Question

I would like to take some stuff from file A and reformat it to stick into file B using regular expressions. I am kind of new to vim so this may be a dumb question but I could not find the solution to this anywhere. I guess I am searching for the wrong phrases. Anyway, here are the details of what I want to do. I have a static html page that I would like to have an RSS feed for. Luckily, this page is mostly links to various news items, so creating the RSS will be pretty easy.

I have the regular expression ready:

:%s/^<a href="\(.\{-}\)".title="\(.\{-}\)">\(.\{-}\)<\/a>/<title>\3<\/title>\r<link>\1<\/link>\r<description>\2<\/description> 

My problem is I do not want to make the changes in the html file that I am searching. I want the changes to occur in another file, new or existing. How do I make this happen? Or is this method completely off.

Oh and by the way, this expression takes something like this in the html file:

<a href="http://linktosomesite.com" title="Description of link">Title of Link</a>

and turns it into this in the xml file:

<title>Title of Link</title>
<link>http://linktosomesite.com</link>
<description>Description of link</description>

Bonus: It would be really nice if I can place this within another file, say starting at line 5.

PS: I know this is a vim and regex question but posting it in html and rss because I noticed people have static html to rss questions there.

Answer 1

It sounds like you want to write a transform. There are many transform tools. You certainly could do it with sed & awk for example. But I think the easiest way would be xslt. (you could use xsltproc or saxon…)

Here’s an example template:

<xsl:template match="a">
    <title><xsl:value-of select="text()"/></title>
    <link><xsl:value-of select="@href"/></link>
    <description><xsl:value-of select="@title"/></description>
</xsl:template>

It finds each a element, and outputs the results, with the text() node and attributes filled in.