All Downloads are FREE. Search and download functionalities are using the official Maven repository.

org.elasticsearch.rest.support.RestUtils Maven / Gradle / Ivy

The newest version!
/*
 * Licensed to Elasticsearch under one or more contributor
 * license agreements. See the NOTICE file distributed with
 * this work for additional information regarding copyright
 * ownership. Elasticsearch licenses this file to you under
 * the Apache License, Version 2.0 (the "License"); you may
 * not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *    http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing,
 * software distributed under the License is distributed on an
 * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
 * KIND, either express or implied.  See the License for the
 * specific language governing permissions and limitations
 * under the License.
 */

package org.elasticsearch.rest.support;

import com.google.common.base.Charsets;
import org.elasticsearch.common.Nullable;
import org.elasticsearch.common.Strings;
import org.elasticsearch.common.path.PathTrie;

import java.nio.charset.Charset;
import java.util.Map;
import java.util.regex.Pattern;

/**
 *
 */
public class RestUtils {

    public static final PathTrie.Decoder REST_DECODER = new PathTrie.Decoder() {
        @Override
        public String decode(String value) {
            return RestUtils.decodeComponent(value);
        }
    };

    public static void decodeQueryString(String s, int fromIndex, Map params) {
        if (fromIndex < 0) {
            return;
        }
        if (fromIndex >= s.length()) {
            return;
        }

        int queryStringLength = s.contains("#") ? s.indexOf("#") : s.length();

        String name = null;
        int pos = fromIndex; // Beginning of the unprocessed region
        int i;       // End of the unprocessed region
        char c = 0;  // Current character
        for (i = fromIndex; i < queryStringLength; i++) {
            c = s.charAt(i);
            if (c == '=' && name == null) {
                if (pos != i) {
                    name = decodeComponent(s.substring(pos, i));
                }
                pos = i + 1;
            } else if (c == '&') {
                if (name == null && pos != i) {
                    // We haven't seen an `=' so far but moved forward.
                    // Must be a param of the form '&a&' so add it with
                    // an empty value.
                    addParam(params, decodeComponent(s.substring(pos, i)), "");
                } else if (name != null) {
                    addParam(params, name, decodeComponent(s.substring(pos, i)));
                    name = null;
                }
                pos = i + 1;
            }
        }

        if (pos != i) {  // Are there characters we haven't dealt with?
            if (name == null) {     // Yes and we haven't seen any `='.
                addParam(params, decodeComponent(s.substring(pos, i)), "");
            } else {                // Yes and this must be the last value.
                addParam(params, name, decodeComponent(s.substring(pos, i)));
            }
        } else if (name != null) {  // Have we seen a name without value?
            addParam(params, name, "");
        }
    }

    private static void addParam(Map params, String name, String value) {
        params.put(name, value);
    }

    /**
     * Decodes a bit of an URL encoded by a browser.
     * 

* This is equivalent to calling {@link #decodeComponent(String, Charset)} * with the UTF-8 charset (recommended to comply with RFC 3986, Section 2). * * @param s The string to decode (can be empty). * @return The decoded string, or {@code s} if there's nothing to decode. * If the string to decode is {@code null}, returns an empty string. * @throws IllegalArgumentException if the string contains a malformed * escape sequence. */ public static String decodeComponent(final String s) { return decodeComponent(s, Charsets.UTF_8); } /** * Decodes a bit of an URL encoded by a browser. *

* The string is expected to be encoded as per RFC 3986, Section 2. * This is the encoding used by JavaScript functions {@code encodeURI} * and {@code encodeURIComponent}, but not {@code escape}. For example * in this encoding, é (in Unicode {@code U+00E9} or in UTF-8 * {@code 0xC3 0xA9}) is encoded as {@code %C3%A9} or {@code %c3%a9}. *

* This is essentially equivalent to calling * {@link java.net.URLDecoder URLDecoder}.{@link * java.net.URLDecoder#decode(String, String)} * except that it's over 2x faster and generates less garbage for the GC. * Actually this function doesn't allocate any memory if there's nothing * to decode, the argument itself is returned. * * @param s The string to decode (can be empty). * @param charset The charset to use to decode the string (should really * be {@link Charsets#UTF_8}. * @return The decoded string, or {@code s} if there's nothing to decode. * If the string to decode is {@code null}, returns an empty string. * @throws IllegalArgumentException if the string contains a malformed * escape sequence. */ @SuppressWarnings("fallthrough") public static String decodeComponent(final String s, final Charset charset) { if (s == null) { return ""; } final int size = s.length(); boolean modified = false; for (int i = 0; i < size; i++) { final char c = s.charAt(i); switch (c) { case '%': i++; // We can skip at least one char, e.g. `%%'. // Fall through. case '+': modified = true; break; } } if (!modified) { return s; } final byte[] buf = new byte[size]; int pos = 0; // position in `buf'. for (int i = 0; i < size; i++) { char c = s.charAt(i); switch (c) { case '+': buf[pos++] = ' '; // "+" -> " " break; case '%': if (i == size - 1) { throw new IllegalArgumentException("unterminated escape" + " sequence at end of string: " + s); } c = s.charAt(++i); if (c == '%') { buf[pos++] = '%'; // "%%" -> "%" break; } else if (i == size - 1) { throw new IllegalArgumentException("partial escape" + " sequence at end of string: " + s); } c = decodeHexNibble(c); final char c2 = decodeHexNibble(s.charAt(++i)); if (c == Character.MAX_VALUE || c2 == Character.MAX_VALUE) { throw new IllegalArgumentException( "invalid escape sequence `%" + s.charAt(i - 1) + s.charAt(i) + "' at index " + (i - 2) + " of: " + s); } c = (char) (c * 16 + c2); // Fall through. default: buf[pos++] = (byte) c; break; } } return new String(buf, 0, pos, charset); } /** * Helper to decode half of a hexadecimal number from a string. * * @param c The ASCII character of the hexadecimal number to decode. * Must be in the range {@code [0-9a-fA-F]}. * @return The hexadecimal value represented in the ASCII character * given, or {@link Character#MAX_VALUE} if the character is invalid. */ private static char decodeHexNibble(final char c) { if ('0' <= c && c <= '9') { return (char) (c - '0'); } else if ('a' <= c && c <= 'f') { return (char) (c - 'a' + 10); } else if ('A' <= c && c <= 'F') { return (char) (c - 'A' + 10); } else { return Character.MAX_VALUE; } } /** * Determine if CORS setting is a regex * * @return a corresponding {@link Pattern} if so and o.w. null. */ public static Pattern checkCorsSettingForRegex(String corsSetting) { if (corsSetting == null) { return null; } int len = corsSetting.length(); boolean isRegex = len > 2 && corsSetting.startsWith("/") && corsSetting.endsWith("/"); if (isRegex) { return Pattern.compile(corsSetting.substring(1, corsSetting.length() - 1)); } return null; } /** * Return the CORS setting as an array of origins. * * @param corsSetting the CORS allow origin setting as configured by the user; * should never pass null, but we check for it anyway. * @return an array of origins if set, otherwise {@code null}. */ public static String[] corsSettingAsArray(String corsSetting) { if (Strings.isNullOrEmpty(corsSetting)) { return new String[0]; } String[] arr = corsSetting.split(","); for (int i = 0; i < arr.length; i++) { arr[i] = arr[i].trim(); } return arr; } }





© 2015 - 2024 Weber Informatics LLC | Privacy Policy