v. 0.8.0
| Revision History | ||
|---|---|---|
| Revision 3.0 | 18 November 2009 | dn |
| Update to 0.8.0 | ||
| Revision 2.9 | 6 November 2009 | dn |
| Update to 0.7.7 | ||
| Revision 2.8 | 13 July 2009 | dn |
| Update to 0.7.6 | ||
| Revision 2.7 | 22 June 2009 | dn |
| Update to 0.7.5 | ||
| Revision 2.6 | 4 April 2009 | dn |
| Update to 0.7.4 | ||
| Revision 2.4 | 3 December 2008 | dn |
| Update to 0.7.2 | ||
| Revision 2.2 | 6 November 2008 | dn |
| Update to 0.7.1 | ||
| Revision 2.1 | 17 July 2008 | dn |
| Update to 0.7.0 | ||
| Revision 2.0 | 2 March 2007 | dn |
| Update to 0.6.2 | ||
| Revision 1.9 | 8 February 2007 | dn |
| Update to 0.6.1 | ||
| Revision 1.8 | 19 January 2007 | dn |
| Update to 0.6.0 | ||
| Revision 1.7 | 10 October 2006 | dn |
| Update to 0.5.3 | ||
| Revision 1.6 | 15 September 2006 | dn |
| Update to 0.5.2 | ||
| Revision 1.5 | 8 June 2006 | dn |
| Update to 0.5.1 | ||
| Revision 1.4 | 7 May 2006 | dn |
| Update to 0.5.0, extensive updates, new content | ||
| Revision 1.3 | 30 Nov 2005 | dn |
| Update to 0.4.2 | ||
| Revision 1.2 | 26 Oct 2005 | dn |
| Update to 0.4.0 | ||
| Revision 1.1 | 14 Oct 2005 | hw |
| Converted to docbook.xml | ||
| Revision 1.0 | 2 Aug 2005 | hw |
| First release | ||
Table of Contents
The Qore programming language is a powerful, thread-capable, embeddable weakly-typed language with optional strong typing and procedural and object-oriented features designed for anything from quick scripting to complex multithreaded, network-aware application development to embedded application scripting. Qore was initially designed to facilitate the rapid implementation of sophisticated interfaces in embedded code in an enterprise environment, and has since grown into a general-purpose language as well.
Qore exports a C++ API to allow programs or libraries to embed Qore code; this manual documents Qore's user-level features, for more information about Qore's C++ API, see the Qore's home page.
Qore features database, XML, and JSON integration into the syntax, functions, data structures, and operators of the language.
Flexible character encoding support is also built-in to Qore strings, and automatic character encoding conversions are supported, enabling correct behavior when working in an environment with mixed character encoding requirements (see Qore Strings and Character Encoding).
Qore includes the following design points:
Support for Embedded Logic
Qore was designed to support embedding logic in applications; this also applies to applications written in Qore as well as applications using the Qore library's public C++ API. By using the Program class, discrete objects can be created and destroyed at will containing embedded code to extend or modify the behavior of your application in user-defined ways. The Program class allows the capabilities of embedded code to be arbitrarily restricted as well.
Thread Safety and SMP Scalability
All elements of Qore are thread-safe, and the language in general has been designed with SMP scalability in mind. The internal design and implementation of Qore favors multithreaded performance over single-threaded performance, so multithreaded Qore programs can count on an efficient and stable execution platform, and do not have to limit themselves to a subset of Qore's functionality (see Threading). Additionally, Qore includes optimizations designed to reduce the number of SMP cache invalidations that provide a substantial performance boost on SMP machines.
Qore supports deadlock detection in complex locking scenarios and will throw an exception rather than allow an operation to be performed that would cause a deadlock. Furthermore, Qore's threading primitives detect threading errors and throw exceptions in these cases as well.
Database Integration and DBI Layer
Retrieving, comparing, and manipulating data in a consistent manner from heterogenous database types is made possible by Qore's built-in database integration. Qore was designed with a database independent interfacing (DBI) layer, providing a standard interface for Qore programs to access any database supported by a Qore DBI driver (see the Datasource Class).
XML and JSON Integration
Qore uses the libxml2 library to provide fast and efficient XML functionality. Qore's XML integration enables serialization from Qore data structures to XML strings and deserialization from XML strings to Qore data structures, making it trivial to work with data in XML format (see XML Integration). Qore provides JSON and JSON-RPC integration as well. There is also a module providing YAML serialization and deserialization support.
Function and Class Library
Qore's basic functionality covers areas such as: POSIX-compliant command-line parsing (ex: GetOpt Class), XML and JSON serialization/deserialization, strong encryption and digest calculation, thread synchronization (ex: Queue Class, Mutex Class, Condition Class, etc), working with files (File class), socket, HTTP, and higher-level protocol communication (Socket, HTTPClient, XmlRpcClient, JsonRpcClient, FtpClient classes, optionally with TLS/SSL encryption), support for dynamic embedded application logic (Program Class). Additionally, Qore's functionality is extended with modules delivered separately from the Qore library (see Qore's home page for more information.
Logical Syntax
Qore syntax is similar to other programming languages, allowing new programmers to rapidly come up to speed in Qore. Qore borrows features from languages such as: C++ (ex: multiple inheritance, exception handling, static methods), Java (ex: the synchronized keyword, the instanceof operator, object and class implementation), Perl (ex: the foreach statement, splice, push, pop, chomp, splice operators, perl5-compatible regular expressions, and more), the D Programming Language (the on_exit, on_success, and on_error statements provide exception-aware functionality similar to scope(exit), scope(failure), allowing exception-aware cleanup code to be placed next to the code requiring cleanup), and others, also with many features unique to Qore. Furthermore, Qore supports closures (including binding local variables in the closure in a way that is safe to use even in multithreaded contexts) and features for advanced list processing (map, foldl, foldr, and select).
Qore's operators are designed to produce the expected results for the programmer even when data types are mixed, a feature meant to further flatten the learning curve for new programmers.
Simple or complex interfaces involving data extraction and manipulation from multiple Oracle, MySQL, PostgreSQL and other databases and/or multiple applications connected with a supported messaging bus or through lightweight web service protocols bus can be developed quickly and easily with Qore, particularly in comparison with Java, Perl, or C/C++. Furthermore, solutions based on Qore are transparent and easily maintainable, bringing advantages for the developer and end-user alike.
Additionally, Qore is thoroughly tested with valgrind on Linux and dbx on Solaris for memory leaks and memory errors. While it is possible that there are bugs in some less tested code paths, every effort is made to eliminate all bugs before each new release, particularly memory bugs and race conditions, in order to ensure that Qore releases are of the highest possible quality.
Examples of Qore code are given in the following font:
#!/usr/bin/qore # # this is an example of Qore code
keywords are given in bold:
# "if" is a key word
if ($a == 1) print("yes\n");
placeholders are given in italics:
if (expression)statement
where expressions and statements are defined in this document. The text above indicates that any valid expression and any valid statement may be used in the positions indicated.
Optional text is given in square brackets "[" and "]" as follows:
sub name([$variable1, $variable2, ...])
indicating that an arbitrarily long list of variable names separated by commas (or nothing at all) may appear in subroutine (function) declarations.
A Qore program is composed of a series of declarations, statements, subroutine definitions, and/or class definitions. Non-block statements are terminated by a semi-colon ";". Block statements are grouped by using curly brackets ("{" and "}"), as in C, C++, Java, and Perl.
Programmers familiar with C, C++, Java, and/or Perl should find the standard Qore syntax intuitive and should be productive fairly quickly with the language. However Qore has unique features that differentiate it from other languages, and these features must be mastered in order to leverage the full power of Qore.
Qore programs/scripts are free form. Formatting does not affect the execution of the program; formatting is at the discretion of the programmer and should be used to enhance readability and clarity of the source code.
Qore was creaetd as a weakly typed language. That means that variables without type restructions can hold values of any type and subroutines with out type restrictions can return any value type (or none at all, see Variables) and take arguments of any type. Furthermore list elements can be of any type (they do not have to be uniform), and multidimensional lists can have a different number of elements in each list. The same type flexibility holds true of hashes, objects, and all combinations of container types.
Qore can be used a a traditional subroutine-based scripting language or as a pure object-oriented language, where the application is defined as a class. Aside from traditional local and global variables, constants, and subroutines, Qore also supports nested namespaces, classes, multiple inheritance, overriding base class constructor arguments, private members and methods, and static class methods.
All elements of Qore are designed to work together seamlessly: database access, XML transformations, socket communication, embedding logic in subprograms, regular expressions, operators, functions, and all other elements are thread-safe and built on an execution engine that was designed for SMP scalability.
Qore automatically converts data types when necessary when evaluating operators. The goal is to provide the expected result for the programmer without requiring the programmer to explicitly convert data types. Please see Operators for more information.
Qore supports signal handling by dispatching Qore-language signal handlers in a safe manner, asynchronously to the actual receipt of the signal.
UNIX operating systems allow an executable script to specify their interpreter. This is done by setting the first line in the program to a special string indicating the location of the Qore binary. For the purposes of this document, the location for the Qore binary is assumed to be /usr/bin/qore. The first line of Qore scripts in this case should look as follows:
#!/usr/bin/qore
If another installation directory is used (such as /usr/local/bin), then the correct path must be reflected in the first line of the Qore script.
Qore convention dictates that Qore script file names end with .q.
This section will outline the environment variables that are used by Qore.
Table 2.1. Qore Environment Variables
Environment Variable | Description |
|---|---|
| This environment variable should contain a colon-separated list of directories which will be searched for Qore modules when Qore starts. If any modules are found, they are loaded automatically before any parsing starts. |
| This environment variable should contain a colon-separated list of directories which will be searched when modules are loaded with the |
| This variable should be a colon-separated list of directories where the Qore binary should look for include files |
| If this variable is set, then the default character encoding name for the process will be the value of this variable. This variable takes precedence over the |
| If this variable is set and includes a character encoding specification, then, if the |
Qore modules allow the Qore language to be extended at run-time. Qore modules must conform to the Qore Module API and have the file extension *.qmod. Qore modules normally depend on other shared libraries and therefore can only be loaded if the libraries they require are present on the system and can be found in the library path.
Please note that as of version 0.7.0, modules are no longer delivered with the Qore library. Modules must be downloaded separately; see Qore's home page for more information.
To load a module at parse time (normally required for most modules), use the %requires parse directive. If the named feature is not already present in Qore, Qore looks for a module with this name in the directories listed in the QORE_MODULE_DIR environment variable.
Use the load_module() function to load Qore modules at run-time; however, note that any module providing parse support (classes, constants, functions, etc) must be loaded at parse time using the %requires directive.
From Qore 0.7.1, you can specify a comparison operator (one of < <=, =, >=, or >) and version information after the module name as well. Version numbers are compared via integer comparisons of each element, where elements are separated by a '.'. If one of the versions does not have as many elements as another, the missing elements are assumed to be '0' (i.e. version "1.0" compared with version "1.0.1" will be exteneded to "1.0.0").
Also note that DBI drivers are loaded on demand by the Datasource and DatasourcePool classes.
At the time of writing this documentation, the following modules exist for Qore:
Table 2.2. Qore Modules
Module | Description |
|---|---|
| Provides ASN.1 functionality to Qore |
| Provides |
| Provides SSH2 and SFTP functionality to Qore. |
| Provides TIBCO Rendezvous(R) (TIBCO, Inc) functionality to Qore. |
| Provides TIBCO ActiveEnterprise(TM) (TIBCO, Inc) functionality to Qore. |
| Provides Oracle (ex Bea) Tuxedo functionality to Qore. |
| Provides a MySQL DBI driver to Qore. |
| Provides a PostgreSQL DBI driver to Qore. |
| Provides a SQLite3 DBI driver to Qore. |
| Provides a Sybase DBI driver to Qore. |
| Provides a FreeTDS-based DBI driver to Qore. |
| Provides an Oracle DBI driver to Qore. |
| Provides an IBM DB2 driver to Qore. |
| Provides GLUT functionality to Qore. |
| Provides an OpenGL API to Qore. |
| BLACKLISTED: old QT4 modules: please use the much more complete qt4 module. These old modules (never officially released) use a namespace management trick that does not work with qore 0.8.0+, so they will not load with newer versions of qore. |
| Provides Nokia (formerly Trolltech) QT4 APIs for GUI development in Qore. |
| Provides curses APIs to Qore. |
| Provides YAML functionality to Qore. |
A Qore program can include other program code to be used and executed in the current program by using the %include directive. The %include directive must be the first text on the line, and the file name to include must follow. All text on the line after the %include directive will be used for the file name to include. The file name should not be quoted.
Here is an example:
#!/usr/bin/qore %include /usr/qore/lib/functions.lib
After this, any variable, subroutine, namespace, constant, or object declared in the above file can be used by the Qore program.
The QORE_INCLUDE_DIR environment variable determines the search path for include files.
Qore identifiers must start with an alphabetic character, and then may contain any number of alphabetic, numeric, or "_" characters. There is no length limit on Qore identifiers.
All Qore identifiers are case-sensitive, therefore the identifier hello_there is not the same as Hello_There or HELLO_THERE.
The following are examples of valid Qore identifiers:
Table 2.3. Examples of Valid Qore Identifiers
Identifier | Description |
|---|---|
| Simple one-character identifier |
| Identifier with number |
| Long identifier with underline characters |
| Identifier with underline and number |
| Mixed case identifier name |
| Identifier in all capital letters |
The following are invalid identifiers:
Table 2.4. Examples of Invalid Qore Identifiers
Identifier | Description |
|---|---|
| Does not start with an alphabetic character |
| Contains "-" characters |
| Contains "#" character |
Comments are allowed in Qore scripts; line comments are preceded by a hash "#", and block comments are made C-style, ex:
# this is a line comment /* this is a block comment */
For line comments, all text following a hash until the end-of-line character "\n" is considered a part of the comment.
For block comments, all text between the /* and */ is ignored by the parser.
Here is an example Qore script containing comments:
#!/usr/bin/qore
#
# these are line comments
# another line comment
/*
--- this text is in block comments
print("hello"); <- this won't get executed
--- because it's in the block comment
*/Variables are Qore identifiers prefixed by a "$" sign, similar to Perl. The data type of variables does not have to be declared in advance, and variable types are assigned and changed automatically as appropriate in the context of the operations being performed on them or the values being assigned. Any Qore variable can hold any Qore data type including container types (lists, hashes, or objects).
A few variables are set by the Qore language during the execution of Qore programs. These are normal variables that can be reassigned to other values by the user if necessary.
Table 2.5. Special Qore Variables
Variable | Type | Data Type | Explanation |
|---|---|---|---|
| Local | List | automatically assigned local variable containing the list of subroutine or method arguments that were not assigned to parameter variables (see Subroutines and Implicit Arguments for supporting related information) |
| Global | List | script command-line arguments (use the GetOpt Class to parse command-line arguments) |
| Global | List | complete qore command-line arguments |
| Global | Hash | UNIX program environment |
As of version 0.5.0, $STDERR and $STDOUT have been removed from Qore. Use the I/O constants stderr, stdout, and stdin constants of the File Class instead.
Variables not in a parameter list automatically have global scope unless the first reference is prefixed with my. Variable names in a parameter list are always local to their associated subroutine, method, or catch block. Global variables can be explicitly declared with our. The our keyword is required if the parse option PO_REQUIRE_OUR (-O or --require-our command-line option) is set for the parent program. See the section on Parse Options for more information.
Variables may be assigned any value unless restricted with a type declaration. If no type declaration is given, then the variable is assumed to be type any. Note that type declarations are required for all variables (and for function and method parameters and class members) when the PO_REQUIRE_TYPES parse option is set.
Note that variable types can only be declared when the my or our keywords are used; it is a syntax error to declare a variable's type without declaring its scope at the same time with one of these keywords.
Local variables are not shared between threads (local variables have a distinct value in each thread), however global variables are. See Threading (and in particular Threading and Variables) for more information.
For example (in the following script, the our keyword is optional):
#!/usr/bin/qore # # variable scoping example our int $a = 1; # this is a global variable our (string $b, any $c, hash $d); # list of global variables if ($a == 1) { my int $a = 2; my (string $b, any $c); # $a, $b, and $c are local variables, # the use of which will not affect the # global variables of the same name print("local a = %d\n", $a); } print("global a = %d\n", $a);
The first print() statement will output:
local a = 2
The second print() statement will output:
global a = 1
The following are the basic data types in Qore (see Container Data Types for container data types):
Table 2.6. Basic data types
Type | Description | Example | Default Value |
|---|---|---|---|
True or |
| False | |
A sequence of characters |
| Empty string | |
A 64-bit signed integer |
| 0 | |
A double-precision floating-point number |
| 0.0 | |
A date/time value with an optional time zone component, with resolution to the microsecond. |
| 1970-01-01Z | |
An opaque binary object | n/a | an empty object of size 0 | |
Corresponds to a NULL value in a database query (not equivalent to NOTHING) |
|
| |
Represents the state of a variable having no value or function returning no value (not equivalent to NULL) |
|
|
The Boolean type can have two values, True and False. When converting other types to a Boolean, any value that converts to a non-zero integer will be evaluated as True. Otherwise the result of the conversion will be False.
String values are specified with text between double or single quotes. Text between double quotes is subject to interpretation of escape characters. Please see String Formatting for more information.
Strings are assumed by default to have the encoding given by the QORE_CHARSET or the LANG environment variable. If neither of these variables is set, then all strings will be assumed to have UTF-8 encoding.
For detailed information on Qore character encoding handling, please see Qore Strings and Character Encoding.
