zulip/frontend_tests/node_tests/search_suggestion_legacy.js

1143 lines
31 KiB
JavaScript
Raw Normal View History

set_global('page_params', {
search_pills_enabled: false,
});
set_global('message_store', {
user_ids: () => [],
});
const settings_config = zrequire('settings_config');
page_params.realm_email_address_visibility =
settings_config.email_address_visibility_values.admins_only.code;
const huddle_data = zrequire('huddle_data');
zrequire('typeahead_helper');
set_global('Handlebars', global.make_handlebars());
zrequire('Filter', 'js/filter');
zrequire('narrow_state');
zrequire('stream_data');
zrequire('stream_topic_history');
zrequire('people');
zrequire('unread');
zrequire('common');
const search = zrequire('search_suggestion');
search.max_num_of_search_results = 15;
const me = {
email: 'myself@zulip.com',
full_name: 'Me Myself',
user_id: 41,
};
const bob = {
email: 'bob@zulip.com',
full_name: 'Bob Roberts',
user_id: 42,
};
const ted =
{
email: 'ted@zulip.com',
user_id: 101,
full_name: 'Ted Smith',
};
const alice =
{
email: 'alice@zulip.com',
user_id: 102,
full_name: 'Alice Ignore',
};
const jeff =
{
email: 'jeff@zulip.com',
user_id: 103,
full_name: 'Jeff Zoolipson',
};
function init() {
people.init();
people.add_active_user(bob);
people.add_active_user(me);
people.add_active_user(ted);
people.add_active_user(alice);
people.add_active_user(jeff);
people.initialize_current_user(me.user_id);
}
init();
set_global('narrow', {});
page_params.is_admin = true;
stream_topic_history.reset();
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
function get_suggestions(base_query, query) {
return search.get_suggestions(base_query, query);
}
run_test('basic_get_suggestions', () => {
const query = 'fred';
global.stream_data.subscribed_streams = function () {
return [];
};
global.narrow_state.stream = function () {
return 'office';
};
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
const suggestions = get_suggestions('', query);
const expected = [
'fred',
];
assert.deepEqual(suggestions.strings, expected);
});
run_test('subset_suggestions', () => {
const query = 'stream:Denmark topic:Hamlet shakespeare';
global.stream_data.subscribed_streams = function () {
return [];
};
global.narrow_state.stream = function () {
return;
};
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
const suggestions = get_suggestions('', query);
const expected = [
"stream:Denmark topic:Hamlet shakespeare",
"stream:Denmark topic:Hamlet",
"stream:Denmark",
];
assert.deepEqual(suggestions.strings, expected);
});
run_test('private_suggestions', () => {
global.stream_data.subscribed_streams = function () {
return [];
};
global.narrow_state.stream = function () {
return;
};
let query = 'is:private';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
let suggestions = get_suggestions('', query);
let expected = [
"is:private",
"pm-with:alice@zulip.com",
"pm-with:bob@zulip.com",
"pm-with:jeff@zulip.com",
"pm-with:myself@zulip.com",
"pm-with:ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
query = 'is:private al';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"is:private al",
"is:private is:alerted",
"is:private sender:alice@zulip.com",
"is:private pm-with:alice@zulip.com",
"is:private group-pm-with:alice@zulip.com",
"is:private",
];
assert.deepEqual(suggestions.strings, expected);
query = 'pm-with:t';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"pm-with:t",
"pm-with:ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
query = '-pm-with:t';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"-pm-with:t",
"is:private -pm-with:ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
query = 'pm-with:ted@zulip.com';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"pm-with:ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
query = 'sender:ted';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"sender:ted",
"sender:ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
query = 'sender:te';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"sender:te",
"sender:ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
query = '-sender:te';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"-sender:te",
"-sender:ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
query = 'sender:ted@zulip.com';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"sender:ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
query = 'is:unread from:ted';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"is:unread from:ted",
"is:unread from:ted@zulip.com",
"is:unread",
];
assert.deepEqual(suggestions.strings, expected);
// Users can enter bizarre queries, and if they do, we want to
// be conservative with suggestions.
