Changeset - 3c503044e9f1
[Not reviewed]
default
0 3 0
Mads Kiilerich (mads) - 6 years ago 2020-07-01 14:58:56
mads@kiilerich.com
mysql: bump sqlalchemy.url MariaDB/MySQL charset to to 'utf8mb4' to get full UTF-8 support

The change in 210e76d69b62 only changed character_set_database, as shown by output after:

--- a/kallithea/model/base.py
+++ b/kallithea/model/base.py
@@ -46,3 +46,8 @@ def init_model(engine):
engine_str = obfuscate_url_pw(str(engine.url))
log.info("initializing db for %s", engine_str)
meta.Base.metadata.bind = engine
+
+ meta.Session.configure(bind=engine)
+ for a, b in meta.Session().execute('''show variables''').fetchall():
+ if 'character_set_' in a:
+ print(a, repr(b))

Before, with charset=utf8, the utf8mb3 charset was used all the way through the stack:

[kallithea.model.base] initializing db for mysql://kallithea-test:XXXXX@localhost/kallithea-test?charset=utf8
character_set_client 'utf8'
character_set_connection 'utf8'
character_set_database 'utf8mb4'
character_set_filesystem 'binary'
character_set_results 'utf8'
character_set_server 'latin1'
character_set_system 'utf8'

With explicit charset=utf8mb4:

[kallithea.model.base] initializing db for mysql://kallithea-test:XXXXX@localhost/kallithea-test?charset=utf8mb4
character_set_client 'utf8mb4'
character_set_connection 'utf8mb4'
character_set_database 'utf8mb4'
character_set_filesystem 'binary'
character_set_results 'utf8mb4'
character_set_server 'latin1'
character_set_system 'utf8'
3 files changed with 4 insertions and 4 deletions:
0 comments (0 inline, 0 general)
development.ini
Show inline comments
 
@@ -268,193 +268,193 @@ celery.worker_max_tasks_per_child = 1
 

	
 
## If true, tasks will never be sent to the queue, but executed locally instead.
 
celery.task_always_eager = false
 

	
 
####################################
 
##          BEAKER CACHE          ##
 
####################################
 

	
 
beaker.cache.data_dir = %(here)s/data/cache/data
 
beaker.cache.lock_dir = %(here)s/data/cache/lock
 

	
 
beaker.cache.regions = long_term,long_term_file
 

	
 
beaker.cache.long_term.type = memory
 
beaker.cache.long_term.expire = 36000
 
beaker.cache.long_term.key_length = 256
 

	
 
beaker.cache.long_term_file.type = file
 
beaker.cache.long_term_file.expire = 604800
 
beaker.cache.long_term_file.key_length = 256
 

	
 
####################################
 
##        BEAKER SESSION          ##
 
####################################
 

	
 
## Name of session cookie. Should be unique for a given host and path, even when running
 
## on different ports. Otherwise, cookie sessions will be shared and messed up.
 
session.key = kallithea
 
## Sessions should always only be accessible by the browser, not directly by JavaScript.
 
session.httponly = true
 
## Session lifetime. 2592000 seconds is 30 days.
 
session.timeout = 2592000
 

	
 
## Server secret used with HMAC to ensure integrity of cookies.
 
#session.secret = VERY-SECRET
 
session.secret = development-not-secret
 
## Further, encrypt the data with AES.
 
#session.encrypt_key = <key_for_encryption>
 
#session.validate_key = <validation_key>
 

	
 
## Type of storage used for the session, current types are
 
## dbm, file, memcached, database, and memory.
 

	
 
## File system storage of session data. (default)
 
#session.type = file
 

	
 
## Cookie only, store all session data inside the cookie. Requires secure secrets.
 
#session.type = cookie
 

	
 
## Database storage of session data.
 
#session.type = ext:database
 
#session.sa.url = postgresql://postgres:qwe@localhost/kallithea
 
#session.table_name = db_session
 

	
 
####################################
 
##        ERROR HANDLING          ##
 
####################################
 

	
 
## Show a nice error page for application HTTP errors and exceptions (default true)
 
#errorpage.enabled = true
 

	
 
