data: Replace custom formats with msgpack#374
Draft
fstachura wants to merge 1 commit intobootlin:masterfrom
Draft
data: Replace custom formats with msgpack#374fstachura wants to merge 1 commit intobootlin:masterfrom
fstachura wants to merge 1 commit intobootlin:masterfrom
Conversation
tleb
reviewed
Feb 14, 2025
| self.families = parsed_data[1] | ||
| else: | ||
| self.entries = [] | ||
| self.families = "" |
Member
There was a problem hiding this comment.
Could we work only on the raw data, and parsing it when things are requested? Goal is to store only a bytes buffer without taking loads of memory if we want to have loads in memory.
tleb
reviewed
Feb 14, 2025
| else: | ||
| return self.ctype(p) | ||
| else: | ||
| return None |
Member
There was a problem hiding this comment.
Early return the if p is None case.
tleb
reviewed
Feb 14, 2025
| if type(key) is str: | ||
| key = key.encode() | ||
| elif type(key) is int: | ||
| key = msgpack.dumps(key) |
Member
There was a problem hiding this comment.
Do we want to do this? Isn't it rather an error if someone gives us a string or int? A key is of type bytes, callers that don't respect that have a bug IMO.
tleb
reviewed
Feb 14, 2025
| elif type(arg) is int: | ||
| arg = str(arg).encode() | ||
| arg = msgpack.dumps(arg) | ||
| return arg |
Member
There was a problem hiding this comment.
Same comment: is this used? Shouldn't callers know what they have and do the right thing themselves.
tleb
reviewed
Feb 14, 2025
| previous_type = sd.type; | ||
| } | ||
| let ln = sd.line.toString().split(','); | ||
| let ln = [sd.line]; |
tleb
reviewed
Feb 14, 2025
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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.
I have previously attempted to refactor data converters used in data.py in this branch. In the end I wasn't happy with the result, because I believe that writing custom parsers for each database is the wrong approach.
This PR replaces all data.py parsers with msgpack. Plain python objects can be serialized and deserialized into the databases.
The main advantage is convenience - values can be manipulated like normal Python objects, no string parsing is required anywhere in the codebase that interacts with the database.
From what I remember, larger databases were also a bit smaller, mainly because large ints take less space in msgpack than in base10 representation. But to be fair, there is some storage overhead for other datatypes.
I also wouldn't be surprised if average serialization/deserialization times were a bit smaller, although I don't have numbers on that and I doubt it's a major bottleneck anywhere.
Leaving this as a draft - I tested it only a little bit.