It is legal to specify a string literal with newline characters like the following:
$str = "this string is a multiline string";
Internally, strings are stored as a pointer to the string data, an unsigned integer giving the length of the string, and a pointer to an encoding object, giving the string's character encoding.
Qore floats are double precision floating-point numbers (C/C++ type double), normally a 64-bit value.
Qore dates have a time component supporting a resolution to the microsecond and can be either absolute or relative.
See Date and Time Functions for a list of functions related to date/time processing.
Absolute date/time values specify a specific point in time in a certain time zone, such as January 1, 2005 10:35:00 +01:00. They are stored interally as a 64-bit signed offset from the Qore epoch (1970-01-01Z), a positive 4-byte integer for microseconds, and a pointer to a time zone description object that provides the UTC offset and daylight savings time information (see Time Zones for more information). Note that all absolute date/time values in Qore are stored internally in UTC and are converted for display purposes to the representation of wall time in their tagged time zone.
Absolute date/time values can be specified with a syntax based on ISO-8601 date formats as follows:
YYYY-MM-DD[THH:mm:SS[.n*]][Z|[+-]HH[:mm[:SS]]]Note that if no time zone information is given, the local time zone will be assumed. If a time zone UTC offset is given, it is given in units of time east of UTC (i.e. +05:00 means five hours east of UTC).
Or an alternative format (with a '-' instead ofa 'T' to separate the time component):
YYYY-MM-DD[-HH:mm:SS[.n*]][Z|[+-]HH[:mm[:SS]]]for example, for just the date in UTC, without a time component:
2010-05-26or, for just the time, without a date component (note that in this case the date component will be set to Jan 1, 1970, in order for time arithmetic to function properly and will also be tagged with the local time zone):
20:05:10.458342Some further examples (note that the date/time values without a time zone specification here are tagged with the local time zone):
prompt%qore -X '2005-03-29-18:12:25' 2005-03-29 18:12:25 Tue +02:00 (CEST)prompt%qore -X '0512-01-01T01:49:59.002213Z' 0512-01-01 01:49:59.002213 Fri Z (UTC)prompt%qore -X '2005-03-29' 2005-03-29 00:00:00 Tue +02:00 (CEST)prompt%qore -X '18:35:26+08:00' 1970-01-01 18:35:26 Thu +08:00 (+08)
The year must be a four-digit number, and all other values except microseconds must be two-digit numbers. If microseconds are present, at least one and up to 6 digits may be given after the decimal point. Pad the numbers with leading zeros if the numbers are smaller than the required number of digits. The hour component must be in 24-hour time format. Except for the month and day values, all other values start with 0 (hour = 00 - 23, minute and second: 00 - 59). Any deviation from this format will cause a parse exception.
When a date/time value is converted to an integer or vice-versa, a 64-bit offset in seconds from the start of the "epoch" is used for the conversion. Qore's "zero date" (the start of Qore's "epoch") is January 1, 1970 UTC. When calculating second offsets from this date, a 64-bit integer is used.
The default local time zone for qore is set when the qore library is initialized; see Time Zones for more information.
Relative dates (durations) are normally used for date addition and subtraction. See Date/Time Arithmetic for more information.
Internally, durations are stored as a set of seven discrete signed integer values, one each for years, months, days, hours, minutes, seconds, and microseconds.
There are 3 different formats understood by the Qore parser for describing literal durations in Qore.
A single relative date/time value (or a duration) may be specified as follows (note that this format is specific to Qore and not based on ISO-8601):
<integer><date component specifier>Table 2.7. Date Specifiers For Single Values For Relative Dates (non-ISO-8601 syntax)
Component | Meaning | Example | Description |
|---|---|---|---|
Y | Years |
| 2 Years |
M | Months |
| 3 Months |
D | Days |
| 10 Days |
h | Hours |
| 15 hours |
m | Minutes |
| 25 minutes |
s | Seconds |
| 19 seconds |
ms | Milliseconds |
| 250 milliseconds |
us | Microseconds |
| 21194 microseconds |
This and the next duration format for composite relative date/time values are both based on ISO-8601.
This first format has the following syntax:
PnYnMnDTnHnMnSnuEach element above is optional, but at least one element must be present. Note that "M" means months when before the "T" and minutes when found after the "T". The other elements are years, days, hours, seconds, and, as an extension to ISO-8601, "u" for microseconds. Additionally, the values may be negative.
Here are some examples (using qore's -X command-line option to evaluate and expression and print out the result):
prompt%qore -X 'P1Y3MT4S' <time: 1 year 3 months 4 seconds>prompt%qore -X 'PT4M551u' <time: 4 minutes 551 microseconds>prompt%qore -X 'P3DT21H' <time: 3 days 21 hours>
The second ISO-8601-based format for specifing complex durations with multiple time units has the following syntax:
PYYYY-MM-DDTHH:mm:SSThis format is more limited than the first format, in that all values must be positive, and furthermore, all values must be present (although they may be zero).
Here are some examples of the second format (equivalent to the first examples):
prompt%qore -X 'P0001-03-00T00:00:04' <time: 1 year 3 months 4 seconds>prompt%qore -X 'P0000-00-00T00:04:00.000551' <time: 4 minutes 551 microseconds>prompt%qore -X 'P0000-00-03T21:00:00' <time: 3 days 21 hours>
The binary data type is used to hold binary arbitrary binary data. Internally it is represented by a pointer to a memory location for the data and a size indicator.
Binary data can be concatenated with the + and += operators.
This data can be manipulated by being written and read from File, Socket, and Datasource objects, or converted and parsed to/from base64 encoded strings using the makeBase64String() and parseBase64String() functions, or compressed and decompressed using the compress(), gzip(), bzip2(), etc. functions, and processed by most cryptographic funtions, among others.
Binary objects can be read from a File object using the File::readBinary() method and can be written using the File::write() method. Please see the File Class for more information.
Binary objects can be read from a Socket object using the Socket::recvBinary() method and can be written using the Socket::send() method. Please see the Socket Class for more information.
The Datasource and DatasourcePool classes can also be used to read and write Binary objects as BLOBs.
Note that this is not an exhaustive list; see the function and class library documentation for more examples.
This data type represents an SQL NULL value. Note that NULL is not equivalent to NOTHING.
This special data type represents no value.
The exists operator will return False when given NOTHING as an argument; for example:
prompt% qore -X 'exists NOTHING'
FalseBoolean, string, integer, date, and floating point data types can be freely converted from one type to the other, although data loss is possible depending on the conversion (particularly when converting to the boolan type as only two possible values are supported).
The special types NULL and NOTHING are not equivalent and cannot be converted to or from any other type.
When date types are converted from strings, any of the following formats can be used: "YYYYMMDDHHmmSS[.us][Z|+-HH[:MM[:SS]]]", "YYYY-MM-DD HH:mm:SS.us", "YYYY-MM-DDTHH:mm:SS", "YYYY-MM-DDTHH:mm:SS[.us][Z|+-HH[:MM[:SS]]]", and most reasonable combinations thereof. If the time zone component is missing, then the local time zone will be assumed (see Time Zones).
When dates are converted to and from integer values, the a 64-bit second offset from January 1, 1970 in the local time zone is used for the conversion. For example int(2006-01-01) gives 1136073600 (regardless of the local time zone the date is in). To get the second offset of a date from 1970-01-01Z (i.e. the true epoch offset), call get_epoch_seconds() instead.
When an expression requires a certain data type and the source data type cannot be converted to the desired data type, the default value for the desired data type will be used. The default values are given here.
Qore supports three types of container types: lists, hashes (associative arrays), and objects (see Objects and Classes for more information). These container types can be combined to make arbitrarily complex data structures.
The data type of any element can be any basic type or another aggregate type. The types do not have to be uniform in one container structure.
Lists (or arrays) are simply ordered containers of values. A list element can be any Qore type (even another list, hash, or object).
Lists are specified by giving expressions separated by commas as follows:
$list = (expression,expression[,expression...]);
Here is a concrete example:
my list $list = (1, 2, "three", 4.0, 5, 6, 2001-01-15Z);Note that trailing commas can be left on the end of a list (or a hash, for that matter). This makes it easier to comment-out the last element of a multi-line list without having to worry about removing the trailing comma.
List elements are dereferenced using square brackets: "[" and "]". The first element in a list has index zero.
$element3 = $list[2];
The following operators perform special processing on lists: elements, shift, unshift, push, pop, splice, [], +, +=, map, foldl, foldr, and select.
Hashes are containers that associate values to a string key.
Note that Qore hashes preserve the insertion order in order to be able to guarantee the order of keys when hashes are serialized to XML strings (see XML Integration), therefore the keys operator will always return the hash keys in insertion/creation order.
Hashes are specified using the following syntax:
$hash = ( "key1" :expression, "key2" :expression, ... );
Here is a concrete example:
my hash $hash = ( "apple" : 1 + 1,
"pear" : "good" );Hashes are dereferenced in one of two ways, either using curly brackets: "{" and "}", where any valid Qore expression can be used, or using the dot "." hash member dereferencing operator, where literal strings can be used.
$element3 = $hash{"pe" + "ar"};Is equivalent to:
$element3 = $hash.pear;
and:
$element3 = $hash."pear";
and:
$element3 = $hash.("pe" + "ar");Hash members can have the names of keywords, but in this case to dereference them, you cannot use the dot operator with a literal string, otherwise a parse error will be raised. Use quotes around the member name when dereferencing hash members with the same name as a qore keyword as follows:
$element3 = $hash."keys";
$element3 = $hash{"keys"};A literal string after the dot "." hash member dereferencing operator must be a valid Qore identifier; therefore if you want to use a hash key that's not a valid identifier, enclose the string in quotes.
If you want to use the result of an expression to dereference the hash, then the curly bracket syntax must be used.
Note that hash keys can also be given by constants (as long as the constant resolves to a string) when using curly brackets.
Qore objects are instantiations of a class. They have members (like a hash - values associated to string keys), and methods. The class definition specifies the methods that run on objects of that class, private members, and static methods associated with the class (however note that static methods do not run in the scope of an object). Qore classes are declared with a special syntax.
The recommend way to instantiate an object is to declare its type and give constructor arguments after the variable name in parentheses as follows:
myclass_name_or_path $var_name([argument list])
For example (for a constructor taking no arguments or having only default values for the aguments, the list is empty):
my Mutex $m();Objects can also be instantiated using the new operator as follows.
newclass_identifier([argument list])
For example:
my $m = new Mutex();
Objects have named data members that are referenced like hash elements, although this behavior can be modified for objects using the memberGate() method. Object members are accessed by appending a dot '.' and the member name to the object reference as follows:
object_reference.member_nameFor more information, see Object Members.
Object methods are called by appending a dot '.' and a method name to the object reference as follows:
object_reference.method_name([argument_list])
Or, from within the class code itself to call another method from inside the same class hierarchy:
$.method_name([argument_list])
For more information, see Object Method Calls.
The object references above are normally variable references holding an object, but could be any expression that returns an object, such as a new expression or even a subroutine call.
Objects are treated differently than other Qore data types; they are only explicitly copied (see Object References for more informaion). Any object instantiated with the new operator will remain unique until deleted or explicitly copied. An explicit copy is made with the copy method, and does not always guarantee an exact copy of the source object (it depends on the definition of the copy method for the class in question).
Objects exist until they go out of scope, are explicitly deleted, or their last thread exits. For detailed information, see the section Classes on Qore classes.
Starting in Qore 0.8.0, it is possible to restrict variables, class members, and function and method parameters to certain data types. This allows programmers to write safer code, as many more errors can be caught at parse time that would otherwise be caught at run time. Furthermore, providing type information to the parser allows Qore to implement performance optimizations by performing lookups and resolutions once at parse time rather than every time a variable or class member is accessed at run time.
When types are declared in a parameter list, functions and methods can be overloaded as well.
The types in the following table can be used:
Table 2.8. Data Type Declaration Names
Name | Accepts Qore Type(s) | Returns Qore Type(s) | Description |
|---|---|---|---|
| Restricts values to Qore's Integer type. | ||
| Restricts values to Qore's Float type. | ||
| Restricts values to Qore's Boolean type. | ||
| Restricts values to Qore's String type. | ||
| Restricts values to Qore's Date type; values may be either absolute or relative date/time values. | ||
| Restricts values to Qore's Binary type. | ||
| Restricts values to Qore's Hash type. | ||
| Restricts values to Qore's List type. | ||
| Restricts values to Qore's Object type. | ||
| Restricts values to objects of the specific class given; either the class name can be given (ex: | ||
| Restricts values to Qore's NULL type; this type has few (if any) practical applications and has been included for completeness' sake. | ||
| Restricts values to Qore's NOTHING type; this type is mostly useful for declaring that a function or method returns no value. | ||
| Accepts Integer, Float, Boolean, String and converts non-integer values to an integer and returns the integer. | ||
| Accepts Integer, Float, Boolean, String and converts non-float values to a float and returns the new value. | ||
| Accepts Integer, Float, Boolean, String and converts non-boolean values to a boolean and returns the new value. | ||
| Accepts Integer, Float, Boolean, String and converts non-string values to a string and returns the new value. | ||
| any | same as received | Provides no restrictions on the type of value it receives and returns the same value. |
| same as received | Restricts input to String and Binary and returns the same type. | |
| same as received | Restricts values to closures and call references. | |
| reference to an lvalue | same as received | Restricts values to references to lvalues; currently only usable in function or method parameters. |
| same as received | Does not restrict value to just closures, but rather also allows call references. Synonym for | |
| same as received | Does not restrict value to just call references, but rather also allows closures. Synonym for |
Complex types (hash of lists, reference to string, etc) are currently not possible to declare.
Functions and methods can be overloaded if parameter types are declared as in the following example:
sub example(int $i) returns int { printf("i=%d\n", $i); return $i + 1; } sub example(string $str) returns string { printf("str=%s\n", $str); return $str + "foo"; }
In this case, the first version (example(int)) will be executed if called with an integer argument, and the second (example(string)) if called with a string argument.
Class methods may also be overloaded, but note that destructor(), copy(), methodGate(), memberGate(), and memberNotification() methods may not be overloaded (see Classes for more information).
Qore assumes a default time zone for all programs when it starts up. The rules for determining the default time zone are similar to those for the C library in most UNIX or UNIX-like operating systems.
If the TZ environment variable is defined, then the contents of that variable are used to find a zoneinfo file that contains the time zone definition. If this file cannot be found, then the default time zone will default to UTC.
If the TZ environment variable is not defined or is empty, then the Qore library tries to find the default zoneinfo definition file (normally /etc/localtime). If found, this file is read in and provides the information about the local time zone. If not found, the default time zone will default to UTC.
When a zoneinfo file is found, information about local time zone names and daylight savings time is available for times tagged with that time zone.
See the TimeZone class for information about retrieving, setting, and querying time zone information; see Date and Time Functions for a list of functions related to date/time processing.
Here are some examples using Qore's '-X' option for evaluating an expression and displaying the result immediately:
prompt%TZ=America/Chicago qore -X 'now_us()' 2010-05-11 06:14:28.845857 Tue -05:00 (CDT)prompt%TZ=Europe/Rome qore -X 'now_us()' 2010-05-11 13:14:35.070568 Tue +02:00 (CEST)prompt%TZ=Australia/Sydney qore -X 'now_us()' 2010-05-11 21:14:45.422222 Tue +10:00 (EST)prompt%TZ=Asia/Tokyo qore -X 'now_us()' 2010-05-11 20:14:59.609249 Tue +09:00 (CJT)
Note that posix-style time zone rules are not understood if assigned to the TZ environment variable, only file names to a zoneinfo file can be processed at the moment. Furthermore if the zoneinfo file contains leap second information, it is currently ignored.
The Qore language is character-encoding aware. All strings are assumed to have the default character encoding, unless the program explicitly specified another encoding for certain objects and operations. Every Qore string has a character encoding ID attached to it, so, when another encoding is required, the Qore language will attempt to do an encoding translation.
Qore uses the operating system's iconv library functions to perform any encoding conversions.
Qore supports character encodings that are backwards compatible with 7-bit ASCII. This includes all ISO-8859-* character encodings, UTF-8, KOIR-8, KOIU-8, and KOI7, among others (see the table below: Character Encodings Known to Qore).
However, mutibyte character encodings are currently only properly supported for UTF-8. For UTF-8 strings, the length(), index(), rindex(), substr(), reverse(), the splice operator, print formatting (regarding field lengths) functions and methods taking format strings, and regular expression operators and functions, all work with character offsets, which may be different than byte offsets. For all character encodings other than UTF-8, a 1 byte=1 character relationship is assumed.
Qore will accept any encoding name given to it, even if it is not a known encoding name or alias. In this case, Qore will tag the strings with this encoding, and pass this user-defined encoding name to the iconv library when encodings must be converted. This allows programmers to use encodings known by the system's iconv library, but unknown to qore. In this case, Qore will assume that the strings are backwards compatible with ASCII, meanin that that one character is represented by one byte and that the strings are null-terminated.
Note that when Qore matches an encoding name to a code or alias in the following table, the comparison is not case-sensitive.
Table 2.9. Character Encodings Known to Qore
Code | Aliases | Description |
|---|---|---|
| latin-1, Western European character set | |
| latin-2, Central European character set | |
| latin-3, Southern European character set | |
| latin-4, Northern European character set | |
| Cyrillic character set | |
| Arabic character set | |
| Greek character set | |
| Hebrew character set | |
| latin-5, Turkish character set | |
| latin-6, Nordic character set | |
| Thai character set | |
| latin-7, Baltic rim character set | |
| latin-8, Celtic character set | |
| latin-9, Western European with euro symbol | |
| latin-10, Southeast European character set | |
n/a | Russian: Kod Obmena Informatsiey, 7 bit characters | |
| Russian: Kod Obmena Informatsiey, 8 bit | |
| Ukrainian: Kod Obmena Informatsiey, 8 bit | |
| 7-bit ASCII character set | |
| variable-width universal character set |
The default character encoding for Qore is determined by environment variables.
