must not be freed by the caller. For safety and convenience, consider assigning the return value to astd::string.
This method returns the complete dot-separated path to the setting. Settings which do not have a name (list and array elements) are represented by their index in square brackets.
This method returns the parent setting of the setting. If the setting is the root setting, a
SettingNotFoundExceptionis thrown.
This method returns true if the setting is the root setting, and false otherwise.
This method returns the index of the setting within its parent setting. When applied to the root setting, this method returns -1.
This method returns the type of the setting. The
Setting::Typeenumeration consists of the following constants:TypeInt,TypeInt64,TypeFloat,TypeString,TypeBoolean,TypeArray,TypeList, orTypeGroup.
These methods get and set the external format for the setting.
The Setting::Format enumeration consists of the following constants:
FormatDefault,FormatHex. All settings support theFormatDefaultformat. TheFormatHexformat specifies hexadecimal formatting for integer values, and hence only applies to settings of typeTypeIntandTypeInt64. If format is invalid for the given setting, it is ignored.
These methods test if the setting has a child setting with the given name. They return
trueif the setting exists, and false otherwise. These methods do not throw exceptions.
This method returns the number of settings in a group, or the number of elements in a list or array. For other types of settings, it returns 0.
These convenience methods test if a setting is of a given type.
These convenience methods test if a setting is of an aggregate type (a group, array, or list), of a scalar type (integer, 64-bit integer, floating point, boolean, or string), and of a number (integer or floating point), respectively.
This method returns the line number of the configuration file or stream at which the setting was parsed. This information is useful for reporting application-level errors. If the setting was not read from a file or stream, or if the line number is otherwise unavailable, the method returns 0.
Below is the BNF grammar for configuration files. Comments are not part of the grammar, and hence are not included here.
configuration = setting-list | empty
empty =
setting-list = setting | setting-list setting
setting = name (":" | "=") value ";"
value = scalar-value | array | list | group
value-list = value | value-list "," value
scalar-value = boolean | integer | integer64 | hex | hex64 | float
| string
scalar-value-list = scalar-value | scalar-value-list "," scalar-value
array = "[" (scalar-value-list | empty) "]"
list = "(" (value-list | empty) ")"
group = "{" (setting-list | empty) "}"
Terminals are defined below as regular expressions:
boolean |
([Tt][Rr][Uu][Ee])|([Ff][Aa][Ll][Ss][Ee])
|
string |
\"([^\"\\]|\\.)*\"
|
name |
[A-Za-z\*][-A-Za-z0-9_\*]*
|
integer |
[-+]?[0-9]+
|
integer64 |
[-+]?[0-9]+L(L)?
|
hex |
0[Xx][0-9A-Fa-f]+
|
hex64 |
0[Xx][0-9A-Fa-f]+L(L)?
|
float |
([-+]?([0-9]*)?\.[0-9]*([eE][-+]?[0-9]+)?)|([-+]([0-9]+)(\.[0-9]*)?[eE][-+]?[0-9]+)
|
Copyright © 1991, 1999 Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed.
[This is the first released version of the Lesser GPL. It also counts as the successor of the GNU Library Public License, version 2, hence the version number 2.1.]
The licenses for most software are designed to take away your freedom to share and change it. By contrast, the GNU General Public Licenses are intended to guarantee your freedom to share and change free software–to make sure the software is free for all its users.
This license, the Lesser General Public License, applies to some specially designated software packages–typically libraries–of the Free Software Foundation and other authors who decide to use it. You can use it too, but we suggest you first think carefully about whether this license or the ordinary General Public License is the better strategy to use in any particular case, based on the explanations below.
When we speak of free software, we are referring to freedom of use, not price. Our General Public Licenses are designed to make sure that you have the freedom to distribute copies of free software (and charge for this service if you wish); that you receive source code or can get it if you want it; that you can change the software and use pieces of it in new free programs; and that you are informed that you can do these things.
To protect your rights, we need to make restrictions that forbid distributors to deny you these rights or to ask you to surrender these rights. These restrictions translate to certain responsibilities for you if you distribute copies of the library or if you modify it.
