common/net/InternetDomainName.java

/*
 * Copyright (C) 2009 The Guava Authors
 *
 * Licensed 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 com.google.common.net;

import static com.google.common.base.Preconditions.checkArgument;
import static com.google.common.base.Preconditions.checkNotNull;
import static com.google.common.base.Preconditions.checkState;

import com.google.common.annotations.Beta;
import com.google.common.annotations.GwtCompatible;
import com.google.common.base.Ascii;
import com.google.common.base.CharMatcher;
import com.google.common.base.Joiner;
import com.google.common.base.Splitter;
import com.google.common.collect.ImmutableList;
import com.google.thirdparty.publicsuffix.PublicSuffixPatterns;

import java.util.List;

import javax.annotation.Nullable;

/**
 * An immutable well-formed internet domain name, such as {@code com} or {@code
 * foo.co.uk}. Only syntactic analysis is performed; no DNS lookups or other
 * network interactions take place. Thus there is no guarantee that the domain
 * actually exists on the internet.
 *
 * <p>One common use of this class is to determine whether a given string is
 * likely to represent an addressable domain on the web -- that is, for a
 * candidate string {@code "xxx"}, might browsing to {@code "http://xxx/"}
 * result in a webpage being displayed? In the past, this test was frequently
 * done by determining whether the domain ended with a {@linkplain
 * #isPublicSuffix() public suffix} but was not itself a public suffix. However,
 * this test is no longer accurate. There are many domains which are both public
 * suffixes and addressable as hosts; {@code "uk.com"} is one example. As a
 * result, the only useful test to determine if a domain is a plausible web host
 * is {@link #hasPublicSuffix()}. This will return {@code true} for many domains
 * which (currently) are not hosts, such as {@code "com"}, but given that any
 * public suffix may become a host without warning, it is better to err on the
 * side of permissiveness and thus avoid spurious rejection of valid sites.
 *
 * <p>During construction, names are normalized in two ways:
 * <ol>
 * <li>ASCII uppercase characters are converted to lowercase.
 * <li>Unicode dot separators other than the ASCII period ({@code '.'}) are
 * converted to the ASCII period.
 * </ol>
 * <p>The normalized values will be returned from {@link #toString()} and
 * {@link #parts()}, and will be reflected in the result of
 * {@link #equals(Object)}.
 *
 * <p><a href="http://en.wikipedia.org/wiki/Internationalized_domain_name">
 * Internationalized domain names</a> such as {@code 网络.cn} are supported, as
 * are the equivalent <a
 * href="http://en.wikipedia.org/wiki/Internationalized_domain_name">IDNA
 * Punycode-encoded</a> versions.
 *
 * @author Craig Berry
 * @since 5.0
 */
@Beta
@GwtCompatible
public final class InternetDomainName {

  private static final CharMatcher DOTS_MATCHER =
      CharMatcher.anyOf(".\u3002\uFF0E\uFF61");
  private static final Splitter DOT_SPLITTER = Splitter.on('.');
  private static final Joiner DOT_JOINER = Joiner.on('.');

  /**
   * Value of {@link #publicSuffixIndex} which indicates that no public suffix
   * was found.
   */
  private static final int NO_PUBLIC_SUFFIX_FOUND = -1;

  private static final String DOT_REGEX = "\\.";

  /**
   * Maximum parts (labels) in a domain name. This value arises from
   * the 255-octet limit described in
   * <a href="http://www.ietf.org/rfc/rfc2181.txt">RFC 2181</a> part 11 with
   * the fact that the encoding of each part occupies at least two bytes
   * (dot plus label externally, length byte plus label internally). Thus, if
   * all labels have the minimum size of one byte, 127 of them will fit.
   */
  private static final int MAX_PARTS = 127;

  /**
   * Maximum length of a full domain name, including separators, and
   * leaving room for the root label. See
   * <a href="http://www.ietf.org/rfc/rfc2181.txt">RFC 2181</a> part 11.
   */
  private static final int MAX_LENGTH = 253;

  /**
   * Maximum size of a single part of a domain name. See
   * <a href="http://www.ietf.org/rfc/rfc2181.txt">RFC 2181</a> part 11.
   */
  private static final int MAX_DOMAIN_PART_LENGTH = 63;

  /**
   * The full domain name, converted to lower case.
   */
  private final String name;

  /**
   * The parts of the domain name, converted to lower case.
   */
  private final ImmutableList<String> parts;

  /**
   * The index in the {@link #parts()} list at which the public suffix begins.
   * For example, for the domain name {@code www.google.co.uk}, the value would
   * be 2 (the index of the {@code co} part). The value is negative
   * (specifically, {@link #NO_PUBLIC_SUFFIX_FOUND}) if no public suffix was
   * found.
   */
  private final int publicSuffixIndex;

