start implementing more efficient querying of URL patterns

5 files changed, 635 insertions, 12 deletions

diff --git a/common/patterns.js b/common/patterns.js
index 635b128..054e610 100644
--- a/common/patterns.js
+++ b/common/patterns.js
@@ -17,16 +17,25 @@ const MAX = {
 const proto_regex = /^(\w+):\/\/(.*)$/;
 
 const user_re = "[^/?#@]+@"
-const domain_re = "[.a-zA-Z0-9-]+";
+const domain_re = "[.*a-zA-Z0-9-]+";
 const path_re = "[^?#]*";
 const query_re = "\\??[^#]*";
 
 const http_regex = new RegExp(`^(${domain_re})(${path_re})(${query_re}).*`);
 
-const file_regex = new RegExp(`^(${path_re}).*`);
+const file_regex = new RegExp(`^(/${path_re}).*`);
 
 const ftp_regex = new RegExp(`^(${user_re})?(${domain_re})(${path_re}).*`);
 
+function match_or_throw(regex, string, error_msg)
+{
+    const match = regex.exec(string);
+    if (match === null)
+	throw error_msg;
+
+    return match;
+}
+
 function deconstruct_url(url, use_limits=true)
 {
     const max = MAX;
@@ -35,21 +44,19 @@ function deconstruct_url(url, use_limits=true)
 	    max[key] = Infinity;
     }
 
-    const proto_match = proto_regex.exec(url);
-    if (proto_match === null)
-	throw `bad url '${url}'`;
+    const matcher = (re, str) => match_or_throw(re, str, `bad url '${url}'`)
 
+    const proto_match = matcher(proto_regex, url);
     const deco = {proto: proto_match[1]};
 