For example, if you distribute copies of the library, whether gratis or for a fee, you must give the recipients all the rights that we gave you. You must make sure that they, too, receive or can get the source code. If you link other code with the library, you must provide complete object files to the recipients, so that they can relink them with the library after making changes to the library and recompiling it. And you must show them these terms so they know their rights.
We protect your rights with a two-step method: (1) we copyright the library, and (2) we offer you this license, which gives you legal permission to copy, distribute and/or modify the library.
To protect each distributor, we want to make it very clear that there is no warranty for the free library. Also, if the library is modified by someone else and passed on, the recipients should know that what they have is not the original version, so that the original author's reputation will not be affected by problems that might be introduced by others.
Finally, software patents pose a constant threat to the existence of any free program. We wish to make sure that a company cannot effectively restrict the users of a free program by obtaining a restrictive license from a patent holder. Therefore, we insist that any patent license obtained for a version of the library must be consistent with the full freedom of use specified in this license.
Most GNU software, including some libraries, is covered by the ordinary GNU General Public License. This license, the GNU Lesser General Public License, applies to certain designated libraries, and is quite different from the ordinary General Public License. We use this license for certain libraries in order to permit linking those libraries into non-free programs.
When a program is linked with a library, whether statically or using a shared library, the combination of the two is legally speaking a combined work, a derivative of the original library. The ordinary General Public License therefore permits such linking only if the entire combination fits its criteria of freedom. The Lesser General Public License permits more lax criteria for linking other code with the library.
We call this license the “Lesser” General Public License because it does Less to protect the user's freedom than the ordinary General Public License. It also provides other free software developers Less of an advantage over competing non-free programs. These disadvantages are the reason we use the ordinary General Public License for many libraries. However, the Lesser license provides advantages in certain special circumstances.
For example, on rare occasions, there may be a special need to encourage the widest possible use of a certain library, so that it becomes a de-facto standard. To achieve this, non-free programs must be allowed to use the library. A more frequent case is that a free library does the same job as widely used non-free libraries. In this case, there is little to gain by limiting the free library to free software only, so we use the Lesser General Public License.
In other cases, permission to use a particular library in non-free programs enables a greater number of people to use a large body of free software. For example, permission to use the GNU C Library in non-free programs enables many more people to use the whole GNU operating system, as well as its variant, the GNU/Linux operating system.
Although the Lesser General Public License is Less protective of the users' freedom, it does ensure that the user of a program that is linked with the Library has the freedom and the wherewithal to run that program using a modified version of the Library.
The precise terms and conditions for copying, distribution and modification follow. Pay close attention to the difference between a “work based on the library” and a “work that uses the library”. The former contains code derived from the library, whereas the latter must be combined with the library in order to run.
A “library” means a collection of software functions and/or data prepared so as to be conveniently linked with application programs (which use some of those functions and data) to form executables.
The “Library”, below, refers to any such software library or work which has been distributed under these terms. A “work based on the Library” means either the Library or any derivative work under copyright law: that is to say, a work containing the Library or a portion of it, either verbatim or with modifications and/or translated straightforwardly into another language. (Hereinafter, translation is included without limitation in the term “modification”.)
“Source code” for a work means the preferred form of the work for making modifications to it. For a library, complete source code means all the source code for all modules it contains, plus any associated interface definition files, plus the scripts used to control compilation and installation of the library.
Activities other than copying, distribution and modification are not covered by this License; they are outside its scope. The act of running a program using the Library is not restricted, and output from such a program is covered only if its contents constitute a work based on the Library (independent of the use of the Library in a tool for writing it). Whether that is true depends on what the Library does and what the program that uses the Library does.
You may charge a fee for the physical act of transferring a copy, and you may at your option offer warranty protection in exchange for a fee.
(For example, a function in a library to compute square roots has a purpose that is entirely well-defined independent of the application. Therefore, Subsection 2d requires that any application-supplied function or table used by this function must be optional: if the application does not supply it, the square root function must still compute square roots.)
These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Library, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Library, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it.
Thus, it is not the intent of this section to claim rights or contest your rights to work written entirely by you; rather, the intent is to exercise the right to control the distribution of derivative or collective works based on the Library.
In addition, mere aggregation of another work not based on the Library with the Library (or with a work based on the Library) on a volume of a storage or distribution medium does not bring the other work under the scope of this License.
