new ways for scraping data.

[hsfzxjy]

Due to the weird design of WikiMedia, Igem Parts Registry may use non-semantic tags for documentation rendering (e.g. use <table> for typography). This may lead to incorrect parsing by html2markdown, and mess up the final result. Comparatively, directly accessing the Edit page (take BBa_K2042000 as an example) can fetch the raw page code in WikiText format, which can make the parsing simpler and preciser.

Besides, the History page in wikitools will list out recent changes on the page, so there is no need to re-fetch the whole page entirely during upgrading.