summaryrefslogtreecommitdiffstats
path: root/Auth
Commit message (Collapse)AuthorAgeFilesLines
* #130: PHP 7 compatibility (fixes #129)HEADorigin/masterorigin/HEADmasterOleksandr Zahorulia2016-09-3025-67/+69
|
* Fixes CVE-2014-8150Marco Ceppi2015-07-301-1/+11
| | | | Thanks for reporting and patching this Andrey!
* loop over directories "the right way"Lawrence Brown2015-01-031-1/+1
| | | | | | | | | | | | | | | | See http://php.net/manual/en/function.readdir.php - specifically /* This is the correct way to loop over the directory. */ while (false !== ($entry = readdir($handle))) { echo "$entry\n"; } /* This is the WRONG way to loop over the directory. */ while ($entry = readdir($handle)) { echo "$entry\n"; } Looping over a directory the wrong way will cause the loop to stop if it comes to a directory named 0 (or anything else which evaluates to false)
* Provide several modifications on the PredisStore implementation.drmalex072014-03-111-4/+7
| | | | | | | | - Fixed some typos - Fixed the way an association is retreived from the store (getAssociation()). The existing implementation was poping (LPOP) from a list, causing unneeded re-associations. If we need to discard stale association handles, i think we should rather set the appropriate expiry intervals.
* Merge pull request #115 from marcoceppi/warn-when-person-is-doing-harmMarco Ceppi2014-02-131-2/+7
|\ | | | | Add warning when you disable SSL, don't be silly please
| * Add warning when you disable SSL, don't be silly pleaseMarco Ceppi2014-01-221-2/+7
| |
* | Reject OpenID XML responses with a doctype CVE-2012-4554Marco Ceppi2014-01-231-0/+4
| |
* | Clean upMarco Ceppi2014-01-231-3/+3
|/
* Merge pull request #113 from thanhhh/masterMarco Ceppi2014-01-201-1/+5
|\ | | | | Add Auth_OpenID_DISABLE_SSL_VERIFY opt to disable SSL verify
| * Remove wasted white space for #113ThanhHH2014-01-161-2/+2
| |
| * Add Auth_OpenID_DISABLE_SSL_VERIFY opt to disable SSL verifyThanhHH2014-01-091-3/+7
| |
* | Yadis: drop custom replaceEntities() function and use html_entity_decodeAdam Williamson2014-01-091-24/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Yadis' ParseHTML.php has a replaceEntities() function for replacing HTML entities, with a comment that appears to explain its existence by stating "Replace numeric entities because html_entity_decode doesn't do it for us". This is breaking with PHP 5.5 because it uses the deprecated /e modifier for preg_replace() - https://github.com/openid/php-openid/issues/108 . I think this custom function is no longer needed at least with PHP 5. PHP 5 has had support for replacing numeric entities since 2003, and its entity handling code has been refined quite a lot since then. replaceEntities() has been there since 2006, and probably earlier. I guess at that time PHP 4 compatibility was still important so this was needed - I don't think PHP 4's html_entity_decode() has ever had numeric entity support - but now PHP 5 is a decade old and there's a separate PHP 4 branch of php-openid, I think we can ditch replaceEntities() in the main branch!
* | Merge pull request #111 from stutiredboy/masterMarco Ceppi2014-01-091-0/+3
|\ \ | |/ |/| Update Consumer.php for Auth_OpenID_DumbStore
| * Update Consumer.phpstutiredboy2013-12-171-0/+3
| | | | | | | | | | | | | | | | ``` $this->_use_assocs = (is_null($this->store) ? false : true); ``` always return true, even the store is Auth_OpenID_DumbStore, and do an association with the OpenID Server. but OpenID Server will reject the check_authentication, if RP have do an association.
* | Merge pull request #94 from etienneq/pass-by-reference-fixMarco Ceppi2013-12-175-14/+14
|\ \ | |/ |/| issue #8 fix (PHP 5.4 call-time pass-by-reference removed)
| * Call-time pass-by-reference fixEtienne Quiel2013-04-095-14/+14
| |
* | Merge pull request #105 from vmattila/predis-storeMarco Ceppi2013-09-031-2/+2
|\ \ | | | | | | Fixed wrong expiration command (expiresat -> expireat)
| * | Fixed wrong expiration command (expiresat -> expireat)Ville Mattila2013-09-021-2/+2
| | |
* | | correction for PHP < 5.2.11nao-pon2013-08-251-6/+11
| | | | | | | | | libxml_disable_entity_loader (PHP 5 >= 5.2.11)
* | | Merge pull request #103 from vmattila/predis-storeMarco Ceppi2013-08-171-0/+208
|\ \ \ | |/ / | | | Implemented PredisStore to provide association store for Redis servers.
| * | Implemented PredisStore to provide association store for Redis servers.Ville Mattila2013-08-131-0/+208
| | |
* | | disable external XML entities and libxml errorsWill Norris2013-08-111-1/+8
|/ / | | | | | | thanks to Kousuke Ebihara for the report and patch.