Once this change is made in a given copy, it is irreversible for that copy, so the ordinary GNU General Public License applies to all subsequent copies and derivative works made from that copy.
This option is useful when you wish to copy part of the code of the Library into a program that is not a library.
If distribution of object code is made by offering access to copy from a designated place, then offering equivalent access to copy the source code from the same place satisfies the requirement to distribute the source code, even though third parties are not compelled to copy the source along with the object code.
However, linking a “work that uses the Library” with the Library creates an executable that is a derivative of the Library (because it contains portions of the Library), rather than a “work that uses the library”. The executable is therefore covered by this License. Section 6 states terms for distribution of such executables.
When a “work that uses the Library” uses material from a header file that is part of the Library, the object code for the work may be a derivative work of the Library even though the source code is not. Whether this is true is especially significant if the work can be linked without the Library, or if the work is itself a library. The threshold for this to be true is not precisely defined by law.
If such an object file uses only numerical parameters, data structure layouts and accessors, and small macros and small inline functions (ten lines or less in length), then the use of the object file is unrestricted, regardless of whether it is legally a derivative work. (Executables containing this object code plus portions of the Library will still fall under Section 6.)
Otherwise, if the work is a derivative of the Library, you may distribute the object code for the work under the terms of Section 6. Any executables containing that work also fall under Section 6, whether or not they are linked directly with the Library itself.
You must give prominent notice with each copy of the work that the Library is used in it and that the Library and its use are covered by this License. You must supply a copy of this License. If the work during execution displays copyright notices, you must include the copyright notice for the Library among them, as well as a reference directing the user to the copy of this License. Also, you must do one of these things:
For an executable, the required form of the “work that uses the Library” must include any data and utility programs needed for reproducing the executable from it. However, as a special exception, the materials to be distributed need not include anything that is normally distributed (in either source or binary form) with the major components (compiler, kernel, and so on) of the operating system on which the executable runs, unless that component itself accompanies the executable.
It may happen that this requirement contradicts the license restrictions of other proprietary libraries that do not normally accompany the operating system. Such a contradiction means you cannot use both them and the Library together in an executable that you distribute.
If any portion of this section is held invalid or unenforceable under any particular circumstance, the balance of the section is intended to apply, and the section as a whole is intended to apply in other circumstances.
It is not the purpose of this section to induce you to infringe any patents or other property right claims or to contest validity of any such claims; this section has the sole purpose of protecting the integrity of the free software distribution system which is implemented by public license practices. Many people have made generous contributions to the wide range of software distributed through that system in reliance on consistent application of that system; it is up to the author/donor to decide if he or she is willing to distribute software through any other system and a licensee cannot impose that choice.
This section is intended to make thoroughly clear what is believed to be a consequence of the rest of this License.
Each version is given a distinguishing version number. If the Library specifies a version number of this License which applies to it and “any later version”, you have the option of following the terms and conditions either of that version or of any later version published by the Free Software Foundation. If the Library does not specify a license version number, you may choose any version ever published by the Free Software Foundation.
If you develop a new library, and you want it to be of the greatest possible use to the public, we recommend making it free software that everyone can redistribute and change. You can do so by permitting redistribution under these terms (or, alternatively, under the terms of the ordinary General Public License).
To apply these terms, attach the following notices to the library. It is safest to attach them to the start of each source file to most effectively convey the exclusion of warranty; and each file should have at least the “copyright” line and a pointer to where the full notice is found.
<one line to give the library's name and a brief idea of what it does.> Copyright (C) <year> <name of author> This library is free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details. You should have received a copy of the GNU Lesser General Public License along with this library; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Also add information on how to contact you by electronic and paper mail.
