fix: resolve duplicated status parsing in ltsvreader

[kazeburo]

PR Type

Bug fix, Enhancement

---

Description

• Preserve first duplicate LTSV status value.

• Prioritize configured status keys in both readers.

• Return no values when parsing fails.

• Replace parse flags with value checks.

---

File Walkthrough

	Relevant files
Enhancement	2 files axslog.go Remove parse flags from the reader interface                          +10/-19  parser.go Validate parsed values instead of status flags                      +3/-3
Tests	3 files axslog_test.go Remove tests for obsolete parse flags                                        +0/-12    jsonreader_test.go Update JSON tests for two-value parse results                        +3/-14    ltsvreader_test.go Add duplicate-key coverage and update parse assertions      +15/-18
Error handling	1 files jsonreader.go Simplify JSON results and handle parse errors                        +8/-10
Bug fix	1 files ltsvreader.go Ignore duplicate fields and prioritize configured statuses +16/-8
Configuration changes	1 files Makefile Declare additional build and check targets phony                  +1/-1

---

[github-actions]

PR Reviewer Guide 🔍

(Review updated until commit 0e7faa8)

Here are some key observations to aid the review process:

⏱️ Estimated effort to review: 2 🔵🔵⚪⚪⚪

🧪 PR contains tests

🔒 No security concerns identified

⚡ Recommended focus areas for review Duplicate Tracking With at least 64 configured status keys, shifting a uint64 by an index of 64 or more produces zero. Repeated occurrences of those keys are therefore counted as new matches, allowing remaining to reach zero before a higher-priority status key appears. A log entry can then return the wrong status. bit := uint64(1) << idx if found&bit != 0 { return nil } found |= bit remaining--

✅ Resolved findings

jsonreader/jsonreader.go:73-75

Malformed JSON

A JSON parse error now returns values found before the error instead of discarding them. With status keys status and code, malformed input such as {"reqtime":0.03,"code":200, can be counted as a valid request. Return empty values on unexpected parse errors.

ltsvreader/ltsvreader.go:41-46

Bitset Overflow

For a configured key at index 64 or higher, this shift produces a zero bit. Repeated occurrences are then counted as new matches, so enough duplicates can make remaining reach zero and stop parsing before a later processing-time or higher-priority status field.

ltsvreader/ltsvreader.go:43-48

Duplicate Tracking

For configurations with at least 65 keys, shifting a uint64 by an index of 64 or more produces zero, so duplicates of those keys are never marked as found. A duplicate can decrement remaining twice and stop parsing before a later reqtime field is read.

[github-actions]

Benchmark Comparison

Base: master
PR: fix/ltsvparser-duplicated-status

BenchmarkMainParse results

goos: linux
goarch: amd64
pkg: github.com/monitoring-forge/mackerel-plugin-axslog
cpu: AMD EPYC 9V45 96-Core Processor                
                             │ base-bench.txt │            pr-bench.txt            │
                             │     sec/op     │   sec/op     vs base               │
MainParse_jsonl-4                18.69m ±  4%   18.66m ± 6%       ~ (p=0.971 n=10)
MainParse_ltsv-4                 13.23m ± 17%   12.98m ± 8%       ~ (p=0.631 n=10)
MainParse_jsonl_and_output-4     20.24m ±  8%   20.58m ± 4%       ~ (p=0.218 n=10)
MainParse_ltsv_and_output-4      14.13m ±  4%   13.79m ± 4%  -2.37% (p=0.011 n=10)
geomean                          16.31m         16.19m       -0.70%

                             │ base-bench.txt │            pr-bench.txt             │
                             │      B/op      │     B/op      vs base               │
MainParse_jsonl-4                4.289Mi ± 0%   4.289Mi ± 0%       ~ (p=0.796 n=10)
MainParse_ltsv-4                 4.289Mi ± 0%   4.289Mi ± 0%       ~ (p=0.424 n=10)
MainParse_jsonl_and_output-4     5.059Mi ± 0%   5.059Mi ± 0%       ~ (p=0.093 n=10)
MainParse_ltsv_and_output-4      5.059Mi ± 0%   5.059Mi ± 0%  +0.00% (p=0.000 n=10)
geomean                          4.658Mi        4.658Mi       +0.00%

                             │ base-bench.txt │            pr-bench.txt             │
                             │   allocs/op    │ allocs/op   vs base                 │
MainParse_jsonl-4                  96.00 ± 0%   96.00 ± 0%       ~ (p=1.000 n=10)
MainParse_ltsv-4                   96.00 ± 0%   96.00 ± 0%       ~ (p=1.000 n=10)
MainParse_jsonl_and_output-4       157.0 ± 1%   157.0 ± 0%       ~ (p=1.000 n=10)
MainParse_ltsv_and_output-4        157.0 ± 0%   157.0 ± 0%       ~ (p=1.000 n=10) ¹
geomean                            122.8        122.8       +0.00%
¹ all samples are equal

[github-actions]

Persistent review updated to latest commit c7049f8

[sonarqubecloud]

Quality Gate passed

Issues
1 New issue
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
1.8% Duplication on New Code

See analysis details on SonarQube Cloud

[github-actions]

Persistent review updated to latest commit 0e7faa8