3 # This is part of CGI::Auth::Flexible, a perl CGI authentication module.
5 # This program is free software: you can redistribute it and/or modify
6 # it under the terms of the GNU Affero General Public License as published by
7 # the Free Software Foundation, either version 3 of the License, or
8 # (at your option) any later version, with the "CAF Login Exception"
9 # as published by Ian Jackson (version 1, or at your option any
10 # later version) as an Additional Permission.
12 # This program is distributed in the hope that it will be useful,
13 # but WITHOUT ANY WARRANTY; without even the implied warranty of
14 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 # GNU Affero General Public License for more details.
20 CGI::Auth::Flexible - web authentication optionally using cookies
24 my $verifier = CGI::Auth::Flexible->new_verifier(setting => value,...);
25 my $authreq = $verifier->new_request($cgi);
28 $authreq->check_ok() or return;
30 # sophisticated applications
31 my $divert_kind = $authreq->check_divert();
32 if ($divert_kind) { ... print diversion page and quit ... }
34 # while handling the request
35 $user = $authreq->get_username();
36 $authreq->check_mutate();
40 CGI::Auth::Flexible is a library which you can use to add a
41 forms/cookie-based login facility to a Perl web application.
43 CGI::Auth::Flexible doesn't interfere with your application's URL path
44 namespace and just needs a few (configurable) form parameter and
45 cookie name(s) for its own use. It tries to avoid making assumptions
46 about the implementation structure of your application.
48 Because CGI::Auth::Flexible is licenced under the AGPLv3, you will
49 probably need to provide a facility to allow users (even ones not
50 logged in) to download the source code for your web app. Conveniently
51 by default CGI::Auth::Flexible provides (for pure Perl webapps) a
52 mechanism for users to get the source.
54 CGI::Auth::Flexible is designed to try to stop you accidentally
55 granting access by misunderstanding the API. (Also it, of course,
56 guards against cross-site scripting.) You do need to make sure to
57 call CGI::Auth::Flexible before answering AJAX requests as well as
58 before generating HTML pages, of course, and to call it in every
59 entrypoint to your system.
63 As a minimum you need to do all of the things on this checklist, where
64 applicable. The items marked SECURITY are the ones that you might
65 forget: without them your application may appear to work, but will be
72 Call C<new_verifier> (once at application startup)
76 Call C<new_request> (once per request)
80 B<SECURITY>: Call C<check_ok> or C<check_divert> on every request, and
81 honour the return value.
85 If you're using C<check_ok>, implement either the
86 C<username_password_error> or C<login_ok> hook and provide it as
87 a setting to C<new_verifier>.
91 Provide the setting C<dir> (or provide absolute paths for all the
92 other relevant settings).
96 Call C<get_username> when you need to know who's logged in.
100 B<SECURITY>: Call C<check_mutate> or C<mutate_ok>, if you specified
101 C<promise_check_mutate>.
105 B<SECURITY>: Call C<check_nonpage> for every request which is not a page load
106 (if your application has any of those).
110 When generating URLs and forms (including AJAX requests), include the
111 hidden form parameter using C<secret_hidden_val> or
112 C<secret_hidden_html> when appropriate (see below).
116 B<SECURITY>: If you do not override the source provision facility (see
117 L</SOURCE CODE DOWNLOAD>), check that the assumptions it makes aren't
118 going to leak security-critical data.
122 Set up HTTPS on your webserver, or set the C<encrypted_only> setting
127 These points will now be covered in more detail.
129 =head2 INITIALISATION
131 Your application should, on startup (eg, when it is loaded by
133 C<< $verifier = CGI::Auth::Flexible->new_verifier(settings...) >>.
134 This call can be expensive and is best amortised.
136 The resulting verifier object can be used to process individual
137 requests, in each case with
138 C<< $authreq = CGI::Auth::Flexible->new_request($cgi) >>.
142 =head2 CHECKING AND RESPONSE GENERATION
144 If the user is logged in, your application is to handle the request.
145 Otherwise, the user needs to be presented with a login form or error
146 message, as appropriate. CGI::Auth::Flexible provides two alternative
149 =head3 Simple applications
151 The simplest usage is to call C<< $request->check_ok() >> which will
152 check the user's authentication. If the user is not logged in it will
153 generate a login form (or redirection or other appropriate page) and
154 return false; your application should not then process that request
155 any further. If the user is logged in it will return true.
157 Various hooks are provided to customise the responses generated by
160 After C<check_ok> returns true you should go ahead and process the
161 request; you can use C<< $request->get_username >> to find out which
162 user the request came from.
163 You may also need to call C<check_mutate> and/or C<check_nonpage>
166 =head3 Sophisticated applications
168 If you want to handle the control flow and to generate login forms,
169 redirections, etc., yourself, you can say
170 C<< $divert = $request->check_divert >>. This returns undef if
171 the user is logged in, or a I<divert spec> if some kind of login
172 page or diversion should be generated. See L</DIVERT SPEC> below for
173 details of how to deal with the return value.
175 =head2 MUTATING OPERATIONS AND EXTERNAL LINKS
177 =head3 Mutation-ignorant applications
179 By default CGI::Auth::Flexible does not permit external deep links
181 All GET requests give a "click to continue" page which
182 submits a form which loads your app's main page. In this
183 configuration all your application's forms and AJAX requests should
186 Such applications are also not able to provide user-specific CSS
187 stylesheets, javascript, favicons, etc.
189 This restriction arises from complicated deficiencies
190 in the web's security architecture.