     if (deco.proto === "file") {
-	deco.path = file_regex.exec(proto_match[2])[1];
+	deco.path = matcher(file_regex, proto_match[2])[1];
     } else if (deco.proto === "ftp") {
-	[deco.domain, deco.path] = ftp_regex.exec(proto_match[2]).slice(2, 4);
+	[deco.domain, deco.path] =
+	    matcher(ftp_regex, proto_match[2]).slice(2, 4);
     } else if (deco.proto === "http" || deco.proto === "https") {
-	const http_match = http_regex.exec(proto_match[2]);
-	if (!http_match)
-	    return undefined;
-	[deco.domain, deco.path, deco.query] = http_match.slice(1, 4);
+	[deco.domain, deco.path, deco.query] =
+	    matcher(http_regex, proto_match[2]).slice(1, 4);
 	deco.domain = deco.domain.toLowerCase();
     } else {
 	throw `unsupported protocol in url '${url}'`;
diff --git a/common/patterns_query_tree.js b/common/patterns_query_tree.js
new file mode 100644
index 0000000..326c7b7
--- /dev/null
+++ b/common/patterns_query_tree.js
@@ -0,0 +1,271 @@
+/**
+ * This file is part of Haketilo.
+ *
+ * Function: Data structure to query items by URL patterns.
+ *
+ * Copyright (C) 2021 Wojtek Kosior
+ *
+ * This program is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation, either version 3 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * As additional permission under GNU GPL version 3 section 7, you
+ * may distribute forms of that code without the copy of the GNU
+ * GPL normally required by section 4, provided you include this
+ * license notice and, in case of non-source distribution, a URL
+ * through which recipients can access the Corresponding Source.
+ * If you modify file(s) with this exception, you may extend this
+ * exception to your version of the file(s), but you are not
+ * obligated to do so. If you do not wish to do so, delete this
+ * exception statement from your version.
+ *
+ * As a special exception to the GPL, any HTML file which merely
+ * makes function calls to this code, and for that purpose
+ * includes it by reference shall be deemed a separate work for
+ * copyright law purposes. If you modify this code, you may extend
+ * this exception to your version of the code, but you are not
+ * obligated to do so. If you do not wish to do so, delete this
+ * exception statement from your version.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <https://www.gnu.org/licenses/>.
+ *
+ * I, Wojtek Kosior, thereby promise not to sue for violation of this file's
+ * license. Although I request that you do not make use this code in a
+ * proprietary program, I am not going to enforce this in court.
+ */
+
+/*
+ * IMPORTS_START
+ * IMPORT deconstruct_url
+ * IMPORTS_END
+ */
+
+/*
+ * This is a javascript rewrite of Python implementation of the algorithm here:
+ * https://git.koszko.org/pydrilla/tree/src/pydrilla/pydrilla.py?id=f4edcbe7f4739d6f82a2e1bb180960b003b30862#n205
+ */
+
+/* "Pattern Tree" is how we refer to the data structure used for querying
+ * Haketilo patterns. Those look like 'https://*.example.com/ab/***'. The goal
+ * is to make it possible for given URL to quickly retrieve all known patterns
+ * that match it.
+ */
+function make_tree_node() {
+    return {
+	wildcard_matches: [null, null, null],
+	literal_match:    null,
+	children:         {}
+    };
+}
+
+function is_empty_node(tree_node) {
+    const children = tree_node.children;
+    for (const key in children) {
+	if (children.hasOwnProperty(key))
+	    return false;
+    }
+
+    if (Array.reduce(tree_node.wildcard_matches, (a, b) => b && a !== null, 1))
+	return false;
+
+    return tree_node.literal_match === null;
+}
+
+const is_wildcard = segment => ["*", "**", "***"].lastIndexOf(segment) >= 0;
+
+/*
+ * Yields all matches of this segments sequence against the tree that starts at
+ * this node. Results are produces in order from greatest to lowest pattern
+ * specificity.
+ */
+function* search_sequence(tree_node, segments)
+{
+    const nodes = [tree_node];
+
+    for (const segment of segments) {
+	const next_node = nodes[nodes.length - 1].children[segment];
+	if (next_node === undefined)
+	    break;
+
+	nodes.push(next_node);
+    }
+
+    const nsegments = segments.length;
+
+    const conds = [
+	/* literal pattern match */
+	() => nodes.length     == nsegments,
+	/* wildcard pattern matches */
+	() => nodes.length + 1 == nsegments && segments[nsegments - 1] != "*",
+	() => nodes.length + 1 <  nsegments,
+	() => nodes.length + 1 != nsegments || segments[nsegments - 1] != "***"
+    ];
+
+    while (nodes.length) {
+	const node = nodes.pop();
+	const items = [node.literal_match, ...node.wildcard_matches];
+
+	for (let i = 0; i < 4; i++) {
+	    if (items[i] !== null && conds[i]())
+		yield items[i];
+	}
+    }
+}
+
+/*
+ * Make item queryable through (this branch of) the Pattern Tree or remove its