query = 'is:private near:3';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"is:private near:3",
"is:private",
];
assert.deepEqual(suggestions.strings, expected);
query = 'pm-with:ted@zulip.com near:3';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"pm-with:ted@zulip.com near:3",
"pm-with:ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
// Make sure suggestions still work if preceding tokens
query = 'is:alerted sender:ted@zulip.com';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"is:alerted sender:ted@zulip.com",
"is:alerted",
];
assert.deepEqual(suggestions.strings, expected);
query = 'is:starred has:link is:private al';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"is:starred has:link is:private al",
"is:starred has:link is:private is:alerted",
"is:starred has:link is:private sender:alice@zulip.com",
"is:starred has:link is:private pm-with:alice@zulip.com",
"is:starred has:link is:private group-pm-with:alice@zulip.com",
"is:starred has:link is:private",
"is:starred has:link",
"is:starred",
];
assert.deepEqual(suggestions.strings, expected);
// Make sure it handles past context correctly
query = 'stream:Denmark pm-with:';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
'stream:Denmark pm-with:',
'stream:Denmark',
];
assert.deepEqual(suggestions.strings, expected);
query = 'sender:ted@zulip.com sender:';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
'sender:ted@zulip.com sender:',
'sender:ted@zulip.com',
];
assert.deepEqual(suggestions.strings, expected);
});
run_test('group_suggestions', () => {
global.stream_data.subscribed_streams = function () {
return [];
};
global.narrow_state.stream = function () {
return;
};
// Entering a comma in a pm-with query should immediately generate
// suggestions for the next person.
let query = 'pm-with:bob@zulip.com,';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
let suggestions = get_suggestions('', query);
let expected = [
"pm-with:bob@zulip.com,",
"pm-with:bob@zulip.com,alice@zulip.com",
"pm-with:bob@zulip.com,jeff@zulip.com",
"pm-with:bob@zulip.com,ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
// Only the last part of a comma-separated pm-with query should be used to
// generate suggestions.
query = 'pm-with:bob@zulip.com,t';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"pm-with:bob@zulip.com,t",
"pm-with:bob@zulip.com,ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
// Smit should also generate ted@zulip.com (Ted Smith) as a suggestion.
query = 'pm-with:bob@zulip.com,Smit';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"pm-with:bob@zulip.com,Smit",
"pm-with:bob@zulip.com,ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
// Do not suggest "myself@zulip.com" (the name of the current user)
query = 'pm-with:ted@zulip.com,my';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"pm-with:ted@zulip.com,my",
];
assert.deepEqual(suggestions.strings, expected);
// No superfluous suggestions should be generated.
query = 'pm-with:bob@zulip.com,red';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"pm-with:bob@zulip.com,red",
];
assert.deepEqual(suggestions.strings, expected);
// is:private should be properly prepended to each suggestion if the pm-with
// operator is negated.
query = '-pm-with:bob@zulip.com,';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"-pm-with:bob@zulip.com,",
"is:private -pm-with:bob@zulip.com,alice@zulip.com",
"is:private -pm-with:bob@zulip.com,jeff@zulip.com",
"is:private -pm-with:bob@zulip.com,ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
query = '-pm-with:bob@zulip.com,t';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"-pm-with:bob@zulip.com,t",
"is:private -pm-with:bob@zulip.com,ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
query = '-pm-with:bob@zulip.com,Smit';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"-pm-with:bob@zulip.com,Smit",
"is:private -pm-with:bob@zulip.com,ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
query = '-pm-with:bob@zulip.com,red';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"-pm-with:bob@zulip.com,red",
];
assert.deepEqual(suggestions.strings, expected);
// Test multiple operators
query = 'is:starred has:link pm-with:bob@zulip.com,Smit';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"is:starred has:link pm-with:bob@zulip.com,Smit",
"is:starred has:link pm-with:bob@zulip.com,ted@zulip.com",
"is:starred has:link",
"is:starred",
];
assert.deepEqual(suggestions.strings, expected);
query = 'stream:Denmark has:link pm-with:bob@zulip.com,Smit';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"stream:Denmark has:link pm-with:bob@zulip.com,Smit",
"stream:Denmark has:link",
"stream:Denmark",
];
assert.deepEqual(suggestions.strings, expected);
function message(user_ids, timestamp) {
return {
type: 'private',
display_recipient: user_ids.map((id) => ({
id: id,
})),
timestamp: timestamp,
};
}
huddle_data.process_loaded_messages([
message([bob.user_id, ted.user_id], 99),
message([bob.user_id, ted.user_id, jeff.user_id], 98),
]);
// Simulate a past huddle which should now prioritize ted over alice
query = 'pm-with:bob@zulip.com,';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"pm-with:bob@zulip.com,",
"pm-with:bob@zulip.com,ted@zulip.com",
"pm-with:bob@zulip.com,alice@zulip.com",
"pm-with:bob@zulip.com,jeff@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
// bob,ted,jeff is already an existing huddle, so prioritize this one
query = 'pm-with:bob@zulip.com,ted@zulip.com,';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"pm-with:bob@zulip.com,ted@zulip.com,",
"pm-with:bob@zulip.com,ted@zulip.com,jeff@zulip.com",
"pm-with:bob@zulip.com,ted@zulip.com,alice@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
// bob,ted,jeff is already an existing huddle, but if we start with just jeff,
// then don't prioritize ted over alice because it doesn't complete the full huddle.