192 =head3 Mutation-aware applications
194 The alternative is for your application to always make a special check
195 when the incoming request is going to do some kind of action (such as
196 modifying the user's setup, purchasing goods, or whatever) rather than
197 just retrieve and/or display information. We term such requests
200 Then non-mutating pages can be linked to from other, untrustworthy,
203 To support external links, and C<GET> requests, pass
204 C<< promise_check_mutate => 1 >> in I<settings>, and then call
205 C<< $authreq->check_mutate() >> before taking any actions. If the
206 incoming request is not suitable then C<< $authreq->check_mutate() >>
209 There have to be no mutating C<GET> requests in your application (but
210 you shouldn't have any of those anyway); if there are, they won't
211 work. (CGI::Auth::Flexible will spot them and cause them to fail,
212 rather than allow them to be insecure.)
214 =head2 GENERATING URLS, FORMS AND AJAX QUERIES
216 When you generate a URL, C<POST> form or AJAX request you may need to
217 include a secret hidden form parameter for the benefit of
218 CGI::Auth::Generic. This form parameter will be checked by
219 C<check_ok>/C<check_divert> and should be ignored by your application.
221 By default the hidden parameter is called C<caf__assochash>.
223 After calling C<check_ok> or C<check_divert> the value to put in your
224 form can be obtained from C<secret_hidden_val>; C<secret_hidden_html>
225 will generate the whole HTML C<< <input...> >> element.
227 =head3 Mutation-ignorant applications
229 For mutation-ignorant applications (see above), all forms etc. should
230 include the hidden parameter (and as discussed, they must all use
231 POST rather than GET).
233 =head3 Mutation-aware applications
235 For mutation-aware applications, whether to include the secret
236 parameter depends on the kind of request. CGI::Auth::Flexible knows
237 when it is necessary. You should find out by calling
240 If it is inconvenient to call C<need_add_hidden> at runtime, you can
241 rely instead on the following promises: All POST requests (which
242 includes all mutating requests) need the parameter. The return value
243 of need_add_hidden depends only on the $method and $reqtype
244 parameters, so you can query it once and remember the answer.
245 HTML page load GETs do not need the parameter. It is better to
246 err on the side of including the parameter.
248 If you really must, you can call C<need_add_hidden> "on the bench"
249 during development and bake the answer into your application code
250 structure. However, if you do that and a new vulnerability was
251 discovered which is fixed by changing the answer, updating
252 CGI::Auth::Flexible wouldn't be sufficient to fix it.
254 =head3 Mutation-aware applications - non-page requests
256 If your mutation-aware application supports non-page resources (AJAX
257 and JSON requests, stylesheets, favicons, etc.) it must inform
258 CGI::Auth::Flexible when it is handling such a request, by calling
261 Normally C<check_nonpage> will simply return (and you can ignore the
262 return value). However, if there is an attack (or, perhaps, a bug) it
263 will die, stopping the attack.
265 (You do not need to call C<check_nonpage> for POST requests, but it is
268 =head3 Mutation-aware applications - novel kinds of request
270 If you want to support a kind of request perhaps not yet known about
271 by CGI::Auth::Flexible, you can provide information about that new
272 request kind using C<update_get_need_add_hidden>.
276 CGI::Auth::Flexible needs to store various information in plain files;
277 it does this in the directory specified by the C<dir> parameter.
279 =head1 SOURCE CODE DOWNLOAD
281 By default, CGI::Auth::Flexible provides a facility for users to
282 download the source code for the running version of your web
285 This facility makes a number of important assumptions which you need
286 to check. Note that if the provided facility is not sufficient
287 because your application is more sophisticated than it copes with (or
288 if you disable the builtin facility), you may need to implement a
289 functioning alternative to avoid violating the AGPLv3 licence.
291 Here are the most important (default) assumptions:
297 Your app's source code is available by looking at @INC, $0 and
298 S<$ENV{'SCRIPT_FILENAME'}> (the B<source items>). See
299 C<srcdump_listitems>. Where these point to files or directories under
300 revision control, the source item is the whole containing vcs tree.
304 Specifically, there are no compiled or autogenerated Perl
305 files, Javascript resources, etc., which are not contained in one of
306 the source item directories. (Files which came with your operating
307 system install don't need to be shipped as they fall under the system
312 You have not installed any modified versions of system
313 libraries (including system-supplied Perl modules) in C</usr> outside
314 C</usr/local>. See C<srcdump_system_dir>.
318 For each source item in a dvcs, the entire dvcs history does
319 not contain anything confidential (or libellous). Also, all files which
320 contain secrets are in the dvcs's I<.ignore> file. See
321 C<srcdump_vcsscript_git> et al.
325 For each source item NOT in a dvcs, there are no confidential
326 files with the world-readable bit set (being in a world-inaccessible
327 directory is not sufficient). See C<srcdump_excludes>.
331 You have none of your app's source code in C</etc>.
335 You don't regard pathnames on your server as secret.
339 You don't intentionally load Perl code by virtue of C<.>
340 being in C<@INC> by default. (See C<srcdump_filter_cwd>.)
344 =head1 MAIN FUNCTIONS AND METHODS
348 =item C<< CGI::Auth::Flexible->new_verifier(setting => value, ...) >>
350 Initialises an instance and returns a verifier object.
351 The arguments are setting pairs like a hash initialiser.
352 See L</SETTINGS> below.
354 =item C<< $verifier->new_request($cgi) >>
356 Prepares to process a request. I<$cgi> should normally
357 be the query object from L<CGI(3perl)>. Most of the default
358 hook methods assume that it is; however if you replace enough of
359 the hook methods then you can pass any value you like and it
360 will be passed to your hooks.
362 The return value is the authentication request object (I<$authreq>)
363 which is used to check the incoming request and will contain
364 information about its credentials.
366 =item C<< $authreq->check_divert() >>
368 Checks whether the user is logged in. Returns undef if the user is
369 logged in and we should service the request. Otherwise returns a
370 divert spec (see L</DIVERT SPEC>) saying what should happen instead.
