Would you have a rebuild handy for Trusty? Like app developers should, I
try to stick to Ubuntu LTS for development tasks.

On Mon, Feb 16, 2015 at 10:36 AM, Federico Gimenez <email address hidden>
wrote:

> Some inline comments, thanks!
>
>
> Diff comments:
>
> > === modified file 'README'
> > --- README    2013-09-13 08:20:52 +0000
> > +++ README    2015-02-14 05:29:25 +0000
> > @@ -1,9 +1,106 @@
> > -This package provides a set of default web application versions for
> popular web services.
> > +Core Webapps
> > +============
> > +
> > +This package provides a set of default web application versions for
> popular
> > +web services.
> >
> >  The debian package itself serves 2 purposes:
> > -- generate click packages in a clean environment, that can be
> integrated with the rest of our toolchain
> > -- package a set of icons for the webapps, while click doesn't support
> icons directly (temporary)
> > -
> > -To generate the click packages directly, ie without the debian package:
> > -
> > -./build.sh
> > +- generate click packages in a clean environment, that can be
> integrated with
> > +the rest of our toolchain
> > +- package a set of icons for the webapps, while click doesn't support
> icons
> > +directly (temporary)
> > +
> > +Building the click packages
> > +===========================
> > +
> > +To generate the click packages directly, ie without the debian package::
> > +
> > +  $ ./build.sh
> > +
> > +Running the user acceptance tests
> > +=================================
> > +
> > +Run tests locally
> > +-----------------
> > +
> > +The tests need an updated selenium package and the oxide webdriver::
> > +
> > +    $ sudo add-apt-repository ppa:canonical-platform-qa/selenium
> > +    $ sudo apt-get update
> > +    $ sudo apt-get install python3-selenium oxideqt-chromedriver
> > +
> > +To run the user acceptance tests from the source branch::
> > +
> > +    $ cd webapp-amazon/tests/autopilot
> > +    $ autopilot3 run --config from_source amazon_webapp
> > +
> > +Run tests in the phone
> > +----------------------
> > +
> > +First flash the phone with developer mode enabled::
> > +
> > +    $ ubuntu-device-flash --developer-mode --password 0000 \
> > +      --channel="ubuntu-touch/devel-proposed" --wipe
> > +
> > +When the phone is flashed, complete the welcome wizard.
> > +
> > +To run the user accceptance tests into an Ubuntu phone with a read-only
> image::
> > +
> > +    $ adt-run --click-source webapp-amazon/ \
> > +      --click
> build/com.ubuntu.developer.webapps.webapp-amazon_*_all.click \
> > +      -o /tmp/adt-webapps-tests \
> > +      --setup-commands \
> > +      'mount -o remount,rw /' \
> > +      --setup-commands \
> > +      'apt-add-repository -y ppa:canonical-platform-qa/selenium' \
> > +      --setup-commands \
> > +      'apt-get --no-list-cleanup update -o
> Dir::Etc::SourceList=/dev/null' \
> > +      --setup-commands \
> > +      'sync; sleep 2; mount -o remount,ro /' \
> > +      --- ssh -s adb -- -p 0000
> > +
> > +Run tests in a virtual machine
> > +------------------------------
> > +
> > +The user acceptance tests also can be run against a built and installed
> click
> > +package, under qemu. To do so, install autopkgtest and qemu::
> > +
> > +    $ sudo apt-get install autopkgtest qemu
> > +
> > +After installing autopkgtest and qemu, you need to build an image for
> qemu
> > +to use. We use vivid here, as building an image to closely resemble the
> actual
> > +stable phone images is quite difficult. When this is easier in the
> future, we
> > +will switch to using stable phone images for this. We output the image
> to ~/
> > +rather than the current directory, so it will be in a safer place to
> avoid
> > +rebuilding images every time. You can store it in any directory you
> wish.
> > +
> > +::
> > +
> > +    $ adt-buildvm-ubuntu-cloud -r vivid -a amd64 -o ~/
> > +
> > +Then the tests may be run using adt-run with the qemu virtualization
> host.
