@@ -992,19 +992,6 @@ public function getProcessReport(): PHPCrawlerProcessReport
992
992
return $ Report ;
993
993
}
994
994
995
- /**
996
- * Retruns an array with summarizing report-information after the crawling-process has finished
997
- *
998
- * For detailed information on the conatining array-keys see PHPCrawlerProcessReport-class.
999
- *
1000
- * @deprecated Please use getProcessReport() instead.
1001
- * @section 11 Deprecated
1002
- */
1003
- public function getReport (): array
1004
- {
1005
- return $ this ->getProcessReport ()->toArray ();
1006
- }
1007
-
1008
995
/**
1009
996
* Overridable method that will be called after the header of a document was received and BEFORE the content
1010
997
* will be received.
@@ -1324,19 +1311,6 @@ public function addContentTypeReceiveRule($regex): bool
1324
1311
return $ this ->PageRequest ->addReceiveContentType ($ regex );
1325
1312
}
1326
1313
1327
- /**
1328
- * Alias for addContentTypeReceiveRule().
1329
- *
1330
- * @section 11 Deprecated
1331
- * @param $regex
1332
- * @return bool
1333
- * @deprecated
1334
- */
1335
- public function addReceiveContentType ($ regex ): bool
1336
- {
1337
- return $ this ->addContentTypeReceiveRule ($ regex );
1338
- }
1339
-
1340
1314
/**
1341
1315
* Adds a rule to the list of rules that decide which URLs found on a page should be followd explicitly.
1342
1316
*
@@ -1386,32 +1360,6 @@ public function addURLFilterRule($regex): bool
1386
1360
return $ this ->UrlFilter ->addURLFilterRule ($ regex );
1387
1361
}
1388
1362
1389
- /**
1390
- * Alias for addURLFollowRule().
1391
- *
1392
- * @section 11 Deprecated
1393
- * @param $regex
1394
- * @return bool
1395
- * @deprecated
1396
- */
1397
- public function addFollowMatch ($ regex ): bool
1398
- {
1399
- return $ this ->addURLFollowRule ($ regex );
1400
- }
1401
-
1402
- /**
1403
- * Alias for addURLFilterRule().
1404
- *
1405
- * @section 11 Deprecated
1406
- * @param $regex
1407
- * @return bool
1408
- * @deprecated
1409
- */
1410
- public function addNonFollowMatch ($ regex ): bool
1411
- {
1412
- return $ this ->addURLFilterRule ($ regex );
1413
- }
1414
-
1415
1363
/**
1416
1364
* Adds a rule to the list of rules that decides what types of content should be streamed diretly to a temporary file.
1417
1365
*
@@ -1445,19 +1393,6 @@ public function addStreamToFileContentType($regex): bool
1445
1393
return $ this ->PageRequest ->addStreamToFileContentType ($ regex );
1446
1394
}
1447
1395
1448
- /**
1449
- * Has no function anymore.
1450
- *
1451
- * Please use setWorkingDirectory()
1452
- *
1453
- * @param $tmp_file
1454
- * @deprecated This method has no function anymore since v 0.8.
1455
- * @section 11 Deprecated
1456
- */
1457
- public function setTmpFile ($ tmp_file ): void
1458
- {
1459
- }
1460
-
1461
1396
/**
1462
1397
* Defines whether the crawler should parse and obey robots.txt-files.
1463
1398
*
@@ -1496,35 +1431,6 @@ public function obeyRobotsTxt($mode, $robots_txt_uri = null): bool
1496
1431
return true ;
1497
1432
}
1498
1433
1499
- /**
1500
- * Alias for addStreamToFileContentType().
1501
- *
1502
- * @param $regex
1503
- * @return bool
1504
- * @deprecated
1505
- * @section 11 Deprecated
1506
- */
1507
- public function addReceiveToTmpFileMatch ($ regex ): bool
1508
- {
1509
- return $ this ->addStreamToFileContentType ($ regex );
1510
- }
1511
-
1512
- /**
1513
- * Has no function anymore!
1514
- *
1515
- * This method was redundant, please use addStreamToFileContentType().
1516
- * It just still exists because of compatibility-reasons.
1517
- *
1518
- * @param $regex
1519
- * @return bool
1520
- * @deprecated This method has no function anymore since v 0.8.
