### Changed
- Moved `\tl_if_single_token:n(TF)` to stable
- Moved `\file_get_mdfive_hash:nN(TF)`, `\file_get_size:nN(TF)`
and `\file_get_timestamp:nN(TF)` to stable
- Moved `\file_if_exist_input:n` and `\file_if_exist_input:nF` to stable
- Moved `\file_input_stop:` to stable
- Moved `\peek_N_type:TF` to stable
- Experimental `\driver_pdf_compress_objects:n` replaces
`\driver_pdf_objects_(en|dis)able:`
### Fixed
- Support for ConTeXt from mid-December 2018
- Infinite loop in some cases in DVI mode due to link-breaking code
- Category code of output from `\char_<target>_case:N`, and
same issue in `\str_<target>_case:n`
- Some issues in `dvisvgm` driver
### Added
- Experimental support for a range of PDF concepts at the lowest
(driver abstraction) level
- New `l3legacy` module containing
- `\legacy_if:n(TF)`
- `\char_str_<target>_case:N`
- `.muskip:N` property
- Experimental driver-level interfaces for image inclusion
- Experimental `\ior_shell_open:Nn`
pytest 4.5.0:
Features
- A warning is now emitted when unknown marks are used as a decorator.
This is often due to a typo, which can lead to silently broken tests.
- Show XFail reason as part of JUnitXML message field.
- Messages from crash reports are displayed within test summaries now, truncated to the terminal width.
- New flag --strict-markers that triggers an error when unknown markers (e.g. those not registered using the markers option_ in the configuration file) are used in the test suite.
The existing --strict option has the same behavior currently, but can be augmented in the future for additional checks.
- Assertion failure messages for sequences and dicts contain the number of different items now.
- Improve reporting with --lf and --ff (run-last-failure).
- The --cache-show option/action accepts an optional glob to show only matching cache entries.
- Standard input (stdin) can be given to pytester's Testdir.run() and Testdir.popen().
- The -r option learnt about A to display all reports (including passed ones) in the short test summary.
- The short test summary is displayed after passes with output (-rP).
- The --last-failed (--lf) option got smarter and will now skip entire files if all tests
of that test file have passed in previous runs, greatly speeding up collection.
- Introduce new specific warning PytestWarning subclasses to make it easier to filter warnings based on the class, rather than on the message. The new subclasses are:
* PytestAssertRewriteWarning
* PytestCacheWarning
* PytestCollectionWarning
* PytestConfigWarning
* PytestUnhandledCoroutineWarning
* PytestUnknownMarkWarning
- New record_testsuite_property session-scoped fixture allows users to log <property> tags at the testsuite
level with the junitxml plugin.
The generated XML is compatible with the latest xunit standard, contrary to
the properties recorded by record_property and record_xml_attribute.
- The default logging format has been changed to improve readability. Here is an
example of a previous logging message::
test_log_cli_enabled_disabled.py 3 CRITICAL critical message logged by test
This has now become::
CRITICAL root:test_log_cli_enabled_disabled.py:3 critical message logged by test
The formatting can be changed through the log_format <https://docs.pytest.org/en/latest/reference.html#confval-log_format>__ configuration option.
- --fixtures now also shows fixture scope for scopes other than "function".
Bug Fixes
- Deselected items from plugins using pytest_collect_modifyitems as a hookwrapper are correctly reported now.
- With usage errors exitstatus is set to EXIT_USAGEERROR in the pytest_sessionfinish hook now as expected.
- outcome.exit is not used with EOF in the pdb wrapper anymore, but only with quit.
Improved Documentation
- Expand docs on registering marks and the effect of --strict.
Trivial/Internal Changes
- logging.raiseExceptions is not set to False anymore.
- pytest now depends on wcwidth <https://pypi.org/project/wcwidth>__ to properly track unicode character sizes for more precise terminal output.
- pytester's Testdir.popen() uses stdout and stderr via keyword arguments with defaults now (subprocess.PIPE).
- The code for the short test summary in the terminal was moved to the terminal plugin.
- Improved validation of kwargs for various methods in the pytester plugin.
- record_property now emits a PytestWarning when used with junit_family=xunit2: the fixture generates
property tags as children of testcase, which is not permitted according to the most
recent schema <https://github.com/jenkinsci/xunit-plugin/blob/master/
src/main/resources/org/jenkinsci/plugins/xunit/types/model/xsd/junit-10.xsd>__.
