org.python.core.PyArray Maven / Gradle / Ivy
Show all versions of jython-slim Show documentation
// Copyright (c) Corporation for National Research Initiatives
package org.python.core;
import java.io.ByteArrayOutputStream;
import java.io.DataInputStream;
import java.io.DataOutputStream;
import java.io.EOFException;
import java.io.IOException;
import java.io.InputStream;
import java.io.OutputStream;
import java.lang.ref.WeakReference;
import java.lang.reflect.Array;
import java.math.BigInteger;
import java.nio.ByteBuffer;
import org.python.core.PyArray.ItemType;
import org.python.core.buffer.BaseBuffer;
import org.python.core.buffer.SimpleBuffer;
import org.python.core.buffer.SimpleStringBuffer;
import org.python.core.util.ByteSwapper;
import org.python.core.util.StringUtil;
import org.python.expose.ExposedGet;
import org.python.expose.ExposedMethod;
import org.python.expose.ExposedNew;
import org.python.expose.ExposedType;
import org.python.expose.MethodType;
import org.python.modules.gc;
/**
* The type {@code array.array}. This is a wrapper around native Java arrays. Instances of
* {@code PyArray} are created either by Java functions or directly by the {@code jarray} module
* (q.v.).
*
* The range of possible element (item) types exceeds that in Python, since it allows for arbitrary
* Java classes. This extended behaviour is accessible from Python by supplying a Java type (class)
* to the constructor, where one might have used a single character type code. For example:
* >>> ax = array.array(BigDecimal, (BigDecimal(str(n)) for n in range(5)))
* >>> ax
* array(java.math.BigDecimal, [0, 1, 2, 3, 4])
* >>> type(ax[2])
* <type 'java.math.BigDecimal'>
*
*
*
* Supported item types
*
* typecode
* Python type
* Java type
* serialised size
* signed
*
*
* {@code b}
* {@code int}
* {@code byte}
* 1
*
*
* {@code B}
* {@code int}
* {@code byte}
* 1
* unsigned
*
*
* {@code h}
* {@code int}
* {@code short}
* 2
*
*
* {@code H}
* {@code int}
* {@code short}
* 2
* unsigned
*
*
* {@code i}
* {@code int}
* {@code int}
* 4
*
*
* {@code I}
* {@code long}
* {@code int}
* 4
* unsigned
*
*
* {@code l}
* {@code long}
* {@code long}
* 8
*
*
* {@code L}
* {@code long}
* {@code long}
* 8
* unsigned
*
*
* {@code f}
* {@code float}
* {@code float}
* 4
*
*
* {@code d}
* {@code float}
* {@code double}
* 8
*
*
* {@code c}
* {@code str}
* {@code byte}
* 1
* unsigned
*
*
* {@code u}
* {@code unicode}
* {@code int}
* 1
* unsigned
*
*
* {@code z}
* {@code bool}
* {@code boolean}
* 1
*
*
* Types shown as "unsigned" represent positive values encoded in the same number of bits as the
* equivalent signed type. The Java value serialised makes no distinction. The consumer of the
* stream has to know whether to make a signed or unsigned interpretation of the bits. When reading
* a stream, the type code declared for the destination array decides the interpretation of the
* bytes.
*/
@ExposedType(name = "array.array", base = PyObject.class)
public class PyArray extends PySequence implements Cloneable, BufferProtocol, Traverseproc {
public static final PyType TYPE = PyType.fromClass(PyArray.class);
/** The underlying Java array, a Java Array in practice. */
private Object data;
/** The Java class of elements in the {@code data} array. */
private Class itemClass;
/** Everything else we need to know about the type of elements in the {@code data} array. */
private ItemType itemType;
/**
* Mix in the mechanisms for manipulating the underlying array as this "delegate" object. Many
* operations on this {@code array.array} are actually operations on this delegate (an
* {@code AbstractArray} in practice), which in turn manipulates {@link #data}.
*/
private ArrayDelegate delegate;
/**
* Create a default {@code PyArray} of specific Python type (for sub-class use).
*
* @param subtype actual Python type
*/
public PyArray(PyType subtype) {
super(subtype);
}
/**
* Create a {@code PyArray} with the given array item type, specific item class and content.
*
* The primary specification is the {@link ItemType} parameter and if that is
* {@link ItemType#OBJECT} a specific Java class for the items.
*
* In a subtle twist, if {@code itemType =} {@link ItemType#OBJECT} but {@code itemClass} is one
* of the types used to implement the "single letter" type codes, the item type of the array
* will be the first signed type represented by that class, not {@code OBJECT}. This is to
* preserve a legacy behaviour of the {@code PyArray} constructors.
*
* @param subtype actual Python type
* @param itemType of the elements
* @param itemClass when {@code itemType =} {@link ItemType#OBJECT}
* @param data content array
*/
PyArray(PyType subtype, ItemType itemType, Class itemClass, Object data) {
this(subtype);
setElementType(itemType, itemClass);
setData(data);
}
/**
* Create a {@code PyArray} with the given array item type, specific item class and length, but
* zero content.
*
* Roughly equivalent to
* PyArray(itemType, itemClass, Array.newInstance(itemClass, n))
* But with {@code itemClass} for the new array deduced from {@code itemType} as
* explained in {@link PyArray#PyArray(ItemType, Class, Object)}.
*
* @param subtype actual Python type
* @param itemType of the elements
* @param itemClass when {@code itemType =} {@link ItemType#OBJECT}
* @param n length of content array to create
*/
PyArray(PyType subtype, ItemType itemType, Class itemClass, int n) {
this(subtype);
setElementType(itemType, itemClass);
setData(Array.newInstance(this.itemClass, n));
}
/**
* Create a {@code PyArray} with the given array item class and content initialised from a
* Python object (like an iterable).
*
* @param subtype actual Python type
* @param itemType of the elements
* @param itemClass when {@code itemType =} {@link ItemType#OBJECT}
* @param initial provider of initial contents
*/
PyArray(PyType subtype, ItemType itemType, Class itemClass, PyObject initial) {
this(subtype, itemType, itemClass, 0);
useInitial(initial);
}
/**
* Create a {@code PyArray} with the given array item class and content. If {@code itemClass} is
* one of the primitive types used to implement the "single letter" type codes, the type code of
* the array will be a signed zero of that item class.
*
* @param itemClass of elements in the array
* @param data content array
*/
public PyArray(Class itemClass, Object data) {
this(TYPE, ItemType.OBJECT, itemClass, data);
}
/**
* Create a {@code PyArray} with the given array item class and content initialised from a
* Python object (iterable).
*
* @param itemClass of elements in the array
* @param initial provider of initial contents
*/
public PyArray(Class itemClass, PyObject initial) {
this(TYPE, ItemType.OBJECT, itemClass, initial);
}
/**
* Create a {@code PyArray} with the given array item class and number of zero or {@code null}
* elements. If {@code itemClass} is one of the primitive types used to implement the "single
* letter" type codes, the type code of the array will be a signed zero of that item class.
*
* @param itemClass of elements in the array
* @param n number of (zero or {@code null}) elements
*/
public PyArray(Class itemClass, int n) {
this(TYPE, ItemType.OBJECT, itemClass, Array.newInstance(itemClass, n));
}
/**
* Create a {@code PyArray} as a copy of another.
*
* @param toCopy the other array
*/
public PyArray(PyArray toCopy) {
this(TYPE, toCopy.itemType, toCopy.itemClass, toCopy.delegate.copyArray());
}
/**
* Initialise this array from an {@link ItemType} (from a Python {@code array.array} type code
* character) and if that is {@link ItemType#OBJECT} a specific Java class for the items.
*
* If {@code itemType =} {@link ItemType#OBJECT} but {@code itemClass} is one of the types used
* to implement the "single letter" type codes, the item type of the array will be the first
* signed type represented by that class, not {@code OBJECT}. This is to preserve a legacy
* behaviour of the {@code PyArray} constructors.
*
* The way {@link #array_new(PyNewWrapper, boolean, PyType, PyObject[], String[]) array_new}
* works, and the constructors, is to create an instance with the almost parameterless
* {@link #PyArray(PyType)} with sub-type argument.
*
* This blank canvas needs to be inscribed with a consistent state by a call to this method and
* either {@link #setData(Object) setData} or {@link #useInitial(PyObject) useInitial}.
*
* @param itemType of the elements
* @param itemClass when {@code itemType =} {@link ItemType#OBJECT}
*/
private void setElementType(ItemType itemType, Class itemClass) {
if (itemType == ItemType.OBJECT) {
/*
* If itemClass is one of the types used to implement the "single letter" type codes,
* the item type of the array will be the first signed type represented by that class,
* not OBJECT. This is to preserve a legacy behaviour of the PyArray constructors.
*/
this.itemClass = itemClass;
this.itemType = ItemType.fromJavaClass(itemClass);
} else {
// itemType tells the whole story
this.itemType = itemType;
this.itemClass = itemType.itemClass;
}
}
/**
* Make a given object the storage for the array. Normally this is a Java array of type
* consistent with the element type. It will be manipulated by {@link #delegate}.
*
* @param data the storage (or {@code null} to to create at zero length consistent with type).
*/
private void setData(Object data) {
this.data = data != null ? data : Array.newInstance(itemClass, 0);
this.delegate = new ArrayDelegate();
}
/**
* Provide initial values to the internal storage array from one of several types in the broad
* categories of a byte string (which is treated as a machine representation of the data) or an
* iterable yielding values assignable to the elements. There is special treatment for typecode
* 'u', itemClass Unicode.