> > +The output directory option here can be wherever you like, and is where
> the
> > +test artifacts will be placed. The ordering of the arguments to adt-run
> is
> > +important so try to keep them in this order.
> > +
> > +::
> > +
> > +    $ adt-run --click-source webapp-amazon/ \
>
> At this point I'm getting:
>
> adt-run [14:56:55]: ERROR: unexpected error: "click info --user ubuntu
> build/com.ubuntu.developer.webapps.webapp-amazon_*_all.click" failed with
> stderr "click_query_error-quark: Failed to canonicalize
> build/com.ubuntu.developer.webapps.webapp-amazon_*_all.click: No such file
> or directory (0)
>
> Changing '*' with the version in the manifest file doesn't seem to help.
>
> > +      --click
> build/com.ubuntu.developer.webapps.webapp-amazon_*_all.click \
> > +      -o /tmp/adt-webapps-tests \
> > +      --setup-commands ubuntu-touch-session \
> > +      --setup-commands \
> > +      'add-apt-repository -y ppa:canonical-platform-qa/selenium' \
> > +      -U --apt-pocket proposed \
> > +      --- qemu ~/adt-vivid-amd64-cloud.img
> > +
> > +See the detailed results
> > +------------------------
> > +
> > +To examine the test results, which are in subunit format, additional
> tools are
> > +required::
> > +
> > +    $ sudo add-apt-repository ppa:thomir/trv
> > +    $ sudo apt-get update
> > +    $ sudo apt-get install trv
> > +    $ trv /tmp/adt-webapps-tests/artifacts/autopilot.subunit
> >
> > === modified file 'webapp-amazon/manifest.json'
> > --- webapp-amazon/manifest.json       2014-09-19 16:29:06 +0000
> > +++ webapp-amazon/manifest.json       2015-02-14 05:29:25 +0000
> > @@ -11,6 +11,19 @@
> >      "maintainer": "Webapps Team <email address hidden>",
> >      "name": "com.ubuntu.developer.webapps.webapp-amazon",
> >      "title": "webapp-amazon",
> > -    "version": "1.0.10"
> > +    "version": "1.0.10",
> > +    "x-test": {
> > +        "autopilot": {
> > +            "autopilot_module": "amazon_webapp",
> > +            "depends": [
> > +                "click",
> > +                "dpkg-dev",
> > +                "qtdeclarative5-ubuntu-ui-toolkit-plugin",
> > +                "ubuntu-sdk-libs",
> > +                "oxideqt-chromedriver",
> > +                "python3-selenium",
> > +                "webapp-container"
> > +            ]
> > +        }
> > +    }
> >  }
> > -
> >
> > === added directory 'webapp-amazon/tests'
> > === added directory 'webapp-amazon/tests/autopilot'
> > === added directory 'webapp-amazon/tests/autopilot/amazon_webapp'
> > === added file 'webapp-amazon/tests/autopilot/amazon_webapp/__init__.py'
> > --- webapp-amazon/tests/autopilot/amazon_webapp/__init__.py   1970-01-01
> 00:00:00 +0000
> > +++ webapp-amazon/tests/autopilot/amazon_webapp/__init__.py   2015-02-14
> 05:29:25 +0000
> > @@ -0,0 +1,138 @@
> > +# -*- Mode: Python; coding: utf-8; indent-tabs-mode: nil; tab-width: 4
> -*-
> > +
> > +#
> > +# Ubuntu Core Webapps
> > +# Copyright (C) 2015 Canonical
> > +#
> > +# This program is free software: you can redistribute it and/or modify
> > +# it under the terms of the GNU General Public License as published by
> > +# the Free Software Foundation, either version 3 of the License, or
> > +# (at your option) any later version.
> > +#
> > +# This program is distributed in the hope that it will be useful,
> > +# but WITHOUT ANY WARRANTY; without even the implied warranty of
> > +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> > +# GNU General Public License for more details.