+ * path from there.
+ *
+ * item_modifier should be a function that accepts 1 argument, the item stored
+ * in the tree (or `null` if there wasn't any item there), and returns an item
+ * that should be used in place of the first one. It is also legal for it to
+ * return the same item modifying it first. If it returns `null`, it means the
+ * item should be deleted from the Tree.
+ *
+ * If there was not yet any item associated with the tree path designated by
+ * segments and value returned by item_modifier is not `null`, make the value
+ * queryable by this path.
+ */
+function modify_sequence(tree_node, segments, item_modifier)
+{
+    const nodes = [tree_node];
+    let removed = true;
+
+    for (var current_segment of segments) {
+	wildcards = tree_node.wildcard_matches;
+
+	const child = tree_node.children[current_segment] || make_tree_node();
+	tree_node.children[current_segment] = child;
+	tree_node = child;
+	nodes.push(tree_node);
+    }
+
+    tree_node.literal_match = item_modifier(tree_node.literal_match);
+    if (tree_node.literal_match !== null)
+	removed = false;
+
+    let i = segments.length;
+
+    if (is_wildcard(current_segment)) {
+	const asterisks = current_segment.length - 1;
+	const wildcards = nodes[i - 1].wildcard_matches;
+	wildcards[asterisks] = item_modifier(wildcards[asterisks]);
+	if (wildcards[asterisks] !== null)
+	    removed = false;
+    }
+
+    while (!removed)
+	return;
+
+    while (i > 0) {
+	tree_node = nodes[i--];
+	if (is_empty_node(tree_node))
+	    delete nodes[i].children[segments[i]];
+    }
+}
+
+/*
+ * Make item queryable through the Pattern Tree that starts with the protocols
+ * dictionary object passed in the first argument.
+ */
+function pattern_tree_register(patterns_by_proto, pattern, item_name, item)
+{
+    /*
+     * We pass 'false' to disable length limits on URL parts. Length limits are
+     * mostly useful in case of iteration over all patterns matching given URL.
+     * Here we don't do that.
+     */
+    const deco = deconstruct_url(pattern, false);
+
+    const tree = patterns_by_proto[deco.proto] || make_tree_node();
+    patterns_by_proto[deco.proto] = tree;
+
+    let path_trees;
+
+    if (deco.proto === "file") {
+	path_trees = [tree];
+    } else {
+	/* We need an aray of domain labels ordered most-significant-first. */
+	const domain = [...deco.domain].reverse();
+	path_trees = add_sequence(tree, domain, make_tree_node);
+    }
+
+    for (const path_tree of path_trees) {
+        for (const match_obj of add_sequence(path_tree, deco.path, () => {}))
+            match_obj[item_name] = item;
+    }
+}
+
+// /*
+//  * Remove registered item from the Pattern Tree that starts with the protocols
+//  * dictionary object passed in the first argument. The remaining 2 arguments
+//  * should be pattern and name that have been earlier passed to
+//  * pattern_tree_register().
+//  */
+// function pattern_tree_deregister(patterns_by_proto, pattern, item_name)
+// {
+//     const deco = deconstruct_url(pattern, false);
+
+//     const tree = patterns_by_proto[deco.proto] || make_tree_node();
+//     patterns_by_proto[deco.proto] = tree;
+
+//     let path_trees;
+
+//     if (deco.proto === "file") {
+// 	path_trees = [tree];
+//     } else {
+// 	/* We need an aray of domain labels ordered most-significant-first. */
+// 	const domain = [...deco.domain].reverse();
+// 	..........................
+// 	path_trees = add_sequence(tree, domain, make_tree_node);
+//     }
+// }
+
+/*
+ * Yield registered items (as [item_name, item] arrays) that match url. For
+ * cosistent behavior, please don't modify the pattern tree while iterating over
+ * the results.
+ */
+function* pattern_tree_search(patterns_by_proto, url)
+{
+    const deco = deconstruct_url(url, false);
+
+    tree = patterns_by_proto[deco.proto] || make_tree_node();
+
+    let path_trees;
+
+    if (deco.proto === "file") {
+	path_trees = [tree];
+    } else {
+	/* We need an aray of domain labels ordered most-significant-first. */
+	const domain = [...deco.domain].reverse();
+	path_trees = search_sequence(tree, domain);
+    }
+
+    for (const path_tree of path_trees) {
+	for (const match_obj of search_sequence(path_tree, deco.path)) {
+	    for (const entry of Object.entries(match_obj))
+		yield entry;
+	}
+    }
+}
+
+const pattern_tree = {
+    make: () => {},
+    register: pattern_tree_register,
+    // deregister: pattern_tree_deregister,
+    search: pattern_tree_search
+}
+
+/*
+ * EXPORTS_START
+ * EXPORT pattern_tree
+ * EXPORTS_END
+ */
diff --git a/test/unit/conftest.py b/test/unit/conftest.py
index 6877b7a..62cc1a0 100644
--- a/test/unit/conftest.py
+++ b/test/unit/conftest.py
@@ -76,7 +76,7 @@ try {
 return window.haketilo_selenium_return_value;
 '''
 