query = 'pm-with:jeff@zulip.com,';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"pm-with:jeff@zulip.com,",
"pm-with:jeff@zulip.com,alice@zulip.com",
"pm-with:jeff@zulip.com,bob@zulip.com",
"pm-with:jeff@zulip.com,ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
query = "pm-with:jeff@zulip.com,ted@zulip.com hi";
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"pm-with:jeff@zulip.com,ted@zulip.com hi",
"pm-with:jeff@zulip.com,ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
});
init();
run_test('empty_query_suggestions', () => {
const query = '';
global.stream_data.subscribed_streams = function () {
return ['devel', 'office'];
};
global.narrow_state.stream = function () {
return;
};
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
const suggestions = get_suggestions('', query);
const expected = [
"",
"streams:public",
"is:private",
"is:starred",
"is:mentioned",
"is:alerted",
"is:unread",
"sender:myself@zulip.com",
"stream:devel",
"stream:office",
'has:link',
'has:image',
'has:attachment',
];
assert.deepEqual(suggestions.strings, expected);
function describe(q) {
return suggestions.lookup_table.get(q).description;
}
assert.equal(describe('is:private'), 'Private messages');
assert.equal(describe('is:starred'), 'Starred messages');
assert.equal(describe('is:mentioned'), '@-mentions');
assert.equal(describe('is:alerted'), 'Alerted messages');
assert.equal(describe('is:unread'), 'Unread messages');
assert.equal(describe('sender:myself@zulip.com'), 'Sent by me');
assert.equal(describe('has:link'), 'Messages with one or more link');
assert.equal(describe('has:image'), 'Messages with one or more image');
assert.equal(describe('has:attachment'), 'Messages with one or more attachment');
});
run_test('has_suggestions', () => {
// Checks that category wise suggestions are displayed instead of a single
// default suggestion when suggesting `has` operator.
let query = 'h';
global.stream_data.subscribed_streams = function () {
return ['devel', 'office'];
};
global.narrow_state.stream = function () {
return;
};
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
let suggestions = get_suggestions('', query);
let expected = [
"h",
'has:link',
'has:image',
'has:attachment',
];
assert.deepEqual(suggestions.strings, expected);
function describe(q) {
return suggestions.lookup_table.get(q).description;
}
assert.equal(describe('has:link'), 'Messages with one or more link');
assert.equal(describe('has:image'), 'Messages with one or more image');
assert.equal(describe('has:attachment'), 'Messages with one or more attachment');
query = '-h';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"-h",
'-has:link',
'-has:image',
'-has:attachment',
];
assert.deepEqual(suggestions.strings, expected);
assert.equal(describe('-has:link'), 'Exclude messages with one or more link');
assert.equal(describe('-has:image'), 'Exclude messages with one or more image');
assert.equal(describe('-has:attachment'), 'Exclude messages with one or more attachment');
// operand suggestions follow.
