What’s new in Tornado 6.1.0
Oct 30, 2020
Deprecation notice
This is the last release of Tornado to support Python 3.5. Future versions will require Python 3.6 or newer.
General changes
Windows support has been improved. Tornado is now compatible with the proactor event loop (which became the default in Python 3.8) by automatically falling back to running a selector in a second thread. This means that it is no longer necessary to explicitly configure a selector event loop, although doing so may improve performance. This does not change the fact that Tornado is significantly less scalable on Windows than on other platforms.
Binary wheels are now provided for Windows, MacOS, and Linux (amd64 and arm64).
tornado.gen
coroutine now has better support for the Python 3.7+ contextvars module. In particular, the ContextVar.reset method is now supported.
tornado.http1connection
HEAD requests to handlers that used chunked encoding no longer produce malformed output.
Certain kinds of malformed gzip data no longer cause an infinite loop.
tornado.httpclient
Setting decompress_response=False now works correctly with curl_httpclient.
Mixing requests with and without proxies works correctly in curl_httpclient (assuming the version of pycurl is recent enough).
A default User-Agent of Tornado/$VERSION is now used if the user_agent parameter is not specified.
After a 303 redirect, tornado.simple_httpclient always uses GET. Previously this would use GET if the original request was a POST and would otherwise reuse the original request method. For curl_httpclient, the behavior depends on the version of libcurl (with the most recent versions using GET after 303 regardless of the original method).
Setting request_timeout and/or connect_timeout to zero is now supported to disable the timeout.
tornado.httputil
Header parsing is now faster.
parse_body_arguments now accepts incompletely-escaped non-ASCII inputs.
tornado.iostream
ssl.CertificateError during the SSL handshake is now handled correctly.
Reads that are resolved while the stream is closing are now handled correctly.
tornado.log
When colored logging is enabled, logging.CRITICAL messages are now recognized and colored magenta.
tornado.netutil
EADDRNOTAVAIL is now ignored when binding to localhost with IPv6. This error is common in docker.
tornado.platform.asyncio
AnyThreadEventLoopPolicy now also configures a selector event loop for these threads (the proactor event loop only works on the main thread)
tornado.platform.auto
The set_close_exec function has been removed.
tornado.testing
ExpectLog now has a level argument to ensure that the given log level is enabled.
tornado.web
RedirectHandler.get now accepts keyword arguments.
When sending 304 responses, more headers (including Allow) are now preserved.
reverse_url correctly handles escaped characters in the regex route.
Default Etag headers are now generated with SHA-512 instead of MD5.
tornado.websocket
The ping_interval timer is now stopped when the connection is closed.
websocket_connect now raises an error when it encounters a redirect instead of hanging.
For older versions look here: https://www.tornadoweb.org/en/stable/releases.html
* Fix comments.
Changelog:
7.3.0.3
Writer
General improvements
* Hyperlinks can now be attached to shapes tdf#123626 (Samuel Mehrbrodt,
allotropia)
* Footnote and endnote anchors in the end of a paragraph are now treated
exactly as they already were treated in the middle for the purposes of
search/replace with regular expressions, i.e. they match [\p{Control}] or
[:control:] regular expression. That means that they need to be taken into
account when matching text against end-of-paragraph tdf#137737 (Mike
Kaganski, Collabora)
* Start of list level support in Writer paragraph styles blog post (Miklos
Vajna, Collabora)
* Start of linked character and paragraph styles blog post (Miklos Vajna,
Collabora)
Tracked table row deletion/insertion
Add visualization and Manage Changes support for tracked deletion and insertion
of tables and table rows. (L??szl?? N??meth, NISZ)
Basic support of table change tracking (introduced by LibreOffice 7.2, see tdf#
60382, tdf#79069) was extended by the following improvements:
* Hide deleted table rows and tables in Hide Changes mode instead of showing
empty cells tdf#144057
* Show table and table row insertions in cyan and deletions in pink in Show
Changes mode tdf#146120
* In Manage Changes, accept or reject deletion/insertion of tables or
consecutive table rows with a single click tdf#144270
* Clicking on a table row, accept or reject its deletion/insertion tdf#146145
* Add tooltip to table rows with change tracking tdf#146144
* Track insertion of empty table rows tdf#143358
* Track deletion of empty table rows tdf#143359
Highlighted tracked row changes, and improved Manage Changes dialog window with
the new table (row) changes
Tracked text moving
During track changes, show moved text in green color and with double
strikethrough or underlines to speed up reviewing. Add new tooltip captions and
Manage Changes Action icons for tracked text moving. Re-ordered list elements
or changed paragraph or sentence order is more visible this way. Tracked moving
is detected at Move Up/Down, drag & drop, cut & paste and deletion + re-typing,
also during ODT and DOCX import (including documents created with editors which
lack of tracked text moving support, like Google Docs or web version of
Microsoft Office). tdf#145233 tdf#145718 tdf#145719 tdf#145721 (L??szl?? N??
