ecki commented on code in PR #776:
URL: https://github.com/apache/commons-io/pull/776#discussion_r2338658902
##########
src/main/java/org/apache/commons/io/IOUtils.java:
##########
@@ -2637,57 +2645,61 @@ public static BufferedReader toBufferedReader(final
Reader reader, final int siz
}
/**
- * Gets the contents of an {@link InputStream} as a {@code byte[]}.
- * <p>
- * This method buffers the input internally, so there is no need to use a
{@link BufferedInputStream}.
- * </p>
+ * Reads all the bytes from an input stream in a byte array.
*
- * @param inputStream the {@link InputStream} to read.
- * @return the requested byte array.
- * @throws NullPointerException if the InputStream is {@code null}.
- * @throws IOException if an I/O error occurs or reading more
than {@link Integer#MAX_VALUE} occurs.
+ * <p>The memory used by this method is <strong>proportional</strong> to
the number
+ * of bytes read, which is only limited by {@link Integer#MAX_VALUE}. This
makes it unsuitable for
+ * processing large input streams, unless sufficient heap space is
available.</p>
+ *
+ * @param inputStream The {@link InputStream} to read; must not be {@code
null}.
+ * @return A new byte array.
+ * @throws IllegalArgumentException If the size of the stream is greater
than the maximum array size.
+ * @throws IOException If an I/O error occurs while reading.
+ * @throws NullPointerException If {@code inputStream} is {@code null}.
*/
public static byte[] toByteArray(final InputStream inputStream) throws
IOException {
- // We use a ThresholdingOutputStream to avoid reading AND writing more
than Integer.MAX_VALUE.
- try (UnsynchronizedByteArrayOutputStream ubaOutput =
UnsynchronizedByteArrayOutputStream.builder().get();
- ThresholdingOutputStream thresholdOutput = new
ThresholdingOutputStream(Integer.MAX_VALUE, os -> {
- throw new IllegalArgumentException(String.format("Cannot
read more than %,d into a byte array", Integer.MAX_VALUE));
- }, os -> ubaOutput)) {
- copy(inputStream, thresholdOutput);
- return ubaOutput.toByteArray();
+ final UnsynchronizedByteArrayOutputStream output =
copyToOutputStream(inputStream, SOFT_MAX_ARRAY_LENGTH + 1, DEFAULT_BUFFER_SIZE);
+ if (output.size() > SOFT_MAX_ARRAY_LENGTH) {
+ throw new IllegalArgumentException(String.format("Cannot read more
than %,d into a byte array", SOFT_MAX_ARRAY_LENGTH));
}
+ return output.toByteArray();
}
/**
- * Gets the contents of an {@link InputStream} as a {@code byte[]}. Use
this method instead of
- * {@link #toByteArray(InputStream)} when {@link InputStream} size is
known.
+ * Reads exactly {@code size} bytes from the given {@link InputStream}
into a new {@code byte[]}.
*
- * @param input the {@link InputStream} to read.
- * @param size the size of {@link InputStream} to read, where 0 <
{@code size} <= length of input stream.
- * @return byte [] of length {@code size}.
- * @throws IOException if an I/O error occurs or {@link InputStream}
length is smaller than parameter {@code size}.
- * @throws IllegalArgumentException if {@code size} is less than zero.
+ * <p>This variant provides no safeguards against allocating very large
arrays.
+ * For large streams, prefer {@link #toByteArray(InputStream, int, int)},
+ * which enforces stricter memory usage constraints.</p>
+ *
+ * @param input the {@link InputStream} to read; must not be {@code null}.
+ * @param size the exact number of bytes to read; must be {@code >= 0}.
+ * @return a new byte array of length {@code size}.
+ * @throws IllegalArgumentException if {@code size} is negative.
+ * @throws EOFException if the stream ends before {@code size}
bytes are read.
+ * @throws IOException if an I/O error occurs while reading.
+ * @throws NullPointerException if {@code input} is {@code null}.
* @since 2.1
*/
- @SuppressWarnings("resource")
public static byte[] toByteArray(final InputStream input, final int size)
throws IOException {
- return toByteArray(Objects.requireNonNull(input, "input")::read, size);
+ return toByteArray(input::read, size);
}
/**
- * Gets contents of an {@link InputStream} as a {@code byte[]}.
- * Use this method instead of {@link #toByteArray(InputStream)}
- * when {@link InputStream} size is known.
- * <strong>NOTE:</strong> the method checks that the length can safely be
cast to an int without truncation
- * before using {@link IOUtils#toByteArray(InputStream, int)} to read into
the byte array.
- * (Arrays can have no more than Integer.MAX_VALUE entries anyway.)
+ * Reads exactly {@code size} bytes from the given {@link InputStream}
into a new {@code byte[]}.
*
- * @param input the {@link InputStream} to read.
- * @param size the size of {@link InputStream} to read, where 0 <
{@code size} <= min(Integer.MAX_VALUE, length of input stream).
- * @return byte [] the requested byte array, of length {@code size}.
- * @throws IOException if an I/O error occurs or {@link
InputStream} length is less than {@code size}.
- * @throws IllegalArgumentException if size is less than zero or size is
greater than Integer.MAX_VALUE.
- * @see IOUtils#toByteArray(InputStream, int)
+ * <p>This variant provides no safeguards against allocating very large
arrays.
Review Comment:
How about “This variant always allocates the whole requested array size, for
a dynamic growing variant use ..”?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscr...@commons.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