*
* @param initial source of values or {@code null}
*/
private void useInitial(PyObject initial) {
// If we do not yet have a representation array, provide one
if (this.data == null || this.delegate == null) {
setData(Array.newInstance(this.itemClass, 0));
}
// The initialiser may be omitted, or may validly be one of several types.
if (initial == null) {
// Fall through
} else if (initial instanceof PyList) {
fromlist(initial);
} else if (initial instanceof PyString && !(initial instanceof PyUnicode)) {
fromstring(initial.toString());
} else if (itemType == ItemType.UNICHAR) {
if (initial instanceof PyUnicode) {
extendArray((PyUnicode) initial);
} else {
extendUnicodeIter(initial);
}
} else {
extendInternal(initial);
}
}
@ExposedNew
static final PyObject array_new(PyNewWrapper new_, boolean init, PyType subtype,
PyObject[] args, String[] keywords) {
if (new_.for_type != subtype && keywords.length > 0) {
/*
* We're constructing as a base for a derived type (via PyDerived) and there are
* keywords. The effective args locally should not include the keywords.
*/
int argc = args.length - keywords.length;
PyObject[] justArgs = new PyObject[argc];
System.arraycopy(args, 0, justArgs, 0, argc);
args = justArgs;
}
// Create a 'blank canvas' of the appropriate concrete class.
PyArray self =
new_.for_type == subtype ? new PyArray(subtype) : new PyArrayDerived(subtype);
// Build the argument parser for this call
ArgParser ap = new ArgParser("array", args, Py.NoKeywords,
new String[] {"typecode", "initializer"}, 1);
ap.noKeywords();
// Retrieve the mandatory type code that determines the element itemClass
PyObject obj = ap.getPyObject(0);
if (obj instanceof PyString && !(obj instanceof PyUnicode)) {
if (obj.__len__() != 1) {
throw Py.TypeError("array() argument 1 must be char, not str");
}
char typecode = obj.toString().charAt(0);
self.setElementType(ItemType.fromTypecode(typecode), null);
} else if (obj instanceof PyJavaType) {
Class itemClass = ((PyJavaType) obj).getProxyType();
self.setElementType(ItemType.OBJECT, itemClass);
} else {
throw Py.TypeError(
"array() argument 1 must be char, not " + obj.getType().fastGetName());
}
// Fill the array from the second argument (if there is one)
self.useInitial(ap.getPyObject(1, null));
return self;
}
/**
* Create a {@code PyArray} with the given array type code and number of zero elements.
*
* @param typecode of elements in the array
* @param n number of (zero or {@code null}) elements
* @return created array
*/
public static PyArray zeros(int n, char typecode) {
return new PyArray(TYPE, ItemType.fromTypecode(typecode), null, n);
}
/**
* Create a {@code PyArray} with the given array item class and number of zero or {@code null}
* elements. If {@code itemClass} is one of the primitive types used to implement the "single
* letter" type codes, the type code of the array will be a signed zero of that item class.
*
* @param itemClass
* @param n number of (zero or {@code null}) elements
* @return created array
*/
public static PyArray zeros(int n, Class itemClass) {
return new PyArray(TYPE, ItemType.OBJECT, itemClass, n);
}
/**
* Create a {@code PyArray} with the given array item type and content initialised from a Python
* object (iterable).
*
* @param seq to suply content
* @param typecode
* @return created array
*/
public static PyArray array(PyObject seq, char typecode) {
return new PyArray(TYPE, ItemType.fromTypecode(typecode), null, seq);
}
public static Class array_class(Class type) {
return Array.newInstance(type, 0).getClass();
}
/**
* Create a {@code PyArray} storing {@code ctype} types and being initialised with {@code init}.
*
* @param init an initialiser for the array - can be {@code PyString} or {@code PySequence}
* (including {@code PyArray}) or iterable type.
* @param itemClass {@code Class} of the elements stored in the array.
* @return a new PyArray
*/
public static PyArray array(PyObject init, Class itemClass) {
return new PyArray(TYPE, ItemType.OBJECT, itemClass, init);
}
@ExposedMethod(type = MethodType.BINARY)
final PyObject array___ne__(PyObject o) {
return seq___ne__(o);
}
@ExposedMethod(type = MethodType.BINARY)
final PyObject array___eq__(PyObject o) {
return seq___eq__(o);
}
@Override
public int hashCode() {
return array___hash__();
}
@ExposedMethod
final int array___hash__() {
throw Py.TypeError(String.format("unhashable type: '%.200s'", getType().fastGetName()));
}
@ExposedMethod(type = MethodType.BINARY)
final PyObject array___lt__(PyObject o) {
return seq___lt__(o);
}
@ExposedMethod(type = MethodType.BINARY)
final PyObject array___le__(PyObject o) {
return seq___le__(o);
}
@ExposedMethod(type = MethodType.BINARY)
final PyObject array___gt__(PyObject o) {
return seq___gt__(o);
}
@ExposedMethod(type = MethodType.BINARY)
final PyObject array___ge__(PyObject o) {
return seq___ge__(o);
}
@ExposedMethod
final boolean array___contains__(PyObject o) {
return object___contains__(o);
}
@ExposedMethod
final void array___delitem__(PyObject index) {
seq___delitem__(index);
}
@ExposedMethod
final void array___setitem__(PyObject o, PyObject def) {
seq___setitem__(o, def);
}
@ExposedMethod
final PyObject array___getitem__(PyObject o) {
PyObject ret = seq___finditem__(o);
if (ret == null) {
throw Py.IndexError("index out of range: " + o);
}
return ret;
}
@ExposedMethod
final boolean array___nonzero__() {
return seq___nonzero__();
}
@ExposedMethod
public PyObject array___iter__() {
return seq___iter__();
}
@ExposedMethod(defaults = "null")
final PyObject array___getslice__(PyObject start, PyObject stop, PyObject step) {
return seq___getslice__(start, stop, step);
}
@ExposedMethod(defaults = "null")
final void array___setslice__(PyObject start, PyObject stop, PyObject step, PyObject value) {
seq___setslice__(start, stop, step, value);
}
@ExposedMethod(defaults = "null")
final void array___delslice__(PyObject start, PyObject stop, PyObject step) {
seq___delslice__(start, stop, step);
}
@Override
public PyObject __imul__(PyObject o) {
return array___imul__(o);
}
@ExposedMethod(type = MethodType.BINARY)
final PyObject array___imul__(PyObject o) {
if (!o.isIndex()) {
return null;
}
resizeCheck(); // Prohibited if exporting a buffer
if (delegate.getSize() > 0) {
int count = o.asIndex(Py.OverflowError);
if (count <= 0) {
delegate.clear();
return this;
}
Object copy = delegate.copyArray();
delegate.ensureCapacity(delegate.getSize() * count);
for (int i = 1; i < count; i++) {
delegate.appendArray(copy);
}
}
return this;
}
@Override
public PyObject __mul__(PyObject o) {
return array___mul__(o);
}
@ExposedMethod(type = MethodType.BINARY)
final PyObject array___mul__(PyObject o) {
if (!o.isIndex()) {
return null;
}
return repeat(o.asIndex(Py.OverflowError));
}
@Override
public PyObject __rmul__(PyObject o) {
return array___rmul__(o);
}
@ExposedMethod(type = MethodType.BINARY)
final PyObject array___rmul__(PyObject o) {
if (!o.isIndex()) {
return null;
}
return repeat(o.asIndex(Py.OverflowError));
}
@Override
public PyObject __iadd__(PyObject other) {
return array___iadd__(other);
}
@ExposedMethod(type = MethodType.BINARY)
final PyObject array___iadd__(PyObject other) {
try {
PyArray otherArr = arrayChecked(other);
resizeCheck(); // Prohibited if exporting a buffer
delegate.appendArray(otherArr.delegate.copyArray());
return this;
} catch (ClassCastException e) {
// other wasn't a PyArray
return null;
}
}
@Override
public PyObject __add__(PyObject other) {
return array___add__(other);
}
/**
* Adds (appends) two PyArrays together
*
* @param other a {@code PyArray} to be added to the instance
* @return the result of the addition as a new {@code PyArray} instance
*/
@ExposedMethod(type = MethodType.BINARY)
final PyObject array___add__(PyObject other) {
try {
PyArray otherArr = arrayChecked(other);
PyArray ret = new PyArray(this);
ret.delegate.appendArray(otherArr.delegate.copyArray());
return ret;
} catch (ClassCastException e) {
// other wasn't a PyArray
return null;
}
}
/**
* Check the other array is an array and is compatible for element type. Raise {@code TypeError}
* if not.
*
* @param otherObject supposed {@code PyArray}
* @return {@code other}
* @throws ClassCastException if {@code other} not {@code PyArray}
*/
private PyArray arrayChecked(PyObject otherObject) throws ClassCastException {
PyArray other = (PyArray) otherObject;
if (itemType == other.itemType) {
if (itemType != ItemType.OBJECT) {
return other;
} else if (itemClass.isAssignableFrom(other.itemClass)) {
return other;
}
}
throw Py.TypeError(String.format("bad argument types for built-in operation: (%s, %s)",
reprTypecode(), other.reprTypecode()));
}
/**
* Length of the array (as the number of elements, not a storage size).
