CHANGELOG.txt in Link checker 6.2

linkchecker 6.x-2.9, August 6, 2016
-----------------------------------

* Upgrade outdated HTTP user agents.
* Issue #2664260 by hass: num_rows only exists in mysqli
* Fixed two typos
* Issue #2303177: "Back to Top" Page anchor being reported as a 404 error

linkchecker 6.x-2.8, June 7, 2014
---------------------------------

* Issue #2280827: Notice: Undefined index: content-type in _linkchecker_status_handling()
* Issue #2125719 by hass: Links in unpublished nodes are still checked
* Issue #2088461 by fonant, hass: PDF link with #page= results in "URL fragment identifier not found in content"
* Added Smileys filter (http://drupal.org/project/smiley) to filter blacklist
* Issue #1306498 by hass, evolvingweb: Remove unnecessary hook_requirements in linkchecker.install 
* Issue #1923328 by larowlan: Fragment check fails when source page contains spaces between attribute name, equals and value.
* Issue #1926558 by hass: Block extraction re-scans fail.

linkchecker 6.x-2.7, February 14, 2013
--------------------------------------

* Issue #1901018 by hass: ERROR - __clone method called on non-object.
* #1891900: Uninstall of variable 'linkchecker_check_connections_max' failed.
* #1719174: Editing node throws "Undefined property" stdClass::$nid in hook_node_prepare().
* #1875602: Check URL fragment identifiers in content
* Minor code style fixes

linkchecker 6.x-2.6, December 24, 2012
--------------------------------------

* Fixed some typos.
* Upgraded outdated HTTP user agents.
* #380052: Add support with non-blocking parallel link checking.
* linkchecker_update_6211 was missing an update message.
* #1869924: Strict warning: Creating default object from empty value.
* #1867460: Prevent save on automatic updates, if content has not changed.
* #1804842 follow up: Exit if all node types are disabled
* #1450672: Cron task does not impersonate to admin user on automatic content updates.
* Added Drupal 6.14 warning to known issues.
* #1804842: Lots of warnings when comment module is not installed
* #1811908: Allow extracting protocol relative url's
* #1586022: linkchecker generating bad User-Agent string
* Status code 500 on HEAD/GET does not increase fail_count
* #1488572: i18n: "Permission restrictions deny" all broken links in other languages

linkchecker 6.x-2.5, February 15, 2012
--------------------------------------

* Fix for access bypass vulnerability.
* #1429284: Don't follow redirects
* Removed block 'title' for consitency reasons. It's only a title in administration and should not contain URLs
* Show a recommendation next to blacklisted filter names.
* #1149596: Links with spaces are not extracted
* Backported some documentation fixes
* Batch redirect URL added for blocks
* #860700: Disable all non-http(s) links for link checking until we have curl implemented
* #1054118: After batch scan the target url is not pointing to the node view
* #1075390: Added HTTP range support (Performance)
* Links disabled for checking (blacklist, never checked) will be counted as as unchecked and the number of unchecked links will therefore never become zero.
* #1321378: Improve performance of queries
* #953958: User able to see other user broken links
* truncate is ***much*** more faster than delete.
* #725398: variable_get() may not be available in the install.php run
* First decode, than filter as URLs could theoretically consist of encoded blanks only.
* #1067160 by somimi: Minor grammar fixes to some of the descriptions in linkchecker.pages.inc
* Removing translation directories
* #903318: Internal Links in cck, weblinks and links fields are not extracted.
* Remove locale module setUp as notices in _linkchecker_add_node_links() has been fixed.
* Prevent notices if tests run without locale module.
* #780242 by AlexisWilke: Several E_NOTICE when dealing with the URLs
* #500090 by hass: Block removal of RFC domains in exclude links
* #685580 by wheyse: CCK field_name need to be checked with empty() as the value could be empty
* Upgrade line break converter from D5 filter/2 to D6 filter/1.
* #661484: Empty URL exclusion list disables link checking