* | Prevent errors in strict mode due to weird PEAR static method declarationsDavid Simon2013-05-212-14/+14
| |
* | Fixes #54. Pass request data to fix AX.Evgeniya Patuk2013-05-081-3/+8
| | | | | | | | | | | | AX fetch response is created based on AX fetch request data. Need to pass the $request to the Auth_OpenID_AX_FetchResponse getExtensionArgs method.
* | included more checks for file access errorsEtienne Quiel2013-05-061-1/+10
|/
* Better to use isset() in case it doesn't exist.Henrik Sjökvist2013-02-211-1/+1
|
* If $_SESSION isn't set, don't trigger a PHP notice.Henrik Sjökvist2013-02-201-1/+1
|
* Merge pull request #83 from UNINETT/masterMarco Ceppi2012-12-211-3/+3
|\ | | | | Fix encoding errors in some POST forms + fix example server
| * Properly escape form output data.Olav Morken2012-10-191-3/+3
| | | | | | | | | | | | Since this data is inserted as strings in HTML documents, we should call htmlspecialchars() to protect against any values containing special characters, e.g. '"' or '&'.
| * Revert "1. Added urldecoding POST form values (browser will encode them if ↵Olav Morken2012-10-191-1/+1
| | | | | | | | | | | | | | | | | | | | needed)." Since the message values are not URL encoded, calling urldecode() on them does not make sense. Instead, it breaks any OpenID parameter containing anything that looks like URL encoding. This reverts commit 8d53dfe542f321da3a6e6d6e626923d575ed91d2.
* | Merge pull request #61 from JonTheNiceGuy/removeloggingMarco Ceppi2012-12-211-7/+0
|\ \ | | | | | | Removed logging calls for a successful retrieval, as the PlainHTTPFetcher doesn't perform logging on a successful commit.
| * | Removed logging calls for a successful retrieval, as the PlainHTTPFetcher ↵Jon "The Nice Guy" Spriggs2011-10-051-7/+0
| | | | | | | | | | | | doesn't perform logging on a successful commit.
* | | Set proxy options only if Auth_OpenID_HTTP_PROXY is explicitly setitsour2012-10-251-1/+7
| |/ |/|
* | Return string instead of arrayDaniel Hofstetter2012-08-141-1/+5
| |
* | Removed one deprecated runtime call-by-referenceVille Mattila2012-08-091-1/+1
| |
* | Merge pull request #41 from cakebaker/masterMarco Ceppi2012-07-311-1/+1
|\ \ | | | | | | Changing datatype of the server_url column from blob to varchar
| * | Using varchar(2047) instead of blob for server_url columnDaniel Hofstetter2010-12-151-1/+1
| | |
* | | fixed call-time pass-by-referenceArnaud Le Blanc2012-03-043-7/+11
| | |
* | | allow custom CA certs to be used to verify hostsWill Norris2012-01-021-0/+6
| | | | | | | | | | | | partially fixes #38
* | | restructure Auth_OpenID_VERIFY_HOST logicWill Norris2012-01-021-24/+19
| | | | | | | | | | | | make it a little more readable and obvious what we're doing here
* | | In response to Issue 38 and issue 51, this patch allows forcing host ↵Jon "The Nice Guy" Spriggs2011-10-051-3/+24
| |/ |/| | | | | verification on and off, or leaving the defaults as-is.
* | Improve PHP 5 complianceAdrian Lang2010-10-084-3/+9
|/
* Patching issue 17 (http://github.com/openid/php-openid/issues/#issue/17)JonTheNiceGuy2010-07-211-1/+5
| | | | | | This was also logged previously as Trac ticket 356 This patch was submitted at the time by https://me.yahoo.com/doulepov
* fixed Parse::match() returning different types and cleaned call-by-ref warningsTim Preston2010-07-203-5/+5
| | | | | | | | | Auth/OpenID/Parse.php The match() function takes a reference parameter $match and, depending on execution flow, either returns that as an array or a string. This has been modified to now return an array in both cases. Auth/OpenID/Consumer.php Auth/Yadis/Manager.php Fixed two function calls with explicit call-by-reference parameters.