First, the QORE_CHARSET environment variable is checked. If it is set, then this character encoding will be the default character encoding for the process. If not, then the LANG environment variable is checked. If a character encoding is specified in the LANG environment variable, then it will be used as the default character encoding. Otherwise, if no character encoding can be derived from the environment, UTF-8 is assumed.
Character encodings are automatically converted by the Qore language when necessary. Encoding conversion errors will cause a Qore exception to be thrown. The character encoding conversions supported by Qore depend on the Operating System's iconv library function.
The following is a non-exhaustive list of examples in Qore where character encoding processing is performed.
Character encodings can be explicitly performed with the convert_encoding() function, and the encoding attached to a string can be checked with the get_encoding() function. If you have a string with incorrect encoding and want to change the encoding tag of the string (without changing the actual bytes of the string), use the force_encoding() function.
The Datasource and DatasourcePool classes will translate character encodings to the encoding required by the database if necessary as well (this is actually the responsibility of the DBI driver for the database in question).
The File and Socket classes translate character encodings to the encoding specified for the object if necessary, as well as tagging strings received or read with the object's encoding.
The HTTPClient, XmlRpcClient, and JsonRpcClient classes will translate character encodings to the encoding specified for the object if necessary, as well as tag strings received with the object's encoding. Additionally, if an HTTP server response specifies a specific encoding to use, the encoding will be automatically set to this encoding as well.
Additionally you can serialize a Qore hash to an XML string and specify the character encoding to use. If the target encoding is different than the source encoding conversions are automatically performed. See XML Integration for more information about XML handling.
An expression can be any of the following (note that expressions are also recursively defined):
Table 2.10. Expressions
Type | Description |
Examples | |
|---|---|---|---|
An immediate value | Qore values that can be expressed directly (see Basic Data Types for more information) |
True 152 1.2 "a string" 2005-10-27 NULL NOTHING ("key" : $val) 1, 2.0, "three" | |
A variable reference | Qore variables (see Variables for more information) | $var | |
A variable declaration | Qore variable declarations (see Variables for more information) | my int $var | |
An in-class object member reference | References to members of an object from within the class (see Object Members for more information) | $.member | |
An lvalue assignment | Assigns a value to a lvalue (see Assignment Operator for more information) |
$var = 1 ($a, $b, $c, $date) = (1, "two", 3.3, 2005-10-28) | |
A subroutine/function call | Qore subroutine calls (see Subroutines for more information) | calculate($var1, $var2, "string", 4) | |
A method call | Qore object method calls (see Object_Method_Calls for more information) | $object.method("argument") | |
An in-class method call | Qore in-class object method calls (see Object_Method_Calls for more information) | $.method("argument") | |
Qore static method calls (see Static Methods for more information) | ClassName::static_method("argument") | ||
Expressions with operators | Use of Qore operators (see Operators for more information) |
1 + 2
$a || $b
background my_function()
| |
An expression in parentheses | Use of parentheses for clarity or to specify evaluation precedence. |
(1 + 2) (2 * (3 + 1)) | |
A list | a list of values (see Lists for more information) | 1, 2, 3, "four", 5.0 | |
A hash | a hash (associative/keyed array) value container (see Hashes for more information) | ( "key1" : 1, "key2" : "two" ) | |
Finds a value or values in a hash of lists, such as returned by the Datasource::select() method (see find expressions for more information) |
find %name, %id in $data where (%name =~ /Smith/); | ||
A context identifier ( | A contextual reference to the value of a key of a complex data structure within the current row being iterated by a context, summarize, subcontext statement, or a find expression. | %name | |
A context row identifier ( | A contextual reference to the current row of complex data structure being iterated by a context, summarize, subcontext statement, or a find expression. This expression will return a hash of the current row. | %% | |
A reference to a function or object method call (similar to a function pointer in C or C++). Function references are resolved in the second phase of parsing (commit phase), and object method references are resolved at run-time. | \function_call() \$object_expression.method_name() | ||
A closure | An anonymous function used a value; technically in computer science a closure must have at least one bound variable, but in qore a closure is any function used as a value, whether or not it encloses local variables from the scope in which it was created or not. | sub (string $a) returns string { return $a + $b; } | |
A call reference call | Calls the code referenced by the call reference expression or closure using any arguments supplied and returns the result. | $result = $call_reference($arg1, $arg2) | |
References an implicit argument. | $1, $2 $$ |
Calls to static class methods are made by giving the class name followed by two colons and then the method name. The method name must be implemented and accessible (i.e. not private and accessed outside the class) somewhere within the class hierarchy and must be static or a parse exception will occur.
class_name::method_name([argument_expressions...])
class_name
The name of the class implementing the static method.
method_name
The name of the static method to call.
[argument_expressions...]
Expressions passing arguments to the static method.
TimeZone::setRegion("Europe/Prague");The find expression can be used to quickly find data in a hash of lists (such as a query result set returned by the Datasource::select() method). The find expression will loop through a data structure, and for each element in the structure where the where expression is True, it will evaluate and return a result expression. If the where expression only is true for one element in the list, it will return the result of evaluating the result expression directly, otherwise if the where expression is true more than once, then a list of the results of evaluting the result expression for each element is returned. In each expression in the find expression, column names can be referred to by preceding the name with a '%" character (as with context statements).
findresult_expressionindata_expressionwhere (where_expression)
result_expression
This expression will be evaluated and returned when the where_expression evaluates to True.
data_expression
This expression must evaluate to a hash of lists, so that the internal context can be set up for the find loop.
where_expression
This expression will be evaluated for each row in the data_expression. Each time it evaluates to True, the result_expression will be evaulated and used for the return value for the find expression.
$rlist = find %name, %id in $data where (%name =~ /Smith/);
References to functions or object methods are called call references. A call reference can be used like a function pointer; a call reference is a Qore data type that can be returned by functions or methods or assigned to variables.
Note that it is currently not legal to assign a call reference to a constant. This restriction may be lifted in a future version of Qore.
Function Call References
Call references to functions are resolved at parse time; if the function does not exist a parse exception will be thrown.
Object Method Call References
Call references to object methods are executed and resolved at run time; if the object expression does not evaluate to an object at run-time, an OBJECT-METHOD-REFERENCE-ERROR exception will be thrown. If the method does not exist, a METHOD-DOES-NOT-EXIST run-time exception will be thrown.
When called, a call reference to an object method will be executed in the context of the object originally referenced.
Object method call references do not prolong the lifetime of an object; if the object is deleted (for example, by going out of scope), then if called the call reference will cause a OBJECT-ALREADY-DELETED exception to be thrown.
\function_name(); \object_expression.method_name();
\function_name()
This form gives a function call references. The function name can be any valid user or builtin function name. Note the backslash at the beginning and the empty pair of parentheses at the end; these are required when specifying a call reference.
\object_expression.method_name()
This form gives an object method call reference. The object expression can be any valid Qore expression that evaluates to an object. The method_name must be an unquoted string (see examples below) and must represent a valid method name of the object's class.
$call_ref = \func_name(); $call_ref = \$object.method_name();
A closure is an anonymous function used as a value. Closures can be returned from functions or methods, assigned to variables, or passed as arguments to other functions.
Note that it is not legal to assign a closure to a constant.
sub ([[type] variable1, ...]) [returnstype] {code...}
Closures encapsulate the state and value of local variables of the outer code block referenced from within the closure when the closure is created. Whenever local variables are bound within a closure, these variables are subject to concurrent thread access protection (locking) just as with global variables, in order to allow closures to be used in any context without restriction and to preseve thread-safety regarding bound local variables.
Note that returning a closure from within an object method encapsulates the state of the object as well (it's legal to refer to $self and $.<variable> from within closures created from objects) and additionally prolongs the scope of the object for the lifetime of the closure.
Note that parameter and return types are required when the PO_REQUIRE_TYPES or PO_REQUIRE_PROTOTYPES parse options are set.
# if $b is a local variable in the function where the closure is created # then $b will be bound to the closure when the closure is created my code $closure = sub (int $a) returns int { return $a + $b; };
Implicit arguments are arguments not captured by parameter variables as well as automatic arguments in list-processing operator expressions. A special syntax to reference these arguments is documented here.
$<integer> # for a single implicit argument
$$ # for the entire implicit argument list
Implicit arguments can be directly referenced using the dollar sign ($) and either a number from 1 onwards (giving the position in the argument list, where 1 is the first element) or a double dollar sign ($$) giving the entire implicit argument list.
For unassigned arguments to functions or methods, this syntax supplements the automatic $argv variable holding all function arguments not assigned to parameter variables.
This syntax is particularly useful when writing expressions for the map, map, foldr, and select operators, where implicit argument references are the only way the operator expressions can reference the current list values that are populated as implicit arguments as the operators traverse the list.
# extract a list of even numbers from a list my $l = select $list, !($1 % 2);
The following table lists all Qore operators in order of precedence, starting with the highest precedence. The lower the precedence number, the higher the precedence, therefore the operators with precedence level 1 ("{}", "[]", ".") have the highest precedence of all Qore operators. The precedence levels in Qore are roughly equal to the precedence levels of C language operators. To explicitly specify the precedence for expression evaluation, use parentheses ().
Table 2.11. Operators
Operator | Prec. | Description | Example |
|---|---|---|---|
`` | 1 |
| |
{} | 1 | hash element or object member expression dereference operator |
|
. | 1 |
| |
[] | 1 |
| |
++ | 2 |
| |
-- | 2 |
| |
new | 3 |
| |
background | 3 |
| |
delete | 3 |
| |
remove | 3 |
| |
cast<>() | 3 |
| |
! | 4 |
| |
~ | 5 |
| |
- (unary minus) | 6 |
| |
shift | 7 |
| |
pop | 7 |
| |
chomp | 7 |
| |
trim | 7 |
| |
elements | 8 |
| |
keys | 8 |
| |
* | 9 |
| |
/ | 9 |
| |
% | 10 |
| |
+ | 11 | plus operator: string, binary, list, and hash concatenation, integer and float addition |
|
- | 11 |
| |
>> | 12 |
| |
<< | 12 |
| |
exists | 13 |
| |
instanceof | 13 |
| |
< | 14 |
| |
> | 14 |
| |
== | 14 |
| |
!= | 14 |
| |
<= | 14 |
| |
>= | 14 |
| |
<=> | 14 |
| |
=== | 14 |
| |
!== | 14 |
| |
=~ // | 14 |
| |
!~ // | 14 |
| |
=~ s/// | 14 |
| |
=~ x// | 14 |
| |
=~ tr | 14 |
| |
& | 15 |
| |
| | 15 |
| |
^ | 15 |
| |
&& | 16 |
| |
|| | 16 |
| |
? : | 17 |
| |
, | 18 |
| |
unshift | 19 |
| |
push | 19 |
| |
splice | 19 |
| |
extract | 19 |
| |
map | 19 |
| |
foldl | 19 |
| |
foldr | 19 |
| |
select | 19 |
| |
= | 20 |
| |
+= | 21 |
| |
-= | 21 |
| |
&= | 21 |
| |
|= | 21 |
| |
%= | 21 |
| |
*= | 21 |
| |
/= | 21 |
| |
^= | 21 |
| |
<<= | 21 |
| |
>>= | 21 |
|
All Qore operators perform thread-atomic actions with respect to the immediate arguments of the operator. If the operators are used in a complex expression, the entire expression is not thread-atomic unless explicit user-level locking is used. For example: $a += 5 is a thread-atomic action, but $a += $b-- is not atomic, but rather made up of two atomic actions.
When an operator taking more than one argument is used with arguments of different data types, Qore automatically converts one or both data types to a data type supported by the operator in order to evaluate the result, according to the precedence lists in the following section. That is; when an operator operates on mixed types, the types listed first in the following sections have precedence over types listed farther down in the lists. The result type will always be equal to the final operation type after any conversions due to type precedence per operator. If no type of either argument matches a supported data type for the operator, both types will be converted to the highest precedence data type for the operator and then the operator will evaluate the result. For explicit type conversion, please see the boolean(), string(), date(), int(), float(), etc functions.
Executes the shell command in a separate process and returns the stdout as a string. To perform the same action using a Qore expression, see the backquote() function.
`shell_command`$dirlisting = `ls -l`
Table 2.13. Exceptions Thrown by ``
err | desc |
|---|---|
| An error occurred in fork() or creating the output pipe. |
Retrieves the value of hash key or object member by evaulating an expression.
container_expression{expression}printf("%s\n", $hash{getName()});Table 2.14. Arguments Processed by {}
Argument | Processing |
|---|---|
This expression must evaluate to a hash or an object. If not, then the operator returns no value. | |
list | If the expression evaluates to a list, then a slice of the hash or object is returned as a hash containing keys given in the list that are present in the hash or object. If the key as given in the list (converted to a string if necessary) is not present in the hash or object, then it is also not present in the hash returned. |
| This expression is evaluated and converted to a string if necessary. The value of the hash key corresponding to this string will be returned. If the key or member does not exist, then no value is returned. |
Table 2.15. Exceptions Thrown by {}
err | desc |
|---|---|
| Attempt to access a private member outside the class. |
Retrieves the value of a hash key or object member using a literal identifier or an expression.
container_expression.identifiercontainer_expression.method_identifier([arguments...])container_expression.expression
printf("%s\n", $hash.name);$obj.method("argument");Table 2.16. Arguments Processed by .
Argument | Processing |
|---|---|
| This expression must evaluate to a hash or an object. If not, then the operator returns no value. |
| The value of the hash key or object member corresponding to this identifier will be returned. If no such key exists, then no value is returned. In order to use hash keys that are not valid Qore identifiers, please use the {} operator. If the member is a private member and access is made outside the class, a run-time exception will be thrown. |
| The container expression must evaluate to an object, or a run-time exception is thrown. If the method does not exist in the class a run-time exception is thrown. Otherwise the method is called with any optional arguments given. |
| If the expression evaluates to a list, then a slice of the hash or object is returned as a hash containing keys given in the list that are present in the hash or object. If the key as given in the list (converted to a string if necessary) is not present in the hash or object, then it is also not present in the hash returned. |
| This expression is evaluated and converted to a string if necessary. The value of the hash key corresponding to this string will be returned. If the key or member does not exist, then no value is returned. |
Table 2.17. Exceptions Thrown by .
err | desc |
|---|---|
| Attempt to access a private member outside the class. |
| Attempt to access a method not defined for this class. |
| Attempt to access a private method from outside the class. |
| Attempt to access a method of a privately-inherited base class from outside the class. |
| Attempt to execute a method on a non-object. |
Retrieves the value of a list element, the given character of a string, or the integer value of a byte for a binary object. If the index value is not valid for the argument, NOTHING is returned. Note that this operator only works as a list dereferencing operator in lvalue expressions; you cannot assign a character or a byte value to strings or binaries using this operator.
list_expression[expression]string_expression[expression]binary_expression[expression]
printf("%s\n", $list[2]);
printf("%s\n", $str[2]);
printf("0x%x\n", $binary[2]);
Table 2.18. Arguments Processed By []
Argument | Processing |
|---|---|
| If the expression evaluates to a list, then the offset_expression will be used to return the given element from the list. |
| If the expression evaluates to a string, then the offset_expression will be used to return the given character from the list; note that multi-byte characters with UTF-8 are properly respected with this operator. |
| If the expression evaluates to a binary, then the offset_expression will be used to return the integer value of the byte given from the binary object. |
| The expression is evaluated and converted to an integer if necessary. Then the value of the list element given is returned (elements start at position 0). |
This operator does not throw any exceptions; if the first expression does not evaluate to either a list, string, or binary, then no value is returned.
increments an lvalue and returns the incremented value.
++lvalue++$i;
Table 2.19. Arguments Processed By Pre-Increment ++
Argument | Processing |
|---|---|
Integer | First converts the value of |
This operator does not throw any exceptions.
increments an lvalue and returns the value before the increment.
lvalue++$i++;
Table 2.20. Arguments Processed By Post-Increment ++
Argument | Processing |
|---|---|
Integer | First converts the value of |
This operator does not throw any exceptions.
decrements an lvalue and returns the decremented value.
--lvalue--$i;
Table 2.21. Arguments Processed By Pre-Decrement --
Argument | Processing |
|---|---|
Integer | First converts the value of |
This operator does not throw any exceptions.
decrements an lvalue and returns the value before the decrement.
lvalue--$i--;
Table 2.22. Arguments Processed By Post-Decrement --
Argument | Processing |
|---|---|
Integer | First converts the value of |
This operator does not throw any exceptions.
Creates an instance of a class by running the class' constructor on the new class (if any exists) and returns the new object.