query = 'has:';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
'has:link',
'has:image',
'has:attachment',
];
assert.deepEqual(suggestions.strings, expected);
query = 'has:im';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
'has:image',
];
assert.deepEqual(suggestions.strings, expected);
query = '-has:im';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
'-has:image',
];
assert.deepEqual(suggestions.strings, expected);
query = 'att';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
'att',
'has:attachment',
];
assert.deepEqual(suggestions.strings, expected);
query = 'stream:Denmark is:alerted has:lin';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
'stream:Denmark is:alerted has:link',
'stream:Denmark is:alerted',
'stream:Denmark',
];
assert.deepEqual(suggestions.strings, expected);
});
run_test('check_is_suggestions', () => {
global.stream_data.subscribed_streams = function () {
return ['devel', 'office'];
};
global.narrow_state.stream = function () {
return;
};
let query = 'i';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
let suggestions = get_suggestions('', query);
let expected = [
'i',
'is:private',
'is:starred',
'is:mentioned',
'is:alerted',
'is:unread',
'sender:alice@zulip.com',
'pm-with:alice@zulip.com',
'group-pm-with:alice@zulip.com',
'has:image',
];
assert.deepEqual(suggestions.strings, expected);
function describe(q) {
return suggestions.lookup_table.get(q).description;
}
assert.equal(describe('is:private'), 'Private messages');
assert.equal(describe('is:starred'), 'Starred messages');
assert.equal(describe('is:mentioned'), '@-mentions');
assert.equal(describe('is:alerted'), 'Alerted messages');
assert.equal(describe('is:unread'), 'Unread messages');
query = '-i';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
'-i',
'-is:private',
'-is:starred',
'-is:mentioned',
'-is:alerted',
'-is:unread',
];
assert.deepEqual(suggestions.strings, expected);
assert.equal(describe('-is:private'), 'Exclude private messages');
assert.equal(describe('-is:starred'), 'Exclude starred messages');
assert.equal(describe('-is:mentioned'), 'Exclude @-mentions');
assert.equal(describe('-is:alerted'), 'Exclude alerted messages');
assert.equal(describe('-is:unread'), 'Exclude unread messages');
// operand suggestions follow.
query = 'is:';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
'is:private',
'is:starred',
'is:mentioned',
'is:alerted',
'is:unread',
];
assert.deepEqual(suggestions.strings, expected);
query = 'is:st';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
'is:starred',
];
assert.deepEqual(suggestions.strings, expected);
query = '-is:st';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
'-is:starred',
];
assert.deepEqual(suggestions.strings, expected);
query = 'st';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
'st',
'streams:public',
'is:starred',
'stream:',
];
assert.deepEqual(suggestions.strings, expected);
query = 'stream:Denmark has:link is:sta';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
'stream:Denmark has:link is:starred',
'stream:Denmark has:link',
'stream:Denmark',
];
assert.deepEqual(suggestions.strings, expected);
});
run_test('sent_by_me_suggestions', () => {
global.stream_data.subscribed_streams = function () {
return [];
};
global.narrow_state.stream = function () {
return;
};
let query = '';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
let suggestions = get_suggestions('', query);
assert(suggestions.strings.includes('sender:myself@zulip.com'));
assert.equal(suggestions.lookup_table.get('sender:myself@zulip.com').description,
'Sent by me');
query = 'sender';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
let expected = [
"sender",
"sender:myself@zulip.com",
"sender:",
];
assert.deepEqual(suggestions.strings, expected);
query = '-sender';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"-sender",
"-sender:myself@zulip.com",
"-sender:",
];
assert.deepEqual(suggestions.strings, expected);
query = 'from';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"from",
"from:myself@zulip.com",
"from:",
];
assert.deepEqual(suggestions.strings, expected);
query = '-from';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"-from",
"-from:myself@zulip.com",
"-from:",
];
assert.deepEqual(suggestions.strings, expected);
query = 'sender:bob@zulip.com';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"sender:bob@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
query = 'from:bob@zulip.com';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"from:bob@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
query = 'sent';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"sent",
"sender:myself@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
