Added basic support for comments, for now mainly viewing comments from database (no submission forms yet).
Added first simple wikitext filter (render wikitext content as xhtml).
Ensure response content type is text/html with utf-8 charset.
- now, the CMS_PAGE_MODULE has to be declared in the related CMS_SETUP via CMS_EXECUTION.
(See demo for example)
Improved the export facilities.
Implemented blog and page export.
Added import facilities.
Implemented blog and page import.
Improved node revision web interface (allow to edit a past revision, in order to restore it as latest revisionm i.e current).
Removed specific tag from blog module, and reuse the taxonomy module for that purpose.
Added WIKITEXT module that provide a WIKITEXT_FILTER, so now we can have wikitext content.
- for now, no support for wiki links such as [[Foobar]].
The installer can now use a roc.cfg file to define the various needed modules.
Format:
{ "name": "demo", "location": ".",
"modules": {
"foo": { "location": "path-to-source-of-module-foo"; },
"bar": { "location": "path-to-source-of-module-bar"; }
}
}
Fixed typo in auth post password message.
Make sure to keep site_destination set in login form, so that the user is redirected to previous location.
Use default http client in the feel aggregator component, instead of forcing libcurl usage.
(This is important, otherwise it uses the current working directory, but for fcgi, this is not always obvious where it is.)
Allow advanced upload, and also basic html upload, in case the dropzone.js has trouble.
Split get and post handling for files module upload request.
Contribution from Fabian Murer, as part of an ETH student project.
Supervised, refactorized and merged by Jocelyn Fiat.
Signed-off-by: Fabian Murer <fmurer@student.ethz.ch>
Signed-off-by: Jocelyn Fiat <git@djoce.net>
- Removed CMS_REQUEST_UTIL
- centralize a few request related code into CMS_API
Added CMS_API.user, CMS_API.set_user (CMS_USER), ... and user related routines.
Refactored Auth related code
- added various abstractions to factorize implementation and harmonize solutions.
- revisited the logout strategy.
- updated the account info page, and remove info user should not care about.
- simplified the process, and encourage auth module to follow same design.
Added CMS_LINK helper routines to modify the related query string.
Removed CMS_USER.profile (and related routines)
- It was not used so far.
- it will probably a specific module later, if needed.
Update various module to avoid fetching user from sql directly, and let this task to CMS_USER_API.
Removed CMS_NODE_API.node_author (a_node: CMS_NODE): detachable CMS_USER,
- as the info is already in CMS_NODE.author
Added CMS_RESPONSE.redirection_delay, if ever one code want to redirect after a few seconds.
Added the request uri info to the not found cms response.
Improved Auth related module implementation by having a way to change settings like token, max age.
- use CMS_SETUP.site_id and related "auth.$module.token" ... configuration values.
- removed related CMS_..._CONSTANTS classes.
For auth session module, use auth_session as table name, and use VARCHAR(64).
Extracted sql from blog module, and store it under site/scripts/install.sql .
Renamed a few $modulename.sql as install.sql
For now with basic implementation.
It will be improved later
Added SEO related attribute in CMS_RESPONSE.
Added improved Contact module.
Added basic SEO module.
Added optional "mailer.subject_prefix" configuration item.
Added CMS_SETUP.utf_8_site_name for convenience.
Fixed a few potential unicode issues.
Fixed various typos.
as a consequence, removed usage of email_service library.
Updated the meaning for site.email to be sender email addressed.
Added notification.email to set the email address that will received system email notification.
- create term, vocabulary, add or remove term from vocabularies, ...
Fixed content editing related to taxonomy (especially with multiple terms vs tags).
Fixed various SQL storage issue related to taxonomy and vocabularies.
Added CMS_RESPONSE.wsf_theme as helper.
Moved content_types and content_type_webform_managers from CMS_RESPONSE to CMS_API.
Updated the way to output content (node, ...) to html page.
See CMS_CONTENT_TYPE_WEBFORM_MANAGER.append_cointent_as_html_to (...).
Added notion of "teaser" (short version of the content), as opposed to full content.
One can use CMS_API.html_encoder ... when possible, same for `formats', ...
Added bridge from CMS_MODULE_API to CMS_API's encoders.
Added new CMS_TAXONOMY_HOOK used to retrieve list of content associated with a specific term.
Moved up to CMS_RESPONSE a few features which was available only in specific descendants.
Added /taxonomy/term/{termid} implementation.
Updated node web form, to support taxonomy editing if allowed (specific support for CMS_VOCABULARY.is_tags: BOOLEAN).
Added notion of required or optional module dependencies.
Indeed, hooks does not require RESPONSE interface,
and should be setup before, at the system level (i.e CMS_API)
Added exportation solution via CMS_HOOK_EXPORT.
Updated blocks settings for demo example project.
jocelyn