From eee01170fdfcc4f5208f0f20aa6a8752c8c0fd7f Mon Sep 17 00:00:00 2001
From: Mark Davis <mark@unicode.org>
Date: Sat, 27 Jun 2026 12:09:20 +0200
Subject: [PATCH 1/8] ICU-23431 Add link detection/formatting tools

Very rough first cut
---
 .../icu/impl/links/LinkHandlingUtilities.java | 1130 +++++++++++++++++
 .../com/ibm/icu/impl/locale/XCldrStub.java    |   42 +
 .../java/com/ibm/icu/util/LinkUtilities.java  |   99 ++
 .../com/ibm/icu/dev/test/links/LinkTest.java  |   53 +
 .../data/testdata/links/LinkDetectionTest.txt |  432 +++++++
 .../testdata/links/LinkFormattingTest.txt     |  278 ++++
 6 files changed, 2034 insertions(+)
 create mode 100644 icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java
 create mode 100644 icu4j/main/core/src/main/java/com/ibm/icu/util/LinkUtilities.java
 create mode 100644 icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
 create mode 100644 icu4j/main/core/src/test/resources/com/ibm/icu/dev/data/testdata/links/LinkDetectionTest.txt
 create mode 100644 icu4j/main/core/src/test/resources/com/ibm/icu/dev/data/testdata/links/LinkFormattingTest.txt

diff --git a/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java b/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java
new file mode 100644
index 000000000000..e99420751518
--- /dev/null
+++ b/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java
@@ -0,0 +1,1130 @@
+package com.ibm.icu.impl.links;
+
+import com.ibm.icu.impl.IDNA2003;
+import com.ibm.icu.impl.UnicodeMap;
+import com.ibm.icu.impl.Utility;
+import com.ibm.icu.impl.locale.XCldrStub;
+import com.ibm.icu.impl.locale.XCldrStub.Splitter;
+import com.ibm.icu.lang.UCharacter;
+import com.ibm.icu.lang.UProperty;
+import com.ibm.icu.lang.UProperty.NameChoice;
+import com.ibm.icu.text.StringPrepParseException;
+import com.ibm.icu.text.UnicodeSet;
+import com.ibm.icu.text.UnicodeSet.EntryRange;
+import com.ibm.icu.text.UnicodeSet.SpanCondition;
+import com.ibm.icu.util.ICUException;
+import com.ibm.icu.util.Output;
+import java.nio.charset.StandardCharsets;
+import java.util.EnumMap;
+import java.util.EnumSet;
+import java.util.List;
+import java.util.Map;
+import java.util.Map.Entry;
+import java.util.NavigableMap;
+import java.util.Objects;
+import java.util.Set;
+import java.util.Stack;
+import java.util.function.Function;
+import java.util.regex.MatchResult;
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
+import java.util.stream.Collectors;
+
+public class LinkHandlingUtilities {
+
+    private static final UnicodeSet SOFAR = new UnicodeSet();
+
+    /**
+     * Based on https://url.spec.whatwg.org/#percent-encoded-bytes. Is the default for maximal
+     * encoding, but can be customized
+     */
+    private enum WHATWG_PERCENT_ENCODED {
+        /** C0 controls and all code points greater than U+007E (~). */
+        C0(null, "[\\u0000-\\u001F\\u007E-\\x{10FFFF}]"),
+
+        /** C0 and U+0020 SPACE, U+0022 ("), U+003C (<), U+003E (>), andU+0060 (`). */
+        FRAGMENT(C0, "[\\ \"<>`]"),
+
+        /** C0 and U+0020 SPACE, U+0022 ("), U+0023 (#), U+003C (<), and U+003E (>) */
+        QUERY(C0, "[\\ \"<>]"),
+
+        /** query percent-encode set and U+0027 ('). Used for http://... */
+        SPECIAL_QUERY(QUERY, "[']"),
+
+        /**
+         * query percent-encode set and<br>
+         * U+003F (?), U+005E (^), U+0060 (`), U+007B ({), and U+007D (}).
+         */
+        PATH(QUERY, "[?\\^`\\{\\}]"),
+
+        /**
+         * path and<br>
+         * U+002F (/), U+003A (:), U+003B (;), U+003D (=), U+0040 (@), U+005B ([) to U+005D (]),
+         * inclusive, and U+007C (|).
+         */
+        USERINFO(PATH, "[/:;=@\\[-\\]|]"),
+
+        /**
+         * userinfo and <br>
+         * U+0024 ($) to U+0026 (&), inclusive, U+002B (+), and U+002C (,).
+         */
+        COMPONENT(USERINFO, "[\\$-\\&+,]");
+
+        public final UnicodeSet set;
+
+        private WHATWG_PERCENT_ENCODED(WHATWG_PERCENT_ENCODED base, String uset) {
+            set = new UnicodeSet(uset).addAll(base == null ? UnicodeSet.EMPTY : base.set).freeze();
+        }
+    }
+
+    private static String quote(String s) {
+        return s; // later, escape for HTML
+    }
+
+    /** TODO use real property */
+    private static int getOpening(int cp) {
+        return cp == '>' ? '<' : UCharacter.getIntPropertyValue(cp, UProperty.BIDI_PAIRED_BRACKET);
+    }
+
+    /** Defines the LinkTermination property TODO use real property */
+    private enum LinkTermination {
+        Hard("[\\p{whitespace}\\p{NChar}[\\p{C}-\\p{Cf}]\\p{deprecated}]"),
+        Soft("[\\p{Term}\\p{lb=qu}-\\p{deprecated}]"),
+        Close("[\\p{Bidi_Paired_Bracket_Type=Close}[>]-\\p{deprecated}]"),
+        Open("[\\p{Bidi_Paired_Bracket_Type=Open}[<]-\\p{deprecated}]"),
+        Include(null), // all else
+        ;
+
+        public final UnicodeSet base;
+
+        public UnicodeSet getBase() {
+            return base;
+        }
+
+        private LinkTermination(String uset) {
+            if (uset == null) { // only called with Include, the "none of the above" option
+                this.base = SOFAR.complement().freeze();
+            } else {
+                java.text.ParsePosition parsePosition = new java.text.ParsePosition(0);
+                this.base = new UnicodeSet(uset).freeze();
+                SOFAR.addAll(this.base);
+            }
+        }
+
+        private static final UnicodeMap<LinkTermination> PROPERTY_MAP = new UnicodeMap<>();
+
+        static {
+            // Verify consistency
+            for (LinkTermination pv1 : values()) {
+                for (LinkTermination pv2 : values()) {
+                    if (pv1.compareTo(pv2) <= 0) {
+                        continue;
+                    }
+                    if (pv1.base.containsSome(pv2.base)) {
+                        throw new IllegalArgumentException(
+                                "Values in LinkTermination overlap! "
+                                        + pv1
+                                        + ", "
+                                        + pv2
+                                        + ": "
+                                        + new UnicodeSet(pv1.base).retainAll(pv2.base));
+                    }
+                }
+            }
+            for (LinkTermination lt : values()) {
+                PROPERTY_MAP.putAll(lt.base, lt);
+            }
+            PROPERTY_MAP.freeze();
+        }
+
+        private static final Set<LinkTermination> NON_MISSING =
+                XCldrStub.setsDifference(EnumSet.allOf(LinkTermination.class), Set.of(Hard));
+    }
+
+    // https://www.rfc-editor.org/rfc/rfc5322.html#section-3.2.3 has the full list for ASCII part
+    // See also https://en.wikipedia.org/wiki/Email_address#Local-part
+    // We add dot (ascii '.'), and then check after for the special dot constraints.
+
+    static final UnicodeSet EMAIL_ASCII_INCLUDES =
+            new UnicodeSet("[[a-zA-Z][0-9][_ \\- ! ? ' \\{ \\} * / \\& # % ` \\^ + = | ~ \\$]]")
+                    .add('.')
+                    .freeze();
+    static final UnicodeSet validEmailLocalPart =
+            new UnicodeSet("[\\p{XID_Continue}-\\p{block=basic_latin}]")
+                    .addAll(EMAIL_ASCII_INCLUDES)
+                    .freeze();
+
+    public static int scanEmailBackwards(CharSequence source, int hardStart, int domainStart) {
+        if (UCharacter.codePointBefore(source, domainStart) != '@') {
+            throw new IllegalArgumentException("Scanning must start after an '@' sign");
+        }
+        int result = validEmailLocalPart.spanBack(source, domainStart - 1, SpanCondition.SIMPLE);
+        return result >= hardStart ? result : domainStart - 1;
+    }
+
+    private static String getGeneralCategory(int property, int codePoint, int nameChoice) {
+        return UCharacter.getPropertyValueName(
+                property, UCharacter.getIntPropertyValue(codePoint, property), nameChoice);
+    }
+
+    private enum Structure {
+        single,
+        listP("/", "𝑷"),
+        listF(":~:", "𝑭"), // directive
+        listQ2("&", "𝑸", "=", "𝑽");
+        public final String sub;
+        public final String sub2;
+        public final String prefix;
+        public final String prefix2;
+
+        Structure() {
+            this(null, null, null, null);
+        }
+
+        Structure(String sub, String prefix) {
+            this(sub, prefix, null, null);
+        }
+
+        Structure(String sub, String prefix, String sub2, String prefix2) {
+            this.sub = sub;
+            this.sub2 = sub2;
+            this.prefix = prefix;
+            this.prefix2 = prefix2;
+        }
+    }
+
+    /** Parallels the spec parts table */
+    private enum Part {
+        // initiator, terminators, clearStack
+        PROTOCOL(Structure.single, '\u0000', "[{//}]", "[]", "[]", null),
+        HOST(Structure.single, '\u0000', "[/?#]", "[]", "[]", null),
+        PATH(Structure.listP, '/', "[?#]", "[/]", "[]", WHATWG_PERCENT_ENCODED.PATH),
+        QUERY(Structure.listQ2, '?', "[#]", "[=\\&]", "[+]", WHATWG_PERCENT_ENCODED.SPECIAL_QUERY),
+        FRAGMENT(
+                Structure.listF,
+                '#',
+                "[]",
+                "[{:~:}]",
+                "[]",
+                WHATWG_PERCENT_ENCODED.FRAGMENT), // the :~: is handled by code
+    // FRAGMENT_DIRECTIVE(Structure.listF, '#', "[]", "[{:~:}]", "[]"), // the :~: is handled by
+    // code
+    ;
+
+        static final Set<Part> ALL = XCldrStub.SetofOrdered(values());
+
+        static final int[] FRAGMENT_DIRECTIVE_STRING = ":~:".codePoints().toArray();
+
+        public final Structure structure;
+        final int initiator;
+        final UnicodeSet terminators;
+        final UnicodeSet clearStack;
+        final UnicodeSet extraQuoted;
+        final UnicodeSet whatwg_quoted;
+
+        private Part(
+                Structure structure,
+                char initiator,
+                String terminators,
+                String clearStack,
+                String extraQuoted,
+                WHATWG_PERCENT_ENCODED whatwg_quoted) {
+            this.structure = structure;
+            this.initiator = initiator;
+            this.terminators = new UnicodeSet(terminators).freeze();
+            this.clearStack = new UnicodeSet(clearStack).freeze();
+            this.extraQuoted =
+                    new UnicodeSet(extraQuoted)
+                            .addAll(this.clearStack)
+                            .addAll(this.terminators)
+                            .freeze();
+            this.whatwg_quoted = whatwg_quoted == null ? UnicodeSet.EMPTY : whatwg_quoted.set;
+        }
+
+        static Part fromInitiator(int cp) {
+            for (Part part : Part.values()) {
+                if (part.initiator == cp) {
+                    return part;
+                }
+            }
+            return null;
+        }
+
+        /**
+         * Unescape a part. But don't unescape interior characters or terminators because they are
+         * content! For example "a/b%2Fc" as a path should not be turned into a/b/c, because that
+         * b/c is a path-part.
+         *
+         * @param substring
+         * @return
+         */
+        private String unescape(String substring) {
+            return LinkHandlingUtilities.unescape(substring, extraQuoted);
+        }
+
+        private String fullEscape(List<List<String>> source) {
+            if (source.isEmpty()) {
+                return "";
+            }
+            if (structure == Structure.single) {
+                return source.get(0).get(0);
+            }
+            StringBuilder result = new StringBuilder().appendCodePoint(initiator);
+            UnicodeSet allEscape = whatwg_quoted;
+            // new UnicodeSet(0, 0x21, 0x7F, 0x10FFFF).addAll(extraQuoted).freeze();
+            boolean atStart = true;
+            for (List<String> list : source) {
+                if (atStart) {
+                    atStart = false;
+                } else {
+                    result.append(structure.sub);
+                }
+                if (structure.sub2 == null) {
+                    result.append(escape(list.get(0), allEscape));
+                } else {
+                    boolean atStart2 = true;
+                    for (String string : list) {
+                        if (atStart2) {
+                            atStart2 = false;
+                        } else {
+                            result.append(structure.sub2);
+                        }
+                        result.append(escape(string, allEscape));
+                    }
+                }
+            }
+            return result.toString();
+        }
+
+        /** Append an unescaped part, escaping as necessary */
+        private void appendPart(StringBuilder toAppendTo, String partValue) {
+            if (!partValue.isBlank()) {
+                if (initiator != 0) {
+                    toAppendTo.appendCodePoint(initiator);
+                }
+                toAppendTo.append(partValue);
+            }
+        }
+    }
+
+    public static class UrlInternals {
+        private NavigableMap<Part, List<List<String>>>
+                data; // note for some parts, the lists only ever have 1 member
+
+        /**
+         * Pull apart a URL string into Parts. <br>
+         * TODO: unescape the %escapes.
+         *
+         * @param source
+         * @param unescape TODO
+         * @return
+         */
+        public static UrlInternals from(String source) {
+            Map<Part, List<List<String>>> result = new EnumMap<>(Part.class);
+            // quick and dirty
+            int partStart = 0;
+            int partEnd;
+            main:
+            for (Part part : Part.values()) {
+                switch (part) {
+                    case PROTOCOL:
+                        partEnd = source.indexOf("://"); // TODO fix for mailto
+                        if (partEnd > 0) {
+                            partEnd += 3;
+                            result.put(
+                                    Part.PROTOCOL, List.of(List.of(source.substring(0, partEnd))));
+                            partStart = partEnd;
+                        }
+                        break;
+                    default:
+                        partEnd =
+                                part.terminators.span(
+                                        source, partStart, SpanCondition.NOT_CONTAINED);
+                        if (partStart != partEnd) {
+                            if (part != Part.HOST) {
+                                ++partStart;
+                            }
+                            String partString = source.substring(partStart, partEnd);
+                            if (part.structure.sub == null) { // only host, no unescaping
+                                result.put(part, List.of(List.of(partString)));
+                            } else if (part.structure.sub2 == null) { // we have just a list
+                                List<List<String>> cleanList = cleanList(part, partString);
+                                result.put(part, cleanList);
+                            } else { // we have a list of lists (query)
+                                List<List<String>> cleanList =
+                                        Splitter.on(part.structure.sub)
+                                                .splitToStream(partString)
+                                                .map(x -> splitList(part, part.structure.sub2, x))
+                                                .collect(Collectors.toList());
+                                result.put(part, cleanList);
+                            }
+                        }
+                        if (partEnd == source.length()) {
+                            break main;
+                        }
+                        partStart = partEnd;
+                        break;
+                }
+            }
+            return new UrlInternals(result);
+        }
+
+        private UrlInternals(Map<Part, List<List<String>>> data) {
+            this.data = XCldrStub.immutableSortedMap(data);
+        }
+
+        public enum EndStatus {
+            MEDIAL,
+            FINAL
+        }
+
+        /**
+         * Minimally escape. Presumes that the parts use \ for interior quoting.<br>
+         *
+         * @param endStatus TODO
+         * @param escapedCounter TODO
+         */
+        public String minimalEscape(EndStatus endStatus) {
+            StringBuilder output = new StringBuilder();
+            // get the last part
+            List<Entry<Part, List<List<String>>>> ordered = List.copyOf(data.entrySet());
+            Part lastPart = null;
+
+            for (Entry<Part, List<List<String>>> entry : ordered) {
+                if (!entry.getValue().isEmpty()) {
+                    lastPart = entry.getKey();
+                }
+            }
+            // process all parts
+            for (Entry<Part, List<List<String>>> partEntry : ordered) {
+                Part part = partEntry.getKey();
+                final List<List<String>> partParts = partEntry.getValue();
+                if (partParts.isEmpty()) {
+                    continue;
+                }
+                if (part.structure.sub == null) {
+                    output.append(partParts.get(0).get(0)); // just copy
+                    continue;
+                }
+                String unified =
+                        joinListListEscaping(part.structure.sub, part.structure.sub2, partParts);
+
+                int[] cps = unified.codePoints().toArray();
+                int n = cps.length;
+                if (cps[0] != part.initiator) {
+                    output.appendCodePoint(part.initiator);
+                }
+                ;
+                int copiedAlready = 0;
+                Stack<Integer> openingStack = new Stack<>();
+                for (int i = 0; i < n; ++i) {
+                    final int cp = cps[i];
+                    switch (cp) { // was just for test files
+                        case '\\': // if we have \ followed by x, just emit the literal x; otherwise
+                            // \
+                            // This is ONLY used for our test files;
+                            // in production the parts of the path/query
+                            // would be handled separately.
+
+                            // append soft code points
+                            appendCodePointsBetween(output, cps, copiedAlready, i);
+                            if (i < n - 1) {
+                                // append next code point
+                                ++i;
+                                appendPercentEscaped(output, cps[i]);
+                                copiedAlready = i + 1;
+                            } else {
+                                // append '\' alone (at end)
+                                appendPercentEscaped(output, cp);
+                                copiedAlready = i + 1;
+                            }
+                            continue;
+                            //
+                            //                        case '%': // if we have %xy, and x and y are
+                            // hex, escape the %
+                            //                            if (i < n - 2 && HEX.contains(cps[i + 1])
+                            // && HEX.contains(cps[i + 2])) {
+                            //                                // append soft code points
+                            //                                appendCodePointsBetween(output, cps,
+                            // copiedAlready, i);
+                            //                                appendPercentEscaped(output, cp,
+                            // escapedCounter);
+                            //                                copiedAlready = i + 1;
+                            //                                continue;
+                            //                            }
+                            //                            break;
+                    }
+                    LinkTermination lt =
+                            part.terminators.contains(cp)
+                                    ? LinkTermination.Hard
+                                    : LinkTermination.PROPERTY_MAP.get(cp);
+                    switch (lt) {
+                        case Include:
+                            appendCodePointsBetween(output, cps, copiedAlready, i);
+                            output.appendCodePoint(cp);
+                            copiedAlready = i + 1;
+                            break;
+                        case Hard:
+                            appendCodePointsBetween(output, cps, copiedAlready, i);
+                            appendPercentEscaped(output, cp);
+                            copiedAlready = i + 1;
+                            continue;
+                        case Soft: // fix
+                            continue;
+                        case Open:
+                            openingStack.push(cp);
+                            appendCodePointsBetween(output, cps, copiedAlready, i);
+                            output.appendCodePoint(cp);
+                            copiedAlready = i + 1;
+                            continue; // fix
+                        case Close: // fix
+                            if (openingStack.empty()) {
+                                appendCodePointsBetween(output, cps, copiedAlready, i);
+                                appendPercentEscaped(output, cp);
+                            } else {
+                                Integer topOfStack = openingStack.pop();
+                                int matchingOpening = getOpening(cp);
+                                if (matchingOpening == topOfStack) {
+                                    appendCodePointsBetween(output, cps, copiedAlready, i);
+                                    output.appendCodePoint(cp);
+                                } else { // failed to match
+                                    appendCodePointsBetween(output, cps, copiedAlready, i);
+                                    appendPercentEscaped(output, cp);
+                                }
+                            }
+                            copiedAlready = i + 1;
+                            continue;
+                        default:
+                            throw new IllegalArgumentException();
+                    }
+                }
+                if (endStatus == EndStatus.MEDIAL || part != lastPart) {
+                    appendCodePointsBetween(output, cps, copiedAlready, n);
+                } else if (copiedAlready < n) {
+                    appendCodePointsBetween(output, cps, copiedAlready, n - 1);
+                    appendPercentEscaped(output, cps[n - 1]);
+                }
+            }
+
+            return output.toString();
+        }
+
+        public String fullEscape() {
+            Output<Part> last = new Output<>();
+            String result =
+                    data.entrySet().stream()
+                            .map(
+                                    x -> {
+                                        Part part = x.getKey();
+                                        last.value = part;
+                                        List<List<String>> value = x.getValue();
+                                        return part.fullEscape(value);
+                                    })
+                            .collect(Collectors.joining());
+            // handle the very last character, if soft
+            if (last.value != null && last.value != Part.HOST && last.value != Part.PROTOCOL) {
+                int lastCodePoint = result.codePointBefore(result.length());
+                if (LinkTermination.Soft.base.contains(lastCodePoint)) {
+                    // escape final soft code point
+                    int indexBeforeLast = result.length() - Character.charCount(lastCodePoint);
+                    StringBuilder resultBuilder =
+                            new StringBuilder(result.substring(0, indexBeforeLast));
+                    appendPercentEscaped(resultBuilder, lastCodePoint);
+                    result = resultBuilder.toString();
+                }
+            }
+            return result;
+        }
+
+        public List<List<String>> get(Part part) {
+            return data.get(part);
+        }
+
+        @Override
+        public String toString() {
+            return toString(Part.ALL); // show all
+        }
+
+        @Override
+        public boolean equals(Object obj) {
+            if (obj instanceof UrlInternals) {
+                return data.equals(((UrlInternals) obj).data);
+            }
+            return false;
+        }
+
+        @Override
+        public int hashCode() {
+            return Objects.hash(data);
+        }
+
+        public String toString(Set<Part> partsToShow) {
+            StringBuilder result = new StringBuilder("{");
+            for (Entry<Part, List<List<String>>> partListList : data.entrySet()) {
+                Part part = partListList.getKey();
+                List<List<String>> listList = partListList.getValue();
+                int index1 = -1;
+                for (List<String> list1 : listList) {
+                    ++index1;
+                    int index2 = -1;
+                    for (String value : list1) {
+                        ++index2;
+                        if (result.length() != 1) {
+                            result.append(" ");
+                        }
+                        switch (part.structure) {
+                            case single:
+                                // we don't care about interior structure
+                                result.append(part == Part.PROTOCOL ? "𝑺" : "𝑯");
+                                break;
+                            case listP:
+                                result.append(part.structure.prefix);
+                                break;
+                            case listF:
+                                // Hack the difference between Fragment and FragmentDirective for
+                                // now
+                                result.append(index1 == 0 ? part.structure.prefix : "𝑫");
+                                break;
+                            case listQ2:
+                                result.append(
+                                        index2 == 0
+                                                ? part.structure.prefix
+                                                : part.structure.prefix2);
+                                break;
+                        }
+                        result.append('=');
+                        result.append(value);
+                    }
+                }
+            }
+            return result.append("}").toString();
+        }
+    }
+
+    private static String joinListListEscaping(
+            String sep, String sep2, List<List<String>> partParts) {
+        if (sep.codePointCount(0, sep.length()) == 0
+                || (sep2 != null && sep2.codePointCount(0, sep2.length()) == 0)) {
+            throw new IllegalArgumentException();
+        }
+        UnicodeSet toEscape = new UnicodeSet().add(sep);
+        if (sep2 != null) {
+            toEscape.add(sep2);
+        }
+        return partParts.stream()
+                .map(
+                        x -> {
+                            return joinListEscaping(sep2, toEscape, x);
+                        })
+                .collect(Collectors.joining(sep));
+    }
+
+    private static String joinListEscaping(
+            String sep, UnicodeSet toEscape, List<String> partParts) {
+        return sep == null
+                ? escape(partParts.get(0), toEscape)
+                : partParts.stream()
+                        .map(
+                                x -> {
+                                    return escape(x, toEscape);
+                                })
+                        .collect(Collectors.joining(sep));
+    }
+
+    private static List<List<String>> cleanList(Part part, String partString) {
+        return splitList(part, part.structure.sub, partString).stream()
+                .map(
+                        x -> {
+                            return List.of(x);
+                        })
+                .collect(Collectors.toUnmodifiableList());
+    }
+
+    private static List<String> splitList(Part part, String separator, String partString) {
+        return Splitter.on(separator)
+                .splitToStream(partString)
+                .map(x -> unescape(x))
+                .collect(Collectors.toList());
+    }
+
+    //    private static final UnicodeSet idnMapped =
+    //            IUP.getSet("Idn_Status=" + Idn_Status_Values.mapped);
+    //
+    //    private static final UnicodeSet idnValid = IUP.getSet("Idn_Status=" +
+    // Idn_Status_Values.valid);
+    //    static final Pattern protocol = Pattern.compile("(https?://|mailto:)");
+    //    public static final UnicodeSet validHost =
+    //            new UnicodeSet(idnValid)
+    //                    .addAll(idnMapped)
+    //                    .removeAll(new UnicodeSet("[:Block=Basic_Latin:]"))
+    //                    .addAll(new UnicodeSet("[-a-zA-Z0-9.．。]")) // the three dots are
+    // different:
+    //                    .freeze();
+    //    public static final UnicodeSet validHostNoDot =
+    //            new UnicodeSet(validHost).removeAll(".．。").freeze();
+
+    /** Status of link found. Note that if the link is imputed, https or mailto will be returned. */
+    enum LinkStatus {
+        bogus,
+        http,
+        https,
+        mailto,
+        tld
+    }
+
+    /** Immutable class for returning the start/end of link that is found, plus some information */
+    private static final class LinkFound {
+        public final int start;
+        public final int limit;
+        public final LinkStatus linkStatus;
+
+        public LinkFound(int start, int limit, LinkStatus linkStatus) {
+            this.start = start;
+            this.limit = limit;
+            this.linkStatus = linkStatus;
+        }
+
+        public String substring(String source) {
+            return source.substring(start, limit);
+        }
+    }
+
+    // OLDER Code, leaving here for now, for comparison
+    //    /**
+    //     * Parses a restricted set of URLs, for testing the PathQueryFragment portion. That is, it
+    // is of
+    //     * the form &lt;host>&lt;domain_name><pathqueryfragment>? <br>
+    //     * The host is currently just http, https, and mailto. <br>
+    //     * The domain_name is just approximated for testing.
+    //     *
+    //     * @return null if we run out of string, otherwise a LinkFound. If what is found is
+    // malformed in
+    //     *     some way, indicate with LinkFound.bogus.
+    //     */
+    //    private static LinkFound parseLink(String source, int startCodePointOffset) {
+    //        Matcher findStartMatcher = protocol.matcher(source);
+    //
+    //        findStartMatcher.region(startCodePointOffset, source.length());
+    //        if (!findStartMatcher.find()) {
+    //            return null; // we are at the end
+    //        }
+    //        // if we found something, and there was a character before it,
+    //        // and that character was not soft, then exit
+    //        int start = findStartMatcher.start();
+    //        int protocolEnd = findStartMatcher.end();
+    //        if (start != 0) {
+    //            LinkTermination lt =
+    //                    LinkTermination.PROPERTY_MAP.get(UCharacter.codePointBefore(source,
+    // start));
+    //            if (lt == LinkTermination.Include) {
+    //                return new LinkFound(start, protocolEnd, LinkStatus.bogus);
+    //            }
+    //        }
+    //
+    //        String protocolValue = findStartMatcher.group(1);
+    //        if (protocolValue.equals("mailto")) {
+    //            return parseRestOfMailto(source, start, protocolEnd);
+    //        }
+    //
+    //        // dumb search for end of host, doesn't handle .. or edge cases, but this does not
+    // have to
+    //        // be production-quality
+    //        int hostLimit = findEndOfDomain(source, protocolEnd);
+    //        if (protocolEnd == hostLimit) {
+    //            return new LinkFound(start, protocolEnd, LinkStatus.bogus);
+    //        }
+    //        int limit = parsePathQueryFragment(source, hostLimit);
+    //        return new LinkFound(
+    //                start, limit, LinkStatus.valueOf(source.substring(start, protocolEnd)));
+    //    }
+    //
+    //    private static LinkFound parseRestOfMailto(String source, int start, int hostStart) {
+    //        // basic implementation at first
+    //        int atPosition = source.indexOf('@', hostStart);
+    //        if (atPosition == -1) {
+    //            return new LinkFound(start, hostStart, LinkStatus.bogus);
+    //        }
+    //        // TBD we could be in the middle of a quoted string, check for that later
+    //
+    //        // see if what is in front of the @ looks ok
+    //
+    //        int limit = parsePathQueryFragment(source, atPosition + 1);
+    //        return new LinkFound(start, limit, LinkStatus.mailto);
+    //    }
+    //
+    //    // Simple implementation for testing, since the spec doesn't define the content
+    //    private static int findEndOfDomain(String source, int protocolLimit) {
+    //        return validHost.span(source, protocolLimit, SpanCondition.CONTAINED);
+    //    }
+
+    /**
+     * Set lastSafe to 0 — this marks the last code point that is definitely included in the
+     * linkification.<br>
+     * Set closingStack to empty<br>
+     * Set the current code point position i to 0<br>
+     * Loop from i = 0 to n<br>
+     * Set LT to LinkTermination(cp[i])<br>
+     * If LT == none, set lastSafe to be i+1, continue loop<br>
+     * If LT == soft, continue loop<br>
+     * If LT == hard, stop linkification and return lastSafe<br>
+     * If LT == opening, push cp[i] onto closingStack<br>
+     * If LT == closing, set open to the pop of closingStack, or 0 if the closingStack is empty<br>
+     * If LinkPairedOpeners(cp[i]) == open, set lastSafe to be i+1, continue loop.<br>
+     * Otherwise, stop linkification and return lastSafe<br>
+     * If lastSafe == n+1, then the entire part is safe; continue to the next part<br>
+     * Otherwise, stop linkification and return lastSafe<br>
+     */
+    public static int parsePathQueryFragment(String source, int codePointOffset) {
+        // For simplicity, and to match the spec, we just get the code points
+        // Production code would be optimized, of course.
+
+        int[] codePoints = source.codePoints().toArray();
+        int lastSafe = codePointOffset;
+        Part part = null;
+
+        Stack<Integer> openingStack = new Stack<>();
+        LinkTermination lt = LinkTermination.Soft;
+        for (int i = codePointOffset; i < codePoints.length; ++i) {
+            int cp = codePoints[i];
+            if (part == null) {
+                part = Part.fromInitiator(cp);
+                if (part == null) {
+                    return i; // failed, don't move cursor
+                }
+                lastSafe = i + 1;
+                continue;
+            }
+
+            lt = LinkTermination.PROPERTY_MAP.get(cp);
+            // handle terminators and interior syntax
+            if (part.terminators.contains(cp)) {
+                lastSafe = i;
+                part = Part.fromInitiator(cp);
+                if (part == null) {
+                    return lastSafe;
+                }
+                lastSafe = i + 1;
+                continue;
+            } else if (part.clearStack.contains(cp)) {
+                openingStack.clear();
+            } else if (part == Part.FRAGMENT
+                    && matches(codePoints, i, Part.FRAGMENT_DIRECTIVE_STRING)) {
+                // there is one string form, so hard-code it
+                openingStack.clear();
+            }
+            switch (lt) {
+                case Include:
+                    lastSafe = i + 1;
+                    break;
+                case Soft: // no action
+                    break;
+                case Hard:
+                    return lastSafe;
+                case Open:
+                    // We're not testing for stack limits; not needed for test data
+                    openingStack.push(cp);
+                    lastSafe = i + 1;
+                    break;
+                case Close:
+                    if (openingStack.empty()) {
+                        return lastSafe;
+                    }
+                    int matchingOpening = getOpening(cp);
+                    Integer topOfStack = openingStack.pop();
+                    if (matchingOpening == topOfStack) {
+                        lastSafe = i + 1;
+                        break;
+                    } // else failed to match
+                    return lastSafe;
+            }
+        }
+        // if we hit the end, it acts like we hit a hard character ***
+        return lt == LinkTermination.Soft ? lastSafe : codePoints.length;
+    }
+
+    /** Simple utility for finding matching int array regions */
+    private static boolean matches(int[] codePoints, int cpIndex, int[] fragmentDirective) {
+        if (cpIndex + fragmentDirective.length > codePoints.length) {
+            return false;
+        }
+        for (int i = 0; i < fragmentDirective.length; ++i) {
+            if (codePoints[i + cpIndex] != fragmentDirective[i]) {
+                return false;
+            }
+        }
+        return true;
+    }
+
+    private static void appendCodePointsBetween(
+            StringBuilder output, int[] cp, int copyEnd, int notToCopy) {
+        for (int i = copyEnd; i < notToCopy; ++i) {
+            output.appendCodePoint(cp[i]);
+        }
+    }
+
+    /** Regex for percent escaping */
+    private static final Pattern ESCAPED_BYTE_PATTERN =
+            Pattern.compile("(%[a-fA-F0-9][a-fA-F0-9])+");
+
+    private static final Pattern ESCAPED_SINGLE = Pattern.compile("%[a-fA-F0-9][a-fA-F0-9]");
+
+    /** Unescape a string. */
+    private static String unescape(String stringWithEscapes) {
+        return unescape(stringWithEscapes, UnicodeSet.EMPTY); // no characters escaped bak
+    }
+
+    /** Unescape a string; however, code points in toEscape are escaped back. */
+    private static String unescape(String stringWithEscapes, UnicodeSet toEscape) {
+        StringBuilder result = new StringBuilder();
+        Matcher matcher = ESCAPED_BYTE_PATTERN.matcher(stringWithEscapes);
+        int current = 0;
+        while (matcher.find(current)) {
+            result.append(
+                    stringWithEscapes.substring(
+                            current, matcher.start())); // append intervening text
+            String unescaped = percentUnescape(matcher.group());
+            unescaped
+                    .chars()
+                    .forEach(
+                            x -> {
+                                if (toEscape.contains(x)) {
+                                    // quote it
+                                    appendPercentEscaped(result, x);
+                                } else {
+                                    result.appendCodePoint(x);
+                                }
+                            });
+            current = matcher.end();
+        }
+        result.append(stringWithEscapes.substring(current, stringWithEscapes.length()));
+        return result.toString();
+    }
+
+    private static void appendPercentEscaped(StringBuilder output, int cp) {
+        byte[] bytes = Character.toString(cp).getBytes(StandardCharsets.UTF_8);
+        for (int i = 0; i < bytes.length; ++i) {
+            output.append('%');
+            output.append(Utility.hex(bytes[i], 2));
+        }
+    }
+
+    private static void appendPercentEscaped(StringBuilder output, String source) {
+        byte[] bytes = source.getBytes(StandardCharsets.UTF_8);
+        for (int i = 0; i < bytes.length; ++i) {
+            output.append('%');
+            output.append(Utility.hex(bytes[i], 2));
+        }
+    }
+
+    /** percent-escape single char or string. */
+    private static String escape(String source, String cpToEscape) {
+        return escape(source, new UnicodeSet().add(cpToEscape));
+    }
+
+    /** percent-escape all the toEscape characters. */
+    private static String escape(String source, UnicodeSet toEscape) {
+        StringBuilder result = new StringBuilder();
+        int lastEnd = 0;
+        while (true) {
+            int start = toEscape.span(source, lastEnd, SpanCondition.NOT_CONTAINED);
+            appendCheckingPercent(result, source, lastEnd, start);
+            if (start == source.length()) {
+                break;
+            }
+            lastEnd = toEscape.span(source, start, SpanCondition.SIMPLE);
+            appendPercentEscaped(result, source.substring(start, lastEnd));
+            if (lastEnd == source.length()) {
+                break;
+            }
+        }
+        return result.toString();
+    }
+
+    /** If we have in an unescaped string %xy, and xy are hex, then we have to escape the % sign. */
+    private static void appendCheckingPercent(
+            StringBuilder toAppendTo, String source, int lastEnd, int start) {
+        int pos = source.indexOf("%", lastEnd); // quickcheck
+        if (pos < 0 || pos >= start) {
+            toAppendTo.append(source, lastEnd, start);
+            return;
+        }
+        String sub = source.substring(lastEnd, start);
+        Matcher matcher = ESCAPED_SINGLE.matcher(sub);
+        Function<MatchResult, String> replacer =
+                x -> {
+                    return "%25"
+                            + x.group(0)
+                                    .substring(1); // separate the % from the xy, and return %25xy
+                };
+        String fixed = matcher.replaceAll(replacer);
+        toAppendTo.append(fixed);
+    }
+
+    /** We are guaranteed that string is all percent escaped utf8, %a3%c0 ... */
+    private static String percentUnescape(String escapedSource) {
+        byte[] temp = new byte[escapedSource.length() / 3];
+        int tempOffset = 0;
+        for (int i = 0; i < escapedSource.length(); i += 3) {
+            if (escapedSource.charAt(i) != '%') {
+                throw new IllegalArgumentException();
+            }
+            byte b = (byte) Integer.parseInt(escapedSource.substring(i + 1, i + 3), 16);
+            temp[tempOffset++] = b;
+        }
+        return new String(temp, StandardCharsets.UTF_8);
+    }
+
+    /**
+     * The wikipedia languages are not all BCP47. Convert the ones that are not. See:<br>
+     * https://meta.wikimedia.org/wiki/Special_language_codes <br>
+     * https://meta.wikimedia.org/wiki/List_of_Wikipedias#Nonstandard_language_codes
+     *
+     * @param languageCode
+     * @return
+     */
+    private static String fixWiki(String languageCode) {
+        switch (languageCode) {
+            case "als":
+                return "gsw";
+            case "roa-rup":
+                return "rup";
+            case "bat-smg":
+                return "sgs";
+            case "simple":
+                return "en";
+            case "fiu-vro":
+                return "vro";
+            case "zh-classical":
+                return "lzh";
+            case "zh-min-nan":
+                return "nan";
+            case "zh-yue":
+                return "yue";
+            case "cbk-zam":
+                return "cbk";
+            case "map-bms":
+                return "map";
+            case "nrm":
+                return "nrf";
+            case "roa-tara":
+                return "nap";
+            default:
+                return languageCode;
+        }
+    }
+
+    /** Hard-coded list of wikilanguages, for testing */
+    //    private static Set<String> WIKI_LANGUAGES =
+    //            of(
+    //                    SPLIT_COMMA.splitToList(
+    //
+    // "en,ceb,de,fr,sv,nl,ru,es,it,pl,arz,zh,ja,uk,vi,war,ar,pt,fa,ca,id,sr,ko,no,tr,ce,fi,cs,hu,tt,ro,sh,eu,zh-min-nan,ms,he,eo,hy,da,bg,uz,cy,simple,sk,et,be,azb,el,kk,min,hr,lt,gl,ur,az,sl,lld,ka,nn,ta,th,hi,bn,mk,zh-yue,la,ast,lv,af,tg,my,te,sq,mr,mg,bs,oc,be-tarask,ku,br,sw,ml,nds,ky,lmo,jv,pnb,ckb,new,ht,vec,pms,lb,ba,su,ga,is,szl,cv,pa,fy,io,ha,tl,an,mzn,wuu,diq,vo,ig,yo,sco,kn,ne,als,gu,ia,avk,crh,bar,ban,scn,bpy,mn,qu,nv,si,xmf,frr,ps,os,or,tum,sd,bcl,bat-smg,sah,cdo,gd,bug,glk,yi,ilo,am,li,nap,gor,as,fo,mai,hsb,map-bms,shn,zh-classical,eml,ace,ie,wa,sa,hyw,sat,zu,sn,mhr,lij,hif,km,bjn,mrj,mni,dag,ary,hak,pam,rue,roa-tara,ug,zgh,bh,nso,co,tly,so,vls,nds-nl,mi,se,myv,rw,kaa,sc,bo,kw,vep,mt,tk,mdf,kab,gv,gan,fiu-vro,ff,zea,ab,skr,smn,ks,gn,frp,pcd,udm,kv,csb,ay,nrm,lo,ang,fur,olo,lfn,lez,ln,pap,nah,mwl,tw,stq,rm,ext,lad,gom,dty,av,tyv,koi,dsb,lg,cbk-zam,dv,ksh,za,bxr,blk,gag,pfl,bew,szy,haw,tay,pag,pi,awa,tcy,krc,inh,gpe,xh,kge,fon,atj,to,pdc,mnw,arc,shi,om,tn,dga,ki,nia,jam,kbp,wo,xal,nov,kbd,anp,nqo,bi,kg,roa-rup,tpi,tet,guw,jbo,mad,fj,lbe,kcg,pcm,cu,ty,trv,dtp,sm,ami,st,iba,srn,btm,alt,ltg,gcr,ny,kus,mos,ss,chr,ee,ts,got,bbc,gur,bm,pih,ve,rmy,fat,chy,rn,igl,ik,guc,ch,ady,pnt,iu,ann,rsk,pwn,dz,ti,sg,din,tdd,kl,bdr,nr,cr"));
+
+    /** Show termination values, for generating property files */
+    private void showLinkTermination() {
+        for (LinkTermination lt : LinkTermination.values()) {
+            UnicodeSet value = LinkTermination.PROPERTY_MAP.getSet(lt);
+            String name = lt.toString();
+            System.out.println("\n#\tLink_Termination=" + name);
+            if (lt == LinkTermination.Include) {
+                System.out.println("#   " + "(All code points without other values)");
+                continue;
+            } else {
+                System.out.println("#   draft = " + lt.base);
+            }
+            if (lt == LinkTermination.Hard) {
+                value.removeAll(new UnicodeSet("[\\p{Cn}\\p{Cs}]"));
+                System.out.println("#   (not listing Unassigned or Surrogates)");
+            }
+            System.out.println();
+            for (EntryRange range : value.ranges()) {
+                final String rangeString =
+                        Utility.hex(range.codepoint)
+                                + (range.codepoint == range.codepointEnd
+                                        ? ""
+                                        : ".." + Utility.hex(range.codepointEnd));
+                System.out.println(
+                        rangeString
+                                + ";"
+                                + " ".repeat(15 - rangeString.length())
+                                + lt
+                                + "\t# "
+                                + "("
+                                + getGeneralCategory(
+                                        UProperty.GENERAL_CATEGORY,
+                                        range.codepoint,
+                                        NameChoice.SHORT)
+                                + ") "
+                                + quote(UCharacter.getExtendedName(range.codepoint))
+                                + (range.codepoint == range.codepointEnd
+                                        ? ""
+                                        : ".."
+                                                + "("
+                                                + getGeneralCategory(
+                                                        UProperty.GENERAL_CATEGORY,
+                                                        range.codepointEnd,
+                                                        NameChoice.SHORT)
+                                                + ") "
+                                                + quote(
+                                                        UCharacter.getExtendedName(
+                                                                range.codepointEnd))));
+            }
+            System.out.println();
+        }
+    }
+
+    /** Show paired openers, for generating property files */
+    private void showLinkPairedOpeners() {
+        UnicodeSet value = LinkTermination.PROPERTY_MAP.getSet(LinkTermination.Close);
+
+        System.out.println("\n#\tLink_Paired_Opener");
+        System.out.println(
+                "#   draft = BidiPairedBracket + (“&gt;” GREATER-THAN SIGN 🡆  “&lt;” LESS-THAN SIGN)");
+        System.out.println();
+
+        for (String cpString : value) {
+            int cp = cpString.codePointAt(0);
+            String hex = Utility.hex(cp);
+            final int value2 = getOpening(cp);
+            System.out.println(
+                    hex
+                            + ";"
+                            + " ".repeat(7 - hex.length())
+                            + Utility.hex(value2)
+                            + "\t#"
+                            + " “"
+                            + quote(Character.toString(cp))
+                            + "” "
+                            + UCharacter.getExtendedName(cp)
+                            + " 🡆 "
+                            + " “"
+                            + quote(Character.toString(value2))
+                            + "” "
+                            + UCharacter.getExtendedName(value2));
+        }
+    }
+
+    /**
+     * Regex to scan for possible TLDs. The result needs to be checked that there is no
+     * validHostNoDot before and after
+     */
+    //    private static final Pattern TLD_SCANNER;
+    //
+    //    private static final SortedSet<String> TLDS;
+    //
+    private static final String DOTSET_STRING = "[.。]";
+
+    private static final UnicodeSet DOTSET = new UnicodeSet("[.。]").freeze();
+    private static final Splitter SPLIT_LABELS = Splitter.on(Pattern.compile("[.。]"));
+
+    private static String toUnicode2(String x) {
+        try {
+            if (x.startsWith("XN--")) {
+                x = IDNA2003.convertIDNToUnicode(x, 0).toString();
+            }
+            return x;
+        } catch (StringPrepParseException e) {
+            throw new ICUException(e);
+        }
+    }
+}
diff --git a/icu4j/main/core/src/main/java/com/ibm/icu/impl/locale/XCldrStub.java b/icu4j/main/core/src/main/java/com/ibm/icu/impl/locale/XCldrStub.java
index 527be9217823..c57c6228ef7d 100644
--- a/icu4j/main/core/src/main/java/com/ibm/icu/impl/locale/XCldrStub.java
+++ b/icu4j/main/core/src/main/java/com/ibm/icu/impl/locale/XCldrStub.java
@@ -13,6 +13,7 @@
 import java.util.Arrays;
 import java.util.Collection;
 import java.util.Collections;
+import java.util.Comparator;
 import java.util.HashMap;
 import java.util.HashSet;
 import java.util.Iterator;
@@ -21,11 +22,14 @@
 import java.util.List;
 import java.util.Map;
 import java.util.Map.Entry;
+import java.util.NavigableMap;
 import java.util.Set;
 import java.util.TreeMap;
 import java.util.TreeSet;
 import java.util.regex.Matcher;
 import java.util.regex.Pattern;
+import java.util.stream.Collectors;
+import java.util.stream.Stream;
 
