Jump to content

Extension:TextExtracts

From mediawiki.org
This extension comes with MediaWiki 1.34 and above. Thus you do not have to download it again. However, you still need to follow the other instructions provided.
This extension is under code stewardship review and not actively maintained (task T256505). No new feature requests will be considered during this period.
MediaWiki extensions manual
TextExtracts
Release status: stable
Implementation API
Description Provides API with plain-text or limited HTML extracts of page content
Author(s) Max Semenik (MaxSemtalk)
Compatibility policy Snapshots releases along with MediaWiki. Master is not backward compatible.
MediaWiki >= 1.43
Database changes No
License GNU General Public License 2.0 or later
Download
  • $wgExtractsRemoveClasses
  • $wgExtractsExtendOpenSearchXml
  • $wgExtractsExtendRestSearch
Quarterly downloads 106 (Ranked 56th)
Public wikis using 2,046 (Ranked 189th)
Translate the TextExtracts extension if it is available at translatewiki.net
Issues Open tasks · Report a bug

The TextExtracts extension provides an API which allows retrieval of plain-text or limited HTML (HTML with content for some CSS classes removed) extracts of page content.

Installation

[edit]
  • Download and move the extracted TextExtracts folder to your extensions/ directory.
    Developers and code contributors should install the extension from Git instead, using:cd extensions/
    git clone https://gerrit.wikimedia.org/r/mediawiki/extensions/TextExtracts
  • Add the following code at the bottom of your LocalSettings.php file:
    wfLoadExtension( 'TextExtracts' );
    
  • Yes Done – Navigate to Special:Version on your wiki to verify that the extension is successfully installed.

Configuration settings

[edit]
  • $wgExtractsRemoveClasses is an array of <tag>, <tag>.class, .<class>, #<id> which will be excluded from extraction.
    For example, $wgExtractsRemoveClasses[] = 'dl'; removes indented text, often used for non-templated hatnotes that are not desired in summaries.
    extension.json defines the defaults, of which the class "noexcerpt" is one - this may be added to any template to exclude it.
  • $wgExtractsExtendOpenSearchXml defines whether TextExtracts should provide its extracts to the Opensearch API module. The default is false.

API

[edit]


prop=extracts (ex)

(main | query | extracts)

Returns plain-text or limited HTML extracts of the given pages.

Specific parameters:
Other general parameters are available.
exchars

How many characters to return. Actual text returned might be slightly longer.

Type: integer
The value must be between 1 and 1,200.
exsentences

How many sentences to return.

Type: integer
The value must be between 1 and 10.
exlimit

How many extracts to return. (Multiple extracts can only be returned if exintro is set to true.)

Type: integer or max
The value must be between 1 and 20.
Default: 20
exintro

Return only content before the first section.

Type: boolean (details)
explaintext

Return extracts as plain text instead of limited HTML.

Type: boolean (details)
exsectionformat

How to format sections in plaintext mode:

plain
No formatting.
wiki
Wikitext-style formatting (== like this ==).
raw
This module's internal representation (section titles prefixed with <ASCII 1><ASCII 2><section level><ASCII 2><ASCII 1>).
One of the following values: plain, raw, wiki
Default: wiki
excontinue

When more results are available, use this to continue. More detailed information on how to continue queries can be found on mediawiki.org.

Type: integer


Another example
Result
{
    "query": {
        "pages": {
            "9228": {
                "pageid": 9228,
                "ns": 0,
                "title": "Earth",
                "extract": "Earth, also called the world and, less frequently, Gaia, (or Terra in some works of science fiction)..."
            }
        }
    }
}

Caveats

[edit]

There are various things to be aware of when using the API or software that uses the API e.g. Extension:Popups .

  • We do not recommend the usage of `exsentences`. It does not work for HTML extracts and there are many edge cases for which it doesn't exist. For example "Arm. gen. Ing. John Smith was a soldier." will be treated as 4 sentences. We do not plan to fix this.
  • Inline images are stripped from the response (even in HTML mode). This means if you are using the Math extension and using formulae in your lead section they may not appear in the summary output.
  • In HTML mode we cannot guarantee well formed HTML. Resulting HTML may be invalid or malformed.
  • In plaintext mode:
    • citations may not be stripped (see phab:T197266)
    • if a paragraph ends with an HTML tag e.g. ref tag, new lines may be dropped (see phab:T201946)
    • new lines may be dropped after lists phab:T208132
    • Articles must begin with the lead paragraph for an extract to be generated. The use of any template, and/or unclosed or empty HTML element may result in no preview for the article. E.g. "<div></div>hello" will give an empty extract.

FAQ

[edit]

How can I remove content from a page preview/extract?

[edit]

TextExtracts will strip any element that is marked with the class .noexcerpt. This is provided by the $wgExtractsRemoveClasses configuration variable (which also defines some other excluded elements).

See also

[edit]