9092b2a29c
Remove sharing elements in fewer situations ( #545 )
...
* remove fewer share elements
* simplify and fix social-buttons testcase
5 years ago
814f0a3884
Add support for detecting lazy-loaded images ( #542 )
...
Add support for detecting lazy-loaded images using `src` or `srcset` attributes.
5 years ago
cb5771fd4a
Add nested font tags to test _setNodeTag on those (see #59 )
5 years ago
9009f64f9c
Fix table header missing ( #530 )
5 years ago
6761a7e412
Fix embedded videos getting removed ( #526 )
...
Fix embedded videos getting removed
5 years ago
b9cece3e58
add test
5 years ago
a014e0c9c8
exclude graphs from nytimes articles
5 years ago
c942b32945
Revert source files and fix expected results
5 years ago
bd5087d2f1
fix error in testing "wikipedia"
5 years ago
3e025d58e5
fix error in testing "lwn-01"
5 years ago
df95c9d717
fix error in testing "keep-tabulard-data"
5 years ago
6a5066abe2
Fix tabular data got removed
5 years ago
f70d36852b
check itemprop when determining whether a node is a byline
5 years ago
b9f47bcc8d
fix(test-util): fix generate testcase tool
5 years ago
d41de78c26
Close img tag
5 years ago
1187b2dae1
Update test expectations
5 years ago
3ca8c12d87
Update test expectations
5 years ago
4ffd482004
Add medicalnewstoday test case with incorrect results
5 years ago
60ef565b67
Don't choke on <meta> tags that do not have a content attribute
5 years ago
512e1c18a7
Update to latest JSDOM
6 years ago
e8bb7f722f
Fix whitespace normalization in title metadata
...
When switching to a newer version of JSDOM, it is more literal
about listing whitespace as part of textContent, including
newlines and not normalizing multiple spaces.
It seems prudent to just always normalize whitespace for titles,
which are guaranteed to be pretty short anyway.
6 years ago
3610476663
Remove CSS that jsdom struggles to parse
6 years ago
2620542dd1
Split off isProbablyReaderable implementation
6 years ago
8c41d92560
Assert on siteName in all test cases
6 years ago
1bac47c70d
Add newly generated test case
6 years ago
5a69d4a8eb
Improve metadata extraction ( #478 )
...
* Improve metadata extraction
* Recognize meta[property] as a space-separated list
* Recognize Dulin Core (dc|dcterm): metadata.
* Prefer Dublin Core, Open Graph, Twitter, and HTML in that order.
* _getArticleTitle() is now only used as fallback if document
doesn't provide good metadata.
6 years ago
f782bc5f06
Avoid global flag when looking for metadata using regexes
6 years ago
30611cc57f
Fix quotes issues in test and benchmark files
6 years ago
8fec62d246
Strip XML namespaces from tag names to deal with broken serializations
6 years ago
ea4165721f
Remove single-cell tables
6 years ago
bf64b58d90
Update tests
6 years ago
68c9af4ffa
Use numeric encoding for non-XML entities
...
JSDOMParser can't handle HTML named entities like ` `
6 years ago
d4b842c82a
Match headings on trimmed strings to avoid whitespace causing mismatches
6 years ago
8c02a0d34c
Fix #283 and remove hidden nodes
6 years ago
656a6673d9
Don't put non-phrasing content into paragraphs
6 years ago
5ae90930cd
Don't convert DIVs to Ps when more than 25% links
6 years ago
9f2c5cb42e
Put phrasing content into paragraphs
...
This removes the need for `p.readability-styled` elements.
6 years ago
c823a6efb2
Fix generate-testcase.js
6 years ago
f4ab856992
Check for a document being passed
...
This provides a descriptive error message if no document is passed, and
ignores the first argument if the second argument looks like
a reasonable DOM document instance.
6 years ago
7a24801958
Don't include root html node in candidates
...
Fixes #435
6 years ago
acfd3759a1
Generate XHTML-compatible input for test cases
...
Fixes the bug noted in the README
6 years ago
d60184966c
Remove unused URI parameter from constructor
6 years ago
3c76104adb
Fix engadget test case
6 years ago
4b99f41ec9
Add engadget test case
6 years ago
6c5bc62959
Remove aside tags on test cases
6 years ago
f8d9b1c224
Update test expectations
6 years ago
8414158fa9
Fix _replaceBrs
...
Previously, `nextElem` was not actually proceeding to the next element, and therefore aborting the paragraph at the first `<br>` (rather than the first `<br><br>` as the comment indicates).
6 years ago
8525c6af36
Fix relative URIs given <base> tags ( #422 )
6 years ago
d598baf02b
Improve URL handling in JSDOMParser and Readability.js
...
This change ups the required node version to 7.0 because it relies on the builtin url module.
We now pass a url when constructing a jsdom document or JSDOMParser document.
Because this is an API change, I'm increasing the package version.
Ultimately, I would like to remove the argument from the readability constructor. It should
use the documentURI from the document it is passed.
6 years ago
834672ef86
Return longest text after failing to detect text longer than the configured value ( #423 )
...
Save extracted text across attempts and return the longest one when all attempts fail, and add a test case from hukumusume
6 years ago
PalmerAL
Gijs Kruitbosch
Radhi
EvsChen
Andres Rey
Maria Luiza Soares
Daniel Aleksandersen
David A Roberts
Andres Rey
Brad Philips
Andres Rey