query = '-sent';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"-sent",
"-sender:myself@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
query = 'stream:Denmark topic:Denmark1 sent';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"stream:Denmark topic:Denmark1 sent",
"stream:Denmark topic:Denmark1 sender:myself@zulip.com",
"stream:Denmark topic:Denmark1",
"stream:Denmark",
];
assert.deepEqual(suggestions.strings, expected);
query = 'is:starred sender:m';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"is:starred sender:m",
"is:starred sender:myself@zulip.com",
"is:starred",
];
assert.deepEqual(suggestions.strings, expected);
query = 'sender:alice@zulip.com sender:';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"sender:alice@zulip.com sender:",
"sender:alice@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
});
run_test('topic_suggestions', () => {
let suggestions;
let expected;
global.stream_data.subscribed_streams = function () {
return ['office'];
};
global.narrow_state.stream = function () {
return 'office';
};
const devel_id = 44;
const office_id = 77;
global.stream_data.get_stream_id = function (stream_name) {
switch (stream_name) {
case 'office': return office_id;
case 'devel': return devel_id;
}
};
stream_topic_history.reset();
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', 'te');
expected = [
"te",
"sender:ted@zulip.com",
"pm-with:ted@zulip.com",
"group-pm-with:ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
stream_topic_history.add_message({
stream_id: devel_id,
topic_name: 'REXX',
});
js: Automatically convert _.each to for…of. This commit was automatically generated by the following script, followed by lint --fix and a few small manual lint-related cleanups. import * as babelParser from "recast/parsers/babel"; import * as recast from "recast"; import * as tsParser from "recast/parsers/typescript"; import { builders as b, namedTypes as n } from "ast-types"; import { Context } from "ast-types/lib/path-visitor"; import K from "ast-types/gen/kinds"; import { NodePath } from "ast-types/lib/node-path"; import assert from "assert"; import fs from "fs"; import path from "path"; import process from "process"; const checkExpression = (node: n.Node): node is K.ExpressionKind => n.Expression.check(node); const checkStatement = (node: n.Node): node is K.StatementKind => n.Statement.check(node); for (const file of process.argv.slice(2)) { console.log("Parsing", file); const ast = recast.parse(fs.readFileSync(file, { encoding: "utf8" }), { parser: path.extname(file) === ".ts" ? tsParser : babelParser, }); let changed = false; let inLoop = false; let replaceReturn = false; const visitLoop = (...args: string[]) => function(this: Context, path: NodePath) { for (const arg of args) { this.visit(path.get(arg)); } const old = { inLoop }; inLoop = true; this.visit(path.get("body")); inLoop = old.inLoop; return false; }; recast.visit(ast, { visitDoWhileStatement: visitLoop("test"), visitExpressionStatement(path) { const { expression, comments } = path.node; let valueOnly; if ( n.CallExpression.check(expression) && n.MemberExpression.check(expression.callee) && !expression.callee.computed && n.Identifier.check(expression.callee.object) && expression.callee.object.name === "_" && n.Identifier.check(expression.callee.property) && ["each", "forEach"].includes(expression.callee.property.name) && [2, 3].includes(expression.arguments.length) && checkExpression(expression.arguments[0]) && (n.FunctionExpression.check(expression.arguments[1]) || n.ArrowFunctionExpression.check(expression.arguments[1])) && [1, 2].includes(expression.arguments[1].params.length) && n.Identifier.check(expression.arguments[1].params[0]) && ((valueOnly = expression.arguments[1].params[1] === undefined) || n.Identifier.check(expression.arguments[1].params[1])) && (expression.arguments[2] === undefined || n.ThisExpression.check(expression.arguments[2])) ) { const old = { inLoop, replaceReturn }; inLoop = false; replaceReturn = true; this.visit( path .get("expression") .get("arguments") .get(1) .get("body") ); inLoop = old.inLoop; replaceReturn = old.replaceReturn; const [right, { body, params }] = expression.arguments; const loop = b.forOfStatement( b.variableDeclaration("let", [ b.variableDeclarator( valueOnly ? params[0] : b.arrayPattern([params[1], params[0]]) ), ]), valueOnly ? right : b.callExpression( b.memberExpression(right, b.identifier("entries")), [] ), checkStatement(body) ? body : b.expressionStatement(body) ); loop.comments = comments; path.replace(loop); changed = true; } this.traverse(path); }, visitForStatement: visitLoop("init", "test", "update"), visitForInStatement: visitLoop("left", "right"), visitForOfStatement: visitLoop("left", "right"), visitFunction(path) { this.visit(path.get("params")); const old = { replaceReturn }; replaceReturn = false; this.visit(path.get("body")); replaceReturn = old.replaceReturn; return false; }, visitReturnStatement(path) { if (replaceReturn) { assert(!inLoop); // could use labeled continue if this ever fires const { argument, comments } = path.node; if (argument === null) { const s = b.continueStatement(); s.comments = comments; path.replace(s); } else { const s = b.expressionStatement(argument); s.comments = comments; path.replace(s, b.continueStatement()); } return false; } this.traverse(path); }, visitWhileStatement: visitLoop("test"), }); if (changed) { console.log("Writing", file); fs.writeFileSync(file, recast.print(ast).code, { encoding: "utf8" }); } } Signed-off-by: Anders Kaseorg <anders@zulipchat.com>
2020-02-06 06:19:47 +01:00
for (const topic_name of ['team', 'ignore', 'test']) {
stream_topic_history.add_message({
stream_id: office_id,
topic_name: topic_name,
});
js: Automatically convert _.each to for…of. This commit was automatically generated by the following script, followed by lint --fix and a few small manual lint-related cleanups. import * as babelParser from "recast/parsers/babel"; import * as recast from "recast"; import * as tsParser from "recast/parsers/typescript"; import { builders as b, namedTypes as n } from "ast-types"; import { Context } from "ast-types/lib/path-visitor"; import K from "ast-types/gen/kinds"; import { NodePath } from "ast-types/lib/node-path"; import assert from "assert"; import fs from "fs"; import path from "path"; import process from "process"; const checkExpression = (node: n.Node): node is K.ExpressionKind => n.Expression.check(node); const checkStatement = (node: n.Node): node is K.StatementKind => n.Statement.check(node); for (const file of process.argv.slice(2)) { console.log("Parsing", file); const ast = recast.parse(fs.readFileSync(file, { encoding: "utf8" }), { parser: path.extname(file) === ".ts" ? tsParser : babelParser, }); let changed = false; let inLoop = false; let replaceReturn = false; const visitLoop = (...args: string[]) => function(this: Context, path: NodePath) { for (const arg of args) { this.visit(path.get(arg)); } const old = { inLoop }; inLoop = true; this.visit(path.get("body")); inLoop = old.inLoop; return false; }; recast.visit(ast, { visitDoWhileStatement: visitLoop("test"), visitExpressionStatement(path) { const { expression, comments } = path.node; let valueOnly; if ( n.CallExpression.check(expression) && n.MemberExpression.check(expression.callee) && !expression.callee.computed && n.Identifier.check(expression.callee.object) && expression.callee.object.name === "_" && n.Identifier.check(expression.callee.property) && ["each", "forEach"].includes(expression.callee.property.name) && [2, 3].includes(expression.arguments.length) && checkExpression(expression.arguments[0]) && (n.FunctionExpression.check(expression.arguments[1]) || n.ArrowFunctionExpression.check(expression.arguments[1])) && [1, 2].includes(expression.arguments[1].params.length) && n.Identifier.check(expression.arguments[1].params[0]) && ((valueOnly = expression.arguments[1].params[1] === undefined) || n.Identifier.check(expression.arguments[1].params[1])) && (expression.arguments[2] === undefined || n.ThisExpression.check(expression.arguments[2])) ) { const old = { inLoop, replaceReturn }; inLoop = false; replaceReturn = true; this.visit( path .get("expression") .get("arguments") .get(1) .get("body") ); inLoop = old.inLoop; replaceReturn = old.replaceReturn; const [right, { body, params }] = expression.arguments; const loop = b.forOfStatement( b.variableDeclaration("let", [ b.variableDeclarator( valueOnly ? params[0] : b.arrayPattern([params[1], params[0]]) ), ]), valueOnly ? right : b.callExpression( b.memberExpression(right, b.identifier("entries")), [] ), checkStatement(body) ? body : b.expressionStatement(body) ); loop.comments = comments; path.replace(loop); changed = true; } this.traverse(path); }, visitForStatement: visitLoop("init", "test", "update"), visitForInStatement: visitLoop("left", "right"), visitForOfStatement: visitLoop("left", "right"), visitFunction(path) { this.visit(path.get("params")); const old = { replaceReturn }; replaceReturn = false; this.visit(path.get("body")); replaceReturn = old.replaceReturn; return false; }, visitReturnStatement(path) { if (replaceReturn) { assert(!inLoop); // could use labeled continue if this ever fires const { argument, comments } = path.node; if (argument === null) { const s = b.continueStatement(); s.comments = comments; path.replace(s); } else { const s = b.expressionStatement(argument); s.comments = comments; path.replace(s, b.continueStatement()); } return false; } this.traverse(path); }, visitWhileStatement: visitLoop("test"), }); if (changed) { console.log("Writing", file); fs.writeFileSync(file, recast.print(ast).code, { encoding: "utf8" }); } } Signed-off-by: Anders Kaseorg <anders@zulipchat.com>