*
* @return number of elements in the array
*/
@Override
public int __len__() {
return array___len__();
}
@ExposedMethod
final int array___len__() {
return delegate.getSize();
}
@Override
public PyObject __reduce__() {
return array___reduce__();
}
@ExposedMethod
final PyObject array___reduce__() {
PyObject dict = __findattr__("__dict__");
if (dict == null) {
dict = Py.None;
}
PyString typecode = Py.newString(getTypecode());
if (__len__() > 0) {
return new PyTuple(getType(), new PyTuple(typecode, Py.newString(tostring())), dict);
} else {
return new PyTuple(getType(), new PyTuple(typecode), dict);
}
}
@Override
public String toString() {
if (__len__() == 0) {
return String.format("array(%s)", reprTypecode());
}
String value;
switch (itemType) {
case CHAR:
value = PyString.encode_UnicodeEscape(tostring(), true);
break;
case UNICHAR:
value = (new PyUnicode(tounicode())).__repr__().toString();
break;
default:
value = tolist().toString();
}
return String.format("array(%s, %s)", reprTypecode(), value);
}
private String reprTypecode() {
if (itemType == ItemType.OBJECT) {
return getTypecode();
} else {
return "'" + getTypecode() + "'";
}
}
/**
*
* @param c target {@code Class} for the conversion
* @return Java object converted to required class type if possible.
*/
@Override
public Object __tojava__(Class c) {
boolean isArray = c.isArray();
Class componentType = c.getComponentType();
if (c == Object.class || (isArray && componentType.isAssignableFrom(itemClass))) {
if (delegate.capacity != delegate.size) {
// when unboxing, shrink the array first, otherwise incorrect results to Java
return delegate.copyArray();
} else {
return data;
}
}
// rebox: this array is made of primitives but converting to Object[]
if (isArray && componentType == Object.class) {
Object[] boxed = new Object[delegate.size];
for (int i = 0; i < delegate.size; i++) {
boxed[i] = Array.get(data, i);
}
return boxed;
}
if (c.isInstance(this)) {
return this;
}
return Py.NoConversion;
}
@ExposedMethod
public final void array_append(PyObject value) {
resizeCheck(); // Prohibited if exporting a buffer
appendUnchecked(value);
}
/**
* Append new value x to the end of the array.
*
* @param value item to be appended to the array
*/
public void append(PyObject value) {
resizeCheck(); // Prohibited if exporting a buffer
appendUnchecked(value);
}
/**
* Common helper method used internally to append a new value x to the end of the array:
* {@link #resizeCheck()} is not called, so the client must do so in advance.
*
* @param value item to be appended to the array
*/
private final void appendUnchecked(PyObject value) {
int afterLast = delegate.getSize();
delegate.makeInsertSpace(afterLast);
try {
pyset(afterLast, value);
} catch (PyException e) {
delegate.setSize(afterLast);
throw new PyException(e.type, e.value);
}
}
@ExposedMethod
public void array_byteswap() {
byteswap();
}
/**
* "Byteswap" all items of the array. This is only supported for values which are 1, 2, 4, or 8
* bytes in size; for other types of values, {@code RuntimeError} is raised. It is useful when
* reading data from a file written on a machine with a different byte order.
*/
public void byteswap() {
if (itemType == ItemType.OBJECT) {
throw Py.RuntimeError("don't know how to byteswap this array type");
}
ByteSwapper.swap(data);
}
/**
* Implementation of {@code Cloneable} interface.
*
* @return copy of current PyArray
*/
@Override
public Object clone() {
return new PyArray(this);
}
/**
* Converts a character code for the array type to the Java {@code Class} of the elements of the
* implementation array.
*
* @param typecode character code for the array type
* @return {@code Class} of the native itemClass
*/
public static Class char2class(char typecode) throws PyIgnoreMethodTag {
return ItemType.fromTypecode(typecode).itemClass;
}
@ExposedMethod
public final int array_count(PyObject value) {
// note: cpython does not raise type errors based on item type;
int iCount = 0;
int len = delegate.getSize();
for (int i = 0; i < len; i++) {
if (value.equals(itemType.get(this, i))) {
iCount++;
}
}
return iCount;
}
/**
* Return the number of occurrences of x in the array.
*
* @param value instances of the value to be counted
* @return number of time value was found in the array.
*/
public PyInteger count(PyObject value) {
return Py.newInteger(array_count(value));
}
/**
* Delete the element at position {@code i} from the array
*
* @param i index of the item to be deleted from the array
*/
@Override
protected void del(int i) {
resizeCheck(); // Prohibited if exporting a buffer
delegate.remove(i);
}
/**
* Delete the slice defined by {@code start} to {@code stop} from the array.
*
* @param start starting index of slice
* @param stop finishing index of slice
*/
@Override
protected void delRange(int start, int stop) {
resizeCheck(); // Prohibited if exporting a buffer
delegate.remove(start, stop);
}
@ExposedMethod
public final void array_extend(PyObject iterable) {
extendInternal(iterable);
}
/**
* Append items from {@code iterable} to the end of the array. If iterable is another array, it
* must have exactly the same type code; if not, TypeError will be raised. If iterable is not an
* array, it must be iterable and its elements must be the right type to be appended to the
* array.
*
* @param iterable iterable object used to extend the array
*/
public void extend(PyObject iterable) {
extendInternal(iterable);
}
/**
* Internal extend function, provides basic interface for extending arrays. Handles specific
* cases of {@code iterable} being PyStrings or PyArrays. Default behaviour is to defer to
* {@link #extendInternalIter(PyObject) extendInternalIter }
*
* @param iterable object of type PyString, PyArray or any object that can be iterated over.
*/
private void extendInternal(PyObject iterable) {
if (iterable instanceof PyUnicode) {
if (itemType == ItemType.UNICHAR) {
extendUnicodeIter(iterable);
} else if (itemType == ItemType.CHAR) {
throw Py.TypeError("array item must be char");
} else {
throw Py.TypeError("an integer is required");
}
} else if (iterable instanceof PyArray) {
PyArray source = (PyArray) iterable;
if (source.itemType == itemType) {
resizeCheck(); // Prohibited if exporting a buffer
delegate.appendArray(source.delegate.copyArray());
} else {
throw Py.TypeError("can only extend with array of same kind");
}
} else {
extendInternalIter(iterable);
}
}
/**
* Internal extend function to process iterable objects.
*
* @param iterable any object that can be iterated over.
*/
private void extendInternalIter(PyObject iterable) {
// Prohibited operation if exporting a buffer
resizeCheck();
if (iterable.__findattr__("__len__") != null) {
// Make room according to source length
int last = delegate.getSize();
delegate.ensureCapacity(last + iterable.__len__());
for (PyObject item : iterable.asIterable()) {
pyset(last++, item);
delegate.size++;
}
} else {
// iterable has no length property: cannot size the array so append each item.
for (PyObject item : iterable.asIterable()) {
appendUnchecked(item); // we already did a resizeCheck
}
}
}
/**
* Helper used only when the array elements are Unicode characters (typecode=='u').
* (Characters are stored as integer point codes.) The parameter must be an iterable yielding
* {@link PyUnicode}s. Often this will be an instance of {@link PyUnicode}, which is an iterable
* yielding single-character {@code PyUnicode}s. But it is also acceptable to this method for
* the argument to yield arbitrary {@code PyUnicode}s, which will be concatenated in the array.
*
* @param iterable of {@link PyUnicode}s
*/
private void extendUnicodeIter(PyObject iterable) {
// Prohibited operation if exporting a buffer
resizeCheck();
try {
// Append all the code points of all the strings in the iterable
for (PyObject item : iterable.asIterable()) {
PyUnicode uitem = (PyUnicode) item;
// Append all the code points of this item
for (int codepoint : uitem.toCodePoints()) {
int afterLast = delegate.getSize();
delegate.makeInsertSpace(afterLast);
Array.setInt(data, afterLast, codepoint);
}
}
} catch (ClassCastException e) {
// One of the PyUnicodes wasn't
throw notCompatibleTypeError();
}
}
private void extendArray(PyUnicode codepoints) {
// Prohibited operation if exporting a buffer
resizeCheck();
int last = delegate.getSize();
int[] items = codepoints.toCodePoints();
delegate.ensureCapacity(last + items.length);
for (int item : items) {
Array.setInt(data, last++, item);
delegate.size++;
}
}
@ExposedMethod
public final void array_fromfile(PyObject f, int count) {
fromfile(f, count);
}
/**
* Read {@code count} items (as machine values) from the file object {@code f} and append them
* to the end of the array. If less than {@code count} items are available, EOFError is raised,
* but the items that were available are still inserted into the array. {@code f} must be a real
* built-in file object; something else with a read() method won't do.
*
* @param f Python builtin file object to retrieve data
* @param count number of array elements to read
*/
public void fromfile(PyObject f, int count) {
/*
* Prohibit when exporting a buffer. Different from CPython, BufferError takes precedence in
* Jython over EOFError: if there's nowhere to write the data, we don't read it.
*/
resizeCheck();
/*
* Now get the required number of bytes from the file. Guard against non-file or closed.