372 This method may die if it doesn't like the request, in which case
373 the request needs to be rejected.
375 =item C<< $authreq->check_ok() >>
377 Checks whether the user is logged in. Returns true if the user is
378 logged in and we should service the request.
380 Otherwise it handles the request itself, generating any appropriate
381 redirect, login form, or continuation page. It then returns false and
382 the application should not process the request further.
384 =item C<< $verifier->disconnect() >>
386 Discards the resources (open files, etc.) in the verifier object.
390 =head1 REQUEST-RELATED FUNCTIONS AND METHODS
392 All of these are only valid after C<check_divert> or C<check_ok> has
393 been called. (In the case of C<check_ok> it won't normally be sensible
394 to call these functions unless C<check_ok> returned true.)
398 =item C<< $authreq->get_divert() >>
400 Returns the value previously returned by C<check_divert>.
402 =item C<< $authreq->get_username() >>
404 Returns the name of the logged-in user. If the user was not logged
405 in (or their session had timed out, or something), returns undef.
407 =item C<< $authreq->check_mutate() >>
409 Declares to CGI::Auth::Generic that the request being handled will
410 "mutate". That is, it will modify some server-side state (eg, adding
411 items to shopping baskets, posting messages to blogs, sending emails,
414 If you have set the setting C<promise_check_mutate> you must call
415 C<check_mutate> whenever appropriate. If you haven't then it's
416 irrelevant. See L<MUTATING OPERATIONS AND EXTERNAL LINKS>.
418 C<check_mutate> will either return successfully, indicating that all
419 is well and the request should proceed, or it will die. If it dies
420 that means that the request was improper, which can only result from a
421 bug or an attack. So an "internal server error" is a suitable
424 =item C<< $authreq->check_nonpage($method, $reqtype) >>
426 Declares to CGI::Auth::Generic that the request is not a page request,
427 but rather a request of type I<$reqtype>.
429 If your application has set the setting C<promise_check_mutate>,
430 whenever it is handling anything except an HTML page loads, it must
431 call this function. See L</REQUEST TYPES>, and
432 L<GENERATING URLS, FORMS AND AJAX QUERIES>.
434 C<check_nonpage> will either return successfully, indicating that all
435 is well and the request should proceed, or it will die, like
440 =head1 RESPONSE-RELATED FUNCTIONS AND METHODS
444 =item C<< $authreq->url_with_query_params($params, [$nonpagetype]) >>
446 Convenience function which returns a url for a GET request to this
449 I<$params> is a hashref specifying the parameters and the PATH_INFO
450 (not including any parameters related to CGI::Auth::Flexible).
451 The keys are the parameter names, and the values are array refs with
452 the parameter value(s) (as strings, as yet unquoted). (They are array
453 refs because it is possible to pass multiple values for the same
454 parameter in a single request; normally each arrayref would be a
457 The request path will be the path to the application. If a parameter
458 with name C<< '' >> is supplied, it is taken as the PATH_INFO - its
459 value will be appended to the application path. (It should normally
460 start with C<< / >>, and only one value should be supplied.)
462 =item C<< something->need_add_hidden($method, $reqtype) >>
464 Enquires whether a request of type I<$reqtype> using HTTP method
465 I<$method> needs the hidden form parameter. See L</REQUEST TYPES>.
467 =item C<< something->secret_hidden_val() >>
469 Returns the value of the hidden form parameter. This should be
470 included in all POST requests to your application (and thus be a
471 hidden form parameter in all forms).
473 It should also be in some (maybe all) GET requests. If your
474 application is mutation-ignorant, it should be in all GET requests.
475 If you are mutation-aware, you need to consult C<need_add_hidden>.
477 The name of the hidden parameter is the setting C<assoc_param_name>,
478 C<caf_hassochash> by default. xxx rename param and setting
480 =item C<< something->secret_hidden_html() >>
482 Returns the HTML for an C<INPUT> element specifying the hidden form
485 =item C<< something->secret_cookie_val() >>
487 Returns the value of the secret cookie. CGI::Auth::Flexible sets this
488 cookie in the forms generated by C<check_ok>. You may also set it
489 yourself (and indeed you must do so if you use C<check_divert>).
491 =item C<< $authreq->_chain_params() >>
493 Returns a hash of the "relevant" parameters to this request, in a form
494 suitable for C<url_with_query_params>. This is all of the query
495 parameters which are not related to CGI::Auth::Flexible. The
496 PATH_INFO from the request is returned as the parameter C<< '' >>.
500 =head1 OTHER FUNCTIONS AND METHODS
504 =item C<< $verifier_or_authreq->hash($data) >>
506 Hashes the supplied data using the hash function specified by the
507 C<hash_algorithm> setting, and converts the result to a string of hex
510 =item C<< something->update_get_need_add_hidden($reqtype, $value, [$force]) >>
512 Updates CGI::Auth::Generic's knowledge about the various kinds of
513 request, and whether they need the hidden form parameter. This
514 function applies only to GET requests - POST requests always use the
517 I<$reqtype> is the request type (the value which will be passed to
518 C<check_nonpage> and C<need_add_hidden>. If you are supporting a new
519 I<$reqtype> you shouuld coordinate with CGI::Auth::Flexible upstream,
520 or other users, to assign a unique request type name.
522 This method may be called on an authreq or a verifier, in which case
523 it will affect all authreqs using the same verifier. Or it may be
524 called on the class as a whole, in which case it will affect the
525 global default list for all verifiers.
527 If I<$force> is supplied and true, this will override
528 CGI::Auth::Flexible's existing knowledge. Otherwise this new setting
529 will be ignored if CGI::Auth::Flexible already knows about the request
530 type. (When called on a verifier or authreq, it will ignore the
531 update in favour of existing knowledge recorded both globally in the
532 class or due to previous requests on the same verifier.)