Note that it is normally better to declare an object with its type and use the abbreviated form to construct the object as follows:
my Mutex $m();This provides type information to the parser which allows more errors to be caught at parse time (instead of at run time), and furthermore allows Qore improve performance by performing more work once at parse time rather than for every time the object is accessed at run time (for example, method and variant resolution).
new class_identifier(constructor_arguments ...)specific class, an object of the class given
$obj = new Qore::Mutex();Table 2.23. Arguments Processed By new
Argument | Processing |
|---|---|
| The class_identifier must be an existing class name; if so, the operator instantiates an object of this class, executes the constructor for the class (if any exists, along with any base class constructors, if applicable) on the new object, and returns the object (for constructor execution order in an inherited class, see Class Inheritance). If an exception is thrown in the constructor, the object is deleted immediately. |
Table 2.24. Exceptions Thrown by new
err | desc |
|---|---|
depends on class/constructor | See class documentation for possible exceptions. |
Start a background thread and return the TID (thread ID).
background expressionbackground startThread();Table 2.25. Arguments Processed By background
Argument | Processing |
|---|---|
string |
The expression given as an argument will be executed in a new thread. The TID of the new thread will be returned as the return value of the operator. |
Please note the following when using the background operator:
expressions that have no effect cannot be used with the background operator (a parse exception will be thrown)
it is illegal to make changes to a local variable anywhere in a background expression (a parse exception will be thrown)
local variables and find expressions are evaluated before the new thread is started and the result of the evaluation is used in the expression in the new thread.
it is not possible to pass local variables by reference anywhere in a background expression (a parse exception will be thrown)
Table 2.26. Exceptions Thrown by background
err | desc |
|---|---|
| If the thread table is full or if the operating system returns an error while starting the thread, this exception is thrown. |
The delete operator deletes the contents of an lvalue. If the delete operator is called on an object, the object will be destroyed unconditionally. The delete operator does not return any value. When called on a hash key, the key is removed from the hash entirely; when called on a list, the element is assigned NOTHING.
In the case the delete operator operates on an object, any exception can be thrown that is thrown by the class' destructor.
For a similar operator that returns the value that is removed from the data structure, and does not delete objects, see the remove operator.
delete lvalue_expressionDoes not return any value
delete $value;This operator does not throw any exceptions, however exceptions could be thrown in an object's destructor method when deleted by this operator.
The remove operator removes a value from a data structure, or, in the case the operand of the remove operator is a simple value, the value itself is removed from the variable and returned. The remove operator returns the value removed from the lvalue. When called on a hash key, the key is removed from the hash entirely; when called on a list, the element is assigned NOTHING.
The remove operator does not call destructors when operating on objects, but if removing an object from an lvalue or from a data structure within the lvalue causes the object to go out of scope, it will be destroyed, and then its destructor could throw an exception.
For a similar operator that deletes the value that is removed from the data structure, see the delete operator.
remove lvalue_expressionmy $var = remove $hash.value;
This operator does not throw any exceptions, however exception could be thrown in an object's destructor if it goes out of scope due to the action of this operator.
The cast<>() operator provides a way to tell the parser that the type of object is not actually the declared type but rather a subclass as given between the angle brackets.
cast<class_name_or_path>(expression)
specific class, the class given
cast<SubClass>($obj).method();Table 2.27. Arguments Processed By cast<>()
Argument | Processing |
|---|---|
| The object is treated as if it were the class given between the angle brackets; this is mostly useful at parse time to avoid |
Table 2.28. Exceptions Thrown by cast<>()
err | desc |
|---|---|
| The expression given does not evaluate to an object that can be cast to the given class. |
Reverses the logical sense of an expression (True becomes False and False becomes True).
!expressionif (!exists $error_code) do_something();
Table 2.29. Arguments Processed By !
Argument | Processing |
|---|---|
| The expression is evaluated and converted to Boolean, if necessary. Then the value is logically reversed ( |
This operator does not throw any exceptions.
The value of each bit in an integer is reversed (0 becomes 1, 1 becomes 0).
~expression$a = ~$b;
Table 2.30. Arguments Processed By ~
Argument | Processing |
|---|---|
| The argument is converted to an integer (if necessary), and bitwise negation is performed on the argument (ex: 666 & ~27 = 640) |
This operator does not throw any exceptions.
Removes the first element from a list and returns that element.
shift lvalue$a = shift $ARGV;Table 2.32. Arguments Processed By shift
Argument | Processing |
|---|---|
list | Returns the first element of the list, and the list is modified by having the first element removed from the list. |
This operator does not throw any exceptions.
Removes the last element from a list and returns that element.
pop lvalue$a = pop $list;Table 2.33. Arguments Processed By pop
Argument | Processing |
|---|---|
list | Returns the last element of the list, and the list is modified, having the last element removed from the list. |
This operator does not throw any exceptions.
Removes the end-of-line marker(s) ('\n' or '\r\n') from a string, or each string element in a list, or each hash key value in a hash (if the value is a string) and returns the number of characters removed.
To perform this operation on a non-lvalue expression, see the chomp() function.
chomp lvaluechomp $str;Table 2.34. Arguments Processed By chomp
Argument | Processing |
|---|---|
string | Removes any EOL characters from a string and returns the number of characters removed. |
list | Removes any EOL characters from each string element of the list passed and returns the number of characters removed. |
hash | Removes any EOL characters from each hash key's value (where the value is a string) and returns the number of characters removed. |
This operator does not throw any exceptions.
Removes whitespace characters from the beginning and end of a string, or each string element in a list, or each hash key value in a hash (if the value is a string) and returns the value processed (string, list, or hash).
To perform this operation on a non-lvalue expression, see the trim() function.
The following whitespace characters are removed from the beginning and end of strings: ' ', '\n', '\r', '\t', '\v' (vertical tab, ASCII 11), and '\0' (null character).
trim lvaluetrim $str;Table 2.35. Arguments Processed By trim
Argument | Processing |
|---|---|
string | Removes whitespace characters from the beginning and end of a string and returns the value processed. |
list | Removes whitespace characters from the beginning and end of each string element of the list passed and returns the list. |
hash | Removes whitespace characters from the beginning and end of each string value of the hash passed and returns the hash. |
This operator does not throw any exceptions.
Executes (or maps) an expression on a list and returns the result. An optional select expression can be given to filter elements out from the result list.
If the second argument is not a list, then map_expression is executed on the single value and the result is returned, and any select expression is ignored.
map map_expression, list, [select_expression]# returns (2, 4, 6)
map $1 * 2, (1, 2, 3);Table 2.36. Arguments Processed By map
Argument | Processing |
|---|---|
| The expression to map on the list; the implicit argument $1 represents the current element being processed. |
list | The list to process. |
| An optional expression than can be used to filter out elements of the list before the map expression is applied; if this expression evaluates to |
This operator does not throw any exceptions (however note that exceptions could be thrown by expressions executed by this operator).
Folds an operation on a list from left to right and returns the result. The result of each individual operation is used as the first argument in the foldl expression for the next element in the list. The first operation of the fold is made by executing the fold expression on the first and second elements of the list, from this point onwards, the result of each successive operation is used as the first argument for the next operation, the second argument being the next element in the list.
If the list expression does not evaluate to a list, then the evaluated argument is returned immediately with no processing by the fold expression.
foldl expression, list# returns 5
foldl $1 - $2, (10, 4, 1);Table 2.37. Arguments Processed By foldl
Argument | Processing |
|---|---|
| The expression to fold on the list; the implicit argument $1 represents the result of the last operation (or the first element in the list when beginning the fold), and $2 represents the next element of the list. |
list | The list to process. |
This operator does not throw any exceptions (however note that exceptions could be thrown by expressions executed by this operator).
Folds an operation on a list from right to left and returns the result. The result of each individual operation is used as the first argument in the foldr expression for the next element in the list in reverse order. The first operation of the right fold is made by executing the fold expression on the last and penultimate elements of the list, from this point onwards, the result of each successive operation is used as the first argument for the next operation, the second argument being the next element in the list in reverse order.
If the list expression does not evaluate to a list, then the evaluated argument is returned immediately with no processing by the fold expression.
foldr expression, list# returns -13
foldr $1 - $2, (10, 4, 1);Table 2.38. Arguments Processed By foldr
Argument | Processing |
|---|---|
| The expression to fold on the list; the implicit argument $1 represents the result of the last operation (or the last element in the list when beginning the fold), and $2 represents the next element of the list in reverse order. |
list | The list to process. |
This operator does not throw any exceptions (however note that exceptions could be thrown by expressions executed by this operator).
Selects elements from a list that meet the given criteria and returns the new list.
If the list expression does not evaluate to a list, then the select expression is evaluated using the value of the list expression as an argument, if it evalutes to true, then the value is returned, otherwise, no value is returned.
select list, expression# returns (2, 4, 6)
select (1, 2, 3, 4, 5, 6), !($1 % 2);Table 2.39. Arguments Processed By select
Argument | Processing |
|---|---|
list | The list to process. |
| The expression will be evaluated on each element of the list, the implicit argument $1 represents current element of the list; only if the expression evaluates to |
This operator does not throw any exceptions (however note that exceptions could be thrown by the expression executed by this operator).
Returns the number of elements in a list, the number of keys in a hash, the number of characters (not bytes) in a string, or the number of bytes in a binary object.
elements expression$size = elements $list;This operator does not throw any exceptions.
Table 2.41. Arguments Processed By keys
Argument | Processing |
|---|---|
| Returns a list of strings giving the keys in |
This operator does not throw any exceptions.
This operator does not throw any exceptions.
This operator does not throw any exceptions.
Gives the integer remainder after division of one number by another.
expression1%expression2
$mod = $x % $y;
Table 2.44. Arguments Processed By %
Argument | Processing |
|---|---|
Gives |
This operator does not throw any exceptions.
Numeric addition, list, string, binary, and hash concatenation operator.
expression1+expression2
$a = 1 + 2;
$string = "hello" + "-there";
$list = (1, 2) + ("three", "four", "five");$hash = ( "key1" : 1, "key2" : 2) + ( "key3" : "three", "key4": "four");
$bin = $bin1 + $bin2;
Table 2.45. Argument Processing and Conversion Priorities for +
Argument | Processing |
|---|---|
Gives the result of concatenating its arguments, i.e. (1, 2) + (3, 4) = (1, 2, 3, 4) | |
Gives the result of concatenating its arguments. | |
Gives the result of adding date/time values (see Date/Time Arithmetic) | |
Gives the result of adding its arguments. | |
Gives the result of adding its arguments. | |
Gives the result of concatenating/merging its arguments. Any common keys will be overwritten by the values in the second hash ( |
This operator does not throw any exceptions.
With float or integer arguments, subtracts one number from another.
With date arguments, subtracts one date from another; if both date arguments are absolute dates, the result is a relative date (duration) giving the time between them; if the first date argument is an absolute date and the second is a relative date (duration), then the result is an absolute date. If both date arguments are relative dates, then the result is a relative date. If the first argument is a relative date and the second date is an absolute date, the result is an absolute date as if the operands were reversed.
However, if the left-hand side is a hash, and the right-hand side is a string, then the hash key represented by the string will be removed from the hash. If the left-hand side is a hash and the right-hand side is a list, then each element in the list will be converted to a string and any hash key with that name will be deleted from the hash.
expression1-expression2
$num = $x - $y;
$date = 2010-05-13 - P3MT14H10M;
$hash = $hash - "key";
$hash = $hash - ("key1", "key2", "key3");Table 2.46. Argument Processing and Conversion Priorities for -
Argument | Processing |
|---|---|
date subtraction: | |
arithmetic subtraction: | |
arithmetic subtraction: | |
hash key deletion: | |
hash key deletion: |
This operator does not throw any exceptions.
Shifts bits in an integer towards zero (divides an integer by a power of 2)
expression1>>expression2
$a = $x >> $y;
Table 2.47. Arguments Processed By >>
Argument | Processing |
|---|---|
Gives the result of shifting |
This operator does not throw any exceptions.
Shifts bits in an integer towards infinity (multiplies an integer by a power of 2)
expression1<<expression2
$a = $x << $y;
Table 2.48. Arguments Processed By <<
Argument | Processing |
|---|---|
Gives the result of shifting |
This operator does not throw any exceptions.
Tests if an expression is an instance of a class or not.
expressioninstanceofclass_specification
if ($obj instanceof Qore::Mutex) print("object is Mutex\n");
Table 2.49. Arguments Processed By instanceof
Argument | Processing |
|---|---|
| If |
This operator does not throw any exceptions.
Tests if an expression represents a value or not.
exists expressionif (exists $a) printf("a = $n\n", $a);
Table 2.50. Arguments Processed By exists
Argument | Processing |
|---|---|
| If |
This operator does not throw any exceptions.
Tests if a value is less than another; types are converted if necessary (ex: ("1" < 2) is True).
expression1<expression2
if ($x < $y)
printf("%n is less than %n\n", $x, $y);Table 2.51. Argument Processing and Conversion Priorities for <
Argument | Processing |
|---|---|
If | |
If | |
If | |
If |
This operator does not throw any exceptions.
Tests if a value is greater than another; types are converted if necessary (ex: ("2" > 1) is True).
expression1>expression2
if ($x > $y)
printf("%n is less than %n\n", $x, $y);expression1>expression2
Table 2.52. Argument Processing and Conversion Priorities for >
Argument | Processing |
|---|---|
If | |
If | |
If | |
If |
This operator does not throw any exceptions.
Tests if a value is equal to another; types are converted if necessary (ex: ("1" == 1) is True). For absolute equals, where types must also be equal to return true, see the Absolute Equals Operator (===).
expression1==expression2
if ($x == $y)
printf("%n is equal to %n\n", $x, $y);Table 2.53. Argument Processing and Conversion Priorities for ==
Argument | Processing |
|---|---|
If | |
If | |
If | |
If | |
If each element in the each list where order is relevant satisfies this operator, the operator returns | |
If each hash has the same keys and the value of each equal key in each hash satisfies this operator, the operator returns | |
If | |
If | |
If both expressions are NULL, returns | |
If neither expression has a value, returns |
This operator does not throw any exceptions.
Tests if a value is not equal to another; types are converted if necessary (ex: ("1" != 1) is False).
expression1!=expression2
if ($x != $y)
printf("%n is not equal to %n\n", $x, $y);Table 2.54. Argument Processing and Conversion Priorities for !=
Argument | Processing |
|---|---|
If | |
If | |
If | |
If | |
If any element in the each list compared where order is relevant satisfies this operator, the operator returns | |
If the hashes have different key sets, or the values of any equal key in each hash satisfies this operator, the operator returns | |
If either | |
If | |
If one expression is NULL and the other not, returns | |
If one of the expressions has a value, returns |
This operator does not throw any exceptions.
Tests if a value is less than or equals to another value; types are converted if necessary (ex: ("1" <= 2) is True).
expression1<=expression2
if ($x <= $y)
printf("%n is less than or equal to %n\n", $x, $y);Table 2.55. Argument Processing and Conversion Priorities for <=
Argument | Processing |
|---|---|
If | |
If | |
If | |
If |
This operator does not throw any exceptions.
Tests if a value is greater than or equals to another value; types are converted if necessary (ex: ("2" >= 1) is True).
expression1>=expression2
if ($x >= $y)
printf("%n is greater than or equal to %n\n", $x, $y);Table 2.56. Argument Processing and Conversion Priorities for >=
Argument | Processing |
|---|---|
If | |
If | |
If | |
If |
This operator does not throw any exceptions.
Tests if the left-hand value is less than, equal, or greater than the right-hand value; types are converted if necessary (ex: ("1" <=> 2) returns -1).
expression1<=>expression2
switch ($x <=> $y) { case -1: print("$x is less than $y\n"); break; case 0: print("$x is equal to $y\n"); break; case 1: print("$x is greater than $y\n"); break; }
Table 2.57. Argument Processing and Conversion Priorities for <=>
Argument | Processing |
|---|---|
If | |
If | |
If | |
If |
This operator does not throw any exceptions.
Checks two values for equality without doing any data type conversions; if the types do not match, then the result is False.
expression1===expression2
if ($x === $y)
printf("%n is equal to %n and has the same data type as well\n", $x, $y);Table 2.58. Arguments Processed By ===
Argument | Processing |
|---|---|
All | This operator returns |
This operator does not throw any exceptions.
Checks two values for inequality without doing any data type conversions. If the data types do not match, then returns True.
expression1 !== expression2
if ($x !== $y)
printf("%n is not equal to %n and may not have the data type as well\n", $x, $y);Table 2.59. Arguments Processed By !==
Argument | Processing |
|---|---|
All | This operator returns |
This operator does not throw any exceptions.
Checks for a regular expression match; returns True if the expression matches the string, False if not. See Regular Expression Options for the meaning of the i, s, x, and m options after the regular expression.
See Regular Expressions for more information about regular expression support in Qore.
expression=~ [m]/regex/[isxm]
if ($str =~ /hello/)
printf("%s contains 'hello'\n", $str);Table 2.60. Arguments Processed By =~
Argument | Processing |
|---|---|
This operator returns |
This operator does not throw any exceptions.
Checks for a regular expression non match; returns True if the expression does not match the string, False if it does. See Regular Expression Options for the meaning of the i, s, x, and m options after the regular expression.
See Regular Expressions for more information about regular expression support in Qore.
expression!~ [m]/regex/[isxm]
if ($str !~ /hello/)
printf("%s does not contain 'hello'\n", $str);Table 2.61. Arguments Processed By !~
Argument | Processing |
|---|---|
This operator returns |
This operator does not throw any exceptions.
Looks for a regular expression match in a string, and, if found, substitutes the matched string with a new string. Subpattern backreferences are supported in the target string, $1=first subpattern, $2=second subpattern, etc... See Regular Expression Options for the meaning of the i, s, x, and m options after the regular expression.
See Regular Expressions for more information about regular expression support in Qore.
lvalue=~ s/regex_pattern/target_string/[isxmg]
$str =~ s/hello/goodbye/i;
$str =~ s/(\w+) +(\w+)/$2, $1/;
Table 2.62. Arguments Processed By =~ s///
Argument | Processing |
|---|---|
This operator substitutes text in the |
This operator does not throw any exceptions.
Matches regular expression patterns (enclosed in parentheses) in a string and returns a list giving the text matched for each pattern. If the regular expression does not match, then no value (NOTHING) is returned. See Regular Expression Options for the meaning of the i, s, x, and m options after the regular expression.