- Pin pluggy to < 1.0 so we don't update to 1.0 automatically when
it gets released: there are planned breaking changes, and we want to ensure
pytest properly supports pluggy 1.0.
tools/fiptool: Fix UUID parsing in blob handling
Commit 0336486 ("Make TF UUID RFC 4122 compliant") changed the scanf
parsing string to handle endianness correctly.
However that changed the number of items sscanf handles, without
adjusting the sanity check just below.
Increase the expected return value from 11 to 16 to let fiptool handle
UUIDs given as blob parameters correctly again.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
With 37, the build fails with complaints about future and the wrong
version. With 27, it builds. Definitive documentation is scarce, but
signs point to 27 being standard for 2.18, and 3 perhaps working,
perhaps not. Because qgis 2.18 should be updated to 3.x anyway, don't
worry and just set the variables for now.
Changes:
0.9.11
------
- Remove sfeed_tail: it did not worked well with FIFOs and a tail-like
example is now documented in README
- sfeed_plain, start with timestamp, similarly to sfeed_twtxt
- Optimize sfeed_opml_export
- New package option 'disable' has been added.
- New command \addtodelayedfloat has beed added.
- New macros \efloatbegin, \efloatend, \efloatbeginlist, and \efloatendlist
for customization have been added.
- The bugfix regarding the usage of UTF-8 characters in floating
environments has been revised.
- spurious blank in sorting code removed
- package 'pgfrcs' added as required package
- introduction augmented with additional hints for first time users
PostgreSQL 11.3, 10.8, 9.6.13, 9.5.17, and 9.4.22 Released!
The PostgreSQL Global Development Group has released an update to all supported versions of our database system, including 11.3, 10.8, 9.6.13, 9.5.17, and 9.4.22. This release fixes two security issues in the PostgreSQL server, a security issue found in two of the PostgreSQL Windows installers, and over 60 bugs reported over the last three months.
Security Issues
Four security vulnerabilities have been closed by this release:
CVE-2019-10127: BigSQL Windows installer does not clear permissive ACL entries
CVE-2019-10128: EnterpriseDB Windows installer does not clear permissive ACL entries
Due to both the EnterpriseDB and BigSQL Windows installers not locking down the permissions of the PostgreSQL binary installation directory and the data directory, an unprivileged Windows user account and an unprivileged PostgreSQL account could cause the PostgreSQL service account to execute arbitrary code.
This vulnerability is present in all supported versions of PostgreSQL for these installers, and possibly exists in older versions. Both sets of installers have fixed the permissions for these directories for both new and existing installations. If you have installed PostgreSQL on Windows using other methods, we advise that you check that your PostgreSQL binary directories are writable only to trusted users and that your data directories are only accessible to trusted users.
The PostgreSQL project thanks Conner Jones for reporting this problem.
CVE-2019-10129: Memory disclosure in partition routing
Prior to this release, a user running PostgreSQL 11 can read arbitrary bytes of server memory by executing a purpose-crafted INSERT statement to a partitioned table.
CVE-2019-10130: Selectivity estimators bypass row security policies
PostgreSQL maintains statistics for tables by sampling data available in columns; this data is consulted during the query planning process. Prior to this release, a user able to execute SQL queries with permissions to read a given column could craft a leaky operator that could read whatever data had been sampled from that column. If this happened to include values from rows that the user is forbidden to see by a row security policy, the user could effectively bypass the policy. This is fixed by only allowing a non-leakproof operator to use this data if there are no relevant row security policies for the table.
This issue is present in PostgreSQL 9.5, 9.6, 10, and 11. The PostgreSQL project thanks Dean Rasheed for reporting this problem.
Bug Fixes and Improvements
This update also fixes over 60 bugs that were reported in the last several months. Some of these issues affect only version 11, but many affect all supported versions.