*/
if (f instanceof PyFile) {
PyFile file = (PyFile) f;
if (!file.getClosed()) {
// Load required amount or whatever is available into a bytes object
int readbytes = count * itemType.itemsize;
String buffer = file.read(readbytes).toString();
fromstring(buffer);
// check for underflow
if (buffer.length() < readbytes) {
int readcount = buffer.length() / itemType.itemsize;
throw Py.EOFError(String.format(NOT_ENOUGH_IN_FILE, count, readcount));
}
}
return;
}
throw Py.TypeError("arg1 must be open file");
}
private static final String NOT_ENOUGH_IN_FILE =
"not enough items in file. %d requested, %d actually read";
@ExposedMethod
public final void array_fromlist(PyObject obj) {
fromlist(obj);
}
/**
* Append items from the list. This is equivalent to {@code for x in list: a.append(x)} except
* that if there is a type error, the array is unchanged.
*
* @param obj input list object that will be appended to the array
*/
public void fromlist(PyObject obj) {
if (!(obj instanceof PyList)) {
throw Py.TypeError("arg must be list");
}
// store the current size of the internal array
int size = delegate.getSize();
try {
extendInternalIter(obj);
} catch (PyException e) {
// trap any exception - any error invalidates the whole list
delegate.setSize(size);
// re-throw
throw new PyException(e.type, e.value);
}
}
/**
* Fill the current array with primitive values (of the type the array holds) from a stream,
* starting at array index zero, up to the capacity of the array, without resizing. Data are
* read until the array is filled or the stream runs out. If the stream does not contain a whole
* number of items (possible if the item size is not one byte), the behaviour in respect of the
* final partial item and stream position is not defined.
*
* @param is InputStream to source the data from
* @return number of primitives successfully read
* @throws IOException reflecting I/O errors during reading
*/
public int fillFromStream(InputStream is) throws IOException {
return fromStream(is, 0, delegate.getSize());
}
/**
* Read primitive values from a stream into a slice of the array, defined by a start and a
* count. Data are read until the array slice is filled or the stream runs out. Data in the
* array beyond the slice.
*
* This method is behind the manipulation of bytes into the several primitive element types on
* behalf of {@link #fillFromStream(InputStream)} etc.. The storage is resized if the slice
* being written ends beyond the current end of the array, i.e. it is increased to the value of
* {@code limit}. The return value should be checked against the count of items requested. If
* the stream runs out before the request is satisfied, the return will be less than the count,
* and items beyond the last whole item read are not altered.
*
* @param dis data stream source for the values
* @param start first element index to read
* @param count number of primitive elements to read
* @return number of primitives successfully read ({@code =count}, if not ended by EOF)
* @throws IOException reflecting I/O errors during reading
*/
private int fromStream(InputStream is, int start, int count) throws IOException {
// Ensure the array is dimensioned to fit the data expected
int size = delegate.getSize(), limit = start + count;
if (limit > size) {
resizeCheck();
delegate.setSize(limit);
}
// We need a wrapper capable of decoding the data from the representation defined by Java.
DataInputStream dis = new DataInputStream(is);
// itemType.fromStream returns *index* of first element *not* written
return itemType.fromStream(dis, data, start, limit) - start;
}
/**
* Append items from the object, which is a byte string of some kind ({@code PyString} or object
* with the buffer interface providing bytes). The string of bytes is interpreted as an array of
* machine values (as if it had been read from a file using the {@link #fromfile(PyObject, int)
* fromfile()} method).
*
* @param input string of bytes containing array data
*/
public void fromstring(PyObject input) {
array_fromstring(input);
}
/**
* Append items from the string, interpreting the string as an array of bytes (as if it had been
* read from a file using the {@link #fromfile(PyObject, int) fromfile()} method). The bytes
* encode primitive values of the type appropriate to the array,
*
* @param input string of bytes containing array data
*/
public void fromstring(String input) {
fromBytes(delegate.getSize(), StringUtil.toBytes(input));
}
/**
* Read primitive values from a stream into a slice of the array, defined by a start and the
* number of items encoded in the bytes. Data are read until the array slice is filled or the
* stream runs out. Data in the array beyond the slice are not altered. Write a slice of the
* array with primitive values
*
* items from the string, interpreting the string as an array of bytes (as if it had been read
* from a file using the {@link #fromfile(PyObject, int) fromfile()} method). The bytes encode
* primitive values of the type appropriate to the array,
*
* @param start first element index to read into
* @param input string of bytes containing array data
* @return number of primitives successfully read ({@code =count}, if not ended by EOF)
*/
public int fromstring(int start, String input) {
return fromBytes(start, StringUtil.toBytes(input));
}
/**
* Append items from a bytes-like object. The bytes encode primitive values of the type
* appropriate to the array,
*
* @param input string of bytes containing array data
*/
@ExposedMethod
final void array_fromstring(PyObject input) {
// This is an append, so start at the current end.
int start = delegate.getSize();
if (input instanceof BufferProtocol) {
if (input instanceof PyUnicode) {
// Unicode is treated as specifying a byte string via the default encoding.
String s = ((PyUnicode) input).encode();
fromBytes(start, StringUtil.toBytes(s));
} else {
// Access the bytes through the abstract API of the BufferProtocol
try (PyBuffer pybuf = ((BufferProtocol) input).getBuffer(PyBUF.STRIDED_RO)) {
if (pybuf.getNdim() == 1) {
if (pybuf.getStrides()[0] == 1) {
// Data are contiguous in the buffer
fromBytes(start, pybuf.getNIOByteBuffer());
} else {
// As frombytesInternal only knows contiguous bytes, make a copy.
byte[] copy = new byte[pybuf.getLen()];
pybuf.copyTo(copy, 0);
fromBytes(start, ByteBuffer.wrap(copy));
}
} else {
// Currently don't support n-dimensional sources
throw Py.ValueError("multi-dimensional buffer not supported");
}
}
}
} else {
String fmt = "must be string or read-only buffer, not %s";
throw Py.TypeError(String.format(fmt, input.getType().fastGetName()));
}
}
/**
* Copy into this array, starting at the given item index and expanding if necessary, a sequence
* of primitive values decoded from the contents of a byte array. The number of items copied is
* determined by the size of the byte array.
*
*
* Common code supporting Java and Python versions of .fromstring() or
* .frombytes() (Python 3.2+ name).
*
* @param start item-index of first item to read from byte buffer
* @param bytes array encoding the primitive values
* @return number of primitives successfully read
*/
private final int fromBytes(int start, byte[] bytes) {
return fromBytes(start, ByteBuffer.wrap(bytes));
}
/**
* Copy into this array, starting at the given item index and expanding if necessary, a sequence
* of primitive values decoded from the remaining bytes of a {@code ByteBuffer} (from the
* current position to the limit of the source buffer). The number of items copied is determined
* by the size of the data.
*
* This is common code supporting Java and Python versions of .fromstring() or
* .frombytes() (Python 3.2+ name).
*
* @param start item-index of first item to read from byte buffer
* @param bytes buffer encoding the primitive values
* @return number of primitives successfully read
*/
private final int fromBytes(int start, ByteBuffer bytes) {
// Check validity wrt array itemsize
int byteCount = bytes.remaining();
int count = byteCount / itemType.itemsize;
if (byteCount > count * itemType.itemsize) {
throw Py.ValueError("data length not a multiple of item size");
}
try {
// Provide argument as stream of bytes for fromstream method
InputStream is = new ByteBufferBackedInputStream(bytes);
return fromStream(is, start, count);
} catch (IOException ioe) {
// Not really possible since we just wrapped a byte buffer
return 0;
}
}
public void fromunicode(PyUnicode input) {
array_fromunicode(input);
}
@ExposedMethod
final void array_fromunicode(PyObject input) {
if (!(input instanceof PyUnicode)) {
throw Py.ValueError("fromunicode argument must be an unicode object");
} else if (itemType != ItemType.UNICHAR) {
throw Py.ValueError("fromunicode() may only be called on type 'u' arrays");
} else {
extend(input);
}
}
/**
* Get the element at position {@code i} from the array
*
* @param i index of the item to be retrieved from the array
*/
@Override
protected PyObject pyget(int i) {
return itemType.get(this, i);
}
/**
* Return the internal Java array storage of the {@code PyArray} instance
*
* @return the {@code Array} store.
*/
public Object getArray() throws PyIgnoreMethodTag {
return delegate.copyArray();
}
/**
* Getter for the item size of the array element type.
*
* The sizes returned by this method represent the number of bytes used to store the type. In
* the case of streams of primitive values, this is the number of bytes written to, or read from
* a stream. The amount of memory occupied by each item is an internal matter for Java.
*
* This method is used by other methods to define read/write quanta from strings and streams.
*
* @return number of bytes used to store array type, relevant when serialising to an array of
* bytes, or the reverse.
*/
@ExposedGet(name = "itemsize")
public int getItemsize() {
return itemType.itemsize;
}
/**
* Getter for the storage size of the array's type, relevant when serialising to an array of
* bytes, or the reverse.
*
* @return actual storage size
* @deprecated Use {@link #getItemsize()} instead which (since 2.7.3) gives the same result.
*/
@Deprecated
public int getStorageSize() {
return itemType.itemsize;
}
/**
* Retrieve a slice from the array specified by the {@code start}, {@code stop} and
* {@code step}.
*
* @param start start index of the slice
* @param stop stop index of the slice
* @param step stepping increment of the slice
* @return A new PyArray object containing the described slice
*/
@Override
protected PyObject getslice(int start, int stop, int step) {
if (step > 0 && stop < start) {
stop = start;
}
int n = sliceLength(start, stop, step);
// We have to specify both the type and the class (for use when OBJECT)
PyArray ret = new PyArray(TYPE, itemType, itemClass, n);
if (step == 1) {
System.arraycopy(data, start, ret.data, 0, n);
return ret;
}
for (int i = start, j = 0; j < n; i += step, j++) {
Array.set(ret.data, j, Array.get(data, i));
}
return ret;
}
/**
* Type as it would appear in an error message. Simple class name for {@code OBJECT}.