  /**
   * Constructor used to implement {@link #from(String)}, and from subclasses.
   */
  InternetDomainName(String name) {
    // Normalize:
    // * ASCII characters to lowercase
    // * All dot-like characters to '.'
    // * Strip trailing '.'

    name = Ascii.toLowerCase(DOTS_MATCHER.replaceFrom(name, '.'));

    if (name.endsWith(".")) {
      name = name.substring(0, name.length() - 1);
    }

    checkArgument(name.length() <= MAX_LENGTH,
        "Domain name too long: '%s':", name);
    this.name = name;

    this.parts = ImmutableList.copyOf(DOT_SPLITTER.split(name));
    checkArgument(parts.size() <= MAX_PARTS,
        "Domain has too many parts: '%s'", name);
    checkArgument(validateSyntax(parts), "Not a valid domain name: '%s'", name);

    this.publicSuffixIndex = findPublicSuffix();
  }

  /**
   * Returns the index of the leftmost part of the public suffix, or -1 if not
   * found. Note that the value defined as the "public suffix" may not be a
   * public suffix according to {@link #isPublicSuffix()} if the domain ends
   * with an excluded domain pattern such as {@code "nhs.uk"}.
   */
  private int findPublicSuffix() {
    final int partsSize = parts.size();

    for (int i = 0; i < partsSize; i++) {
      String ancestorName = DOT_JOINER.join(parts.subList(i, partsSize));

      if (PublicSuffixPatterns.EXACT.containsKey(ancestorName)) {
        return i;
      }

      // Excluded domains (e.g. !nhs.uk) use the next highest
      // domain as the effective public suffix (e.g. uk).

      if (PublicSuffixPatterns.EXCLUDED.containsKey(ancestorName)) {
        return i + 1;
      }

      if (matchesWildcardPublicSuffix(ancestorName)) {
        return i;
      }
    }

    return NO_PUBLIC_SUFFIX_FOUND;
  }

  /**
   * Returns an instance of {@link InternetDomainName} after lenient
   * validation.  Specifically, validation against <a
   * href="http://www.ietf.org/rfc/rfc3490.txt">RFC 3490</a>
   * ("Internationalizing Domain Names in Applications") is skipped, while
   * validation against <a
   * href="http://www.ietf.org/rfc/rfc1035.txt">RFC 1035</a> is relaxed in
   * the following ways:
   * <ul>
   * <li>Any part containing non-ASCII characters is considered valid.
   * <li>Underscores ('_') are permitted wherever dashes ('-') are permitted.
   * <li>Parts other than the final part may start with a digit.
   * </ul>
   *
   *
   * @param domain A domain name (not IP address)
   * @throws IllegalArgumentException if {@code name} is not syntactically valid
   *     according to {@link #isValid}
   * @since 10.0 (previously named {@code fromLenient})
   */
  public static InternetDomainName from(String domain) {
    return new InternetDomainName(checkNotNull(domain));
  }

  /**
   * Validation method used by {@from} to ensure that the domain name is
   * syntactically valid according to RFC 1035.
   *
   * @return Is the domain name syntactically valid?
   */
  private static boolean validateSyntax(List<String> parts) {
    final int lastIndex = parts.size() - 1;

    // Validate the last part specially, as it has different syntax rules.

    if (!validatePart(parts.get(lastIndex), true)) {
      return false;
    }

    for (int i = 0; i < lastIndex; i++) {
      String part = parts.get(i);
      if (!validatePart(part, false)) {
        return false;
      }
    }

    return true;
  }

  private static final CharMatcher DASH_MATCHER = CharMatcher.anyOf("-_");

  private static final CharMatcher PART_CHAR_MATCHER =
      CharMatcher.JAVA_LETTER_OR_DIGIT.or(DASH_MATCHER);

  /**
   * Helper method for {@link #validateSyntax(List)}. Validates that one part of
   * a domain name is valid.
   *
   * @param part The domain name part to be validated
   * @param isFinalPart Is this the final (rightmost) domain part?
   * @return Whether the part is valid
   */
  private static boolean validatePart(String part, boolean isFinalPart) {

    // These tests could be collapsed into one big boolean expression, but
    // they have been left as independent tests for clarity.

    if (part.length() < 1 || part.length() > MAX_DOMAIN_PART_LENGTH) {
      return false;
    }

    /*
     * GWT claims to support java.lang.Character's char-classification methods,
     * but it actually only works for ASCII. So for now, assume any non-ASCII
     * characters are valid. The only place this seems to be documented is here:
     * http://osdir.com/ml/GoogleWebToolkitContributors/2010-03/msg00178.html
     *
     * <p>ASCII characters in the part are expected to be valid per RFC 1035,
     * with underscore also being allowed due to widespread practice.
     */