* 1. Added urldecoding POST form values (browser will encode them if needed).Shein Alexey2010-07-201-1/+1
|
* callback requires 3 parameters, not 2. fixes issue#1 from ↵Alexey Zakhlestin2010-07-201-1/+1
| | | | github.com/openid/php-openid branch
* no need for reference here, as php-5.x already treats all objects as ↵Alexey Zakhlestin2010-07-201-1/+1
| | | | pseudo-references
* Use constant-time comparison of signatures to mitigate timing attacksCarl Howells2010-07-202-1/+15
|
* Updating version2.2.2Lilli2010-04-291-1/+1
|
* Fixing issue opened by jhr with patch found on bugs.debian.orgs: "A Debian ↵Lilli2010-04-291-1/+1
| | | | user reported that the regular expression in Auth/OpenID/Parse.php is too greedy to cope with big pages. A simple fix is to make .? to . See http://bugs.debian.org/535479 for more info."
* Updated version2.2.1Lilli2010-04-281-1/+1
|
* Merge branch 'master' of github.com:openid/php-openidLilli2010-04-282-14/+14
|\
| * form should autosubmit. fixes issue#2 from github.com/openid/php-openid branchAlexey Zakhlestin2010-04-291-1/+1
| |
| * fix mb_ereg warningAlexey Zakhlestin2010-04-291-11/+11
| |
| * fix typoAlexey Zakhlestin2010-04-291-2/+2
| |
* | Manually added indeyet's commit because it was failing: 'no need for ↵Lilli2010-04-281-1/+1
|/ | | | reference here, as php-5.x already treats all objects as pseudo-references'
* Merge branch 'master' of github.com:openid/php-openidLilli2010-04-283-15/+3
|\
| * Don't throw error when when received arg count is unlimitedAnthony BRETAUDEAU2010-04-291-1/+1
| |
| * As written in doc, don't erase signed args when some of the args are unsignedAnthony BRETAUDEAU2010-04-291-1/+1
| |
| * SQLStore: use is_subclass_of to determine if we have a PEAR instance.Daniel Hahler2010-04-291-13/+1
| | | | | | | | | | | | | | | | This prevents a blank page with no error indication at all, if there appears to happen a fatal error when trying to require DB.php from the global include path (very bad practice!) Instead, using is_subclass_of should be enough, especially when using an own implementation.
* | Added ccazette/indeyet's change to function returnToVerified() manually as ↵Lilli2010-04-281-2/+3
| | | | | | | | the commit was failing as when I tried to pull the change. This fixes the issue opened by ccazette: 'In Server.php, this function calls the Auth_OpenID_verifyReturnTo() function from TrustRoot.php, which takes 3 arguments, but only 2 are given in call_user_func_array.'
* | Fixed minor issue that Ouroboros17 brought to my attention: 'Server.php give ↵Lilli2010-04-281-1/+1
|/ | | | me a notice error at line 1471 because ->code can be undefined. An isset test can correct the problem ;)'
* Removed/changed references to openidenabled.com. At this time, not every ↵2.2.0Lilli2010-03-161-1/+1
| | | | reference has been updated.
* Added the following patch from the trac.openidenabled.com bug tracker:Lilli2010-02-167-8/+10
| | | | | | | | | | | | | | | | | | | | | | | | | http://trac.openidenabled.com/trac/ticket/338 (http://trac.openidenabled.com/trac/attachment/ticket/338/openid-php5.3.diff) Patch makes php-openid library compatible with PHP 5.3, and included changes to many files. Original Message: Reported by: http://paulisageek.com/ Opened 6 months ago Last modified 2 months ago php openid library doesn't seem to work with php 5.3 "I get many PHP warnings and then the error condition on the Consumer. Deprecated: Assigning the return value of new by reference is deprecated in *path*/Auth/OpenID/Consumer.php on line 271 Deprecated: Assigning the return value of new by reference is deprecated in *path*/Auth/OpenID/Consumer.php on line 273 Warning: Parameter 1 to filter_MatchesAnyOpenIDType() expected to be a reference, value given in *path*/Auth/Yadis/XRDS.php on line 431 Warning: Parameter 2 to Auth_OpenID_GenericConsumer::_completeInvalid() expected to be a reference, value given in *path*/Auth/OpenID/Consumer.php on line 669 " It should be noted that 3 hunks across 7 files failed, and I applied most of them manually. 7 files were patched.