534 See L</REQUEST TYPES>.
536 =item C<< CGI::Auth::Flexible::srcdump_dir_cpio($cgi,$verifier,$dumpdir,$dir,$outfn,$how,$script) >>
538 Helper function for implementing the C<srcdump_process_item> hook.
539 Generates a tarball using cpio and includes it in the prepared source
542 The arguments are mostly the same as for that hook. C<$dir> is the
543 root directory at which to start the archive. C<$how> is a short text
544 string which will be mentioned in the log.
546 C<$script> is a shell script fragment which must output a
547 nul-separated list of filenames (e.g. the output of C<find -print0>).
548 It is textually surrounded by C<( )> and will be executed with C<set -e>
549 in force. Its cwd will be C<$dir>.
551 =item C<< $verifier_or_authreq->($data) | CGI::Auth::Flexible-> >>
553 Hashes the supplied data using the hash function specified by the
554 C<hash_algorithm> setting, and converts the result to a string of hex
561 The C<$reqtype> values understood by C<check_nonpage> are strings.
568 A top-level HTML page load. May contain confidential information for
569 the benefit of the logged-in user.
573 An HTML frame. May contain confidential information for
574 the benefit of the logged-in user.
578 An HTML iframe. May contain confidential information for
579 the benefit of the logged-in user.
583 Source dump request, whether for the licence or actual source code
584 tarball; returned value is not secret.
588 CSS stylesheet. B<MUST NOT> contain any confidential data. If the
589 stylesheet depends on the user, then attackers may be able to
590 determine what stylesheet the user is using. Hopefully this is not a
595 "Favicon" - icon for display in the browser's url bar etc. We aren't
596 currently aware of a way that attackers can get a copy of this.
600 C<robots.txt>. Should not contain any confidential data (obviously).
604 Inline image, for an C<< <img src=...> >> element.
606 Unfortunately it is not possible to sensibly show top-level
607 confidential images (that is, have the user's browser directly visit a
608 url which resolves to an image rather than an HTML page with an inline
609 image). This is because images need to have a per-session hidden form
610 parameter to avoid cross-site scripting, which breaks bookmarks etc.
614 JavaScript for a C<< <script> >> element. (Possibly confidential for
619 C<< XMLHttpRequest >> returning XML data. (Possibly
620 confidential for the user.)
624 C<< XMLHttpRequest >> returning JSON data. (Possibly
625 confidential for the user.)
629 C<< XMLHttpRequest >> returning data of some other kind. (Possibly
630 confidential for the user.)
636 The return value from C<check_divert> indicates how the request should
637 be handled. It is C<undef> if all is well and the user is logged in.
639 Otherwise the return value is a hash ref with the following keys:
645 Scalar string indicating the kind of diversion required.
649 Scalar string for display to the user in relation to the diversion.
650 Has already been translated. In HTML but normally does not contain
653 =item C<CookieSecret>
655 The login cookie which should be set along with whatever response is
656 sent to the client. The value in the hash is the actual value
657 of the cookie as a string. C<undef> means no cookie setting header
658 should be sent; C<''> means the cookie should be cleared.
662 Provided with diversion kinds which involve
663 generating a redirection or indirection,
664 perhaps via a login form.
666 The extra hidden form parameters (and the C<PATH_INFO>) which should
667 be set when the subsequent request bounces back from the client, in
668 the form used by C<url_with_query_params>.
670 The contents of this hashref does not include the CAF-specific
671 parameters such as the secret cookie, those which follow from the kind
672 of diversion requested, etc.
674 It is correct to always include the contents of C<Params> as hidden
675 parameters in the urls for all redirections, and as hidden input
676 fields in all generated forms. The specific cases where C<Params> is
677 currently relevant are also mentioned in the text for each divert
682 The values of C<Kind> are:
686 =item C<SRCDUMP->I<item>
688 We should respond by sending our application source code. I<item>
689 (which will contain only word characters, and no lower case) is the
690 specific item to send, normally C<SOURCE> or C<LICENCE>.
692 =item C<REDIRECT-HTTPS>
694 We should respond with an HTTP redirect to the HTTPS instance of our
697 =item C<REDIRECT-LOGGEDOUT>
699 We should redirect to a page showing that the user has been logged
700 out. (Ie, to a url with one of the the C<loggedout_param_names> set.)
702 =item C<SMALLPAGE-LOGGEDOUT>
704 We should generate a page showing that the user has been logged out.
705 There can be a link on the page pointing to the login page so that the
706 user can log back in.
708 =item C<SMALLPAGE-NOCOOKIE>
710 We should generate a page reporting that the user does not have
711 cookies enabled. It should probably contain a link pointing to the
712 login page with additionally all the parameters in C<Params>. When
713 this divert spec is generated, C<Message> will explain the problem
714 with cookies so there is no need to do that again in the page body if
715 you include the contents of C<Message>.
719 The user's session was stale (this is described in C<Message>). We
720 should generate a login form.
724 The user supplied bad login credentials. The details are in
725 C<Message>. We should generate a login form (with additionally the
726 parameters from C<Params> as hidden fields).
728 =item C<LOGIN-INCOMINGLINK>
730 We should generate a login form (with the specified parameters); the
731 user is entering the site via a cross-site link but is not yet logged
736 We should generate a login form. The user is not yet logged in.
738 =item C<REDIRECT-LOGGEDIN>
740 We should redirect to our actual application, with the specified
741 parameters. (The user has just logged in.)
745 The user is logged in but the incoming form submission looks like it
746 was from a stale login session. Alternatively, it may have been
747 generated by an attacker's cross-site-scripting attack.