 /** Stub class to make migration easier until we get either Guava or a higher level of Java. */
 public class XCldrStub {
@@ -314,6 +318,10 @@ public Splitter(char c) {
             this(Pattern.compile("\\Q" + c + "\\E"));
         }
 
+        public Splitter(String s) {
+            this(Pattern.compile("\\Q" + s + "\\E"));
+        }
+
         public Splitter(Pattern p) {
             pattern = p;
         }
@@ -322,6 +330,10 @@ public static Splitter on(char c) {
             return new Splitter(c);
         }
 
+        public static Splitter on(String sub) {
+            return new Splitter(sub);
+        }
+
         public static Splitter on(Pattern p) {
             return new Splitter(p);
         }
@@ -336,6 +348,10 @@ public List<String> splitToList(String input) {
             return Arrays.asList(items);
         }
 
+        public Stream<String> splitToStream(String input) {
+            return splitToList(input).stream();
+        }
+
         public Splitter trimResults() {
             trimResults = true;
             return this;
@@ -458,4 +474,30 @@ public interface Predicate<T> {
          */
         boolean test(T t);
     }
+
+    /** Utilities because regular java doesn't have some nice Guava features */
+    public static <T> Set<T> SetofOrdered(T... source) {
+        return Arrays.stream(source)
+                .collect(
+                        Collectors.collectingAndThen(
+                                Collectors.toCollection(LinkedHashSet::new), Set::copyOf));
+    }
+
+    /** Not as good as Guava, since it makes a new set instead of a view */
+    public static <T> Set<T> setsDifference(Set<T> set1, Set<T> set2) {
+        return set1.stream()
+                .filter(element -> !set2.contains(element))
+                .collect(
+                        Collectors.collectingAndThen(
+                                Collectors.toCollection(LinkedHashSet::new), Set::copyOf));
+    }
+
+    public static <K, V> NavigableMap<K, V> immutableSortedMap(Map<K, V> existingMap) {
+        return Collections.unmodifiableNavigableMap(new TreeMap<>(existingMap));
+    }
+
+    public static final Comparator<String> LONGEST_FIRST =
+            Comparator.<String>comparingInt(x -> x.length())
+                    .reversed()
+                    .thenComparing(Comparator.naturalOrder());
 }
diff --git a/icu4j/main/core/src/main/java/com/ibm/icu/util/LinkUtilities.java b/icu4j/main/core/src/main/java/com/ibm/icu/util/LinkUtilities.java
new file mode 100644
index 000000000000..74eda248f349
--- /dev/null
+++ b/icu4j/main/core/src/main/java/com/ibm/icu/util/LinkUtilities.java
@@ -0,0 +1,99 @@
+package com.ibm.icu.util;
+
+import com.ibm.icu.impl.links.LinkHandlingUtilities;
+import com.ibm.icu.impl.links.LinkHandlingUtilities.UrlInternals;
+import com.ibm.icu.impl.links.LinkHandlingUtilities.UrlInternals.EndStatus;
+import com.ibm.icu.text.UnicodeSet;
+
+/**
+ * Utility class for assisting with detecting links (URLs or emails) in text, and formatting them
+ * for display, implementing the algorithms in https://www.unicode.org/reports/tr58/ to handle
+ * Unicode characters properly. It supplies lower level APIs for use in augmenting existing scanners
+ * and formatters.
+ */
+public class LinkUtilities {
+
+    /**
+     * Lower level utility for finding the end of a PathQueryFragment (PQF) in text. It assumes that
+     * the start position is immediately after an identified domain name. The purpose of this
+     * routine is for fitting into algorithms that are already in use, just taking over for the PQF
+     * scanning. For more information, see https://www.unicode.org/reports/tr58/.
+     *
+     * @param source the text to be scanned
+     * @param start the position in the text to be scanned from. It should be immediately after a
+     *     domain name.
+     * @return the end position of the PQF, or the start value if there is none.
+     */
+    public static int scanPathQueryFragment(CharSequence source, int start, int limit) {
+        return LinkHandlingUtilities.parsePathQueryFragment(source.toString(), start);
+    }
+
+    /**
+     * Lower level utility for finding the start of an email address in text. It assumes that the
+     * limit position is immediately before an '@' + identified domain name. The purpose of this
+     * routine is for fitting into algorithms that are already in use, just handling for the email
+     * `local-part`.
+     *
+     * @param source the text to be scanned
+     * @param start the position that is the earliest that should be considered in a backwards scan
+     * @param limit the position to start scanning backwards from — should be just after @ and just
+     *     before the domain_name.
+     * @return the start of the email locale part, or limit if no email local part is found
+     */
+    public static int scanBackEmailLocalPart(CharSequence s, int start, int limit) {
+        return LinkHandlingUtilities.scanEmailBackwards(s, start, limit);
+    }
+
+    /** Enum for determining whether any percent-escaping is minimal or maximal, for use */
+    public enum Extent {
+        /** Minimal percent-escaping only percent-escapes non-ASCII where necessary. */
+        MINIMAL,
+        /** Maximal percent-escaping percent-escapes all non-ASCII. */
+        MAXIMAL
+    }
+
+    /**
+     * Escapes a URL according to the Extent parameter
+     *
+     * @param source In the source, it is assumed that ASCII syntax characters requiring escaping
+     *     have already been escaped. For example, a literal / in a path segment would already be
+     *     percent-escaped. For more information, see https://www.unicode.org/reports/tr58/.
+     * @param extent either MINIMAL or MAXIMAL
+     * @return an escaped string according to the extent parameter.
+     */
+    public static String escapePathQueryFragment(String source, Extent extent) {
+        UrlInternals ui = UrlInternals.from(source.toString());
+        switch (extent) {
+            case MINIMAL:
+                return ui.minimalEscape(EndStatus.FINAL);
+            case MAXIMAL:
+                return ui.fullEscape();
+            default:
+                throw new InternalError();
+        }
+    }
+
+    // NOTE for reviewers: These are only temporary, until ICU supplies \p{Link_Term=hard} and
+    // \p{Idn_Status≠disallowed}
+
+    /**
+     * Returns a frozen set of Unicode characters that are guaranteed to never be part of a URL or
+     * email address. This allows implementations to make various optimizations because URLs and
+     * email addresses can never span these characters. For example, a span of characters between
+     * safe characters that doesn't have a sequence of domain-character + . + domain-character can
+     * be skipped in processing.
+     */
+    public static UnicodeSet getSafeCharacters() {
+        return null;
+    }
+
+    /**
+     * Returns a frozen set of Unicode characters that are possible characters in a domain name
+     * (pre-mapping) This allows implementations to make various optimizations because URLs and
+     * email addresses must contain a sequence of domain-character + . + domain-character. It is the
+     * same as the set of IDNA Mapping Table character with values ≠ disallowed
+     */
+    public static UnicodeSet getDomainCharacters() {
+        return null;
+    }
+}
diff --git a/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java b/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
new file mode 100644
index 000000000000..063dcd429a90
--- /dev/null
+++ b/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
@@ -0,0 +1,53 @@
+package com.ibm.icu.dev.test.links;
+
+import com.ibm.icu.dev.test.CoreTestFmwk;
+import com.ibm.icu.util.LinkUtilities;
+import com.ibm.icu.util.LinkUtilities.Extent;
+import org.junit.Test;
+import org.junit.runner.RunWith;
+import org.junit.runners.JUnit4;
+
+@RunWith(JUnit4.class)
+public class LinkTest extends CoreTestFmwk {
+
+    /** Constructor */
+    public LinkTest() {}
+
+    @Test
+    public void testSimpleScan() {
+        String source = "/αβγ?def#ghi ";
+        int expected = source.indexOf(" ");
+        int actual = LinkUtilities.scanPathQueryFragment(source, 0, source.length());
+        assertEquals(source, expected, actual);
+    }
+
+    @Test
+    public void testSimpleEscape() {
+        String source = "/%CE%B1%CE%B2%CE%B3/%CE%B4.%CE%B5%CE%B6%2E";
+        String expected = "/αβγ/δ.εζ%2E";
+        String actual = LinkUtilities.escapePathQueryFragment(source, Extent.MINIMAL);
+        assertEquals(source, expected, actual);
+    }
+
+    @Test
+    public void testSimpleScanBackEmail() {
+        String source = "See αβγ@example.com";
+        int start = 0;
+        int end = source.indexOf('@') + 1;
+        int expected = source.indexOf("αβγ");
+        int actual = LinkUtilities.scanBackEmailLocalPart(source, start, end);
+        assertEquals(source, expected, actual);
+    }
+
+    //    @Test
+    //    public void testAgainstFiles() {
+    //    	// for later
+    //    	try (Stream<String> lines = Files.lines(Paths.get("file.txt"))) {
+    //    	    lines.forEach(dosomething);
+    //    	} catch (IOException e) {
+    //    	    e.printStackTrace();
+    //    	}
+    //
+    //    }
+
+}
diff --git a/icu4j/main/core/src/test/resources/com/ibm/icu/dev/data/testdata/links/LinkDetectionTest.txt b/icu4j/main/core/src/test/resources/com/ibm/icu/dev/data/testdata/links/LinkDetectionTest.txt
new file mode 100644
index 000000000000..8da4c44a45ee
--- /dev/null
+++ b/icu4j/main/core/src/test/resources/com/ibm/icu/dev/data/testdata/links/LinkDetectionTest.txt
@@ -0,0 +1,432 @@
+# LinkDetectionTest.txt
+# Date: 2026-02-03, 23:10:02 GMT 
+# © 2026 Unicode®, Inc.
+# Unicode and the Unicode Logo are registered trademarks of Unicode, Inc. in the U.S. and other countries.
+# For terms of use and license, see https://www.unicode.org/terms_of_use.html
+# 
+# The usage and stability of these values is covered in https://www.unicode.org/reports/tr58/
+# 
+# ================================================
+# 
+# Format:
+#   Each line contains zero or more marked links, such as ⸠abc.com⸡
+# 
+# Operation:
+#   For each line.
+#   • Create a copy of the line, with the characters ⸠ and ⸡ removed.
+#   • Run link detection on the line, inserting ⸠ and ⸡ around each detected link.
+#   • Report a failure if the result is not identical to the original line.
+#   Empty lines, and lines starting with # are ignored.
+#   Otherwise # is treated like any other character.
+# ================================================
+# 
+
+# Misc. test cases
+
+# Implementations may differ as to whether they linkify text where the TLD is not listed in  IANA's tlds-alpha-by-domain.txt, such as example.invalid
+See example.😎  on…
+See .fr on…
+
+# Sample uppercase
+See ⸠TEST.COM⸡ on…
+See ⸠FOO.VERMÖGEN.com⸡ on…
+
+# Illegal domain names, so not linked
+See http://.foo.example.com/αβγ on…
+See http://foo..example.com/αβγ on…
+See http://-foo.example-.com. on…
+
+# Legal but unusual. Because we might be at the end of a sentence, we don't include the . unless followed by a Path, Query, or Fragment
+See ⸠http://foo.example.com⸡. on…
+See ⸠http://foo.example.com./αβγ⸡ on…
+
+# Various Path/Query/Fragment cases
+See ⸠example.com⸡! on…
+See ⸠example.com/αβγ⸡ on…
+See ⸠example.com?αβγ⸡ on…
+See ⸠example.com#αβγ⸡ on…
+See ⸠example.com/αβγ/δεζ?θικ#λμν⸡ on…
+See ⸠example.com/αβγ/δεζ?δ.εφ#λμν⸡ on…
+See ⸠https://example.com/αβγ⸡ on…
+See ⸠http://example.com/αβγ⸡ on…
+
+# Illustrations of Soft vs Hard, bracketing
+
+# Break on hard (such as ' ')
+See ⸠example.com/αβγ/δεζ?δ⸡ εφ#λμν on…
+
+# Break on soft (such as '.') followed by hard (such as ' ')
+See ⸠example.com/αβγ/δεζ?δ⸡. εφ#λμν on…
+See ⸠example.com/α/βγ?δ/ε?ζ#λ/μ?ν#π⸡ on…
+
+# Break on unmatched bracket
+See ⸠example.com/αβ⸡) on…
+
+# Include matched bracket
+See ⸠example.com/α(β)⸡ on…
+See ⸠example.com/αβγ/δρς?α=θ&β=κ#λμν:~:text=φχψ⸡ on…
+
+# Continue past matching brackets within same part and not across interior syntax
+See ⸠example.com/α(β)γ/δρς?α=(θ)&(β)=κ#λμν:~:text=(φχψ)⸡ on…
+See ⸠example.com/α[(β)]γ/δρς?α=[(θ)]&[(β)]=κ#λμν:~:text=[(φχψ)]⸡ on…
+
+# Don't match mismatched brackets
+See ⸠example.com/α[(β⸡])γ/δρς?α=[(θ)]&[(β)]=κ#λμν:~:text=φχψ on…
+See ⸠example.com/α[(β)]γ/δρς?α=[(θ⸡])&[(β)]=κ#λμν:~:text=φχψ on…
+See ⸠example.com/α[(β)]γ/δρς?α=[(θ)]&[(β⸡])=κ#λμν:~:text=φχψ on…
+See ⸠example.com/α[(β)]γ/δρς?α=[(θ)]&[(β)]=κ#λμν:~:text=[(φχψ⸡]) on…
+
+# Don't match across interior syntax 
+See ⸠example.com/αβ(γ/δ⸡)ρς?α=θ&β=κ#λμν:~:text=φχψ on…
+See ⸠example.com/αβγ/δρς?α(=⸡)θ&β=κ#λμν:~:text=φχψ on…
+See ⸠example.com/αβγ/δρς?α=θ(&⸡)β=κ#λμν:~:text=φχψ on…
+See ⸠example.com/αβγ/δρς?α=θ&β=κ#λμ(ν:~:text=φχ⸡)ψ on…
+
+⸠https://ja.wikipedia.org/wiki/フィンセント・ファン・ゴッホ⸡
+⸠https://ja.wikipedia.org/wiki/%E3%83%95%E3%82%A3%E3%83%B3%E3%82%BB%E3%83%B3%E3%83%88%E3%83%BB%E3%83%95%E3%82%A1%E3%83%B3%E3%83%BB%E3%82%B4%E3%83%83%E3%83%9B⸡
+⸠abc.com⸡ ⸠deΩ.uk⸡ ⸠foo.accountants⸡ ⸠foo.香港⸡
+⸠abc.com/d/e?xyz#w⸡
+⸠abc.XN--11B4C3D⸡
+
+# Email addresses
+
+See ⸠αβγ.δεζ@example.com⸡ on…
+See ⸠mailto:αβγ.δεζ@example.com⸡ on…
+
+# If a local-part is invalid, skip the domain name
+See mailto:αβγ.δεζ.@example.com on…
+See mailto:αβγ..δεζ@example.com on…
+See mailto:.αβγ.δεζ@example.com on…
+
+#Stop backing up when a space is hit
+See ⸠abcd@example.com⸡
+
+#Include the medial dot
+See ⸠x.abcd@example.com⸡.
+
+#Handle non-ASCII
+See ⸠アルベルト.アルベルト@example.com⸡
+ 
+#Invalid domain name, so not linked
+See @example.😎
+
+# No valid local-part, so not linked
+See @example.com
+See john.@example.com
+See .john@example.com
+See ..john.doe@example.com
+
+#Quoted local-part, so not linked (in the base algorithm)
+See "john\ doe"@example.com
+
+# Note that john.doe@example.com is ambiguous: it could be an email address, or it could be a URL with a userinfo
+# The latter, however, is extremely rare, so always favor the email address.
+# (Even when the context would tip the balance towards the URL interpretation, for consistency.)
+
+See ⸠example.com:1234/forum/questions/?tag=new&order=newest#top⸡ on…
+See ⸠john.doe@example.com⸡:1234/forum/questions/?tag=new&order=newest#top on…
+
+
+⸠john.smith@example.com⸡
+⸠john.smith@example.com⸡/foo/bar
+
+john..smith@example.com
+john..smith@example.com/foo/bar
+
+⸠mailto:john.smith@example.com⸡
+⸠mailto:john.smith@example.com⸡/foo/bar
+
+mailto:john..smith@example.com
+mailto:john..smith@example.com/foo/bar
+
+# The following are strange. the "john.smith" could be a valid userinfo, except that that is deprecated for security
+
+http://john.smith@example.com
+http://john.smith@example.com/foo/bar
+
+http://john..smith@example.com
+http://john..smith@example.com/foo/bar
+
+# Sample pages from https://meta.wikimedia.org/wiki/List_of_articles_every_Wikipedia_should_have
+
+⸠https://als.wikipedia.org/wiki/Wikipedia:Artikel,_wos_i_allne_Wikipedene_gee_sött⸡
+⸠https://an.wikipedia.org/wiki/Wikipedia:Lista_d'articlos_que_toda_Wikipedia_habría_de_tener⸡
+⸠https://ang.wikipedia.org/wiki/Wikipǣdia:List_of_articles_all_languages_should_have⸡
+⸠https://anp.wikipedia.org/wiki/मूलभूत_लेखॊ_के_सूची⸡
+⸠https://ar.wikipedia.org/wiki/ويكيبيديا:قائمة_مقالات_يجب_أن_تحتويها_كل_ويكيبيديا⸡
+⸠https://arc.wikipedia.org/wiki/ܘܝܩܝܦܕܝܐ:ܡܟܬܒܘܬܐ_ܕܡܓܠܐ_ܠܟܠ_ܘܝܩܝܦܕܝܐ⸡
+⸠https://ary.wikipedia.org/wiki/ويكيپيديا:1000_مقالة_خاص_يكونو_ف_جميع_لويكيپيديات⸡
+⸠https://arz.wikipedia.org/wiki/ويكيبيديا:مقالات_لازم_تكون_فى_كل_ويكيبيديا⸡
+⸠https://as.wikipedia.org/wiki/ৱিকিপিডিয়া:অত্যাৱশ্যকীয়_প্ৰবন্ধসমূহৰ_তালিকা⸡
+⸠https://ast.wikipedia.org/wiki/Wikipedia:Llista_d'artículos_que_toles_Wikipedies_habríen_de_tener⸡
+⸠https://av.wikipedia.org/wiki/Википедия:Киналго_Википедиязда_рукӀине_кколел_макъалабазул_сияхӀ⸡
+⸠https://awa.wikipedia.org/wiki/विकिपीडिया:अवधी_भाषाकय_विकिपीडिया_में_होए_वाला_न्युनतम_लेख_कय_लिस्ट⸡
+⸠https://az.wikipedia.org/wiki/Vikipediya:1000_məqalə⸡
+⸠https://azb.wikipedia.org/wiki/ویکی‌پدیا:۱۰۰۰_مقاله⸡
+⸠https://ba.wikipedia.org/wiki/Википедия:Һәр_тел_бүлегендә_булырға_тейеш_мәҡәләләр_исемлеге⸡
+⸠https://bar.wikipedia.org/wiki/Wikipedia:Artikl,_de_wo_jede_Wikipedia_hom_soit⸡