* Added the following patch from the trac.openidenabled.com bug tracker:Lilli2010-02-1622-236/+189
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | http://trac.openidenabled.com/trac/ticket/337 (http://trac.openidenabled.com/trac/attachment/ticket/337/php-openid-php5.3.patch) Patch makes php-openid library compatible with PHP 5.3, and included changes to many files. Original Message: Reported by: http://bfarber.myopenid.com/ Opened 7 months ago Last modified 2 months ago PHP OpenID libraries do not work on PHP 5.3 "In my own application the script simply stops working at function Auth_OpenID_setNoMathSupport() { if (!defined('Auth_OpenID_NO_MATH_SUPPORT')) { define('Auth_OpenID_NO_MATH_SUPPORT', true); } } (Namespace issue perhaps?) I downloaded the latest 2.1.3 libraries available from the site here and set them up on my local server. Using PHP 5.2.9 they work fine (with some notices) however using PHP 5.3 they don't work at all. I just get this one a white screen Deprecated: Assigning the return value of new by reference is deprecated in C:\wamp\www\openid\php-openid-2.1.3\examples\consumer\common.php on line 72 Deprecated: Assigning the return value of new by reference is deprecated in C:\wamp\www\openid\php-openid-2.1.3\Auth\OpenID\Consumer.php on line 271 Deprecated: Assigning the return value of new by reference is deprecated in C:\wamp\www\openid\php-openid-2.1.3\Auth\OpenID\Consumer.php on line 273" It should be noted that 34 hunks across about 22 files failed, and I applied most of them manually. 37 files were patched.
* Added the following patch from the dev@openidenabled.com mailing list:Lilli2010-02-121-0/+20
| | | | | | | | | | | | | | | http://lists.openidenabled.com/pipermail/dev/attachments/20090928/2d616914/attachment.bin Original Message: anthony.lenton at canonical.com anthony.lenton at canonical.com Mon Sep 28 11:30:12 PDT 2009 darcs patch: cURL verify host "With this patch, if you define Auth_Openid_VERIFY_HOST cURL will enable host verification on ssl connections, for added security." This patch was in the form of a Darcs patch, not a normal patch. So solve this, I applied it to the Darcs repository found on openidenabled, then created a new diff file between the original Darcs repo and the new one (with the patch applied) so that I could apply it to this git repo. All hunks were applied successfully.
* Added the following patch from the dev@openidenabled.com mailing list:Lilli2010-02-122-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | http://lists.openidenabled.com/pipermail/dev/attachments/20090206/f0d91666/attachment-0001.bin Original Message: will at willnorris.com will at willnorris.com Fri Feb 6 14:56:12 PST 2009 darcs patch: Properly handle Yadis discovery failure "The library does not curently handle failures during Yadis discovery properly. Specifically, if Yadis discovery fails, and discovery WAS being done with an XRDS document, then non-Yadis discovery should be attempted. The use-case I'm dealing with right now that exposes this bug is when an XRDS document location is advertised (via HTTP response header or <meta /> HTML tag), but the document is not able to be retrieved for whatever reason. The library is failing in this case, instead of continuing to attempt non-Yadis discovery. This patch includes two changes: - change to Yadis/Yadis.php that properly detects if XRDS was used - change to OpenID/Discovery.php that prevents discovery from halting in the above example" This patch was in the form of a Darcs patch, not a normal patch. So solve this, I applied it to the Darcs repository found on openidenabled, then created a new diff file between the original Darcs repo and the new one (with the patch applied) so that I could apply it to this git repo. All hunks were applied successfully.
* Added the following patch from the dev@openidenabled.com mailing list:Lilli2010-02-121-1/+1
| | | | | | | | | | | | | | | http://lists.openidenabled.com/pipermail/dev/attachments/20090113/8ed067b3/attachment.bin Original Message: will at willnorris.com will at willnorris.com Sun Jan 4 11:27:26 PST 2009 darcs patch: I'm not entirely sure which versions of PHP are having... "I've had a number of users report that $this->store is incorrectly evaluating to false, thereby preventing associations from being used. I've been unable to reproduce this myself, but the change is minor enough." This patch was in the form of a Darcs patch, not a normal patch. So solve this, I applied it to the Darcs repository found on openidenabled, then created a new diff file between the original Darcs repo and the new one (with the patch applied) so that I could apply it to this git repo. All hunks were applied successfully.
