Added sound mode support by starkillerOG · Pull Request #43 · ol-iver/denonavr

starkillerOG · 2018-03-05T23:01:19Z

new functions to use:

denonavr.sound_mode (gives the current matched sound mode (after an denonavr.update))
denonavr.sound_mode_raw (gives the current sound mode as received from the AVR (after an denonavr.update))
denonavr.sound_mode_list (gives a list of supported sound mode commands)
denonavr.sound_mode_dict (gives the full dictionarry containing all supported commands including their matching values)
denonavr.set_sound_mode(VALUE) (change the sound mode of the AVR using one of the VALUE's of the denonavr.sound_mode_list)
denonavr.set_sound_mode_dict (change the matching sound mode dict, this will throw an error if the wrong syntax is used with an example of a correct syntax)

new functions to use: - denonavr.sound_mode (gives the current matched sound mode (after an denonavr.update)) - denonavr.sound_mode_raw (gives the current sound mode as received from the AVR (after an denonavr.update)) - denonavr.sound_mode_list (gives a list of supported sound mode commands) - denonavr.sound_mode_dict (gives the full dictionarry containing all supported commands including their matching values) - denonavr.set_sound_mode(VALUE) (change the sound mode of the AVR using one of the VALUE's of the denonavr.sound_mode_list) - denonavr.set_sound_mode_dict (change the matching sound mode dict, this will throw an error if the wrong syntax is used with an example of a correct syntax)

houndci-bot · 2018-03-05T23:01:26Z

+                    return sound_mode
+        except ValueError:
+            pass
+        _LOGGER.warning("Not able to match sound mode, returning raw sound mode.")


line too long (82 > 79 characters)

houndci-bot · 2018-03-05T23:01:26Z

+        else:
+            _LOGGER.error(Error_msg)
+            return False
+


blank line contains whitespace

houndci-bot · 2018-03-05T23:01:26Z

+            return False
+
+    def set_sound_mode_dict(self, sound_mode_dict):
+        Error_msg = "Syntax of sound mode dictionary not valid, use: OrderedDict([('COMMAND', ['VALUE1','VALUE2'])])"


line too long (117 > 79 characters)

houndci-bot · 2018-03-05T23:01:26Z

        # Set all tags to be evaluated
        relevant_tags = {"Power": None, "InputFuncSelect": None, "Mute": None,
-                         "MasterVolume": None}
+                         "MasterVolume": None, "selectSurround": None, "SurrMode": None}


line too long (88 > 79 characters)

fixed houndci-bot typo's

houndci-bot · 2018-03-05T23:07:26Z

+
+    def set_sound_mode_dict(self, sound_mode_dict):
+        Error_msg = "Syntax of sound mode dictionary not valid, "
+                    "use: OrderedDict([('COMMAND', ['VALUE1','VALUE2'])])"


IndentationError: unexpected indent
unexpected indentation

houndci-bot · 2018-03-05T23:09:08Z

+
+    def set_sound_mode_dict(self, sound_mode_dict):
+        Error_msg = ("Syntax of sound mode dictionary not valid, "
+                    "use: OrderedDict([('COMMAND', ['VALUE1','VALUE2'])])")


continuation line under-indented for visual indent

starkillerOG · 2018-03-05T23:20:59Z

@scarface-4711 could you please merge this PR?
It would be greath if you could then push to the next version of denonavr, such that I can use the new sound mode implementation in my PR for HomeAssistant.

starkillerOG · 2018-03-18T17:16:24Z

@scarface-4711 could you please merge this PR?
I am waiting until this gets merged, before I can continue my work with the sound mode support in HomeAssistant and I would really like to finisch this.
It would be greath if you could find the time to accept this PR.

ol-iver · 2018-03-25T14:06:26Z

@starkillerOG , was on vacation and did have lots of work before. I was not able to test your changes until now.

Setting sound modes using the values from sound_mode_list is working fine. The modes for setting seem to be the same over several receiver models.