## Enable Backlash client-side interactive debugger (default false)
 
## WARNING: *THIS MUST BE false IN PRODUCTION ENVIRONMENTS!!!*
 
## This debug mode will allow all visitors to execute malicious code.
 
#debug = false
 
debug = true
 

	
 
## Enable Backlash server-side error reporting (unless debug mode handles it client-side) (default true)
 
#trace_errors.enable = true
 
## Errors will be reported by mail if trace_errors.error_email is set.
 

	
 
## Propagate email settings to ErrorReporter of TurboGears2
 
## You do not normally need to change these lines
 
get trace_errors.smtp_server = smtp_server
 
get trace_errors.smtp_port = smtp_port
 
get trace_errors.from_address = error_email_from
 
get trace_errors.error_email = email_to
 
get trace_errors.smtp_username = smtp_username
 
get trace_errors.smtp_password = smtp_password
 
get trace_errors.smtp_use_tls = smtp_use_tls
 

	
 

	
 
##################################
 
##        LOGVIEW CONFIG        ##
 
##################################
 

	
 
logview.sqlalchemy = #faa
 
logview.pylons.templating = #bfb
 
logview.pylons.util = #eee
 

	
 
#########################
 
##      DB CONFIG      ##
 
#########################
 

	
 
sqlalchemy.url = sqlite:///%(here)s/kallithea.db?timeout=60
 
#sqlalchemy.url = postgresql://kallithea:password@localhost/kallithea
 
#sqlalchemy.url = mysql://kallithea:password@localhost/kallithea?charset=utf8
 
#sqlalchemy.url = mysql://kallithea:password@localhost/kallithea?charset=utf8mb4
 
## Note: the mysql:// prefix should also be used for MariaDB
 

	
 
sqlalchemy.pool_recycle = 3600
 

	
 
################################
 
##   ALEMBIC CONFIGURATION    ##
 
################################
 

	
 
[alembic]
 
script_location = kallithea:alembic
 

	
 
################################
 
##   LOGGING CONFIGURATION    ##
 
################################
 

	
 
[loggers]
 
keys = root, routes, kallithea, sqlalchemy, tg, gearbox, beaker, templates, whoosh_indexer, werkzeug, backlash
 

	
 
[handlers]
 
keys = console, console_color, console_color_sql, null
 

	
 
[formatters]
 
keys = generic, color_formatter, color_formatter_sql
 

	
 
#############
 
## LOGGERS ##
 
#############
 

	
 
[logger_root]
 
level = NOTSET
 
#handlers = console
 
## For coloring based on log level:
 
handlers = console_color
 

	
 
[logger_routes]
 
#level = WARN
 
level = DEBUG
 
handlers =
 
qualname = routes.middleware
 
## "level = DEBUG" logs the route matched and routing variables.
 

	
 
[logger_beaker]
 
#level = WARN
 
level = DEBUG
 
handlers =
 
qualname = beaker.container
 

	
 
[logger_templates]
 
#level = WARN
 
level = INFO
 
handlers =
 
qualname = pylons.templating
 

	
 
[logger_kallithea]
 
#level = WARN
 
level = DEBUG
 
handlers =
 
qualname = kallithea
 

	
 
[logger_tg]
 
#level = WARN
 
level = DEBUG
 
handlers =
 
qualname = tg
 

	
 
[logger_gearbox]
 
#level = WARN
 
level = DEBUG
 
handlers =
 
qualname = gearbox
 

	
 
[logger_sqlalchemy]
 
level = WARN
 
handlers =
 
qualname = sqlalchemy.engine
 
## For coloring based on log level and pretty printing of SQL:
 
#level = INFO
 
#handlers = console_color_sql
 
#propagate = 0
 

	
 
[logger_whoosh_indexer]
 
#level = WARN
 
level = DEBUG
 
handlers =
 
qualname = whoosh_indexer
 

	
 
[logger_werkzeug]
 
level = WARN
 
handlers =
 
qualname = werkzeug
 

	
 
[logger_backlash]
 
level = WARN
 
handlers =
 
qualname = backlash
 

	
docs/contributing.rst
Show inline comments
 
@@ -8,193 +8,193 @@ Kallithea is developed and maintained by
 
your own itch.
 

