How To Convert Compound Names Encoded In Html To Utf-8 (Or Similar)?

Question

[Egon Willighagen]

I am converting data from LycoCyc 3.3 into a BridgeDb file, and the compounds.dat of LycoCyc has names in HTML encoding, such as:

<i>D</i>-glycerate

However, the BridgeDb will have names without HTML, e.g. with UTF-8, though I would be happy to have the italics just removed. Is there a Java library that can convert a HTML string like that given above into a non-HTML string, for example in ASCII or even UTF-8 for some superscripted and subscripted digits?

[Pierre Lindenbaum]

using xslt ?.

[Egon Willighagen]

a possibility...

[Pierre Lindenbaum]

forget it , I thought you only wanted to grab the HTML pages.

Answer 1

[Istvan Albert]

What you are looking for are tools that implement the functionality called "strip html tags". This can be done with regular expressions, for example with sed

sed 's/<[^>]*>//g' index.html  | more

or with various libraries like so:

http://stackoverflow.com/questions/753052/strip-html-from-strings-in-python

now moving that into superscript/subscript will require a custom HTML parsing and formatting.

[Egon Willighagen]

Actually, this does not take care of HTML entities, like α ... other than that, I was hoping for a Java library, though I obviously can do regex's there too...

Answer 2

[Egon Willighagen]

I found Jsoup that may serve my needs, or at least some. The library is small, and the Java code simple (or Groovy in this snippet):

  println "val: " + value
  value = Jsoup.parseBodyFragment(value).text()
  println "val: " + value

This takes care of removing the HTML tags and the decoding of the HTML entities, but not the super/subscripting of digits. Example output:

val: trans-&Delta;<SUP>2</SUP>, cis-&Delta;<SUP>4</SUP>-decadienoyl-CoA
val: trans-Δ2, cis-Δ4-decadienoyl-CoA

I welcome other answers very much, particularly if they are smart about sub- and superscript!