The mapping of sound modes does not work for my receiver. Every time I run the update method there is a Not able to match sound mode, returning raw sound mode. warning, because the mapping of my receiver is totally different than yours. When you look at other XML files from the tests/xml subdirectory of this repository, many other models will have the same issue.
Thus, please think about a more generic mapping (which would be a huge mapping table, as I wrote in your issue thread) or remove it and use the raw values instead.

ol-iver · 2018-03-25T14:11:28Z

+                self._sound_mode = self.match_sound_mode(sound_mode_raw)
+                self._sound_mode_raw = sound_mode_raw
+                if "selectSurround" in relevant_tags.keys():
+                    relevant_tags.pop("selectSurround", None)


Could you pop both keys from the list, when the first one was found?
They seem to cover the same information. At the moment there might be a second HTTP request, which is not needed, because the STATUS_URL is called first and there is a second call to MAINZONE_URL if some tags are not found yet.

I am poping both keys when the first one is found.
when one of the keys is found elif (child.tag == "selectSurround" or child.tag == "SurrMode"): , both keys are poped using relevant_tags.pop("selectSurround", None) and relevant_tags.pop("SurrMode", None).
The if "selectSurround" in relevant_tags.keys(): and if "SurrMode" in relevant_tags.keys(): will in principle always return true, however I put those in to prevent the occasion that later when the code is changed there is no problem with trying to remove a key that does not exist. In the case that if "selectSurround" in relevant_tags.keys(): or if "SurrMode" in relevant_tags.keys(): returns false, the key is not in the list anyway so their is no need to pop it.

Therefore I think this part of the code is correct, and a second HTTP request is not possible unless none of the two keys ("selectSurround" and "SurrMode") are found.

ah ok, I was confused by the if statement.
Btw. you don't need the if statements here. If you pop with an default value, there won't be an exception even if the key is not in the list

starkillerOG · 2018-03-26T13:16:04Z

@scarface-4711 The mapping is indeed quite difficult.
I have searched in all xml files i could find from my receiver but I cannot find the exact spelling of the returned sound mode anywhere. Therefore it seems not possible to get the mapping list from the receiver, I have tryed.

I am happy to include more mapping values and make the mapping structure bigger (just fill in the list), This will make this list pretty big indeed. The problem is that I do not know the values that are returnd. For my receiver I just asked for all the sound modes I use and then put the returned values in the mapping list. However if I compare my retrun values with list such as:
Marantz Command list
Marantz Command list 2
They do not match.
For instance the return value for 'MCH STEREO' is for me 'MULTI CH STEREO', but that is not in any of the above documentation.
Therefore I would need the return values for diffrent types of receivers (if you know them I can included them in the list).

Note however that I implemented the option to change the mapping list:
denonavr.set_sound_mode_dict(DICT_TO_USE)
I also made an optional configuration parameter in my HomeAssistant PR so people can change the mapping list inside HomeAssistant to adapt it to match their type of receiver.

starkillerOG · 2018-03-26T13:20:23Z

@scarface-4711 The raw values can always be obtained using:
denonavr.sound_mode_raw()
(after an update)
I do need the matched values for HomeAssistant, because the input list wil otherwise not show the correct sound mode.
What I will do is move the warning from the denonavr.update() to the denonavr.sound_mode(), such that the warning is not shown if you do not use the matched values but only the raw values.

Moved the warning if the sound mode can not be matched from the update function to when you call the matched sound mode.

Added all the sound mode values that were present in denonavr/tests/xml/

houndci-bot · 2018-03-26T15:20:56Z

                        "Flickr": "FLICKR", "Favorites": "FAVORITES"}

+SOUND_MODE_MAPPING = OrderedDict([('MUSIC', ['PLII MUSIC']),
+                                  ('MOVIE', ['PLII MOVIE','PLII CINEMA',


missing whitespace after ','

starkillerOG · 2018-03-26T15:27:28Z

@scarface-4711 I added all the sound mode values that were present in denonavr/tests/xml/.
Could you provide me with any aditional sound mode values for receivers that you know/own?

