Implement Lazy Directory Parsing for Improved Deserialization Performance by expanded-for-real · Pull Request #13 · imprint-serde/imprint-java

expanded-for-real · 2025-06-06T02:04:05Z

Implemented lazy directory de-serializing. Times were a bit scattered on the runs but consistently saw strong results, and I saw as low as 250 ns/ops at one point for Imprint. Either way we were consistently at least twice as fast as before so I think the additional complexity is worth it.

Before

Library	Score (ns/op)	Error	Relative Performance
FlatBuffers	56.86	±3.97	1.0x (baseline)
Imprint	1,109.83	±105.04	19.5x slower

After

Serialization Format	Average Time (ns/op)	Std Deviation
FlatBuffers	106.512	± 40.800
Imprint	527.664	± 66.763

We should be in a good spot now to implement Merge and Project

…mance tracking; add comprehensive String benchmark

…to remove a bunch of stuff

…ing deserialization a bit faster

Try to enhance string deserialization

A full list of enhancements can be found here - #3

…ome micro-optimizations added that were found along the way

* Full comprehensive comparison tests with a lot of other libraries + some micro-optimizations added that were found along the way * replace deprecated gradle methods with latest --------- Co-authored-by: expand3d <>

# Conflicts: # src/jmh/java/com/imprint/benchmark/ComparisonBenchmark.java # src/main/java/com/imprint/core/ImprintRecord.java # src/main/java/com/imprint/types/TypeHandler.java # src/main/java/com/imprint/types/Value.java

…with nested objects

…es in gradle file. Also fix permission issue

…Comparison tests iterations manually

# Conflicts: # .github/workflows/ci.yml # build.gradle # src/jmh/java/com/imprint/benchmark/ComparisonBenchmark.java # src/main/java/com/imprint/core/ImprintRecord.java # src/main/java/com/imprint/types/TypeHandler.java # src/main/java/com/imprint/types/Value.java

expanded-for-real · 2025-06-06T02:04:45Z

.github/workflows/ci.yml

Massively reduced the ci compared to last time - probably won't have to change this much for a while

# Conflicts: # .github/workflows/ci.yml # build.gradle # src/jmh/java/com/imprint/benchmark/ComparisonBenchmark.java # src/main/java/com/imprint/core/ImprintRecord.java

expanded-for-real · 2025-06-06T02:11:36Z

build.gradle

Massively reduced gradle from last time

expanded-for-real · 2025-06-06T02:14:09Z

src/main/java/com/imprint/util/VarInt.java

Cache VarInt values. Minor performance tweak that I came up with when running comparisons and was low hanging fruit.

expanded-for-real · 2025-06-06T02:15:40Z

src/main/java/com/imprint/core/ImprintRecord.java

I put all the buffer handling logic in a new class to leave the ImprintRecord a bit cleaner (especially since I'm about to put merge and project in there too). I tried to comment it as best I could since the buffer handling logic can be confusing a took me a few tries before tests passed

agavra

skimmed the ImprintRecord changes but otherwise lgtm

agavra · 2025-06-06T17:41:07Z

src/main/java/com/imprint/core/ImprintBuffers.java

+        if (directoryCount >= 0)
+            return directoryCount;


minor, but it seems like overkill to cache the directory count - reading a single varint is pretty speedy

Yeah I over-engineered that for the sake of avoiding directory hits but it ends up just being unnecessary branching here. Will remove in the next PR

agavra · 2025-06-06T17:43:17Z

src/main/java/com/imprint/core/ImprintBuffers.java

+    private final ByteBuffer payload; // Read-only payload view
+
+    // Lazy-loaded directory state
+    private List<DirectoryEntry> parsedDirectory;


why did you prefer a list to a map (idx -> entry) for efficient lookup when parsed? that way we can also lazily fill it in if we wanted to on every lookup (in addition to a single call to eagerly parse)

Good question! The reason is because I didn't think of that lol. I think I mostly wanted to preserve insertion order on these.... but they're already sorted coming in anyways so that's a good idea. I'll implement the map on next PR

agavra · 2025-06-06T17:43:34Z

src/main/java/com/imprint/core/ImprintBuffers.java

+            return -1;
+        int low = 0;
+        int high = parsedDirectory.size() - 1;
+        while (low <= high) {


if we use a map this is just an O(1) lookup into the parsed dir

agavra · 2025-06-06T17:45:24Z

src/main/java/com/imprint/core/ImprintBuffers.java

+            if (fieldId > currentFieldId)
+                return offset;


shouldn't we binsearch here instead of go through field by field?

