Package com.norconex.importer.handler.transformer.impl

Class DomTransformer

java.lang.Object
com.norconex.importer.handler.transformer.impl.DomTransformer
All Implemented Interfaces:
Configurable<DomTransformerConfig>, ConfigurableDocHandler<DomTransformerConfig>, DocHandler
public class DomTransformer extends Object implements ConfigurableDocHandler<DomTransformerConfig>

Extract the value of one or more elements or attributes into a target field, or delete matching elements. Applies to HTML, XHTML, or XML document.

This class constructs a DOM tree from a document or field content. That DOM tree is loaded entirely into memory. Use this tagger with caution if you know you'll need to parse huge files. It may be preferable to use RegexTransformer if this is a concern. Also, to help performance and avoid re-creating DOM tree before every DOM extraction you want to perform, try to combine multiple extractions in a single instance of this Tagger.

The jsoup parser library is used to load a document content into a DOM tree. Elements are referenced using a CSS or JQuery-like syntax.

Should be used as a pre-parse handler.

Storing values in an existing field

If a target field with the same name already exists for a document, values will be added to the end of the existing value list. It is possible to change this default behavior by supplying a PropertySetter.

Content-types

By default, this filter is restricted to (applies only to) documents matching the restrictions returned by CommonRestrictions.domContentTypes(String). You can specify your own content types if you know they represent a file with HTML or XML-like markup tags.

When used as a pre-parse handler, this class attempts to detect the content character encoding unless the character encoding was specified using DomTransformerConfig.setSourceCharset(java.nio.charset.Charset). Since document parsing converts content to UTF-8, UTF-8 is always assumed when used as a post-parse handler.

You can control what gets extracted exactly thanks to the "extract" argument of the new method DomOperation.setExtract(String). Possible values are:

  • text: Default option when extract is blank. The text of the element, including combined children.
  • html: Extracts an element inner HTML (including children).
  • outerHtml: Extracts an element outer HTML (like "html", but includes the "current" tag).
  • ownText: Extracts the text owned by this element only; does not get the combined text of all children.
  • data: Extracts the combined data of a data-element (e.g. <script>).
  • id: Extracts the ID attribute of the element (if any).
  • tagName: Extract the name of the tag of the element.
  • val: Extracts the value of a form element (input, textarea, etc).
  • className: Extracts the literal value of the element's "class" attribute, which may include multiple class names, space separated.
  • cssSelector: Extracts a CSS selector that will uniquely select (identify) this element.
  • attr(attributeKey): Extracts the value of the element attribute matching your replacement for "attributeKey" (e.g. "attr(title)" will extract the "title" attribute).

You can specify a fromField as the source of the HTML to parse instead of using the document content. If multiple values are present for that source field, DOM extraction will be applied to each value.

You can specify a defaultValue on each DOM extraction details. When no match occurred for a given selector, the default value will be stored in the toField (as opposed to not storing anything). When matching blanks (see below) you will get an empty string as opposed to the default value. Empty strings and spaces are supported as default values (the default value is now taken literally).

You can set matchBlanks to true to match elements that are present but have blank values. Blank values are empty values or values containing white spaces only. Because white spaces are normalized by the DOM parser, such matches will always return an empty string (spaces will be trimmed). By default elements with blank values are not matched and are ignored.

You can specify which parser to use when reading documents. The default is "html" and will normalize the content as HTML. This is generally a desired behavior, but this can sometimes have your selector fail. If you encounter this problem, try switching to "xml" parser, which does not attempt normalization on the content. The drawback with "xml" is you may not get all HTML-specific selector options to work. If you know you are dealing with XML to begin with, specifying "xml" should be a good option.

Content deletion

You can specify whether to delete any elements matched by the selector with DomOperation.setDelete(boolean). You can use it with a "toField" or on its own. Some options are ignored by deletions, such as "extract" or "defaultValue". Deletion applies to whatever the operation targets: the document content when reading it directly, or a metadata field's value when using "fromField".

Given this HTML snippet... 

 <div class="firstName">Joe</div>
 <div class="lastName">Dalton</div>

... the above example will store "Joe" in a "firstName" field and "Dalton" in a "lastName" field.

See Also: