Converted db.rs to use prepared statements.

[shawn-zil]

Trying out an idea. Not exactly sure how to interpret the Bencher Report but I believe that this small change has resulted in a modest improvement in performance.

process-empty/process-empty
                        time:   [4.7159 ms 4.7270 ms 4.7387 ms]
                        thrpt:  [211.03  elem/s 211.55  elem/s 212.05  elem/s]
                 change:
                        time:   [-12.425% -12.045% -11.635%] (p = 0.00 < 0.05)
                        thrpt:  [+13.167% +13.694% +14.187%]
                        Performance has improved.
Found 43 outliers among 500 measurements (8.60%)
  17 (3.40%) high mild
  26 (5.20%) high severe

Benchmarking produce-full/produce-full: Warming up for 3.0000 s
Warning: Unable to complete 20 samples in 120.0s. You may wish to increase target time to 159.2s, enable flat sampling, or reduce sample count to 10.
produce-full/produce-full
                        time:   [1.4192 s 1.7721 s 2.1411 s]
                        thrpt:  [0.4671  elem/s 0.5643  elem/s 0.7046  elem/s]
                 change:
                        time:   [-68.031% -44.402% +9.6003%] (p = 0.14 > 0.05)
                        thrpt:  [-8.7593% +79.863% +212.80%]
                        No change in performance detected.

     Running unittests src/lib.rs (target/release/deps/zilliqa_macros-4a76e05b52691afc)

[github-actions]

Bencher Report

Branch	2160-use-prepared-sql
Testbed	self-hosted

Click to view all benchmark results

Benchmark	Latency	Benchmark Result milliseconds (ms) (Result Δ%)	Upper Boundary milliseconds (ms) (Limit %)
process-empty/process-empty	📈 view plot 🚷 view threshold	7.72 (-16.02%)	10.41 (74.16%)
produce-full/produce-full	📈 view plot 🚷 view threshold	2,231.00 (+13.03%)	2,299.74 (97.01%)

🐰 View full continuous benchmarking report in Bencher