	
 

	
 
Infrastructure
 
--------------
 

	
 
The main repository is hosted on Our Own Kallithea (aka OOK) at
 
https://kallithea-scm.org/repos/kallithea/, our self-hosted instance
 
of Kallithea.
 

	
 
For now, we use Bitbucket_ for `pull requests`_ and `issue tracking`_. The
 
issue tracker is for tracking bugs, not for support, discussion, or ideas --
 
please use the `mailing list`_ or :ref:`IRC <readme>` to reach the community.
 

	
 
We use Weblate_ to translate the user interface messages into languages other
 
than English. Join our project on `Hosted Weblate`_ to help us.
 
To register, you can use your Bitbucket or GitHub account. See :ref:`translations`
 
for more details.
 

	
 

	
 
Getting started
 
---------------
 

	
 
To get started with Kallithea development::
 

	
 
        hg clone https://kallithea-scm.org/repos/kallithea
 
        cd kallithea
 
        python3 -m venv ../kallithea-venv
 
        source ../kallithea-venv/bin/activate
 
        pip install --upgrade pip setuptools
 
        pip install --upgrade -e . -r dev_requirements.txt python-ldap python-pam
 
        kallithea-cli config-create my.ini
 
        kallithea-cli db-create -c my.ini --user=user --email=user@example.com --password=password --repos=/tmp
 
        kallithea-cli front-end-build
 
        gearbox serve -c my.ini --reload &
 
        firefox http://127.0.0.1:5000/
 

	
 
If you plan to use Bitbucket_ for sending contributions, you can also fork
 
Kallithea on Bitbucket_ first (https://bitbucket.org/conservancy/kallithea) and
 
then replace the clone step above by a clone of your fork. In this case, please
 
see :ref:`contributing-guidelines` below for configuring your fork correctly.
 

	
 

	
 
Contribution flow
 
-----------------
 

	
 
Starting from an existing Kallithea clone, make sure it is up to date with the
 
latest upstream changes::
 

	
 
        hg pull
 
        hg update
 

	
 
Review the :ref:`contributing-guidelines` and :ref:`coding-guidelines`.
 

	
 
If you are new to Mercurial, refer to Mercurial `Quick Start`_ and `Beginners
 
Guide`_ on the Mercurial wiki.
 

	
 
Now, make some changes and test them (see :ref:`contributing-tests`). Don't
 
forget to add new tests to cover new functionality or bug fixes.
 

	
 
For documentation changes, run ``make html`` from the ``docs`` directory to
 
generate the HTML result, then review them in your browser.
 

	
 
Before submitting any changes, run the cleanup script::
 

	
 
        ./scripts/run-all-cleanup
 

	
 
When you are completely ready, you can send your changes to the community for
 
review and inclusion. Most commonly used methods are sending patches to the
 
mailing list (via ``hg email``) or by creating a pull request on Bitbucket_.
 

	
 
.. _contributing-tests:
 

	
 

	
 
Running tests
 
-------------
 

	
 
After finishing your changes make sure all tests pass cleanly. Run the testsuite
 
by invoking ``py.test`` from the project root::
 

	
 
    py.test
 

	
 
Note that on unix systems, the temporary directory (``/tmp`` or where
 
``$TMPDIR`` points) must allow executable files; Git hooks must be executable,
 
and the test suite creates repositories in the temporary directory. Linux
 
systems with /tmp mounted noexec will thus fail.
 