See Regular Expressions for more information about regular expression support in Qore.
string=~ x/regex_with_patterns/[isxm]
$list =~ x/(\w+):(\w+)/;
$list =~ x/(.*)\.(.*)/;
Table 2.63. Arguments Processed By =~ x//
Argument | Processing |
|---|---|
This operator extracts strings from the |
This operator does not throw any exceptions.
Table 2.64. Arguments Processed By =~ tr//
Argument | Processing |
|---|---|
This operator substitutes characters in the |
This operator does not throw any exceptions.
Performs a bitwise (binary) AND operation on two integers.
expression1&expression2
$a = $x & $y;
Table 2.65. Arguments Processed By &
Argument | Processing |
|---|---|
Gives the result of the binary (bitwise) AND operation between |
This operator does not throw any exceptions.
Performs a bitwise (binary) OR operation on two integers.
expression1|expression2
$a = $x | $y;
Table 2.66. Arguments Processed By |
Argument | Processing |
|---|---|
Gives the result of the binary (bitwise) OR operation between |
This operator does not throw any exceptions.
Performs a bitwise (binary) XOR operation on two integers.
expression1^expression2
$a = $x ^ $y;
Table 2.67. Arguments Processed By ^
Argument | Processing |
|---|---|
Gives the result of the binary (bitwise) EXCLUSIVE OR operation between |
This operator does not throw any exceptions.
Checks to see if two expressions are True with logical short-circuiting.
expression1&&expression2
if ($x && $y)
printf("%n and %n are both True\n", $x, $y);Table 2.68. Arguments Processed By &&
Argument | Processing |
|---|---|
Returns |
This operator does not throw any exceptions.
Returns True if either of the arguments are True with logical short-circuiting.
expression1||expression2
if ($x || $y)
printf("either %n or %n or both are True\n", $x, $y);Table 2.69. Arguments Processed By ||
Argument | Processing |
|---|---|
Returns |
This operator does not throw any exceptions.
Evaluates and returns the value of one of two expressions depending on the value of a conditional expression.
expression ? if_true_expression : if_false_expression$a = ($z > 100 ? "Big" : "Small");
Table 2.70. Arguments Processed By ? :
Argument | Processing |
|---|---|
All | If |
This operator does not throw any exceptions.
Makes a list from more than one element.
expression1, expression2$a = 1, 2, "three";
Table 2.71. Arguments Processed By ,
Argument | Processing |
|---|---|
All | The comma operator builds lists of arguments. |
This operator does not throw any exceptions.
Inserts an element into the first position of a list and moves all other elements up one position.
unshift lvalue, expressionunshift $list, "one";Table 2.72. Arguments Processed By unshift
Argument | Processing |
|---|---|
All | Inserts the value of |
Table 2.73. Arguments Processed By push
Argument | Processing |
|---|---|
All | Appends the value of the expression as the last element in the list given by |
Removes and optionally inserts elements in lists and strings. For a similar operator that returns the values removed, see the extract operator.
splice lvalue, offset_expression, [length_expression, [substitution_expression]]splice $list, 2, 2;splice $string, 2, 2, "-text-";Works on either strings or lists in a similar way; removes elements from a list or characters from a string and optionally inserts new ones. If no length_expression is given, splice removes all elements/characters from the list or string starting at offset_expression) (list and string offsets begin at 0). Otherwise, a number of elements/characters equal to length_expression is removed (or up to the end of the list/string if applicable). If substitution_expression is present, then the removed elements/characters are substituted with the elements/string given by this expression.
Note that string splice takes character offsets, which may not be the same as byte offsets for multi-byte character encodings, such as UTF-8
Table 2.74. Arguments Processed By splice
Argument | Processing |
|---|---|
If the lvalue is a list, list elements are processed, otherwise, if it is a string, characters in the string are processed. For any other data type, no action is taken. | |
| The start element/character position for removing elements/characters from the list or string. |
| The number of elements/characters to remove. If this expression is not present, then all elements/characters from the offset to the end of the list/string are removed. If this expression is present and evaluates to 0, no characters/elements are removed. |
| For list splice, an optional element or list to substitute for the removed elements (to insert a list in a single element's position, make sure that the list to be inserted is the first and only element of another list used as the argument in this position; in other words, pass a list within a single-element list). For string splice, an optional string to substitute for the removed characters. |
Removes and optionally inserts elements in lists and strings. For a similar operator that only removes value and does not return any value, see the splice operator.
extract lvalue, offset_expression, [length_expression, [substitution_expression]]my list $sublist = extract $list, 2, 2;
my string $substring = extract $string, 2, 2, "-text-";
Works on either strings or lists in a similar way; removes elements from a list or characters from a string and optionally inserts new ones. If no length_expression is given, extract removes all elements/characters from the list or string starting at offset_expression) (list and string offsets begin at 0). Otherwise, a number of elements/characters equal to length_expression is removed (or up to the end of the list/string if applicable). If substitution_expression is present, then the removed elements/characters are substituted with the elements/string given by this expression.
When operating on lists, a list is returned of any elements extracted (if no elements are extracted, then an empty list is returned); when operating on strings, a string is extracted of all characters extracted from the string (if no characters are extracted, then an empty string is returned).
Note that string extract takes character offsets, which may not be the same as byte offsets for multi-byte character encodings, such as UTF-8
Table 2.75. Arguments Processed By extract
Argument | Processing |
|---|---|
If the lvalue is a list, list elements are processed, otherwise, if it is a string, characters in the string are processed. For any other data type, no action is taken. | |
| The start element/character position for removing elements/characters from the list or string. |
| The number of elements/characters to remove. If this expression is not present, then all elements/characters from the offset to the end of the list/string are removed. If this expression is present and evaluates to 0, no characters/elements are removed. |
| For list extract, an optional element or list to substitute for the removed elements (to insert a list in a single element's position, make sure that the list to be inserted is the first and only element of another list used as the argument in this position; in other words, pass a list within a single-element list). For string extract, an optional string to substitute for the removed characters. |
Assigns a value to an lvalue and returns the value assigned.
lvalue=expression
$a = 1;
Increments and concatentates an lvalue with the value of an expression depending on the data type of the lvalue, unless the lvalue is NOTHING, in which case this operator acts like the assignment operator (simply assigns the value of the right hand side to the lvalue).
lvalue+=expression
$a += 10;
$date += P1M2DT45M;
$list += $new_element;
$string += ".foo";
$binary += <0c67a374>
$hash += ("new-key" : 1, "other" : "two");$object += $hash;
Table 2.77. Arguments Processed By +=
Argument | Processing |
|---|---|
list | the expression will be evaluated and concatenated to the lvalue. If expression is a list, the lists will be concatenated, to ensure adding a single element to a list, use the push operator (see Push Operator). |
the expression will be evaluated, and, if it is a hash or object, then it's members will be added to the lvalue, any duplicate elements in the lvalue will be overridden by elements in the expression. | |
string | the expression will be evaluated and converted to a string if necessary and concatenated to the lvalue. |
float | the expression will be evaluated and converted to a float if necessary and added to the lvalue. |
binary | the expression will be evaluated and converted to a binary if necessary and added to the lvalue. |
date | the expression will be evaluated and converted to a date if necessary and added to the lvalue. |
nothing | the lvalue will be assigned to the value of |
| the lvalue's type will be converted to an integer, and the expression will be evaluated and converted to an integer if necessary, and then the result will be added to the lvalue. |
For a float or integer argument, decrements the value of an lvalue by the value of an expression. However if the lvalue is a hash and the expression is a string, removes the key represented by the string from the hash.
lvalue-=expression
$a -= 10;
$date -= PT45H213S;
$hash -= "key";
$hash -= ("key1", "key2");$object -= "key";
$object -= $list_of_keys;
Table 2.78. Arguments Processed By -=
Argument | Processing |
|---|---|
float | the expression will be evaluated and converted to a float if necessary and subtracted from the lvalue |
date | the expression will be evaluated and converted to a date if necessary and subtracted from the lvalue |
the hash key represented by | |
each element in the list will be converted to a string (if necessary) and the key represented by each string will be removed from the hash | |
the lvalue will be assigned to | |
| the lvalue's type will be converted to an integer (if necessary), and the expression will be evaluated and converted to an integer (if necessary), and then the result will be subtracted from the lvalue |
Performs a bitwise (binary) AND operation on an lvalue using the value of an expression and returns the new value.
lvalue&=expression
$a &= 0xfe;
Table 2.79. Arguments Processed By &=
Argument | Processing |
|---|---|
All | the lvalue's type will be converted to an integer if necessary, and the expression will be evaluated and converted to an integer as well if necessary, and then the result will be binary and'ed to the lvalue |
Performs a bitwise (binary) OR operation on an lvalue using the value of an expression and returns the new value.
lvalue|=expression
$a |= 0xba;
Table 2.80. Arguments Processed By |=
Argument | Processing |
|---|---|
All | the lvalue's type will be converted to an integer if necessary, and the expression will be evaluated and converted to an integer as well if necessary, and then the result will be binary or'ed to the lvalue |
Performs a modula calculation on an lvalue using the value of an expression and returns the new value.
lvalue%=expression
$a %= 100;
Table 2.81. Arguments Processed By %=
Argument | Processing |
|---|---|
All | the lvalue's type will be converted to an integer if necessary, and the expression will be evaluated and converted to an integer as well if necessary, and then the result will be used to divide the lvalue's value and the remainder will be saved to the lvalue |
Table 2.82. Arguments Processed By *=
Argument | Processing |
|---|---|
All | If either side of the operator is a float, the result will be a float as well. Otherwise the result is an integer value. The expression will be evaluated and multiplied by the lvalue, and the result will be saved to the lvalue. |
Table 2.83. Arguments Processed By *=
Argument | Processing |
|---|---|
All | If either side of the operator is a float, the result will be a float as well. Otherwise the result is an integer value. The expression will be evaluated and used to divide the lvalue, and the result will be saved to the lvalue. |
Table 2.84. Exceptions Thrown by /*
err | desc |
|---|---|
| If the divisor expression evaluates to zero, this exception is thrown. |
Performs an exclusive-or operation on an lvalue using the value of an expression.
lvalue^=expression
$a ^= 0xf9034ba7;
Table 2.85. Arguments Processed By ^=
Argument | Processing |
|---|---|
All | Values are converted to integers if necessary. The expression will be evaluated and exclusive-or'ed with the lvalue, and the result will be saved to the lvalue. |
Performs a shift-left operation on an lvalue using the value of an expression and returns the value assigned.
lvalue<<=expression
$a <<= 3;
Table 2.86. Arguments Processed By <<=
Argument | Processing |
|---|---|
All | Values are converted to integers if necessary. The expression will be evaluated and this value will determine how many bits the lvalue will be shifted left. The result will be saved to the lvalue. |
Performs a shift-right operation on an lvalue using the value of an expression and returns the value assigned.
lvalue>>=expression
$a >>= 3;
Table 2.87. Arguments Processed By >>=
Argument | Processing |
|---|---|
All | Values are converted to integers if necessary. The expression will be evaluated and this value will determine how many bits the lvalue will be shifted right. The result will be saved to the lvalue. |
Regular expression functionality in Qore is provided by PCRE: Perl-Compatible Regular Expression library.
Using this library, Qore implements regular expression pattern matching using the same syntax and semantics as Perl 5.
The following is a list of operators based on regular expressions (or similar to regular expressions in the case of the transliteration operator).
Table 2.88. Regular Expression Operators
Operator | Description |
|---|---|
Returns | |
Returns | |
Substitutes text in a string based on matching a regular expression. | |
Returns a list of substrings in a string based on matching patterns defined by a regular expression. | |
Not a regular expression operator; transliterates one or more characters to other characters in a string. |
See the table below for valid regular expression options.
Table 2.89. Regular Expression Options
Option | Description |
|---|---|
i | Ignores case when matching |
m | makes start-of-line (^) or end-of-line ($) match after or before any newline in the subject string |
s | makes a dot (.) match a newline character |
x | ignores whitespace characters and enables comments prefixed by # |
g | makes global substitutions (only applicable with the substitution operator) |
The following is a list of functions providing regular expression functionality where the pattern may be given at run-time:
Table 2.90. Regular Expression Functions
Function | Description |
|---|---|
Returns | |
Substitutes a pattern in a string based on regular expressions and returns the new string. | |
Returns a list of substrings in a string based on matching patterns defined by a regular expression. |
Date/time arithmetic is relatively straightforward and should normally produce the expected results. However with leap years, months with different lengths, and daylights savings time the situation can be confusing; this section will clarify how Qore does date arithmetic considering these special cases.
Adding or subtracting years and months (ex: $date += 2Y + 3M) will give you the same day on the desired month in the desired year. If the target month has fewer days than the source month, then you will get the last day of the month in that year. For example:
prompt% qore -X '2004-02-29Z - 1Y'
2003-02-28 00:00:00 Fri Z (UTC)Adding or subtracting days means adding or subtracting 24h periods; i.e. you will get the same time in the result of subtracting days, for example:
prompt% qore -X '2004-02-29T10:15:00Z - 10D'
2004-02-19 10:15:00 Thu Z (UTC)Subtracting one absolute date from another will result in a relative date, normalized to the hour (that is, microseconds over 999,999 are converted to seconds, seconds over 59 to minutes, and minutes over 59 to hours; days, months, and years will not appear in the result as they do not indicate a fixed period of time but rather can vary in length depending on the absolute date/time starting point. For example:
prompt% qore -X '2007-02-29T10:15:03.255Z - 2004-02-29T10:14:02.100Z'
<time: 26304 hours 1 minute 1 second 155 milliseconds>To find the difference in seconds between two dates, convert each date value to an integer and subtract as follows:
prompt% qore -X 'int(2004-02-29Z) - int(2004-02-28Z)'
86400Or use the get_duration_seconds() function as follows:
prompt% qore -X 'get_duration_seconds(2004-02-29Z - 2004-02-28Z)'
86400Time zones and daylight savings time information is supplied by the system's zoneinfo database (if any exists; see Time Zones for more information).
To find out if the current time zone has daylight savings time, execute the following:
prompt% qore -X 'TimeZone::get().hasDST()'
TrueSee the TimeZone class for more information on time zone information.
Qore is capable of representing and performing calculations on dates before the adoption of the Gregorian calendar (proposed in 1582 and adopted at various times in Europe after this point). However all calculations are made as if the Gregorian calendar were always in effect (Qore implements a proleptic Gregorian calendar).
Non-block statements in Qore are always terminated by a semi-colon ";" as in Perl, C, or Java. Statements can be grouped into blocks, which are delimited by curly brackets "{" and "}" containing zero or more semi-colon delimited statements, as in C or Java. Like C, C++, and Java, but unlike perl, any Qore statement taking a statement modifier will accept a single statement or a statement block.
A statement can be any of the following (note that statements are also recursively defined):
Table 2.91. Qore Statements
Type | Examples | Reference |
|---|---|---|
An expression that changes an lvalue | $var = 1;
$var += 5;
$var[1].count++;
shift $var.key[$i]; | |
An expression with the new operator | new ObjectClass(1, 2, 3); | |
An expression with the background operator | background function(); | |
A call reference or closure call | $call_reference($arg1, $arg2); | |
A method call | $object.method(1, 2, 3); | |
An if statement | if ($var == 3) | if and else statements |
An "if ... else" statement | if ($var == 3) | if and else statements |
A while statement | while ($var < 10)
| while statements |
A do while statement | do | do while statements |
A for statement | for ( | for statements |
A foreach statement | foreach | foreach statements |
A switch statement | switch ( | switch statements |
A return statement | return | return statements |
A local variable declaration |
my $var; my ($a, $b, $c); | |
A global variable declaration |
our $var; our ($a, $b, $c); | |
A function call | calculate($this, $that, $the_other); | |
A continue statement | continue; | continue statements |
A break statement | break; | break statements |
A statement block | { | one or more statements enclosed in curly brackets. |
A throw statement | throw | throw statements |
try and catch statements | try | try and catch statements |
A rethrow statement | rethrow; | rethrow statements |
A thread_exit statement | thread_exit; | thread_exit statements |
A context statement |
context | context statements |
A summarize statement |
summarize | summarize statements |
A subcontext statement | subcontext
| subcontext statements |
An on_exit statement | on_exit
| on_exit statements |
An on_success statement | on_success
| on_success statements |
An on_error statement | on_error
| on_error statements |
The if statement allows for conditional logic in a Qore program's flow; the syntax is similar to that of C, C++, or Java.
if (expression)statement[elsestatement]
Qore if statements work like if statements in C or Java. If the result of evaluating the expression converted to a Boolean value is True, then the first statement (which can also be a block) is executed. If the result is False, and there is an else keyword after the first statement, the following statement is executed.
Any expression that evaluates to a non-zero integer value will be converted to a Boolean True. Any expression that evaluates to zero value is interpreted as False. This is more like C and Java's behavior and not like Perl's (where any non-null string except "0" is True).
The Qore for statement is most similar to the for statement in C and Java, or the non array iterator for statement in Perl. This statement is ideal for loops that should execute a given number of times, then complete. Each of the three expressions in the for statement is optional and may be omitted. To iterate through a list without directly referencing list index values, see the foreach statement.
for ([initial_expression];[test_expression];[iterator_expression])statement
[initial_expression]
The initial_expression is executed only once at the start of each for loop. It is typically used to initialize a loop variable.
[test_expression]
The test_expression is executed at the start of each for loop iteration. If this expression evaluates to Boolean False, the loop will terminate.
[iterator_expression]
The iterator_expression is executed at the end of each for loop iteration. It is typically used to increment or decrement a loop variable that will be used in the test_expression.