+⸠https://bat-smg.wikipedia.org/wiki/Vikipedėjė:Būtėnū_straipsniu_sārašos_vikipedėjuo⸡
+⸠https://be-tarask.wikipedia.org/wiki/Вікіпэдыя:Істотныя_артыкулы⸡
+⸠https://be.wikipedia.org/wiki/Вікіпедыя:Артыкулы,_якія_мусяць_быць_у_кожнай_Вікіпедыі⸡
+⸠https://bew.wikipedia.org/wiki/Wikipédi:Daptar_makalah_nyang_saban_Wikipédi_kudu_gableg⸡
+⸠https://bg.wikipedia.org/wiki/Уикипедия:1000_статии⸡
+⸠https://bh.wikipedia.org/wiki/विकिपीडिया:जरूरी_लेख⸡
+⸠https://bn.wikipedia.org/wiki/উইকিপিডিয়া:আবশ্যকীয়_নিবন্ধ⸡
+⸠https://bo.wikipedia.org/wiki/Wikipedia:སྐད་རིགས་ཆ་ཚང་ཡོད་པའི་རྩོམ་ཡིག་ཟིན་ཐོ⸡།
+⸠https://bpy.wikipedia.org/wiki/উইকিপিডিয়া:থানা_থকিসে_নিবন্ধ⸡
+⸠https://br.wikipedia.org/wiki/Wikipedia:Roll_ar_pennadoù_a_rankfe_pep_wikipedia_kaout⸡
+⸠https://bs.wikipedia.org/wiki/Wikipedia:1000_članaka⸡
+⸠https://btm.wikipedia.org/wiki/Wikipedia:Artikel_ponting⸡
+⸠https://bug.wikipedia.org/wiki/Wikipedia:Daftar_artikel_ᨆᨊᨛᨂ_ᨅᨔ_ᨄᨑᨛᨒᨘ_ᨂᨛᨀ⸡
+⸠https://bxr.wikipedia.org/wiki/Википеэди:1000_үгүүлэл⸡
+⸠https://ca.wikipedia.org/wiki/Viquipèdia:Llista_d'articles_que_totes_les_llengües_haurien_de_tenir⸡
+⸠https://cdo.wikipedia.org/wiki/Wikipedia:Ék-chiék_Wikipedia_dŭ_găi-dŏng_ô_gì_ùng⸡
+⸠https://ce.wikipedia.org/wiki/Википеди:Массо_а_маттахь_хилайеза_агӀонаш⸡
+⸠https://ceb.wikipedia.org/wiki/Wikipedia:Talaan_sa_mga_artikulo_nga_angayng_anaa_sa_matag_Wikipedya⸡
+⸠https://ch.wikipedia.org/wiki/Wikipedia:List_of_articles_all_languages_should_have⸡
+⸠https://ckb.wikipedia.org/wiki/ویکیپیدیا:وتارە_سەرەکییەکان⸡
+⸠https://co.wikipedia.org/wiki/Wikipedia:Lista_di_l'articuli_chì_ogni_Wikipedia_duverìa_avè⸡
+⸠https://crh.wikipedia.org/wiki/Vikipediya:Er_bir_Vikipediyada_olması_kerek_olğan_maqaleler⸡
+⸠https://cs.wikipedia.org/wiki/Wikipedie:1000_nejdůležitějších_článků⸡
+⸠https://cu.wikipedia.org/wiki/Википєдїꙗ:Важьни_члѣни⸡
+⸠https://cv.wikipedia.org/wiki/Википеди:Ку_статьясем_кашни_Википеди_уйрăмĕсенче_пулмалла⸡
+⸠https://cy.wikipedia.org/wiki/Wicipedia:Rhestr_erthyglau_sy'n_angenrheidiol_ym_mhob_iaith⸡
+⸠https://da.wikipedia.org/wiki/Wikipedia:Artikler_vi_bør_have⸡
+⸠https://de.wikipedia.org/wiki/Wikipedia:Artikel,_die_es_in_allen_Wikipedias_geben_sollte⸡
+⸠https://dz.wikipedia.org/wiki/Wikipedia:List_of_articles_all_Wikipedias_should_have⸡
+⸠https://ee.wikipedia.org/wiki/Wikipedia:Vital_articles⸡
+⸠https://eml.wikipedia.org/wiki/Wikipedia:Léssta_ed_artéccuel_necesàri⸡
+⸠https://en.wikipedia.org/wiki/Wikipedia:List_of_articles_all_languages_should_have⸡
+⸠https://eo.wikipedia.org/wiki/Vikipedio:Listo_de_havendaj_artikoloj⸡
+⸠https://es.wikipedia.org/wiki/Wikipedia:Lista_de_artículos_que_toda_Wikipedia_debería_tener⸡
+⸠https://et.wikipedia.org/wiki/Vikipeedia:Oluliste_artiklite_loend⸡
+⸠https://eu.wikipedia.org/wiki/Wikipedia:Wikipedia_guztiek_izan_beharreko_artikuluen_zerrenda/3._maila⸡
+⸠https://ext.wikipedia.org/wiki/Lista_d'artícalus_que_toa_Wikipedia_deviera_de_tenel⸡
+⸠https://fat.wikipedia.org/wiki/Atekels_a_ɔsɛ_dɛ_Wikipedia_biara_enya⸡
+⸠https://ff.wikipedia.org/wiki/Wikipedia:Lissafin_labaran_duk_harsuna_yakamata_su_kasance_suna_da⸡
+⸠https://fi.wikipedia.org/wiki/Wikipedia:Luettelo_keskeisistä_tietosanakirja-artikkeleista⸡
+⸠https://fiu-vro.wikipedia.org/wiki/Wikipedia:Artikliq,_miä_egan_keelen_pidänü_olõma⸡
+⸠https://fr.wikipedia.org/wiki/Wikipédia:Liste_d'articles_que_toutes_les_encyclopédies_devraient_avoir⸡
+⸠https://frr.wikipedia.org/wiki/Wikipedia:1000_wichtag_artiikler⸡
+⸠https://fur.wikipedia.org/wiki/Vichipedie:Liste_dai_articui_che_dutis_lis_vichipedîs_a_varessin_di_vê⸡
+⸠https://fy.wikipedia.org/wiki/Wikipedy:Artikels_dy't_wichtich_binne_foar_eltse_Wikipedy⸡
+⸠https://ga.wikipedia.org/wiki/Vicipéid:Liosta_d'ábhair_riachtanach_do_gach_uile_Vicipéid⸡
+⸠https://gan.wikipedia.org/wiki/Wikipedia:基礎文章⸡
+⸠https://gl.wikipedia.org/wiki/Wikipedia:Lista_de_artigos_que_toda_Wikipedia_debería_ter⸡
+⸠https://gor.wikipedia.org/wiki/Wikipedia:Artikel_u_musti_woluwo_to_timi'idu_Wikipedia⸡
+⸠https://gv.wikipedia.org/wiki/Wikipedia:Artyn_er-laccal⸡
+⸠https://ha.wikipedia.org/wiki/Wikipedia:Vital_articles⸡
+⸠https://he.wikipedia.org/wiki/ויקיפדיה:מיזמי_ויקיפדיה/מיזם_ערכים_חשובים⸡
+⸠https://hi.wikipedia.org/wiki/विकिपीडिया:कुछ_प्रारंभिक_लेख_जो_कि_हर_भाषा_के_विकिपीडिया_में_होने_चाहिए⸡
+⸠https://hif.wikipedia.org/wiki/Wikipedia:Articles_jiske_sab_wikipedia_me_hoe_ke_chaahi_ke_suchi⸡
+⸠https://hr.wikipedia.org/wiki/Wikipedija:Wikiprojekt_10000/Predloženi_članci⸡
+⸠https://hsb.wikipedia.org/wiki/Wikipedija:1000_wažnych_nastawkow⸡
+⸠https://ht.wikipedia.org/wiki/Wikipedya:Lis_atik_ke_tout_ansiklopedi_yo_ta_dwe_genyen⸡
+⸠https://hu.wikipedia.org/wiki/Wikipédia:Ezer_fontos_cikk⸡
+⸠https://hy.wikipedia.org/wiki/Վիքիպեդիա:Կարևորագույն_հոդվածներ⸡
+⸠https://ia.wikipedia.org/wiki/Wikipedia:Articulos_vital⸡
+⸠https://id.wikipedia.org/wiki/Wikipedia:Artikel_penting⸡
+⸠https://ie.wikipedia.org/wiki/Liste_de_articules_quel_omni_Wikipedia_deve_haver⸡
+⸠https://ilo.wikipedia.org/wiki/Wikipedia:Listaan_dagiti_artikulo_a_nasken_nga_adda_ti_amin_a_Wikipedia⸡
+⸠https://inh.wikipedia.org/wiki/Википеди:Массайолча_эршашка_хила_езаш_йола_оагӀонаш⸡
+⸠https://io.wikipedia.org/wiki/Wikipedio:Obligita_artikli⸡
+⸠https://is.wikipedia.org/wiki/Wikipedia:Grundvallargreinar⸡
+⸠https://it.wikipedia.org/wiki/Wikipedia:Lista_delle_voci_che_tutte_le_Wikipedie_dovrebbero_avere⸡
+⸠https://ja.wikipedia.org/wiki/Wikipedia:すべての言語版にあるべき項目の一覧⸡
+⸠https://jam.wikipedia.org/wiki/Wikipidia:Di_nesiseri_aatikl_dem⸡
+⸠https://jbo.wikipedia.org/wiki/uikipedi'as:liste_be_le_tcevai_ckupau⸡
+⸠https://jv.wikipedia.org/wiki/Wikipédia:Daftar_Artikel_Dhasar⸡
+⸠https://ka.wikipedia.org/wiki/ვიკიპედია:აუცილებელი_სტატიები⸡
+⸠https://kaa.wikipedia.org/wiki/Wikipedia:Hámme_Wikipedialarda_bolıwı_tiyis_maqalalar_dizimi⸡
+⸠https://kbd.wikipedia.org/wiki/Уикипедиэ:Бзэ_версиэу_хъуам_яӀэн_хуей_тхыгъэхэм_я_напэ⸡
+⸠https://kcg.wikipedia.org/wiki/A̱tsatsak_a̱yaati̱kut⸡
+⸠https://kg.wikipedia.org/wiki/Wikipedia:Vital_articles⸡
+⸠https://kk.wikipedia.org/wiki/Уикипедия:Барлық_Уикипедияларда_болуы_тиіс_мақала_тізімі⸡
+⸠https://kn.wikipedia.org/wiki/ವಿಕಿಪೀಡಿಯ:ಅಗತ್ಯ_ಲೇಖನಗಳು⸡
+⸠https://ko.wikipedia.org/wiki/위키백과:모든_언어의_위키백과마다_꼭_있어야_하는_문서_목록⸡
+⸠https://krc.wikipedia.org/wiki/Википедия:Хар_Википедияда_болургъа_керек_статьяланы_тизмеси⸡
+⸠https://ks.wikipedia.org/wiki/وِکیٖپیٖڈیا:Vital_articles⸡
+⸠https://ksh.wikipedia.org/wiki/Wikipedia:Sigge_di_en_alle_Sproche_drin_sin_sullte⸡
+⸠https://ku.wikipedia.org/wiki/Wîkîpediya:Gotarên_ku_ji_bo_hemû_Wîkîpediyayan_pêwîst_in⸡
+⸠https://kw.wikipedia.org/wiki/Wikipedia:1,000_erthygel⸡
+⸠https://ky.wikipedia.org/wiki/Википедия:Vital_articles⸡
+⸠https://la.wikipedia.org/wiki/Vicipaedia:Paginae_quas_omnibus_Wikipediis_contineri_oportet⸡
+⸠https://lad.wikipedia.org/wiki/Vikipedya:Lista_de_artikolos_ke_toda_Vikipedya_kale_tener⸡
+⸠https://lb.wikipedia.org/wiki/Wikipedia:Artikelen,_déi_et_op_all_Wikipedia_sollt_ginn⸡
+⸠https://lez.wikipedia.org/wiki/Википедия:1000⸡
+⸠https://lfn.wikipedia.org/wiki/Vicipedia:Lista_de_articles_nesesada_per_vicipedia⸡
+⸠https://lg.wikipedia.org/wiki/Wikipedia:Vital_articles⸡
+⸠https://li.wikipedia.org/wiki/Wikipedia:Lies_van_artikele_die_eder_Wikipedia_moot_höbbe⸡
+⸠https://lij.wikipedia.org/wiki/Wikipedia:Vôxe_de_bâze⸡
+⸠https://lmo.wikipedia.org/wiki/Wikipedia:Lista_di_articoi_che_tutt_i_Wikipedii_dovarissen_avègh⸡
+⸠https://lt.wikipedia.org/wiki/Vikipedija:Sąrašas_straipsnių,_būtinų_visomis_kalbomis⸡
+⸠https://lv.wikipedia.org/wiki/Vikipēdija:Nozīmīgi_raksti⸡
+⸠https://mad.wikipedia.org/wiki/Wikipèḍia:Artikel_penting⸡
+⸠https://mai.wikipedia.org/wiki/विकिपिडिया:महत्वपूर्ण_लेखसभ⸡
+⸠https://mdf.wikipedia.org/wiki/Википедиесь:1000_эрявикс_сёрматфксне⸡
+⸠https://min.wikipedia.org/wiki/Wikipedia:Artikel_nan_paralu_ado_di_satiok_Wikipedia⸡
+⸠https://mk.wikipedia.org/wiki/Википедија:Потребни_статии⸡
+⸠https://ml.wikipedia.org/wiki/വിക്കിപീഡിയ:സുപ്രധാന_ലേഖനങ്ങൾ⸡
+⸠https://mni.wikipedia.org/wiki/ꯋꯤꯀꯤꯄꯦꯗꯤꯌꯥ_ꯈꯨꯗꯤꯡꯗ_ꯌꯥꯎꯒꯗꯕ_ꯋꯥꯔꯦꯡꯁꯤꯡ⸡
+⸠https://mr.wikipedia.org/wiki/विकिपीडिया:सगळ्या_विकिपीडियांवर_अपेक्षित_लेखांची_यादी/आंतरभाषीय_परिपेक्ष⸡
+⸠https://ms.wikipedia.org/wiki/Wikipedia:Senarai_rencana_semua_bahasa_perlu_ada⸡
+⸠https://mt.wikipedia.org/wiki/Wikipedija:Lista_ta'_artikli_li_kull_Wikipedija_għandu_jkollha⸡
+⸠https://mwl.wikipedia.org/wiki/Biquipédia:1000_artigos_bitales⸡
+⸠https://my.wikipedia.org/wiki/ဝီကီပီးဒီးယား:မြန်မာဝီကီတွင်_ရှိသင့်သော_ဆောင်းပါးများ⸡
+⸠https://myv.wikipedia.org/wiki/Википедиясь:Лопань_потмокс,_конатат_эрявить_эрьва_келень_пельксэ⸡
+⸠https://nah.wikipedia.org/wiki/Huiquipedia:Amatlahcuilolli_tlen_monequi_ipan_nochi_Huiquipedia_ica_nahuatlahtolli⸡
+⸠https://nap.wikipedia.org/wiki/Wikipedia:Lista_d"e_vvoce_ca_tutt'_'e_Wikipedie_aveno_a_avé⸡
+⸠https://nds.wikipedia.org/wiki/Wikipedia:List_vun_Artikels,_de_in_all_Wikipedias_binnen_ween_schöölt⸡
+⸠https://ne.wikipedia.org/wiki/विकिपिडिया:प्रत्येक_भाषामा_हुनु_पर्ने_न्यूनतम_लेखहरू⸡
+⸠https://ng.wikipedia.org/wiki/Wikipedia:Vital_articles⸡
+⸠https://nl.wikipedia.org/wiki/Wikipedia:Archief/Wikipedia:Artikelen_die_elke_Wikipedia_zou_moeten_hebben⸡
+⸠https://nn.wikipedia.org/wiki/Wikipedia:Artiklar_vi_bør_ha⸡
+⸠https://no.wikipedia.org/wiki/Wikipedia:Liste_over_artikler_vi_bør_ha⸡
+⸠https://nov.wikipedia.org/wiki/Wikipedia:Liste_de_artikles_kel_chaki_wikipedia_deve_kontena⸡
+⸠https://nso.wikipedia.org/wiki/Wikipedia:Ditemana_tše_Bohlokwa⸡
+⸠https://ny.wikipedia.org/wiki/Wikipedia:Vital_articles⸡
+⸠https://oc.wikipedia.org/wiki/Wikipèdia:Tièra_de_1000_articles_que_totas_las_Wikipèdias_deurián_aver⸡
+⸠https://olo.wikipedia.org/wiki/Wikipedii:Vältämättömien_Wikipedii-kirjutuksien_luvettelo⸡
+⸠https://om.wikipedia.org/wiki/Wikipedia:Tarreeffama_barruu_Wiikiipiidiyaa_hundi_qabaachuu_qabu⸡
+⸠https://os.wikipedia.org/wiki/Википеди:Æппæт_Википедиты_чи_хъуамæ_уа,_уыцы_уацтæ⸡
+⸠https://pag.wikipedia.org/wiki/Wikipedia:Listaan_na_nakaukulan_ed_balang_salita⸡
+⸠https://pap.wikipedia.org/wiki/Wikipedia:Artíkulonan_importante⸡
+⸠https://pl.wikipedia.org/wiki/Wikipedia:Strony,_które_powinna_mieć_każda_Wikipedia⸡
+⸠https://pms.wikipedia.org/wiki/Wikipedia:Lista_dj'artìcoj_che_minca_Wikipedia_a_dovrìa_avèj⸡
+⸠https://pnt.wikipedia.org/wiki/Βικιπαίδεια:1000_σελίδας⸡
+⸠https://pt.wikipedia.org/wiki/Wikipédia:Lista_dos_artigos_que_toda_Wikipédia_deve_ter⸡
+⸠https://qu.wikipedia.org/wiki/Wikipidiya:Lliw_Wikipidiyapaq_qillqanakuna⸡
+⸠https://rm.wikipedia.org/wiki/Wikipedia:Artitgels_che_mintga_Vichipedia_duai_aver⸡
+⸠https://ro.wikipedia.org/wiki/Wikipedia:Articole_vitale⸡
+⸠https://roa-rup.wikipedia.org/wiki/Wikipedia:Frândzâ_anânghioasi_trâ_unâ_enciclopedie⸡
+⸠https://ru.wikipedia.org/wiki/Википедия:Список_статей,_которые_должны_быть_во_всех_языковых_версиях⸡
+⸠https://rue.wikipedia.org/wiki/Вікіпедія:Статї,_якы_повинны_быти⸡
+⸠https://rw.wikipedia.org/wiki/Wikipedia:Vital_articles⸡
+⸠https://sa.wikipedia.org/wiki/विकिपीडिया:प्रारम्भिक-विषयाः⸡
+⸠https://sah.wikipedia.org/wiki/Бикипиэдьийэ:Ханнык_баҕарар_Бикипиэдьийэҕэ_баар_буолуохтаах_ыстатыйалар_тиһиликтэрэ⸡
+⸠https://sat.wikipedia.org/wiki/ᱞᱟᱹᱠᱛᱤᱭᱟᱱ_ᱚᱱᱚᱞᱠᱚ⸡
+⸠https://sc.wikipedia.org/wiki/Wikipedia:Lista_de_sos_artìculos_chi_totu_sas_Wikipedias_bisonzat_a_tenner⸡
+⸠https://scn.wikipedia.org/wiki/Wikipedia:Artìculi_nicissari⸡
+⸠https://sco.wikipedia.org/wiki/Wikipedia:Leet_o_airticles_that_aa_Wikipedias_shuid_hae⸡
+⸠https://sh.wikipedia.org/wiki/Wikipedija:Spisak_članaka_koje_sve_Wikipedije_trebaju_imati⸡
+⸠https://simple.wikipedia.org/wiki/Wikipedia:List_of_articles_all_languages_should_have⸡
+⸠https://sk.wikipedia.org/wiki/Wikipédia:1_000_najdôležitejších_článkov⸡
+⸠https://sl.wikipedia.org/wiki/Wikipedija:Članki,_ki_bi_jih_morala_imeti_vsaka_Wikipedija⸡
+⸠https://so.wikipedia.org/wiki/Wikipedia:Maqaalada_aan_u_baahan_laheen⸡
+⸠https://sq.wikipedia.org/wiki/Wikipedia:Artikujt_që_secila_Wikipedia_duhet_t'i_ketë⸡
+⸠https://sr.wikipedia.org/wiki/Википедија:Списак_чланака_које_свака_Википедија_треба_да_има⸡
+⸠https://srn.wikipedia.org/wiki/Wikipedia:Rei_fu_Peprewoysi_ala_Wikipedia_mu_Abi⸡
+⸠https://ss.wikipedia.org/wiki/Wikipedia:List_of_articles_all_languages_should_have⸡
+⸠https://st.wikipedia.org/wiki/Wikipedia:Vital_articles⸡
+⸠https://stq.wikipedia.org/wiki/Wikipedia:Artikkele_do't_in_älke_Wikipedia_reeke_skäl⸡
+⸠https://su.wikipedia.org/wiki/Wikipedia:Artikel_nu_sadaya_basa_Wikipédia_kedah_gaduh⸡
+⸠https://sv.wikipedia.org/wiki/Wikipedia:Basartiklar⸡
+⸠https://sw.wikipedia.org/wiki/Wikipedia:Makala_za_msingi_za_kamusi_elezo⸡
+⸠https://ta.wikipedia.org/wiki/விக்கிப்பீடியா:முக்கிய_கட்டுரைகள்⸡
+⸠https://te.wikipedia.org/wiki/వికీపీడియా:వికీపీడియాలో_తప్పకుండా_ఉండవలసిన_వ్యాసాలు⸡
+⸠https://tet.wikipedia.org/wiki/Wikipedia:Lista_husi_artigu_sira_ne'ebé_Wikipédia_hotu-hotu_bele_iha⸡
+⸠https://tg.wikipedia.org/wiki/Википедиа:Феҳристи_мақолаҳои_муҳим⸡
+⸠https://th.wikipedia.org/wiki/วิกิพีเดีย:รายการบทความที่วิกิพีเดียทุกภาษาควรมี⸡
+⸠https://tk.wikipedia.org/wiki/Wikipediýa:Wikipediýada_bolmagy_gerek_sahypalar⸡
+⸠https://tl.wikipedia.org/wiki/Wikipedia:Talaan_ng_mga_artikulong_kailangang_mayroon_ang_bawat_Wikipedia⸡
+⸠https://tly.wikipedia.org/wiki/Vikipediá:Məǧolon_sijohi,_komon_bəpe_har_zyvonədə_bybun⸡
+⸠https://tn.wikipedia.org/wiki/Wikipedia:List_of_articles_we_should_have⸡
+⸠https://tpi.wikipedia.org/wiki/Wikipedia:Ol_pes_i_mas_stap_long_Wikipedia⸡
+⸠https://tr.wikipedia.org/wiki/Vikipedi:Her_Vikipedi'de_olması_gereken_maddeler⸡
+⸠https://ts.wikipedia.org/wiki/Wikipedia:Matsalwa_ya_Nkoka/xiyenge/3⸡
+⸠https://tt.wikipedia.org/wiki/Википедия:Һәрбер_тел_бүлегендә_булырга_тиешле_мәкаләләр⸡
+⸠https://tum.wikipedia.org/wiki/Wikipedia:Vital_articles⸡
+⸠https://tw.wikipedia.org/wiki/Wikipidia:Atwerɛsɛm_dodoɔ_a_ɛwɔ_sɛ_Wikipidia_ɛnya⸡
+⸠https://ty.wikipedia.org/wiki/Wikipedia:Te_àpi_faufaa⸡
+⸠https://tyv.wikipedia.org/wiki/Википедия:Шупту_дылдарга_бижээн_турар_чүүлдер⸡
+⸠https://udm.wikipedia.org/wiki/Википедия:Тужгес_кулэ_статьяос⸡
+⸠https://uk.wikipedia.org/wiki/Вікіпедія:Статті,_які_повинні_бути_в_усіх_Вікіпедіях⸡
+⸠https://ur.wikipedia.org/wiki/ویکیپیڈیا:مضامین_تازہ⸡
+⸠https://uz.wikipedia.org/wiki/Vikipediya:Muhim_maqolalar⸡
+⸠https://ve.wikipedia.org/wiki/Wikipedia:Vital_articles⸡
+⸠https://vec.wikipedia.org/wiki/Wikipedia:Lista_de_łe_voze_che_tuta_Wikipèdia_garia_da_ver⸡
+⸠https://vep.wikipedia.org/wiki/Vikipedii:Tarbhaižed_lehtpoled,_kudambad_tarbiž_säta_kaikihe_Vikipedijan_versijoihe⸡
+⸠https://wuu.wikipedia.org/wiki/Wikipedia:必有个页面⸡
+⸠https://xal.wikipedia.org/wiki/Wikipedia:Википедьт_иим_төрмүд_бəəх_зөвтə⸡
+⸠https://xh.wikipedia.org/wiki/Wikipedia:Vital_articles⸡
+⸠https://xmf.wikipedia.org/wiki/ვიკიპედია:უციო_სტატიეფი⸡
+⸠https://yi.wikipedia.org/wiki/װיקיפּעדיע:וויכטיגע_ארטיקלען⸡
+⸠https://yo.wikipedia.org/wiki/Wikipedia:Àwọn_àyọkà_tó_ṣe_kóko⸡
+⸠https://zea.wikipedia.org/wiki/Wikipedia:Artikels_die_ieleken_Wikipedia_zou_motte_ha⸡
+⸠https://zgh.wikipedia.org/wiki/ⵡⵉⴽⵉⴱⵉⴷⵢⴰ:ⵜⴰⵍⴳⴰⵎⵜ_ⵏ_ⵉⵎⴳⵔⴰⴷⵏ_ⵏⵏⴰ_ⴷ_ⵉⵇⵇⴰⵏ_ⴰⴷ_ⵉⵍⵉⵏ_ⴳ_ⴽⵔⴰⵢⴳⴰⵜⵜ_ⵡⵉⴽⵉⴱⵉⴷⵢⴰ⸡
+⸠https://zh-classical.wikipedia.org/wiki/維基大典:當有之文⸡
+⸠https://zh-min-nan.wikipedia.org/wiki/Wikipedia:Só͘-ū_ê_Wikipedia_pán-pún_lóng_èng-kai_ū_ê_bûn-chiuⁿ⸡
+⸠https://zh-yue.wikipedia.org/wiki/Wikipedia:基礎文章⸡
+⸠https://zh.wikipedia.org/wiki/Wikipedia:基礎條目⸡
+⸠https://zu.wikipedia.org/wiki/Wikipedia:Vital_articles⸡
+
+# Test cases contributed by ICANN
+
+a ⸠blogspot.com⸡ b
+a com b
+a ⸠example.com:443⸡ b
+a ⸠example.com/123⸡ b
+a ⸠example.com?123⸡ b
+a ⸠example.com#123⸡ b
+a ⸠example.com/123?123#123⸡ b
+a (⸠example.com/123?123#123⸡) b
+a (⸠example.com/123?123#(123)⸡) b
+a ⸠example.com/123.html⸡ b
+a ⸠example.com/123⸡. HTML is great
+a ⸠example.com/123⸡... HTML is so great
+a ⸠example.com/123⸡.
+a ⸠xn-----ctdbabcfhu9c2b9l1acccr4c.xn--mgbah1a3hjkrd⸡ b
+Lorem ipsum ⸠universal-acceptance-test.international⸡ dolor sit amet
+Lorem ipsum ⸠universal-acceptance-test.icu⸡ dolor sit amet
+Lorem ipsum ⸠تجربة-القبول-الشامل.موريتانيا⸡ dolor sit amet
+Lorem ipsum ⸠համընդհանուր-ընկալում-թեստ.հայ⸡ dolor sit amet
+Lorem ipsum ⸠সর্বজনীন-স্বীকৃতির-পরীক্ষা.ভারত⸡ dolor sit amet
+Lorem ipsum ⸠универсальное-принятие-тест.москва⸡ dolor sit amet
+Lorem ipsum ⸠सार्वभौमिक-स्वीकृति-परीक्षण.संगठन⸡ dolor sit amet
+Lorem ipsum ⸠უნივერსალური-თავსობადობის-ტესტი.გე⸡ dolor sit amet
+Lorem ipsum ⸠καθολική-αποδοχή-δοκιμή.ευ⸡ dolor sit amet
