Hopiu/linkchecker
1
0
Fork 0
mirror of https://github.com/Hopiu/linkchecker.git synced 2026-04-03 14:40:32 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions
linkchecker/debian/changelog
calvin a2ceebf3eb python 2.0 fix 
git-svn-id: https://linkchecker.svn.sourceforge.net/svnroot/linkchecker/trunk/linkchecker@224 e7d03fd6-7b0d-0410-9947-9c21f3af8025
2001-01-11 20:32:11 +00:00

457 lines
15 KiB
Text
Raw Blame History

linkchecker (1.2.14) unstable; urgency=low
* fixed Python 2.0 bug: TypeError: object has read-only attributes
-- Bastian Kleineidam <calvin@users.sourceforge.net> Thu, 11 Jan 2001 21:29:00 +0100
linkchecker (1.2.13) unstable; urgency=low
* linkcheck/HttpUrlData.py:
- better redirection handling
- really use host variable in "Host:" header
- support response code "305 Use proxy"
* linkcheck/robotparser2.py:
- better redirection handling
- cope with user-agent: lines without a preceding blank line
- printed "Allow" when I meant "Disallow". This has been fixed.
- two licenses: GPL and Python 2.0
* linkcheck/__init__.py: user friendly warning for keyboard interrupts
* debian/control: standards version 3.2.1
* support Mozilla-specific find: links
* print link names for:
- <a href> links (<a href="..">link name</a>)
- <img> links (the alt= attribute)
this makes it easier for bookmarks.html files to find the link
WARNING: this changed the output format!
-- Bastian Kleineidam <calvin@users.sourceforge.net> Sun, 7 Jan 2001 12:51:19 +0100
linkchecker (1.2.12) unstable; urgency=low
* MANIFEST.in: include rpm_build_script in source distribution
* linkcheck/FileUrlData.py:
- use only expanduser, not normpath in FileUrl constructor. I dont
want to norm because the user gave me this URL and he should know
better than me.
* debian/{postinst,prerm}: use #!/bin/sh -e because it is shorter
than set -e
* linkcheck/Config.py: use append(...) for list of config files, not
insert(0,...)
* linkcheck/Logging.py: tidy HTML output:
- use summary attribute in HTML tables
- fix <center> positioning
- use DOCTYPE tag
- use CSS (advantage: makes smaller file size, looks good)
- finally found the &amp; bug: dont use StringUtil.htmlify
on the url names.
-- Bastian Kleineidam <calvin@users.sourceforge.net> Fri, 22 Dec 2000 14:41:12 +0100
linkchecker (1.2.11) unstable; urgency=low
* debian/rules:
- use dh_installexamples to install example files in the correct
directory
- move test files into a common directory
- move linkcheckerrc in /etc, make no symlink
* rpm_build_script: shell script to help build an RPM package
* setup.cfg: use rpm_build_script
* MANIFEST.in: exclude LinkCheckerConf.py from distribution to enforce
a run of 'python setup.py config'
-- Bastian Kleineidam <calvin@users.sourceforge.net> Sun, 17 Dec 2000 02:22:58 +0100
linkchecker (1.2.10) unstable; urgency=low
* broken *.mo i18n files fixed
* test/test1.html: add test for missing quotes
* debian/control: updated description field
* moved all single .py files in linkcheck/ except LinkCheckerConf.py
which is automatically generated
-- Bastian Kleineidam <calvin@users.sourceforge.net> Thu, 14 Dec 2000 23:50:01 +0100
linkchecker (1.2.9) unstable; urgency=low
* changed specification and better documentation of intern/extern
checking; see the FAQ for detailed information
* linkchecker.1: added a man page
* setup.py: overloaded distutils debugging function dump_dirs
* setup.py: handle man pages correctly with bdist_rpm target