You should also get your employer (if you work as a programmer) or your school, if any, to sign a “copyright disclaimer” for the library, if necessary. Here is a sample; alter the names:
Yoyodyne, Inc., hereby disclaims all copyright interest in the library `Frob' (a library for tweaking knobs) written by James Random Hacker. <signature of Ty Coon>, 1 April 1990 Ty Coon, President of Vice
That's all there is to it!
add on Setting: The C++ APIConfig on Config: The C++ APIconfig_destroy: The C APIconfig_error_line: The C APIconfig_error_text: The C APIconfig_get_auto_convert: The C APIconfig_init: The C APIconfig_lookup: The C APIconfig_lookup_bool: The C APIconfig_lookup_float: The C APIconfig_lookup_int: The C APIconfig_lookup_int64: The C APIconfig_lookup_string: The C APIconfig_read: The C APIconfig_read_file: The C APIconfig_root_setting: The C APIconfig_set_auto_convert: The C APIconfig_set_destructor: The C APIconfig_setting_add: The C APIconfig_setting_get_bool: The C APIconfig_setting_get_bool_elem: The C APIconfig_setting_get_elem: The C APIconfig_setting_get_float: The C APIconfig_setting_get_float_elem: The C APIconfig_setting_get_format: The C APIconfig_setting_get_hook: The C APIconfig_setting_get_int: The C APIconfig_setting_get_int64: The C APIconfig_setting_get_int64_elem: The C APIconfig_setting_get_int_elem: The C APIconfig_setting_get_member: The C APIconfig_setting_get_string: The C APIconfig_setting_get_string_elem: The C APIconfig_setting_index: The C APIconfig_setting_is_aggregate: The C APIconfig_setting_is_array: The C APIconfig_setting_is_group: The C APIconfig_setting_is_list: The C APIconfig_setting_is_number: The C APIconfig_setting_is_root: The C APIconfig_setting_is_scalar: The C APIconfig_setting_length: The C APIconfig_setting_name: The C APIconfig_setting_parent: The C APIconfig_setting_remove: The C APIconfig_setting_remove_elem: The C APIconfig_setting_set_bool: The C APIconfig_setting_set_bool_elem: The C APIconfig_setting_set_float: The C APIconfig_setting_set_float_elem: The C APIconfig_setting_set_format: The C APIconfig_setting_set_hook: The C APIconfig_setting_set_int: The C APIconfig_setting_set_int64: The C APIconfig_setting_set_int64_elem: The C APIconfig_setting_set_int_elem: The C APIconfig_setting_set_string: The C APIconfig_setting_set_string_elem: The C APIconfig_setting_source_line: The C APIconfig_setting_type: The C APIconfig_write: The C APIconfig_write_file: The C APIexists on Config: The C++ APIexists on Setting: The C++ APIgetAutoConvert on Config: The C++ APIgetError on ParseException: The C++ APIgetFormat on Setting: The C++ APIgetIndex on Setting: The C++ APIgetLength on Setting: The C++ APIgetLine on ParseException: The C++ APIgetName on Setting: The C++ APIgetParent on Setting: The C++ APIgetPath on Setting: The C++ APIgetRoot on Config: The C++ APIgetSourceLine on Setting: The C++ APIgetType on Setting: The C++ APIisAggregate on Setting: The C++ APIisArray on Setting: The C++ APIisGroup on Setting: The C++ APIisList on Setting: The C++ APIisNumber on Setting: The C++ APIisRoot on Setting: The C++ APIisScalar on Setting: The C++ APIlookup on Config: The C++ APIlookupValue on Config: The C++ APIlookupValue on Setting: The C++ APIoperator bool() on Setting: The C++ APIoperator const char *() on Setting: The C++ APIoperator double() on Setting: The C++ APIoperator float() on Setting: The C++ APIoperator int() on Setting: The C++ APIoperator long long() on Setting: The C++ APIoperator long() on Setting: The C++ APIoperator std::string() on Setting: The C++ APIoperator unsigned int() on Setting: The C++ APIoperator unsigned long long() on Setting: The C++ APIoperator unsigned long() on Setting: The C++ APIoperator= on Setting: The C++ APIoperator[] on Setting: The C++ APIread on Config: The C++ APIreadFile on Config: The C++ APIremove on Setting: The C++ APIsetAutoConvert on Config: The C++ APIsetFormat on Setting: The C++ APIwrite on Config: The C++ APIwriteFile on Config: The C++ API~Config on Config: The C++ APIConfig: The C++ APIconfig_setting_t: The C APIconfig_t: The C APIConfigException: The C++ APIFileIOException: The C++ APIParseException: The C++ APISetting: The C++ APISetting::Format: The C++ APISetting::Type: The C++ APISettingFormat: The C APISettingNameException: The C++ APISettingNotFoundException: The C++ APISettingTypeException: The C++ API