+Lorem ipsum ⸠સાર્વત્રિક-સ્વીકૃતિ-પરીક્ષણ.ભારત⸡ dolor sit amet
+Lorem ipsum ⸠ਸਰਵਵਿਆਪਕ-ਪ੍ਰਵਾਨਗੀ-ਪਰਖ.ਭਾਰਤ⸡ dolor sit amet
+Lorem ipsum ⸠다국어도메인이용환경테스트.한국⸡ dolor sit amet
+Lorem ipsum ⸠מבחן-קבלה-אוניברסלי.קום⸡ dolor sit amet
+Lorem ipsum ⸠どこでもつかえる.みんな⸡ dolor sit amet
+Lorem ipsum ⸠ಸಾರ್ವತ್ರಿಕ-ಸ್ವೀಕಾರಾರ್ಹತೆ-ಪರೀಕ್ಷೆ.ಭಾರತ⸡ dolor sit amet
+Lorem ipsum ⸠ユニバーサルアクセプタンス.クラウド⸡ dolor sit amet
+Lorem ipsum ⸠ສາກົນ-ການຍອມຮັບ-ທົດລອງ.ລາວ⸡ dolor sit amet
+Lorem ipsum ⸠സാർവത്രിക-സ്വീകാര്യതാ-പരിശോധന.ഭാരതം⸡ dolor sit amet
+Lorem ipsum ⸠ଯୁନିଭରସାଲ-ଏକସେପ୍ଟନ୍ସ-ଟେଷ୍ଟ.ଭାରତ⸡ dolor sit amet
+Lorem ipsum ⸠විශ්ව-සම්මුති-පිරික්සුම.ලංකා⸡ dolor sit amet
+Lorem ipsum ⸠பொது-ஏற்பு-சோதனை.சிங்கப்பூர்⸡ dolor sit amet
+Lorem ipsum ⸠యూనివర్సల్-ఆమోదం-పరీక్ష.భారత్⸡ dolor sit amet
+Lorem ipsum ⸠ยูเอทดสอบ.ไทย⸡ dolor sit amet
+Lorem ipsum ⸠普遍适用测试.我爱你⸡ dolor sit amet
+Lorem ipsum ⸠普遍適用測試.台灣⸡ dolor sit amet
+Lorem ipsum ⸠ሁለንአቀፍ-ተቀባይነት-ሙከራ.com⸡ dolor sit amet
+Lorem ipsum ⸠ការសាកល្បងទទួលយកជាអន្តរជាតិ.com⸡ dolor sit amet
+Lorem ipsum ⸠အလုံးစုံလက်ခံမှုစမ်းသပ်ချက်.com⸡ dolor sit amet
+Lorem ipsum ⸠ދުނިޔެ-ގަބޫލުކުރާ-ޓެސްޓު.com⸡ dolor sit amet
+Lorem ipsum ⸠universal-acceptance-test.קום⸡ dolor sit amet
+Lorem ipsum ⸠épreuve-acceptation-universelle.org⸡ dolor sit amet
+Lorem ipsum ⸠ཡོངས་ཁྱབ་ངོས་ལེན་བརྟག་དཔྱད.com⸡ dolor sit amet
+Lorem ipsum ⸠Universales-Akzeptanz-Test.vermögensberatung⸡ dolor sit amet
+Lorem ipsum ⸠épreuve-acceptation-universelle.org⸡ dolor sit amet
+Lorem ipsum ⸠普遍适用测试。我爱你⸡ dolor sit amet
+Lorem ipsum موريتانيا.xn-----ctdbabcfhu9c2b9l1acccr4c dolor sit amet
+Lorem ipsum ⸠تجربة-القبول-الشامل.xn--mgbah1a3hjkrd⸡ dolor sit amet
+Lorem ipsum ⸠xn-----ctdbabcfhu9c2b9l1acccr4c.xn--mgbah1a3hjkrd⸡ dolor sit amet
+Lorem ipsum ⸠universal-acceptance-test.icu/测试⸡ dolor sit amet
+Lorem ipsum ⸠普遍适用测试.我爱你/测试⸡ dolor sit amet
+Lorem ipsum ⸠تجربة-القبول-الشامل.موريتانيا/تجربة⸡ dolor sit amet
+blah ⸠en.wikipedia.org/wiki/The_Lovemakers_(film)⸡ blah
+blah ⸠example.com/?foo[1]=a&amp;foo[2]=b⸡ blah
+Lorem ipsum ⸠comoyo.com/play/S(123)⸡ dolor sit amet
+Lorem ipsum ⸠example.com/knutsen_ludvigsen/ver(k)ste(d)/brilleslange.mp3⸡ dolor sit amet
+Lorem ipsum ⸠example.com/Bob_Marley/Rastaman_Vibration/11_Jah_Live_(originally_issued_as_Island_Single_(WIP_6265))_(bonus_track⸡ dolor sit amet
+Lorem ipsum ⸠business.timesonline.co.uk/article/0,,9065-2473189,00.html⸡ dolor sit amet
+Lorem ipsum ⸠www.mail-archive.com/ruby-talk@ruby-lang.org/⸡ dolor sit amet
+Lorem ipsum ⸠tools.ietf.org/html/rfc3986⸡ dolor sit amet
+Lorem ipsum ⸠www.amazon.com/Testing-Equal-Sign-In-Path/ref=pd_bbs_sr_1?ie=UTF8&s=books&qid=1198861734&sr=8-1⸡ dolor sit amet
+Lorem ipsum ⸠www.google.com/doku.php?id=gps:resource:scs:start⸡ dolor sit amet
+Lorem ipsum ⸠maps.google.co.uk/maps?f=q&q=the+london+eye&ie=UTF8&ll=51.503373,-0.11939&spn=0.007052,0.012767&z=16&iwloc=A⸡ dolor sit amet
+Lorem ipsum ⸠www.rubyonrails.com/foo.cgi?trailing_hyphen=value-⸡ dolor sit amet
+Lorem ipsum ⸠www.rubyonrails.com/foo.cgi?trailing_forward_slash=value⸡ dolor sit amet
diff --git a/icu4j/main/core/src/test/resources/com/ibm/icu/dev/data/testdata/links/LinkFormattingTest.txt b/icu4j/main/core/src/test/resources/com/ibm/icu/dev/data/testdata/links/LinkFormattingTest.txt
new file mode 100644
index 000000000000..eb3ff5169136
--- /dev/null
+++ b/icu4j/main/core/src/test/resources/com/ibm/icu/dev/data/testdata/links/LinkFormattingTest.txt
@@ -0,0 +1,278 @@
+# LinkFormattingTest.txt
+# Date: 2026-05-20, 21:41:06 GMT 
+# © 2026 Unicode®, Inc.
+# Unicode and the Unicode Logo are registered trademarks of Unicode, Inc. in the U.S. and other countries.
+# For terms of use and license, see https://www.unicode.org/terms_of_use.html
+# 
+# The usage and stability of these values is covered in https://www.unicode.org/reports/tr58/
+# 
+# ================================================
+# 
+# Format: This is not a file with semicolon-delimited fields.
+# Instead, it consists of paired lines, separated by empty lines and/or comment lines.
+# The first line of each pair is the source, fully escaped.
+# The second line of each pair is the result, minimally escaped.
+# 
+# Comments are lines that begin with a #.
+# 
+# The fully-escaped field percent-escapes all code points based on https://url.spec.whatwg.org/#percent-encoded-bytes.
+# This means all literal syntax characters in each Part and all code points above ASCII.
+# It also percent-escapes the last character, if it is Link_Term=Soft.
+# 
+# The minimally-escaped field is the more readable format described in UTS #58.
+# 
+# Each pair also has a comment line for the internal structure of the URL.
+# 𝑺 = the schema
+# 𝑯 = the host (typically just a domain name) the internal structure is not broken down.
+# 𝑷 = indicates one of the labels in the path. So /seg1/seg2 becomes 𝑷=seg1 𝑷=seg1
+# 𝑸 = each key in the query. So &θ=ικλ&μ=γξο becomes 𝑸=θ 𝑽=ικλ 𝑸=μ 𝑽=γξο.
+# 𝑽 = a value for the preceding key
+# 𝑭 = a fragment. So #reserved:~:text=Reserved,into:~:text=open,closed becomes 𝑭=reserved 𝑫=Reserved,into 𝑫=text=open,closed
+# 𝑫 = a fragment-directive
+# ================================================
+# 
+
+# Selected test cases
+
+# Path only					
+# {𝑺=https:// 𝑯=example.com 𝑷=α}
+https://example.com/%CE%B1
+https://example.com/α
+
+# Query only					
+# {𝑺=https:// 𝑯=example.com 𝑸=α}
+https://example.com?%CE%B1
+https://example.com?α
+
+# Fragment only					
+# {𝑺=https:// 𝑯=example.com 𝑭=α}
+https://example.com#%CE%B1
+https://example.com#α
+
+# All parts					
+# {𝑺=https:// 𝑯=example.com 𝑷=αβγ 𝑷=δεζ 𝑸=θ 𝑽=ικλ 𝑸=μ 𝑽=γξο 𝑭=πρς}
+https://example.com/%CE%B1%CE%B2%CE%B3/%CE%B4%CE%B5%CE%B6?%CE%B8=%CE%B9%CE%BA%CE%BB&%CE%BC=%CE%B3%CE%BE%CE%BF#%CF%80%CF%81%CF%82
+https://example.com/αβγ/δεζ?θ=ικλ&μ=γξο#πρς
+
+# Escape soft at end of Path (with nothing following)				
+# {𝑺=https:// 𝑯=example.com 𝑷=αβγ 𝑷=δ.εζ.}
+https://example.com/%CE%B1%CE%B2%CE%B3/%CE%B4.%CE%B5%CE%B6%2E
+https://example.com/αβγ/δ.εζ%2E
+
+# Escape soft at end of Query (with nothing following)				
+# {𝑺=https:// 𝑯=example.com 𝑷=αβγ 𝑷=δ.εζ. 𝑸=θ 𝑽=ικλ 𝑸=μ 𝑽=γ.ξο.}
+https://example.com/%CE%B1%CE%B2%CE%B3/%CE%B4.%CE%B5%CE%B6.?%CE%B8=%CE%B9%CE%BA%CE%BB&%CE%BC=%CE%B3.%CE%BE%CE%BF%2E
+https://example.com/αβγ/δ.εζ.?θ=ικλ&μ=γ.ξο%2E
+
+# Escape soft at end of Fragment (with nothing following)				
+# {𝑺=https:// 𝑯=example.com 𝑷=αβγ 𝑷=δ.εζ 𝑸=θ 𝑽=ικλ 𝑸=μ 𝑽=γ.ξο 𝑭=π.ρς.}
+https://example.com/%CE%B1%CE%B2%CE%B3/%CE%B4.%CE%B5%CE%B6?%CE%B8=%CE%B9%CE%BA%CE%BB&%CE%BC=%CE%B3.%CE%BE%CE%BF#%CF%80.%CF%81%CF%82%2E
+https://example.com/αβγ/δ.εζ?θ=ικλ&μ=γ.ξο#π.ρς%2E
+
+# Escape ? in Path					
+# {𝑺=https:// 𝑯=example.com 𝑷=α?μπ}
+https://example.com/%CE%B1%3F%CE%BC%CF%80
+https://example.com/α%3Fμπ
+
+# Escape #, =, & in Path/Query					
+# {𝑺=https:// 𝑯=example.com 𝑷=α#β 𝑸=γ 𝑽=δ#ε}
+https://example.com/%CE%B1#%CE%B2?%CE%B3=%CE%B4#%CE%B5
+https://example.com/α%23β?γ=δ%23ε
+
+# Escape hard (' ')					
+# {𝑺=https:// 𝑯=example.com 𝑷=αβ γ 𝑷=δεζ 𝑸=θ 𝑽=ικ λ 𝑸= 𝑽=γξο 𝑭=πρ σ}
+https://example.com/%CE%B1%CE%B2%20%CE%B3/%CE%B4%CE%B5%CE%B6?%CE%B8=%CE%B9%CE%BA%20%CE%BB&=%CE%B3%CE%BE%CE%BF#%CF%80%CF%81%20%CF%83
+https://example.com/αβ%20γ/δεζ?θ=ικ%20λ&=γξο#πρ%20σ
+
+# Escape soft ('.') unless followed by include					
+# {𝑺=https:// 𝑯=example.com 𝑷=αβγ. 𝑷=δεζ. 𝑸=θ 𝑽=ικ.λ 𝑸= 𝑽=γξο. 𝑭=πρς.}
+https://example.com/%CE%B1%CE%B2%CE%B3./%CE%B4%CE%B5%CE%B6.?%CE%B8=%CE%B9%CE%BA.%CE%BB&=%CE%B3%CE%BE%CE%BF.#%CF%80%CF%81%CF%82%2E
+https://example.com/αβγ./δεζ.?θ=ικ.λ&=γξο.#πρς%2E
+
+# Escape unmatched brackets					
+# {𝑺=https:// 𝑯=example.com 𝑷=α(β) 𝑸=γ(δ) 𝑭=ε(ζ))}
+https://example.com/%CE%B1(%CE%B2)?%CE%B3(%CE%B4)#%CE%B5(%CE%B6))
+https://example.com/α(β)?γ(δ)#ε(ζ)%29
+
+# {𝑺=https:// 𝑯=example.com 𝑷=α(β) 𝑸=γ(δ))}
+https://example.com/%CE%B1(%CE%B2)?%CE%B3(%CE%B4))
+https://example.com/α(β)?γ(δ)%29
+
+# {𝑺=https:// 𝑯=example.com 𝑷=α(β) 𝑭=ε(ζ))}
+https://example.com/%CE%B1(%CE%B2)#%CE%B5(%CE%B6))
+https://example.com/α(β)#ε(ζ)%29
+
+# {𝑺=https:// 𝑯=example.com 𝑷=α(β))}
+https://example.com/%CE%B1(%CE%B2))
+https://example.com/α(β)%29
+
+# {𝑺=https:// 𝑯=example.com 𝑸=γ(δ) 𝑭=ε(ζ))}
+https://example.com?%CE%B3(%CE%B4)#%CE%B5(%CE%B6))
+https://example.com?γ(δ)#ε(ζ)%29
+
+# {𝑺=https:// 𝑯=example.com 𝑸=γ(δ))}
+https://example.com?%CE%B3(%CE%B4))
+https://example.com?γ(δ)%29
+
+# {𝑺=https:// 𝑯=example.com 𝑭=ε(ζ))}
+https://example.com#%CE%B5(%CE%B6))
+https://example.com#ε(ζ)%29
+
+# No escape for matched brackets					
+# {𝑺=https:// 𝑯=example.com 𝑷=α(β) 𝑸=γ(δ) 𝑭=ε(ζ)}
+https://example.com/%CE%B1(%CE%B2)?%CE%B3(%CE%B4)#%CE%B5(%CE%B6)
+https://example.com/α(β)?γ(δ)#ε(ζ)
+
+# {𝑺=https:// 𝑯=example.com 𝑷=α(β) 𝑸=γ(δ)}
+https://example.com/%CE%B1(%CE%B2)?%CE%B3(%CE%B4)
+https://example.com/α(β)?γ(δ)
+
+# {𝑺=https:// 𝑯=example.com 𝑷=α(β) 𝑭=ε(ζ)}
+https://example.com/%CE%B1(%CE%B2)#%CE%B5(%CE%B6)
+https://example.com/α(β)#ε(ζ)
+
+# {𝑺=https:// 𝑯=example.com 𝑷=α(β)}
+https://example.com/%CE%B1(%CE%B2)
+https://example.com/α(β)
+
+# {𝑺=https:// 𝑯=example.com 𝑸=γ(δ) 𝑭=ε(ζ)}
+https://example.com?%CE%B3(%CE%B4)#%CE%B5(%CE%B6)
+https://example.com?γ(δ)#ε(ζ)
+
+# {𝑺=https:// 𝑯=example.com 𝑸=γ(δ)}
+https://example.com?%CE%B3(%CE%B4)
+https://example.com?γ(δ)
+
+# {𝑺=https:// 𝑯=example.com 𝑭=ε(ζ)}
+https://example.com#%CE%B5(%CE%B6)
+https://example.com#ε(ζ)
+
+# Path with escaped separator
+# {𝑺=https:// 𝑯=example.com 𝑷=α 𝑷=β/γ}
+https://example.com/%CE%B1/%CE%B2/%CE%B3
+https://example.com/α/β%2Fγ
+
+# Path with escapes (% is escaped only if followed by 2 hex digits [0-9A-fa-f)
+# {𝑺=https:// 𝑯=example.com 𝑷=α 𝑷=β%41γ%ε%}
+https://example.com/%CE%B1/%CE%B2%2541%CE%B3%%CE%B5%
+https://example.com/α/β%2541γ%ε%
+
+# Query with escapes (& separates key-value pairs, so handle literal &)	
+# {𝑺=https:// 𝑯=example.com 𝑸=α& 𝑽=β 𝑸=γ 𝑽=&δ}
+https://example.com?%CE%B1&=%CE%B2&%CE%B3=&%CE%B4
+https://example.com?α%26=β&γ=%26δ
+
+# Query with escapes (= separates keys and values, so handle literal =)	
+# {𝑺=https:// 𝑯=example.com 𝑸=α=β 𝑽=γ=δ}
+https://example.com?%CE%B1=%CE%B2=%CE%B3=%CE%B4
+https://example.com?α%3Dβ=γ%3Dδ
+
+# Query with escapes (% is escaped only if followed by 2 hex digits [0-9A-fa-f)
+# {𝑺=https:// 𝑯=example.com 𝑸=α%β 𝑽=γ%δ}
+https://example.com?%CE%B1%%CE%B2=%CE%B3%%CE%B4
+https://example.com?α%β=γ%δ
+
+# Path and Query with quoted syntax characters
+# {𝑺=https:// 𝑯=example.com 𝑷=α 𝑷=b/?#c 𝑸=αβ 𝑽=γ&ζ=#Ξ 𝑸=k 𝑽=v 𝑭=frag}
+https://example.com/%CE%B1/b/%3F#c?%CE%B1%CE%B2=%CE%B3&%CE%B6=#%CE%9E&k=v#frag
+https://example.com/α/b%2F%3F%23c?αβ=γ%26ζ%3D%23Ξ&k=v#frag
+
+
+# Wikipedia test cases
+
+# {𝑺=https:// 𝑯=ru.wikinews.org 𝑷=wiki 𝑷=Категория:Вселенная}
+https://ru.wikinews.org/wiki/%D0%9A%D0%B0%D1%82%D0%B5%D0%B3%D0%BE%D1%80%D0%B8%D1%8F:%D0%92%D1%81%D0%B5%D0%BB%D0%B5%D0%BD%D0%BD%D0%B0%D1%8F
+https://ru.wikinews.org/wiki/Категория:Вселенная
+
+# {𝑺=https:// 𝑯=av.wikipedia.org 𝑷=wiki 𝑷=Ракь_(планета)}
+https://av.wikipedia.org/wiki/%D0%A0%D0%B0%D0%BA%D1%8C_(%D0%BF%D0%BB%D0%B0%D0%BD%D0%B5%D1%82%D0%B0)
+https://av.wikipedia.org/wiki/Ракь_(планета)
+
+# {𝑺=https:// 𝑯=bo.wikipedia.org 𝑷=wiki 𝑷=སའི་གོ་ལ།}
+https://bo.wikipedia.org/wiki/%E0%BD%A6%E0%BD%A0%E0%BD%B2%E0%BC%8B%E0%BD%82%E0%BD%BC%E0%BC%8B%E0%BD%A3%E0%BC%8D
+https://bo.wikipedia.org/wiki/སའི་གོ་ལ%E0%BC%8D
+
+# {𝑺=https:// 𝑯=fiu-vro.wikipedia.org 𝑷=wiki 𝑷=Maa_(hod'otäht)}
+https://fiu-vro.wikipedia.org/wiki/Maa_(hod'ot%C3%A4ht)
+https://fiu-vro.wikipedia.org/wiki/Maa_(hod'otäht)
+
+# {𝑺=https:// 𝑯=ty.wikipedia.org 𝑷=wiki 𝑷=’Afirita}
+https://ty.wikipedia.org/wiki/%E2%80%99Afirita
+https://ty.wikipedia.org/wiki/’Afirita
+
+# {𝑺=https:// 𝑯=ab.wikipedia.org 𝑷=wiki 𝑷=Вашингтон,_Џьорџь}
+https://ab.wikipedia.org/wiki/%D0%92%D0%B0%D1%88%D0%B8%D0%BD%D0%B3%D1%82%D0%BE%D0%BD,_%D0%8F%D1%8C%D0%BE%D1%80%D1%9F%D1%8C
+https://ab.wikipedia.org/wiki/Вашингтон,_Џьорџь
+
+# {𝑺=https:// 𝑯=mni.wikipedia.org 𝑷=wiki 𝑷=ꯅ꯭ꯌꯨ_ꯌꯣꯔ꯭ꯛ_ꯁꯤꯇꯤꯒꯤ_ꯌꯨ.ꯑꯦꯁ.}
+https://mni.wikipedia.org/wiki/%EA%AF%85%EA%AF%AD%EA%AF%8C%EA%AF%A8_%EA%AF%8C%EA%AF%A3%EA%AF%94%EA%AF%AD%EA%AF%9B_%EA%AF%81%EA%AF%A4%EA%AF%87%EA%AF%A4%EA%AF%92%EA%AF%A4_%EA%AF%8C%EA%AF%A8.%EA%AF%91%EA%AF%A6%EA%AF%81%2E
+https://mni.wikipedia.org/wiki/ꯅ꯭ꯌꯨ_ꯌꯣꯔ꯭ꯛ_ꯁꯤꯇꯤꯒꯤ_ꯌꯨ.ꯑꯦꯁ%2E
+
+# {𝑺=https:// 𝑯=azb.wikipedia.org 𝑷=wiki 𝑷=واشینقتن،_دی.سی.}
+https://azb.wikipedia.org/wiki/%D9%88%D8%A7%D8%B4%DB%8C%D9%86%D9%82%D8%AA%D9%86%D8%8C_%D8%AF%DB%8C.%D8%B3%DB%8C%2E
+https://azb.wikipedia.org/wiki/واشینقتن،_دی.سی%2E
+
+# {𝑺=https:// 𝑯=mad.wikipedia.org 𝑷=wiki 𝑷=Tasè’}
+https://mad.wikipedia.org/wiki/Tas%C3%A8%E2%80%99
+https://mad.wikipedia.org/wiki/Tasè%E2%80%99
+
+# {𝑺=https:// 𝑯=wuu.wikipedia.org 𝑷=wiki 𝑷=聖保羅（巴西）}
+https://wuu.wikipedia.org/wiki/%E8%81%96%E4%BF%9D%E7%BE%85%EF%BC%88%E5%B7%B4%E8%A5%BF%EF%BC%89
+https://wuu.wikipedia.org/wiki/聖保羅（巴西）
+
+# {𝑺=https:// 𝑯=vep.wikipedia.org 𝑷=wiki 𝑷=Brüssel'}
+https://vep.wikipedia.org/wiki/Br%C3%BCssel%27
+https://vep.wikipedia.org/wiki/Brüssel%27
+
+# {𝑺=https:// 𝑯=tw.wikipedia.org 𝑷=wiki 𝑷=Wiase_Nyinaa_Wɛbsaet_(_World_Wide_Web;_WWW_)}
+https://tw.wikipedia.org/wiki/Wiase_Nyinaa_W%C9%9Bbsaet_(_World_Wide_Web;_WWW_)
+https://tw.wikipedia.org/wiki/Wiase_Nyinaa_Wɛbsaet_(_World_Wide_Web;_WWW_)
+
+# {𝑺=https:// 𝑯=ja.wikibooks.org 𝑷=wiki 𝑷=植物学 𝑷=植物とはどのような生き物か？}
+https://ja.wikibooks.org/wiki/%E6%A4%8D%E7%89%A9%E5%AD%A6/%E6%A4%8D%E7%89%A9%E3%81%A8%E3%81%AF%E3%81%A9%E3%81%AE%E3%82%88%E3%81%86%E3%81%AA%E7%94%9F%E3%81%8D%E7%89%A9%E3%81%8B%EF%BC%9F
+https://ja.wikibooks.org/wiki/植物学/植物とはどのような生き物か%EF%BC%9F
+
+# {𝑺=https:// 𝑯=bn.wikibooks.org 𝑷=wiki 𝑷=উইকিশৈশব:দেশসমূহ_(অ-হ) 𝑷=ইসরায়েল}
+https://bn.wikibooks.org/wiki/%E0%A6%89%E0%A6%87%E0%A6%95%E0%A6%BF%E0%A6%B6%E0%A7%88%E0%A6%B6%E0%A6%AC:%E0%A6%A6%E0%A7%87%E0%A6%B6%E0%A6%B8%E0%A6%AE%E0%A7%82%E0%A6%B9_(%E0%A6%85-%E0%A6%B9)/%E0%A6%87%E0%A6%B8%E0%A6%B0%E0%A6%BE%E0%A6%AF%E0%A6%BC%E0%A7%87%E0%A6%B2
+https://bn.wikibooks.org/wiki/উইকিশৈশব:দেশসমূহ_(অ-হ)/ইসরায়েল
+
+# {𝑺=https:// 𝑯=haw.wikipedia.org 𝑷=wiki 𝑷=Puke_noi‘i_kū‘ikena}
+https://haw.wikipedia.org/wiki/Puke_noi%E2%80%98i_k%C5%AB%E2%80%98ikena
+https://haw.wikipedia.org/wiki/Puke_noi‘i_kū‘ikena
+
+# {𝑺=https:// 𝑯=new.wikipedia.org 𝑷=wiki 𝑷=विन्सेन्ट_भ्यान_ग:}
+https://new.wikipedia.org/wiki/%E0%A4%B5%E0%A4%BF%E0%A4%A8%E0%A5%8D%E0%A4%B8%E0%A5%87%E0%A4%A8%E0%A5%8D%E0%A4%9F_%E0%A4%AD%E0%A5%8D%E0%A4%AF%E0%A4%BE%E0%A4%A8_%E0%A4%97%3A
+https://new.wikipedia.org/wiki/विन्सेन्ट_भ्यान_ग%3A
+
+# {𝑺=https:// 𝑯=he.wikipedia.org 𝑷=wiki 𝑷=נאט"ו}
+https://he.wikipedia.org/wiki/%D7%A0%D7%90%D7%98%22%D7%95
+https://he.wikipedia.org/wiki/נאט"ו
+
+# {𝑺=https:// 𝑯=ks.wikipedia.org 𝑷=wiki 𝑷=شارٕک۔}
+https://ks.wikipedia.org/wiki/%D8%B4%D8%A7%D8%B1%D9%95%DA%A9%DB%94
+https://ks.wikipedia.org/wiki/شارٕک%DB%94
+
+# {𝑺=https:// 𝑯=zh.wikipedia.org 𝑷=wiki 𝑷=联合国教育、科学及文化组织}
+https://zh.wikipedia.org/wiki/%E8%81%94%E5%90%88%E5%9B%BD%E6%95%99%E8%82%B2%E3%80%81%E7%A7%91%E5%AD%A6%E5%8F%8A%E6%96%87%E5%8C%96%E7%BB%84%E7%BB%87
+https://zh.wikipedia.org/wiki/联合国教育、科学及文化组织
+
+# {𝑺=https:// 𝑯=am.wikipedia.org 𝑷=wiki 𝑷=«የሰብዓዊ_መብት_አቀፋዊ_መግለጽ»}
+https://am.wikipedia.org/wiki/%C2%AB%E1%8B%A8%E1%88%B0%E1%89%A5%E1%8B%93%E1%8B%8A_%E1%88%98%E1%89%A5%E1%89%B5_%E1%8A%A0%E1%89%80%E1%8D%8B%E1%8B%8A_%E1%88%98%E1%8C%8D%E1%88%88%E1%8C%BD%C2%BB
+https://am.wikipedia.org/wiki/«የሰብዓዊ_መብት_አቀፋዊ_መግለጽ%C2%BB
+
+# {𝑺=https:// 𝑯=bo.wikipedia.org 𝑷=wiki 𝑷=༼མ་ཧ་བ་ར་ཏ།༽}
+https://bo.wikipedia.org/wiki/%E0%BC%BC%E0%BD%98%E0%BC%8B%E0%BD%A7%E0%BC%8B%E0%BD%96%E0%BC%8B%E0%BD%A2%E0%BC%8B%E0%BD%8F%E0%BC%8D%E0%BC%BD
+https://bo.wikipedia.org/wiki/༼མ་ཧ་བ་ར་ཏ།༽
+
+# {𝑺=https:// 𝑯=am.wikipedia.org 𝑷=wiki 𝑷=አፈ፡ታሪክ}
+https://am.wikipedia.org/wiki/%E1%8A%A0%E1%8D%88%E1%8D%A1%E1%89%B3%E1%88%AA%E1%8A%AD
+https://am.wikipedia.org/wiki/አፈ፡ታሪክ
+
+# {𝑺=https:// 𝑯=it.wikibooks.org 𝑷=wiki 𝑷=Questo_è_l'ebraismo!}
+https://it.wikibooks.org/wiki/Questo_%C3%A8_l'ebraismo%21
+https://it.wikibooks.org/wiki/Questo_è_l'ebraismo%21
+
+# {𝑺=https:// 𝑯=my.wikipedia.org 𝑷=wiki 𝑷=ခို၊_ချိုးနှင့်_အလားတူငှက်များ}
+https://my.wikipedia.org/wiki/%E1%80%81%E1%80%AD%E1%80%AF%E1%81%8A_%E1%80%81%E1%80%BB%E1%80%AD%E1%80%AF%E1%80%B8%E1%80%94%E1%80%BE%E1%80%84%E1%80%B7%E1%80%BA_%E1%80%A1%E1%80%9C%E1%80%AC%E1%80%B8%E1%80%90%E1%80%B0%E1%80%84%E1%80%BE%E1%80%80%E1%80%BA%E1%80%99%E1%80%BB%E1%80%AC%E1%80%B8
+https://my.wikipedia.org/wiki/ခို၊_ချိုးနှင့်_အလားတူငှက်များ
+

