Hi folks, Sorry for posting a question unrelated to boost. I know pretty well that boost has perfect solutions to it. But I am working on a legacy system using Hpricot of Ruby on Rail. So only Hpricot-specific suggestions please. Thank you. In my HTML parser, I can parse a html file with the following hpricot commands: (1) doc = open( "MyFileToParse.html" ) { |f| Hpricot(f) } (2) elements = (doc.search("/html/body/table/tr/td/table/tr/td/font") ) (3) puts (elements[13]).inner_html to get the following output: Giaever G, et al (2002). Functional profiling of the Saccharomyces c erevisiae genome. Nature, 418:387-91. [<a href=" http://www.ncbi.nlm.nih.gov/entr ez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=12140549&dopt=Abstract" target="_ blank">PubMed</a>] How can I proceed to get the following results (3) and (4) respectively? (3) Giaever G, et al (2002). Functional profiling of the Saccharomyces c erevisiae genome. Nature, 418:387-91. (4) http://www.ncbi.nlm.nih.gov/pubmed/12140549?dopt=Abstract NOTE: to get (4) I need to take two more steps: (5) replace "&" with "?" (6) replace "PubMed" with "pubmed" (this might be trivial, but how?) in the process of parsing in addition to "normal" HTML parsing. Thanks a lot in advance. Robert