> > +#
> > +# You should have received a copy of the GNU General Public License
> > +# along with this program.  If not, see <http://www.gnu.org/licenses/>.
> > +#
> > +
> > +import contextlib
> > +import logging
> > +
> > +import autopilot.logging
> > +import ubuntuuitoolkit
> > +from autopilot import platform
> > +
> > +logger = logging.getLogger(__name__)
> > +
> > +
> > +class PageObject():
> > +
> > +    """Base page object."""
> > +
> > +    def __init__(self, proxy, driver):
> > +        """Initialize the page object.
> > +
> > +        :param proxy: The autopilot application proxy.
> > +        :param driver: The webdriver used to introspect and control the
> web
> > +            container.
> > +
> > +        """
> > +        super().__init__()
> > +        self.proxy = proxy
> > +        self.driver = driver
> > +
> > +    def get_container_proxy(self):
> > +        return self.proxy.select_single('WebappContainerWebview')
> > +
> > +    def find_element_by_css_in_desktop_or_mobile(
> > +            self, desktop_css_selector, mobile_css_selector):
> > +        return self.driver.find_element_by_css_selector(
> > +            '{}, {}'.format(desktop_css_selector, mobile_css_selector))
> > +
> > +    @contextlib.contextmanager
> > +    def focused_type(self, input_web_element):
> > +        """Type into an input web element.
> > +
> > +        This context manager takes care of making sure a particular
> > +        *input_target* UI control is selected before any text is
> entered.
> > +
> > +        """
> > +
> > +        pointer = ubuntuuitoolkit.get_pointing_device()
> > +        pointer.move(*self.get_center_point(input_web_element))
> > +        pointer.click()
> > +
> > +        yield ubuntuuitoolkit.get_keyboard()
> > +
> > +    def get_center_point(self, web_element):
> > +        web_element_center_x = web_element.location.get('x') + (
> > +            web_element.size.get('width') // 2)
> > +        web_element_center_y = web_element.location.get('y') + (
> > +            web_element.size.get('height') // 2)
> > +
> > +        container_x, container_y, _, _ =
> self.get_container_proxy().globalRect
> > +
> > +        return (
> > +            container_x + web_element_center_x,
> > +            container_y + web_element_center_y
> > +        )
> > +
> > +
> > +class AmazonMainPage(PageObject):
> > +
> > +    """Page object for the main page of the Amazon website."""
> > +
> > +    @autopilot.logging.log_action(logger.info)
> > +    def search(self, query):
> > +        search_bar = self._get_search_bar()
> > +        self._enter_search_query(search_bar, query)
> > +        go_button = self._get_go_button(search_bar)
> > +        go_button.click()
> > +        return AmazonResultsPage(self.proxy, self.driver)
> > +
> > +    def _get_search_bar(self):
> > +        desktop_css_selector = '#nav-searchbar'
> > +        mobile_css_selector = '.nav-searchbar'
> > +        return self.find_element_by_css_in_desktop_or_mobile(
> > +            desktop_css_selector, mobile_css_selector)
> > +
> > +    def _enter_search_query(self, search_bar, query):
> > +        desktop_css_selector = '#twotabsearchtextbox'
> > +        mobile_css_selector = 'input.nav-input[type="text"]'
> > +        search_text_field =
> self.find_element_by_css_in_desktop_or_mobile(
> > +            desktop_css_selector, mobile_css_selector)
> > +        if platform.model() == 'Desktop':
> > +            with self.focused_type(search_text_field) as keyboard:
> > +                keyboard.type(query)
> > +        else:
> > +            # The location returned by the driver in the phone is not
> accurate.
> > +            # TODO report a bug.
> > +            # XXX We can't send keys because of bug
> http://pad.lv/1421423
> > +            # Remove the workaround once that bug is fixed.