2020-02-06 06:19:47 +01:00
}
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', 'te');
expected = [
"te",
"sender:ted@zulip.com",
"pm-with:ted@zulip.com",
"group-pm-with:ted@zulip.com",
"stream:office topic:team",
"stream:office topic:test",
];
assert.deepEqual(suggestions.strings, expected);
function describe(q) {
return suggestions.lookup_table.get(q).description;
}
assert.equal(describe('te'), "Search for te");
assert.equal(describe('stream:office topic:team'), "Stream office &gt; team");
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', 'topic:staplers stream:office');
expected = [
'topic:staplers stream:office',
'topic:staplers',
];
assert.deepEqual(suggestions.strings, expected);
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', 'stream:devel topic:');
expected = [
'stream:devel topic:',
'stream:devel topic:REXX',
'stream:devel',
];
assert.deepEqual(suggestions.strings, expected);
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', 'stream:devel -topic:');
expected = [
'stream:devel -topic:',
'stream:devel -topic:REXX',
'stream:devel',
];
assert.deepEqual(suggestions.strings, expected);
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', '-topic:te');
expected = [
'-topic:te',
'stream:office -topic:team',
'stream:office -topic:test',
];
assert.deepEqual(suggestions.strings, expected);
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', 'is:alerted stream:devel is:starred topic:');
expected = [
'is:alerted stream:devel is:starred topic:',
'is:alerted stream:devel is:starred topic:REXX',
'is:alerted stream:devel is:starred',
'is:alerted stream:devel',
'is:alerted',
];
assert.deepEqual(suggestions.strings, expected);
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', 'is:private stream:devel topic:');
expected = [
'is:private stream:devel topic:',
'is:private stream:devel',
'is:private',
];
assert.deepEqual(suggestions.strings, expected);
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', 'topic:REXX stream:devel topic:');
expected = [
'topic:REXX stream:devel topic:',
'topic:REXX stream:devel',
'topic:REXX',
];
assert.deepEqual(suggestions.strings, expected);
});
run_test('whitespace_glitch', () => {
const query = 'stream:office '; // note trailing space
global.stream_data.subscribed_streams = function () {
return ['office'];
};
global.narrow_state.stream = function () {
return;
};
stream_topic_history.reset();
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
const suggestions = get_suggestions('', query);
const expected = [
"stream:office",
];
assert.deepEqual(suggestions.strings, expected);
});
run_test('stream_completion', () => {
global.stream_data.subscribed_streams = function () {
return ['office', 'dev help'];
};
global.narrow_state.stream = function () {
return;
};
stream_topic_history.reset();
let query = 'stream:of';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
let suggestions = get_suggestions('', query);
let expected = [
"stream:of",
"stream:office",
];
assert.deepEqual(suggestions.strings, expected);
query = '-stream:of';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"-stream:of",
"-stream:office",
];
assert.deepEqual(suggestions.strings, expected);
query = 'hel';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"hel",
"stream:dev+help",
];
assert.deepEqual(suggestions.strings, expected);
});
run_test('people_suggestions', () => {
let query = 'te';
global.stream_data.subscribed_streams = function () {
return [];
};
global.narrow_state.stream = function () {
return;
};
const ted = {
email: 'ted@zulip.com',
user_id: 201,
full_name: 'Ted Smith',
};
const bob = {
email: 'bob@zulip.com',
user_id: 202,
full_name: 'Bob Térry',
};
const alice = {
email: 'alice@zulip.com',
user_id: 203,
full_name: 'Alice Ignore',
};
people.add_active_user(ted);
people.add_active_user(bob);
people.add_active_user(alice);
stream_topic_history.reset();
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
let suggestions = get_suggestions('', query);
let expected = [
"te",
"sender:bob@zulip.com",
"sender:ted@zulip.com",
"pm-with:bob@zulip.com", // bob térry
"pm-with:ted@zulip.com",
"group-pm-with:bob@zulip.com",
"group-pm-with:ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
function describe(q) {
return suggestions.lookup_table.get(q).description;
}
assert.equal(describe('pm-with:ted@zulip.com'),
"Private messages with <strong>Te</strong>d Smith &lt;<strong>te</strong>d@zulip.com&gt;");