-def _execute_in_page_context(driver, script, *args):
+def _execute_in_page_context(driver, script, args):
     script = script + '\n;\nwindow.haketilo_selenium_exception = false;'
     try:
         return driver.execute_script(script_injecting_script, script, args)
diff --git a/test/unit/test_patterns.py b/test/unit/test_patterns.py
index 4162fc0..4cfc10c 100644
--- a/test/unit/test_patterns.py
+++ b/test/unit/test_patterns.py
@@ -89,3 +89,65 @@ def test_regexes(execute_in_page, patterns_code):
     match = execute_in_page('returnval(ftp_regex.exec(arguments[0]));',
                             '@bad.url/')
     assert match is None
+
+def test_deconstruct_url(execute_in_page, patterns_code):
+    """
+    patterns.js contains deconstruct_url() function that handles URL parsing.
+    Verify it works properly.
+    """
+    execute_in_page(patterns_code, page='https://gotmyowndoma.in')
+
+    deco = execute_in_page('returnval(deconstruct_url(arguments[0]));',
+                           'https://eXaMpLe.com/a/b?ver=1.2.3#heading2')
+    assert deco
+    assert deco['trailing_dash'] == False
+    assert deco['proto']         == 'https'
+    assert deco['domain']        == ['example', 'com']
+    assert deco['path']          == ['a', 'b']
+
+    deco = execute_in_page('returnval(deconstruct_url(arguments[0]));',
+                           'http://**.example.com/')
+    assert deco
+    assert deco['trailing_dash'] == True
+    assert deco['proto']         == 'http'
+    assert deco['domain']        == ['**', 'example', 'com']
+    assert deco['path']          == []
+
+    deco = execute_in_page('returnval(deconstruct_url(arguments[0]));',
+                           'ftp://user@ftp.example.com/all///passwords.txt/')
+    assert deco
+    assert deco['trailing_dash'] == True
+    assert deco['proto']         == 'ftp'
+    assert deco['domain']        == ['ftp', 'example', 'com']
+    assert deco['path']          == ['all', 'passwords.txt']
+
+    deco = execute_in_page('returnval(deconstruct_url(arguments[0]));',
+                           'ftp://mirror.edu.pl.eu.org')
+    assert deco
+    assert deco['trailing_dash'] == False
+    assert deco['proto']         == 'ftp'
+    assert deco['domain']        == ['mirror', 'edu', 'pl', 'eu', 'org']
+    assert deco['path']          == []
+
+    deco = execute_in_page('returnval(deconstruct_url(arguments[0]));',
+                           'file:///mnt/parabola_chroot///etc/passwd')
+    assert deco
+    assert deco['trailing_dash'] == False
+    assert deco['proto']         == 'file'
+    assert deco['path']          == ['mnt', 'parabola_chroot', 'etc', 'passwd']
+
+    for bad_url in [
+            '://bad-url.missing/protocol',
+            'http:/example.com/a/b',
+            'unknown://example.com/a/b',
+            'idontfancypineapple',
+            'ftp://@example.org/',
+            'https:///some/path/',
+            'file://non-absolute/path'
+    ]:
+        with pytest.raises(Exception, match=r'Error in injected script'):
+            deco = execute_in_page('returnval(deconstruct_url(arguments[0]));',
+                                   bad_url)
+
+    # at some point we might also consider testing url deconstruction with
+    # length limits...
diff --git a/test/unit/test_patterns_query_tree.py b/test/unit/test_patterns_query_tree.py
new file mode 100644
index 0000000..9fbc0c3
--- /dev/null
+++ b/test/unit/test_patterns_query_tree.py
@@ -0,0 +1,283 @@
+# SPDX-License-Identifier: CC0-1.0
+
+"""
+Haketilo unit tests - URL patterns
+"""
+
+# This file is part of Haketilo
+#
+# Copyright (C) 2021, Wojtek Kosior
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the CC0 1.0 Universal License as published by
+# the Creative Commons Corporation.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# CC0 1.0 Universal License for more details.
+
+import pytest
+
+from ..script_loader import load_script
+
+@pytest.fixture(scope="session")
+def patterns_tree_code():
+    yield load_script('common/patterns_query_tree.js', ['common'])
+
+def test_modify_branch(execute_in_page, patterns_tree_code):
+    """
+    patterns_query_tree.js contains Patterns Tree data structure that allows
+    arrays of string labels to be mapped to items.
+    Verify operations modifying a single branch of such tree work properly.
+    """
+    execute_in_page(patterns_tree_code, page='https://gotmyowndoma.in')
+    execute_in_page(
+        '''
+        let items_added;
+        let items_removed;
+
+        function _item_adder(item, array)
+        {
+            items_added++;
+            return [...(array || []), item];
+        }
+
+        function item_adder(item)
+        {
+            items_added = 0;