> > +            # --elopio - 2015-02-12
> > +            # search_text_field.send_keys('Test')
> > +            self.driver.execute_script(
> > +                "document.getElementsByClassName("
> > +                "'nav-searchbar')[0].getElementsByTagName("
> > +                "'input')[1].setAttribute('value', '{}')".format(query)
> > +            )
> > +
> > +    def _get_go_button(self, search_bar):
> > +        desktop_css_selector = '.nav-submit-input'
> > +        mobile_css_selector = '.nav-input'
> > +        return self.find_element_by_css_in_desktop_or_mobile(
> > +            desktop_css_selector, mobile_css_selector)
> > +
> > +
> > +class AmazonResultsPage(PageObject):
> > +
> > +    """Page object for the main page of the Amazon Website."""
> > +
> > +    def get_search_results_count_label(self):
> > +        desktop_css_selector = '#s-result-count'
> > +        mobile_css_selector = '#results span'
> > +        return self.find_element_by_css_in_desktop_or_mobile(
> > +            desktop_css_selector, mobile_css_selector).text
> >
> > === added file
> 'webapp-amazon/tests/autopilot/amazon_webapp/test_amazon_webapp.py'
> > --- webapp-amazon/tests/autopilot/amazon_webapp/test_amazon_webapp.py
> 1970-01-01 00:00:00 +0000
> > +++ webapp-amazon/tests/autopilot/amazon_webapp/test_amazon_webapp.py
> 2015-02-14 05:29:25 +0000
> > @@ -0,0 +1,146 @@
> > +# -*- Mode: Python; coding: utf-8; indent-tabs-mode: nil; tab-width: 4
> -*-
> > +
> > +#
> > +# Ubuntu Core Webapps
> > +# Copyright (C) 2015 Canonical
> > +#
> > +# This program is free software: you can redistribute it and/or modify
> > +# it under the terms of the GNU General Public License as published by
> > +# the Free Software Foundation, either version 3 of the License, or
> > +# (at your option) any later version.
> > +#
> > +# This program is distributed in the hope that it will be useful,
> > +# but WITHOUT ANY WARRANTY; without even the implied warranty of
> > +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> > +# GNU General Public License for more details.
> > +#
> > +# You should have received a copy of the GNU General Public License
> > +# along with this program.  If not, see <http://www.gnu.org/licenses/>.
> > +#
> > +
> > +import os
> > +import shutil
> > +import subprocess
> > +
> > +from autopilot import (
> > +    get_test_configuration,
> > +    testcase
> > +)
> > +from selenium import webdriver
> > +from selenium.webdriver.chrome import options
> > +from ubuntuuitoolkit import fixture_setup
> > +
> > +import amazon_webapp
> > +
> > +
> > +DESKTOP_FILE_NAME = 'webapp-amazon.desktop'
> > +
> > +DEFAULT_IMPLICIT_WAIT = 10  # seconds
> > +
> > +
> > +def get_source_desktop_file_path():
> > +    return os.path.join(
> > +        os.path.dirname(amazon_webapp.__file__), '..', '..', '..',
> > +        DESKTOP_FILE_NAME)
> > +
> > +
> > +def get_local_desktop_file_directory():
> > +    return os.path.join(
> > +        os.environ.get('HOME'), '.local', 'share', 'applications')
> > +
> > +
> > +def get_chrome_driver_path():
> > +    env = os.environ.copy()
> > +    adt_artifacts_path = os.environ.get('ADT_ARTIFACTS', None)
> > +    if adt_artifacts_path:
>
> The deps path is reused later, it could be determined in an extracted
> method.
>
> > +        # XXX When running with adt-run, this is needed if the image is
> > +        # read-only so we can find the chromium driver path.
> > +        env['PERL5LIB'] = os.path.join(
> > +            os.path.dirname(adt_artifacts_path),
> > +            'deps', 'usr', 'share', 'perl5')
> > +
> > +    architecture = subprocess.check_output(
> > +        ['dpkg-architecture', '-qDEB_HOST_MULTIARCH'],
> > +        env=env, universal_newlines=True).strip()
> > +
> > +    chrome_driver_path = '/usr/lib/{}/oxide-qt/chromedriver'.format(
> > +        architecture)
> > +
> > +    if adt_artifacts_path and not os.path.exists(chrome_driver_path):
> > +        # Try the read-only install path.