I suggest we merge this PR, and over time add more mapping values to the list when people owning a diffrent model open an issue for missing mapping values. In the mean time those people can customly define their mapping values with denonavr.set_sound_mode_dict(DICT_TO_USE), or when using HomeAssistant define it in the options of HomeAssistant.

ol-iver

Please check my remarks. I'll check additional sound mode of my receiver.

ol-iver · 2018-04-02T21:21:40Z

                relevant_tags.pop(child.tag, None)
+            elif (child.tag == "selectSurround" or child.tag == "SurrMode"):
+                sound_mode_raw = child[0].text.rstrip()
+                self._sound_mode = self.match_sound_mode(sound_mode_raw)


Could you move the match_sound_mode to the sound_mode property?
You won't need the self. _sound_mode_match_warning attribute then.

just did it.

ol-iver · 2018-04-02T21:24:11Z

+                self._sound_mode = self.match_sound_mode(sound_mode_raw)
+                self._sound_mode_raw = sound_mode_raw
+                if "selectSurround" in relevant_tags.keys():
+                    relevant_tags.pop("selectSurround", None)


ah ok, I was confused by the if statement.
Btw. you don't need the if statements here. If you pop with an default value, there won't be an exception even if the key is not in the list

ol-iver · 2018-04-02T21:44:19Z

+
+    def match_sound_mode(self, sound_mode_raw):
+        try:
+            mode_list = list(self._sound_mode_dict.values())


When I look at these lines, I get the impression you created the SOUND_MODE_MAPPING and self._sound_mode_dict the wrong way and mixed key and values.
You are using it to get the sound modes from the raw values. For me this is a normal dictionary with each sound_mode_raw as key and its sound_mode as value. That would simplify this part here a lot.
sound_mode_list would be a set set(self._sound_mode_dict.values()) then.

In principle you would be right, and it would be nicer to do it that way if each sound_mode_raw would correspond to exactly 1 matched sound mode.
Unfortunately this is not the case, and multiple sound_mode_raw's can be matched to the same matched sound_mode.

I did think about this before I set it up, and in my opinion it is more elegant to do it like I did it because now I have a single matched sound_mode as key, and then all corresponding raw_sound_modes as list in the value corresponding to that key. In this way I do not need to repeat the matched sound_mode as the value for several raw_sound_mode keys (if you would do it the other way around).

Therefore my method seemed more compact and in my opinion more elegant for this situation.

In principle I chose for keeping the self._sound_mode_dict as compact and most eligant as I could make it, and having a bit more complicated code in the match_sound_mode function.

I did this because users of this library and Home Assistant will probably want to tweak the self._sound_mode_dict to have the exact mapping values for their model of receiver, if it is not yet incorperated in the default mapping dict.
However those users will in principle never have to deal with the code of the match_sound_mode function, so it is no problem if that gets a bit more complicated.

It does not look too efficient to replace a dictionary key access by a nested loop across all values of a dictionary during run time.

Maintaining own mapping rules is indeed easier using your way of modelling.

Thus, I would use a dictionary with key: sound_mode_raw and value: sound_mode during run time as I suggested.
This dictionary could be created by the set_sound_mode_dict method you already created from a dictionary with the structure you suggested (key: sound_mode, value: list of sound_mode_raw)

@scarface-4711 smart, it is indead more efficient to use a simple dict key acces during runtime, I have implemented this change in the code.

moved match_sound_mode to sound_mode property. Therefore I could get rid of self._sound_mode and self. _sound_mode_match_warning.

starkillerOG · 2018-04-02T22:32:09Z

@scarface-4711 shall I remove the if "selectSurround" in relevant_tags.keys(): check's, or just keep them just to be sure?

ol-iver · 2018-04-02T22:34:20Z

The sound modes are different at my receiver
When I switch to MOVIE, MUSIC and GAME the sound_mode will be the one I have chosen the last time on my remote control for this category.
"Dolby D + Dolby Surround", "Dolby Digital", "Multi Ch Stereo", "Stereo", "Virtual" could be assigned from each of those three categories.

