All releases of LinkChecker
Release Notes: A new GUI client for checking has been added, and invalid handling of persistent connections has been fixed. Python >= 2.5 is now required.
Release Notes: This release adds options to check the HTML and CSS syntax of downloaded pages, and parses Shockwave Flash (SWF) files for URLs to check. Pressing Ctrl-C twice aborts the program immediately. Also included are minor bugfixes regarding HTTP pipelining, HTML parsing, and mail address checks.
Release Notes: This release correctly handles HTTP/1.1 persistent connections, ignores commented out CSS links, and fixes open file leaks. The documentation has been improved by fixing typographic errors, documenting the --configfile option, and adding a linkcheckerrc(5) manual page.
Release Notes: Log output with certain Unicode characters is fixed now. XML output has been improved. Gopher URLs are now deprecated.
Release Notes: Errors during log file opening are now ignored. A problem with certain URL redirects has been fixed. The default config file syntax no longer has broken comment lines.
Release Notes: HTTP connections are now reused properly, which fixes the occasional empty result warning and bad file descriptor errors. Keyboard interrupts are ignored when the application is already terminating. The JavaScript URL syntax check of the CGI Web interface now allows digits and underscores. Old references to the Psyco optimizer have been removed from the documentation.
Release Notes: The CGI script does not restrict the recursion limit anymore. The HTML documentation is now included in the source tarball.
Release Notes: When running as the root user under Unix systems, privileges are dropped before checking starts. The --profile option has been fixed. The robots.txt parser enforces the character set encoding. Base reference URLs are now always normed correctly. On systems where the default character encoding could not be determined, a fallback encoding is now used.
Release Notes: When this release exits, it waits for all background checker threads to stop in order to prevent shutdown errors.
Release Notes: The threading algorithm has been improved, and all previously known race conditions have been fixed. A new option --cookiefile parses initial cookie data to be sent to HTTP servers. The "Crawl-delay" option of robots.txt files is now supported. The --pause option only waits between two requests to the same server, not between all requests. DNS lookups for HTTP servers are now cached. The output for the DOT graph format has been fixed. URLs with invalid syntax are now cached and are matched by the --ignore-url option.