From 1bdbf54b697cc0b0d7470c7e08d148e9d1d26910 Mon Sep 17 00:00:00 2001
From: Mark Davis <mark@unicode.org>
Date: Sat, 27 Jun 2026 17:13:30 +0200
Subject: [PATCH 2/8] ICU-23431 Moved test files, fleshing out tests WIP

---
 .../com/ibm/icu/dev/test/links/LinkTest.java  | 86 ++++++++++++++++---
 .../links/LinkDetectionTest.txt               |  0
 .../links/LinkFormattingTest.txt              |  0
 3 files changed, 76 insertions(+), 10 deletions(-)
 rename icu4j/main/core/src/test/resources/com/ibm/icu/dev/data/{testdata => unicode}/links/LinkDetectionTest.txt (100%)
 rename icu4j/main/core/src/test/resources/com/ibm/icu/dev/data/{testdata => unicode}/links/LinkFormattingTest.txt (100%)

diff --git a/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java b/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
index 063dcd429a90..3b9101a843ed 100644
--- a/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
+++ b/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
@@ -1,8 +1,12 @@
 package com.ibm.icu.dev.test.links;
 
 import com.ibm.icu.dev.test.CoreTestFmwk;
+import com.ibm.icu.dev.test.TestUtil;
+import com.ibm.icu.impl.locale.XCldrStub.Splitter;
 import com.ibm.icu.util.LinkUtilities;
 import com.ibm.icu.util.LinkUtilities.Extent;
+import java.io.BufferedReader;
+import java.io.IOException;
 import org.junit.Test;
 import org.junit.runner.RunWith;
 import org.junit.runners.JUnit4;
@@ -39,15 +43,77 @@ public void testSimpleScanBackEmail() {
         assertEquals(source, expected, actual);
     }
 
-    //    @Test
-    //    public void testAgainstFiles() {
-    //    	// for later
-    //    	try (Stream<String> lines = Files.lines(Paths.get("file.txt"))) {
-    //    	    lines.forEach(dosomething);
-    //    	} catch (IOException e) {
-    //    	    e.printStackTrace();
-    //    	}
-    //
-    //    }
+    @Test
+    public void testAgainstFiles() {
+        //        	# All parts
+        //        	# {𝑺=https:// 𝑯=example.com 𝑷=αβγ 𝑷=δεζ 𝑸=θ 𝑽=ικλ 𝑸=μ 𝑽=γξο 𝑭=πρς}
+        //
+        //	https://example.com/%CE%B1%CE%B2%CE%B3/%CE%B4%CE%B5%CE%B6?%CE%B8=%CE%B9%CE%BA%CE%BB&%CE%BC=%CE%B3%CE%BE%CE%BF#%CF%80%CF%81%CF%82
+        //        	https://example.com/αβγ/δεζ?θ=ικλ&μ=γξο#πρς
+
+        String expectedPrefix = "https://example.com"; // for now, all the same
 
+        try (BufferedReader idnaTestFile =
+                TestUtil.getUtf8DataReader("unicode/links/LinkFormattingTest.txt"); ) {
+            String pqf = null;
+            boolean firstLine = true;
+            for (String line : (Iterable<String>) idnaTestFile.lines()::iterator) {
+                if (line.startsWith("# {")) {
+                    pqf = assemblePqf(line);
+                    firstLine = true;
+                } else if (line.isBlank() || line.startsWith("#")) {
+                    continue;
+                } else if (!line.startsWith(expectedPrefix)) {
+                    throw new IllegalArgumentException("Expand the test to cover other S/H");
+                } else if (firstLine) {
+                    String actual =
+                            expectedPrefix
+                                    + LinkUtilities.escapePathQueryFragment(pqf, Extent.MAXIMAL);
+                    assertEquals("maximum " + pqf, line, actual);
+                    firstLine = false;
+                } else {
+                    String actual =
+                            expectedPrefix
+                                    + LinkUtilities.escapePathQueryFragment(pqf, Extent.MINIMAL);
+                    assertEquals("maximum " + pqf, line, actual);
+                    firstLine = false;
+                }
+            }
+        } catch (IOException e) {
+            e.printStackTrace();
+        }
+    }
+
+    private String assemblePqf(String line) {
+        line = line.substring(3, line.length() - 1); // subtract the # { and the }
+        StringBuilder sb = new StringBuilder();
+        char query = '?';
+        for (String item : Splitter.on(' ').splitToList(line)) {
+            int eq = item.indexOf('=');
+            String type = item.substring(0, eq);
+            String rest = item.substring(eq + 1);
+            //	        	# {𝑺=https:// 𝑯=example.com 𝑷=αβγ 𝑷=δεζ 𝑸=θ 𝑽=ικλ 𝑸=μ 𝑽=γξο 𝑭=πρς}
+            switch (type) {
+                case "𝑺":
+                case "𝑯":
+                    break;
+                case "𝑷":
+                    sb.append('/').append(rest);
+                    break;
+                case "𝑸":
+                    sb.append(query).append(rest);
+                    query = '&';
+                    break;
+                case "𝑽":
+                    sb.append('=').append(rest);
+                    break;
+                case "𝑭":
+                    sb.append('#').append(rest);
+                    break;
+                default:
+                    throw new IllegalArgumentException("Unexpected line: " + item);
+            }
+        }
+        return sb.toString();
+    }
 }
diff --git a/icu4j/main/core/src/test/resources/com/ibm/icu/dev/data/testdata/links/LinkDetectionTest.txt b/icu4j/main/core/src/test/resources/com/ibm/icu/dev/data/unicode/links/LinkDetectionTest.txt
similarity index 100%
rename from icu4j/main/core/src/test/resources/com/ibm/icu/dev/data/testdata/links/LinkDetectionTest.txt
rename to icu4j/main/core/src/test/resources/com/ibm/icu/dev/data/unicode/links/LinkDetectionTest.txt
diff --git a/icu4j/main/core/src/test/resources/com/ibm/icu/dev/data/testdata/links/LinkFormattingTest.txt b/icu4j/main/core/src/test/resources/com/ibm/icu/dev/data/unicode/links/LinkFormattingTest.txt
similarity index 100%
rename from icu4j/main/core/src/test/resources/com/ibm/icu/dev/data/testdata/links/LinkFormattingTest.txt
rename to icu4j/main/core/src/test/resources/com/ibm/icu/dev/data/unicode/links/LinkFormattingTest.txt

From cb1ea2bf2463029ad8268a995e8ad1afa6e2c5a3 Mon Sep 17 00:00:00 2001
From: Mark Davis <mark@unicode.org>
Date: Wed, 1 Jul 2026 21:41:21 +0300
Subject: [PATCH 3/8] ICU-23431 More progress but not done

---
 .../icu/impl/links/LinkHandlingUtilities.java |  18 +-
 .../java/com/ibm/icu/util/LinkUtilities.java  |   2 +-
 .../com/ibm/icu/dev/test/links/LinkTest.java  | 361 +++++++++++++++++-
 3 files changed, 371 insertions(+), 10 deletions(-)

diff --git a/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java b/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java
index e99420751518..c498a99a4fbe 100644
--- a/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java
+++ b/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java
@@ -83,7 +83,7 @@ private static String quote(String s) {
 
     /** TODO use real property */
     private static int getOpening(int cp) {
-        return cp == '>' ? '<' : UCharacter.getIntPropertyValue(cp, UProperty.BIDI_PAIRED_BRACKET);
+        return cp == '>' ? '<' : UCharacter.getBidiPairedBracket(cp);
     }
 
