Add PHP support (#149)

.travis.yml

@@ -9,7 +9,7 @@ python:
 env:
   - TEST=API
   - TEST=E2E LANG="c or python or java or go or javascript or r_lang"
-  - TEST=E2E LANG="c_sharp or visual_basic or powershell"
+  - TEST=E2E LANG="c_sharp or visual_basic or powershell or php"
 
 before_install:
   - bash .travis/setup.sh

.travis/setup.sh

@@ -23,3 +23,9 @@ if [[ $LANG == *"r_lang"* ]]; then
   sudo apt-get update
   sudo apt-get install --no-install-recommends -y r-base
 fi
+
+# Install PHP.
+if [[ $LANG == *"php"* ]]; then
+  sudo apt-get update
+  sudo apt-get install --no-install-recommends -y php
+fi

Dockerfile

@@ -19,7 +19,8 @@ RUN apt-get update && \
         golang-go \
         dotnet-sdk-3.0 \
         powershell \
-        r-base && \
+        r-base \
+        php && \
     rm -rf /var/lib/apt/lists/*
 
 WORKDIR /m2cgen

README.md

@@ -6,7 +6,7 @@
 [![Python Versions](https://img.shields.io/pypi/pyversions/m2cgen.svg?logo=python&logoColor=white)](https://pypi.org/project/m2cgen)
 [![PyPI Version](https://img.shields.io/pypi/v/m2cgen.svg?logo=pypi&logoColor=white)](https://pypi.org/project/m2cgen)
 
-**m2cgen** (Model 2 Code Generator) - is a lightweight library which provides an easy way to transpile trained statistical models into a native code (Python, C, Java, Go, JavaScript, Visual Basic, C#, PowerShell, R).
+**m2cgen** (Model 2 Code Generator) - is a lightweight library which provides an easy way to transpile trained statistical models into a native code (Python, C, Java, Go, JavaScript, Visual Basic, C#, PowerShell, R, PHP).
 
 * [Installation](#installation)
 * [Supported Languages](#supported-languages)
@@ -30,6 +30,7 @@ pip install m2cgen
 - Go
 - Java
 - JavaScript
+- PHP
 - PowerShell
 - Python
 - R

m2cgen/__init__.py

@@ -10,6 +10,7 @@
     export_to_c_sharp,
     export_to_powershell,
     export_to_r,
+    export_to_php,
 )
 
 __all__ = [
@@ -22,6 +23,7 @@
     export_to_c_sharp,
     export_to_powershell,
     export_to_r,
+    export_to_php,
 ]
 
 with open(os.path.join(os.path.dirname(os.path.realpath(__file__)),

m2cgen/cli.py

@@ -28,6 +28,7 @@
                 ["indent", "class_name", "namespace"]),
     "powershell": (m2cgen.export_to_powershell, ["indent"]),
     "r": (m2cgen.export_to_r, ["indent"]),
+    "php": (m2cgen.export_to_php, ["indent"]),
 }
 
 

m2cgen/exporters.py

@@ -232,6 +232,25 @@ def export_to_r(model, indent=4):
     return _export(model, interpreter)
 
 
+def export_to_php(model, indent=4):
+    """
+    Generates a PHP code representation of the given model.
+
+    Parameters
+    ----------
+    model : object
+        The model object that should be transpiled into code.
+    indent : int, optional
+        The size of indents in the generated code.
+
+    Returns
+    -------
+    code : string
+    """
+    interpreter = interpreters.PhpInterpreter(indent=indent)
+    return _export(model, interpreter)
+
+
 def _export(model, interpreter):
     assembler_cls = assemblers.get_assembler_cls(model)
     model_ast = assembler_cls(model).assemble()

m2cgen/interpreters/__init__.py

@@ -7,6 +7,7 @@
 from .c_sharp.interpreter import CSharpInterpreter
 from .powershell.interpreter import PowershellInterpreter
 from .r.interpreter import RInterpreter
+from .php.interpreter import PhpInterpreter
 