assert.equal(describe('sender:ted@zulip.com'),
"Sent by <strong>Te</strong>d Smith &lt;<strong>te</strong>d@zulip.com&gt;");
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', 'Ted '); // note space
expected = [
"Ted",
"sender:ted@zulip.com",
"pm-with:ted@zulip.com",
"group-pm-with:ted@zulip.com",
];
assert.deepEqual(suggestions.strings, expected);
query = 'sender:ted sm';
expected = [
'sender:ted+sm',
'sender:ted@zulip.com',
];
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
assert.deepEqual(suggestions.strings, expected);
query = 'sender:ted@zulip.com new';
expected = [
'sender:ted@zulip.com new',
'sender:ted@zulip.com',
];
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
assert.deepEqual(suggestions.strings, expected);
query = 'sender:ted@tulip.com new';
expected = [
'sender:ted@tulip.com+new',
];
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
assert.deepEqual(suggestions.strings, expected);
});
run_test('operator_suggestions', () => {
// Completed operator should return nothing
let query = 'stream:';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
let suggestions = get_suggestions('', query);
let expected = [
'stream:',
];
assert.deepEqual(suggestions.strings, expected);
query = 'st';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
'st',
'streams:public',
'is:starred',
'stream:',
];
assert.deepEqual(suggestions.strings, expected);
query = 'group-';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
'group-',
'group-pm-with:',
];
assert.deepEqual(suggestions.strings, expected);
query = '-s';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
'-s',
'-streams:public',
'-sender:myself@zulip.com',
'-stream:',
'-sender:',
];
assert.deepEqual(suggestions.strings, expected);
query = 'stream:Denmark is:alerted -f';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
'stream:Denmark is:alerted -f',
'stream:Denmark is:alerted -from:myself@zulip.com',
'stream:Denmark is:alerted -from:',
'stream:Denmark is:alerted',
'stream:Denmark',
];
assert.deepEqual(suggestions.strings, expected);
});
run_test('queries_with_spaces', () => {
global.stream_data.subscribed_streams = function () {
return ['office', 'dev help'];
};
global.narrow_state.stream = function () {
return;
};
stream_topic_history.reset();
// test allowing spaces with quotes surrounding operand
let query = 'stream:"dev he"';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
let suggestions = get_suggestions('', query);
let expected = [
"stream:dev+he",
"stream:dev+help",
];
assert.deepEqual(suggestions.strings, expected);
// test mismatched quote
query = 'stream:"dev h';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"stream:dev+h",
"stream:dev+help",
];
assert.deepEqual(suggestions.strings, expected);
// test extra space after operator still works
query = 'stream: offi';
search suggestions: De-duplicate legacy search codepath. This merges the `exports.get_search_result_legacy` and `exports.get_search_result` function. The key differences between the two code paths are as follows: * We only want to generate suggestions for the queries which the user is typing or can edit. For the legacy version, suggestions are displayed for the entire search string in the searchbox. (`all_operators`) For the pills enabled version, suggestions are displayed only for the input which hasn't been converted to pills. (`query_operators`) `all_operators` = `base_query_operators` + " " + `query_operators`. trim is added at the end just to handle the legacy case where we pass the `base_query` as ''. * It is not possible to detect whether the user wants to continue typing in the legacy version. However if the the searchbox is still focused even after pill creation we can assume the user still wants to continue typing. To handle this we push an empty term as the `last` operator. This is possible since the previous queries have been completely entered as evident from it's generated pill. * When using the legacy version, `search_operators` are the same as `all_operators`, as mentioned in point 1. In the pills enabled version we perform most of the computations from the `query_operators`, but we do require all `all_operators`, only for filtering the last query's suggestion. * And there is just one block unique to the legacy search system. More details are mentioned in the comments of that block. We also refactor both the search suggestions node tests, mainly to make them similar and easier to detect differences when we switch over to the new version.
2020-06-01 15:00:42 +02:00
suggestions = get_suggestions('', query);
expected = [
"stream:offi",
"stream:office",
];
assert.deepEqual(suggestions.strings, expected);
});