> > +        chrome_driver_path = os.path.join(
> > +            os.path.dirname(adt_artifacts_path), 'deps',
> > +            chrome_driver_path.lstrip('/'))
> > +
> > +    return chrome_driver_path
> > +
> > +
> > +class AmazonWebappTestCase(testcase.AutopilotTestCase):
> > +
> > +    DEFAULT_WEBVIEW_INSPECTOR_IP = '127.0.0.1'
> > +    DEFAULT_WEBVIEW_INSPECTOR_PORT = 9221
> > +
> > +    def setUp(self):
> > +        super().setUp()
> > +        self.setup_webdriver_environment()
> > +
> > +    def setup_webdriver_environment(self):
> > +        self.useFixture(fixture_setup.InitctlEnvironmentVariable(
> > +            global_=True,
> > +
> UBUNTU_WEBVIEW_DEVTOOLS_HOST=self.DEFAULT_WEBVIEW_INSPECTOR_IP,
> > +            UBUNTU_WEBVIEW_DEVTOOLS_PORT=str(
> > +                self.DEFAULT_WEBVIEW_INSPECTOR_PORT)
> > +        ))
> > +
> > +    def launch_application(self):
> > +        if (get_test_configuration().get('from_source', None) and
> > +                os.path.exists(get_source_desktop_file_path())):
> > +            return self.launch_application_from_source()
> > +        else:
> > +            return self.launch_installed_application()
> > +
> > +    def launch_application_from_source(self):
> > +        test_desktop_file_path = os.path.join(
> > +            get_local_desktop_file_directory(), 'webapps-test.desktop')
> > +        shutil.copy(get_source_desktop_file_path(),
> test_desktop_file_path)
> > +        self.addCleanup(os.remove, test_desktop_file_path)
> > +
> > +        return self.launch_upstart_application('webapps-test')
> > +
> > +    def launch_installed_application(self):
> > +        return self.launch_click_package(
> > +            'com.ubuntu.developer.webapps.webapp-amazon')
> > +
> > +    def get_webdriver(self):
> > +        driver_options = options.Options()
> > +        driver_options.binary_location = ''
> > +        driver_options.debugger_address = '{}:{}'.format(
> > +            self.DEFAULT_WEBVIEW_INSPECTOR_IP,
> > +            self.DEFAULT_WEBVIEW_INSPECTOR_PORT)
> > +        driver = webdriver.Chrome(
> > +            executable_path=get_chrome_driver_path(),
> > +            chrome_options=driver_options
> > +        )
> > +
> > +        self.assertIsNot(driver, None)
> > +
> > +        self.addCleanup(self.close_webdriver, driver)
> > +
> > +        driver.implicitly_wait(DEFAULT_IMPLICIT_WAIT)
> > +        return driver
> > +
> > +    def close_webdriver(self, driver):
>
> I wasn't able to make this work at [1], the webapp-container remains
> opened. Maybe the webdriver can only close processes that has itself
> spawned, not others? You can always count on good old 'kill -9' :)
>
> [1]
> https://code.launchpad.net/~fgimenez/ubuntu-sanity-tests/oxide-proof-of-concept/+merge/249388
>
> > +        driver.close()
> > +        driver.quit()
> > +
> > +    def test_search_in_amazon_must_show_results(self):
> > +        proxy = self.launch_application()
> > +        driver = self.get_webdriver()
> > +
> > +        amazon = amazon_webapp.AmazonMainPage(proxy, driver)
> > +        results_page = amazon.search('Test')
> > +
> > +        results = results_page.get_search_results_count_label()
> > +        self.assertTrue(
> > +            results.endswith('results for "Test"') or
> > +            results.endswith('Results'))
> >
>
>
> --
>
> https://code.launchpad.net/~canonical-platform-qa/webapps-core/x-test/+merge/249580
> You are subscribed to branch lp:webapps-core.
>