     /** Defines the LinkTermination property TODO use real property */
@@ -159,7 +159,19 @@ public static int scanEmailBackwards(CharSequence source, int hardStart, int dom
             throw new IllegalArgumentException("Scanning must start after an '@' sign");
         }
         int result = validEmailLocalPart.spanBack(source, domainStart - 1, SpanCondition.SIMPLE);
-        return result >= hardStart ? result : domainStart - 1;
+        if (result == domainStart - 1) {
+            return domainStart;
+        } else if (result < hardStart) {
+            result = hardStart;
+        }
+        String localPart = source.subSequence(result, domainStart - 1).toString();
+        if (localPart.startsWith(".") || localPart.endsWith(".") || localPart.contains("..")) {
+            return domainStart;
+        }
+        if (source.toString().substring(0, result).endsWith("mailto:")) {
+            result -= "mailto:".length();
+        }
+        return result;
     }
 
     private static String getGeneralCategory(int property, int codePoint, int nameChoice) {
@@ -912,7 +924,7 @@ private static void appendPercentEscaped(StringBuilder output, String source) {
         byte[] bytes = source.getBytes(StandardCharsets.UTF_8);
         for (int i = 0; i < bytes.length; ++i) {
             output.append('%');
-            output.append(Utility.hex(bytes[i], 2));
+            output.append(Utility.hex(bytes[i] & 0xFF, 2));
         }
     }
 
diff --git a/icu4j/main/core/src/main/java/com/ibm/icu/util/LinkUtilities.java b/icu4j/main/core/src/main/java/com/ibm/icu/util/LinkUtilities.java
index 74eda248f349..b944ab2f06c5 100644
--- a/icu4j/main/core/src/main/java/com/ibm/icu/util/LinkUtilities.java
+++ b/icu4j/main/core/src/main/java/com/ibm/icu/util/LinkUtilities.java
@@ -32,7 +32,7 @@ public static int scanPathQueryFragment(CharSequence source, int start, int limi
      * Lower level utility for finding the start of an email address in text. It assumes that the
      * limit position is immediately before an '@' + identified domain name. The purpose of this
      * routine is for fitting into algorithms that are already in use, just handling for the email
-     * `local-part`.
+     * `local-part`. It does not scan back through "mailto:".
      *
      * @param source the text to be scanned
      * @param start the position that is the earliest that should be considered in a backwards scan
diff --git a/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java b/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
index 3b9101a843ed..cdde9458efdb 100644
--- a/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
+++ b/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
@@ -3,24 +3,69 @@
 import com.ibm.icu.dev.test.CoreTestFmwk;
 import com.ibm.icu.dev.test.TestUtil;
 import com.ibm.icu.impl.locale.XCldrStub.Splitter;
+import com.ibm.icu.lang.UCharacter;
+import com.ibm.icu.text.IDNA;
+import com.ibm.icu.text.UnicodeSet;
+import com.ibm.icu.text.UnicodeSet.SpanCondition;
 import com.ibm.icu.util.LinkUtilities;
 import com.ibm.icu.util.LinkUtilities.Extent;
 import java.io.BufferedReader;
 import java.io.IOException;
+import java.util.ArrayList;
+import java.util.List;
+import java.util.Set;
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
 import org.junit.Test;
 import org.junit.runner.RunWith;
 import org.junit.runners.JUnit4;
 
+/**
+ * Test the APIs in LinkUtilties using some basic tests, plus the test data from UTS58 Note: the
+ * UTC58 tests check whole links, since that is easiest for implementers to write tests against.<br>
+ * However, the Unicode testdata currently takes certain positions on edge cases that some client
+ * software might not agree with. A simple case is that it only recognizes TLDs that are valid, not
+ * just well formed. It mostly handles this by not using invalid TLDs. So it doesn't use
+ * "example.somefuturetld" as a test case. But it does take some positions that theoretically
+ * clients might disagree with, such as not recognizing "efg.com" in "abc..efg.com". This currently
+ * tests with code that duplicates that.<br>
+ * There are also some edge cases
+ *
+ * <table>
+ * <tr><th>Text</th><th>Approach</th></tr>
+ * <tr><td>dav.is@dav.is/dav.is
+ * </td><td>Note that a string may both be a valid local-part and a valid domain name.
+ * The UTS58 test data recognizes the start of a link;
+ * recognizing the email address (dav.is@dav.is), and skipping the PDF (/dav.is).
+ * </td></tr>
+ * <tr><td>mailto:αβγ..δεζ@example.com
+ * </td><td>A failing email local-part (αβγ..δεζ) fails as a whole (eg δεζ@example.com is not recognized)
+ * </td></tr>
+ * </table>
+ */
 @RunWith(JUnit4.class)
 public class LinkTest extends CoreTestFmwk {
 
     /** Constructor */
     public LinkTest() {}
 
+    @Test
+    public void test1PairedBracket() {
+        // int actual = // UCharacter.getIntPropertyValue(')', UProperty.BIDI_PAIRED_BRACKET);
+        int actual = UCharacter.getBidiPairedBracket(')');
+        assertEquals("", '(', actual);
+    }
+
     @Test
     public void testSimpleScan() {
-        String source = "/αβγ?def#ghi ";
-        int expected = source.indexOf(" ");
+        checkScan("/α(β) on…", "α(β)");
+        checkScan("/αβγ?def#ghi ", "ghi");
+        // checkScan("⸠https://ja.wikipedia.org/wiki/フィンセント・ファン・ゴッホ", "ゴッホ");
+    }
+
+    private void checkScan(String toCheck, String expectedAfter) {
+        String source = toCheck;
+        int expected = source.indexOf(expectedAfter) + expectedAfter.length();
         int actual = LinkUtilities.scanPathQueryFragment(source, 0, source.length());
         assertEquals(source, expected, actual);
     }
@@ -35,16 +80,187 @@ public void testSimpleEscape() {
 
     @Test
     public void testSimpleScanBackEmail() {
-        String source = "See αβγ@example.com";
-        int start = 0;
+        checkEmail("See αβγ@", 0, "αβγ");
+        checkEmail("See mailto:αβγ@", 0, "mailto:α");
+        checkEmail("See mailto:αβγ.δεζ.@", 0, null);
+        checkEmail("See mailto:αβγ.δεζ@", 0, "mailto:α");
+    }
+
+    private void checkEmail(String source, int start, String localPartStart) {
         int end = source.indexOf('@') + 1;
-        int expected = source.indexOf("αβγ");
+        int expected = localPartStart == null ? source.length() : source.indexOf(localPartStart);
         int actual = LinkUtilities.scanBackEmailLocalPart(source, start, end);
         assertEquals(source, expected, actual);
     }
+    @Test
+    public void testHackDomainName() {
+    	String source = // "普遍适用测试。我爱你";
+    	"موريتانيا.xn-----ctdbabcfhu9c2b9l1acccr4c";
+        IntRange domain = new IntRange();
+		boolean actual = hackFindDomainName(source, 0, source.length(), false, domain);
+		assertEquals(source, true, actual);
+		assertTrue(source, domain.start == 0 && domain.end == source.length());
+    }
+
+    // Would like to use the following for JUnit, but can't
+    //  @ParameterizedTest
+    //  @MethodSource("testItemsProvider")
+
+    Pattern SCHEME = Pattern.compile("https?://$");
+    Pattern PORT = Pattern.compile(":\\d+");
+
+    final Set<String> skip = Set.of("See http://-foo.example-.com. on…");
 
     @Test
-    public void testAgainstFiles() {
+    public void testAgainstLinkDetectionTest() {
+        List<String> failures = new ArrayList<>();
+        try (BufferedReader idnaTestFile =
+                TestUtil.getUtf8DataReader("unicode/links/LinkDetectionTest.txt"); ) {
+            for (final String line : (Iterable<String>) idnaTestFile.lines()::iterator) {
+                if (line.startsWith("#") || line.isBlank() || skip.contains(line)) {
+                    continue;
+                }
+                StringBuilder actual = new StringBuilder();
+                String rawLine2 = line.replace("⸠", "").replace("⸡", "");
+                int start = 0;
+                IntRange domain = new IntRange();
+                IntRange domainAtDomain = new IntRange();
+                try {
+                    while (hackFindDomainName(rawLine2, start, rawLine2.length(), false, domain)) {
+                        boolean haveEmail = false;
+                        // check for '@' before the domain name
+                        // if there is one, we see if there is an email address. If not,
+                        // we skip over the domain name
+                        if (domain.start > 0
+                                && UCharacter.codePointBefore(rawLine2, domain.start) == '@') {
+                            // handle email
+                            // check backwards for email
+                            int local_part_start =
+                                    LinkUtilities.scanBackEmailLocalPart(
+                                            rawLine2, start, domain.start);
+                            if (local_part_start != domain.start) {
+                                domain.start = local_part_start;
+                                haveEmail = true;
+                            } else {
+                                // we skip over the domain name
+                                actual.append(rawLine2.subSequence(start, domain.end));
+                                start = domain.end;
+                                continue;
+                            }
+
+                        } else { // we could have chanced on a case where the domain name is also
+                            // a local_part (like john.uk@foo.com)
+                            // For that case we check if there is an @ sign *after* the end of the
+                            // domain name.
+                            // If so:
+                            // - we scan forwards (after the @) to see if there is a domain name,
+                            // AND
+                            // - we scan backwards (before the @) to see if there is local_part, 
+                        	// AND
+                            // - we scan backwards before that to ensure that there is no https?://
+                            // if both are true, we have an email address.
+                            // if only the first is true, we skip over the 2nd domain name;
+                            // ie, we skip over " @example.com"
+                            if (domain.end < rawLine2.length()
+                                    && UCharacter.codePointAt(rawLine2, domain.end) == '@') {
+                                if (hackFindDomainName(
+                                        rawLine2,
+                                        domain.end + 1,
+                                        rawLine2.length(),
+                                        true,
+                                        domainAtDomain)) {
+                                    domain.start =
+                                            LinkUtilities.scanBackEmailLocalPart(
+                                                    rawLine2, start, domain.end + 1);
+                                    boolean failing = domain.start == domain.end + 1;
+                                    if (!failing) {
+                                        // slash is a valid email character, so check
+                                        // two characters after
+                                        Matcher schemeMatcher = SCHEME.matcher(rawLine2);
+                                        schemeMatcher.region(start, domain.start + 2);
+                                        failing = schemeMatcher.find();
+                                    }
+                                    if (failing) {
+                                        // failed to find local part, or found scheme
+                                        // we failed, so copy up to here so we don't rescan
+                                        actual.append(
+                                                rawLine2.subSequence(start, domainAtDomain.end));
+                                        start = domainAtDomain.end;
+                                        continue;
+                                    }
+                                    if (UCharacter.codePointBefore(rawLine2, domainAtDomain.end)
+                                            == '.') {
+                                        // backup; might be end of sentence
+                                        domainAtDomain.end--;
+                                    }
+                                    domain.end = domainAtDomain.end;
+                                    haveEmail = true;
+                                }
+                            }
+                        }
+                        // we only check for PQF if we don't have an email address
+                        if (!haveEmail) {
+                            // handle PQF
+                            // check backwards for scheme "[a-z]://"
+                            Matcher schemeMatcher = SCHEME.matcher(rawLine2);
+                            schemeMatcher.region(start, domain.start);
+                            if (schemeMatcher.find()) {
+                                domain.start = schemeMatcher.start();
+                            } else {
+                                domain.start = domain.start;
+                            }
+
+                            Matcher portMatcher = PORT.matcher(rawLine2);
+                            portMatcher.region(domain.end, rawLine2.length());
+                            if (portMatcher.lookingAt()) {
+                                domain.end = portMatcher.end();
+                            }
+
+                            // check forwards for PQF
+                            int pqfEnd =
+                                    LinkUtilities.scanPathQueryFragment(
+                                            rawLine2, domain.end, rawLine2.length());
+                            if (pqfEnd == domain.end) {
+                                // special case dot at end of domain name and no PQF
+                                if (UCharacter.codePointBefore(rawLine2, domain.end) == '.') {
+                                    // backup; might be end of sentence
+                                    pqfEnd--;
+                                }
+                            }
+                            domain.end = pqfEnd;
+                        }
+                        if (start > domain.start) {
+                            int debug = 0;
+                        }
+                        actual.append(rawLine2.subSequence(start, domain.start))
+                                .append("⸠")
+                                .append(rawLine2.subSequence(domain.start, domain.end))
+                                .append("⸡");
+                        start = domain.end;
+                    }
+                    actual.append(rawLine2.substring(start));
+
+                    assertEquals("", line, actual.toString());
+                    System.out.println("OK\t" + line);
+                } catch (AssertionError e) {
+                    failures.add("FAIL " + e.getMessage());
+                    System.out.println("FAIL\t" + line + "\t  " + actual.toString());
+                }
+            }
+            if (!failures.isEmpty()) {
+                fail(
+                        "Test finished with "
+                                + failures.size()
+                                + " failures:\n"
+                                + String.join("\n", failures));
+            }
+        } catch (IOException e) {
+            e.printStackTrace();
+        }
+    }
+
+    @Test
+    public void testAgainstLinkFormattingTest() {
         //        	# All parts
         //        	# {𝑺=https:// 𝑯=example.com 𝑷=αβγ 𝑷=δεζ 𝑸=θ 𝑽=ικλ 𝑸=μ 𝑽=γξο 𝑭=πρς}
         //
@@ -116,4 +332,137 @@ private String assemblePqf(String line) {
         }
         return sb.toString();
     }
+
+    class IntRange {
+        int start;
+        int end;
+
+        @Override
+        public String toString() {
+            return start + ";" + end;
+        }
+    }
+
+    Pattern DOT = Pattern.compile("[.。]");
+    Pattern DOT2 = Pattern.compile("[.。][.。]");
+    /**
+     * The UTS58 tests take whole domain names, because that is the easiest for clients to test. We
+     * have to hack it because ICU IDN support doesn't provide enough API. Returns true if found,
+     * with the range in result.
+     *
+     * @param mustBeAtStart 
+     * TODO check with isValidDomain
+     */
+    boolean hackFindDomainName(
+            CharSequence source, int start, int limit, boolean mustBeAtStart, IntRange result) {
+        if (mustBeAtStart && start < limit) {
+            int firstCodePoint = UCharacter.codePointAt(source, start);
+            if (!HACK_DOMAIN_SET.contains(firstCodePoint)) {
+                return false;
+            }
+        }
+        Matcher dot = DOT.matcher(source);
+        Matcher dot2 = DOT2.matcher(source);
+        while (start < limit) {
+            int endNot = HACK_DOMAIN_SET.span(source, start, SpanCondition.NOT_CONTAINED);
+            if (endNot > limit) {
+                return false;
+            }
+            int end = HACK_DOMAIN_SET.span(source, endNot, SpanCondition.SIMPLE);
+            if (end > limit) {
+                end = limit;
+            }
+            if (end == endNot) {
+            	return false;
+            }
+            // Check that the possible domain name (endNot..end) is of the right form
+            // otherwise continue after end.
+            // Must not start with [.。], and must find a [.。] before the end
+            dot.region(endNot, end);
+            boolean found = dot.find();
+            if (!found || dot.start() == endNot || dot.end() == end) {
+                start = end;
+                continue;	
+            }
+            // Must not find [.。][.。]
+            dot2.region(endNot, end);
+            if (dot2.find()) {
+                start = end;
+                continue;
+            }
+            result.start = endNot;
+            result.end = end;
+            return true;
+        }
+        return false;
+    }
+
+    private final UnicodeSet HACK_DOMAIN_SET =
+            new UnicodeSet(
+                            "[。.\\-0-9a-z·ß-öø-ÿāăą ćĉċčďđēĕėęěĝğġģĥħĩ"
+                                    + "īĭį ıĵķĸĺļľłńņňŋōŏőœŕŗřśŝ şšţťŧũūŭůűųŵŷźżžƀƃƅƈƌ ƍƒƕƙ-ƛƞơƣƥƨƪƫƭưƴƶƹ-ƻƽ-ǃ ǎǐ"
+                                    + "ǒǔǖǘǚǜǝǟǡǣǥǧǩǫǭǯǰǵǹ ǻǽǿȁȃȅȇȉȋȍȏȑȓȕȗșțȝȟȡȣ ȥȧȩȫȭȯȱȳ-ȹȼȿɀɂɇɉɋɍɏ-ʯ ʹ-ˁˆ-ˑˬˮ̀-̿͂͆-͎͐-ͯͱͳ͵ ͷ-"
+                                    + "͹ ͻ-ͽ΀-΃΋΍ ΐ΢ ά-ώϗϙϛϝϟϡϣϥϧϩϫϭϯϳϸϻϼа-џ ѡѣѥѧѩѫѭѯѱѳѵѷѹѻѽѿҁ҃-҇ҋ ҍҏґғҕҗҙқҝҟҡңҥҧҩҫ"
+                                    + "ҭүұҳҵ ҷҹһҽҿӂӄӆӈӊӌӎӏӑӓӕӗәӛӝӟ ӡӣӥӧөӫӭӯӱӳӵӷӹӻӽӿԁԃԅԇԉ ԋԍԏԑԓԕԗԙԛԝԟԡԣԥԧԩԫԭԯ԰՗-ՙ ՠ-ֆֈ"
+                                    + "֋֌֐-ׇֽֿׁׂׅׄ-׿ؐ-ؚ ؠ-ؿف-٩ٮ-ٴٹ-ۓە-ۜ۟-۪ۨ-ۿ܎ ܐ-ߵ߻-߽ ࠀ-࠯࠿-࡝࡟-ࢇ ࢉ-࢏࢒-ࣣ࣡-ॗ ॠ-ॣ०-९ॱ-৛৞ ৠ-ৱৼ"
+                                    + "৾-ਲ਴ ਵ਷-੘ ੜ੝੟-ੵ੷-૯૲-୛୞-୯ ୱ୸-௯௻-౶ ಀ-ಃಅ-ൎ൐-ൗ ൟ-൯ൺ-ෳ෵-าิ-฾ เ-๎๐-๙๜-າິ-"
+                                    + "໛ ໞ-ༀ་༘༙༠-༩༹༵༷༾-གང-ཌཎ-ད ན-བམ-ཛཝ-ཨཪ-ིེུ-ྀྂ-྄྆-ྒྔ-ྜྞ-ྡྣ-ྦྨ-ྫྭ-ྸྺ-྽࿆࿍࿛-၉ ၐ-ႝ჆჈-჌჎-ჺ ჽ-ჿሀ-፟፽-ᎏ᎚-᏷᏾᏿ ᐁ-"
+                                    + "ᙬᙯ-ᙿᚁ-ᚚ᚝-ᛪ ᛱ-᜴᜷-ឳា-៓ ៗៜ-៯៺-៿ ᠐-᤿᥁-᥃ ᥆-᧙᧛-᧝ ᨀ-᨝ ᨠ-᪟ ᪧ᪮-᪽ᪿ-᭍ ᭐-᭙᭫-᭳ᮀ-᯻ "
+                                    + "ᰀ-᰺ ᱀-ᱽᲊ-᲏᲻᲼᳈-᳔᳒-ᴫ ᴯᴻᵎᵫ-ᵷᵹ-ᶚ᷀-᷿ḁḃḅḇḉḋḍḏḑ ḓḕḗḙḛḝḟḡḣḥḧḩḫḭḯḱḳḵḷḹḻ ḽḿṁṃṅṇṉṋṍṏṑṓṕṗ"
+                                    + "ṙṛṝṟṡṣṥ ṧṩṫṭṯṱṳṵṷṹṻṽṿẁẃẅẇẉẋẍẏ ẑẓẕ-ẙẜẝẟạảấầẩẫậắằẳẵặẹ ẻẽếềểễệỉịọỏốồổỗộớờởỡợ ụủứừử"
+                                    + "ữựỳỵỷỹỻỽỿ-ἇἐ-἗἞-ἧ ἰ-ἷὀ-὇὎-὘὚὜὞ ὠ-ὧὰὲὴὶὸὺὼ὾὿ ᾰᾱ᾵ ᾶ῅ ῆῐ-ῒ῔-ῗ῜ ῠ-ῢῤ-ῧ"
+                                    + "῰῱῵ ῶ῿ ‌‍⁥⁲⁳₏₝-₟⃂-⃏⃱-⃿ ⅎↄ↌-↏␪-␿⑋-⑟⭴⭵ ⰰ-ⱟⱡⱥⱦⱨⱪⱬⱱⱳⱴⱶ-ⱻⲁⲃⲅⲇⲉⲋ ⲍⲏⲑⲓⲕⲗ"
+                                    + "ⲙⲛⲝⲟⲡⲣⲥⲧⲩⲫⲭⲯⲱⲳⲵ ⲷⲹⲻⲽⲿⳁⳃⳅⳇⳉⳋⳍⳏⳑⳓⳕⳗⳙⳛⳝⳟ ⳡⳣⳤⳬⳮ-⳱ⳳ-⳸ ⴀ-⵮⵱-ⷿ ⸯ⹞-⹿⺚⻴-⻿⿖-⿯ 々-〇〪-〭"
+                                    + "〼぀-゚ ゝゞァ-ヾ㄀-㄰㆏ ㆠ-ㆿ㇦-㇮ ㇰ-ㇿ㈟ 㐀-䶿一-꒏꓇-ꓽ ꔀ-ꘌꘐ-꘿ ꙁꙃꙅꙇꙉꙋꙍꙏꙑꙓꙕꙗꙙꙛꙝꙟꙡꙣ"
+                                    + "ꙥꙧꙩ ꙫꙭ-꙯ꙴ-꙽ꙿꚁꚃꚅꚇꚉꚋꚍꚏꚑꚓꚕꚗꚙ ꚛꚞ-ꛥ꛰꛱꛸-꛿ ꜗ-ꜟꜣꜥꜧꜩꜫꜭꜯ-ꜱꜳꜵꜷꜹꜻꜽꜿꝁꝃ ꝅꝇꝉꝋꝍꝏꝑꝓꝕꝗꝙꝛꝝꝟꝡꝣꝥꝧꝩꝫꝭ ꝯꝱ-ꝸꝺꝼꝿ"
+                                    + "ꞁꞃꞅꞇꞈꞌꞎꞏꞑꞓ-ꞕꞗꞙ ꞛꞝꞟꞡꞣꞥꞧꞩꞯꞵꞷꞹꞻꞽꞿꟁꟃꟈꟊꟍ꟏ ꟑꟓꟕꟗꟙꟛ꟝-꟰ ꟶꟷꟺ-ꠧ꠬-꠯꠺-ꡳ꡸-꣍ ꣐-ꣷꣻꣽ-꤭ꤰ-꥞꥽-꧀꧎-꧝ "
+                                    + "ꧠ-꩛ ꩠ-ꩶꩺ-ꫝꫠ-ꫯꫲ-ꭚꭠ-ꭨ꭬-꭯ ꯀ-ꯪ꯬-힯퟇-퟊퟼-퟿ 﨎﨏﨑﨓﨔﨟﨡﨣﨤﨧-﨩﩮﩯﫚-﫿﬇-﬒"
+                                    + "﬘-﬜ﬞ﬷﬽﬿﭂﭅︚-︯﹓﹧﹬-﹯ ﹳ﹵﻽﻾＀﾿-￁￈￉￐￑￘￙￝-￟￧￯-￸ 𐀀-𐃿𐄃-𐄆𐄴-𐄶𐆏𐆝"
+                                    + "-𐆟𐆡-𐇏𐇽-𐋠𐋼-𐌟𐌤-𐍀 𐍂-𐍉𐍋-𐎞 𐎠-𐏏𐏖-𐏿 𐐨-𐒯𐓔-𐕮𐕻𐖋𐖓𐖖-𐞀𐞆𐞱𐞻-𐡖 𐡠-𐡶𐢀-𐢦𐢰-"
+                                    + "𐣺 𐤀-𐤕𐤜-𐤞 𐤠-𐤾 𐥀-𐦻 "
+                                    + "𐦾𐦿𐧐𐧑 𐨀-𐨿𐩉-𐩏𐩙-𐩼 "
+                                    + "𐪀-𐪜𐪠-𐫇 𐫉-𐫪𐫷-𐬸 𐭀-"
+                                    + "𐭗 𐭠-𐭷 𐮀-𐮘𐮝-𐮨𐮰"
+                                    + "-𐱿𐲳-𐳹 𐴀-𐵏𐵦-𐵭 𐵯-"
+                                    + "𐶍𐶐-𐹟𐹿-𐺬𐺮-𐻏𐻙-𐼜"
+                                    + " 𐼧-𐽐𐽚-𐾅𐾊-𐿄𐿌-𑁆𑁎-𑁑 𑁦-𑂺𑃂-𑃌𑃎-𑄿 𑅄-𑅳𑅶-𑇄𑇉-𑇌𑇎-𑇚𑇜𑇠𑇵-𑈷𑈾-𑊨𑊪-𑏓𑏖𑏙-𑑊 𑑐-𑑙𑑜𑑞-"
+                                    + "𑓅 𑓇-𑗀𑗘-𑙀𑙄-𑙟𑙭-𑚸𑚺-𑜹 𑝀-𑠺𑠼-𑢟 𑣀-𑣩𑣳-𑥃𑥇-𑧡 𑧣-𑨾𑩇-𑪙𑪝𑪣-𑫿𑬊-𑯠𑯢-𑱀𑱆-𑱙𑱭-𑱯 "
+                                    + "𑱲-𑻶𑻹-𑽂 𑽐-𑾿𑿲-𑿾 𒀀-𒏿𒑯𒑵-𒿰𒿳-𓐯𓑀-𖩭 𖩰-𖫴𖫶-𖬶 𖭀-𖭃𖭆-𖭚𖭢-𖵬 𖵰-𖸿 𖹠-𖹿𖺛-𖺟𖺹-𖿡 "
+                                    + "𖿣-𖿳𖿷-𛲛𛲝𛲞𛲤-𜯿𜳽-𜳿𜺴-𜺹𜻑-𜻟𜻱-𜽏𜿄-𜿿𝃶-𝃿𝄧𝄨𝇫-𝇿𝉆-𝊿𝋔-𝋟𝋴-𝋿𝍗-𝍟𝍹-𝏿𝑕𝒝𝒠𝒡𝒣𝒤"
+                                    + "𝒧𝒨𝒭𝒺𝒼𝓄𝔆𝔋𝔌𝔕𝔝𝔺𝔿𝕅𝕇-𝕉𝕑𝚦𝚧𝟌𝟍𝨀-𝨶𝨻-𝩬𝩵𝪄𝪌-𞀯𞁮-𞅎𞅐-𞋾𞌀-𞗾𞘀-𞣆𞣐-𞣿 𞤢-𞥝𞥠-𞱰𞲵-𞴀𞴾-𞷿𞸄𞸠𞸣𞸥𞸦𞸨𞸳𞸸𞸺𞸼-𞹁𞹃-𞹆𞹈𞹊𞹌𞹐𞹓𞹕𞹖𞹘𞹚𞹜𞹞𞹠𞹣𞹥𞹦𞹫𞹳𞹸𞹽𞹿𞺊𞺜"
+                                    + "-𞺠𞺤𞺪𞺼-𞻯𞻲-𞿿🀬-🀯🂔-🂟🂯🂰🃀🃐🃶-🃿🆮-🇥🈃-🈏🈼-🈿🉉-🉏🉒-🉟🉦-🋿🛙-🛛🛭-🛯🛽-🛿🟚-🟟🟬-🟯🟱-"
+                                    + "🟿🠌-🠏🡈-🡏🡚-🡟🢈-🢏🢮🢯🢼-🢿🣂-🣏🣙-🣿🩘-🩟🩮🩯🩽-🩿🪋-🪍🫇🫉-🫌🫝🫞🫫-🫮🫹-🫿🮓🯻-🿽 𠀀-𯟿𯨞-𯿽"
+                                    + " 𰀀-𿿽񀀀-񏿽񐀀-񟿽񠀀-񯿽񰀀-񿿽򀀀-򏿽򐀀-򟿽򠀀-򯿽򰀀-򿿽󀀀-󏿽󐀀-󟿽󠀀󠀂-󠀟󠂀-󠃿󠇰-󯿽]")
+                    .closeOver(UnicodeSet.CASE_INSENSITIVE)
+                    .freeze();
+
+    static boolean isValidDomain(String domainName) {
+        if (domainName == null || domainName.isEmpty()) {
+            return false;
+        }
+
+        // Initialize the UTS#46 standard IDNA instance
+        IDNA idna = IDNA.getUTS46Instance(
+            IDNA.NONTRANSITIONAL_TO_ASCII | 
+            IDNA.NONTRANSITIONAL_TO_UNICODE | 
+            IDNA.CHECK_BIDI | 
+            IDNA.CHECK_CONTEXTJ | 
+            IDNA.CHECK_CONTEXTO | 
+            IDNA.USE_STD3_RULES
+        );
+
+        StringBuilder dest = new StringBuilder();
+        IDNA.Info info = new IDNA.Info();
+
+        try {
+            // Perform name-to-ASCII conversion/validation
+            idna.nameToASCII(domainName, dest, info);
+        } catch (IllegalArgumentException e) {
+            // Fails on severe structural issues
+            return false;
+        }
+
+        // Return true only if no errors were encountered during validation
+        return !info.hasErrors();
+    }
 }