* Added the following patch from the dev@openidenabled.com mailing list:Lilli2010-02-123-4/+32
| | | | | | | | | | | | | | | http://lists.openidenabled.com/pipermail/dev/attachments/20090109/7e344691/attachment-0001.bin Original Message: ketmar at ketmar.no-ip.org ketmar at ketmar.no-ip.org Fri Jan 9 07:51:35 PST 2009 darcs patch: fix for incomplete URIs in "location" http field (for ... "* fix for incomplete URIs in "location" http field (for technorati and maybe others)" This patch was in the form of a Darcs patch, not a normal patch. So solve this, I applied it to the Darcs repository found on openidenabled, then created a new diff file between the original Darcs repo and the new one (with the patch applied) so that I could apply it to this git repo. Hunks were applied successfully.
* Added the following patch from the dev@openidenabled.com mailing list:Lilli2010-02-121-2/+2
| | | | | | | | | | | | | | | | | | | | | http://lists.openidenabled.com/pipermail/dev/attachments/20090101/217f2590/attachment.bin Original Message: dev-list-openidenabled at thequod.de dev-list-openidenabled at thequod.de Thu Jan 1 18:14:12 PST 2009 darcs patch: php-openid-urldecode-urlparamkey "This patch also decodes the key of URL params, which is required for param names like 'action[foo]' when the browser sends those urlencoded (Konqueror 4.2 does so). I would like to propose using rawurldecode instead of urldecode, too, but that causes the tests to fail and may be against the OpenID spec (the difference is only that "+" gets decoded with urldecode() but not with rawurldecode IIRC)" This patch was in the form of a Darcs patch, not a normal patch. So solve this, I applied it to the Darcs repository found on openidenabled, then created a new diff file between the original Darcs repo and the new one (with the patch applied) so that I could apply it to this git repo. All hunks were applied successfully.
* Added the following patch from the dev@openidenabled.com mailing list:Lilli2010-02-122-3/+39
| | | | | | | | | | | | | | | | | http://lists.openidenabled.com/pipermail/dev/attachments/20080821/1eded806/attachment-0001.bin Original Message: will at willnorris.com Thu Aug 21 18:43:28 PDT 2008 darcs patch: seperate provider and consumer xrds types "Currently, the five OpenID Provider service types (OpenID 1.0 - 1.2, and the two for OpenID 2.0) are combined with the one OpenID Consumer service type (return_to). This causes a potential false positive when performing discovery on an endpoint. If a php-openid consumer performs discovery on another consumer which has a published return_to service type, php-openid will think it is a valid OpenID Server service type and send an authentication request to the URI. Similarly, when checking the return_to URL of a consumer, if the consumer has any of the OpenID Server service types published, a php-openid server object will try to treat one of those URIs as a published return_to URL. This scenario is actually fairly common... many personal blogs serve as both an OpenID Server (for the blog owner) as well as an OpenID consumer (for people who comment on the blog)." This patch was in the form of a Darcs patch, not a normal patch. So solve this, I applied it to the Darcs repository found on openidenabled, then created a new diff file between the original Darcs repo and the new one (with the patch applied) so that I could apply it to this git repo. It should be noted that all 3 hunks for Discover.php and the 1st hunk for TrustRoot.php were applied successfully, but the second hunk for TrustRoot failed. It seems that all it was doing was removing trailing endlines from the file, so I think it should be ok.
