feat: improve the performance and add tests

[sirily11]

No description provided.

[github-actions]

PDF conversion benchmark

namespace=markitdown-server  memory_limit=2048Mi (2Gi)  budget_scale=1.0
(pod restarted before each book, so PEAK_MB is that book's own high-water mark)

BOOK                           SECONDS   BUDGET  PEAK_MB LIMIT_PCT     CHARS  VERDICT
-----------------------------------------------------------------------------------------
I Robot.pdf                       12.8       35      533      26%    410434  ok
Prelude to Foundation.pdf         20.7       55      581      28%    768223  ok
Foundation and Earth.pdf          71.8      180      631      31%    919223  ok
Foundation Trilogy.pdf            32.7       85      556      27%   1260851  ok

All books converted within time and memory budgets.
(per-book pod logs in /tmp/benchmark-pod-logs)

Commit b27d7ce · workflow run

[sirily11]

🎉 This PR is included in version 1.1.0 🎉

The release is available on GitHub release

Your semantic-release bot 📦🚀