+            return array => _item_adder(item, array);
+        }
+
+        function _item_remover(array)
+        {
+            if (array !== null) {
+                items_removed++;
+                array.pop();
+            }
+            return (array && array.length > 0) ? array : null;
+        }
+
+        function item_remover()
+        {
+            items_removed = 0;
+            return _item_remover;
+        }''')
+
+    # Let's construct some tree branch while checking that each addition gives
+    # the right result.
+    branch = execute_in_page(
+        '''{
+        const branch = make_tree_node();
+        modify_sequence(branch, ['com', 'example'], item_adder('some_item'));
+        returnval(branch);
+        }''')
+    assert branch == {
+        'literal_match': None,
+        'wildcard_matches': [None, None, None],
+        'children': {
+            'com': {
+                'literal_match': None,
+                'wildcard_matches': [None, None, None],
+                'children': {
+                    'example': {
+                        'literal_match': ['some_item'],
+                        'wildcard_matches': [None, None, None],
+                        'children': {
+                        }
+                    }
+                }
+            }
+        }
+    }
+
+    branch, items_added = execute_in_page(
+        '''{
+        const branch = arguments[0];
+        modify_sequence(branch, ['com', 'example'], item_adder('other_item'));
+        returnval([branch, items_added]);
+        }''', branch)
+    assert items_added == 1
+    assert branch['children']['com']['children']['example']['literal_match'] \
+            == ['some_item', 'other_item']
+
+    for i in range(3):
+        for expected_array in [['third_item'], ['third_item', '4th_item']]:
+            wildcard = '*' * (i + 1)
+            branch, items_added = execute_in_page(
+                '''{
+                const branch = arguments[0];
+                modify_sequence(branch, ['com', 'sample', arguments[1]],
+                                item_adder(arguments[2]));
+                returnval([branch, items_added]);
+                }''',
+                branch, wildcard, expected_array[-1])
+            assert items_added == 2
+            sample = branch['children']['com']['children']['sample']
+            assert sample['wildcard_matches'][i] == expected_array
+            assert sample['children'][wildcard]['literal_match'] \
+                == expected_array
+
+    branch, items_added = execute_in_page(
+        '''{
+        const branch = arguments[0];
+        modify_sequence(branch, ['org', 'koszko', '***', '123'],
+                        item_adder('5th_item'));
+        returnval([branch, items_added]);
+        }''',
+        branch)
+    assert items_added == 1
+    assert branch['children']['org']['children']['koszko']['children']['***']\
+        ['children']['123']['literal_match'] == ['5th_item']
+
+    # Let's verify that removing a nonexistent element doesn't modify the tree.
+    branch2, items_removed = execute_in_page(
+        '''{
+        const branch = arguments[0];
+        modify_sequence(branch, ['com', 'not', 'registered', '*'],
+                        item_remover());
+        returnval([branch, items_removed]);
+        }''',
+        branch)
+    assert branch == branch2
+    assert items_removed == 0
+
+    # Let's remove all elements in the tree branch while checking that each
+    # removal gives the right result.
+    branch, items_removed = execute_in_page(
+        '''{
+        const branch = arguments[0];
+        modify_sequence(branch, ['org', 'koszko', '***', '123'],
+                        item_remover());
+        returnval([branch, items_removed]);
+        }''',
+        branch)
+    assert items_removed == 1
+    assert 'org' not in branch['children']
+
+    for i in range(3):
+        for expected_array in [['third_item'], None]:
+            wildcard = '*' * (i + 1)
+            branch, items_removed = execute_in_page(
+                '''{
+                const branch = arguments[0];
+                modify_sequence(branch, ['com', 'sample', arguments[1]],
+                                item_remover());
+                returnval([branch, items_removed]);
+                }''',
+                branch, wildcard)
+            assert items_removed == 2
+            if i == 2 and expected_array == []:
+                break
+            sample = branch['children']['com']['children'].get('sample', {})
+            assert sample.get('wildcard_matches', [None, None, None])[i] \
+                == expected_array
+            assert sample.get('children', {}).get(wildcard, {})\