749 Naive applications should generate a small page with a form or link to
750 our own main page without any parameters.
752 A sophisticated application could infer from the submitted form
753 parameters what the user was allegedly trying to do. We could then
754 generate a fresh page showing what the intended action was, with a
755 fresh form which (if the user confirm) would resubmit that action.
756 B<Great care> must be taken to avoid relying on the sanity and
757 coherence of the incoming form parameters. We B<MUST NOT> simply
758 reproduce the incoming parameters in the new form. It is essential
759 that the visual appearance of the generated form correctly shows to
760 the user what the action is that will be taken if the form is
761 submitted. If that action is dangerous, the form should not look like
762 the kind of confirmation pages which the user is likely to simply
763 click through without thinking.
765 =item C<MAINPAGEONLY>
767 We should generate our main page but B<ignoring all form parameters>
768 and B<ignoring the path_info>. Most applications will find this
769 difficult to implement.
771 An alternative is to generate a small page with a form or link which
772 submits our own main page without any parameters.
774 (Applications which set C<promise_check_mutate> do not see this divert
779 Applications should die if they are presented with a divert kind that
780 they don't recognise.
784 C<new_verifier> and C<new_request> each take a list of settings, as
785 a list of pairs C<< key => value >> (like a Perl hash assignment).
787 The settings supplied to C<new_verifier> are stored in the verifier
788 and will apply to all authreqs made from it unless overridden in the
789 call to C<new_request>
791 When a setting is described as a hook function, it should be a
792 coderef. The first argument will be the query object from
793 L<CGI(3perl)> (strictly, it will be whatever value was passed to
794 C<new_request>). The second argument will be the authreq object (the
795 return value from C<new_request>).
796 Ie, C<< sub some_hook ($$...) { my ($cgi,$authreq,@stuff) = @_ ... >>
798 In bullet point headings, the hook functions are shown in the form
799 C<< some_hook($cgi,$authreq,@stuff) >> even though this would not be
800 legal syntax. This should be read to mean that the
801 %implicit_settings_hash{'some_hook'}($cgi,$authreq,@stuff)
802 would be a legal call. (However, the settings hash is not exposed.)
804 When a hook's default implementation is mentioned and named, that
805 function won't also be described in the section on the module's
808 =head2 GENERAL SETTINGS
814 The directory CGI::Auth::Generic should use for its data storage.
815 This is actually just a default absolute path used when the other
816 path settings are relative values.
818 Must be an absolute filename.
822 CGI::Auth::Flexible needs a database for recording users' login
823 session. This database needs to be shared across all instances of the
824 web application, so in a multi-node cluster it needs to be your actual
827 CGI::Auth::Flexible will create the table and index it needs if they
828 don't already exist, and will manage their contents. You do not need
829 to integrate them into the rest of your webapp's data storage. (In
830 particular, there is no need for transactional integrity across
831 changes made by CAF and your own application.)
833 By default, CAF uses a sqlite3 database stored on local disk in the
834 file named by C<db_path>. This will be suitable for all
835 applications which run on a single host.
837 This value, if supplied, should be a DBI handle for the database.
841 This is the DSN to pass to C<< DBI->connect >>. Used only if
842 C<db_dbh> is not supplied.
846 Path to the sqlite3 database used for CAF's session storage. The
847 default is C<caf.db>.
849 Used only if neither C<db_dbh> or C<db_dsn> are supplied.
851 If this is a relative path, it is in C<dir>.
855 Prefix for the SQL tables and indices to use (and to create, if
858 See L</DATABASE TABLES>.
862 Path to the keys file used by CAF. This arrangement will change in
863 the future. See L</BUGS>.
865 =item C<random_source>
867 Special file to read random numbers from. Should return
868 cryptographically secure (pseudo)-random bytes, unpredictable to
869 adversaries (even ones on the same machine).
871 On Linux, there is no device which is properly suitable. This is a
872 bug in Linux. You can use C</dev/random> which can block
873 unnecessarily even though the kernel PRNG has been properly seeded and
874 is fine, or C</dev/urandom> which might return values which attackers
875 can predict if the kernel PRNG has not been properly seeded.
877 The default is C</dev/urandom>.
881 Length of the assoc secret. Defaults to 128.
883 =item C<hash_algorithm>
885 Must be a string suitable for use with C<new Digest>.
886 Defaults to C<SHA-256>.
888 =item C<login_timeout>
890 A user will be logged out this many seconds after they first logged
891 in. Default: 86400 (one day).
893 =item C<login_form_timeout>
895 A login form becomes invalid this many seconds after it has been sent.
896 Default: 3600 seconds (one hour).
898 =item C<key_rollover>
900 The key used for generating assoc secrets is rolled over approximately
901 this often (in seconds). Default: 86400.
903 =item C<assoc_param_name>
905 Name of the hidden form parameter. Default: C<caf_assochash>.
909 Name of the cookie used for login sessions. Default:
912 =item C<password_param_name>
914 Name of the password field in the login form. Default: C<password>.
916 Used by C<login_ok_password> (the default C<login_ok> hook),
917 C<gen_plain_login_form> and the default C<is_login> hook.
919 =item C<username_param_names>
921 Arrayref of name(s) of username form parameters.
923 The first entry is used by C<login_ok_password> (the default
924 C<login_ok> hook) to pass to the C<username_password_error> hook and
925 used as the username if all is well.
927 All the entries are used by C<gen_plain_login_fork> (the default
928 C<gen_login_form> hook for C<check_ok>) to generate form entry fields.
930 The default is C<['username']>.
932 =item C<logout_param_names>
934 Arrayref of name(s) of form parameters indicating that the request is
937 Used by the default C<is_logout> hook.
939 If you want users to be able to explicitly log out, you need to
940 provide a logout button, something like
941 C<< <input type="submit" name="caf_logout" ... >>
943 The default is C<['caf_logout']>
945 =item C<logged_param_names>
947 Arrayref of name(s) of form parameters indicating that user has just
948 logged out. (During the logout process, the actual logout action is a
949 POST request, whose response redirects to the "you have been logged
950 out" page; these form parameters are for this second page.)