 __all__ = [
     JavaInterpreter,
@@ -18,4 +19,5 @@
     CSharpInterpreter,
     PowershellInterpreter,
     RInterpreter,
+    PhpInterpreter,
 ]

m2cgen/interpreters/php/code_generator.py

@@ -0,0 +1,50 @@
+import contextlib
+
+from m2cgen.ast import CompOpType
+from m2cgen.interpreters.code_generator import CLikeCodeGenerator
+from m2cgen.interpreters.code_generator import CodeTemplate
+
+
+class PhpCodeGenerator(CLikeCodeGenerator):
+
+    tpl_array_index_access = CodeTemplate("$$${array_name}[${index}]")
+
+    def __init__(self, *args, **kwargs):
+        super(PhpCodeGenerator, self).__init__(*args, **kwargs)
+
+    def add_function_def(self, name, args):
+        function_def = "function " + name + "("
+        function_def += ", ".join([
+            ("array " if is_vector else "") + "$" + n
+            for is_vector, n in args])
+        function_def += ") {"
+        self.add_code_line(function_def)
+        self.increase_indent()
+
+    @contextlib.contextmanager
+    def function_definition(self, name, args):
+        self.add_function_def(name, args)
+        yield
+        self.add_block_termination()
+
+    def get_var_name(self):
+        return "$" + super().get_var_name()
+
+    def add_var_declaration(self, size):
+        var_name = self.get_var_name()
+        self.add_var_assignment(
+            var_name=var_name,
+            value="array()" if size > 1 else "null",
+            value_size=size)
+        return var_name
+
+    def vector_init(self, values):
+        return "array(" + ", ".join(values) + ")"
+
+    def _comp_op_overwrite(self, op):
+        if op == CompOpType.EQ:
+            return "==="
+        elif op == CompOpType.NOT_EQ:
+            return "!=="
+        else:
+            return op.value

m2cgen/interpreters/php/interpreter.py

@@ -0,0 +1,44 @@
+import os
+
+from m2cgen import ast
+from m2cgen.interpreters import utils, mixins
+from m2cgen.interpreters.php.code_generator import PhpCodeGenerator
+from m2cgen.interpreters.interpreter import ToCodeInterpreter
+
+
+class PhpInterpreter(ToCodeInterpreter, mixins.LinearAlgebraMixin):
+
+    supported_bin_vector_ops = {
+        ast.BinNumOpType.ADD: "add_vectors",
+    }
+
+    supported_bin_vector_num_ops = {
+        ast.BinNumOpType.MUL: "mul_vector_number",
+    }
+
+    exponent_function_name = "exp"
+    power_function_name = "pow"
+    tanh_function_name = "tanh"
+
+    def __init__(self, indent=4, *args, **kwargs):
+        cg = PhpCodeGenerator(indent=indent)
+        super(PhpInterpreter, self).__init__(cg, *args, **kwargs)
+
+    def interpret(self, expr):
+        self._cg.reset_state()
+        self._reset_reused_expr_cache()
+
+        with self._cg.function_definition(
+                name="score",
+                args=[(True, self._feature_array_name)]):
+            last_result = self._do_interpret(expr)
+            self._cg.add_return_statement(last_result)
+
+        if self.with_linear_algebra:
+            filename = os.path.join(
+                os.path.dirname(__file__), "linear_algebra.php")
+            self._cg.prepend_code_lines(utils.get_file_content(filename))
+
+        self._cg.prepend_code_line("<?php")
+
+        return self._cg.code

m2cgen/interpreters/php/linear_algebra.php

@@ -0,0 +1,14 @@
+function add_vectors(array $v1, array $v2) {
+    $result = array();
+    for ($i = 0; $i < count($v1); ++$i) {
+        $result[] = $v1[$i] + $v2[$i];
+    }
+    return $result;
+}
+function mul_vector_number(array $v1, $num) {
+    $result = array();
+    for ($i = 0; $i < count($v1); ++$i) {
+        $result[] = $v1[$i] * $num;
+    }
+    return $result;
+}

m2cgen/interpreters/powershell/code_generator.py

@@ -45,9 +45,7 @@ def math_function_invocation(self, function_name, *args):
         return function_name + "(" + ", ".join(map(str, args)) + ")"
 
     def get_var_name(self):
-        var_name = "$var" + str(self._var_idx)
-        self._var_idx += 1
-        return var_name
+        return "$" + super().get_var_name()
 
     def add_var_declaration(self, size):
         var_name = self.get_var_name()

tests/e2e/executors/__init__.py

@@ -7,6 +7,7 @@
 from tests.e2e.executors.c_sharp import CSharpExecutor
 from tests.e2e.executors.powershell import PowershellExecutor
 from tests.e2e.executors.r import RExecutor
+from tests.e2e.executors.php import PhpExecutor
 
 __all__ = [
     JavaExecutor,
@@ -18,4 +19,5 @@
     CSharpExecutor,
     PowershellExecutor,
     RExecutor,
+    PhpExecutor,
 ]

tests/e2e/executors/php.py

@@ -0,0 +1,71 @@
+import os
+import string
+
+from m2cgen import assemblers, interpreters
+from tests import utils
+from tests.e2e.executors import base
+
+EXECUTOR_CODE_TPL = """
+<?php
+$$input_array = array();
+for ($$i = 1; $$i < $$argc; ++$$i) {
+    $$input_array[] = floatval($$argv[$$i]);
+}
+
+require '${model_file}.php';
+
+$$res = score($$input_array);
+
+${print_code}
+"""
+
+EXECUTE_AND_PRINT_SCALAR = """
+echo($res);
+"""
+
+EXECUTE_AND_PRINT_VECTOR = """
+echo(implode(" ", $res));
+"""
+
+
+class PhpExecutor(base.BaseExecutor):
+
+    executor_name = "score"
+    model_name = "model"
+
+    def __init__(self, model):
+        self.model = model
+        self.interpreter = interpreters.PhpInterpreter()
+
+        assembler_cls = assemblers.get_assembler_cls(model)
+        self.model_ast = assembler_cls(model).assemble()
+
+        self._php = "php"
+
+    def predict(self, X):
+        file_name = os.path.join(self._resource_tmp_dir,
+                                 "{}.php".format(self.executor_name))
+        exec_args = [self._php,
+                     "-f",
+                     file_name,
+                     *map(str, X)]
+        return utils.predict_from_commandline(exec_args)
+
+    def prepare(self):
+        if self.model_ast.output_size > 1:
+            print_code = EXECUTE_AND_PRINT_VECTOR
+        else:
+            print_code = EXECUTE_AND_PRINT_SCALAR
+        executor_code = string.Template(EXECUTOR_CODE_TPL).substitute(
+            model_file=self.model_name,
+            print_code=print_code)
+        model_code = self.interpreter.interpret(self.model_ast)
+
+        executor_file_name = os.path.join(
+            self._resource_tmp_dir, "{}.php".format(self.executor_name))
+        model_file_name = os.path.join(
+            self._resource_tmp_dir, "{}.php".format(self.model_name))
+        with open(executor_file_name, "w") as f:
+            f.write(executor_code)
+        with open(model_file_name, "w") as f:
+            f.write(model_code)

tests/e2e/test_e2e.py

@@ -24,6 +24,7 @@
 C_SHARP = pytest.mark.c_sharp
 POWERSHELL = pytest.mark.powershell
 R = pytest.mark.r_lang
+PHP = pytest.mark.php
 REGRESSION = pytest.mark.regr
 CLASSIFICATION = pytest.mark.clf
 
@@ -116,6 +117,7 @@ def classification_binary_random(model):
         (executors.CSharpExecutor, C_SHARP),
         (executors.PowershellExecutor, POWERSHELL),
         (executors.RExecutor, R),
+        (executors.PhpExecutor, PHP),
     ],
 
     # These models will be tested against each language specified in the