Here is an example of a for loop using a local variable:
for (my $i = 0; $i < 10; $i++) print("%d\n", $i);
The Qore foreach statement is most similar to the for or foreach array iterator statement in Perl. To iterate an action until a condition is true, use the for statement instead.
foreach [my]$variablein (expression)statement
If the expression does not evaluate to a list, then the variable will be assigned the value of the expression evaluation and the statement will only execute one time. Otherwise the variable will be assigned to each value of the list and the statement will be called once for each value.
Here is an example of a foreach loop using a local variable:
# if $str_list is a list of strings, this will remove all whitespace from the # strings; the reference in the list expression ensures that changes # to the iterator variable are written back to the list foreach my $string in (\$str_list) trim $string;
Note that if a reference (\$lvalue_expression) is used as the list expression, any changes made to the foreach iterator variable will be written back to the list.
The Qore switch statement is similar to the switch statement in C and C++, except that the case values can be any expression that does not need run-time evaluation and can also be expressions with simple relational operators or regular expressions using the switch value as an implied operand.
switch (expression) { casecase_expression:statement(s)... [ default :statement(s)] }
switch ($val) { case < -1: printf("less than -1\n"); break; case "string": printf("string\n"); break; case > 2007-01-22T15:00:00: printf("greater than 2007-01-22 15:00:00\n"); break; case /abc/: printf("string with 'abc' somewhere inside\n"); break; default: printf("default\n"); break; }
The first expression is evaluated and then compared to the value of each case expression in declaration order until one of the case expressions matches or is evaluated to True. In this case all code up to a break statement is executed, at which time execution flow exits the switch statement. Unless relational operators are used, the comparisons are "hard" comparisons; no type conversions are done, so in order for a match to be made, both the value and types of the expressions must match exactly. When relational operators are used, the operators are executed exactly as they are in the rest of qore, so type conversions may be performed if nesessary.
To use soft comparisons, you must explicitly specify the soft equals operator as follows:
switch (1) { case == "1": print("true\n"); break; }
If no match is found and a default label has been given, then any statements after the default label will be executed. If a match is made, then the statements following that case label are executed.
To break out of the switch statement, use the break statement.
Table 2.92. Valid Case Expression Operators
Operator | Description |
|---|---|
> | |
>= | |
< | |
<= | |
== | |
~= | regular expression match operator (in this case the regular expression may be optionally given without the operator) |
!~ |
while statements in Qore are similar to while statements in Perl, C and Java. They are used to loop while a given condition is True.
while (expression)statement
First the expression will be evaluated; if it evaluates to True, then statement will be executed. If it evaluates to False, the loop terminates.
$a = 1;
while ($a < 10)
$a++;do while statements in Qore are similar to do while statements in C. They are used to guarantee at least one iteration and loop until a given expression evaluates to False.
dostatementwhile (expression);
First, statement will be executed, then expression will be evaluated; if it evaluates to True, then the loop iterates again. If it evaluates to False, the loop terminates. The difference between do while statements and while statements is that the do while statement evaluates its loop expression at the end of the loop, and therefore guarantees at least one iteration of the loop.
$a = 1; do $a++; while ($a < 10);
Skips the rest of a loop and jumps right to the evaluation of the iteration expression.
continue;The continue statement affects loop processing; that is; it has an affect on for, foreach, while, do while, context, summarize, and subcontext loop processing. When this statement is encountered while executing a loop, execution control jumps immediately to the evaluation of the iteration expression, skipping any other statements that might otherwise be executed.
Exits immediately from a loop statement or switch block.
break;The break statement affects loop processing; that is; it has an affect on for, while, do while, context, summarize, and subcontext loop processing. When this statement is encountered while executing a loop, the loop is immediately exited, and execution control passes to the next statement outside the loop.
In order to throw an exception explicitly, the throw statement must be used.
throw expression;The expression will be passed to the catch block of a try/catch statement, if the throw is executed in a try block. Otherwise the default system exception handler will be run and the currently running thread will terminate.
Qore convention dictates that a direct list is thrown with at least two string elements, the error code and a description. All system exceptions have this format. See try and catch statements for information on how to handle exceptions, and see Exception Handling for information about how throw arguments are mapped to the exception hash.
Some error conditions can only be detected and handled using exception handlers. To catch exceptions, try and catch statements can to be used. When an exception occurs while executing the try block, execution control will immediately be passed to the catch block, which can capture information about the exception.
trystatementcatch ([$exception_hash_variable])statement
A single variable can be specified in the catch block to be instantiated with the exception hash, giving information about the exception that has occurred. For detailed information about the exception hash, see Exception Handling.
If no variable is given in the catch declaration, it will not be possible to access any information about the exception in the catch block. However, the rethrow statement can be used to rethrow exceptions at any time in a catch block.
A rethrow statement can be used to rethrow an exception in a catch block. In this case a entry tagged as a rethrow entry will be placed on the exception call stack. This statement can be used to maintain coherent call stacks even when exceptions are handled by more than one catch block (for detailed information about the exception hash and the format of call stacks, see Exception Handling).
rethrow;The rethrown exception will be either passed to the next higher-level catch block, or to the system default exception handler, as with a throw statement. Note that it is an error to call rethrow outside of a catch block.
thread_exit statements cause the current thread to exit immediately. Use this statement instead of the exit() function when only the current thread should exit.
thread_exit;This statement will cause the current thread to stop executing immediately.
To easily iterate through multiple rows in a hash of arrays (such as a query result set returned by the Datasource::select() method), the context statement can be used. Column names can be referred to directly in expressions in the scope of the context statement by preceding the name with a '%" character.
context [name] (data_expression )[ where (expression) ] [ sortBy (expression) ] [ sortDescendingBy (expression) ]statement
data_expression
This must evaluate to a hash of arrays in order for the context statement to execute.
[ where ( expression ) ]
An optional where expression may be given, in which case for each row in the hash,
the expression will be executed, and if the where expression evaluates to True,
the row will be iterated in the context loop. If this expression evaluates
to False, then the row will not be iterated. This option is given so the
programmer can create multiple views of a single data structure (such as a query
result set) in memory rather than build different data structures by
hand.
[ sortBy ( expression ) ]
An optional sort_by expression may also be given. In this case, the expression will be evaluated for each row of the query given, and then the result set will be sorted in ascending order by the results of the expressions according to the resulting type of the evaluated expression (i.e. if the result of the evaluation of the expression gives a string, then string order is used to sort, if the result of the evaluation is an integer, then integer order is used, etc).
[ sortDescendingBy ( expression ) ]
Another optional modifier to the context statement that behaves the same as above except that the results are sorted in descending order.
# note that "%service_type" and "%effective_start_date" represent values # in the $service_history hash of arrays. context ($service_history) where (%service_type == "voice") sortBy (%effective_start_date) { printf("%s: start date: %s\n", %msisdn, format_date("YYYY-MM-DD HH:mm:SS", %effective_start_date)); }
summarize statements are like context statements with one important difference: results sets are grouped by a by expression, and the statement is executed only once per discrete by expression result. This statement is designed to be used with the subcontext statement.
summarize (expression) by (expression) [ where (expression) ] [ sortBy (expression) ] [ sortDescendingBy (expression) ]statement
summarize statements modifiers have the same effect as those for the context statement, except for the following:
by ( expression )
The by expression is executed for each row in the data structure indicated. The set of unique results defines groups of result rows. For each group of result rows, each row having an identical result of the evaluation of the by expression, the statement is executed only once.
# note that "%service_type" and "%effective_start_date" represent values # in the $services hash of arrays. summarize ($services) by (%effective_start_date) where (%service_type == "voice") sortBy (%effective_start_date) { printf("account has %d service(s) starting on %s\n", context_rows(), format_date("YYYY-MM-DD HH:mm:SS", %effective_start_date)); }
Statement used to loop through values within a summarize statement.
subcontext [ where (expression) ] [ sortBy (expression) ] [ sortDescendingBy (expression) ]statement
The subcontext statement is used in conjunction with summarize statements. When result rows of a query should be grouped, and then each row in the result set should be individually processed, the Qore programmer should first use a summarize statement, and then a subcontext statement. The summarize statement will group rows, and then the nested subcontext statement will iterate through each row in the current summary group.
summarize ($services) by (%effective_start_date) where (%service_type == "voice") sortBy (%effective_start_date) { printf("account has %d service(s) starting on %s\n", context_rows(), format_date("YYYY-MM-DD HH:mm:SS", %effective_start_date)); subcontext sortDescendingBy (%effective_end_date) { printf("\tservice %s: ends: %s\n", %msisdn, format_date("YYYY-MM-DD HH:mm:SS", %effective_end_date)); } }
return statements causes the flow of execution of the subroutine, method or program to stop immediately and return to the caller. This statement can take an optional expression to return a value to the caller as well.
return [expression];This statement causes execution of the current subroutine, method, or program to cease and optionalls returns a value to the caller.
sub getName() returns string { return "Barney"; } my string $name = getName();
Queues a statement or statement block for unconditional execution when the block is exited, even in the case of exceptions or return statements. For similar statement that queue code for execution depending on the exception status when the block exits, see on_success statements and on_error statements.
on_exit
statementThe on_exit statement provides a clean way to do exception-safe cleanup within Qore code. Any single statment (or statement block) after the on_exit keyword will be executed when the current block exits (as long as the statement itself is reached when executing - on_exit statements that are never reached when executing will have no effect). The the position of the on_exit statement in the block is important, as the immediate effect of this statement is to queue its code for execution when the block is exited. Even if an exception is raised or a return statement is executed, any on_exit code that is queued will be executed. Therefore it's ideal for putting cleanup code right next to the code that requires the cleanup.
Note that if this statement is reached when executing in a loop, the on_exit code will be executed for each iteration of the loop.
By using this statement, programmers ensure that necessary cleanup will be performed regardless of the exit status of the block (exception, return, etc).
{
$mutex.lock();
# here we queue the unlock of the mutex when the block exits, even if an exception is thrown below
on_exit $mutex.unlock();
if ($error)
throw "ERROR", "Scary error happened";
print("everything's OK!\n");
return "OK";
}
# when the block exits for any reason, the mutex will be unlocked
Queues a statement or statement block for execution when the block is exited in the case that no exception is active. Used often in conjunction with the on_error statement and related to the on_exit statement.
on_success
statementThe on_success statement provides a clean way to do block-level cleanup within Qore code in the case that no exception is thrown in the block. Any single statment (or statement block) after the on_success keyword will be executed when the current block exits as long as no unhandled exception has been thrown (and as long as the statement itself is reached when executing - on_success statements that are never reached when executing will have no effect). The the position of the on_success statement in the block is important, as the immediate effect of this statement is to queue its code for conditional execution when the block is exited. Even if a return statement is executed later in the block, any on_success code that is queued will be executed as long as there is no active (unhandled) exception. Therefore it's ideal for putting cleanup code right next to the code that requires the cleanup, along with on_error statements, which are executed in a manner similar to on_success statements, except on_error statements are only executed when there is an active exception when the block is exited.
Note that if this statement is reached when executing in a loop, the on_success code will be executed for each iteration of the loop (as long as there is no active exception).
{
$db.beginTransaction();
# here we queue the commit in the case there are no errors
on_success $db.commit();
# here we queue a rollback in the case of an exception
on_error $db.rollback();
$db.select("select * from table where id = %v for update", $id);
# .. more code
return "OK";
}
# when the block exits. the transaction will be either committed or rolled back,
# depending on if an exception was raised or not
Queues a statement or statement block for execution when the block is exited in the case that no exception is active. Used often in conjunction with the on_success statement and related to the on_exit statement.
on_error
statementThe on_error statement provides a clean way to do block-level cleanup within Qore code in the case that an exception is thrown in the block. Any single statment (or statement block) after the on_error keyword will be executed when the current block exits as long as an unhandled exception has been thrown (and as long as the statement itself is reached when executing - on_error statements that are never reached when executing will have no effect). The the position of the on_error statement in the block is important, as the immediate effect of this statement is to queue its code for conditional execution when the block is exited. Even if a return statement is executed later in the block, any on_error code that is queued will be executed as long as there is an active (unhandled) exception. Therefore it's ideal for putting cleanup code right next to the code that requires the cleanup, along with on_success statements, which are executed in a manner similar to on_error statements, except on_success statements are only executed when there is no active exception when the block is exited.
Note that the code in this statement can only be executed once in any block, as a block (even a block within a loop) can only exit the loop once with an active exception (in contrast to on_success and on_exit statements, which are executed for every iteration of a loop).
{
$db.beginTransaction();
# here we queue the commit in the case there are no errors
on_success $db.commit();
# here we queue a rollback in the case of an exception
on_error $db.rollback();
$db.select("select * from table where id = %v for update", $id);
# .. more code
return "OK";
}
# when the block exits. the transaction will be either committed or rolled back,
# depending on if an exception was raised or not
A subroutine is declared in Qore by using the key word sub as follows:
subsubroutine_name([[type] variable1, ...]) [returnstype] {statements;}
Variables listed in parentheses after the subroutine name are the parameters to the subrouting and automatically get local lexical scoping. In order to process a variable number of arguments to a function, the $argv variable (local variable) is instantiated as a list with the remaining arguments passed to the subroutine. Type declarations optionally precede the parameter variable and will restrict any arguments passed to the type declared. The same subrouting can be declared multiple times if each declaration has different parameter types; this is called overloading the subroutine.
Subroutines can use the return statement to provide a return value. Subroutine names must be valid Qore identifiers.
The return type of the subroutine can be given by listing a type declaration after the returns keyword after the parentheses after the subroutine name.
Note that parameter and return types are required when the PO_REQUIRE_TYPES or PO_REQUIRE_PROTOTYPES parse options are set.
Variables passed as function arguments are passed by value by default, unless the caller places a "\" character before an lvalue in the argument list. In this case the subroutine must have a parameter defined to accept the variable passed by reference. Any changes to the local variable will be reflected in the original variable for variables passed by reference. Also note that it is illegal to pass an argument by reference in a background expression.
Subroutines can return values to the calling expression by using the return statement, with the following syntax:
return expression;Here is an example subroutine declaration for a function returning a value:
#!/usr/bin/qore # # subroutine declaration example sub print_string(string $string) returns int { print("%s\n", $string); return 1; }
Subroutines may also be recursive. Here is an example of a recursive Qore subroutine definition implementing the Fibonacci function:
#!/usr/bin/qore # # recursive subroutine example sub fibonacci(int $num) returns int { if ($num == 1) return 1; return $num * fibonacci($num - 1); }
Function names are resolved during the second parse pass; therefore functions do not need to be declared before being referenced. This allows an easy definition of 2 or more self-referencing functions.
Namespaces allow constants, classes, and even other namespaces with the same name to co-exist in the same program by defining them in separate namespaces. Constants, classes, and sub-namespaces can be declared to belong to a particular namespace either by defining them in-line within a namespace declaration, or by including the namespace name/path prepended to the constant, class, or namespace declaration separated by two colons "::".
If the user does not specify the parent namespace with a namespace path in constant, class, or namespace declarations, the declaration will be by default in the unnamed default root namespace.
In-line namespace declaration:
namespace [namespace_path::]namespace_identifier{ [constant_declarations] [class_declarations] [sub-namespace-declarations] }
Out of line namespace declaration:
namespace [namespace_path::]namespace_identifier;
Namespaces can either be resolved by giving a path to the constant, class, or namespace desired, or by leaving out the namespace path and allowing the system to search for the constant, class, or namespace. In either case, a depth-first search of the namespace tree is made for a match. If a namespace path is included, then the tree is searched for the first namespace match. and, if the rest of the declaration cannot be matched, the search continues in the entire namespace tree until a complete match is found.
Namespace paths look like the following:
starting_namespace::[sub-namespaces::]constant|class|namespace
All Qore-language constants and classes are defined in the Qore namespace or in a subnamespace of the Qore namespace. The Qore namespace is a direct child of the unnamed default root namespace (::).
For detailed information on all constants and classes defined in system namespaces, see System Namespaces and Class Library.
Constant definitions allow programmers to refer to values with Qore identifiers rather than using the value or the expression that generates the value. Constants are defined with the following syntax:
const [namespace_path::]constant_identifier=expr;
The expression cannot be an expression that has side effects or a parse exception will be raised. Only functions tagged with the CONST can be used to initialize a constant.
Objects are instantiations of a Qore class. Classes can define private members and methods, which are functions that operate only on the objects of that class.
Classes are declared with the following syntax:
class [namespace_path::...]class_identifier[inherits [private|public] [namespace_path::...]class_identifier[, ...]] { [private$.var1[, ...];] [private { [type]$.member_name[ =expr]; ... }] [public { [type]$.member_name[ =expr]; ... }] [static] [synchronized] [private] [namespace_path::]method_name_identifier([[type] $var1, ...]) {statements; } ... }
Alternatively objects can be defined out of line as follows:
class [namespace_path::]class_identifier[inherits [private|public] [namespace_path::...]class_identifier[, ...]]; [static] [synchronized] [private] [namespace_path::]class_identifier::method_identifier([[type] $var1, ...]) {statements; }
Public and private members can only be declared in an in-line class definition (the first example above). If a class has at least one public member declared (or inherits a class with at least one public member declared), then only those members declared as public can be accessed from outside the class, and from within the class only members explicitly declared can be accessed as well (unless the class also defines a memberGate() method). In this way typographical errors in member names can be caught (at parse time if types are declared).
In a class hierarchy, base class constructor methods can be explicitly specified using a special syntax unique to subclass constructor methods. Please see Class Inheritance for more information.
It's possible to write purely object-oriented scripts/programs in Qore by defining an application class and using the -x or --exec-class command-line arguments to tell Qore to instantiate the class instead of doing normal top-level execution (in fact, the --exec-class arguments disallow the use of top-level statements entirely). For more information, please see Command-Line Parsing and Parse Directives.