	
 
Tests can be run on PostgreSQL like::
 

	
 
    sudo -u postgres createuser 'kallithea-test' --pwprompt  # password password
 
    sudo -u postgres createdb 'kallithea-test' --owner 'kallithea-test'
 
    REUSE_TEST_DB='postgresql://kallithea-test:password@localhost/kallithea-test' py.test
 

	
 
Tests can be run on MariaDB/MySQL like::
 

	
 
    echo "GRANT ALL PRIVILEGES ON \`kallithea-test\`.* TO 'kallithea-test'@'localhost' IDENTIFIED BY 'password'" | sudo -u mysql mysql
 
    TEST_DB='mysql://kallithea-test:password@localhost/kallithea-test?charset=utf8' py.test
 
    TEST_DB='mysql://kallithea-test:password@localhost/kallithea-test?charset=utf8mb4' py.test
 

	
 
You can also use ``tox`` to run the tests with all supported Python versions.
 

	
 
When running tests, Kallithea generates a `test.ini` based on template values
 
in `kallithea/tests/conftest.py` and populates the SQLite database specified
 
there.
 

	
 
It is possible to avoid recreating the full test database on each invocation of
 
the tests, thus eliminating the initial delay. To achieve this, run the tests as::
 

	
 
    gearbox serve -c /tmp/kallithea-test-XXX/test.ini --pid-file=test.pid --daemon
 
    KALLITHEA_WHOOSH_TEST_DISABLE=1 KALLITHEA_NO_TMP_PATH=1 py.test
 
    kill -9 $(cat test.pid)
 

	
 
In these commands, the following variables are used::
 

	
 
    KALLITHEA_WHOOSH_TEST_DISABLE=1 - skip whoosh index building and tests
 
    KALLITHEA_NO_TMP_PATH=1 - disable new temp path for tests, used mostly for testing_vcs_operations
 

	
 
You can run individual tests by specifying their path as argument to py.test.
 
py.test also has many more options, see `py.test -h`. Some useful options
 
are::
 

	
 
    -k EXPRESSION         only run tests which match the given substring
 
                          expression. An expression is a python evaluable
 
                          expression where all names are substring-matched
 
                          against test names and their parent classes. Example:
 
    -x, --exitfirst       exit instantly on first error or failed test.
 
    --lf                  rerun only the tests that failed at the last run (or
 
                          all if none failed)
 
    --ff                  run all tests but run the last failures first. This
 
                          may re-order tests and thus lead to repeated fixture
 
                          setup/teardown
 
    --pdb                 start the interactive Python debugger on errors.
 
    -s, --capture=no      don't capture stdout (any stdout output will be
 
                          printed immediately)
 

	
 
Performance tests
 
^^^^^^^^^^^^^^^^^
 

	
 
A number of performance tests are present in the test suite, but they are
 
not run in a standard test run. These tests are useful to
 
evaluate the impact of certain code changes with respect to performance.
 

	
 
To run these tests::
 

	
 
    env TEST_PERFORMANCE=1 py.test kallithea/tests/performance
 

	
 
To analyze performance, you could install pytest-profiling_, which enables the
 
--profile and --profile-svg options to py.test.
 

	
 
.. _pytest-profiling: https://github.com/manahl/pytest-plugins/tree/master/pytest-profiling
 

	
 
.. _contributing-guidelines:
 

	
 

	
 
Contribution guidelines
 
-----------------------
 

	
 
Kallithea is GPLv3 and we assume all contributions are made by the
 
committer/contributor and under GPLv3 unless explicitly stated. We do care a
 
lot about preservation of copyright and license information for existing code
 
that is brought into the project.
 

	
 
Contributions will be accepted in most formats -- such as pull requests on
 
Bitbucket, something hosted on your own Kallithea instance, or patches sent by
 
email to the `kallithea-general`_ mailing list.
 

	
 
When contributing via Bitbucket, please make your fork of
 
https://bitbucket.org/conservancy/kallithea/ `non-publishing`_ -- it is one of
 
the settings on "Repository details" page. This ensures your commits are in
 
"draft" phase and makes it easier for you to address feedback and for project
 
maintainers to integrate your changes.
 

	
 
.. _non-publishing: https://www.mercurial-scm.org/wiki/Phases#Publishing_Repository
 

	
 
Make sure to test your changes both manually and with the automatic tests
 
before posting.
 

	
 
We care about quality and review and keeping a clean repository history. We
 
might give feedback that requests polishing contributions until they are
 
"perfect". We might also rebase and collapse and make minor adjustments to your
 
changes when we apply them.
 