There are other sound modes which are specific to one category:
"Dolby D +Neo:X C": MOVIE
"Dolby D +Neo:X M": MUSIC
"Dolby D +Neo:X G": GAME
"Rock Arena": MUSIC
"Jazz Club": MUSIC
"Matrix": MUSIC
"Mono Movie": MOVIE
"Video Game": GAME

ol-iver · 2018-04-02T22:36:20Z

@starkillerOG : you can remove the if "selectSurround" in relevant_tags.keys()

Added aditional sound mode mapping values removed the if tag in list checks for the pop function.

houndci-bot · 2018-04-03T11:03:13Z

                        "Media Server": "SERVER", "Spotify": "SPOTIFY",
                        "Flickr": "FLICKR", "Favorites": "FAVORITES"}

+SOUND_MODE_MAPPING = OrderedDict \


whitespace before '('

instead of looping over the original nice syntax self._sound_mode_dict, a second dict is created based on self._sound_mode_dict, this new dict is called self._SM_match_dict. In this bigger dictionary the key value structure is reversed (but it contains duplicate values). This self._SM_match_dict is used to match the raw sound mode with a simple dict value request based on key.

houndci-bot · 2018-04-03T12:52:59Z

+    def set_sound_mode_dict(self, sound_mode_dict):
+        Error_msg = ("Syntax of sound mode dictionary not valid, "
+                     "use: OrderedDict([('COMMAND', ['VALUE1','VALUE2'])])")
+        if (type(sound_mode_dict) == OrderedDict or type(sound_mode_dict) == dict):


line too long (83 > 79 characters)

houndci-bot · 2018-04-03T12:53:00Z

+
+    @property
+    def sound_mode_match_dict(self):
+        """Return a dictionary that is used to map each sound_mode_raw to it's matched sound_mode."""


line too long (101 > 79 characters)

houndci-bot · 2018-04-03T12:53:00Z

+
+    @property
+    def sound_mode_dict(self):
+        """Return a dictionary of available sound modes with their mapping values."""


line too long (85 > 79 characters)

houndci-bot · 2018-04-03T13:05:52Z

+        Error_msg = ("Syntax of sound mode dictionary not valid, "
+                     "use: OrderedDict([('COMMAND', ['VALUE1','VALUE2'])])")
+        if (type(sound_mode_dict) == OrderedDict or\
+            type(sound_mode_dict) == dict):


visually indented line with same indent as next logical line

houndci-bot · 2018-04-03T13:05:52Z

+    def set_sound_mode_dict(self, sound_mode_dict):
+        Error_msg = ("Syntax of sound mode dictionary not valid, "
+                     "use: OrderedDict([('COMMAND', ['VALUE1','VALUE2'])])")
+        if (type(sound_mode_dict) == OrderedDict or\


the backslash is redundant between brackets

houndci-bot · 2018-04-03T13:05:52Z

+    @property
+    def SM_match_dict(self):
+        """
+        Return a dictionary that is used to 


trailing whitespace

starkillerOG · 2018-04-03T13:18:19Z

@scarface-4711 I implemented all your change requests. Thanks for the feedback.
I also included all the mapping values of your receiver.

Your receiver works the same as mine, I can also define "stereo" from each of the menu's (MUSIC, MOVIE, GAME etc. However this is actually the same sound mode regardless from wich menu you selected it. You can see this if you look at the phone application or at the web-interface of the receiver. Their the sound mode menu's are organized in a diffrent way and their there is only 1 "stereo" option.

starkillerOG · 2018-04-03T13:19:17Z

@scarface-4711 could you have a final look at the code and give any aditional comments or merge it?

ol-iver · 2018-04-03T21:29:41Z

@starkillerOG changes are looking good.

There are some small comments and warnings from pylint and pydocstyle. Could you please fix them? I'll push a new version for you tomorrow evening then.
Thanks.

