Question

This is the question I asked yesterday. I was able to get the required data. The final data is like this. Please follow this link.

I tried with the following code to get all the infobox data

                                content = content.split("}}\n");
                                for(k in content)
                                {
                                    if(content[k].search("Infobox")==2)
                                    {
                                        var infobox  = content[k];
                                        alert(infobox); 
                                        infobox = infobox.replace("{{","");
                                        alert(infobox);
                                        infobox = infobox.split("\n|");
                                        //alert(infobox[0]);
                                        var infohtml="";
                                        for(l in infobox)
                                        {
                                            if(infobox[l].search("=")>0)
                                            {
                                                var line = infobox[l].split("=");

                                                infohtml = infohtml+"<tr><td>"+line[0]+"</td><td>"+line[1]+"</td></tr>";

                                            }
                                        }
                                        infohtml="<table>"+infohtml+"</table>";
                                        $('#con').html(infohtml);
                                        break;
                                    }
                                }

I initially thought each element is enclosed in {{ }}. So I wrote this code. But what I see is, I was not able to get the entire infobox data with this. There is this element

{{Sfn|National Informatics Centre|2005}}

occuring which ends my infobox data.

It seems to be far simpler without using json. Please help me

Answer 1

Have you tried DBpedia? Afaik they provide template usage information. There is also a toolserver tool named Templatetiger, which does template extraction from the static dumps (not live).

However, I once wrote a tiny snippet to extract templates from wikitext in javascript:

var title; // of the template
var wikitext; // of the page
var templateRegexp = new RegExp("{{\\s*"+(title.indexOf(":")>-1?"(?:Vorlage:|Template:)?"+title:title)+"([^[\\]{}]*(?:{{[^{}]*}}|\\[?\\[[^[\\]]*\\]?\\])?[^[\\]{}]*)+}}", "g");
var paramRegexp = /\s*\|[^{}|]*?((?:{{[^{}]*}}|\[?\[[^[\]]*\]?\])?[^[\]{}|]*)*/g;
wikitext.replace(templateRegexp, function(template){
    // logabout(template, "input ");
    var parameters = template.match(paramRegexp);
    if (!parameters) {
        console.log(page.title + " ohne Parameter:\n" + template);
        parameters  = [];
        }
    var unnamed = 1;
    var p = parameters.reduce(function(map, line) {
        line = line.replace(/^\s*\|/,"");
        var i = line.indexOf("=");
        map[line.substr(0,i).trim() || unnamed++] = line.substr(i+1).trim();
        return map;
    }, {});
    // you have an object "p" in here containing the template parameters
});

It features one-level nested templates, but still is very error-prone. Parsing wikitext with regexp is as evil as trying to do it on html 🙂

It may be easier to query the parse-tree from the api: api.php?action=query&prop=revisions&rvprop=content&rvgeneratexml=1&titles=….
From that parsetree you will be able to extract the templates easily.