Debug and enhance a custom web crawler

Project Description:
Need a skilled Java developer to debug, fix code of and enhance an existing image web crawler. Successful completion of the following tasks is requested:
Debug:
1) debug and fix a memory leak (java heap error) in the code of the crawler.
2) ever once in a while the app hangs, while memory is still available. Need to understand if this is related to the memory leak.

Enhance:
1) when user stops the app, subsequent restart should continue from the last scanned URL and not from the starting project URL. For example, if starting URL is amazon.com and the crawler was stopped when it was scanning http://www.amazon.com/Kindle-eReader-eBook-Reader-e-Reader-Special-Offers/dp/B0051QVESA/ref=amb_link_356991982_1?pf_rd_m=ATVPDKIKX0DER&pf_rd_s=browse&pf_rd_r=10CQ1DAT4KJ1GZASSBYJ&pf_rd_t=101&pf_rd_p=1330783002&pf_rd_i=283155, the restart should begin from the last URL and not fresh from amazon.com
2) Should be able to capture and scan through relative URLs
3) When given a sub-directory as a starting point, the scan should begin from that directory inwards, and not from the main domain. For example, if bbc.co.uk/international is the starting URL, then the scan starts from there, stays/scans deeper within that directory (i.e. does not go up to bbc.co.uk or bbc.co.uk/domestic) and neither starts from bbc.co.uk