linkchecker 6.x-2.4, November 24, 2009
--------------------------------------

* #566388 by wimh, hass: Re-check link status on next cron run
* #632732 by sleepcamel, hass: Add a message to $node->log when saving a new revision
* #627774 by charmer: Linkchecker Czech Translation
* Replace word "node" in text with "content" for easier translation
* #580248: Prevent link check failure if 'max_execution_time' = 0 (unlimited).
* Blacklisted "smileys/0" filter as the URLs are maintained in the Smiley module settings
* Fixed wrong variable name in single comment import function
* Updated README.txt
* Cleanup / refactoring _linkchecker_extract_links()
* #563464: Internal links reported as broken incorrectly on SSL only site. Use $base_root as site URL for prefixing absolute internal URLs.
* E_ALL saver...
* Better menu item description, D7 backport.
* Dynamic destination for user broken links page
* Use theme function for item_list
* Dynamic colspan for no data
* Joined wrong fields in user report with comments enabled
* Filters do not run on content as expected.
* db_last_insert_id('boxes', 'bid') could return a wrong value if other contrib modules add block hooks and insert something in other tables.
* Make all count() upper case
* Show items in a list
* Filter $link->error for arbitrary code
* #532178 by hass: Database update #6209 failed
* #196862 by hass: COUNT(*) is an expensive query in InnoDB.
* Added links extraction .test
* E_ALL: Fixed notice for undefined $node->language.
* E_ALL: Fixed notice caused by array_map(md5, $links)
* Typo fixed
* Changed unique key and column name 'token' into 'urlhash'.
* Added filter blacklist for link extraction.
* Remove reference to $form_state in hook_form_alter()
* Fix comments
* Add $node->type check
* Added support for node type 'weblinks' from 'weblinks' modules
* D5: check_plain() all '@' string values for the use with strtr().


linkchecker 6.x-2.3, August 1, 2009
-----------------------------------

By this release linkchecker no longer depends on job_queue!

* by wimh: Copy an paste error, used _linkchecker_node_links_missing() in _linkchecker_add_comment_links()
* #536744 by wimh, hass: Add broken links warning message to block edit page on view, but not after save.
* #536744 by wimh, hass: Fix comment edit form to show the broken links warning message only on 'view' and 'preview', but not after save.
* Fix inline documenation typos
* Code style, variable names, constant for limit, doc fixes, others
* #536672 by wimh, hass: Link references not updated when last link of a node/comment/block is removed
* More comments added.
* Updated translation files
* Use update_sql() for permission updates
* Save form settings on link scan and clear.
* #497096: Support links generated by input filter
* Remove duplicate lid's in SQL results
* #465414: Followup for relative links extracted incorrectly
* Note that custom link settings are lost on every clean and re-analyse task
* Added GUI for link check URL filter
* #524380: Do not check example.com/net/org domain links
* Updated installation steps in README.txt
* Added sample URLs for internal and external to description
* Changed permission name "access linkchecker" to "access broken links report"
* Added the status code 302 to the list of ignored response codes.
* Select from comment table only, if comment modules is active.
* Remove ignored links from broken links view


linkchecker 6.x-2.2, July 14, 2009
----------------------------------

* Removed confusing watchdog message.
* Use default node type revision setting and save a revision on 301 auto-update.
* UX: Provide more descriptive help what the HTML tags are for.
* #451456: 301 auto-update could break links
* #452080: Update permanently moved internal links is broken
* Removed duplicate _linkchecker_add_xxx_links() calls. node_save automaticallly calls hook_nodeapi and comment_save automatically calls hook_comment with 'update'.
* Added a link to the job_queue module project page
* #465414: Stop incorrect extraction of relative links
* Typo fixed in batch finished message
* Removed unused default case
* Removed UI setting no longer having a real use case (and reduce complexity)
* Deleted the associated variable.
* One more chance to extract a URL (for e.g. Java Applet)
* Added support for HTML5 tags (video|audio|source)
* Uncomment cURL requirement until implemented.
* #402742: Integration with path_redirect
* #427906, #417354: Extract links from various configurable HTML tags (a|area|img|iframe|object|param|embed)
* #504332: Cannot install, HTTP request status fails
* Show edit link only if user have permission to change settings
* New permission to edit link settings
* Add UI to change request method
* Add ignore filter for links pointing to buggy servers


linkchecker 6.x-2.1, March 26, 2009
-----------------------------------

* #409324 by Dave Reid: _linkchecker_delete_node_links() expects a nid, but passed a node object
* #386612 by hass: Prevent possible cron failures caused by too many link checks
* #408860 by Dave Reid: PHP notice: Undefined property: stdClass::$moderate in linkchecker.module on line 428.
* #373485 by hass: DB errors when adding a new node
* #387228: linkchecker.module fails in PHP4 on line 910
* #375767: Wrong table aliases and column names for comments and boxes
* Added MySQL 5.0 requirement check
* Show broken link message on node prepare, not view
* #380548: No results after configuration changes
* #380476: Descriptive text in Admin confusing
* #381986: Broken link does not immediately disappear in broken links list after link has been fixed in content
* #380872 by hass: Show informative message about current check status on reports page.