Summary: Shows fragment of page around given word(s).
Version: 2007-09-30 - v1.2
Prerequisites: PmWiki2
Status: beta
Questions answered by this recipe
Can I show (for example in my search results) a fragment of a page around given word, with the word shown in bold, as it's done in Google search results?
Description
After installation, two new markups are available:
Single excerpt
(:excerpt Group.Page "search terms" :) or
(:excerpt Group.Page term :)
This syntax displays one line, showing where in specified page is the "search terms" phrase, or "term" word.
Example:
(:excerpt PmWiki.InitialSetupTasks "when you":)
- will find words around the first occurence of "when you" phrase in the PmWiki.InitialSetupTasks page, which will show as:
...n file (''local/config.php'') When you first install PmWiki, the ''l...
A pack of excerpts
(:excerpts Group.Page first_term second_term ... :)
The plural syntax (note the 's' in 'excerpts') allows to display several excerpts of the same page using one directive. It creates a single excerpt for each specified term.
Example:
(:excerpts PmWiki.WikiGroup foo bar:)
- will find an excerpt for each of "foo" and "bar":
...e a default page in the group Foo, create a link to [=[[Foo/]]=...
....]]=]). To make a page called Bar in the group Foo, create a li...
Installation
Installation is standard: download excerpts.phpΔ, save it in your cookbook/ directory, add include_once("$FarmD/cookbook/excerpts.php"); to your local/config.php file.
Configuration parameters (optional):
$ExcerptSideWidth = 30;- How many characters of the page are included at each side of the matched fragment.
Notes
- It's case insensitive as intended, but not for UTF-8 characters. I haven't yet came with a good solution; see source code for details. If you want to improve the code, please feel welcome. I'm also open for suggestions - please put them in comments, below.
- Todo: The recipe should first try to match whole words, and match parts of the words only if that failed.
- Todo: Test if the
(:excerpts:) directive works properly both with single (') and double quotes ("). Then test Cookbook:SortByScore about the same (or doesn't the latter need to work with singles?).
- Todo: The Markup() regexp should require whitespaces between
(:excerpt or (:excerpts and characters written immediately afterwards.
Release Notes
- 2007-09-30 - v1.2
- removed unused ExcertpsToupper() function
- added $ExcerptSideWidth
- 2007-09-26 - v1.1 - bug fix, which helps to make Cookbook:SortByScore's lists cleaner when excerpts not required
- 2006-12-21 - v0.9 - initial release, by MateuszCzaplinski
Comments
- Is there any way to hide (or process) markup in the excerpts? Francis March 02, 2007, at 03:51 AM
- There is no simple way for that currently. I wanted to do this the way you suggest, but problems arose and I found no simple solution yet. MateuszCzaplinski September 26, 2007, at 01:50 PM
- In more details:
- To parse markup correctly, I think at least one line of page text should be included. But that can result in a veeery long excerpt.
- This recipe is designed to work with Cookbook:SortByScore. The latter currently searches whole page text (as does PmWiki's normal search). And if markup was processed, some matches (like 'include') would disappear, and that might lead reader to confusion.
- That said, I don't claim that's completely impossible. I suppose that might be a good goal for some 1.5 or 2.0 version.
I wonder if there is a solution to the difficulty I have in that the search text may appear more than once on a page and I would like to show a line for each occurrence in the search results. Is there a switch or setting to cause either of the methods to show all of the hits on the same page?
See Also
Contributors
User notes?: If you use, used or reviewed this recipe, you can
add your name. The following format is recognized:
* (+) Optional positive comment. Name, date
* (-) Optional negative comment. Name, date
These statistics appear in the Cookbook listings and will help newcomers browsing through the wiki.
