#773448: refactor session and 'no connection' handling, by introducing proper web session. We should now be able to see page even when no anon is configured, and be redirected to the login form as soon as one tries to do a query.

[doc/book] more flesh to views/primary view

[doc/book] remove merged chapter

[doc/book] complete chapter on hooks & ops

[storage] missing qrefresh in previous patch applied: fix comment, error message, and use a storage specified encoding, not cubicweb's encoding

[fix] fix path unicity process in BytesFileSystemStorage.new_fs_path The previous implementation was bugged (prefixing the whole path with '_' instead of the base name). A new version (using number) replace it. * * * Improve BytesFileSystemStorage.new_fs_path to use available metadata This version try to get some hint about how to name the file using metadata. Using the real file name and extension when available. Keeping the extension might be usefull for exemple in the case of processing that use filename extension to detect content-type.

[source] small refactoring to avoid being doomed by a wrong assertion message on buggy storage raising attribute error

[twisted] fix LongTimeExpiringFile and getChild implementation

merge

fix cwctl db-init -d on SQL Server

[repository] fix edited_attributes management in multiple SET queries entity.edited_attributes was messed up with previous edited entities. The UpdateStep now gives a new fresh edited_attributes set to repository.glob_update_entity() so that each entity can modify it during its "transaction".

[testlib] fix dummy bug in reset_test_database(): s/if/elif

merge

merge

merge

enable tests against sqlserver2005 database

fix bad indentation

merge

merge

merge

merge

[doc/book] more about hooks (simple examples with entities and relations)

backport stable

[web] on logout, use the base_url argument of build_url instead of hacking the request to get proper url. This fix a bug on site with http/https versions: the session cookie is badly removed on logout

[entity] move req.create_entity content to a cw_instantiate factory method on entity classes, so cube developpers will be able to easily customize instance creation

[controller] refactor send mail controller to use cwconfig.sendmails

[test] properly close session

backport stable

le patch move-pdf-export-to-cubes a été importé

[hooks] fix set_operation options passing (** -> *)

[session] no way for queries from an InternalSession to be a dbapi query

[merge]

[doc/book] begin chapter on Hooks/Operations

[doc/book] dbapi: talk about IN exception, simplify

[doc/book] rql/dbapi cleanup, rip cachekey (prematurely ?)

[doc/css] cleanup a bit the sphinx default css

[repository] fix bad indentation in glob_update_entity, source.update_entity() should be called once per update

[doc] note on pytestconf, fixlets

[test] avoid weird bug when test are changing the cwd

[c-c newcube] #1192: simpler cubicweb-ctl newcube, and more * new --layout option, to get 'simple' or 'full' cube * new --licence option to specify license for the cube * added README and sobjects.py template (the later only added in full layout) * fixed python file templates to get proper headers according to the config

[storage] consider fs_importing on update operations too

[primary] fix name error

[migration] propagate TrInfo.comment permissions fix to existing instances

[xthml] fix #696079, potential empty table cause xhtml validation error

[schema] #847739: test and fix bad permission on TrInfo.comment

[iprogress] use xml_escape instead of buggy replace

backport stable

[querier] don't need a Cache instance for rql cacke key cache

[testlib] properly close connections opened during test in tearDown

#759035: Automate addition of eid cachekey in RQL analysis

[rql annotation] add test and fix for bug introduced by 5004:4cc020ee70e2. While it sounded good, it wasn't because stinfo['relations'] didn't include anymore stinfo['typerels']

merge

merge

merged back old local changeset

fixed indentation error in devtools/__init__.py

SQL Server port: temporary table handling Together with support from logilab.database, handle temp tables for SQLServer (which are denoted by a name starting with '#', just don't ask) Note updated dependency on logilab.database 1.0.2

log RQL query in case of unexpected failure (priceless when porting to new database)

enable live test against sqlserver 2005 database

fixed skeleton help strings

[wf] should use is_in_group to avoid error with internal manager

[pkginfo] when only __depends__ is specified, deduced __depends_cubes__ from it (same thing for recommends) ; fix ad skel pkginfo on the way

[hook] fix name error in case of error on hook's .event checking

[ms] fix planning bug introduced with recent rql refactoring for queries with only eids (fix unittest_multisources.test_in_eid)

[hook] proper bad .events detection. Fix unittest_hook

backport stable

[test] _gc_debug deleted from etwist.server

[doc/book] talk about reloadComponent, misc tweaks and notes

[doc/book] add q/a on i18n customisation, misc tweaks

Switched from TwistedWeb2 to TwistedWeb - added HTTPResponse class in etwist/http.py (could be then abstracted in cubicweb/web) - added twisted.web2 http_headers.py file in cubicweb/web to handle HTTP headers conversion between raw headers and python object - deleted caching for base views (except for startup views). A better solution would be using weak entity tags (but they don't seem to be implemented in twisted.web). - added forbidden access message when browsing static local directories - tested with TwistedWeb 8, 9 and 10 TODO: ===== - Handle file uploading in forms. twisted.web seems to keep very little information (only file content) about uploaded files in twisted_request.args['input_field_name']. But it doesn't seem to keep track of filenames. Possible solutions : - use web2 code to parse raw request content still stored and available in twisted_request.content - find a magic function in twisted.web API to get the filenames - More tests.

merge

[ldap] fix email synchronization code: bad rql query + avoid integrity error on creating the address + nicer use_email / primary_email handling

[doc/book] adjust a bit

[doc/book] expand the js chapter with an ajax story

[doc] remove useless directories

temporary move

[web components] refactor as book recommends :$

more / cleaner / in code documentation of vreg, selectors and appobject

add note about running repository / client code

added information about the naming conventions in schema.py also: * some typo fix * note about ObjectRelation soon being deprecated * clarified a few hazy points

[doc/book] a new chapter on how to use the ORM

improved doc on selectors an vregistry

fix rest formatting

typo fix

merge

catch and log errors caused by bad json data

list allowed values for ldap protocol setting in help, so that the generated file includes this in a comment

[doc/book] enhance i18n section #656191

[doc] more (doc oriented) feature in list

[doc] enhanced setup section: fix sections'level, more postgres tips, merge environment doc with cwconfig.py module docstring using custom autodocstring directive

[doc] enhanced concepts section

remove ref to no more existing file

[cleaning] simpler implementation of cmd_create_entity

[storages] fix fs_importing side-effect on entity.data When creating a new File object, if fs_importing is set, we want entity.data to be the file content instead of the filepath for the rest of the transaction. (see test_bfss_fs_importing_transparency) for test implementation To make this possible, the storage hooks (entity_added / entity_updated) must return the correct value to set in the entity dict.

nicer warning

[doc] started full features list

[python] take care to this detail of @contextmanager: if an unhandled exception occurs in the block, it is reraised inside the generator at the point where the yield occurred