    String asciiChars = CharMatcher.ASCII.retainFrom(part);

    if (!PART_CHAR_MATCHER.matchesAllOf(asciiChars)) {
      return false;
    }

    // No initial or final dashes or underscores.

    if (DASH_MATCHER.matches(part.charAt(0))
        || DASH_MATCHER.matches(part.charAt(part.length() - 1))) {
      return false;
    }

    /*
     * Note that we allow (in contravention of a strict interpretation of the
     * relevant RFCs) domain parts other than the last may begin with a digit
     * (for example, "3com.com"). It's important to disallow an initial digit in
     * the last part; it's the only thing that stops an IPv4 numeric address
     * like 127.0.0.1 from looking like a valid domain name.
     */

    if (isFinalPart && CharMatcher.DIGIT.matches(part.charAt(0))) {
      return false;
    }

    return true;
  }

  /**
   * Returns the individual components of this domain name, normalized to all
   * lower case. For example, for the domain name {@code mail.google.com}, this
   * method returns the list {@code ["mail", "google", "com"]}.
   */
  public ImmutableList<String> parts() {
    return parts;
  }

  /**
   * Indicates whether this domain name represents a <i>public suffix</i>, as
   * defined by the Mozilla Foundation's
   * <a href="http://publicsuffix.org/">Public Suffix List</a> (PSL). A public
   * suffix is one under which Internet users can directly register names, such
   * as {@code com}, {@code co.uk} or {@code pvt.k12.wy.us}. Examples of domain
   * names that are <i>not</i> public suffixes include {@code google}, {@code
   * google.com} and {@code foo.co.uk}.
   *
   * @return {@code true} if this domain name appears exactly on the public
   *     suffix list
   * @since 6.0
   */
  public boolean isPublicSuffix() {
    return publicSuffixIndex == 0;
  }

  /**
   * Indicates whether this domain name ends in a {@linkplain #isPublicSuffix()
   * public suffix}, including if it is a public suffix itself. For example,
   * returns {@code true} for {@code www.google.com}, {@code foo.co.uk} and
   * {@code com}, but not for {@code google} or {@code google.foo}. This is
   * the recommended method for determining whether a domain is potentially an
   * addressable host.
   *
   * @since 6.0
   */
  public boolean hasPublicSuffix() {
    return publicSuffixIndex != NO_PUBLIC_SUFFIX_FOUND;
  }

  /**
   * Returns the {@linkplain #isPublicSuffix() public suffix} portion of the
   * domain name, or {@code null} if no public suffix is present.
   *
   * @since 6.0
   */
  public InternetDomainName publicSuffix() {
    return hasPublicSuffix() ? ancestor(publicSuffixIndex) : null;
  }

  /**
   * Indicates whether this domain name ends in a {@linkplain #isPublicSuffix()
   * public suffix}, while not being a public suffix itself. For example,
   * returns {@code true} for {@code www.google.com}, {@code foo.co.uk} and
   * {@code bar.ca.us}, but not for {@code google}, {@code com}, or {@code
   * google.foo}.
   *
   * <p><b>Warning:</b> a {@code false} result from this method does not imply
   * that the domain does not represent an addressable host, as many public
   * suffixes are also addressable hosts. Use {@link #hasPublicSuffix()} for
   * that test.
   *
   * <p>This method can be used to determine whether it will probably be
   * possible to set cookies on the domain, though even that depends on
   * individual browsers' implementations of cookie controls. See
   * <a href="http://www.ietf.org/rfc/rfc2109.txt">RFC 2109</a> for details.
   *
   * @since 6.0
   */
  public boolean isUnderPublicSuffix() {
    return publicSuffixIndex > 0;
  }

  /**
   * Indicates whether this domain name is composed of exactly one subdomain
   * component followed by a {@linkplain #isPublicSuffix() public suffix}. For
   * example, returns {@code true} for {@code google.com} and {@code foo.co.uk},
   * but not for {@code www.google.com} or {@code co.uk}.
   *
   * <p><b>Warning:</b> A {@code true} result from this method does not imply
   * that the domain is at the highest level which is addressable as a host, as
   * many public suffixes are also addressable hosts. For example, the domain
   * {@code bar.uk.com} has a public suffix of {@code uk.com}, so it would
   * return {@code true} from this method. But {@code uk.com} is itself an
   * addressable host.
   *
   * <p>This method can be used to determine whether a domain is probably the
   * highest level for which cookies may be set, though even that depends on
   * individual browsers' implementations of cookie controls. See
   * <a href="http://www.ietf.org/rfc/rfc2109.txt">RFC 2109</a> for details.
   *
   * @since 6.0
   */
  public boolean isTopPrivateDomain() {
    return publicSuffixIndex == 1;
  }