From cfd2d4a77c48fd23ec0d92f51893b2c61e6e2562 Mon Sep 17 00:00:00 2001
From: Mark Davis <mark@unicode.org>
Date: Fri, 3 Jul 2026 12:42:32 +0300
Subject: [PATCH 4/8] ICU-23431 Fixed remaining test failures

---
 .../icu/impl/links/LinkHandlingUtilities.java |  18 +-
 .../com/ibm/icu/impl/locale/XCldrStub.java    |  14 +-
 .../java/com/ibm/icu/util/LinkUtilities.java  |  10 +-
 .../com/ibm/icu/dev/test/links/LinkTest.java  | 257 +++++++++++++-----
 4 files changed, 229 insertions(+), 70 deletions(-)

diff --git a/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java b/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java
index c498a99a4fbe..7e1188ecb616 100644
--- a/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java
+++ b/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java
@@ -3,6 +3,7 @@
 import com.ibm.icu.impl.IDNA2003;
 import com.ibm.icu.impl.UnicodeMap;
 import com.ibm.icu.impl.Utility;
+import com.ibm.icu.impl.links.LinkHandlingUtilities.UrlInternals;
 import com.ibm.icu.impl.locale.XCldrStub;
 import com.ibm.icu.impl.locale.XCldrStub.Splitter;
 import com.ibm.icu.lang.UCharacter;
@@ -17,6 +18,7 @@
 import java.nio.charset.StandardCharsets;
 import java.util.EnumMap;
 import java.util.EnumSet;
+import java.util.LinkedHashMap;
 import java.util.List;
 import java.util.Map;
 import java.util.Map.Entry;
@@ -916,7 +918,7 @@ private static void appendPercentEscaped(StringBuilder output, int cp) {
         byte[] bytes = Character.toString(cp).getBytes(StandardCharsets.UTF_8);
         for (int i = 0; i < bytes.length; ++i) {
             output.append('%');
-            output.append(Utility.hex(bytes[i], 2));
+            output.append(Utility.hex(bytes[i] & 0xFF, 2));
         }
     }
 
@@ -934,7 +936,7 @@ private static String escape(String source, String cpToEscape) {
     }
 
     /** percent-escape all the toEscape characters. */
-    private static String escape(String source, UnicodeSet toEscape) {
+    public static String escape(String source, UnicodeSet toEscape) {
         StringBuilder result = new StringBuilder();
         int lastEnd = 0;
         while (true) {
@@ -1139,4 +1141,16 @@ private static String toUnicode2(String x) {
             throw new ICUException(e);
         }
     }
+
+    public static Map<String, List<List<String>>> getInternals(String source) {
+        UrlInternals ui = UrlInternals.from(source.toString());
+        Map<String, List<List<String>>> result = new LinkedHashMap<>();
+        for (Part p : Part.values()) {
+            List<List<String>> list = ui.get(p);
+            if (list != null) {
+                result.put(p.toString(), list);
+            }
+        }
+        return result;
+    }
 }
diff --git a/icu4j/main/core/src/main/java/com/ibm/icu/impl/locale/XCldrStub.java b/icu4j/main/core/src/main/java/com/ibm/icu/impl/locale/XCldrStub.java
index c57c6228ef7d..437f9300aaea 100644
--- a/icu4j/main/core/src/main/java/com/ibm/icu/impl/locale/XCldrStub.java
+++ b/icu4j/main/core/src/main/java/com/ibm/icu/impl/locale/XCldrStub.java
@@ -136,6 +136,15 @@ public boolean equals(Object obj) {
         public int hashCode() {
             return map.hashCode();
         }
+
+        @Override
+        public String toString() {
+            return "["
+                    + asMap().entrySet().stream()
+                            .map(Entry::toString)
+                            .collect(Collectors.joining(", "))
+                    + "]";
+        }
     }
 
     public static class Multimaps {
@@ -183,6 +192,7 @@ public Entry<K, V> next() {
                 Entry<K, Set<V>> e = it1.next();
                 entry.key = e.getKey();
                 it2 = e.getValue().iterator();
+                entry.value = it2.next();
             }
             return entry;
         }
@@ -199,8 +209,8 @@ public void remove() {
     }
 
     private static class ReusableEntry<K, V> implements Entry<K, V> {
-        K key;
-        V value;
+        private K key;
+        private V value;
 
         @Override
         public K getKey() {
diff --git a/icu4j/main/core/src/main/java/com/ibm/icu/util/LinkUtilities.java b/icu4j/main/core/src/main/java/com/ibm/icu/util/LinkUtilities.java
index b944ab2f06c5..dd51149a8f01 100644
--- a/icu4j/main/core/src/main/java/com/ibm/icu/util/LinkUtilities.java
+++ b/icu4j/main/core/src/main/java/com/ibm/icu/util/LinkUtilities.java
@@ -73,8 +73,8 @@ public static String escapePathQueryFragment(String source, Extent extent) {
         }
     }
 
-    // NOTE for reviewers: These are only temporary, until ICU supplies \p{Link_Term=hard} and
-    // \p{Idn_Status≠disallowed}
+    // NOTE for reviewers: These are only temporary, until ICU supplies \p{Link_Term=X} and
+    // \p{Idn_Status=X}
 
     /**
      * Returns a frozen set of Unicode characters that are guaranteed to never be part of a URL or
@@ -82,7 +82,10 @@ public static String escapePathQueryFragment(String source, Extent extent) {
      * email addresses can never span these characters. For example, a span of characters between
      * safe characters that doesn't have a sequence of domain-character + . + domain-character can
      * be skipped in processing.
+     *
+     * @internal
      */
+    @Deprecated
     public static UnicodeSet getSafeCharacters() {
         return null;
     }
@@ -92,7 +95,10 @@ public static UnicodeSet getSafeCharacters() {
      * (pre-mapping) This allows implementations to make various optimizations because URLs and
      * email addresses must contain a sequence of domain-character + . + domain-character. It is the
      * same as the set of IDNA Mapping Table character with values ≠ disallowed
+     *
+     * @internal
      */
+    @Deprecated
     public static UnicodeSet getDomainCharacters() {
         return null;
     }
diff --git a/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java b/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
index cdde9458efdb..881d7981510e 100644
--- a/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
+++ b/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
@@ -2,7 +2,7 @@
 
 import com.ibm.icu.dev.test.CoreTestFmwk;
 import com.ibm.icu.dev.test.TestUtil;
-import com.ibm.icu.impl.locale.XCldrStub.Splitter;
+import com.ibm.icu.impl.links.LinkHandlingUtilities;
 import com.ibm.icu.lang.UCharacter;
 import com.ibm.icu.text.IDNA;
 import com.ibm.icu.text.UnicodeSet;
@@ -11,6 +11,7 @@
 import com.ibm.icu.util.LinkUtilities.Extent;
 import java.io.BufferedReader;
 import java.io.IOException;
+import java.io.UncheckedIOException;
 import java.util.ArrayList;
 import java.util.List;
 import java.util.Set;
@@ -46,6 +47,24 @@
 @RunWith(JUnit4.class)
 public class LinkTest extends CoreTestFmwk {
 
+    private static final boolean SHOW_LINK_DETECTION = false;
+
+    /**
+     * There are 2 format test cases that need to be fixed in UTS58 In each of these, the = in a 𝑽
+     * doesn't need to be quoted, because it doesn't initiate a different component.
+     */
+    static final Set<String> SKIP_FORMAT_TEST =
+            Set.of(
+                    "# {𝑺=https:// 𝑯=example.com 𝑸=α=β 𝑽=γ=δ}",
+                    "# {𝑺=https:// 𝑯=example.com 𝑷=α 𝑷=b/?#c 𝑸=αβ 𝑽=γ&ζ=#Ξ 𝑸=k 𝑽=v 𝑭=frag}");
+
+    /**
+     * There is 1 detection test case that need to be removed in UTS58 UTS58 checks for valid domain
+     * names, and for neutrality it should avoid listing such cases
+     */
+    static final Set<String> SKIP_DETECTION_TEST =
+            Set.of("Lorem ipsum موريتانيا.xn-----ctdbabcfhu9c2b9l1acccr4c dolor sit amet");
+
     /** Constructor */
     public LinkTest() {}
 
@@ -72,10 +91,37 @@ private void checkScan(String toCheck, String expectedAfter) {
 
     @Test
     public void testSimpleEscape() {
-        String source = "/%CE%B1%CE%B2%CE%B3/%CE%B4.%CE%B5%CE%B6%2E";
-        String expected = "/αβγ/δ.εζ%2E";
-        String actual = LinkUtilities.escapePathQueryFragment(source, Extent.MINIMAL);
-        assertEquals(source, expected, actual);
+        List<List<String>> tests = // source, minimal escaping, maximal escaping
+                List.of(
+                        List.of(
+                                "https://example.com/αβγ/δεζ?θ=ικλ&μ=γξο#πρς",
+                                "https://example.com/αβγ/δεζ?θ=ικλ&μ=γξο#πρς",
+                                "https://example.com/%CE%B1%CE%B2%CE%B3/%CE%B4%CE%B5%CE%B6?%CE%B8=%CE%B9%CE%BA%CE%BB&%CE%BC=%CE%B3%CE%BE%CE%BF#%CF%80%CF%81%CF%82"),
+                        List.of("https://example.com/α/β%2Fγ", "https://example.com/α/β%2Fγ", ""),
+                        List.of(
+                                "https://example.com/α%23β?γ=δ%23ε",
+                                "https://example.com/α%23β?γ=δ%23ε", ""),
+                        List.of(
+                                "ex.com/%CE%B1%CE%B2%CE%B3/%CE%B4.%CE%B5%CE%B6%2E",
+                                "ex.com/αβγ/δ.εζ%2E", ""),
+                        List.of(
+                                "https://example.com/α%3Fμπ",
+                                "", "https://example.com/%CE%B1%3F%CE%BC%CF%80"));
+        for (List<String> test : tests) {
+            String source = test.get(0);
+            String expectedMin = test.get(1);
+            String expectedMax = test.get(2);
+            if (!expectedMin.isEmpty()) {
+                String actual = LinkUtilities.escapePathQueryFragment(source, Extent.MINIMAL);
+                assertEquals(
+                        "MIN " + LinkHandlingUtilities.getInternals(source), expectedMin, actual);
+            }
+            if (!expectedMax.isEmpty()) {
+                String actual = LinkUtilities.escapePathQueryFragment(source, Extent.MAXIMAL);
+                assertEquals(
+                        "MAX " + LinkHandlingUtilities.getInternals(source), expectedMax, actual);
+            }
+        }
     }
 
     @Test
@@ -92,14 +138,15 @@ private void checkEmail(String source, int start, String localPartStart) {
         int actual = LinkUtilities.scanBackEmailLocalPart(source, start, end);
         assertEquals(source, expected, actual);
     }
+
     @Test
     public void testHackDomainName() {
-    	String source = // "普遍适用测试。我爱你";
-    	"موريتانيا.xn-----ctdbabcfhu9c2b9l1acccr4c";
+        String source = // "普遍适用测试。我爱你";
+                "موريتانيا.xn-----ctdbabcfhu9c2b9l1acccr4c";
         IntRange domain = new IntRange();
-		boolean actual = hackFindDomainName(source, 0, source.length(), false, domain);
-		assertEquals(source, true, actual);
-		assertTrue(source, domain.start == 0 && domain.end == source.length());
+        boolean actual = hackFindDomainName(source, 0, source.length(), false, domain);
+        assertEquals(source, true, actual);
+        assertTrue(source, domain.start == 0 && domain.end == source.length());
     }
 
     // Would like to use the following for JUnit, but can't
@@ -109,15 +156,13 @@ public void testHackDomainName() {
     Pattern SCHEME = Pattern.compile("https?://$");
     Pattern PORT = Pattern.compile(":\\d+");
 
-    final Set<String> skip = Set.of("See http://-foo.example-.com. on…");
-
     @Test
     public void testAgainstLinkDetectionTest() {
         List<String> failures = new ArrayList<>();
         try (BufferedReader idnaTestFile =
                 TestUtil.getUtf8DataReader("unicode/links/LinkDetectionTest.txt"); ) {
             for (final String line : (Iterable<String>) idnaTestFile.lines()::iterator) {
-                if (line.startsWith("#") || line.isBlank() || skip.contains(line)) {
+                if (line.startsWith("#") || line.isBlank() || SKIP_DETECTION_TEST.contains(line)) {
                     continue;
                 }
                 StringBuilder actual = new StringBuilder();
@@ -155,8 +200,8 @@ public void testAgainstLinkDetectionTest() {
                             // If so:
                             // - we scan forwards (after the @) to see if there is a domain name,
                             // AND
-                            // - we scan backwards (before the @) to see if there is local_part, 
-                        	// AND
+                            // - we scan backwards (before the @) to see if there is local_part,
+                            // AND
                             // - we scan backwards before that to ensure that there is no https?://
                             // if both are true, we have an email address.
                             // if only the first is true, we skip over the 2nd domain name;
@@ -241,10 +286,11 @@ public void testAgainstLinkDetectionTest() {
                     actual.append(rawLine2.substring(start));
 
                     assertEquals("", line, actual.toString());
-                    System.out.println("OK\t" + line);
+                    if (SHOW_LINK_DETECTION) System.out.println("OK\t" + line);
                 } catch (AssertionError e) {
                     failures.add("FAIL " + e.getMessage());
-                    System.out.println("FAIL\t" + line + "\t  " + actual.toString());
+                    if (SHOW_LINK_DETECTION)
+                        System.out.println("FAIL\t" + line + "\t  " + actual.toString());
                 }
             }
             if (!failures.isEmpty()) {
@@ -255,7 +301,7 @@ public void testAgainstLinkDetectionTest() {
                                 + String.join("\n", failures));
             }
         } catch (IOException e) {
-            e.printStackTrace();
+            throw new UncheckedIOException(e);
         }
     }
 
@@ -267,72 +313,150 @@ public void testAgainstLinkFormattingTest() {
         //	https://example.com/%CE%B1%CE%B2%CE%B3/%CE%B4%CE%B5%CE%B6?%CE%B8=%CE%B9%CE%BA%CE%BB&%CE%BC=%CE%B3%CE%BE%CE%BF#%CF%80%CF%81%CF%82
         //        	https://example.com/αβγ/δεζ?θ=ικλ&μ=γξο#πρς
 
-        String expectedPrefix = "https://example.com"; // for now, all the same
-
+        List<String> failures = new ArrayList<>();
+        // the test case is broken onto 3 lines for ease of reading
+        // so we assemble each one before testing.
+        List<String> testCase = new ArrayList<>();
         try (BufferedReader idnaTestFile =
                 TestUtil.getUtf8DataReader("unicode/links/LinkFormattingTest.txt"); ) {
+            String parts = null;
             String pqf = null;
-            boolean firstLine = true;
+            StringBuilder schemeHost = new StringBuilder();
+            String minimalPqf = null;
+            String maximalPqf = null;
             for (String line : (Iterable<String>) idnaTestFile.lines()::iterator) {
-                if (line.startsWith("# {")) {
-                    pqf = assemblePqf(line);
-                    firstLine = true;
-                } else if (line.isBlank() || line.startsWith("#")) {
-                    continue;
-                } else if (!line.startsWith(expectedPrefix)) {
-                    throw new IllegalArgumentException("Expand the test to cover other S/H");
-                } else if (firstLine) {
-                    String actual =
-                            expectedPrefix
-                                    + LinkUtilities.escapePathQueryFragment(pqf, Extent.MAXIMAL);
-                    assertEquals("maximum " + pqf, line, actual);
-                    firstLine = false;
-                } else {
-                    String actual =
-                            expectedPrefix
-                                    + LinkUtilities.escapePathQueryFragment(pqf, Extent.MINIMAL);
-                    assertEquals("maximum " + pqf, line, actual);
-                    firstLine = false;
+                try {
+                    if (line.startsWith("# {")) {
+                        testCase.clear();
+                        testCase.add(line);
+                    } else if (line.isBlank() || line.startsWith("#")) {
+                        continue;
+                    } else if (testCase.size() == 1) {
+                        testCase.add(line);
+                    } else { // we have assembled the first two lines, so we can process
+                        String partsLine = testCase.get(0);
+                        if (SKIP_FORMAT_TEST.contains(partsLine)) {
+                            continue;
+                        }
+                        parts =
+                                partsLine.substring(
+                                        3, partsLine.length() - 1); // just used in reporting
+                        pqf = assemblePqf(partsLine, schemeHost);
+                        String maximalLine = testCase.get(1);
+                        String minimalLine = line;
+                        System.out.println(partsLine);
+                        minimalPqf = LinkUtilities.escapePathQueryFragment(pqf, Extent.MINIMAL);
+                        assertEquals(
+                                Extent.MINIMAL + " " + parts, minimalLine, schemeHost + minimalPqf);
+                        maximalPqf = LinkUtilities.escapePathQueryFragment(pqf, Extent.MAXIMAL);
+                        assertEquals(
+                                Extent.MAXIMAL + " " + parts, maximalLine, schemeHost + maximalPqf);
+                    }
+                    System.out.println("OK " + line);
+                } catch (AssertionError e) {
+                    failures.add("FAIL " + e.getMessage());
+                    System.out.println("FAIL\t" + line);
                 }
             }
         } catch (IOException e) {
-            e.printStackTrace();
+            throw new UncheckedIOException(e);
+        }
+
+        if (!failures.isEmpty()) {
+            fail(
+                    String.join(
+                            " ",
+                            "Test finished with",
+                            failures.size() + "",
+                            "failures:\n" + String.join("\n", failures)));
         }
     }
 
-    private String assemblePqf(String line) {
+    @Test
+    public void testParts() {
+        String source =
+                "# {𝑺=https:// 𝑯=example.com 𝑷=αβγ 𝑷=δεζ 𝑸=θ 𝑽=ικλ 𝑸=μ 𝑽=γξο 𝑭=πρς}";
+        List<List<String>> expected = new ArrayList<>();
+        expected.add(List.of("𝑺", "https://"));
+        expected.add(List.of("𝑯", "example.com"));
+        expected.add(List.of("𝑷", "αβγ"));
+        expected.add(List.of("𝑷", "δεζ"));
+        expected.add(List.of("𝑸", "θ"));
+        expected.add(List.of("𝑽", "ικλ"));
+        expected.add(List.of("𝑸", "μ"));
+        expected.add(List.of("𝑽", "γξο"));
+        expected.add(List.of("𝑭", "πρς"));
+        List<List<String>> actual = getParts(source);
+        assertEquals(source, expected, actual);
+    }
+
+    Pattern UrlTypes = Pattern.compile("([𝑺𝑯𝑷𝑸𝑽𝑭])=([^𝑺𝑯𝑷𝑸𝑽𝑭]+)");
+
+    private List<List<String>> getParts(String line) {
+        // # {𝑺=https:// 𝑯=example.com 𝑷=αβγ 𝑷=δεζ 𝑸=θ 𝑽=ικλ 𝑸=μ 𝑽=γξο 𝑭=πρς}
         line = line.substring(3, line.length() - 1); // subtract the # { and the }
+        Matcher urlType = UrlTypes.matcher(line);
+        List<List<String>> result = new ArrayList<>();
+        while (urlType.find()) {
+            String key = urlType.group(1);
+            String value = urlType.group(2);
+            if (value.endsWith(" ")) {
+                value = value.substring(0, value.length() - 1);
+            }
+            result.add(List.of(key, value));
+        }
+        return List.copyOf(result);
+    }
+
+    /**
+     * When composing a URL, the following have to be escaped within each part. They are the
+     * characters that could initiate a different component
+     */
+    final UnicodeSet PATH_ESCAPE = new UnicodeSet("[/?#]").freeze();
+
+    final UnicodeSet QUERY_ESCAPE = new UnicodeSet("[\\&=#]").freeze();
+    final UnicodeSet VALUE_ESCAPE = new UnicodeSet("[\\&#]").freeze();
+
+    private String assemblePqf(String line, StringBuilder schemeHost) {
         StringBuilder sb = new StringBuilder();
+        schemeHost.setLength(0);
+        List<List<String>> parts = getParts(line);
         char query = '?';
-        for (String item : Splitter.on(' ').splitToList(line)) {
-            int eq = item.indexOf('=');
-            String type = item.substring(0, eq);
-            String rest = item.substring(eq + 1);
-            //	        	# {𝑺=https:// 𝑯=example.com 𝑷=αβγ 𝑷=δεζ 𝑸=θ 𝑽=ικλ 𝑸=μ 𝑽=γξο 𝑭=πρς}
+        for (List<String> entry : parts) {
+            String type = entry.get(0);
+            String value = entry.get(1);
+            // # {𝑺=https:// 𝑯=example.com 𝑷=αβγ 𝑷=δεζ 𝑸=θ 𝑽=ικλ 𝑸=μ 𝑽=γξο 𝑭=πρς}
             switch (type) {
                 case "𝑺":
                 case "𝑯":
+                    if (value != null) {
+                        schemeHost.append(value);
+                    }
                     break;
                 case "𝑷":
-                    sb.append('/').append(rest);
+                    sb.append('/').append(escape(value, PATH_ESCAPE));
                     break;
                 case "𝑸":
-                    sb.append(query).append(rest);
+                    sb.append(query).append(escape(value, QUERY_ESCAPE));
                     query = '&';
                     break;
                 case "𝑽":
-                    sb.append('=').append(rest);
+                    sb.append('=').append(escape(value, VALUE_ESCAPE));
                     break;
                 case "𝑭":
-                    sb.append('#').append(rest);
+                    sb.append('#').append(value);
                     break;
                 default:
-                    throw new IllegalArgumentException("Unexpected line: " + item);
+                    throw new IllegalArgumentException("Unexpected line: " + line);
             }
         }
         return sb.toString();
     }
 
+    private Object escape(String value, UnicodeSet unicodeSet) {
+        return LinkHandlingUtilities.escape(value, unicodeSet);
+    }
+
     class IntRange {
         int start;
         int end;
@@ -345,13 +469,13 @@ public String toString() {
 
     Pattern DOT = Pattern.compile("[.。]");
     Pattern DOT2 = Pattern.compile("[.。][.。]");
+
     /**
      * The UTS58 tests take whole domain names, because that is the easiest for clients to test. We
      * have to hack it because ICU IDN support doesn't provide enough API. Returns true if found,
      * with the range in result.
      *
-     * @param mustBeAtStart 
-     * TODO check with isValidDomain
+     * @param mustBeAtStart TODO check with isValidDomain
      */
     boolean hackFindDomainName(
             CharSequence source, int start, int limit, boolean mustBeAtStart, IntRange result) {
@@ -373,7 +497,7 @@ boolean hackFindDomainName(
                 end = limit;
             }
             if (end == endNot) {
-            	return false;
+                return false;
             }
             // Check that the possible domain name (endNot..end) is of the right form
             // otherwise continue after end.
@@ -382,7 +506,7 @@ boolean hackFindDomainName(
             boolean found = dot.find();
             if (!found || dot.start() == endNot || dot.end() == end) {
                 start = end;
-                continue;	
+                continue;
             }
             // Must not find [.。][.。]
             dot2.region(endNot, end);
@@ -390,6 +514,11 @@ boolean hackFindDomainName(
                 start = end;
                 continue;
             }
+            // and a final test for valid
+            if (!isValidDomain(source.subSequence(endNot, end).toString())) {
+                start = end;
+                continue;
+            }
             result.start = endNot;
             result.end = end;
             return true;
@@ -442,14 +571,14 @@ static boolean isValidDomain(String domainName) {
         }
 
         // Initialize the UTS#46 standard IDNA instance
-        IDNA idna = IDNA.getUTS46Instance(
-            IDNA.NONTRANSITIONAL_TO_ASCII | 
-            IDNA.NONTRANSITIONAL_TO_UNICODE | 
-            IDNA.CHECK_BIDI | 
-            IDNA.CHECK_CONTEXTJ | 
-            IDNA.CHECK_CONTEXTO | 
-            IDNA.USE_STD3_RULES
-        );
+        IDNA idna =
+                IDNA.getUTS46Instance(
+                        IDNA.NONTRANSITIONAL_TO_ASCII
+                                | IDNA.NONTRANSITIONAL_TO_UNICODE
+                                | IDNA.CHECK_BIDI
+                                | IDNA.CHECK_CONTEXTJ
+                                | IDNA.CHECK_CONTEXTO
+                                | IDNA.USE_STD3_RULES);
 