*
* @return single character type code or simple class name
*/
private CharSequence description() {
if (itemType == ItemType.OBJECT) {
return itemClass.getName();
} else {
return itemType.description();
}
}
/**
* Return either a Python-style {@code array.array} type code for the element (item) type or the
* Java class name.
*
* @return single character type code or simple class name
*/
@ExposedGet(name = "typecode")
public String getTypecode() {
if (itemType == ItemType.OBJECT) {
return itemClass.getName();
} else {
return itemType.typecode;
}
}
@ExposedMethod
public final int array_index(PyObject value) {
int index = indexInternal(value);
if (index != -1) {
return index;
}
throw Py.ValueError("array.index(" + value + "): " + value + " not found in array");
}
/**
* Return the smallest i such that i is the index of the first occurrence of
* {@code value} in the array.
*
* @param value value to find the index of
* @return index of the first occurrence of {@code value}
*/
public PyObject index(PyObject value) {
return Py.newInteger(array_index(value));
}
/**
* Return the smallest i such that i is the index of the first occurrence of
* {@code value} in the array.
*
* @param value value to find the index of
* @return index of the first occurrence of {@code value}
*/
private int indexInternal(PyObject value) {
// note: cpython does not raise type errors based on item type
int len = delegate.getSize();
for (int i = 0; i < len; i++) {
if (value.equals(pyget(i))) {
return i;
}
}
return -1;
}
@ExposedMethod
public final void array_insert(int index, PyObject value) {
insert(index, value);
}
/**
* Insert a new item with value {@code value} in the array before position {@code index}.
* Negative values are treated as being relative to the end of the array.
*
* @param index insert position
* @param value value to be inserted into array
*/
public void insert(int index, PyObject value) {
resizeCheck(); // Prohibited operation if exporting a buffer
index = boundToSequence(index);
delegate.makeInsertSpace(index);
pyset(index, value);
}
/**
* Removes the item at the index {@code i} from the array and returns it. The optional argument
* defaults to -1, so that by default the last item is removed and returned.
*/
@ExposedMethod(defaults = "-1")
public final PyObject array_pop(int i) {
return pop(i);
}
/**
* Removes the last item from the array and return it.
*/
public PyObject pop() {
return pop(-1);
}
/**
* Removes the item with the index {@code index} from the array and returns it.
*
* @param index array location to be popped from the array
* @return array element popped from index
*/
public PyObject pop(int index) {
if (delegate.getSize() == 0) {
throw Py.IndexError("pop from empty array");
}
index = delegator.fixindex(index);
if (index == -1) {
throw Py.IndexError("pop index out of range");
}
resizeCheck(); // Prohibit when exporting a buffer
PyObject ret = itemType.get(this, index);
delegate.remove(index);
return ret;
}
@ExposedMethod
public final void array_remove(PyObject value) {
remove(value);
}
/**
* Remove the first occurrence of {@code value} from the array.
*
* @param value array value to be removed
*/
public void remove(PyObject value) {
int index = indexInternal(value);
if (index != -1) {
resizeCheck(); // Prohibit when exporting a buffer
delegate.remove(index);
return;
}
throw Py.ValueError("array.remove(" + value + "): " + value + " not found in array");
}
/**
* Repeat the array {@code count} times.
*
* @param count number of times to repeat the array
* @return A new PyArray object containing the source object repeated {@code count} times.
*/
@Override
protected PyObject repeat(int count) {
Object arraycopy = delegate.copyArray();
PyArray ret = new PyArray(itemClass, 0);
ret.setElementType(itemType, itemClass);
for (int i = 0; i < count; i++) {
ret.delegate.appendArray(arraycopy);
}
return ret;
}
@ExposedMethod
public final void array_reverse() {
reverse();
}
/**
* Reverse the elements in the array
*/
public void reverse() {
// build a new reversed array and set this.data to it when done
Object array = Array.newInstance(itemClass, Array.getLength(data));
for (int i = 0, lastIndex = delegate.getSize() - 1; i <= lastIndex; i++) {
Array.set(array, lastIndex - i, Array.get(data, i));
}
data = array;
}
/**
* Set an element in the array - the index needs to exist, this method does not automatically
* extend the array. See {@link AbstractArray#setSize(int) AbstractArray.setSize()} or
* {@link AbstractArray#ensureCapacity(int) AbstractArray.ensureCapacity()} for ways to extend
* capacity.
*
* This code specifically checks for overflows of the integral types: byte, short, int and long.
*
* @param i index of the element to be set
* @param value value to set the element to
*/
public void set(int i, PyObject value) {
pyset(i, value);
}
@Override
protected void pyset(int i, PyObject value) {
try {
itemType.set(this, i, value);
} catch (ClassCastException cce) {
throw Py.TypeError(String.format("array item must be %s", description()));
}
}
/**
* Set element to integer value, tolerating primitive integer values in arrays of Unicode
* character, {@code int} or {@code long}. Negative values assigned to unsigned elements adopt
* their wrapped unsigned values.
*
* @param i index to set
* @param value to set
*/
public void set(int i, int value) {
if (itemType == ItemType.UNICHAR || itemClass == Integer.TYPE || itemClass == Long.TYPE) {
Array.setInt(data, i, value);
} else {
throw notCompatibleTypeError();
}
}
/**
* Set element to integer value given as a Java {@code char}, tolerating primitive integer
* values in arrays of Unicode character, {@code int} or {@code long}.
*
* @param i index to set
* @param value to set
*/
public void set(int i, char value) {
if (itemType == ItemType.CHAR || itemClass == Integer.TYPE || itemClass == Long.TYPE) {
Array.setChar(data, i, value);
} else {
throw notCompatibleTypeError();
}
}
/**
* Set element in an array of element type 'b','B', or 'c' to a Java {@code byte}.
*
* @param i index to set
* @param value to set
*/
public void set(int i, byte value) {
if (itemClass == Byte.TYPE) {
Array.setByte(data, i, value);
} else {
throw notCompatibleTypeError();
}
}
/**
* Sets a slice of the array. {@code value} can be a string (for {@code byte} and {@code char}
* types) or {@code PyArray}. If a {@code PyArray}, its type must be convertible into the type
* of the target {@code PyArray}.
*
* @param start start index of the delete slice
* @param stop end index of the delete slice
* @param step stepping increment of the slice
*/
@Override
protected void setslice(int start, int stop, int step, PyObject value) {
if (stop < start) {
stop = start;
}
if (itemClass == Character.TYPE && value instanceof PyString) {
char[] chars = null;
// if (value instanceof PyString) {
if (step != 1) {
throw Py.ValueError("invalid bounds for setting from string");
}
chars = value.toString().toCharArray();
if (start + chars.length != stop) {
// This is a size-changing operation: check for buffer exports
resizeCheck();
}
delegate.replaceSubArray(start, stop, chars, 0, chars.length);
} else {
if (value instanceof PyString && itemClass == Byte.TYPE) {
byte[] chars = ((PyString) value).toBytes();
if (chars.length == stop - start && step == 1) {
System.arraycopy(chars, 0, data, start, chars.length);
} else {
throw Py.ValueError("invalid bounds for setting from string");
}
} else if (value instanceof PyArray) {
PyArray array = arrayChecked(value);
if (step == 1) {
Object arrayDelegate;
if (array == this) {
arrayDelegate = array.delegate.copyArray();
} else {
arrayDelegate = array.delegate.getArray();
}
int len = array.delegate.getSize();
if (start + len != stop) {
// This is a size-changing operation: check for buffer exports
resizeCheck();
}
try {
delegate.replaceSubArray(start, stop, arrayDelegate, 0, len);
} catch (IllegalArgumentException e) {
throw Py.TypeError("Slice typecode " + array.reprTypecode()
+ " is not compatible with this array (typecode " + reprTypecode()
+ ")");
}
} else if (step > 1) {
int len = array.__len__();
for (int i = 0, j = 0; i < len; i++, j += step) {
Array.set(data, j + start, Array.get(array.data, i));
}
} else if (step < 0) {
if (array == this) {
array = (PyArray) array.clone();
}
int len = array.__len__();
for (int i = 0, j = start; i < len; i++, j += step) {
Array.set(data, j, Array.get(array.data, i));
}
}
} else {
throw Py.TypeError(
String.format("can only assign array (not \"%.200s\") to array " + "slice",
value.getType().fastGetName()));
}
}
}
@ExposedMethod
public final void array_tofile(PyObject f) {
tofile(f);
}
@ExposedMethod
public void array_write(PyObject f) {
tofile(f);
}
/**
* Write all items (as machine values) to the file object {@code f}.
*
* @param f Python builtin file object to write data
*/
public void tofile(PyObject f) {
if (!(f instanceof PyFile)) {
throw Py.TypeError("arg must be open file");
}
PyFile file = (PyFile) f;
file.write(tostring());
}
@ExposedMethod
public final PyObject array_tolist() {
return tolist();
}
/**
* Convert the array to an ordinary list with the same items.
*
* @return array contents as a list
*/
public PyObject tolist() {
PyList list = new PyList();
int len = delegate.getSize();
for (int i = 0; i < len; i++) {
list.append(itemType.get(this, i));
}
return list;
}
/**
* Generic stream writer to write the entire contents of the array to the stream as primitive
* types.