  /**
   * Returns the portion of this domain name that is one level beneath the
   * public suffix. For example, for {@code x.adwords.google.co.uk} it returns
   * {@code google.co.uk}, since {@code co.uk} is a public suffix.
   *
   * <p>If {@link #isTopPrivateDomain()} is true, the current domain name
   * instance is returned.
   *
   * <p>This method should not be used to determine the topmost parent domain
   * which is addressable as a host, as many public suffixes are also
   * addressable hosts. For example, the domain {@code foo.bar.uk.com} has
   * a public suffix of {@code uk.com}, so it would return {@code bar.uk.com}
   * from this method. But {@code uk.com} is itself an addressable host.
   *
   * <p>This method can be used to determine the probable highest level parent
   * domain for which cookies may be set, though even that depends on individual
   * browsers' implementations of cookie controls.
   *
   * @throws IllegalStateException if this domain does not end with a
   *     public suffix
   * @since 6.0
   */
  public InternetDomainName topPrivateDomain() {
    if (isTopPrivateDomain()) {
      return this;
    }
    checkState(isUnderPublicSuffix(), "Not under a public suffix: %s", name);
    return ancestor(publicSuffixIndex - 1);
  }

  /**
   * Indicates whether this domain is composed of two or more parts.
   */
  public boolean hasParent() {
    return parts.size() > 1;
  }

  /**
   * Returns an {@code InternetDomainName} that is the immediate ancestor of
   * this one; that is, the current domain with the leftmost part removed. For
   * example, the parent of {@code www.google.com} is {@code google.com}.
   *
   * @throws IllegalStateException if the domain has no parent, as determined
   *     by {@link #hasParent}
   */
  public InternetDomainName parent() {
    checkState(hasParent(), "Domain '%s' has no parent", name);
    return ancestor(1);
  }

  /**
   * Returns the ancestor of the current domain at the given number of levels
   * "higher" (rightward) in the subdomain list. The number of levels must be
   * non-negative, and less than {@code N-1}, where {@code N} is the number of
   * parts in the domain.
   *
   * <p>TODO: Reasonable candidate for addition to public API.
   */
  private InternetDomainName ancestor(int levels) {
    return from(DOT_JOINER.join(parts.subList(levels, parts.size())));
  }

  /**
   * Creates and returns a new {@code InternetDomainName} by prepending the
   * argument and a dot to the current name. For example, {@code
   * InternetDomainName.from("foo.com").child("www.bar")} returns a new
   * {@code InternetDomainName} with the value {@code www.bar.foo.com}. Only
   * lenient validation is performed, as described {@link #from(String) here}.
   *
   * @throws NullPointerException if leftParts is null
   * @throws IllegalArgumentException if the resulting name is not valid
   */
  public InternetDomainName child(String leftParts) {
    return from(checkNotNull(leftParts) + "." + name);
  }

  /**
   * Indicates whether the argument is a syntactically valid domain name using
   * lenient validation. Specifically, validation against <a
   * href="http://www.ietf.org/rfc/rfc3490.txt">RFC 3490</a>
   * ("Internationalizing Domain Names in Applications") is skipped.
   *
   * <p>The following two code snippets are equivalent:
   *
   * <pre>   {@code
   *   domainName = InternetDomainName.isValid(name)
   *       ? InternetDomainName.from(name)
   *       : DEFAULT_DOMAIN;}</pre>
   *
   * <pre>   {@code
   *   try {
   *     domainName = InternetDomainName.from(name);
   *   } catch (IllegalArgumentException e) {
   *     domainName = DEFAULT_DOMAIN;
   *   }}</pre>
   *
   * @since 8.0 (previously named {@code isValidLenient})
   */
  public static boolean isValid(String name) {
    try {
      from(name);
      return true;
    } catch (IllegalArgumentException e) {
      return false;
    }
  }

  /**
   * Does the domain name match one of the "wildcard" patterns (e.g.
   * {@code "*.ar"})?
   */
  private static boolean matchesWildcardPublicSuffix(String domain) {
    final String[] pieces = domain.split(DOT_REGEX, 2);
    return pieces.length == 2 && PublicSuffixPatterns.UNDER.containsKey(pieces[1]);
  }

  /**
   * Returns the domain name, normalized to all lower case.
   */
  @Override
  public String toString() {
    return name;
  }

  /**
   * Equality testing is based on the text supplied by the caller,
   * after normalization as described in the class documentation. For
   * example, a non-ASCII Unicode domain name and the Punycode version
   * of the same domain name would not be considered equal.
   *
   */
  @Override
  public boolean equals(@Nullable Object object) {
    if (object == this) {
      return true;
    }

    if (object instanceof InternetDomainName) {
      InternetDomainName that = (InternetDomainName) object;
      return this.name.equals(that.name);
    }

    return false;
  }

  @Override
  public int hashCode() {
    return name.hashCode();
  }
}