Speed up string parsing in ODSP binary snapshot flow

[vladsud]

Shows ~ 25% improvement for structure parsing part, or about 10% overall improvement. The only risk here - strings containing zeros, and possibly some differences in parsing code (browser vs. node) how zeros are handled.

Also converting recursion into loop. By itself it does not buy us much, but it localizes all parsing code in one function, which is big plus.

Other changes:

• Propagated logger through layers
• Simplified a bit APIs and removed unneeded exports

[msfluid-bot]

⯅ @fluid-example/bundle-size-tests: +1.08 KB

Metric Name	Baseline Size	Compare Size	Size Diff
aqueduct.js	397.39 KB	397.39 KB	■ No change
connectionState.js	680 Bytes	680 Bytes	■ No change
containerRuntime.js	196.33 KB	196.33 KB	■ No change
loader.js	153.74 KB	153.74 KB	■ No change
map.js	42.89 KB	42.89 KB	■ No change
matrix.js	131.83 KB	131.83 KB	■ No change
odspDriver.js	151.26 KB	151.8 KB	⯅ +554 Bytes
odspPrefetchSnapshot.js	39.34 KB	39.88 KB	⯅ +551 Bytes
sharedString.js	153.04 KB	153.04 KB	■ No change
Total Size	1.27 MB	1.27 MB	⯅ +1.08 KB

---

Baseline commit: dc536059677bca0cf636cef1d7072a589451ef46

Generated by :no_entry_sign: dangerJS against 829d3655ea29730ef4ea8efb97020a070bc0613b

[github-actions[bot]]

This commit is queued for merging with the next branch! Please ignore this PR for now. Contact @microsoft/fluid-cr-infra for help.