*
* @param os OutputStream to sink the array data to
* @return number of bytes successfully written
* @throws IOException
*/
public int toStream(OutputStream os) throws IOException {
DataOutputStream dos = new DataOutputStream(os);
itemType.toStream(dos, data, delegate.getSize());
return dos.size(); // bytes written
}
@ExposedMethod
public final PyObject array_tostring() {
return new PyString(tostring());
}
/**
* Convert the array to an array of machine values and return the string representation (the
* same sequence of bytes that would be written to a file by the {@link #tofile(PyObject)
* tofile()} method.)
*/
public String tostring() {
ByteArrayOutputStream bos = new ByteArrayOutputStream();
try {
toStream(bos);
} catch (IOException e) {
throw Py.IOError(e);
}
return StringUtil.fromBytes(bos.toByteArray());
}
@ExposedMethod
public final PyUnicode array_tounicode() {
if (itemType != ItemType.UNICHAR) {
throw Py.ValueError("tounicode() may only be called on type 'u' arrays");
}
int len = delegate.getSize();
int[] codepoints = new int[len];
for (int i = 0; i < len; i++) {
codepoints[i] = Array.getInt(data, i);
}
return new PyUnicode(codepoints);
}
public String tounicode() {
return array_tounicode().getString();
}
// PyArray can't extend anymore, so delegate
private class ArrayDelegate extends AbstractArray {
/**
* Construct a delegate that manages the {@link PyArray#data} in the containing
* {@code PyArray} or a private array of length zero it creates (if {@link PyArray#data} is
* {@code null} at the time of the call).
*/
private ArrayDelegate() {
// When do we need the data==null path?
super(data == null ? 0 : Array.getLength(data));
}
@Override
protected Object getArray() {
return data;
}
@Override
protected void setArray(Object array) {
data = array;
}
@Override
protected Object createArray(int size) {
Class baseType = data.getClass().getComponentType();
return Array.newInstance(baseType, size);
}
}
/**
* An enumeration of the supported array element (item) types and their properties (type code,
* representation, range, etc.). One member {@code OBJECT} covers all other types, to support
* Jython's ability to represent a (homogeneous) array of an arbitrary class.
*/
static enum ItemType {
// Put the signed case before an unsigned case of same Java class (fromJavaClass depends)
BYTE("b", "byte", Byte.TYPE, 1, Byte.MIN_VALUE, Byte.MAX_VALUE) {
@Override
void set(PyArray a, int i, PyObject value) {
Array.setByte(a.data, i, (byte) checkedInteger(value));
}
@Override
PyObject get(PyArray a, int i) {
return Py.newInteger(Array.getByte(a.data, i));
}
@Override
void toStream(DataOutputStream dos, Object data, int n) throws IOException {
for (int i = 0; i < n; i++) {
dos.writeByte(Array.getByte(data, i));
}
}
@Override
int fromStream(DataInputStream dis, Object data, int index, int limit)
throws IOException {
try {
while (index < limit) {
byte val = dis.readByte();
Array.setByte(data, index++, val);
}
} catch (EOFException eof) {}
return index;
}
},
UBYTE("B", "byte", Byte.TYPE, 1, Byte.MAX_VALUE * 2L + 1) {
@Override
void set(PyArray a, int i, PyObject value) {
Array.setByte(a.data, i, (byte) checkedInteger(value));
}
@Override
PyObject get(PyArray a, int i) {
// Negative values must be masked to positive int
return Py.newInteger(0xff & Array.getByte(a.data, i));
}
@Override
void toStream(DataOutputStream dos, Object data, int n) throws IOException {
for (int i = 0; i < n; i++) {
dos.writeByte(Array.getByte(data, i));
}
}
@Override
int fromStream(DataInputStream dis, Object data, int index, int limit)
throws IOException {
try {
while (index < limit) {
byte val = dis.readByte();
Array.setByte(data, index++, val);
}
} catch (EOFException eof) {}
return index;
}
},
SHORT("h", "short integer", Short.TYPE, 2, Short.MIN_VALUE, Short.MAX_VALUE) {
@Override
void set(PyArray a, int i, PyObject value) {
Array.setShort(a.data, i, (short) checkedInteger(value));
}
@Override
PyObject get(PyArray a, int i) {
return Py.newInteger(Array.getShort(a.data, i));
}
@Override
void toStream(DataOutputStream dos, Object data, int n) throws IOException {
for (int i = 0; i < n; i++) {
dos.writeShort(Array.getShort(data, i));
}
}
@Override
int fromStream(DataInputStream dis, Object data, int index, int limit)
throws IOException {
try {
while (index < limit) {
short val = dis.readShort();
Array.setShort(data, index++, val);
}
} catch (EOFException eof) {}
return index;
}
},
USHORT("H", "short integer", Short.TYPE, 2, Short.MAX_VALUE * 2L + 1) {
@Override
void set(PyArray a, int i, PyObject value) {
Array.setShort(a.data, i, (short) checkedInteger(value));
}
@Override
PyObject get(PyArray a, int i) {
// Negative values must be masked to positive int
return Py.newInteger(0xffff & Array.getShort(a.data, i));
}
@Override
void toStream(DataOutputStream dos, Object data, int n) throws IOException {
for (int i = 0; i < n; i++) {
dos.writeShort(Array.getShort(data, i));
}
}
@Override
int fromStream(DataInputStream dis, Object data, int index, int limit)
throws IOException {
try {
while (index < limit) {
short val = dis.readShort();
Array.setShort(data, index++, val);
}
} catch (EOFException eof) {}
return index;
}
},
INT("i", "int", Integer.TYPE, 4, Integer.MIN_VALUE, Integer.MAX_VALUE) {
@Override
void set(PyArray a, int i, PyObject value) {
Array.setInt(a.data, i, (int) checkedInteger(value));
}
@Override
PyObject get(PyArray a, int i) {
return Py.newInteger(Array.getInt(a.data, i));
}
@Override
void toStream(DataOutputStream dos, Object data, int n) throws IOException {
for (int i = 0; i < n; i++) {
dos.writeInt(Array.getInt(data, i));
}
}
@Override
int fromStream(DataInputStream dis, Object data, int index, int limit)
throws IOException {
try {
while (index < limit) {
int val = dis.readInt();
Array.setInt(data, index++, val);
}
} catch (EOFException eof) {}
return index;
}
},
UINT("I", "int", Integer.TYPE, 4, Integer.MAX_VALUE * 2L + 1) {
@Override
void set(PyArray a, int i, PyObject value) {
Array.setInt(a.data, i, (int) checkedInteger(value));
}
@Override
PyObject get(PyArray a, int i) {
int val = Array.getInt(a.data, i);
if (val >= 0) {
return Py.newInteger(val);
} else {
// Negative values must be interpreted as positive
return new PyLong(0xffffffffL & val);
}
}
@Override
void toStream(DataOutputStream dos, Object data, int n) throws IOException {
for (int i = 0; i < n; i++) {
dos.writeInt(Array.getInt(data, i));
}
}
@Override
int fromStream(DataInputStream dis, Object data, int index, int limit)
throws IOException {
try {
while (index < limit) {
int val = dis.readInt();
Array.setInt(data, index++, val);
}
} catch (EOFException eof) {}
return index;
}
},
LONG("l", "long", Long.TYPE, 8, Long.MIN_VALUE, Long.MAX_VALUE) {
@Override
void set(PyArray a, int i, PyObject value) {
Array.setLong(a.data, i, checkedSignedLong(value));
}
@Override
PyObject get(PyArray a, int i) {
return new PyLong(Array.getLong(a.data, i));
}
@Override
void toStream(DataOutputStream dos, Object data, int n) throws IOException {
for (int i = 0; i < n; i++) {
dos.writeLong(Array.getLong(data, i));
}
}
@Override
int fromStream(DataInputStream dis, Object data, int index, int limit)
throws IOException {
try {
while (index < limit) {
long val = dis.readLong();
Array.setLong(data, index++, val);
}
} catch (EOFException eof) {}
return index;
}
},
ULONG("L", "long", Long.TYPE, 8, Long.MAX_VALUE * 2L + 1) {
@Override
void set(PyArray a, int i, PyObject value) {
Array.setLong(a.data, i, checkedUnsignedLong(value));
}
@Override
PyObject get(PyArray a, int i) {
long val = Array.getLong(a.data, i);
if (val >= 0) {
return new PyLong(val);
} else {
// Negative values must be interpreted as positive
return new PyLong(BigInteger.valueOf(val).add(TWO_TO_64));
}
}
@Override
void toStream(DataOutputStream dos, Object data, int n) throws IOException {
for (int i = 0; i < n; i++) {
dos.writeLong(Array.getLong(data, i));
}
}
@Override
int fromStream(DataInputStream dis, Object data, int index, int limit)
throws IOException {
try {
while (index < limit) {
long val = dis.readLong();
Array.setLong(data, index++, val);
}
} catch (EOFException eof) {}
return index;
}
},
FLOAT("f", "float", Float.TYPE, 4) {
@Override
void set(PyArray a, int i, PyObject value) {
Array.setFloat(a.data, i, (float) value.asDouble());
}
@Override
PyObject get(PyArray a, int i) {
return new PyFloat(Array.getFloat(a.data, i));
}
@Override
void toStream(DataOutputStream dos, Object data, int n) throws IOException {
for (int i = 0; i < n; i++) {
dos.writeFloat(Array.getFloat(data, i));
}
}
@Override
int fromStream(DataInputStream dis, Object data, int index, int limit)
throws IOException {
try {
while (index < limit) {
float val = dis.readFloat();
Array.setFloat(data, index++, val);
}
} catch (EOFException eof) {}
return index;
}
},
DOUBLE("d", "double", Double.TYPE, 8) {
@Override
void set(PyArray a, int i, PyObject value) {
Array.setDouble(a.data, i, value.asDouble());
}
@Override
