Fix message parsing for documents containing Content-Length keyword #80
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This commit fixes the logic for parsing a batch of messages in data_received(), where each message body may contain a literal 'Content-Length' keyword. Previously, if a document contained this keyword, the the server would crash, because it attempted to split messages on this keyword. Now, instead of splitting the messages on the keyword, it parses content length from the first keyword, and then extracts the message based on the length. After that, the parsing loop is repeated, until no more messages can be read.
This commit enables receival of partial messages, where the entire message may arrive in chunks. We achieve this by maintaining _message_buf [], which builds the message as the parts arrive, until the entire message (as specified in its Content-Length) can be parsed. Then, the loop continues for any other messages that may arrive together with the first one. The updated regex allows extra headers in the message. Additionally, we change aio_readline() logic to always pass the entire message (with headers) to data_received(). This allows us to avoid extraneous logic than required otherwise. Finally, we add tests for this functionality.
|
@dinvlad Thanks for the PR, great work! Is it ready for the review or you plan to make more changes? |
|
I think it's ready, thanks! |
danixeee
reviewed
Aug 31, 2019
CONTRIBUTORS.md
Outdated
| @@ -5,3 +5,4 @@ | |||
| - [Max O'Cull](https://github.com/Maxattax97) | |||
| - [Tomoya Tanjo](https://github.com/tom-tan) | |||
| - [yorodm](https://github.com/yorodm) | |||
| - [Denis Loginov](https://github.com/dinvlad) | |||
There was a problem hiding this comment.
Can you make this alphabetical?
danixeee
reviewed
Aug 31, 2019
pygls/protocol.py
Outdated
| @@ -397,18 +397,43 @@ def connection_made(self, transport: asyncio.Transport): | |||
| """Method from base class, called when connection is established""" | |||
| self.transport = transport | |||
|
|
|||
| MESSAGE_PATTERN = re.compile( | |||
There was a problem hiding this comment.
Can you move this up where other class constants are (alphabetically)?
danixeee
suggested changes
Aug 31, 2019
|
Awesome, thanks for reviewing! I've made the changes. |
|
I have tested everything and it works. I needed to test json-extension after installing from I merged another PR today, can you just resolve conflicts and include link to your PR? CHANGELOG should look like this: - Fix parsing of partial messages and those with Content-Length keyword ([#80])
- Fix Full SyncKind for servers accepting Incremental SyncKind ([#78])
[#80]: https://github.com/openlawlibrary/pygls/pull/80
[#78]: https://github.com/openlawlibrary/pygls/pull/78 |
|
Sounds good - resolved, please take a look. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
This PR fixes the logic for parsing a batch of messages in
data_received(), where each message body may contain a literalContent-Lengthkeyword.Previously, if a document contained this keyword, then the server would crash, because it attempted to split messages on this keyword.
Now, instead of splitting the messages on the keyword, it parses content length from the first keyword, and then extracts the message based on the length. After that, the parsing loop is repeated, until no more messages can be read.
Update: I added handling of partial messages, since some messages may arrive in chunks.
Code review checklist (for code reviewer to complete)