1521
- * @section 11 Deprecated
1522
- */
1523
- public function addReceiveToMemoryMatch ($ regex ): bool
1524
- {
1525
- return true ;
1526
- }
1527
-
1528
1434
/**
1529
1435
* Sets a limit to the total number of requests the crawler should execute.
1530
1436
*
@@ -1552,20 +1458,6 @@ public function setRequestLimit($limit, $only_count_received_documents = false):
1552
1458
return true ;
1553
1459
}
1554
1460
1555
- /**
1556
- * Alias for setRequestLimit() method.
1557
- *
1558
- * @section 11 Deprecated
1559
- * @param $limit
1560
- * @param bool $only_count_received_documents
1561
- * @return bool
1562
- * @deprecated Please use setRequestLimit() method!
1563
- */
1564
- public function setPageLimit ($ limit , $ only_count_received_documents = false ): bool
1565
- {
1566
- return $ this ->setRequestLimit ($ limit , $ only_count_received_documents );
1567
- }
1568
-
1569
1461
/**
1570
1462
* Sets the content-size-limit for content the crawler should receive from documents.
1571
1463
*
@@ -1629,19 +1521,6 @@ public function enableCookieHandling($mode): bool
1629
1521
return true ;
1630
1522
}
1631
1523
1632
- /**
1633
- * Alias for enableCookieHandling()
1634
- *
1635
- * @section 11 Deprecated
1636
- * @param $mode
1637
- * @return bool
1638
- * @deprecated Please use enableCookieHandling()
1639
- */
1640
- public function setCookieHandling ($ mode ): bool
1641
- {
1642
- return $ this ->enableCookieHandling ($ mode );
1643
- }
1644
-
1645
1524
/**
1646
1525
* Enables or disables agressive link-searching.
1647
1526
*
@@ -1664,19 +1543,6 @@ public function enableAggressiveLinkSearch($mode): bool
1664
1543
return $ this ->PageRequest ->enableAggressiveLinkSearch ($ mode );
1665
1544
}
1666
1545
1667
- /**
1668
- * Alias for enableAggressiveLinkSearch()
1669
- *
1670
- * @section 11 Deprecated
1671
- * @param $mode
1672
- * @return bool
1673
- * @deprecated Please use enableAggressiveLinkSearch()
1674
- */
1675
- public function setAggressiveLinkExtraction ($ mode ): bool
1676
- {
1677
- return $ this ->enableAggressiveLinkSearch ($ mode );
1678
- }
1679
-
1680
1546
/**
1681
1547
* Sets the list of html-tags the crawler should search for links in.
1682
1548
*
@@ -1698,24 +1564,6 @@ public function setLinkExtractionTags($tag_array): bool
1698
1564
return $ this ->PageRequest ->setLinkExtractionTags ($ tag_array );
1699
1565
}
1700
1566
1701
- /**
1702
- * Sets the list of html-tags from which links should be extracted from.
1703
- *
1704
- * This method was named wrong in previous versions of phpcrawl.
1705
- * It does not ADD tags, it SETS the tags from which links should be extracted from.
1706
- *
1707
- * Example
1708
- * <code>$crawler->addLinkExtractionTags("href", "src");</code>
1709
- *
1710
- * @section 11 Deprecated
1711
- * @deprecated Please use setLinkExtractionTags()
1712
- */
1713
- public function addLinkExtractionTags (): bool
1714
- {
1715
- $ tags = func_get_args ();
1716
- return $ this ->setLinkExtractionTags ($ tags );
1717
- }
1718
-
1719
1567
/**
1720
1568
* Adds a basic-authentication (username and password) to the list of basic authentications that will be send with requests.
1721
1569
*
@@ -1752,19 +1600,6 @@ public function setUserAgentString($user_agent): bool
1752
1600
return true ;
1753
1601
}
1754
1602
1755
- /**
1756
- * Has no function anymore.
1757
- *
1758
- * Thes method has no function anymore, just still exists because of compatibility-reasons.
1759
- *
1760
- * @section 11 Deprecated
1761
- * @param $mode
1762
- * @deprecated
1763
- */
1764
- public function disableExtendedLinkInfo ($ mode ): void
1765
- {
1766
- }
1767
-
1768
1603
/**
1769
1604
* Sets the working-directory the crawler should use for storing temporary data.
1770
1605
*
0 commit comments