com.univocity.api.entity.html.builders

Interface Group

All Superinterfaces:

BasicElementFilter<Group>, ElementFilter<Group>, ElementFilterStart<Group>, FieldDefinition, UpDown<Group>

public interface Group
extends ElementFilter<Group>, ElementFilterStart<Group>, FieldDefinition

A group defines the boundaries where a given set of fields should be processed. The parser will only collect values from elements of a group if a matching condition is satisfied.

For example, given the following HTML:

<div id="55">
  <h1>random text</h1>
  <p>random paragraph</p>
  <article>
    <h1>first</h1>
    <p>lorem</p>
  </article>
  <h1>random text 2</h1>
  <p>random paragraph 2</p>
  <article>
    <h1>second</h1>
    <p>ipsum</p>
  </article>
</div>

To get all values of every h1 and p element inside an article, the following rules can be created:

HtmlEntityList entities = new HtmlEntityList();
HtmlEntitySettings entity = entities.configureEntity("test");

Group group = entity
    .newGroup()
        .startAt("article")
        .endAtClosing("article");

group.addField("title").match("h1").getText();
group.addField("text").match("p").getText();

In the example above, the group starts at every article element, and ends when the parser finds a closing article element. Fields must added to the group. The h1 and p elements outside of article are simply ignored by the parser.

The example above should produce the following two records:

[first, lorem]
[second, ipsum]

Author:

uniVocity Software Pty Ltd - dev@univocity.com

Method Summary

Modifier and Type	Method and Description
RecordTriggerStart	addRecordTrigger() Starts the definition of a RecordTrigger, which is essentially a path to an element that when found makes the parser generate a record will all values accumulated so far.
T	copyPath() Copies the current path allowing new matching to be added to a common path without changing the original one.

Methods inherited from interface com.univocity.api.entity.html.builders.BasicElementFilter

attribute, attribute, childOf, classes, containedBy, containedBy, containing, containing, containing, filter, followedBy, followedBy, followedByText, followedImmediatelyBy, id, matchNext, not, parentOf, precededBy, precededBy, precededByText, precededImmediatelyBy, under, underHeader, withExactText, withExactTextMatchCase, withText, withTextMatchCase

Methods inherited from interface com.univocity.api.entity.html.builders.UpDown

downTo, downToFooter, upTo, upToHeader

Methods inherited from interface com.univocity.api.entity.html.builders.ElementFilterStart

match, match, match, matchCurrent, matchFirst, matchLast, select

Methods inherited from interface com.univocity.api.entity.html.builders.FieldDefinition

addField, addField, addPersistentField, addSilentField

Method Detail

copyPath

T copyPath()

Copies the current path allowing new matching to be added to a common path without changing the original one. Useful for when the path to multiple elements share a common structure. An example can be shown with this HTML: ```html

first

lorem

second

ispum

first

article

``` A technique to get the text of the `p` element of the first and second `article`s of the first `div` is: ```java HtmlEntityList entities = new HtmlEntityList(); HtmlEntitySettings entity = entities.configureEntity("test"); //creates a new path with a common matching sequence. PartialPath path = entity.newPath() .match("article").containing("h1"); //the copied path ends with 'h1', we further specify it should start with text 'first'. path.copyPath().withText("first") .addField("one").match("p").getText(); //the original path is unchanged. We can copy it again and match 'h1' elements now starting with text 'second' path.copyPath().withText("second") .addField("two").match("p").getText(); ``` In this example, the first path that is created is used as a basis. This path is then copied to specify the the text to be matched from a `h1` element. Fields can be added to paths and in the example above each copy matches any `p` element starting from the path. This will make the parser capture the values: ``` [lorem, ispum] ```

Returns:

a PartialPath that allows the specification of a path and does not affect the path that it is built upon.

addRecordTrigger

RecordTriggerStart addRecordTrigger()

Starts the definition of a RecordTrigger, which is essentially a path to an element that when found makes the parser generate a record will all values accumulated so far. For example, assume you have to capture email address and home address fields from a HTML document of customer details. In the document, any of the customers fields may or may not exist (i.e. be `null` when parsed). An example of this HTML is shown below: ```html

Customer A address

Email Address	bla@email.com
Home Address

Customer B address

Email Address
Home Address	123 real street

``` Consider the code without a RecordTrigger: ```java HtmlEntityList entityList = new HtmlEntityList(); HtmlEntitySettings contact = entityList.configureEntity("contact"); contact.addField("emailAddress") .match("td").precededBy("td").withText("Email Address").getText(); contact.addField("homeAddress") .match("td").precededBy("td").withText("Home Address").getText(); ``` This would produce one row with mixed results: ``` [bla.@email.com, 123 real street] ``` Instead of the expected output: ``` [bla.@email.com, null] [null, 123 real street] ``` The incorrect output is produced because there is no `` after the first 'Home Address' and no `` after 'E-mail Address' in the details of the second customer. We can instruct the parser to generate a record with all values collected so far when a `` with text "Email Address" is found with the following: ```java path.addRecordTrigger() .match("td").withText("Email Address"); ``` Now, when the parser hits the second email address entry, the record trigger will be activated and a new row will be created. Now, when the second Home Address is found, it adds the value to the second row instead of the first row.

Returns:

a RecordTriggerStart that defines the path for the trigger