+                .get('literal_match') == expected_array
+
+    for i in range(2):
+        branch, items_removed = execute_in_page(
+            '''{
+            const branch = arguments[0];
+            modify_sequence(branch, ['com', 'example'], item_remover());
+            returnval([branch, items_removed]);
+            }''',
+            branch)
+        assert items_removed == 1
+        if i == 0:
+            assert branch['children']['com']['children']['example']\
+                ['literal_match'] == ['some_item']
+        else:
+            assert branch == {
+                'literal_match': None,
+                'wildcard_matches': [None, None, None],
+                'children': {
+                }
+            }
+
+def test_search_branch(execute_in_page, patterns_tree_code):
+    """
+    patterns_query_tree.js contains Patterns Tree data structure that allows
+    arrays of string labels to be mapped to items.
+    Verify searching a single branch of such tree work properly.
+    """
+    execute_in_page(patterns_tree_code, page='https://gotmyowndoma.in')
+    execute_in_page(
+        '''
+        const item_adder = item => (array => [...(array || []), item]);
+        ''')
+
+    # Let's construct some tree branch to test on.
+    execute_in_page(
+        '''
+        var branch = make_tree_node();
+
+        for (const [item, sequence] of [
+            ['(root)', []],
+            ['***',    ['***']],
+            ['**',     ['**']],
+            ['*',      ['*']],
+
+            ['a',      ['a']],
+            ['A',      ['a']],
+            ['b',      ['b']],
+
+            ['a/***',  ['a', '***']],
+            ['A/***',  ['a', '***']],
+            ['a/**',   ['a', '**']],
+            ['A/**',   ['a', '**']],
+            ['a/*',    ['a', '*']],
+            ['A/*',    ['a', '*']],
+            ['a/sth',  ['a', 'sth']],
+            ['A/sth',  ['a', 'sth']],
+
+            ['b/***',  ['b', '***']],
+            ['b/**',   ['b', '**']],
+            ['b/*',    ['b', '*']],
+            ['b/sth',  ['b', 'sth']],
+        ])
+            modify_sequence(branch, sequence, item_adder(item));
+        ''')
+
+    # Let's make the actual searches on our testing branch.
+    for sequence, expected in [
+            ([],      [{'(root)'},                            {'***'}]),
+            (['a'],   [{'a', 'A'}, {'a/***', 'A/***'}, {'*'}, {'***'}]),
+            (['b'],   [{'b'},      {'b/***'},          {'*'}, {'***'}]),
+            (['c'],   [                                {'*'}, {'***'}]),
+            (['***'], [{'***'},                        {'*'}         ]),
+            (['**'],  [{'**'},                         {'*'}, {'***'}]),
+            (['**'],  [{'**'},                         {'*'}, {'***'}]),
+            (['*'],   [{'*'},                                 {'***'}]),
+
+            (['a', 'sth'], [{'a/sth', 'A/sth'}, {'a/*', 'A/*'}, {'a/***', 'A/***'}, {'**'}, {'***'}]),
+            (['b', 'sth'], [{'b/sth'},          {'b/*'},        {'b/***'},          {'**'}, {'***'}]),
+            (['a', 'hts'], [                    {'a/*', 'A/*'}, {'a/***', 'A/***'}, {'**'}, {'***'}]),
+            (['b', 'hts'], [                    {'b/*'},        {'b/***'},          {'**'}, {'***'}]),
+            (['a', '***'], [{'a/***', 'A/***'}, {'a/*', 'A/*'},                     {'**'}, {'***'}]),
+            (['b', '***'], [{'b/***'},          {'b/*'},                            {'**'}, {'***'}]),
+            (['a', '**'],  [{'a/**', 'A/**'},   {'a/*', 'A/*'}, {'a/***', 'A/***'}, {'**'}, {'***'}]),
+            (['b', '**'],  [{'b/**'},           {'b/*'},        {'b/***'},          {'**'}, {'***'}]),
+            (['a', '*'],   [{'a/*', 'A/*'},                     {'a/***', 'A/***'}, {'**'}, {'***'}]),
+            (['b', '*'],   [{'b/*'},                            {'b/***'},          {'**'}, {'***'}]),
+
+            (['a', 'c', 'd'], [{'a/**', 'A/**'}, {'a/***', 'A/***'}, {'**'}, {'***'}]),
+            (['b', 'c', 'd'], [{'b/**'},         {'b/***'},          {'**'}, {'***'}])
+    ]:
+        result = execute_in_page(
+            '''
+            returnval([...search_sequence(branch, arguments[0])]);
+            ''',
+            sequence)
+
+        try:
+            assert len(result) == len(expected)
+
+            for expected_set, result_array in zip(expected, result):
+                assert len(expected_set) == len(result_array)
+                assert expected_set      == set(result_array)
+        except Exception as e:
+            import sys
+            print('sequence:', sequence, '\nexpected:', expected,
+                  '\nresult:', result, file=sys.stderr)
+            raise e from None