* Added the following patch from the dev@openidenabled.com mailing list:Lilli2010-02-101-0/+413
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | http://lists.openidenabled.com/pipermail/dev/attachments/20080221/f087874e/attachment.obj Original Messages: Ryan Patterson cgamesplay at cgamesplay.com Thu Feb 21 10:57:56 PST 2008 OpenIDStore for PEAR::MDB2 "On Thu, Feb 21, 2008 at 10:24 AM, Thomas Harning <thomas.harning at trustbearer.com> wrote: > Shortening the URLs to 255 chars has some ugly problems since the spec states > that the max Server URL can technically be up to 2047 bytes (per OpenID 1.1-Appendix D) Ah, I glanced at the spec but didn't see anything. > One option for databases that can't have a key large enough is using a prefix-based key... > I know MySQL supports this, but I'm not so sure on others. This is what the current SQLStore implementation does. I've modified the patch to add specific SQL for mysql backends, but use MDB2's generic table creation for other stores. The test cases in the patch pass for MDB2 using MySQL, but I'd like to verify that it works with PostgreSQL. -- Regards, Ryan Patterson <mailto:cgamesplay at cgamesplay.com>" Thomas Harning thomas.harning at trustbearer.com Thu Feb 21 07:24:02 PST 2008 OpenIDStore for PEAR::MDB2 "Ryan Patterson wrote: > I've created an Auth_OpenID_OpenIDStore implementation that utilizes > PEAR::MDB2 instead of PEAR::DB, called Auth_OpenID_MDB2Store. The > advantages to this store are that it supports all database back ends > that PEAR::MDB2 supports, and does not require the user to know which > database back end is being used when instantiating the store. Because > of the complete database abstraction, the entire store is > approximately 350 lines shorter than the SQLStore family. > > As a note to users planning to test this store: the table schema is > incompatible with the other SQLStores. Specifically, the length of > server URLs has been shortened to 255 characters to allow cross-RDBMS > compatibility, and the field types have been changed to NOT NULL. You > may safely delete your associations and nonces tables with no > consequences. Shortening the URLs to 255 chars has some ugly problems since the spec states that the max Server URL can technically be up to 2047 bytes (per OpenID 1.1-Appendix D) One option for databases that can't have a key large enough is using a prefix-based key... I know MySQL supports this, but I'm not so sure on others. Another option would be to use a unique value as the primary key and use some ugly mechanics so that you can store a 20 byte sha1 in a non-unique index and the server URL not in an index. The basic select would be SELECT * FROM Sites WHERE URLHash = ? AND URL = ?; The URLHash would narrow it down to likely one, where URL would clobber potential collisions and keep the data around if needed... -- Thomas Harning @ TrustBearer Labs (http://www.trustbearer.com) Secure OpenID: https://openid.trustbearer.com/harningt 3201 Stellhorn Road 260-399-1656 Fort Wayne, IN 46815" OpenIDStore for PEAR::MDB2 Ryan Patterson cgamesplay at cgamesplay.com Thu Feb 21 07:19:02 PST 2008 "I've created an Auth_OpenID_OpenIDStore implementation that utilizes PEAR::MDB2 instead of PEAR::DB, called Auth_OpenID_MDB2Store. The advantages to this store are that it supports all database back ends that PEAR::MDB2 supports, and does not require the user to know which database back end is being used when instantiating the store. Because of the complete database abstraction, the entire store is approximately 350 lines shorter than the SQLStore family. As a note to users planning to test this store: the table schema is incompatible with the other SQLStores. Specifically, the length of server URLs has been shortened to 255 characters to allow cross-RDBMS compatibility, and the field types have been changed to NOT NULL. You may safely delete your associations and nonces tables with no consequences. A test case has been added that tests the MDB2 store with a MySQL backend. I would appreciate it if someone capable could modify the test case and test it using PostgreSQL, to ensure that the database abstraction I've used is correct. -- Regards, Ryan Patterson <mailto:cgamesplay at cgamesplay.com>" All of the patch's hunks were applied successfully.
* Added the following patch from the dev@openidenabled.com mailing list:Lilli2010-02-101-3/+18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | http://lists.openidenabled.com/pipermail/dev/attachments/20080117/5d35b90c/attachment.obj Original Message: artemy tregoubenko me at arty.name Thu Jan 17 01:58:39 PST 2008 php: raw bytestrings vs. mb_string.func_overload and strlen "Hello. After all, patch for unicode support in 'html parser' is quite small. It solves my problems for now. On 1/16/08, artemy tregoubenko <me at arty.name> wrote: > Hello again! > > Today I switched to php-openid 2.0.0 and some things began working. > However, I experienced more problems with multibyte strings. Now they > happen at stage of fetching and parsing html documents. > ParanoidFetcher was easy to fix, patch for it attached. Currently I'm > hacking into ParseHTML, and using regexes in php-unicode environment > is a pain! > > -- > arty ( http://arty.name ) > > -- arty ( http://arty.name )" It should be noted that hunks 1 and 4, out of the patch's 4, were applied successfully, I manually applied hunk 2, and hunk 3 could no longer be applied for the same reason as the previous patch: In both cases, the patches modified the same function, and the current function no longer contains the lines of code that the 3rd hunk is trying to modify. On that note, both patches are modifying the same lines of code in the same function. From what I can tell from the email messages, they are fixing different issues, but it should be confirmed that the fixes are both needed and will play nicely together.
