Expressions =========== Node values can also be defined indirectly using expressions. Notation of expression values is similar to the string notation wrapped in additional parentheses: .. code-block:: DIPSchema :caption: Expression schema # using double quotes ("") # using single quotes ('') # using block notation (""" """) There are 4 different types of expressions in DIP. Three of them (logical, numerical and textual, i.e. templates) are described below. Dimensional expressions, i.e. :doc:`units `, are described in a separate chapter. .. note:: All operators used in expressions must be separated from values with at least one empty space. Possible input values used by expressions are summarized in the following table: .. list-table:: Input values :widths: 30 30 70 :header-rows: 1 * - Syntax - Name - Example * - - boolean - true/false * - - numerical - 24 cm * - - reference - {?energy} * - - expression - 12 cm == {?width} && true Operators summarized below are evaluated according to their priority from 1 to 4 and can be nested accordingly. Logical ------- Logical expressions are used to generate values for boolean nodes. The expressions can be composed of multiple nested logical operators that always return a true or false value. An example of a logical expression is given below. .. code-block:: DIP a bool = true b float = 23.43 cm c bool = (""" false || ({?b} == 23.43 cm || ~{?a}) && {?a} || ~!{?c} """) # Node 'c' will be evaluated as 'true' Expression operators evaluate only boolean values, either given directly or as a result of a nested expression. .. list-table:: Logical operators :widths: 35 20 100 :header-rows: 1 * - Syntax - Priority - Description * - A || B || C ... - 4 - Logical OR operator returns true if at least one expression (`A`, `B`, `C`, ...) is true * - A && B && C ... - 3 - Logical AND operator returns true if all expressions (`A`, `B`, `C`, ...) are true .. list-table:: Parenthesis operator :widths: 35 20 100 :header-rows: 1 * - Syntax - Priority - Description * - \() - 2 - Parentheses operator evaluates expression `A` in a separate thread and returns its value Numerical values with dimensions compared using comparison operations are automatically converted into same units. The result of such comparison is always a boolean value. .. note:: At the moment, it is possible to compare only scalar numerical values. Support for array comparison is planned to be provided in next versions of DIP. .. list-table:: Comparison operators :widths: 35 20 100 :header-rows: 1 * - Syntax - Priority - Description * - A == B - 1 - Equality operator returns true if `A` and `B` have same dimensions and numeraical value up to EQUAL_PRECISION * - A != B - 1 - Inequality operator returns true if `A` and `B` do not have same dimension or numerical value up to EQUAL_PRECISION * - A >= B - 1 - Greater or equal operator returns true if `A` is greater or equal (up to EQUAL_PRECISION) than `B` * - A <= B - 1 - Smaller or equal operator returns true if `A` is smaller or equal (up to EQUAL_PRECISION) than `B` * - A > B - 1 - Greather than operator returns true if `A` is greater than `B` * - A < B - 1 - Smaller than operator returns true if `A` is smaller than `B` .. list-table:: Single value operators :widths: 35 20 100 :header-rows: 1 * - Syntax - Priority - Description * - ~ - 1 - Negation operator returns true if value `A` is false * - ! - 1 - Definition operator returns true if node exists * - ~! - 1 - Non-definition operator returns true if node does not exist Numerical --------- Numerical expressions are used to generate values for numerical node types. If given, the expression result is automatically converted into node units. .. code-block:: DIP const c float = 299792458 m/s energy float = (""" 2 kg * pow({?const.c},2) """) eV # Node 'energy' will be parsed as 1.79751 eV .. note:: DIP does not aim to substitute advanced numerical programming languages. Numerical expressions in DIP are supposed to give a quick tool for generation of input values that can be easily derived from other parameters. Therefore, it implements only the most basic mathematical operations on scalar values. More advanced operations can be added in the future versions of DIP. Operators used in numerical expressions are summarized below: .. list-table:: Basic operations :widths: 35 20 100 :header-rows: 1 * - Syntax - Priority - Description * - A + B - 3 - Addition of two values of a same dimension * - A - B - 3 - Substraction of two values of a same dimension * - A * B - 2 - Multiplication of two values * - A / B - 2 - Division of two values Parentheses operators evaluate expressions in a separate thread and return its final value. Most of the following operators require, that the final value has no dimensions. .. list-table:: Parentheses operators :widths: 35 20 100 :header-rows: 1 * - Syntax - Priority - Description * - \() - 1 - Standard parenthesis operator * - exp() - 1 - Returns exponential value of a dimensionless expression. * - pow(,) - 1 - Returns first expression risen on a power of second dimensionless expression. * - ln() - 1 - Returns natural logarithmic value of a dimensionless expression. * - log10() - 1 - Returns common logarithmic value of a dimmensionless expression. * - sin() - 1 - Returns sine value of a dimensionless expression. * - cos() - 1 - Returns cosine value of a dimensionless expression. Templates --------- Templates are used to parse node values into a text form. All value types can be parsed into text using standard python formatting notation. .. code-block:: DIP id int = 345 name str = 'Tina' body weight float = 62.3 kg height float = 177 cm married bool = true person str = (""" ID: {{?id}:05d} Name: {{?name}} Weight: {{?body.weight}:.3e} Height: {{?body.height}:.2f} Married: {{?married}} \{not a reference} """) # Node 'person' will be parsed as: # # ID: 00345 # Name: Tina # Weight: 6.230e+01 # Height: 177.00 # Married: True String expressions interpret starting double curly brackets always as a reference. Single curly brackets are interpreted as a plain test. Basic syntax of parsing operators is described below: .. list-table:: Parsing operators :widths: 45 100 :header-rows: 1 * - Syntax - Description * - {} - Default reference of a node value. * - {:} - Formatted reference of a node value. Arrays can also be parsed using templates, however, without specifying a format. The format depends on the default Python string casting functions: .. code-block:: DIP name str = "Will Smith" widths float[2,3] = [[23.4,235.4,34],[1e10,2e23,5e20]] preview str = (""" Surname: {{?name}[5:]} Scalar: {{?widths}[1,1]:.2e} Array: {{?widths}[:,1:]} """) # Node 'preview' will be parsed as: # # Surname: Smith # Scalar: 2.00e+23 # Array: # [[2.354e+02 3.400e+01] # [2.000e+23 5.000e+20]]