Search results no longer include delimiter pipe

[r0ckarong]

My page titles are build like

Product Version | Section | Topic

Until some recent change which I can not fathom the search result would look like that.
Now the search result is only "Topic".

I'm trying to find out if your indexer does something different.

My implementation looks like this, pretty much the reference from your instructions:

Script insertion in the actual page template

  <!-- JS-Search -->
  <script src="<%= File.join(javascripts_path, "jssearch.js") %>" /></script>
  <script src="<%= File.join(javascripts_path, "jssearch.index.js") %>" /></script>

  <script type="text/javascript">
    $(document).on("keyup", "#searchbox", function() {
      var result = jssearch.search($(this).val());

      $('#query').html(jssearch.queryWords.join(' '));

      $('#results').html('');
      var i = 0;
      result.forEach(function(item) {
        if (i++ > 20) {
          return;
        }

        var div = $('#results');

        var version = "<%= version %>"
        var location = window.location.href;
        var pos = location.substr(0, location.search(version));
        var oerl = item.file.url;
        var clean = oerl.substr(2, oerl.length)

        div.html(div.html() + '<li>' + '<a href="' + pos + version + clean + '">' + item.file.title + '</a>' + '</li>');
      });
    });
  </script>

I've also updated to 0.9.3 and I see the same behavior.

[cebe]

Until some recent change which I can not fathom the search result would look like that.

there have not been significant changes to this package since 2014, are you sure its a problem with the code in this package?

[r0ckarong]

That is what I'm trying to debug. Suddenly the titles recorded into jssearch index no longer contain the longer notation with the pipes. Where does the indexer grab the title from and how is it processed?

[cebe]

js-search/lib/Indexer.php

Lines 57 to 71 in 3756a8b

	protected function generateFileInfo($file, $contents, $basePath, $baseUrl)
	{
	// create file entry
	if (preg_match('~<h1>(.*?)</h1>~s', $contents, $matches)) {
	$title = strip_tags($matches[1]);
	} elseif (preg_match('~<title>(.*?)</title>~s', $contents, $matches)) {
	$title = strip_tags($matches[1]);
	} else {
	$title = '<i>No title</i>';
	}
	return [
	'url' => $baseUrl . str_replace('\\', '/', substr($file, strlen(rtrim($basePath, '\\/')))),
	'title' => $title,
	];
	}

[cebe]

have you added <h1> tags that were not present before?

[r0ckarong]

I did actually convert a previously h2 into an h1. So it takes the text from the first h1 it finds over the HTML <title> tag from what I can read? This would explain it.

[r0ckarong]

Ok yeah that was the cause. Ideally could you add an option to pass with the script that allows me to select if the first h1 or the actual <title> tag is used?

[cebe]

I can add it but I do not have time for it right now. If you want to have it earlier I am happy to accept a pull request.