* Added the following patch from the dev@openidenabled.com mailing list:Lilli2010-02-101-1/+12
| | | | | | | | | | | | | | | | | | | | | | http://lists.openidenabled.com/pipermail/dev/attachments/20070117/14e9fc7d/attachment.bin Original Message: dAniel hAhler dev-list-openidenabled at thequod.de Wed Jan 17 14:02:44 PST 2007 [PHP-openid] PCRE backtrack error in Auth_OpenID_Parse::parseLinkAttrs() / todo (patch) "Hi, I find that the Auth_OpenID_Parse::parseLinkAttrs() method is quite ineffective. It failed here (silently), because of a PREG_BACKTRACK_LIMIT_ERROR in preg_match() (introduced with PHP 5.2). The attached patch adds a TODO with an idea of a better/more efficient implementation and a dirty workaround, by disabling this limit. It seems like the parsing fails if the HTML to get parsed is > 100KB or something similar. The default backtrack limit is 100000. ..." It should be noted that hunks 1, 4, and 5 out of the patch's 5 were applied successfully, I manually applied hunk 2, and hunk 3 could no longer be applied.
* Added the changes to consumer and consumer test that were made to ↵Lilli2010-02-081-1/+2
| | | | openidenabled.com's repo after the repo was moved to git hub
* still use in_array, just require strict type checkingWill Norris2009-05-281-4/+2
| | | Signed-off-by: Will Norris <will@willnorris.com>
* [ few random cleanup changes ]Luke Shepard2009-05-286-9/+44
| | | | | | | | A few minor changes made over the past few months. This adds some additional logging, an extra helper function for error messages, and fixes a small bug with the association handler ... not to mention standardizing on "false" instead of "False". Signed-off-by: Will Norris <will@willnorris.com>
* [ remove closing ?> tag from all files ]Luke Shepard2009-05-2840-40/+15
| | | | | | | | | | | | Reviewed By: Will Norris Facebook has a policy to not have closing ?> tags at the bottom of PHP files. It can unintentionally introduce newlines which causes ridiculous bugs. This is recommended in the Zend style guide: http://framework.zend.com/manual/en/coding-standard.php-file-formatting.html Signed-off-by: Will Norris <will@willnorris.com>
* Revert "[ remove closing ?> tag from all files ]"Luke Shepard2009-05-2740-15/+40
| | | | | | This reverts commit a2bdc5c6129ff11b1c1e29a207b1b80153ac6fb0. I'm going to stop playing in bce's fork, and start messing in my own. Sorry for the mess Brian :)
* [ remove closing ?> tag from all files ]Luke Shepard2009-05-2740-40/+15
| | | | | | | | | | Reviewed By: Will Norris Facebook has a policy to not have closing ?> tags at the bottom of PHP files. It can unintentionally introduce newlines which causes ridiculous bugs. This is recommended in the Zend style guide: http://framework.zend.com/manual/en/coding-standard.php-file-formatting.html
* [ Upgrading php-openid test framework to use PHPUnit 3.3 ]Luke Shepard2009-05-273-7/+7
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Reviewed By: Will Norris The set of tests assumes PHPUnit 2.3, which conflicts with my environment setup. I think PHPUnit 3.3 is pretty widespread now, and under active maintenance (latest rev 4/10/2009). There's a few changes included here. I did a few mostly automated code transformations: replace "PHPUnit_TestCase" "PHPUnit_Framework_TestCase" replace "PHPUnit_TestSuite" "PHPUnit_Framework_TestSuite" removed all require_once "PHPUnit.php" - they are unnecessary. "call by reference" is now deprecated, so changed "&$" to "$" I also made a few changes to admin/testtext.php and Tests/TestDriver.php to accomodate the new syntax. Note that this just upgrades the framework; it does not guarantee that all tests pass. I'll submit future diffs to fix or disable failing unit tests. I wasn't able to get the tests running with the previous version, so I don't know if these have always been failing. I doubt that the github has automatically running unit tests, but if so then we'll have to make sure that the central repo is running the same version of PHPUnit.