Methods declared with the private keyword can only be called by other member functions of the same class. Any attempt to call these methods from outside the class will result in a run-time exception.
Methods declared with the synchronized keyword will only run in one thread at a time.
Methods declared with the static keyword are like regular subroutines that are attached to the class. These methods are not associated with a particular object's state and therefore are not allowed to refer to object members or call non-static methods. Also, no reference to the special $self variable is allowed within static methods.
Static method calls take a special syntax as documented here.
All class methods are optional, but some methods have a special meaning.
Table 2.93. Special Methods
Name | Description |
|---|---|
|
Called when objects are created when instantiated by a variable declaration with a class type and constructor arguments or explicitly with the new operator. User code may not explicitly call
|
|
When a user explicitly calls a copy method, Qore will generate a new object with references to the same members as the source object. Then, if there are any base classes, base class
|
|
Called when objects go out of scope or are explicitly deleted. User code may not explicitly call
|
|
If this method is implemented in the class, it is called when read access is attempted to private member or members that do not exist in the current object; the return value of this method is returned as the value of the member.
| |
|
If this method is implemented in the class, it is called when methods are called on the object that do not exist in the current object and the return value of this method is returned as the value of the method call.
| |
|
If this method is implemented in the class, it is called when an object member is updated outside the class with the member name as the argument. Note that this method is called after the member has been updated and without locking; the call is not atomic respective to other threads that also may update the same member simultaneously.
|
When defining a class, members of the current object can be referred to with a special syntax as follows:
$.member_name_identifierFurthermore, the automatic variable $self is instantiated which represents the current object (similar to the this in C++ or Java). Therefore if you need to access hash members which are not valid Qore identifiers, then enclose the member name in double quotes after the dot operator as follows:
$self."&member-name"
The automatic $argv local variable is instantiated as usual in all class methods where there are more arguments than variables declared in the method declaration.
If the class implements a memberGate() method, then whenever a non-existant member of the class is accessed (read), this method will be called with the name of the member as the sole argument, so that the class can create the member (or react in some other way) on demand. This method is also called when methods of the same class try to access (read) non-existant methods, but is not called from within the memberGate() method itself.
To monitor writes to the object, the class can implement a memberNotification() method, which is called whenever an object member is modified from outside class member code. In this case, the memberNotification() method is called with the name of the member that was updated so that an object can automatically react to changes to its members (writes to members) from outside the class. This method is not called when members are updated from within class member code.
If a class has at least one public member declared (or inherits a class with at least one public member declared), then only those members declared as public can be accessed from outside the class, and from within the class only members explicitly declared can be accessed as well (unless the class also defines a memberGate() method). In this way typographical errors in member names can be caught (at parse time if types are declared).
Within a class method definition, calls to methods in the same class hierarchy (of the current class or a base class) can be defined as follows:
[namespace_path::]$.method_name([arg, ...])
For example:
# to call base class Mutex::lock() Thread::Mutex::$.lock(); # to call lock() in the current (or lower base) class $.lock();
This syntax can only be used to call methods in the current class or in base classes. This is because these calls are resolved at parse time, and only these classes are known and accessible at parse time. To call a derived class method from a base class, you must use the $self variable to call the method, so that the call will be resolved at run-time, for example:
# this way, "member" can be resolved to a derived class method $self.member();
Calls to object methods can be made outside the class by using the above syntax as well. All such calls are resolved at run-time, therefore if the call is made to a private function outside the defining class, then a run-time METHOD-IS-PRIVATE (if the method is private) or BASE-CLASS-IS-PRIVATE (if the method resolves to a privately-inherited base class) exception will be raised.
Class inheritance is a powerful concept for easily extending and resuing object-oriented code, but is also subject to some limitations. This section will explain how class inheritance works in Qore.
Classes inherit the methods of a parent class by using the inherits as specified above. Multiple inheritance is supported; a single Qore class can inherit one or more classes. When a class is inherited by another class, it is called a base class. Private inheritance is speficied by including the keyword private before the inherited class name. When a class is privately inherited, it means that the inherited class' public members are treated as private members in the context of accesses outside the class.
Inheritance is public by default, to inherit a class privately, use the private keyword before the class name or class path to inherit.
It is not legal to directly inherit the same class more than once; that is; it is not legal to list the same class more than once after the inherits keyword. However, it is possible that a base class could appear more than once in the inheritance tree if that class is inherited separately by two or more classes in the tree. In this case, the base class will actually only be inherited once in the subclass, even though it appears in the inheritance tree more than once. This must be taken into consideration when designing class hierarchies, particularly if base class constructor parameters for that class are explicitly provided in a different way by the inheriting classes.
Class members only exist once for each object; therefore if classes in an inheritance tree have different uses for members with the same name, then a class hierarchy built of such classes will probably not function properly.
Subclasses can give explicit arguments to their base class constructors using a special syntax (only available to subclass constructors) similar to the C++ syntax for the same purpose as follows:
class_name::constructor([[type] $var1, ...) :base_class_identifier(expression(s))[, ...] {statements; }
Here is a concrete example of giving arguments to an inherited base class:
class XmlRpcClient inherits Qore::HTTPClient { # calls the base class HTTPClient constructor, overrides the "protocols" key to "xmlrpc" constructor(hash $opts = hash()) : Qore::HTTPClient($opts + ( "protocols" : "xmlrpc" )) ... }
Because base class constructors are executed before subclass constructors, the only local variables in the constructor that can be referenced are those declared in the subclass constructor declaration (if any). What this means is that if you declare local variables in the expressions giving base class arguments, these local variables are not accessible from the constructor body.
Base classes that give explicit arguments to their base class constructors can be overridden by subclasses by simply listing the base class in the base class constructor list and providing new arguments.
Like Java, in Qore, objects are treated differently from all other data types in that they are by default passed as arguments to functions and methods by passing a copy of a reference to the object. That means that passing an object to a function that modifies the object will by default modify the original object and not a copy, however reassigning a local parameter variable assigned an object passed as an argument (that is only assigned to a local variable in the calling function) will not result in deleting the object, but rather decrement its scope reference count (note that if the object were created as a part of the call and reassigning the variable would cause the object's scope reference count to reach zero, then the object would be deleted in this case).
Assigning an object to a variable has the same effect; a copy of a reference to the object is assigned to the variable. This results in prolonging the object's scope (by owning a new copy of a reference to the object).
An example:
sub test2(any $x) { # we can modify the original object like so: $x.member = "tree"; # here we re-assign $x, but since the object is also assigned # to $o in the calling function, the object's scope is still # valid, and therefore nothing happens so the object $x = 1; } sub test() { my TestObject $o(); # here we pass a copy of a reference to the object in $o test2($o); # this will print out "ok\n", because the object is still # valid and the member has been set by test2() if ($o.member == "tree") print("ok\n"); } # when test() exits, the object in $o will go out of scope # and be deleted
If, however, an object is passed by reference, then the local variable of the called function that accepts the object owns the scope reference of the calling functions's variable.
An example:
sub test2(any $x) { # we can modify the original object like so: $x.member = "tree"; # here we re-assign $x, and since we own the only scope # reference to the object, the object will go out of # scope here and be deleted $x = 1; } sub test() { my TestObject $o(); # here we pass a reference to the object in $o test2(\$o); # the object has already been deleted in test2() and # therefore nothing will be printed out if ($o.member == "tree") print("ok\n"); }
Objects are automatically deleted when their scope-relevant reference count reaches zero (note that objects can be deleted manually at any time by using the delete operator). Whenever an object is deleted, the object's class' destructor method is run on the object.
The following affect objects' scope:
Variable Assignments
An object's automatic scope is prolonged as long as the object is assigned to a local variable.
Existence of a Closure Created Within the Object
Any closures created from within the object encapsulate the object's state (along with any local variables referenced within the closure) and also prolong the object's automatic scope as long as the closure exists.
Object Method Thread Launched Within the Object
If a member function thread was launched from within the object using the background operator, the object's automatic scope is prolonged to the life of the new thread. Object threads started externally to the object (i.e. not directly from an expression with the background operator within a method) will not prolong the scope of the object.
If an object with running threads is explicitly deleted, and this case is not handled in the object's destructor() method (by ensuring that all other running threads terminate gracefully), exceptions will be thrown in other threads at any attempt to access the already-deleted object.
For more information about threading, please see the following section Threading
The fact that object threads and closures can prolong object scope means, for example, that objects assigned to local variables can exist for longer than the scope of their host variable if they have one or more methods running in other threads or if closures created from within the object still exist at the time the local variable goes out of scope.
To explicitly generate a copy of an object, the copy() constructor must be called. This is a special method that exists implicitly for every class even if it is not explicitly defined (like constructor() and destructor() methods). The implicit behavior of the copy() constructor is to create a new object with new members that are copies of the original members (except objects are once again referenced). Then, if any copy() method exists, it will be executed in the new object, passing a reference to the old object as the first paramter.
In a class hierarchy copy() methods are called in the same order as constructor() methods.
Not all built-in classes can be copied. Classes not supporting copying will throw an exception when the copy() methods are called. See the documentation for each class for more information.
A thread is an independent sequence of execution of Qore code within a Qore program or script. Each thread has a thread ID or TID.
The first thread of execution in a Qore program has TID 1. TID 0 is always reserved for the special signal handler thread.
The Qore language is designed to be thread-safe and Qore programs should not crash the Qore executable due to threading errors. Threading errors should only cause exceptions to be thrown or application errors to occur.
Threading functionality in Qore is provided by the operating system's POSIX threads library.
New threads are created with the background operator. This operator executes the expression given as an argument in a new thread and returns the TID of the new thread to the calling thread. This is most useful for calling user subroutines or object methods designed to run in a separate thread.
To terminate a thread, the thread_exit statement should be called, as calling the exit() function will terminate the entire UNIX process (and therefore all threads) immediately.
All global variables are shared in Qore programs, while local variables (declared with my) are generally local to each thread (and thus accessed without any mutual-exclusion locking), regardless of location. This means that if a variable is declared with my at the top level, it will actually have global scope, but also each thread will have its own copy of the variable. In effect, declaring a top-level local variable with my actually creates a global thread-local variable.
The following code gives an example of declaring a global thread-local variable by using my at the top-level:
%require-our sub t() { printf("x=%n\n", $x); } my $x = 2; t(); background t();
This will print out:
x=2 x=<NOTHING>
Note that the second time the local variable is accessed in the background thread, it has no value.
Due to the way Qore's local variables work, it is illegal to declare a top-level local variable after first block is parsed in the program; that is; if any call to parse() or Program::parse() is made in an existing program (where a top-level block already exists), and an attempt to declare a new top-level local variable is made, then a ILLEGAL-TOP-LEVEL-LOCAL-VARIABLE parse exception will be raised.
Access to global variables in qore is wrapped in mutual-exclusion locks to guarantee safe access to global variable data in a multithreaded context. Local variables are thread-local and therefore not locked, except when referenced in a closure expression, in which case the local variable's scope is extended to that of the closure's, and all accesses to the bound local variable are made within mutual-exclusion locks as these variables may be used in multithreaded contexts.
An alternative to global thread-local variables is offered by the save_thread_data() and get_thread_data() functions (documented in Threading Functions).
The synchronized keyword can be used before subroutine or class method definitions in order to guarantee that the function or method call will only be executed in one thread at a time. As in Java, this keyword can also be used safely with recursive functions and methods (internally a Gate-like object is used to guarantee thread-exclusivity and allow recursion).
The following classes are useful when developing multi-threaded Qore programs:
Table 2.94. Classes Useful With Threading
Class | Description |
|---|---|
A mutual-exclusion thread lock. | |
A recursive thread lock. | |
A read-write thread lock. | |
Allows Qore programs to block until a certain condition becomes true. | |
A blocking counter class. | |
A thread-safe, blocking queue class (useful for message passing). | |
DEPRECATED in favor of the Gate class: A recursive mutual-exclusion thread lock. | |
A simple, thread-atomic sequence object (increment-only). | |
A helper class to automatically release Mutex locks when the AutoLock object is deleted. | |
A helper class to automatically exit Gate locks when the AutoGate object is deleted. | |
A helper class to automatically release read locks when the AutoReadLock object is deleted. | |
A helper class to automatically release read locks when the AutoWriteLock object is deleted. |
The following functions assist writing safe and efficient multi-threaded Qore programs:
Table 2.95. Thread Functions
Function | Description |
|---|---|
Saves a thread-local value against a key. | |
Retrieves the entire thread-local hash. | |
Retrieves a thread-local value based on a key. | |
Deletes the entire thread-local data hash. | |
Delete the value of a key in the thread-local data hash. | |
Gets the thread's TID (thread identifier) | |
Returns a list of TIDs of running threads | |
Returns the number of running threads |
Qore supports deadlock detection in complex locking scenarios and will throw a THREAD-DEADLOCK exception rather than allow an operation to be performed that would cause a deadlock. Deadlock detection is implemented for internal locking (global variable and object access), synchronized methods and subroutinges, etc, as well as for all Qore threading classes.
Qore can only detect deadlocks when a lock resource acquired by one thread is required by another who holds a lock that the first thread also needs. Other errors such as forgetting to unlock a global lock and trying to acquire that lock in another thread cannot be differentiated from valid use of threading primitives and will result in a process that never terminates. However, common threading errors such as trying to lock the same Mutex twice in the same thread without unlocking it between the two Mutex::lock() calls are caught in Qore and exceptions are thrown. Additionally, locks are tracked as thread resources, so if a thread terminates while holding a lock, an exception will be thrown and the lock will be automatically released.
Exceptions are errors that can only be handled using a try catch block. Any exception that is thrown in a try block will immediately cause execution of that thread to begin with the first statement of the catch block, regardless of the position of the program pointer of the running thread, even if nested subroutines or object method calls have been made.
Exceptions can be thrown by the Qore system for a number of reasons, see the documentation for each function and object method for details.
Programmers can also throw exceptions explicitly by using the throw and rethrow statements.
Information about the exception, including the context in which the exception occurred, is saved in the exception hash, which can be retrieved by using a parameter variable in the catch block (for more information about try catch blocks, see try and catch statements).
The exception hash contains the following members:
Table 2.96. Exception Hash Keys
Name | Type | Description |
|---|---|---|
| string | "System" or "User" depending on exception type |
| string | File name of file where exception occurred |
| integer | Line number where exception occurred |
| list of hashes | Backtrace information |
| any | This key is populated with the value of the first expression of the throw statement. For system exceptions, this is a string giving the exception code. |
| any | This key is populated with the value of the second expression of the throw statement (if a list was thrown). For system exceptions, this is a string giving a text description of the error. |
| any | This key is populated with the value of the third expression of the throw statement (if a list was thrown). For system exceptions, this is populated for some exceptions where additional information is provided. |
Table 2.97. Call Stack Description
Name | Type | Description |
|---|---|---|
| string | function name |
| integer | line number |
| string | file name |
| string | Exception Type (ET_*) constants; see Exception Constants for values. |
| integer | Call Type (CT_*) constants; see Exception Constants for values. |
System exceptions always throw 2 values, populating the "err" and "desc" keys of the exception hash, giving the exception string code and the exception description string, respectively, and occassionally, depending on the function, the "arg" key may be populated with supporting information. User exceptions have no restrictions, any values given in the throw statement will be mapped to exception keys as per the table above.
See the on_exit, on_success statement, and on_error statement for statements that allow for exception-safe and exception-dependent cleanup in Qore code.
Classes that assist in exception-safe lock handling are the AutoLock class, the AutoGate class, the AutoReadLock class, and the AutoWriteLock class.
XML functionality in Qore is provided by the libxml2 library, which provides a powerful, stable, clean, and thread-safe basis for XML integration in Qore.
XML provides an excellent way to describe hierarchical data, and thanks to libxml2, Qore can allow for easy serialization and deserialization between XML strings and Qore data structures.
XML serialization (conversion from Qore data structures to XML strings) in Qore relies on the fact that Qore hashes retain insertion order, which means that conversion to and from Qore data structures and XML strings can be done without data loss and without reordering the XML elements. In general, XML serialization is relatively straighforward, but there are a few issues to be aware of, particularly regarding element attributes and lists. These issues are described in the following paragraphs.
First, a straightforward example:
$h = ( "record" : ( "name" : ( "first" : "Fred", "last" : "Smith" ) ) );
printf("%s\n", makeFormattedXMLString($h));This produces the following result
<?xml version="1.0" encoding="UTF-8"?>
<record>
<name>
<first>Fred</first>
<last>Smith</last>
</name>
</record>To set XML attributes, the Qore value must be a hash and the attributes are stored in another hash in the key ^attributes^. That is; the value of the ^attributes^ key must be a hash, and each member of this hash will represent an attribute-value pair.