952 Used by the default C<is_loggedout> hook.
954 The first entry is used by C<check_ok> to generate the redirection.
956 The default is C<['caf_loggedout']>
958 =item C<promise_check_mutate>
960 Boolean. If true, is a declaration by the application that it is
961 mutatin-aware. See L</MUTATING OPERATIONS AND EXTERNAL LINKS>.
965 =item C<encrypted_only>
967 Boolean. If true, CAF will insist that all transactions be done over
968 an encrypted http connection. It will redirect unencrypted requests
969 to the https instance of the applicattion, and will set the encrypted
970 only flag on its cookie.
974 =item C<< get_url($cgi,$authreq) >>
976 Hook which returns the URL of this web application. By default, we
977 call C<< $cgi->url() >> for each request, but you can fix this if you
980 =item C<< is_login,is_logout,is_loggedout($cgi,$authreq) >>
982 Hook which returns a boolean indicating whether the request was,
983 respectively: a login form submission (ie, username and password); a
984 logout request (submission resulting from the user pressing the
985 "logout" button); "logged out" page (redirection from the logout
988 The default is to check whether any of the corresponding request
989 parameters (C<< login_param_names >> etc.) was supplied, using the
994 =head2 SETTINGS (HOOKS) RELATED TO THE CGI REQUEST OBJECT
998 =item C<< get_param($cgi,$authreq,$param) >>
1000 Returns the value of a single-valued form parameter.
1001 The default is to call C<< $cgi->param($param) >>.
1002 The semantics are the same as that of C<CGI::param>.
1004 =item C<< get_params($cgi,$authreq) >>
1006 Returns a hash of the parameters. The return value is a hashref whose
1007 keys are the parameter names and whose values are arrayrefs, one entry
1008 in the arrayref for each value.
1010 The default is to call C<< $cgi->Vars() >>, expect the
1011 results to look like those from C<CGI::Vars>, and massage them into
1012 the required form with split.
1014 =item C<< get_path_info($cgi,$authreq) >>
1016 Returns the PATH_INFO of the request. The default is to
1017 call C<< $cgi->path_info() >>.
1019 =item C<< get_cookie($cgi,$authreq) >>
1021 Returns the value of the CAF cookie sent with the request, or undef if
1022 none was supplied. The default is to call C<<
1023 $cgi->cookie($cookie_name) >> (where C<$cookie_name> is from the
1024 setting of the same name). The return value should be the scalar
1025 value of the cookie.
1027 =item C<< get_method($cgi,$authreq) >>
1029 Returns the HTTP method as a string. The default is to call
1030 C<< $cgi->request_method() >>.
1032 =item C<< is_https($cgi,$authreq) >>
1034 Returns a boolean indicating whether the request was over an encrypted
1035 channel. The default is C<< !!$cgi->https() >>. See C<encrypted_only>.
1039 =head2 SETTINGS RELATED TO HTML GENERATION
1041 These are only used if you call C<check_ok> (or other functions
1042 mentioned in this section).
1044 Settings whose names are of the form C<gen_...> are hooks which each
1045 return an array of strings, normally HTML strings, for use by
1046 C<check_ok> (or, in turn, other hooks, or your application). These
1047 are often documented simply by showing the output produced. In many
1048 cases parts of the output are in turn obtained from other hooks. In
1049 some cases the default implementations have been given names for
1050 convenient use by your application. They will be called in array
1053 We'll write C<gettext(something)> even though actually there is a hook
1054 to control the translation function used.
1058 =item C<handle_divert>($cgi,$authreq,$divert))
1060 C<check_ok> calls this hook before producing output of its own. If
1061 you want to handle some but not all diversions yourself, you may set
1062 this hook. The hook should either do nothing and return false, or
1063 return true if it has handled the request (or arrange for the request
1064 to be handled). If the hook returns true then C<check_ok> simply
1067 =item C<gen_login_form>($cgi,$authreq,$divert))
1069 Default: a table (used mostly for layout) containing input fields for
1070 a login form. Must be within a C<< <form> >> element, but doesn't
1071 generate it. Has text fields for every entry in
1072 C<username_param_names> (in each case associated with a description
1073 C<< gettext(ucfirst $parameter_name) >>, a password field (with
1074 description C<gettext("Password")>, and a login submit button (with
1075 description C<gettext("Login")>.
1077 Default is available as the module function C<gen_plain_login_form>.
1079 =item C<gen_login_link>($cgi,$authreq))
1083 <a href="http:...">gettext(Log in again to continue.)</a>
1085 Default is available as the module function C<gen_plain_login_link>.
1087 =item C<gen_postmainpage_form>($cgi,$authreq,$params))
1089 Default: form contents (but not the C<< <form> >> element):
1091 C<$params> (in the form returned by the C<get_params> hook) as hidden
1094 <input type="submit" ... value=getext('Continue')>
1096 Default is available as the module function C<gen_postmainpage_form>.
1098 =item C<gen_start_html>($cgi,$authreq,$title)
1100 Default: C<$cgi->start_html($title)>
1102 =item C<gen_end_html>($cgi,$authreq,$title)
1104 Default: C<$cgi->end_html($title)>
1106 =item C<gen_footer_html>($cgi,$authreq)>
1111 Powered by Free / Libre / Open Source Software
1112 according to the [gen_licence_link_html].
1113 [gen_source_link_html].