	
 
We try to make sure we have consensus on the direction the project is taking.
 
Everything non-sensitive should be discussed in public -- preferably on the
 
mailing list.  We aim at having all non-trivial changes reviewed by at least
 
one other core developer before pushing. Obvious non-controversial changes will
 
be handled more casually.
 

	
 
There is a main development branch ("default") which is generally stable so that
 
it can be (and is) used in production. There is also a "stable" branch that is
 
almost exclusively reserved for bug fixes or trivial changes. Experimental
 
changes should live elsewhere (for example in a pull request) until they are
 
ready.
 

	
kallithea/lib/paster_commands/template.ini.mako
Show inline comments
 
@@ -370,195 +370,195 @@ beaker.cache.long_term_file.key_length =
 
session.key = kallithea
 
<%text>##</%text> Sessions should always only be accessible by the browser, not directly by JavaScript.
 
session.httponly = true
 
<%text>##</%text> Session lifetime. 2592000 seconds is 30 days.
 
session.timeout = 2592000
 

	
 
<%text>##</%text> Server secret used with HMAC to ensure integrity of cookies.
 
session.secret = ${uuid()}
 
<%text>##</%text> Further, encrypt the data with AES.
 
#session.encrypt_key = <key_for_encryption>
 
#session.validate_key = <validation_key>
 

	
 
<%text>##</%text> Type of storage used for the session, current types are
 
<%text>##</%text> dbm, file, memcached, database, and memory.
 

	
 
<%text>##</%text> File system storage of session data. (default)
 
#session.type = file
 

	
 
<%text>##</%text> Cookie only, store all session data inside the cookie. Requires secure secrets.
 
#session.type = cookie
 

	
 
<%text>##</%text> Database storage of session data.
 
#session.type = ext:database
 
#session.sa.url = postgresql://postgres:qwe@localhost/kallithea
 
#session.table_name = db_session
 

	
 
<%text>##</%text>##################################
 
<%text>##</%text>        ERROR HANDLING          ##
 
<%text>##</%text>##################################
 

	
 
<%text>##</%text> Show a nice error page for application HTTP errors and exceptions (default true)
 
#errorpage.enabled = true
 

	
 
<%text>##</%text> Enable Backlash client-side interactive debugger (default false)
 
<%text>##</%text> WARNING: *THIS MUST BE false IN PRODUCTION ENVIRONMENTS!!!*
 
<%text>##</%text> This debug mode will allow all visitors to execute malicious code.
 
#debug = false
 

	
 
<%text>##</%text> Enable Backlash server-side error reporting (unless debug mode handles it client-side) (default true)
 
#trace_errors.enable = true
 
<%text>##</%text> Errors will be reported by mail if trace_errors.error_email is set.
 

	
 
<%text>##</%text> Propagate email settings to ErrorReporter of TurboGears2
 
<%text>##</%text> You do not normally need to change these lines
 
get trace_errors.smtp_server = smtp_server
 
get trace_errors.smtp_port = smtp_port
 
get trace_errors.from_address = error_email_from
 
get trace_errors.error_email = email_to
 
get trace_errors.smtp_username = smtp_username
 
get trace_errors.smtp_password = smtp_password
 
get trace_errors.smtp_use_tls = smtp_use_tls
 

	
 
%if error_aggregation_service == 'sentry':
 
<%text>##</%text>##############
 
<%text>##</%text>  [sentry]  ##
 
<%text>##</%text>##############
 

	
 
<%text>##</%text> sentry is a alternative open source error aggregator
 
<%text>##</%text> you must install python packages `sentry` and `raven` to enable
 

	
 
sentry.dsn = YOUR_DNS
 
sentry.servers =
 
sentry.name =
 
sentry.key =
 
sentry.public_key =
 
sentry.secret_key =
 
sentry.project =
 
sentry.site =
 
sentry.include_paths =
 
sentry.exclude_paths =
 

	
 
%endif
 

	
 