Some of these fixes include:
Several catalog corruption fixes, including one related to running ALTER TABLE on a partitioned table
Several fixes for partitioning
Avoid server crash when an error occurs while trying to persist a cursor query across a transaction commit
Avoid O(N^2) performance issue when rolling back a transaction that created many tables
Fix possible “could not access status of transaction” failures in txid_status()
Fix updatable views to handle explicit DEFAULT items in INSERT .. VALUES statements where there are multiple VALUES rows
Fix CREATE VIEW to allow zero-column views
Add missing support for the CREATE TABLE IF NOT EXISTS .. AS EXECUTE .. statement
Ensure that sub-SELECTs appearing in row-level-security policy expressions are executed with the correct user's permissions
Accept XML documents as valid values of type xml when xmloption is set to content, as required by SQL:2006 and later
Fix incompatibility of GIN-index WAL records that were introduced in 11.2, 10.7, 9.6.12, 9.5.16, and 9.4.21 that affected replica servers running these versions reading in changes to GIN indexes from primary servers of older versions
Several memory leak fixes as well as fixes to management of dynamic shared memory
Relax panics on fsync and sync_file_range failures for certain cases where a failure indicated "operation not supported"
Several fixes to the query planner, several of which should lead to planning improvements
Fix race condition in which a hot-standby postmaster could fail to shut down after receiving a smart-shutdown request
Several fixes for SCRAM authentication
Fix handling of lc_time settings that imply an encoding different from the database's encoding
Create the current_logfiles file with the same permissions as other files in the server's data directory
Several ecpg fixes
Make pg_verify_checksums verify that the data directory it's pointed at is of the right PostgreSQL version
Several fixes for contrib/postgres_fdw, including one for remote partitions where an UPDATE could lead to incorrect results or a crash
Several Windows fixes
This update also contains tzdata release 2019a for DST law changes in Palestine and Metlakatla, plus historical corrections for Israel. Etc/UCT is now a backward-compatibility link to Etc/UTC, instead of being a separate zone that generates the abbreviation UCT, which nowadays is typically a typo. PostgreSQL will still accept UCT as an input zone abbreviation, but it won't output it.
Changelog:
2019.1.2
This update comes with a bunch of notable fixes:
Fixed several major Gradle issues: IDEA-210487, IDEA-210487, IDEA-178417.
The “Open Recent” menu can now display the path to a project: IDEA-168223.
The terminal tabs can now be reordered with Ctrl+Shift+Left/Right on Windows / Linux, and cmd+Shift+Left/Right on macOS: IDEA-202809.
It’s now possible to build a Groovy project using the Groovy-Eclipse compiler: IDEA-207921.
The Recent Locations popup now opens all the selected locations: IDEA-209286
The Run Anything popup shows the correct list of Maven goals: IDEA-199283, and shows correct suggestions on the first run: IDEA-210716.
2019.1.1
Out of dozens of important fixes this update delivers, here are the most important ones:
Performance issues: the IDE doesn’t freeze on Auto-import (IDEA-206649) and we’ve fixed the issue that was causing very high memory usage on relatively small projects (IDEA-210047).
Regressions fixed: The ‘Start Failed’ dialog doesn’t appear on closing the import settings dialog, and VCS refreshes don’t take very long with Mercurial anymore: IDEA-209262.
Updating plugins in an offline environment now works with the new marketplace and custom plugin repository: IDEA-205032.
No confirmation dialog is shown when you try to close a Terminal without running processes: IDEA-205690.
It’s now possible to switch back from the Terminal tool window to the editor by pressing Esc: IDEA-116221.
In the Run Configuration dialog, you can paste just the name or just the value of an Environment Variable: IDEA-206952.
The IDE now correctly creates a branch when you open a task with the ‘create branch’ option enabled: IDEA-205194.
It’s now possible to type characters that require AltGr (IntelliJ IDEA with JBR11): IDEA-209426.
For those using the IdeaVim plugin, the Esc key in insert mode closes the autocomplete popup and exits the insert mode: IDEA-209779.
The bundled Kotlin plugin was updated to v1.3.21-release-IJ2019.1-3, with the following fixes integrated: KT-30117, KT-29427, KT-30137, and KT-9618.
There is no changes file or list of changes that I can find.
This version does generate files that are more compatible with newer
versions of eliom (removes deprecated types).
Inkscape 0.92.4 is a stability and bugfix release. It also brings some
notable performance improvements for filter rendering, measure tool,
extensions, document saving and moving paths (with a live path effect)
and includes a couple of small, but impactful usage improvements.
3.0.1:
Allow 'null' in CORS_ORIGIN_WHITELIST check.
3.0.0:
CORS_ORIGIN_WHITELIST now requires URI schemes, and optionally ports. This is part of the CORS specification (Section 3.2) that was not implemented in this library, except from with the CORS_ORIGIN_REGEX_WHITELIST setting. It fixes a security issue where the CORS middleware would allow requests between schemes, for example from insecure http:// Origins to a secure https:// site.
You will need to update your whitelist to include schemes, for example from this:
CORS_ORIGIN_WHITELIST = ['example.com']
...to this:
CORS_ORIGIN_WHITELIST = ['https://example.com']
Removed the CORS_MODEL setting, and associated class. It seems very few, or no users were using it, since there were no bug reports since its move to abstract in version 2.0.0 (2017-01-07). If you are using this functionality, you can continue by changing your model to not inherit from the abstract one, and add a signal handler for check_request_enabled that reads from your model. Note you'll need to handle the move to include schemes for Origins.