1116 Default is available as the module function C<gen_plain_footer_html>.
1118 =item C<gen_licence_link_html>($cgi,$authreq)>
1120 Default: uses C<url_with_query_params> to generate a URL for
1121 downloading the licence, and returns:
1122 <a href="...">GNU Affero GPL</a>
1124 Default is available as the module function C<gen_plain_licence_link_html>.
1126 =item C<gen_source_link_html>($cgi,$authreq)>
1128 Default: uses C<url_with_query_params> to generate a URL for
1129 downloading the source, and returns:
1130 <a href="...">Source available</a>
1132 Default is available as the module function C<gen_plain_source_link_html>.
1134 =item C<form_entry_size>
1136 Size of generated text entry fields. Default is 60.
1138 =item C<dummy_param_name_prefix>
1140 Some of CAF's HTML-generating functions need to invent form parameter
1141 names. They will all start with this string. Default: C<caf__>.
1145 =head2 SETTINGS FOR SOURCE CODE DOWNLOAD FACILITY
1149 =item C<srcdump_param_name>
1151 Form parameter name used to indicate that this is a source download
1152 request. If this parameter is supplied, C<check_ok> and
1153 C<check_divert> will arrange for the applicaton source code to be
1154 delivered as the response (in C<check_ok>'s case by doing it itself
1155 and in C<check_divert>'s case by asking your application to do so.
1157 Default is C<caf_srcdump>.
1159 =item C<srcdump_path>
1161 Path to the directory used for storing pre-prepared source downloads.
1162 Defaults to C<caf-srcdump>.
1164 If this is a relative path, it is in C<dir>.
1166 =item C<srcdump_dump($cgi,$authreq,$srcobj)>
1168 Dump the source code (C<$srcobj='source'> or licence data
1169 (C<$srcobj='licence'>). The default implementation checks that
1170 C<$srcobj> has reasonable syntax and uses the files C<$srcobj.data>
1171 and C<$srcobj.ctype> with the C<dump> hook.
1173 =item C<dump($cgi,$authreq,$contenttype,$datafilehandle)>
1175 Responds to the request by sending the contents of $datafilehandle
1176 (which should just have been opened) and specifying a content type of
1179 The default implmentation uses the C<print> hook, and also calls
1180 C<$cgi->header('-type' => $contenttype>, and is available as the
1181 module function C<dump_plain>.
1183 =item C<srcdump_prepare($cgi,$verifier)>
1185 Prepares the source code for download when requested. Invoked by
1186 C<new_verifier>, always, immediately before it returns the
1187 just-created verifier object.
1189 The default implementation is the module function
1190 C<srcdump_dirscan_prepare>, which prepares a manifest, licence file
1191 and source code tarball of tarballs, as follows:
1193 It processes each entry in the return value from C<srcdump_listitems>.
1194 These are the software's include directories and any other directories
1195 containing source code. It handles C<.> specially (see
1196 C<srcdump_filter_cwd>).
1198 For each entry it looks, relative to that, for the licence as a file
1199 with a name mentioned in C<srcdump_licence_files>. The first such
1200 file found is considered to be the licence. It then calls the hook
1201 C<srcdump_process_item> for the entry.
1203 The licence, a manifest file, and all the outputs generated by the
1204 calls to C<srcdump_process_item>, are tarred up and compressed as a
1205 single source tarball.
1207 It uses the directory named by C<srcdump_path> as its directory for
1208 working and output files. It uses the filename patterns
1209 C<generate.*>, C<licence.*>, C<s.[a-z][a-z][a-z].*>, C<manifest.*>,
1210 C<source.*> in that directory.
1212 =item C<srcdump_process_item>($cgi,$verifier,$dumpdir,$item,\&outfn,\$needlicence,\%dirsdone)>
1214 Processes a single include directory or software entry, so as to
1215 include the source code found there. Called only by the default
1216 implementation of C<srcdump_prepare>.
1218 C<$dumpdir> is the directory for working and output files. C<$item>
1219 is the real (no symlinks) absolute path to the item.
1221 C<\$needlicence> is a ref to a scalar: this scalar is undef if we have
1222 already found the licence file; otherwise it is the filename to which
1223 the licence should be copied. If the referent is undef on entry,
1224 C<srcdump_process_item> needs to see if it finds the licence; if it
1225 does it should copy it to the named file and then set the scalar to
1228 C<\%dirsdone> is a ref to the hash used by C<srcdump_prepare> to avoid
1229 including a single directory more than once. If
1230 C<srcdump_process_item> decides to process a directory other than
1231 C<$item> it should check this hash with the real absolute path of the
1232 other directoy as a key: if the hash entry is true, it has already
1233 been done and should be skipped; otherwise the hash entry should be set.
1235 C<\&outfn> is a coderef which C<srcdump_process_item> should call each
1236 time it wants to generate a file which should be included as part of
1237 the source code. It should be called using one of these patterns:
1238 $outfn->("message for manifest");
1239 $outfile = $outfn->("message for manifest", "extension");
1240 The former simply prints the message into the manifest in the form
1241 none: message for manifest
1242 The latter generates and returns a filename which should then
1243 be created and filled with some appropriate data. C<"extension">
1244 should be a string for the file extension, eg C<"txt">. The output
1245 can be written directly to the named file: there is no need to
1246 write to a temporary file and rename. C<$outfn> writes the filename
1247 and the message to the manifest, in the form
1248 filename leaf: message
1249 In neither case is the actual name of C<$dir> on the system
1250 disclosed per se although of course some of the contents of some of
1251 the files in the source code dump may mention it.
1253 The default implementation is the module function
1254 C<srcdump_process_item>.
1256 It skips directories for which C<srcdump_system_dir> returns true.
1258 It then searches the item and its parent
1259 directories for a vcs metadata directory (one of the names in
1260 C<srcdump_vcs_dirs>); if found, it calls the C<srcdump_byvcs> hook
1261 (after checking and updaeing C<%dirsdone>).