* linkcheck/Config.py: correct order with user/password authentication
entries (report and fix by Jamie Heilman <jamie@audible.transient.net>)
* linkchecker: default password is joe@, not guest@ (report and fix by
Jamie Heilman <jamie@audible.transient.net>)
* test/*.html: comments in the test suite files
* Config.py: dont use internals of the UserDict class
-- Bastian Kleineidam <calvin@users.sourceforge.net> Mon, 4 Dec 2000 11:56:03 +0100
linkchecker (1.2.8) unstable; urgency=low
* INSTALL: more documentation for the CGI scripts
* Makefile: better cleaning (clean, cleandeb, distclean)
* XML output (idea and patch from Nicolas Chauvat
<Nicolas.Chauvat@logilab.fr>)
-- Bastian Kleineidam <calvin@users.sourceforge.net> Wed, 15 Nov 2000 23:27:37 +0100
linkchecker (1.2.7) unstable; urgency=low
* new robot.txt parser module which is interface compatible with the
one in Python 2.0
* debian/control: new fields Build-Depends, Build-Depends-Indep
* debian/control: Architecture is any, not i386
* debian/control: Standards version 3.1.1,
* debian/control: depend on Python 1.5.2
* debian/copyright: first line was too long
* debian/postinst: compile .py files
* debian/rules: new configuration target
* debian/prerm: new file, delete compiled .py files
* use Python tools for i18n (backported from Python 2.0)
* proxy configuration is now detected automatically from system
(environment) variables. NOTE: this means the --proxy options
are gone!
* add <script src=> to url list
* include ssl.c again (was missing in 1.2.6)
-- Bastian Kleineidam <calvin@users.sourceforge.net> Sat, 4 Nov 2000 10:45:08 +0100
linkchecker (1.2.6) unstable; urgency=low
* made a FAQ
* configuration changes: distutils are now required; because of that
we have no more .tmpl files
* fix db name in create.sql
* fix tag parsing when a quoted tag attribute value contains a >
character
-- Bastian Kleineidam <calvin@users.sourceforge.net> Sat, 28 Oct 2000 17:52:53 +0200
linkchecker (1.2.5) unstable; urgency=low
* fix /etc/resolv.conf parsing again (this time for real?)
* tuple param in DNS/Lib.py for Python 1.6 compatibility
(Adam Feuer <adamf@pobox.com>)
* adjust #! line in linkchecker script to correct python interpreter
(Adam Feuer <adamf@pobox.com>)
* proper proxy config handling
* distribution restructuring (now with .rpm and .zip packages)
* catch EOFError with ftp links (reported by Stan Mulder
<stan@intrepidsoftware.com>)
-- Bastian Kleineidam <calvin@users.sourceforge.net> Mon, 11 Sep 2000 10:04:32 +0200
linkchecker (1.2.4) unstable; urgency=low
* fixed parsing of /etc/resolv.conf with empty lines
* french translation from Jerome Couderc <j.couderc@ifrance.com>
* fix for HTTP HEAD requests from bad/dumb/old servers
* CGI script fixes
* LinkChecker Online HTML pages added to source
* news: link support for links which specify a NNTP server
* fixed parsing of extern? config file options (was endless loop)
* look more carefully for Zope servers
-- Bastian Kleineidam <calvin@users.sourceforge.net> Sat, 19 Aug 2000 17:17:12 +0200
linkchecker (1.2.3) unstable; urgency=low
* typo fix for adjustWinPath
* added some source code documentation
* improved error messages for wrong options
* configuration file options for logger output
* linkchecker.bat installation support for Windows
* included test suite in distribution
* blacklist output support
* CSV output support
* SSL autodetection in setup.py
* added GPL copyright header to each of my .py files
* i18n support
