Viktor Oreshkin
1b71c6fbf1
Fix vector regex in parser ( #347 )
2017-10-16 19:19:16 +02:00
Lonami Exo
83677fc927
Enhance and use .get_input_photo on the generated code
2017-10-08 13:45:14 +02:00
Lonami Exo
91a5d20e93
Replace .on_response with static .from_reader for all types
2017-10-07 13:26:09 +02:00
Lonami Exo
06bb09b95c
Fix wrong .empty() due to variable shadowing (from ce48c97)
2017-10-01 10:45:59 +02:00
Lonami Exo
ce48c9752e
Assert flag params with same flag index are all set/unset
2017-09-30 19:47:19 +02:00
Lonami Exo
afc4bd9cab
Rename constructor/subclass_of_id to upper case, keep only static
2017-09-29 13:11:33 +02:00
Lonami Exo
a4fa652d35
Precompute the flag index on the generated code
2017-09-29 13:03:35 +02:00
Lonami Exo
76d14b2dd9
Make generated .to_bytes() more readable
2017-09-29 13:00:22 +02:00
Lonami Exo
195bba6f21
Fix bug where booleans were always serialized as false
2017-09-29 12:57:53 +02:00
Lonami Exo
423efc4360
Fix optional vector arguments failing
2017-09-29 11:56:43 +02:00
Lonami Exo
4bedbfe945
Use bytes instead string type on the auth-key methods
2017-09-28 10:32:57 +02:00
Lonami Exo
14791e1df1
Fix wrong type hint for int128 and int256
2017-09-28 10:12:39 +02:00
Lonami Exo
748c2560de
Fix auto-generated code didn't actually write int128/int256
2017-09-28 10:12:15 +02:00
Lonami Exo
fb0898b9cb
Don't distinguish between str and bytes when serializing
...
This makes it easier to use some requests like ReqPqRequest which
needs a string of bytes, not a valid utf-8 string per se.
2017-09-28 09:55:29 +02:00
Lonami Exo
9b7733d6b9
Avoid the use of starred expressions on autogen code ( fix #266 )
2017-09-27 12:05:35 +02:00
Lonami Exo
b83cd98ba0
Replace TLObject.on_send with the new .to_bytes()
...
This also replaces some int.to_bytes() calls with a faster
struct.pack (up to x4 faster). This approach is also around
x3 faster than creating a BinaryWriter just to serialize a
TLObject as bytes.
2017-09-26 14:36:02 +02:00
Lonami Exo
b40708a8c7
Import get_input_* only once on the generated code
2017-09-25 14:05:13 +02:00
Lonami Exo
dc73744fc4
Add get_input_media calls on the generated code
2017-09-25 13:47:25 +02:00
Lonami Exo
1593a4415e
Fix None datetime being type hinted as invalid
2017-09-23 20:07:40 +02:00
Lonami Exo
5701029fbf
Stop treating .to_dict() with no args as a special case
2017-09-23 10:59:08 +02:00
Lonami Exo
0b48b1ec8a
Add support for non-recursive TLObject.to_dict()
2017-09-23 10:41:36 +02:00
Lonami Exo
81d4d76d16
Fix generated code showing the incorrent type hint for datetime
2017-09-19 13:20:34 +02:00
Lonami Exo
f7e4f3f678
Make type hinting on the generated code more IDE-friendly
2017-09-18 21:00:06 +02:00
Lonami Exo
2595d45bd7
Import os only once on the generated code
2017-09-18 20:23:03 +02:00
Lonami Exo
2c5c674531
Stop using TLObject.__repr__ to show the TL definition
2017-09-18 20:19:39 +02:00
Lonami Exo
b14a7d9e68
Revert layer being a property
2017-09-17 16:19:53 +02:00
Lonami Exo
aea10b103b
Rename generated layer variable to uppercase readonly property
2017-09-17 16:17:55 +02:00
Lonami Exo
3a5eab2e83
Fix infering random_id lists failing for ForwardMessages
2017-09-12 10:32:11 +02:00
Lonami Exo
97cab7347b
Make lint happier
2017-09-04 17:10:04 +02:00
Lonami Exo
0e45f46ad5
Properly ignore core types when generating code
2017-09-04 16:38:03 +02:00
Lonami Exo
e8cdcf369d
Sort generated code by name for more predictable results
2017-09-04 15:11:46 +02:00
Lonami Exo
c5b562d4f6
Move the from tlobject import TLObject to only the top of the file
2017-09-04 14:08:36 +02:00
Lonami Exo
64f1031a45
Move comments from auto-generated code to the generator itself
2017-09-04 14:05:30 +02:00
Lonami Exo
c3f679597c
Move the auto-generated notice to only the top of the file
2017-09-04 13:57:40 +02:00
Lonami Exo
e83b250a22
Replace generated directory modules with simply files
2017-09-04 13:51:47 +02:00
Lonami Exo
7eab28206b
Change the code generator to use a single file per namespace
2017-09-04 13:44:39 +02:00
Lonami Exo
a5d86b4572
Update to layer 71
2017-09-04 11:54:37 +02:00
Lonami Exo
160a3699ac
Fix confusing names "MtProtoRequest" and ".confirmed" ( #176 )
...
This also fixes the annoyingly confusing message:
"Odd msg_seqno expected (relevant message), but even received."