1262 Otherwise it calls the C<srcdump_novcs> hook.
1264 =item C<srcdump_novcs($cgi,$verifier,$dumpdir,$item,$outfn)>
1266 Called by the default implementation of C<srcdump_process_item>, with
1267 the same arguments, if it doesn't find vcs metadata.
1269 The default implementation is the module function C<srcdump_novcs>.
1271 If C<$item> is a directory, it uses C<srcdump_dir_cpio> to prepare a
1272 tarball of all the files under C<$item> which have the world read bit
1273 set. Directories are not included (and their permissions are
1274 disregarded). The contents of C<srcdump_excludes> are excluded.
1276 If it's a plain file it uses C<srcdump_file> to include the file.
1278 =item C<srcdump_byvcs($cgi,$verifier,$dumpdir,$item,$outfn,$vcs)>
1280 Called by the default implementation of C<srcdump_process_item>, with
1281 the same arguments, if it finds vcs metadata. The additional argument
1282 C<$vcs> is derived from the entry of C<srcump_vcs_dirs> which was
1283 used: it's the first sequence of word characters, lowercased.
1285 The default implementation is the module function C<srcdump_byvcs>.
1286 It simply calls C<srcdump_dir_cpio> with a script from the setting
1287 C<srcdump_vcsscript>.
1289 =item C<srcdump_vcs_dirs>
1291 Array ref of leaf names of vcs metadata directories. Used by the
1292 default implementation of C<srcdump_process_item>. The default value
1293 is C<['.git','.hg','.bzr','.svn']>.
1295 =item C<srcdump_vcs_script>
1297 Hash ref of scripts for generating vcs metadata. Used by the default
1298 implementation of C<srcdump_byvcs>. The keys are values of C<$vcs>
1299 (see C<srcdump_byvcs>); the values are scripts as for
1300 C<srcdump_dir_cpio>.
1302 The default has an entry only for C<git>:
1304 git ls-files -z --others --exclude-from=.gitignore
1307 =item C<srcdump_excludes>
1309 Array ref of exclude glob patterns, used by the default implementation
1310 of C<srcdump_novcs>. The default value is C<['*~','*.bak','*.tmp','#*#']>.
1312 Entries must not contain C<'> or C<\>.
1314 =item C<srcdump_listitems($cgi,$verifier)>
1316 Returns an array of directories which might contain source code of the
1317 web application and which should be therefore be considered for
1318 including in the source code delivery.
1320 Used by the default implementation of C<srcdump_prepare>.
1322 Entries must be directories, plain files, or nonexistent; they may
1323 also be symlinks which resolve to one of those.
1325 If C<.> is included it may be treated specially - see
1326 C<srcdump_filter_cwd>.
1328 The default implementation returns
1329 C<(@INC, $ENV{'SCRIPT_FILENAME'}, $0)>.
1331 =item C<srcdump_system_dir($cgi,$verifier,$dir)>
1333 Determines whether C<$dir> is a "system directory", in which any
1334 source code used by the application should nevertheless not be
1335 included in the source code dump.
1337 Used by the default implementation of C<srcdump_item>.
1339 The default implementation is as follows: Things in C</etc/> are
1340 system directories. Things in C</usr/> are too, unless they are in
1341 C</usr/local/> or C</usr/lib/cgi*>.
1343 =item C<srcdump_filter_cwd>
1345 Boolean which controls the handling of C<.> if it appears in the
1346 return value from C<srcdump_listitems>. Used only by the default
1347 implementation of C<srcdump_prepare>.
1349 If set to false, C<.> is treated normally and no special action is
1352 However often the current directory may be C</>, or a data directory,
1353 or some other directory containing data which is confidential, or
1354 should not be included in the public source code distribution for
1355 other reasons. And for historical reasons Perl has C<@INC> containing
1356 C<.> by default (which is arguably dangerous and wrong).
1358 So the default this setting is true, which has the following effects:
1360 C<.> is not searched for source code even if it appears in C<@INC>.
1361 C<.> is removed from C<@INC> and C<%INC> is checked to see if any
1362 modules appear to have already been loaded by virtue of C<.> appearing
1363 in C<@INC> and if they have it is treated as a fatal error.
1365 Only the literal string C<.> is affected. If the cwd is included by
1366 any other name it is not treated specially regardless of this setting.
1370 =head1 DATABASE TABLES
1372 In a simple application, you do not need to worry about this. But if
1373 your application runs on multiple frontend hosts with a shared
1374 database, you may need to create for yourself the tables and indices
1375 used by CGI::Auth::Flexible.
1377 By default, every time CAF starts up, it attempts to execute certain
1378 fixed database statements to create the tables and indices it needs.
1379 These are run with C<$dbh->{PrintError}> set to 0. The effect with
1380 sqlite (the default database) is that the tables and indices are
1381 created iff they do not already exist, and that no spurious errors are
1384 If you use a different database, or just prefer to do things
1385 differently, you can set up the tables yourself and/or disable or
1386 modify the default setup statements, via the C<db_setup_stmts>
1389 The tables needed are:
1392 xxx document _db_setup_do
1393 xxx make _db_setup_do explicitly overrideable
1396 xxx remaining settings
1398 username_password_error
1405 xxx document cookie usage
1406 xxx document construct_cookie fn
1408 xxx document @default_db_setup_statements
1410 xxx bugs wrong default random on Linux
1412 xxx bugs wrong default random on *BSD
1414 xxx bugs keys not shared should be in db
1416 xxx rename caf_assocsecret default cookie name
1418 xxx mention relationship between login_timeout and cookies