PS C:\Users\Oliver\python-venv\python35-testing\denonavr-patch-1> pylint denonavr
No config file found, using default configuration
************* Module denonavr.denonavr
C:1056, 0: Unnecessary parens after 'elif' keyword (superfluous-parens)
C:1306, 0: Wrong continued indentation before block (add 4 spaces).
            type(sound_mode_dict) == dict):
            ^   | (bad-continuation)
W:1307, 0: Bad indentation. Found 16 spaces, expected 12 (bad-indentation)
W:1308, 0: Bad indentation. Found 16 spaces, expected 12 (bad-indentation)
W:1309, 0: Bad indentation. Found 20 spaces, expected 16 (bad-indentation)
W:1310, 0: Bad indentation. Found 24 spaces, expected 20 (bad-indentation)
W:1311, 0: Bad indentation. Found 28 spaces, expected 24 (bad-indentation)
W:1312, 0: Bad indentation. Found 32 spaces, expected 28 (bad-indentation)
W:1313, 0: Bad indentation. Found 32 spaces, expected 28 (bad-indentation)
W:1314, 0: Bad indentation. Found 20 spaces, expected 16 (bad-indentation)
W:1315, 0: Bad indentation. Found 24 spaces, expected 20 (bad-indentation)
W:1316, 0: Bad indentation. Found 24 spaces, expected 20 (bad-indentation)
C:239, 8: Invalid attribute name "_SM_match_dict" (invalid-name)
C:1155, 4: Invalid attribute name "SM_match_dict" (invalid-name)
C:1302, 4: Missing method docstring (missing-docstring)
C:1303, 8: Invalid variable name "Error_msg" (invalid-name)
C:1305,12: Using type() instead of isinstance() for a typecheck. (unidiomatic-typecheck)
C:1306,12: Using type() instead of isinstance() for a typecheck. (unidiomatic-typecheck)
C:1309,23: Using type() instead of isinstance() for a typecheck. (unidiomatic-typecheck)
C:1311,31: Using type() instead of isinstance() for a typecheck. (unidiomatic-typecheck)
C:1324, 4: Invalid method name "construct_SM_match_dict" (invalid-name)
C:1324, 4: Missing method docstring (missing-docstring)
R:1324, 4: Method could be a function (no-self-use)
C:1332, 4: Missing method docstring (missing-docstring)
W:1293,16: Attribute '_sound_mode' defined outside __init__ (attribute-defined-outside-init)

-------------------------------------------------------------------
Your code has been rated at 9.72/10 (previous run: 10.00/10, -0.28)

PS C:\Users\Oliver\python-venv\python35-testing\denonavr-patch-1> pydocstyle denonavr
denonavr\denonavr.py:1147 in public method `sound_mode_dict`:
        D205: 1 blank line required between summary line and description (found 0)
denonavr\denonavr.py:1147 in public method `sound_mode_dict`:
        D400: First line should end with a period (not 's')
denonavr\denonavr.py:1155 in public method `SM_match_dict`:
        D205: 1 blank line required between summary line and description (found 0)
denonavr\denonavr.py:1155 in public method `SM_match_dict`:
        D400: First line should end with a period (not 'o')
denonavr\denonavr.py:1302 in public method `set_sound_mode_dict`:
        D102: Missing docstring in public method
denonavr\denonavr.py:1324 in public method `construct_SM_match_dict`:
        D102: Missing docstring in public method
denonavr\denonavr.py:1332 in public method `match_sound_mode`:
        D102: Missing docstring in public method
PS C:\Users\Oliver\python-venv\python35-testing\denonavr-patch-1> flake8 denonavr
denonavr\denonavr.py:40:33: E211 whitespace before '('
denonavr\denonavr.py:1306:13: E129 visually indented line with same indent as next logical line
PS C:\Users\Oliver\python-venv\python35-testing\denonavr-patch-1>

houndci-bot · 2018-04-04T08:07:12Z

+        # sent command
+        try:
+            if self.send_get_command(command_url):
+                self.sound_mode() = sound_mode


SyntaxError: can't assign to function call

houndci-bot · 2018-04-04T08:07:13Z

                        "Flickr": "FLICKR", "Favorites": "FAVORITES"}

+SOUND_MODE_MAPPING = OrderedDict\
+([('MUSIC', ['PLII MUSIC', 'DTS NEO:6 MUSIC', 'DOLBY D +NEO:X M',


continuation line missing indentation or outdented

houndci-bot · 2018-04-04T10:11:21Z

                        "Media Server": "SERVER", "Spotify": "SPOTIFY",
                        "Flickr": "FLICKR", "Favorites": "FAVORITES"}

