CODEWITHSUNDEEP
xpath
Lupius
Lupius

Reputation: 178

Using XPath to select all rows following a header in a table

I'm having a hard time figuring out how to do something that seems very simple. Let's say I have an HTML table such as the following:

<table><tbody>
  <tr><th>First header</th></tr>
  <tr../>
  <tr../>
  <tr../>
  <tr><th>Second header</th></tr>
  <tr../>
</tbody></table>

I want all three rows that immediately follow the "First header". So far I have '/table/tbody/tr[preceding-sibling::tr/th[1]/text()="First header"]', but it's giving me every single row in the table after the "First header". What am I doing wrong?

Edit: I'm working with code that passes in the header as a variable, so I'm parsing the table without knowing if there is another header at the end or what it would be. More generically, given a header string, retrieve all following rows until the next header or end of table.

Upvotes: 1

Views: 2361

Answers (2)

Lupius
Lupius

Reputation: 178

I got this after more trial and error: '/table/tbody/tr[preceding-sibling::tr[th/text()="First header"] = preceding-sibling::tr[th][1]]' Which translates to English: get all rows preceded by the "First header" row where that row is also the first preceding row that contains a header.

Upvotes: 0

alecxe
alecxe

Reputation: 474221

You can get every tr tag that has the preceding-sibling's th text = First header and a following-sibling contains th tag:

//tr[preceding-sibling::tr/th = 'First header' and following-sibling::tr/th]

Demo (using xmllint):

$ xmllint index.html --xpath "//tr[preceding-sibling::tr/th = 'First header' and following-sibling::tr/th]"
<tr>1</tr><tr>2</tr><tr>3</tr>

where index.html contains:

<table>
    <tbody>
    <tr>
        <th>First header</th>
    </tr>
    <tr>1</tr>
    <tr>2</tr>
    <tr>3</tr>
    <tr>
        <th>Second header</th>
    </tr>
    <tr>4</tr>
    </tbody>
</table>

Upvotes: 1

Related Questions