PyObject get(PyArray a, int i) {
return new PyFloat(Array.getDouble(a.data, i));
}
@Override
void toStream(DataOutputStream dos, Object data, int n) throws IOException {
for (int i = 0; i < n; i++) {
dos.writeDouble(Array.getDouble(data, i));
}
}
@Override
int fromStream(DataInputStream dis, Object data, int index, int limit)
throws IOException {
try {
while (index < limit) {
double val = dis.readDouble();
Array.setDouble(data, index++, val);
}
} catch (EOFException eof) {}
return index;
}
},
CHAR("c", "char", Byte.TYPE, 1, Byte.MAX_VALUE * 2L + 1) {
// Has to be after BYTE so fromJavaClass returns that for Byte.TYPE
@Override
void set(PyArray a, int i, PyObject value) {
String s = ((PyString) value).getString();
if (s.length() != 1) {
throw new ClassCastException();
}
Array.setByte(a.data, i, (byte) s.charAt(0));
}
@Override
PyObject get(PyArray a, int i) {
return Py.makeCharacter((char) (0xff & Array.getByte(a.data, i)));
}
@Override
void toStream(DataOutputStream dos, Object data, int n) throws IOException {
for (int i = 0; i < n; i++) {
dos.writeByte(Array.getByte(data, i));
}
}
@Override
int fromStream(DataInputStream dis, Object data, int index, int limit)
throws IOException {
try {
while (index < limit) {
byte val = dis.readByte();
Array.setByte(data, index++, val);
}
} catch (EOFException eof) {}
return index;
}
},
UNICHAR("u", "unicode character", Integer.TYPE, 4, Character.MAX_CODE_POINT) {
// Has to be after INT so fromJavaClass returns that for Integer.TYPE
@Override
void set(PyArray a, int i, PyObject value) {
String s = ((PyUnicode) value).getString();
if (s.codePointCount(0, Math.min(s.length(), 4)) != 1) {
throw new ClassCastException();
}
Array.setInt(a.data, i, s.codePointAt(0));
}
@Override
PyObject get(PyArray a, int i) {
return new PyUnicode(Array.getInt(a.data, i));
}
@Override
void toStream(DataOutputStream dos, Object data, int n) throws IOException {
for (int i = 0; i < n; i++) {
dos.writeInt(Array.getInt(data, i));
}
}
@Override
int fromStream(DataInputStream dis, Object data, int index, int limit)
throws IOException {
try {
while (index < limit) {
int val = dis.readInt();
Array.setInt(data, index++, val);
}
} catch (EOFException eof) {}
return index;
}
},
BOOLEAN("z", "boolean", Boolean.TYPE, 1, 0) {
@Override
void set(PyArray a, int i, PyObject value) {
Array.setBoolean(a.data, i, value.__nonzero__());
}
@Override
PyObject get(PyArray a, int i) {
return Py.newBoolean(Array.getBoolean(a.data, i));
}
@Override
void toStream(DataOutputStream dos, Object data, int n) throws IOException {
for (int i = 0; i < n; i++) {
dos.writeBoolean(Array.getBoolean(data, i));
}
}
@Override
int fromStream(DataInputStream dis, Object data, int index, int limit)
throws IOException {
try {
while (index < limit) {
boolean val = dis.readBoolean();
Array.setBoolean(data, index++, val);
}
} catch (EOFException eof) {}
return index;
}
},
OBJECT() {
@Override
void set(PyArray a, int i, PyObject value) {
Object val = value.__tojava__(a.itemClass);
if (val == Py.NoConversion) {
throw a.notCompatibleTypeError();
}
Array.set(a.data, i, val);
}
@Override
PyObject get(PyArray a, int i) {
return Py.java2py(Array.get(a.data, i));
}
@Override
void toStream(DataOutputStream dos, Object data, int n) throws IOException {
// Not supported: silently ignored (in versions so far)
}
@Override
int fromStream(DataInputStream dis, Object data, int index, int limit)
throws IOException {
// Not supported: silently ignored (in versions so far)
return index;
}
};
/**
* Implementation of setting an element in the data array, specific to this
* {@code ItemType}.
*
* @param a the PyArray on which to operate
* @param i index of element
* @param value to set
*/
abstract void set(PyArray a, int i, PyObject value);
/**
* Implementation of getting an element from the data array, specific to this
* {@code ItemType}.
*
* @param a the PyArray on which to operate
* @param i index of element
* @return value got
*/
abstract PyObject get(PyArray a, int i);
/**
* Generic stream writer to write the entire contents of the array to the stream as
* primitive types. After the call returns, {@code dos.size()} indicates the number of bytes
* successfully written.
*
* @param dos to sink the array data to
* @param data array to write from
* @param n number of items to write
* @throws IOException reflecting I/O errors during writing
*/
abstract void toStream(DataOutputStream dos, Object data, int n) throws IOException;
/**
* Implementation of reading primitive values from a stream into a slice of the array. Data
* are read until the array slice is filled or the stream runs out. Return the index of the
* first item not written: if this is less than the limit, it is because the read ended
* early on an end-of-file.
*
* Each item type provides its own manipulation of bytes into the several primitive element
* types on behalf of {@link #fillFromStream(InputStream)} etc..
*
* If the stream does not contain a whole number of items (possible if the item size is not
* one byte), the behaviour in respect of the final partial item and stream position is not
* defined.
*
* @param dis data stream source for the values
* @param data destination array
* @param index of first array element to read
* @param limit first element not to read
* @return index of first element not read ({@code =limit}, if not ended by EOF)
* @throws IOException reflecting I/O errors during reading
*/
abstract int fromStream(DataInputStream dis, Object data, int index, int limit)
throws IOException;
/** Type name as it would appear in an {@code array.array} constructor. */
String typecode;
/** Type as it would appear in an error message. "" for OBJECT. */
final String description;
/** Class representing elements of this type (for primitive cases). "" for OBJECT. */
Class itemClass;
/** Number of bytes in serialised form (not necessarily in memory). */
int itemsize;
/** Minimum value (only valid in integral and character types). */
long min;
/** Maximum value (only valid in integral and character types). */
long max;
/**
* Do-it-all constructor.
*
* @param typecode type name (as seen in {@code array.array} constructor)
* @param itemClass class of the elements of the array (when primitive)
* @param itemsize size of items in the array for serialisation
* @param min minimum value (for integer item types)
* @param max maximum value (for integer item types)
*/
ItemType(String typecode, String description, Class itemClass, int itemsize, long min,
long max) {
this.typecode = typecode;
this.description = description;
this.itemClass = itemClass;
this.itemsize = itemsize;
this.min = min;
this.max = max;
}
/** Signed types with values representable by a Java {@code double}. */
ItemType(String typecode, String description, Class itemClass, int itemsize) {
this(typecode, description, itemClass, itemsize, 0, 0);
}
/** Unsigned types with values representable by a Java {@code long}. */
ItemType(String typecode, String description, Class itemClass, int itemsize, long max) {
this(typecode, description, itemClass, itemsize, 0L, max);
}
/** Object element types (represented by the singular {@code ItemType OBJECT}. */
ItemType() {
this("", "Java Object", Object.class, 0, 0, 0);
}
/**
* Converts a character code for the array type to the corresponding {@code ItemType} for
* the elements of the implementation array. If the code is not one of the valid options,
* {@code ValueError} is raised.
*
* @param typecode character code for the array item type
* @return {@code ItemType} of the array elements
*/
static ItemType fromTypecode(char typecode) {
switch (typecode) {
case 'z':
return BOOLEAN;
case 'b':
return BYTE;
case 'B':
return UBYTE;
case 'u':
return UNICHAR;
case 'c':
return CHAR;
case 'h':
return SHORT;
case 'H':
return USHORT;
case 'i':
return INT;
case 'I':
return UINT;
case 'l':
return LONG;
case 'L':
return ULONG;
case 'f':
return FLOAT;
case 'd':
return DOUBLE;
default:
throw Py.ValueError("typecode must be " + reminder());
}
}
/**
* Map a Java class to the {@code ItemType} type code that represents it. Where that may be
* ambiguous, the method assumes signed representation (so for example {@code Integer.TYPE}
* maps to {@code 'i'} not {@code 'I'}). Classes other than those map to their Java class
* name. This supports the extended repertoire {@code array.array} has in Jython.
*
* @param cls element class
* @return the {@code array.array} type code that representing {@code cls}
*/
static ItemType fromJavaClass(Class cls) {
for (ItemType i : ItemType.values()) {
if (cls.equals(i.itemClass)) {
return i;
}
}
return OBJECT;
}
protected static final BigInteger TWO_TO_64 = BigInteger.ONE.shiftLeft(64);
/**
* Convert an integer value to a Java long and test it against the bounds {@link #min} and
* {@link #max}.
*
* An {@code OverflowError} is raised if these bounds are exceeded and a
* {@code ClassCastException} if the type is not integral (cannot be converted to a
* {@code long}).
