diff options
| author | ArthurHoaro <arthur@hoa.ro> | 2016-01-04 10:45:54 +0100 |
|---|---|---|
| committer | ArthurHoaro <arthur@hoa.ro> | 2016-01-11 21:19:31 +0100 |
| commit | 1557cefbd76257ceb830f65806831b490faf0acc (patch) | |
| tree | 787f6d8fdabe8ea2fc0c37b61d616e667cdfbda5 /tests/LinkUtilsTest.php | |
| parent | c0a50f3663e207d5df007e0fa321219c1b32d6ea (diff) | |
| download | Shaarli-1557cefbd76257ceb830f65806831b490faf0acc.tar.gz Shaarli-1557cefbd76257ceb830f65806831b490faf0acc.tar.zst Shaarli-1557cefbd76257ceb830f65806831b490faf0acc.zip | |
Fixes #410 - Retrieve title fails in multiple cases
* `get_http_url()` renamed to `get_http_response()`.
* Use the same HTTP context to retrieve response headers and content.
* Follow HTTP 301 and 302 redirections to retrieve the title (default max 3 redirections).
* Add `LinkUtils` to extract titles and charset.
* Try to retrieve charset from HTTP headers first (new), then HTML content.
* Use mb_string to re-encode title if necessary.
Diffstat (limited to 'tests/LinkUtilsTest.php')
| -rw-r--r-- | tests/LinkUtilsTest.php | 85 |
1 files changed, 85 insertions, 0 deletions
diff --git a/tests/LinkUtilsTest.php b/tests/LinkUtilsTest.php new file mode 100644 index 00000000..c2257590 --- /dev/null +++ b/tests/LinkUtilsTest.php | |||
| @@ -0,0 +1,85 @@ | |||
| 1 | <?php | ||
| 2 | |||
| 3 | require_once 'application/LinkUtils.php'; | ||
| 4 | |||
| 5 | /** | ||
| 6 | * Class LinkUtilsTest. | ||
| 7 | */ | ||
| 8 | class LinkUtilsTest extends PHPUnit_Framework_TestCase | ||
| 9 | { | ||
| 10 | /** | ||
| 11 | * Test html_extract_title() when the title is found. | ||
| 12 | */ | ||
| 13 | public function testHtmlExtractExistentTitle() | ||
| 14 | { | ||
| 15 | $title = 'Read me please.'; | ||
| 16 | $html = '<html><meta>stuff</meta><title>'. $title .'</title></html>'; | ||
| 17 | $this->assertEquals($title, html_extract_title($html)); | ||
| 18 | } | ||
| 19 | |||
| 20 | /** | ||
| 21 | * Test html_extract_title() when the title is not found. | ||
| 22 | */ | ||
| 23 | public function testHtmlExtractNonExistentTitle() | ||
| 24 | { | ||
| 25 | $html = '<html><meta>stuff</meta></html>'; | ||
| 26 | $this->assertFalse(html_extract_title($html)); | ||
| 27 | } | ||
| 28 | |||
| 29 | /** | ||
| 30 | * Test get_charset() with all priorities. | ||
| 31 | */ | ||
| 32 | public function testGetCharset() | ||
| 33 | { | ||
| 34 | $headers = array('Content-Type' => 'text/html; charset=Headers'); | ||
| 35 | $html = '<html><meta>stuff</meta><meta charset="Html"/></html>'; | ||
| 36 | $default = 'default'; | ||
| 37 | $this->assertEquals('headers', get_charset($headers, $html, $default)); | ||
| 38 | $this->assertEquals('html', get_charset(array(), $html, $default)); | ||
| 39 | $this->assertEquals($default, get_charset(array(), '', $default)); | ||
| 40 | $this->assertEquals('utf-8', get_charset(array(), '')); | ||
| 41 | } | ||
| 42 | |||
| 43 | /** | ||
| 44 | * Test headers_extract_charset() when the charset is found. | ||
| 45 | */ | ||
| 46 | public function testHeadersExtractExistentCharset() | ||
| 47 | { | ||
| 48 | $charset = 'x-MacCroatian'; | ||
| 49 | $headers = array('Content-Type' => 'text/html; charset='. $charset); | ||
| 50 | $this->assertEquals(strtolower($charset), headers_extract_charset($headers)); | ||
| 51 | } | ||
| 52 | |||
| 53 | /** | ||
| 54 | * Test headers_extract_charset() when the charset is not found. | ||
| 55 | */ | ||
| 56 | public function testHeadersExtractNonExistentCharset() | ||
| 57 | { | ||
| 58 | $headers = array(); | ||
| 59 | $this->assertFalse(headers_extract_charset($headers)); | ||
| 60 | |||
| 61 | $headers = array('Content-Type' => 'text/html'); | ||
| 62 | $this->assertFalse(headers_extract_charset($headers)); | ||
| 63 | } | ||
| 64 | |||
| 65 | /** | ||
| 66 | * Test html_extract_charset() when the charset is found. | ||
| 67 | */ | ||
| 68 | public function testHtmlExtractExistentCharset() | ||
| 69 | { | ||
| 70 | $charset = 'x-MacCroatian'; | ||
| 71 | $html = '<html><meta>stuff2</meta><meta charset="'. $charset .'"/></html>'; | ||
| 72 | $this->assertEquals(strtolower($charset), html_extract_charset($html)); | ||
| 73 | } | ||
| 74 | |||
| 75 | /** | ||
| 76 | * Test html_extract_charset() when the charset is not found. | ||
| 77 | */ | ||
| 78 | public function testHtmlExtractNonExistentCharset() | ||
| 79 | { | ||
| 80 | $html = '<html><meta>stuff</meta></html>'; | ||
| 81 | $this->assertFalse(html_extract_charset($html)); | ||
| 82 | $html = '<html><meta>stuff</meta><meta charset=""/></html>'; | ||
| 83 | $this->assertFalse(html_extract_charset($html)); | ||
| 84 | } | ||
| 85 | } | ||