<%text>##</%text>################################
 
<%text>##</%text>        LOGVIEW CONFIG        ##
 
<%text>##</%text>################################
 

	
 
logview.sqlalchemy = #faa
 
logview.pylons.templating = #bfb
 
logview.pylons.util = #eee
 

	
 
<%text>##</%text>#######################
 
<%text>##</%text>      DB CONFIG      ##
 
<%text>##</%text>#######################
 

	
 
%if database_engine == 'sqlite':
 
sqlalchemy.url = sqlite:///%(here)s/kallithea.db?timeout=60
 
%else:
 
#sqlalchemy.url = sqlite:///%(here)s/kallithea.db?timeout=60
 
%endif
 
%if database_engine == 'postgres':
 
sqlalchemy.url = postgresql://kallithea:password@localhost/kallithea
 
%else:
 
#sqlalchemy.url = postgresql://kallithea:password@localhost/kallithea
 
%endif
 
%if database_engine == 'mysql':
 
sqlalchemy.url = mysql://kallithea:password@localhost/kallithea?charset=utf8
 
sqlalchemy.url = mysql://kallithea:password@localhost/kallithea?charset=utf8mb4
 
%else:
 
#sqlalchemy.url = mysql://kallithea:password@localhost/kallithea?charset=utf8
 
#sqlalchemy.url = mysql://kallithea:password@localhost/kallithea?charset=utf8mb4
 
%endif
 
<%text>##</%text> Note: the mysql:// prefix should also be used for MariaDB
 

	
 
sqlalchemy.pool_recycle = 3600
 

	
 
<%text>##</%text>##############################
 
<%text>##</%text>   ALEMBIC CONFIGURATION    ##
 
<%text>##</%text>##############################
 

	
 
[alembic]
 
script_location = kallithea:alembic
 

	
 
<%text>##</%text>##############################
 
<%text>##</%text>   LOGGING CONFIGURATION    ##
 
<%text>##</%text>##############################
 

	
 
[loggers]
 
keys = root, routes, kallithea, sqlalchemy, tg, gearbox, beaker, templates, whoosh_indexer, werkzeug, backlash
 

	
 
[handlers]
 
keys = console, console_color, console_color_sql, null
 

	
 
[formatters]
 
keys = generic, color_formatter, color_formatter_sql
 

	
 
<%text>##</%text>###########
 
<%text>##</%text> LOGGERS ##
 
<%text>##</%text>###########
 

	
 
[logger_root]
 
level = NOTSET
 
handlers = console
 
<%text>##</%text> For coloring based on log level:
 
#handlers = console_color
 

	
 
[logger_routes]
 
level = WARN
 
handlers =
 
qualname = routes.middleware
 
<%text>##</%text> "level = DEBUG" logs the route matched and routing variables.
 

	
 
[logger_beaker]
 
level = WARN
 
handlers =
 
qualname = beaker.container
 

	
 
[logger_templates]
 
level = WARN
 
handlers =
 
qualname = pylons.templating
 

	
 
[logger_kallithea]
 
level = WARN
 
handlers =
 
qualname = kallithea
 

	
 
[logger_tg]
 
level = WARN
 
handlers =
 
qualname = tg
 

	
 
[logger_gearbox]
 
level = WARN
 
handlers =
 
qualname = gearbox
 

	
 
[logger_sqlalchemy]
 
level = WARN
 
handlers =
 
qualname = sqlalchemy.engine
 
<%text>##</%text> For coloring based on log level and pretty printing of SQL:
 
#level = INFO
 
#handlers = console_color_sql
 
#propagate = 0
 

	
 
[logger_whoosh_indexer]
 
level = WARN
 
handlers =
 
qualname = whoosh_indexer
 

	
 
[logger_werkzeug]
 
level = WARN
 
handlers =
 
qualname = werkzeug
 

	
 
[logger_backlash]
 
level = WARN
 
handlers =
 
qualname = backlash
 

	
 
<%text>##</%text>############
 
<%text>##</%text> HANDLERS ##
 
<%text>##</%text>############
 

	
 
[handler_console]
 
class = StreamHandler
0 comments (0 inline, 0 general)