For example:
$h = ( "record" : ( "^attributes^" : ( "type" : "customer" ) ,
"name" : ( "first" : "Fred", "last" : "Smith" ) ) );
printf("%s\n", makeFormattedXMLString($h));This produces the following results:
<?xml version="1.0" encoding="UTF-8"?>
<record type="customer">
<name>
<first>Fred</first>
<last>Smith</last>
</name>
</record>If instead we wanted to have text instead of child data under the "record" node, we must set the ^value^ key of the hash along with the ^attributes^ key as follows:
$h = ( "record" : ( "^attributes^" : ( "type" : "customer" ) ,
"^value^" : "NO-RECORD" ) );
printf("%s\n", makeFormattedXMLString($h));Giving the following results:
<?xml version="1.0" encoding="UTF-8"?> <record type="customer">NO-RECORD</record>
Arrays are serialized with repeating node names as follows:
$h = ( "record" : ( "part" : ( "part-02-05", "part-99-23", "part-34-28" ) ) );
printf("%s\n", makeFormattedXMLString($h));Producing the following results:
<?xml version="1.0" encoding="UTF-8"?> <record type="customer"> <part>part-02-05</part> <part>part-99-23</part> <part>part-34-28</part> </record>
It gets a little trickier when a key should repeated at the same level in an XML string, but other keys come between, for example, take the following XML string:
<?xml version="1.0" encoding="UTF-8"?> <para>Keywords: <code>this</code>, <code>that</code>, and <code>the_other</code>.</para>
It's not possible to use a list, because text is required in between. As described earlier, the ^value^ hash key can be used to serialize text in an XML string. In this case, we need to have several text nodes and several code nodes in a mixed-up order to give us the XML string we want. Because qore hases have unique keys (we can't use the same key twice in the same hash), we resort to a key naming trick that allows us to virtually duplicate our key names and therefore arrive at the XML string we want. We do this by appending a '^' character to the end of the key name and then some unique text. When serializing hash keys, any text after (and including) the '^' character is ignored. For the special key name ^value^, we do not need to duplicate the final '^' character. Instead we just add unique text to ensure that our hash can contain all the data we want and that it will be serialized in the right order to the XML string as follows:
$h = ( "para" : ( "^value^" : "Keywords: ",
"code" : "this",
"^value^1" : ", ",
"code^1" : "that",
"^value^2" : ", and ",
"code^2" : "the_other",
"^value^3" : "." ) );
printf("%s\n", makeFormattedXMLString($h));By ignoring the text after the '^' character, the above code will serialize to the XML string we want. In general, by using this convention, we can properly serialize multiple out-of-order keys without losing data and still have unique names for our hash keys.
Note than when deserializing XML strings to Qore data structures, the above rules are applied in reverse. If any out-of-order duplicate keys are detected, Qore will automatically generate unique hash key names based on the above rules.
Also note that CDATA text will be generated if a hash key starts with '^cdata'; such text will not be processed for escape code substitution. When deserializing XML strings to qore data structures, CDATA text will be placed unmodified under such a hash key as well.
Table 2.98. Functions For XML Serialization and Deserialization
Function Name | Description |
|---|---|
Serializes a hash into an XML string with formatting without an XML header. | |
Serializes a hash into an XML string with formatting and an XML header. | |
Serializes a hash into an XML string without an XML header or formatting. | |
Serializes a hash into a complete XML string with an XML header and without formatting. | |
parses an XML string as data (duplicate, out-of-order XML elements are collapsed into lists) and returns a Qore hash structure. | |
parses an XML string as data (duplicate, out-of-order XML elements are collapsed into lists) and validates against an XSD schema string and returns a Qore hash structure. | |
parses an XML string (XML element order is preserved by appending numeric suffixes to Qore hash key names when necessary) and returns a Qore hash structure. | |
parses an XML string (XML element order is preserved by appending numeric suffixes to Qore hash key names when necessary) and validates against an XSD schema string and returns a Qore hash structure. |
XML-RPC is a lightweight but powerful XML over HTTP web service protocol. Qore includes builtin support for this protocol. You can find more information about XML-RPC, including specifications and examples at http://xmlrpc.org.
Information about Qore's XML-RPC serialization can be found below.
Table 2.100. XML-RPC Serialization in Qore
Qore Type | XML-RPC Type | Description | |
|---|---|---|---|
| direct conversion to UTF-8 string | ||
| If the integer requires more than 32 bits to represent, then it is sent as a string | ||
| direct conversion | ||
| direct conversion | ||
| Absolute date/time values will convert to the default time zone for the calling context for the output string if necessary. Note that relative date/time values (durations) will be serialized with the same format as absolute date/time values. | ||
| direct conversion | ||
| direct conversion | ||
| direct conversion | ||
all others | n/a | All other types will cause an |
Table 2.101. Functions Providing XML-RPC Functionality
Function Name | Description |
|---|---|
Serializes a hash into an XML string formatted for an XML-RPC call with formatting. | |
Serializes a hash into an XML string formatted for an XML-RPC call with formatting, taking a single list argument for the argument list. | |
Serializes a hash into an XML string formatted for an XML-RPC fault response with formatting. | |
Serializes a hash into an XML string formatted for an XML-RPC response with formatting. | |
Serializes a hash into an XML string in XML-RPC value format with formatting. | |
Serializes a hash into an XML string formatted for an XML-RPC call without formatting. | |
Serializes a hash into an XML string formatted for an XML-RPC call without formatting, taking a single list argument for the argument list. | |
Serializes a hash into an XML string formatted for an XML-RPC fault response without formatting. | |
Serializes a hash into an XML string formatted for an XML-RPC response without formatting. | |
Serializes a hash into an XML string in XML-RPC value format without formatting. | |
deserializies an XML-RPC call string, returning a Qore hash respresenting the call information. | |
deserializies an XML-RPC response string, returning a Qore hash respresenting the response information. | |
deserializies an XML-RPC value tree, returning a Qore hash respresenting the information. |
Qore implements safe signal handling. Signals do not interrupt Qore threads, rather Qore uses a special signal handling thread with TID 0, dedicated to handling signals. The signal handling thread uses very few resources; it stays blocked (using no processor time and very little memory) until a signal with a Qore signal handler is raised; it then executes the handler and resumes waiting for signals.
Because the signal Qore's signal handling thread is not a normal thread, it does not affect num_threads() and does not appear in the list returned by thread_list().
Internally, Qore masks (blocks) all signals in every thread except the signal handling thread. In the signal handling thread, all signals are unmasked, except those with Qore-language handlers, then an internal call to sigwait() is made to receive and process signals raised one at a time.
Qore-language signal handlers are installed by passing a signal constant and a closure or call reference to the code to execute when the signal is raised to the set_signal_handler() function. Signal handlers are removed by passing a signal constant to the remove_signal_handler() function.
When a signal has been raised and the signal handler is called, the signal number is passed as the sole argument to the signal handler code.
Table 2.102. Signal Handling Functions
Function Name | Description |
|---|---|
Sets up a Qore signal handler using a signal number and a call reference. | |
Removes a Qore signal handler using a signal number. |
See Signal Constants for a list of signal constants and Signal Mapping Hash Constants for two hash constants that can be used to map signal names to numbers and vice-versa. Note that signal constants are system-dependent; not all signals will be available in all systems; in case of doubt, see your system documentation for information on which signals are available.
The above functions are atomic, meaning that when they return to the caller, the signal handling thread has already acknowledged the changes.
It is not possible to set signal masks per thread; all signals are delivered to the signal handling thread. Signals not handled with a Qore signal handler are handled with their default action. It is not possible to catch SIGPIPE. SIGPIPE is always ignored in Qore.
Some issues to be aware of in signal handlers:
Thread-local storage is not persistent in signal handlers; it is deleted after every signal handler is run.
A signal handler that does not terminate will block the execution of further signal handlers and will block signal handling changes (such as updating the signal mask), resulting in a Qore process that must be killed manually. Because all Qore signal handling code is executed serially in a single thread, Qore signal handlers should execute and return quickly to give time to execute other handlers.
Signal handlers may install or remove signal handlers using set_signal_handler() or remove_signal_handler(), however in this case, changes to signal handling are made after the signal handler returns.
Signal handlers cannot call fork(); any attempt to call fork() in a signal handler will result in an an exception.
fork() (called externally to a signal handler) is handled as follows: the signal handling thread is terminated, fork() is executed, all signals are masked in the primary thread in the new process, then the signal handling thread is resumed in both processes. No signals can be received or handled while the signal handling thread is terminated. After the fork(), the new process will have exactly the same signal handlers and signal masks as the parent process.
Unhandled exceptions in signal handlers will simply be displayed on stderr as an unhandled exception and will have no other effect on Qore or Qore code (in particular, unhandled exceptions will not cause the signal handling thread to terminate).
If a signal handler executes the thread_exit statement, execution of the signal handler will terminate immediately, but the signal handling thread will not be stopped. Execution of further signal handlers (including that for the same signal being handled when thread_exit is executed) will not be affected.
Qore supports a simple event-handling mechanism to provide notification and details of socket and network events in higher-level classes. Classes currently supporting events are the Socket, HTTPClient, and FtpClient classes.
See Event Constants for a list of all event constants; details about each event are documented in the following sections.
Event information is placed on the event queue (which must be a Queue object) in the form of a hash. Each event has at least the following keys:
Table 2.103. Event Hash Common Keys
Key | Value |
|---|---|
| This key holds the event code; see information for individual events in the following sections |
| This key holds the event source code |
| The value of this key is a unique integer that can be used to uniquely identify the object generating the event. |
EVENT_PACKET_READ
SOURCE_SOCKET
This event is raised immediately after a network packet is received. The event hash contains the following keys:
Table 2.104. EVENT_PACKET_READ Event Hash
Key | Value |
|---|---|
|
|
|
|
| A unique integer ID for the socket object. |
| The number of bytes read in the packet. |
| The total number of bytes read in the read loop. |
[ | The total number of bytes to read in the read loop (this key is only present if the total number of bytes to read is known). |
EVENT_PACKET_SENT
SOURCE_SOCKET
This event is raised immediately after a network packet is sent. The event hash contains the following keys:
Table 2.105. EVENT_PACKET_SENT Event Hash
Key | Value |
|---|---|
|
|
|
|
| A unique integer ID for the socket object. |
| The file descriptor number of the socket. |
| The number of bytes sent in the packet. |
| The total number of bytes sent in the send loop. |
| The total number of bytes to send in the send loop. |
EVENT_HTTP_CONTENT_LENGTH
SOURCE_HTTPCLIENT
This event is raised immediately after an HTTP header is received containing a content length header line, but before the message body is received. The event hash contains the following keys:
Table 2.106. EVENT_HTTP_CONTENT_LENGTH Event Hash
Key | Value |
|---|---|
|
|
|
|
| A unique integer ID for the socket object. |
| The number of bytes given for the content length. |
EVENT_HTTP_CHUNKED_START
SOURCE_HTTPCLIENT
This event is raised after receiving an HTTP header with Transfer-Encoding set to chunked and before the chunked data is read. The event hash contains the following keys:
Table 2.107. EVENT_HTTP_CHUNKED_START Event Hash
Key | Value |
|---|---|
|
|
|
|
| A unique integer ID for the socket object. |
EVENT_HTTP_CHUNKED_END
SOURCE_HTTPCLIENT
This event is raised after all chunked data is read from the socket. The event hash contains the following keys:
Table 2.108. EVENT_HTTP_CHUNKED_END Event Hash
Key | Value |
|---|---|
|
|
|
|
| A unique integer ID for the socket object. |
EVENT_HTTP_REDIRECT
SOURCE_HTTPCLIENT
This event is raised after a redirect response is received from an HTTP server. The event hash contains the following keys:
Table 2.109. EVENT_HTTP_REDIRECT Event Hash
Key | Value |
|---|---|
|
|
|
|
| A unique integer ID for the socket object. |
| The redirect location given by the HTTP server |
[ | Any status message sent by the HTTP server; if no message was sent, then this key will not be present in the event hash. |
EVENT_CHANNEL_CLOSED
SOURCE_SOCKET
This event is raised immediately after the socket is closed. The event hash contains the following keys:
Table 2.110. EVENT_CHANNEL_CLOSED Event Hash
Key | Value |
|---|---|
|
|
|
|
| A unique integer ID for the socket object. |
EVENT_DELETED
SOURCE_SOCKET
This event is raised when the socket object is deleted. The event hash contains the following keys:
Table 2.111. EVENT_DELETED Event Hash
Key | Value |
|---|---|
|
|
|
|
| A unique integer ID for the socket object. |
EVENT_FTP_SEND_MESSAGE
SOURCE_FTPCLIENT
This event is raised immediately before a message is sent on the FTP control channel. The event hash contains the following keys:
Table 2.112. EVENT_FTP_SEND_MESSAGE Event Hash
Key | Value |
|---|---|
|
|
|
|
| A unique integer ID for the socket object. |
| A string giving the FTP command sent (ex: |
[ | The argument to the command; if no argument is sent, then this key will not be present. |
EVENT_FTP_MESSAGE_RECEIVED
SOURCE_FTPCLIENT
This event is raised immediately after a message is received on the FTP control channel. The event hash contains the following keys:
Table 2.113. EVENT_FTP_MESSAGE_RECEIVED Event Hash
Key | Value |
|---|---|
|
|
|
|
| A unique integer ID for the socket object. |
| A string giving the FTP command sent (ex: |
[ | The argument to the command; if no argument is sent, then this key will not be present. |
EVENT_HOSTNAME_LOOKUP
SOURCE_SOCKET
This event is raised immediately before a hostname lookup is made. The event hash contains the following keys:
Table 2.114. EVENT_HOSTNAME_LOOKUP Event Hash
Key | Value |
|---|---|
|
|
|
|
| A unique integer ID for the socket object. |
| A string giving the name to be looked up. |
EVENT_HOSTNAME_RESOLVED
SOURCE_SOCKET
This event is raised immediately after a successful hostname resolution. The event hash contains the following keys:
Table 2.115. EVENT_HOSTNAME_RESOLVED Event Hash
Key | Value |
|---|---|
|
|
|
|
| A unique integer ID for the socket object. |
| A string giving the network address the name was resolved to. |
EVENT_HTTP_SEND_MESSAGE
SOURCE_HTTPCLIENT or SOURCE_SOCKET
This event is raised immediately before an HTTP message is sent. The event hash contains the following keys:
Table 2.116. EVENT_HTTP_SEND_MESSAGE Event Hash
Key | Value |
|---|---|
|
|
|
|
| The first string in the HTTP message (ex: |
| A hash of all headers to send in the message. |
EVENT_HTTP_MESSAGE_RECEIVED
SOURCE_HTTPCLIENT or SOURCE_SOCKET
This event is raised immediately after an HTTP message is received. The event hash contains the following keys:
Table 2.117. EVENT_HTTP_MESSAGE_RECEIVED Event Hash
Key | Value |
|---|---|
|
|
|
|
| A hash of all headers received in the message, plus the following headers giving additional information about the message: |
EVENT_HTTP_FOOTERS_RECEIVED
SOURCE_HTTPCLIENT
This event is raised immediately after HTTP footers are received after receiving chunked data. The event hash contains the following keys:
Table 2.118. EVENT_HTTP_FOOTERS_RECEIVED Event Hash
Key | Value |
|---|---|
|
|
|
|
| A hash of all footers received in the message. |
EVENT_HTTP_CHUNKED_DATA_RECEIVED
SOURCE_HTTPCLIENT
This event is raised immediately after chunked data is received. The event hash contains the following keys:
Table 2.119. EVENT_HTTP_CHUNKED_DATA_RECEIVED Event Hash
Key | Value |
|---|---|
|
|
|
|
| An integer giving the number of bytes read in the chunk. |
| An integer giving the total number of bytes of chunked data read in the current message. |
EVENT_HTTP_CHUNK_SIZE
SOURCE_HTTPCLIENT
This event is raised immediately after chunk information is received providing the size of the next chunk. The event hash contains the following keys:
Table 2.120. EVENT_HTTP_CHUNK_SIZE_RECEIVED Event Hash
Key | Value |
|---|---|
|
|
|
|
| An integer giving the number of bytes in the next chunk. |
| An integer giving the total number of bytes of chunked data read in the current message. |
EVENT_CONNECTING
SOURCE_SOCKET
This event is raised immediately before a socket connection is attempted. The event hash contains the following keys:
Table 2.121. EVENT_CONNECTING Event Hash
Key | Value |
|---|---|
|
|
|
|
| A unique integer ID for the socket object. |
| The type of address for the socket; one of the Network Address Constants. |
| The target address for the connection. |
[ | The target port for the connection; if not applicable for the address family then this hash key is not included. |
EVENT_CONNECTED
SOURCE_SOCKET
This event is raised immediately after a socket connection is established. The event hash contains the following keys:
Table 2.122. EVENT_CONNECTED Event Hash
Key | Value |
|---|---|
|
|
|
|
| A unique integer ID for the socket object. |
EVENT_START_SSL
SOURCE_SOCKET
This event is raised immediately before SSL negotiation is attempted. The event hash contains the following keys:
Table 2.123. EVENT_START_SSL Event Hash
Key | Value |
|---|---|
|
|
|
|
| A unique integer ID for the socket object. |
EVENT_SSL_ESTABLISHED
SOURCE_SOCKET
This event is raised immediately after SSL negotiation has been successfully established. The event hash contains the following keys:
Table 2.124. EVENT_SSL_ESTABLISHED Event Hash
Key | Value |
|---|---|
|
|
|
|
| A unique integer ID for the socket object. |
| A string giving the name of the cipher algorithm used for the connection. |
| A string giving the version of the cipher algorithm used for the connection. |
EVENT_OPEN_FILE
SOURCE_FILE
This event is raised immediately before a file is opened. The event hash contains the following keys:
Table 2.125. EVENT_OPEN_FILE Event Hash
Key | Value |
|---|---|
|
|
|
|
| A unique integer ID for the file object. |
| The file's name. |
| The flags used to open the file. |
| The mode to open the file with. |
| The character encoding given used for reading from or writing to the file. |
EVENT_FILE_OPENED
SOURCE_FILE
This event is raised immediately after a file has been successfully opened. The event hash contains the following keys:
Table 2.126. EVENT_FILE_OPENED Event Hash
Key | Value |
|---|---|
|
|
|
|
| A unique integer ID for the file object. |
| The file's name. |
| The flags used to open the file. |
| The mode to open the file with. |
| The character encoding given used for reading from or writing to the file. |