* german translation
* use http_proxy environment variable if present
* be RFC822 and RFC2368 compliant when scanning mail syntax
* fix for incorrect line number in logger output (reported by Michael
Schmitz)
* Debian package is now lintian clean
* Only catch some exceptions in main check loop so the KeyboardInterrupt
exception propagates through
* Renice the main thread loop by sleep()ing some time
* New function config.reset()
* fix dead lock with news: urls
* print how much links we checked
-- Bastian Kleineidam <calvin@users.sourceforge.net> Thu, 22 Jun 2000 11:23:49 +0200
linkchecker (1.2.2) unstable; urgency=low
* fix for anchor checking (Martin Herbener)
-- Bastian Kleineidam <calvin@users.sourceforge.net> Thu, 13 Apr 2000 00:08:18 +0200
linkchecker (1.2.1) unstable; urgency=low
* fix for Netscape Enterprise servers
* fixed congif typo (Thomas Stinner <t.stinner@billiton.de>)
* fix for authentication (Thomas Stinner <t.stinner@billiton.de>)
-- Bastian Kleineidam <calvin@users.sourceforge.net> Mon, 10 Apr 2000 18:53:52 +0200
linkchecker (1.2.0) unstable; urgency=low
* stable release
* fixed usage text glitches
-- Bastian Kleineidam <calvin@users.sourceforge.net> Sat, 8 Apr 2000 07:09:23 +0200
linkchecker (1.1.4) unstable; urgency=low
* Merged changelogs
-- Bastian Kleineidam <calvin@users.sourceforge.net> Fri, 24 Mar 2000 19:59:08 +0100
6.4.2000
* forgot to include linkchecker script in .deb package
30.3.2000
* support for NNTP news: links
* forgot to put lc.fcgi and lc.sz_fcgi in release
30.3.2000 Version 1.1.4
* fixed missing self.mime assignment in HttpUrlData.py
28.3.2000
* fixed wrong position of configuration file option in Config.py
* option -W is now -F, -W is a new option for inspecting
the content of a link
* Remove the distutils package. You have to download it now
separately if you want to. Otherwise install manually.
* cache the contents of links while checking
27.3.2000
* Fixed misspelled exception in UrlData.py (BugTracker)
* Fixed wrong _time argument in HttpsUrlData.py
(Peter <peterw@waybeat.com.au>)
* Fixed missing env argument in lc.cgi
* allow a quote (') in mail adresses (e.g. o'hara@foo.com)
26.3.2000
* FastCGI fixes
* simplified sz_fcgi.py
25.3.2000
* FastCGI script fixes
* do not close the logging file descriptor
* patch for fcgi.py: flush output immediately
24.3.2000
* first version of FastCGI script
* do not install distutils
22.3.2000 Version 1.1.3
* added README,USAGE for distutils
21.3.2000 Version 1.1.2
* distutils fixes
* linkchecker.bat: quote the path
20.3.2000
* fix a bug in reporting download time
* added the distutils package
19.3.2000
* report the duration of checking a link (Check Time)
* rename httplib.py to http11lib.py so it does not silently
replace the standard library httplib.py
12.3.2000
* first version of distutils setup.py
11.3.2000
* Better debugging:
-D implies -t0
HTML data in debug output
print line number in UrlData.__str__()
* set HTTP Host: header for virtual domains
comment out the Host: header in httplib.py
* only send HTTP authorization header when its needed
8.3.2000 Version 1.1.1
* FastCGI modules added (no script yet)
* CGI script fixes
* supply strict/non-strict flag for each external filtering rule
7.3.2000
* support for multiple user/password pairs
3.3.2000
* ignore HTTP status replies < 200 because some pages return -1
and binary data (Gabor Liptak <gaborliptak@usa.net>)
* if HEAD reveals MIME type "application/octet", use GET method.