         StringBuilder dest = new StringBuilder();
         IDNA.Info info = new IDNA.Info();

From 6bd77efa3d84052d26a8d6e5ad475a74abfd7553 Mon Sep 17 00:00:00 2001
From: Mark Davis <mark@unicode.org>
Date: Fri, 3 Jul 2026 13:07:52 +0300
Subject: [PATCH 5/8] ICU-23431 Copyright

---
 .../java/com/ibm/icu/impl/links/LinkHandlingUtilities.java     | 3 ++-
 .../core/src/main/java/com/ibm/icu/util/LinkUtilities.java     | 2 ++
 .../src/test/java/com/ibm/icu/dev/test/links/LinkTest.java     | 2 ++
 3 files changed, 6 insertions(+), 1 deletion(-)

diff --git a/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java b/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java
index 7e1188ecb616..a7fe0fd4378a 100644
--- a/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java
+++ b/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java
@@ -1,9 +1,10 @@
+// © 2026 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
 package com.ibm.icu.impl.links;
 
 import com.ibm.icu.impl.IDNA2003;
 import com.ibm.icu.impl.UnicodeMap;
 import com.ibm.icu.impl.Utility;
-import com.ibm.icu.impl.links.LinkHandlingUtilities.UrlInternals;
 import com.ibm.icu.impl.locale.XCldrStub;
 import com.ibm.icu.impl.locale.XCldrStub.Splitter;
 import com.ibm.icu.lang.UCharacter;
diff --git a/icu4j/main/core/src/main/java/com/ibm/icu/util/LinkUtilities.java b/icu4j/main/core/src/main/java/com/ibm/icu/util/LinkUtilities.java
index dd51149a8f01..af98b4c2fca5 100644
--- a/icu4j/main/core/src/main/java/com/ibm/icu/util/LinkUtilities.java
+++ b/icu4j/main/core/src/main/java/com/ibm/icu/util/LinkUtilities.java
@@ -1,3 +1,5 @@
+// © 2026 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
 package com.ibm.icu.util;
 
 import com.ibm.icu.impl.links.LinkHandlingUtilities;
diff --git a/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java b/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
index 881d7981510e..7a22b6e1612e 100644
--- a/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
+++ b/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
@@ -1,3 +1,5 @@
+// © 2026 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
 package com.ibm.icu.dev.test.links;
 
 import com.ibm.icu.dev.test.CoreTestFmwk;

From 7e7032b60e7c85cf49a6c8272d174bc5275c373d Mon Sep 17 00:00:00 2001
From: Mark Davis <mark@unicode.org>
Date: Sat, 4 Jul 2026 16:30:02 +0300
Subject: [PATCH 6/8] ICU-23431 Cleanp unused code from port

---
 .../icu/impl/links/LinkHandlingUtilities.java | 306 ------------------
 .../com/ibm/icu/dev/test/links/LinkTest.java  |  30 +-
 2 files changed, 21 insertions(+), 315 deletions(-)

diff --git a/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java b/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java
index a7fe0fd4378a..72ceec174386 100644
--- a/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java
+++ b/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java
@@ -2,23 +2,16 @@
 // License & terms of use: http://www.unicode.org/copyright.html
 package com.ibm.icu.impl.links;
 
-import com.ibm.icu.impl.IDNA2003;
 import com.ibm.icu.impl.UnicodeMap;
 import com.ibm.icu.impl.Utility;
 import com.ibm.icu.impl.locale.XCldrStub;
 import com.ibm.icu.impl.locale.XCldrStub.Splitter;
 import com.ibm.icu.lang.UCharacter;
-import com.ibm.icu.lang.UProperty;
-import com.ibm.icu.lang.UProperty.NameChoice;
-import com.ibm.icu.text.StringPrepParseException;
 import com.ibm.icu.text.UnicodeSet;
-import com.ibm.icu.text.UnicodeSet.EntryRange;
 import com.ibm.icu.text.UnicodeSet.SpanCondition;
-import com.ibm.icu.util.ICUException;
 import com.ibm.icu.util.Output;
 import java.nio.charset.StandardCharsets;
 import java.util.EnumMap;
-import java.util.EnumSet;
 import java.util.LinkedHashMap;
 import java.util.List;
 import java.util.Map;
@@ -100,15 +93,10 @@ private enum LinkTermination {
 
         public final UnicodeSet base;
 
-        public UnicodeSet getBase() {
-            return base;
-        }
-
         private LinkTermination(String uset) {
             if (uset == null) { // only called with Include, the "none of the above" option
                 this.base = SOFAR.complement().freeze();
             } else {
-                java.text.ParsePosition parsePosition = new java.text.ParsePosition(0);
                 this.base = new UnicodeSet(uset).freeze();
                 SOFAR.addAll(this.base);
             }
@@ -139,9 +127,6 @@ private LinkTermination(String uset) {
             }
             PROPERTY_MAP.freeze();
         }
-
-        private static final Set<LinkTermination> NON_MISSING =
-                XCldrStub.setsDifference(EnumSet.allOf(LinkTermination.class), Set.of(Hard));
     }
 
     // https://www.rfc-editor.org/rfc/rfc5322.html#section-3.2.3 has the full list for ASCII part
@@ -265,18 +250,6 @@ static Part fromInitiator(int cp) {
             return null;
         }
 
-        /**
-         * Unescape a part. But don't unescape interior characters or terminators because they are
-         * content! For example "a/b%2Fc" as a path should not be turned into a/b/c, because that
-         * b/c is a path-part.
-         *
-         * @param substring
-         * @return
-         */
-        private String unescape(String substring) {
-            return LinkHandlingUtilities.unescape(substring, extraQuoted);
-        }
-
         private String fullEscape(List<List<String>> source) {
             if (source.isEmpty()) {
                 return "";
@@ -310,16 +283,6 @@ private String fullEscape(List<List<String>> source) {
             }
             return result.toString();
         }
-
-        /** Append an unescaped part, escaping as necessary */
-        private void appendPart(StringBuilder toAppendTo, String partValue) {
-            if (!partValue.isBlank()) {
-                if (initiator != 0) {
-                    toAppendTo.appendCodePoint(initiator);
-                }
-                toAppendTo.append(partValue);
-            }
-        }
     }
 
     public static class UrlInternals {
@@ -662,116 +625,6 @@ private static List<String> splitList(Part part, String separator, String partSt
                 .collect(Collectors.toList());
     }
 
-    //    private static final UnicodeSet idnMapped =
-    //            IUP.getSet("Idn_Status=" + Idn_Status_Values.mapped);
-    //
-    //    private static final UnicodeSet idnValid = IUP.getSet("Idn_Status=" +
-    // Idn_Status_Values.valid);
-    //    static final Pattern protocol = Pattern.compile("(https?://|mailto:)");
-    //    public static final UnicodeSet validHost =
-    //            new UnicodeSet(idnValid)
-    //                    .addAll(idnMapped)
-    //                    .removeAll(new UnicodeSet("[:Block=Basic_Latin:]"))
-    //                    .addAll(new UnicodeSet("[-a-zA-Z0-9.．。]")) // the three dots are
-    // different:
-    //                    .freeze();
-    //    public static final UnicodeSet validHostNoDot =
-    //            new UnicodeSet(validHost).removeAll(".．。").freeze();
-
-    /** Status of link found. Note that if the link is imputed, https or mailto will be returned. */
-    enum LinkStatus {
-        bogus,
-        http,
-        https,
-        mailto,
-        tld
-    }
-
-    /** Immutable class for returning the start/end of link that is found, plus some information */
-    private static final class LinkFound {
-        public final int start;
-        public final int limit;
-        public final LinkStatus linkStatus;
-
-        public LinkFound(int start, int limit, LinkStatus linkStatus) {
-            this.start = start;
-            this.limit = limit;
-            this.linkStatus = linkStatus;
-        }
-
-        public String substring(String source) {
-            return source.substring(start, limit);
-        }
-    }
-
-    // OLDER Code, leaving here for now, for comparison
-    //    /**
-    //     * Parses a restricted set of URLs, for testing the PathQueryFragment portion. That is, it
-    // is of
-    //     * the form &lt;host>&lt;domain_name><pathqueryfragment>? <br>
-    //     * The host is currently just http, https, and mailto. <br>
-    //     * The domain_name is just approximated for testing.
-    //     *
-    //     * @return null if we run out of string, otherwise a LinkFound. If what is found is
-    // malformed in
-    //     *     some way, indicate with LinkFound.bogus.
-    //     */
-    //    private static LinkFound parseLink(String source, int startCodePointOffset) {
-    //        Matcher findStartMatcher = protocol.matcher(source);
-    //
-    //        findStartMatcher.region(startCodePointOffset, source.length());
-    //        if (!findStartMatcher.find()) {
-    //            return null; // we are at the end
-    //        }
-    //        // if we found something, and there was a character before it,
-    //        // and that character was not soft, then exit
-    //        int start = findStartMatcher.start();
-    //        int protocolEnd = findStartMatcher.end();
-    //        if (start != 0) {
-    //            LinkTermination lt =
-    //                    LinkTermination.PROPERTY_MAP.get(UCharacter.codePointBefore(source,
-    // start));
-    //            if (lt == LinkTermination.Include) {
-    //                return new LinkFound(start, protocolEnd, LinkStatus.bogus);
-    //            }
-    //        }
-    //
-    //        String protocolValue = findStartMatcher.group(1);
-    //        if (protocolValue.equals("mailto")) {
-    //            return parseRestOfMailto(source, start, protocolEnd);
-    //        }
-    //
-    //        // dumb search for end of host, doesn't handle .. or edge cases, but this does not
-    // have to
-    //        // be production-quality
-    //        int hostLimit = findEndOfDomain(source, protocolEnd);
-    //        if (protocolEnd == hostLimit) {
-    //            return new LinkFound(start, protocolEnd, LinkStatus.bogus);
-    //        }
-    //        int limit = parsePathQueryFragment(source, hostLimit);
-    //        return new LinkFound(
-    //                start, limit, LinkStatus.valueOf(source.substring(start, protocolEnd)));
-    //    }
-    //
-    //    private static LinkFound parseRestOfMailto(String source, int start, int hostStart) {
-    //        // basic implementation at first
-    //        int atPosition = source.indexOf('@', hostStart);
-    //        if (atPosition == -1) {
-    //            return new LinkFound(start, hostStart, LinkStatus.bogus);
-    //        }
-    //        // TBD we could be in the middle of a quoted string, check for that later
-    //
-    //        // see if what is in front of the @ looks ok
-    //
-    //        int limit = parsePathQueryFragment(source, atPosition + 1);
-    //        return new LinkFound(start, limit, LinkStatus.mailto);
-    //    }
-    //
-    //    // Simple implementation for testing, since the spec doesn't define the content
-    //    private static int findEndOfDomain(String source, int protocolLimit) {
-    //        return validHost.span(source, protocolLimit, SpanCondition.CONTAINED);
-    //    }
-
     /**
      * Set lastSafe to 0 — this marks the last code point that is definitely included in the
      * linkification.<br>
@@ -931,11 +784,6 @@ private static void appendPercentEscaped(StringBuilder output, String source) {
         }
     }
 
-    /** percent-escape single char or string. */
-    private static String escape(String source, String cpToEscape) {
-        return escape(source, new UnicodeSet().add(cpToEscape));
-    }
-
     /** percent-escape all the toEscape characters. */
     public static String escape(String source, UnicodeSet toEscape) {
         StringBuilder result = new StringBuilder();
@@ -989,160 +837,6 @@ private static String percentUnescape(String escapedSource) {
         return new String(temp, StandardCharsets.UTF_8);
     }
 
-    /**
-     * The wikipedia languages are not all BCP47. Convert the ones that are not. See:<br>
-     * https://meta.wikimedia.org/wiki/Special_language_codes <br>
-     * https://meta.wikimedia.org/wiki/List_of_Wikipedias#Nonstandard_language_codes
-     *
-     * @param languageCode
-     * @return
-     */
-    private static String fixWiki(String languageCode) {
-        switch (languageCode) {
-            case "als":
-                return "gsw";
-            case "roa-rup":
-                return "rup";
-            case "bat-smg":
-                return "sgs";
-            case "simple":
-                return "en";
-            case "fiu-vro":
-                return "vro";
-            case "zh-classical":
-                return "lzh";
-            case "zh-min-nan":
-                return "nan";
-            case "zh-yue":
-                return "yue";
-            case "cbk-zam":
-                return "cbk";
-            case "map-bms":
-                return "map";
-            case "nrm":
-                return "nrf";
-            case "roa-tara":
-                return "nap";
-            default:
-                return languageCode;
-        }
-    }
-
-    /** Hard-coded list of wikilanguages, for testing */
-    //    private static Set<String> WIKI_LANGUAGES =
-    //            of(
-    //                    SPLIT_COMMA.splitToList(
-    //
-    // "en,ceb,de,fr,sv,nl,ru,es,it,pl,arz,zh,ja,uk,vi,war,ar,pt,fa,ca,id,sr,ko,no,tr,ce,fi,cs,hu,tt,ro,sh,eu,zh-min-nan,ms,he,eo,hy,da,bg,uz,cy,simple,sk,et,be,azb,el,kk,min,hr,lt,gl,ur,az,sl,lld,ka,nn,ta,th,hi,bn,mk,zh-yue,la,ast,lv,af,tg,my,te,sq,mr,mg,bs,oc,be-tarask,ku,br,sw,ml,nds,ky,lmo,jv,pnb,ckb,new,ht,vec,pms,lb,ba,su,ga,is,szl,cv,pa,fy,io,ha,tl,an,mzn,wuu,diq,vo,ig,yo,sco,kn,ne,als,gu,ia,avk,crh,bar,ban,scn,bpy,mn,qu,nv,si,xmf,frr,ps,os,or,tum,sd,bcl,bat-smg,sah,cdo,gd,bug,glk,yi,ilo,am,li,nap,gor,as,fo,mai,hsb,map-bms,shn,zh-classical,eml,ace,ie,wa,sa,hyw,sat,zu,sn,mhr,lij,hif,km,bjn,mrj,mni,dag,ary,hak,pam,rue,roa-tara,ug,zgh,bh,nso,co,tly,so,vls,nds-nl,mi,se,myv,rw,kaa,sc,bo,kw,vep,mt,tk,mdf,kab,gv,gan,fiu-vro,ff,zea,ab,skr,smn,ks,gn,frp,pcd,udm,kv,csb,ay,nrm,lo,ang,fur,olo,lfn,lez,ln,pap,nah,mwl,tw,stq,rm,ext,lad,gom,dty,av,tyv,koi,dsb,lg,cbk-zam,dv,ksh,za,bxr,blk,gag,pfl,bew,szy,haw,tay,pag,pi,awa,tcy,krc,inh,gpe,xh,kge,fon,atj,to,pdc,mnw,arc,shi,om,tn,dga,ki,nia,jam,kbp,wo,xal,nov,kbd,anp,nqo,bi,kg,roa-rup,tpi,tet,guw,jbo,mad,fj,lbe,kcg,pcm,cu,ty,trv,dtp,sm,ami,st,iba,srn,btm,alt,ltg,gcr,ny,kus,mos,ss,chr,ee,ts,got,bbc,gur,bm,pih,ve,rmy,fat,chy,rn,igl,ik,guc,ch,ady,pnt,iu,ann,rsk,pwn,dz,ti,sg,din,tdd,kl,bdr,nr,cr"));
-
-    /** Show termination values, for generating property files */
-    private void showLinkTermination() {
-        for (LinkTermination lt : LinkTermination.values()) {
-            UnicodeSet value = LinkTermination.PROPERTY_MAP.getSet(lt);
-            String name = lt.toString();
-            System.out.println("\n#\tLink_Termination=" + name);
-            if (lt == LinkTermination.Include) {
-                System.out.println("#   " + "(All code points without other values)");
-                continue;
-            } else {
-                System.out.println("#   draft = " + lt.base);
-            }
-            if (lt == LinkTermination.Hard) {
-                value.removeAll(new UnicodeSet("[\\p{Cn}\\p{Cs}]"));
-                System.out.println("#   (not listing Unassigned or Surrogates)");
-            }
-            System.out.println();
-            for (EntryRange range : value.ranges()) {
-                final String rangeString =
-                        Utility.hex(range.codepoint)
-                                + (range.codepoint == range.codepointEnd
-                                        ? ""
-                                        : ".." + Utility.hex(range.codepointEnd));
-                System.out.println(
-                        rangeString
-                                + ";"
-                                + " ".repeat(15 - rangeString.length())
-                                + lt
-                                + "\t# "
-                                + "("
-                                + getGeneralCategory(
-                                        UProperty.GENERAL_CATEGORY,
-                                        range.codepoint,
-                                        NameChoice.SHORT)
-                                + ") "
-                                + quote(UCharacter.getExtendedName(range.codepoint))
-                                + (range.codepoint == range.codepointEnd
-                                        ? ""
-                                        : ".."
-                                                + "("
-                                                + getGeneralCategory(
-                                                        UProperty.GENERAL_CATEGORY,
-                                                        range.codepointEnd,
-                                                        NameChoice.SHORT)
-                                                + ") "
-                                                + quote(
-                                                        UCharacter.getExtendedName(
-                                                                range.codepointEnd))));
-            }
-            System.out.println();
-        }
-    }
-
-    /** Show paired openers, for generating property files */
-    private void showLinkPairedOpeners() {
-        UnicodeSet value = LinkTermination.PROPERTY_MAP.getSet(LinkTermination.Close);
-
-        System.out.println("\n#\tLink_Paired_Opener");
-        System.out.println(
-                "#   draft = BidiPairedBracket + (“&gt;” GREATER-THAN SIGN 🡆  “&lt;” LESS-THAN SIGN)");
-        System.out.println();
-
-        for (String cpString : value) {
-            int cp = cpString.codePointAt(0);
-            String hex = Utility.hex(cp);
-            final int value2 = getOpening(cp);
-            System.out.println(
-                    hex
-                            + ";"
-                            + " ".repeat(7 - hex.length())
-                            + Utility.hex(value2)
-                            + "\t#"
-                            + " “"
-                            + quote(Character.toString(cp))
-                            + "” "
-                            + UCharacter.getExtendedName(cp)
-                            + " 🡆 "
-                            + " “"
-                            + quote(Character.toString(value2))
-                            + "” "
-                            + UCharacter.getExtendedName(value2));
-        }
-    }
-
-    /**
-     * Regex to scan for possible TLDs. The result needs to be checked that there is no
-     * validHostNoDot before and after
-     */
-    //    private static final Pattern TLD_SCANNER;
-    //
-    //    private static final SortedSet<String> TLDS;
-    //
-    private static final String DOTSET_STRING = "[.。]";
-
-    private static final UnicodeSet DOTSET = new UnicodeSet("[.。]").freeze();
-    private static final Splitter SPLIT_LABELS = Splitter.on(Pattern.compile("[.。]"));
-
-    private static String toUnicode2(String x) {
-        try {
-            if (x.startsWith("XN--")) {
-                x = IDNA2003.convertIDNToUnicode(x, 0).toString();
-            }
-            return x;
-        } catch (StringPrepParseException e) {
-            throw new ICUException(e);
-        }
-    }
-
     public static Map<String, List<List<String>>> getInternals(String source) {
         UrlInternals ui = UrlInternals.from(source.toString());
         Map<String, List<List<String>>> result = new LinkedHashMap<>();
diff --git a/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java b/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
index 7a22b6e1612e..2244e0b11cef 100644
--- a/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
+++ b/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
@@ -49,7 +49,7 @@
 @RunWith(JUnit4.class)
 public class LinkTest extends CoreTestFmwk {
 
-    private static final boolean SHOW_LINK_DETECTION = false;
+    private static final boolean VERBOSE = false;
 
     /**
      * There are 2 format test cases that need to be fixed in UTS58 In each of these, the = in a 𝑽
@@ -288,11 +288,14 @@ public void testAgainstLinkDetectionTest() {
                     actual.append(rawLine2.substring(start));
 
                     assertEquals("", line, actual.toString());
-                    if (SHOW_LINK_DETECTION) System.out.println("OK\t" + line);
+                    if (isVerbose()) {
+                        System.out.println("OK\t" + line);
+                    }
                 } catch (AssertionError e) {
                     failures.add("FAIL " + e.getMessage());
-                    if (SHOW_LINK_DETECTION)
+                    if (isVerbose()) {
                         System.out.println("FAIL\t" + line + "\t  " + actual.toString());
+                    }
                 }
             }
             if (!failures.isEmpty()) {
@@ -346,18 +349,27 @@ public void testAgainstLinkFormattingTest() {
                         pqf = assemblePqf(partsLine, schemeHost);
                         String maximalLine = testCase.get(1);
                         String minimalLine = line;
-                        System.out.println(partsLine);
-                        minimalPqf = LinkUtilities.escapePathQueryFragment(pqf, Extent.MINIMAL);
-                        assertEquals(
-                                Extent.MINIMAL + " " + parts, minimalLine, schemeHost + minimalPqf);
+
+                        // Check explicit test cases
+
                         maximalPqf = LinkUtilities.escapePathQueryFragment(pqf, Extent.MAXIMAL);
                         assertEquals(
                                 Extent.MAXIMAL + " " + parts, maximalLine, schemeHost + maximalPqf);
+
+                        minimalPqf = LinkUtilities.escapePathQueryFragment(pqf, Extent.MINIMAL);
+                        assertEquals(
+                                Extent.MINIMAL + " " + parts, minimalLine, schemeHost + minimalPqf);
+
+                        // TODO Look at roundtripping both of these.
+                    }
+                    if (isVerbose()) {
+                        System.out.println("OK " + line);
                     }
-                    System.out.println("OK " + line);
                 } catch (AssertionError e) {
                     failures.add("FAIL " + e.getMessage());
-                    System.out.println("FAIL\t" + line);
+                    if (isVerbose()) {
+                        System.out.println("FAIL\t" + line);
+                    }
                 }
             }
         } catch (IOException e) {

From 67c6da768d09ccd0a03bc5ba1fadc3059617c741 Mon Sep 17 00:00:00 2001
From: Mark Davis <mark@unicode.org>
Date: Sat, 4 Jul 2026 17:23:56 +0300
Subject: [PATCH 7/8] ICU-23431 Cleanup

---
 .../icu/impl/links/LinkHandlingUtilities.java    | 16 +---------------
 .../com/ibm/icu/dev/test/links/LinkTest.java     |  6 ------
 2 files changed, 1 insertion(+), 21 deletions(-)

diff --git a/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java b/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java
index 72ceec174386..d67a9bb9be4b 100644
--- a/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java
+++ b/icu4j/main/core/src/main/java/com/ibm/icu/impl/links/LinkHandlingUtilities.java
@@ -73,10 +73,6 @@ private WHATWG_PERCENT_ENCODED(WHATWG_PERCENT_ENCODED base, String uset) {
         }
     }
 
-    private static String quote(String s) {
-        return s; // later, escape for HTML
-    }
-
     /** TODO use real property */
     private static int getOpening(int cp) {
         return cp == '>' ? '<' : UCharacter.getBidiPairedBracket(cp);
@@ -162,11 +158,6 @@ public static int scanEmailBackwards(CharSequence source, int hardStart, int dom
         return result;
     }
 
-    private static String getGeneralCategory(int property, int codePoint, int nameChoice) {
-        return UCharacter.getPropertyValueName(
-                property, UCharacter.getIntPropertyValue(codePoint, property), nameChoice);
-    }
-
     private enum Structure {
         single,
         listP("/", "𝑷"),
@@ -219,7 +210,6 @@ private enum Part {
         final int initiator;
         final UnicodeSet terminators;
         final UnicodeSet clearStack;
-        final UnicodeSet extraQuoted;
         final UnicodeSet whatwg_quoted;
 
         private Part(
@@ -233,11 +223,7 @@ private Part(
             this.initiator = initiator;
             this.terminators = new UnicodeSet(terminators).freeze();
             this.clearStack = new UnicodeSet(clearStack).freeze();
-            this.extraQuoted =
-                    new UnicodeSet(extraQuoted)
-                            .addAll(this.clearStack)
-                            .addAll(this.terminators)
-                            .freeze();
+            new UnicodeSet(extraQuoted).addAll(this.clearStack).addAll(this.terminators).freeze();
             this.whatwg_quoted = whatwg_quoted == null ? UnicodeSet.EMPTY : whatwg_quoted.set;
         }
 
diff --git a/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java b/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
index 2244e0b11cef..3f96cff2ee7f 100644
--- a/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
+++ b/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
@@ -48,9 +48,6 @@
  */
 @RunWith(JUnit4.class)
 public class LinkTest extends CoreTestFmwk {
-
-    private static final boolean VERBOSE = false;
-
     /**
      * There are 2 format test cases that need to be fixed in UTS58 In each of these, the = in a 𝑽
      * doesn't need to be quoted, because it doesn't initiate a different component.
@@ -276,9 +273,6 @@ public void testAgainstLinkDetectionTest() {
                             }
                             domain.end = pqfEnd;
                         }
-                        if (start > domain.start) {
-                            int debug = 0;
-                        }
                         actual.append(rawLine2.subSequence(start, domain.start))
                                 .append("⸠")
                                 .append(rawLine2.subSequence(domain.start, domain.end))

From 24790ac83c9e96ece257245c1a5b7aededf09c25 Mon Sep 17 00:00:00 2001
From: Mark Davis <mark@unicode.org>
Date: Sat, 4 Jul 2026 19:51:46 +0300
Subject: [PATCH 8/8] ICU-23431 Cleanup

---
 .../core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java | 2 --
 1 file changed, 2 deletions(-)

diff --git a/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java b/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
index 3f96cff2ee7f..32599bfe3222 100644
--- a/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
+++ b/icu4j/main/core/src/test/java/com/ibm/icu/dev/test/links/LinkTest.java
@@ -250,8 +250,6 @@ public void testAgainstLinkDetectionTest() {
                             schemeMatcher.region(start, domain.start);
                             if (schemeMatcher.find()) {
                                 domain.start = schemeMatcher.start();
-                            } else {
-                                domain.start = domain.start;
                             }
 
                             Matcher portMatcher = PORT.matcher(rawLine2);