+SOUND_MODE_MAPPING = OrderedDict(\


the backslash is redundant between brackets

starkillerOG · 2018-04-04T10:16:18Z

@scarface-4711 I think I fixed all small comments and warnings.
Could you check again, I do not know how I can check it.

ol-iver · 2018-04-04T21:16:40Z

@starkillerOG version 0.7.0 is now available on pypi including your sound mode support.

starkillerOG · 2018-04-04T21:21:19Z

@scarface-4711 Thank you!
I have my HomeAssistant code almost ready, will push a PR tommorow using this new version.
I just tested all the sound mode functions in version 0.7.0 and it seems to work perfectly.
Thank you for your help!

houndci-bot reviewed Mar 5, 2018

View reviewed changes

Add sound mode support (typo-fix)

50b891d

fixed houndci-bot typo's

houndci-bot reviewed Mar 5, 2018

View reviewed changes

Add sound mode support (indent fix)

0c26783

houndci-bot reviewed Mar 5, 2018

View reviewed changes

Add sound mode support (indent fix)

302cb9f

starkillerOG mentioned this pull request Mar 5, 2018

Sound mode support media_player general and denonavr home-assistant/core#11764

Closed

9 tasks

ol-iver reviewed Mar 25, 2018

View reviewed changes

starkillerOG added 2 commits March 26, 2018 17:02

Move sound mode matching warning

8349bda

Moved the warning if the sound mode can not be matched from the update function to when you call the matched sound mode.

Added sound mode mapping values

ca63f87

Added all the sound mode values that were present in denonavr/tests/xml/

houndci-bot reviewed Mar 26, 2018

View reviewed changes

white space

7d5cf51

ol-iver requested changes Apr 2, 2018

View reviewed changes

move match_sound_mode to sound_mode property

3cfd304

moved match_sound_mode to sound_mode property. Therefore I could get rid of self._sound_mode and self. _sound_mode_match_warning.

Added mapping values & removed if check

11c81f0

Added aditional sound mode mapping values removed the if tag in list checks for the pop function.

houndci-bot reviewed Apr 3, 2018

View reviewed changes

starkillerOG added 2 commits April 3, 2018 13:04

white space

489ebfd

houndci-bot reviewed Apr 3, 2018

View reviewed changes

fix typo

c561221

fix line to long

d9ba123

houndci-bot reviewed Apr 3, 2018

View reviewed changes

fix indent

387d9d6

ol-iver approved these changes Apr 3, 2018

View reviewed changes

pylint fixes

ed4fba1

houndci-bot reviewed Apr 4, 2018

View reviewed changes

fix indent & set_sound_mode

0b88140

houndci-bot reviewed Apr 4, 2018

View reviewed changes

delete unnesesary \

0498541

ol-iver merged commit f2d9087 into ol-iver:master Apr 4, 2018

Conversation

starkillerOG commented Mar 5, 2018

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

starkillerOG commented Mar 5, 2018

Uh oh!

starkillerOG commented Mar 18, 2018

Uh oh!

ol-iver commented Mar 25, 2018

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

starkillerOG commented Mar 26, 2018

Uh oh!

starkillerOG commented Mar 26, 2018

Uh oh!

Choose a reason for hiding this comment

Uh oh!

starkillerOG commented Mar 26, 2018

Uh oh!

ol-iver left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

starkillerOG Apr 2, 2018 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

starkillerOG commented Apr 2, 2018 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

ol-iver commented Apr 2, 2018

Uh oh!

ol-iver commented Apr 2, 2018

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

starkillerOG commented Apr 3, 2018

starkillerOG Apr 2, 2018 •

edited

Loading

starkillerOG commented Apr 2, 2018 •

edited

Loading

ol-iver commented Apr 3, 2018 •

edited

Loading