this is needed for Zope pages (Bug Tracker submission)
2.3.2000
* two small comment changes in GML and SQL output
* linkchecker -V works again :)
29.2.2000 Version 1.1.0
* GML output additions
* HTTPS support
* Colored output fixes
28.2.2000
* the patched PyLR parser generator works
* wrote a GML parser
25.2.2000
* changed the name to LinkChecker; my old Java LinkChecker will
disappear because I do not maintain it anymore
21.2.2000
* add -q, --quiet option
* convert all url hosts to lowercase
* log the download time for urls
20.2.2000
* add Graph Modelling Language (GML) output for sitemaps
* add SQL output
19.2.2000
* second try with HTTP/1.1: additionally close response
* remove deprecated options
* new option -W, --file-output
* fix typo for --password option
18.2.2000
* add "-" to mail adress syntax (Baz <B.Rowlingson@lancaster.ac.uk>)
* fix typo in pylice (David J. MacKenzie <djm@web.us.uu.net>)
10.2.2000 Version 0.8.0
* clean the CVS dir
* fixes for configuration
* first version of configuration parsing
9.2.2000
* do not pass anchor in HTTP requests
* fixes for configuration parsing
8.2.2000
* fixed bad finished_NoThreads function
* backed out HTTP/1.1 support. This library is buggy and
does not close some filehandles. Eventually you will get
a "Too many open files" error
* strip whitespace from parsed urls
6.2.2000
* fixed some bugs, the test suite is running again
5.2.2000
* made "LinkChecker" module
* configuration is dynamic; no more class variables
* print line number
* more agressive closing of filehandles
27.1.2000 Version 0.7.0
* put pylicerc in /etc for .deb package
* HTTP/1.1 support with httplib.py from Greg Stein
* DNS MX lookup for mail adresses
use the DNS module from Guido van Rossum and Anthony Baxter
MX lookup was a suggestion to LinkChecker from
Jimmy Engelbrecht <jimmy@e.kth.se>
26.1.2000 Version 0.6.2
* refined HTML link syntax to handle non-quoted URLs
* fix: set urlTuple to None if we cannot check anchors
* fixed anchor checking again
25.1.2000 Version 0.6.1
* fixed the HTML link syntax
24.1.2000
* fix: -e option did not work properly
* fix: reenabled LinkChecker Online, updated to 0.6.0
21.1.2000 Version 0.6.0
* fix: add hostname for relative redirections
* Added TODO list
20.1.2000
* Added documentation for the LinkChecker class
19.1.2000
* HTTP Proxy support
* CGI logging
18.1.2000 Version 0.5.0
* anchor checking in local HTML files
* configuration file
* HTTP Authorization support
* Send HTTP HEAD method to check and GET method to get contents
* Still missing: Proxy support (including HTTP status code 305)
17.1.2000
* cut parameter, query and fragment of local file names
* limit number of redirections to 5
14.1.2000 Version 0.4.3
* pylice.bat fix: now it really works
* fix for local Windows file arguments
14.1.2000 Version 0.4.2
* StringUtil.indentWith: use string multiplying
* Still missing: HTTP authorization and Proxy support
* pylice.bat fix: pass parameters
13.1.2000 Version 0.4.1
* Windows python.bat script
* installation updates
* additional .zip package for Windows
12.1.2000 Version 0.4.0
* fixed LinkChecker.NumThreads setting: if the platform
does not support threading, it is disabled automagically
* robots.txt parsing
* split up UrlData.py
* simplified option parsing
* strip optional quotes from urls
* use quit() not close() to disconnect from FTP servers
11.1.2000 Version 0.3.0
* try to finger for mailto: links
* try to connect for telnet: links
* removed time.sleep(1) commands, they are not necessary
* restrict CGI to recursion level 3
* make UrlCache and RobotsTxtCache thread safe
* fixed the 'No more open files' bug by closing all connections
* fixed thread synchronization in LinkChecker while loop
* you can specify -t 0 on the commandline to disable threading
* STILL MISSING:
HTTP authorization, Proxy and robots.txt parsing
10.1.2000 Version 0.2.0
* configure option to disable threading: LinkChecker.threadsupport
* do not rely on self.mime in HttpUrlData, this could be None
* flush stdout after each log entry
* use LinkChecker.User and LinkChecker.Password in FTP connections
* make sure redirection is not cyclic
9.1.2000 Version 0.1.0
* HTTP request
* FTP request
* fixed MaxRecursionLevel setting
* fixed name clash of variable and function warning
* ColoredLogger
* small doc changes
* CGI and HTML files for LinkChecker Online,
but I still have to install Python on my http server
(will try this tomorrow)
8.1.2000
* Properties, Threader, LinkChecker, UrlData, Logging
7.1.2000 Version 0.0.1
* Option processing
Local variables:
mode: debian-changelog
End:
Reference in a new issue View git blame Copy permalink