* [project @ Unify method signatures to reduce E_STRICT warnings]tailor2008-12-122-2/+2
|
* [project @ Move signed assertions code into contrib/]tailor2008-12-082-400/+0
|
* [project @ OpenID Signed Assertions(Implementation of old sxip draft)]tailor2008-11-142-0/+400
| | | | | | | | | | | | | | | | | | | | | | | | | | In our solution, one party, which we call the Attribute Provider (AP), provides a signed certificate that the the user possesses some attribute (e.g. is over 18). This certificate is stored as an attribute at the user's OP, and other RPs can request this certificate when they want to verify attributes of the user. For the implementation, we have followed the OpenID Signed Assertions draft: http://www.mail-archive.com/specs@openid.net/msg00907.html The Signed Assertions Draft did not specify how signed assertions are stored at the OP, so we adopted the following scheme: Attribute: http://X Certificate: http://X/signature This enables RPs that don't care about certificates to completely ignore them. Assertions are SAML documents as specified in the OpenID Signed Assertions old draft. We are developing a demo application in which a university issues certificates verifying students' age, student-hood, and even their photo (also potentially useful to dating sites). So basically the university acts as an attribute provider, signing assertions about user claims. These claims are stored as an attribute in the OpenId provider and we can use the OpenID AX protocol to pass assertions as attributes. The data flow is: User requests assertion --- University(Attribute provider) --- (store request) --- Openid provider Relying Party(Dating site) --- (fetch request) --- OpenID Provider The RP gets the assertion, verifies the signature, and takes actions depending on the result. In some scenarios, the RP may deny the user request if the attribute verification fails (e.g. the dating site may forbid users under 18). In other scenarios the RP may treat them differently (e.g. the dating site could tag certified photos as "Verified Photo"). Note that the RP must have some sort of trust relationship with the AP. We've tried to keep the system as open as possible. Our protocol and implementation do not specify how this trust relationship is created or managed. For example, there could be a PKI specifically set up for verifying claims about student-hood, another trust system set up for verifying claims about age, etc. Santosh Subramanian Shishir Randive Michael Hart Rob Johnson
* [project @ Message: indentation]tailor2008-11-071-2/+2
|
* [project @ getAliasedArg() returns OpenID namespace when $aliased_key is 'ns']tailor2008-11-071-0/+5
| | | | | | | | | | This fixes an rather cryptic error when using stateless mode via the DumbStore. The 'ns' key can not be found in the alias/namespace mapping (its stored as the "Null Namespace"), it must be returned explicitly. The inability to find the key in the mapping results in a "Server Denied check_authentication" error, but the error is caused before any callback to the server is made. This also brings the PHP lib more in line with the ruby and python libs.
* [project @ Don't use Range header for ID page requests]tailor2008-10-312-4/+0
|
* [project @ Update version to 2.1.2]Kevin Turner2008-09-091-1/+1
|
* [project @ If we're generating user_setup_urls, let's at least do it correctly]tailor2008-07-111-3/+9
|
* [project @ Provide the user_setup_url in the Auth_OpenID_SetupNeededResponse ↵tailor2008-07-101-1/+3
| | | | even in OpenID 2]
* [project @ Update version to 2.1.1]Kevin Turner2008-06-271-1/+1
|
* [project @ Updated TLD list in TrustRoot.php]Kevin Turner2008-06-251-13/+18
|
* [project @ remove Auth_OpenID::quoteMInimal. Resolves #262.]Kevin Turner2008-06-251-32/+0
|
* [project @ ↵dAniel hAhler2008-06-221-14/+6
| | | | | | | | | | ParanoidHTTPFetcher-strpos-instead-of-preg_match-and-code-improvements] ParanoidHTTPFetcher: - Fix corner case in get() and post() for headers without any value (e.g. 'Expires:') - Use strpos() instead of preg_match() when looking for ":" (and look for ": " now instead)
* [project @ fix accept header]tailor2008-06-231-1/+1
|
* [project @ Fix Consumer's OpenID 1 checkid_immediate failure handling ↵http://j3h.us/2008-06-201-1/+2
| | | | (thanks to Evan Prodromou)]
* [project @ Test and fix HMAC-SHA256]http://j3h.us/2008-06-201-1/+1
|
* [project @ Rename files and tests related to HMACSHA1 and HMACSHA256]http://j3h.us/2008-06-206-5/+4
|
* [project @ etxrd.getCanonicalID: case-insensitive comparsions, patch from =wil]Kevin Turner2008-06-131-8/+8
| | | | | | | 1. There should only be a single CanonicalID in each XRD (in the latest XRI resolution spec), so I made it use the first CID found instead of the last. 2. Use case-insensitive comparison when comparing CanonicalIDs.
* [project @ Accept response code 206 from fetcher results (part 2). Fixes #260]Kevin Turner2008-06-112-3/+3
|
generated by cgit v1.1 at 2025-05-10 16:10:01 +0000