expanded-for-real and others added 30 commits June 1, 2025 13:23

Update .gitignore

5cb33c9

initial commit for imprint-java

63313e4

initial commit for imprint-java

dd4fdbc

Add GitHub Actions CI workflow for automated testing

bce1d13

Merge remote-tracking branch 'origin/dev' into dev

f5d90b5

Update GitHub Actions workflow to use upload-artifact@v4

72c468f

Add Gradle wrapper validation to CI workflow

468d682

Fix gitignore to include gradle-wrapper.jar for CI

cf05b13

Force add gradle-wrapper.jar to repository

d0d7983

Update wrapper validation action to v3

f2cdd1b

Fix Javadoc syntax errors and disable strict Javadoc checking

57c8249

Add JMH benchmark .bat and .sh for full suite benchmarking and perfor…

edb3057

…mance tracking; add comprehensive String benchmark

fix map serialization error in benchmark test and streamline ci file …

2853e3f

…to remove a bunch of stuff

Add execute permissions back for gradlew in CI

3a5a113

Add some more string based performance benchmarks and try to make str…

50a288b

…ing deserialization a bit faster

Merge pull request #2 from imprint-serde/faster-strings

ea1c4c4

Try to enhance string deserialization

second main commit to address initial commits

43cab28

A full list of enhancements can be found here - #3

additional cleanup to address concerns in #3

fdb8a56

minor style fixes

2e56688

minor style fixes again

9353388

minor style fixes on benchmark tests and supress unused

09d0377

minor reordering

6209bb1

Merge branch 'main' into dev

ace7c67

Full comprehensive comparison tests with a lot of other libraries + s…

4632e01

…ome micro-optimizations added that were found along the way

replace deprecated gradle methods with latest

3738861

Merge Comparisons into dev branch (#8)

12d2823

* Full comprehensive comparison tests with a lot of other libraries + some micro-optimizations added that were found along the way * replace deprecated gradle methods with latest --------- Co-authored-by: expand3d <>

Lazy load of directory and header data

f7a6e8e

Merge remote-tracking branch 'origin/main' into dev

2834dbb

# Conflicts: # src/jmh/java/com/imprint/benchmark/ComparisonBenchmark.java # src/main/java/com/imprint/core/ImprintRecord.java # src/main/java/com/imprint/types/TypeHandler.java # src/main/java/com/imprint/types/Value.java

minor cleanup

83ed961

minor cleanup

a605b65

expand3d added 7 commits June 5, 2025 15:23

minor cleanup

aacddeb

Actually fixes offsets and read Byte Values for Maps and Arrays even …

3bf81ad

…with nested objects

change CI file to use JMH plugin to respect iteration and warmup valu…

7eaa6e9

…es in gradle file. Also fix permission issue

ok plugin didn't work apparently so reverting that and just reducing …

32640cd

…Comparison tests iterations manually

trying to update github ci to make jmh actually work correctly

880aeb0

lazy directory deserialization

8831922

expanded-for-real commented Jun 6, 2025

View reviewed changes

Merge branch 'main' into dev

e361cf0

# Conflicts: # .github/workflows/ci.yml # build.gradle # src/jmh/java/com/imprint/benchmark/ComparisonBenchmark.java # src/main/java/com/imprint/core/ImprintRecord.java

expanded-for-real commented Jun 6, 2025

View reviewed changes

build.gradle

Copy link

Collaborator Author

expanded-for-real Jun 6, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Massively reduced gradle from last time

remove extra comments

73eade6

expanded-for-real commented Jun 6, 2025

View reviewed changes

remove extra comments

02866d5

expanded-for-real requested a review from agavra June 6, 2025 02:17

expanded-for-real assigned agavra Jun 6, 2025

agavra approved these changes Jun 6, 2025

View reviewed changes

expanded-for-real merged commit e1b6dcf into main Jun 6, 2025
3 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Implement Lazy Directory Parsing for Improved Deserialization Performance #13

Implement Lazy Directory Parsing for Improved Deserialization Performance #13
expanded-for-real merged 40 commits intomainfrom
dev

expanded-for-real commented Jun 6, 2025 •

edited

Loading

Uh oh!

expanded-for-real Jun 6, 2025

Uh oh!

expanded-for-real Jun 6, 2025

Uh oh!

expanded-for-real Jun 6, 2025

Uh oh!

expanded-for-real Jun 6, 2025

Uh oh!

agavra left a comment

Uh oh!

agavra Jun 6, 2025

Uh oh!

expanded-for-real Jun 6, 2025

Uh oh!

agavra Jun 6, 2025

Uh oh!

expanded-for-real Jun 6, 2025

Uh oh!

agavra Jun 6, 2025

Uh oh!

agavra Jun 6, 2025

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

expanded-for-real commented Jun 6, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Before

After

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!

agavra 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!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

expanded-for-real commented Jun 6, 2025 •

edited

Loading