2017-07-24 16:54:48 +02:00
Lonami Exo
38e0888ea0
Update to layer 70
2017-07-23 17:12:29 +02:00
Lonami Exo
9b17888e3e
Make generated code slightly smaller
2017-07-11 11:36:00 +02:00
Lonami Exo
cfea0f80da
Consider vector attributes for get_input_* utils ( closes #166 )
2017-07-11 11:14:58 +02:00
Lonami Exo
bdee94eaf3
Implement automatic cast to InputUser too ( closes #159 )
2017-07-10 16:04:10 +02:00
Lonami Exo
95a989be2c
Automatically cast Channel to InputChannel (similar to InputPeer)
2017-07-07 09:48:06 +02:00
Lonami Exo
8fd0d7eadd
Add a new .stringify() function to visualize TLObjects more easily
2017-07-04 21:15:47 +02:00
Lonami Exo
23e2802215
Update to layer 68
2017-06-30 11:48:45 +02:00
Lonami Exo
79ee98a4dd
Infer the object ID from its TL definition when not given
2017-06-30 11:28:54 +02:00
Lonami Exo
b0b814bdb9
Use setup.py to run tl_generator.py instead (plus some cleanup)
2017-06-21 11:27:22 +02:00
Lonami Exo
e7b0c06ca5
Make .get_input_user faster when the right type is given
2017-06-16 10:01:05 +02:00
Lonami Exo
86d45cc276
Improve .get_input_peer and use it only when creating requests*
...
This avoids cyclic dependencies, so types requiring an InputPeer
as a parameter will NOT convert faulty types to the right ones.
2017-06-15 17:03:59 +02:00
Lonami Exo
b8e46446ba
Automatically call .get_input_peer on the requests that need it
2017-06-15 16:35:40 +02:00
Lonami Exo
9e07a2580c
Fix extra line being appended on empty .to_dict()'s
2017-06-13 19:46:16 +02:00
Lonami Exo
769692959f
Make the TLGenerator class a lot more readable
2017-06-12 10:43:43 +02:00
Lonami Exo
68a625b82b
Try to adhere to the 80-characters limit
2017-06-12 10:16:24 +02:00
Goblenus
76166cd1ec
Add .to_dict() to the generated types
2017-06-12 09:53:26 +02:00
Lonami Exo
1ca41b5854
Automatically infer the value for 'random_id' parameters
2017-06-11 19:16:59 +02:00
Lonami Exo
c6acd6adc5
Use only ASCII characters on the code ( closes #104 )
2017-06-10 13:39:37 +02:00
Lonami Exo
9b5d1b02e0
Add return type to the docs of the generated code
2017-05-23 10:10:01 +02:00
Lonami Exo
02a847b64a
Make lint happier
2017-05-21 13:59:16 +02:00
Lonami Exo
63c89af983
Use relative imports always where possible
2017-05-21 13:02:54 +02:00
Lonami Exo
ca80b05694
Add shebangs and set proper file permissions
2017-05-20 15:58:44 +02:00
Lonami Exo
befc7a5657
Fix layer 66 not being the real layer 66
...
Telegram decided to update the scheme.tl without increasing
the layer number, so it had been unnoticed until now.
2017-05-20 11:59:41 +02:00
Lonami Exo
c4447bf226
Distinguish between vector type and constructor ( fix #81 )
2017-05-19 09:41:22 +02:00
Lonami Exo
b84641023a
Handle more cases where the type is a date
2017-05-19 09:30:30 +02:00
Birger Jarl
930dc99f31
Serialize flag_indicator argument to flags:#
2017-05-13 09:38:49 +02:00
Dmitry D. Chernov
6baad93938
SourceBuilder: Don't perform stream flushing twice
...
stream.close() always perform flush by itself
2017-05-09 19:50:46 +02:00
Lonami Exo
4421c3ac7a
Update to layer 66
2017-04-15 18:10:27 +02:00
Belegnar
0e44657963
* doc improvement
...
* relative imports not allowed when module is not loaded (https://www.python.org/dev/peps/pep-0366/ )
2017-03-10 14:13:22 +03:00
Lonami Exo
b5dfc5d514
Fix requests returning vectors not working ( closes #26 )
2017-02-25 11:28:55 +01:00
Lonami Exo
a5bd740b10
Updated to layer 62
2017-01-04 15:40:53 +01:00
Fadi Hadzh
d087941bd0
Many code-style improvements
2016-11-30 00:29:42 +03:00
Lonami Exo
be94bff576
Patched code generator and updated README.rst removing markdown leftovers
2016-11-23 21:03:58 +01:00
Lonami
6a32c6cd7a
Moved requests .constructor_id to be a class-level variable
2016-10-04 18:55:02 +02:00
Lonami
f2728f47c3
Updated scheme to layer 57 and fixes to README links
2016-10-03 19:57:21 +02:00
Lonami
77aa37d2ad
Added an .empty() function to the objects, fixes to msgs_ack
...
The .empty() function returns an "empty" instance of the object
(attributes set to None). This is used rather than using reflection.
The msgs_ack handling broke stuff (InvokeWithLayer when there were
updates), so this is now handled; yet there may be a better fix
2016-09-26 17:18:25 +02:00
Lonami
36b8a9026f
Fixed revert by hand (I hope)
2016-09-26 13:13:11 +02:00
Lonami
b68772aab5
Fixed JSON encoding and decoding for lists, added empty() method
...
The new empty() method retrieves an empty instance of the
given object, with all the attributes set to None
2016-09-25 10:50:48 +02:00
Lonami
fddb3e9aac
Added ability to encode/decode TLObjects, updated README
2016-09-24 21:17:41 +02:00
Lonami
6642f73a3d
Detect current MTProtoLayer automatically
2016-09-17 21:09:20 +02:00
Lonami
51a531225f
Totally refactored source files location
...
Now it *should* be easier to turn Telethon
into a pip package
2016-09-17 20:45:12 +02:00