*
* @param value to convert
* @return {@code value} as a {@code long}
* @throws PyException {@code OverflowError} if out of range for item type
* @throws ClassCastException if not integral (cannot be converted to a {@code long}).
*/
protected long checkedInteger(PyObject value) {
long val;
if (min == 0) {
val = checkedUnsignedLong(value);
} else {
val = checkedSignedLong(value);
if (val < min) {
throw lessThanMinimum();
}
}
if (val > max) {
throw moreThanMaximum();
} else {
return val;
}
}
/**
* Check the range of an integer value is correct for a Java {@code long} and convert it to
* that type. An {@code OverflowError} is raised if these bounds are exceeded.
*
* @param value to convert
* @return {@code value} as a {@code long}
* @throws PyException {@code OverflowError} if out of range for signed long
* @throws ClassCastException if not integral (cannot be converted to a {@code long}).
*/
protected long checkedSignedLong(PyObject value) throws PyException, ClassCastException {
if (value instanceof PyInteger) {
return ((PyInteger) value).getValue();
} else if (value instanceof PyLong) {
BigInteger val = ((PyLong) value).getValue();
if (PyLong.MAX_LONG.compareTo(val) < 0) {
throw moreThanMaximum();
} else if (PyLong.MIN_LONG.compareTo(val) > 0) {
throw lessThanMinimum();
} else {
return val.longValue();
}
} else {
Long val = (Long) value.__tojava__(Long.TYPE);
if (val == null) {
throw new ClassCastException();
}
return val.longValue();
}
}
/**
* Check the range of an integer value is correct for an unsigned 64-bit integer, and
* convert it to the Java {@code long} with the same bit pattern. (This is what represents
* an {@code unsigned long} in the array.) An {@code OverflowError} is raised if the
* allowable range is exceeded.
*
* @param value to convert
* @return {@code value} as a {@code long}
* @throws PyException {@code OverflowError} if out of range for unsigned long
* @throws ClassCastException if not integral (cannot be converted to a {@code long}).
*/
protected long checkedUnsignedLong(PyObject value) throws ClassCastException {
if (value instanceof PyInteger) {
int val = ((PyInteger) value).getValue();
if (val < 0) {
throw lessThanMinimum();
} else {
return val;
}
} else {
BigInteger val;
if (value instanceof PyLong) {
val = ((PyLong) value).getValue();
} else {
val = (BigInteger) value.__tojava__(BigInteger.class);
if (val == null) {
throw new ClassCastException();
}
}
if (BigInteger.ZERO.compareTo(val) > 0) {
throw lessThanMinimum();
} else if (PyLong.MAX_ULONG.compareTo(val) < 0) {
throw moreThanMaximum();
} else if (PyLong.MAX_LONG.compareTo(val) < 0) {
// In range 2^63 <= value < 2^64: represent as negative long.
return val.subtract(TWO_TO_64).longValue();
} else {
return val.longValue();
}
}
}
/**
* Provide a description of the element type, typically for use in an error message.
*
* @return a description of the element type
*/
CharSequence description() {
StringBuilder buf = new StringBuilder(description.length() + 10);
switch (this) {
case BOOLEAN:
case CHAR:
case DOUBLE:
case FLOAT:
case OBJECT:
case UNICHAR:
break;
default:
buf.append(min < 0 ? "signed " : "unsigned ");
}
buf.append(description);
return buf;
}
/** @return a reminder of the valid item types. */
private static CharSequence reminder() {
StringBuilder buf = new StringBuilder(100);
for (ItemType i : ItemType.values()) {
if (i != OBJECT) {
if (buf.length() != 0) {
buf.append(", ");
}
buf.append(i.typecode);
}
}
buf.append(" or a Java class");
return buf;
}
/**
* Create throwable {@code OverflowError} along the lines "TYPE-array value is less than
* minimum", where TYPE is the element type of the array.
*
* @return the {@code OverflowError}
*/
protected PyException lessThanMinimum() {
return Py.OverflowError(
String.format("%s array value is less than minimum", description()));
}
/**
* Create throwable {@code OverflowError} along the lines "TYPE-array value is more than
* maximum", where TYPE is the element type of the array.
*
* @return the {@code OverflowError}
*/
protected PyException moreThanMaximum() {
return Py.OverflowError(
String.format("%s array value is more than maximum", description()));
}
}
/*
* ============================================================================================
* Support for the Buffer API
* ============================================================================================
*
* The buffer API allows other classes to access the storage directly.
*
* This is a close duplicate of the same mechanism in PyByteArray. There is perhaps scope for a
* shared helper class to implement this logic. For type code 'b', the workings are almost
* identical. The fully-fledged buffer interface for PyArray is richer, more like the Python 3
* memoryview, as it must cope with items of size other than one byte. This goes beyond the
* capabilities of the Jython BufferProtocol at this stage of its development.
*/
/**
* Hold weakly a reference to a PyBuffer export not yet released, used to prevent untimely
* resizing.
*/
private WeakReference export;
/**
* {@inheritDoc}
*
* The {@link PyBuffer} returned from this method is a one-dimensional array of single byte
* items that allows modification of the object state. The existence of this export prohibits
* resizing the array. This prohibition is not only on the consumer of the view but extends
* to operations on the underlying array, such as {@link #insert(int, PyObject)} or
* {@link #pop()}.
*/
@Override
public synchronized PyBuffer getBuffer(int flags) {
if ((flags & ~PyBUF.WRITABLE) == PyBUF.SIMPLE) {
// Client requests a flat byte-oriented read-view, typically from buffer(a).
// If we have already exported a buffer it may still be available for re-use
BaseBuffer pybuf = getExistingBuffer(flags);
if (pybuf == null) {
// No existing export we can re-use: create a new one (acts as unsigned)
if (itemClass == Byte.TYPE) {
// This is byte data, so we can export directly
byte[] storage = (byte[]) data;
int size = delegate.getSize();
pybuf = new SimpleBuffer(flags, this, storage, 0, size);
} else {
// As the client only intends to read, fake the answer with a String
pybuf = new SimpleStringBuffer(flags, this, tostring());
}
// Hold a reference for possible re-use
export = new WeakReference(pybuf);
}
return pybuf;
} else {
// Client request goes beyond Python 2 capability, typically from memoryview(a).
throw new ClassCastException("'array' supports only a byte-buffer view");
}
}
/**
* Try to re-use an existing exported buffer, or return {@code null} if we can't.
*
* @throws PyException {@code BufferError} if the the flags are incompatible with the buffer
*/
private BaseBuffer getExistingBuffer(int flags) throws PyException {
BaseBuffer pybuf = null;
if (export != null) {
// A buffer was exported at some time.
pybuf = export.get();
if (pybuf != null) {
/*
* We do not test for pybuf.isReleased() as, if any operation had taken place that
* invalidated the buffer, resizeCheck() would have set export=null. The exported
* buffer (navigation, buf member, etc.) remains valid through any operation that
* does not need a resizeCheck.
*/
pybuf = pybuf.getBufferAgain(flags);
}
}
return pybuf;
}
/**
* Test to see if the array may be resized and raise a BufferError if not. This must be called
* by the implementation of any operation that changes the number of elements in the array.
*
* @throws PyException {@code BufferError} if there are buffer exports preventing a resize
*/
private void resizeCheck() throws PyException {
if (export != null) {
// A buffer was exported at some time and we have not explicitly discarded it.
PyBuffer pybuf = export.get();
if (pybuf != null && !pybuf.isReleased()) {
// A consumer still has the exported buffer
throw Py.BufferError("cannot resize an array that is exporting buffers");
} else {
/*
* Either the reference has expired or all consumers have released it. Either way,
* the weak reference is useless now.
*/
export = null;
}
}
}
/**
* Wrap a {@link ByteBuffer} in an InputStream. Reference:
* Stackoverflow question 4332264.
*/
private class ByteBufferBackedInputStream extends InputStream {
ByteBuffer buf;
public ByteBufferBackedInputStream(ByteBuffer buf) {
this.buf = buf;
}
/**
* Return the number of bytes remaining in the underlying buffer.
*/
@Override
public int available() throws IOException {
return buf.remaining();
}
@Override
public int read() {
return buf.hasRemaining() ? buf.get() & 0xff : -1;
}
@Override
public int read(byte[] bytes, int off, int len) {
int n = buf.remaining();
if (n >= len) {
// There are enough bytes remaining to satisfy the request.
buf.get(bytes, off, len);
return len;
} else if (n > 0) {
// There are some bytes remaining: truncate request.
buf.get(bytes, off, n);
return n;
} else {
// Signal that there are no bytes left
return -1;
}
}
}
/* Traverseproc implementation */
@Override
public int traverse(Visitproc visit, Object arg) {
if (data == null || !gc.canLinkToPyObject(data.getClass(), true)) {
return 0;
}
return gc.traverseByReflection(data, visit, arg);
}
@Override
public boolean refersDirectlyTo(PyObject ob) throws UnsupportedOperationException {
if (data == null || !gc.canLinkToPyObject(data.getClass(), true)) {
return false;
}
throw new UnsupportedOperationException();
}
/**
* Create throwable {@code TypeError} along the lines "Type not compatible with array of
* TYPECODE", where TYPECODE is the element type of the array.
*
* @return the {@code TypeError}
*/
PyException notCompatibleTypeError() {
return Py.TypeError(String.format("Type not compatible with array of '%s'", description()));
}
}