meth, NISZ)
Tracked list item moved by Move Down, with new tooltip caption and Manage
Changes Action icons
Other improvements in track changes
Related to changes of paragraph formatting etc. (L??szl?? N??meth, NISZ).
* Track change of paragraph style tdf#144272
* Fixed table formula calculation at track changes tdf#39828
* Track changes: fix table deletion at paragraph join tdf#144058
* Track changes: keep format of hints. When a text has multiple character
formatting settings enabled and a new one is applied with tracked changes,
only one of the former settings is restored upon rejecting the changes. tdf
#144173
* Track format changes of the actual word. If there is no text selection,
character formattings are applied on the word under the text cursor, now
with storing the original direct character formatting in a redline "extra
data" for rejection of the tracked formatting change. tdf#143939
* Show bullet and numbering changes at the moved list item, not at the next
one. tdf#145068
A list item moved by Move Down: show changes only in two items, not in four
items.
Performance
* Increased a PDF export speed of some complex documents tdf#125892 (Noel
Grandin, Collabora)
* Speed up loading large RTL documents tdf#64991 (Noel Grandin, Collabora)
* Improved rendering speed with some complex documents core commit 71429b93,
core commit d13b63a9 (Lubo? Lu????k, Collabora)
Calc
General improvements
* The "Link to External Data" dialogue (Sheet ? Link to External Data...) now
lists HTML tables in the order they appear in the source. tdf#142600
(Andreas Heinisch)
HTML table list in the "Link to External Data" dialogue. LibreOffice 7.3 (on
the right) now lists the tables in the order they appear in the source.
(LibreOffice 7.2 on the left.)
* Neumaier summation algorithm?Wikipedia logo v3.svg was implemented for
using modern SIMD CPU instructions like AVX2 tdf#142307 (Dante Dom??nech)
* Whitespace line feed and character tabulation in cell formula expressions
are now preserved and survive round-tripping between OOXML and ODF file
formats. Tabs however can not be entered in the cell input UI and are
converted to one space character. tdf#76310 (Eike Rathke, Red Hat)
* CSV import/export support a sep=; and "sep=;" field separator setting. When
reading CSV the separator is taken from an initial sep=; or "sep=;" single
field if that is the only row content. The quoted form is preserved as
(unquoted) cell content, of the unquoted form the separator is discarded as
contextually it is a real field separator. When writing CSV an existing
single top left cell's content sep=; if that is the only cell in the row is
adapted to the current separator in the quoted form "sep=;" (if quotes /
text delimiters aren't set empty ...) and always uses the ASCII " double
quote character. If the line containing the sep=; is not to be imported as
data (and thus also is not preserved when re-saving) then the From row
number in the dialog must be set to 2. tdf#117868 (Eike Rathke, Red Hat)
* A new Evaluate formulas option in the CSV Import and Paste Special and Text
to Column dialog determines whether formula expressions starting with a =
equal sign character are to be evaluated as formulas or imported as textual
data. In the programmatic FilterOptions string this is the 13th parameter
(token 12) with arguments either false or true. tdf#114878 (Eike Rathke,
Red Hat)
* Bash like autocompletion for Calc autoinput was implemented. tdf#145198
(Dennis Francis, Collabora)
Consider an example of the following data in a column:
ABCD123xyz
ABCD345qwel
ABCD123pqr
ABCD123xyz
PQR
When user types A, it will show the partial suggestion BCD.
User can accept the suggestion with the right arrow key and then the cursor
will be placed after the letter D, waiting for more input. User can choose
to not accept the suggestion either by typing more or by ending the edit
mode by pressing Esc key.
If the user accepts the suggestion BCD by right arrow key, and types 1, it
will show a partial suggestion of 23. User can accept this by pressing the
right arrow key.
If the user accepts the suggestion in the 3rd step and types x it will show
the final suggestion yz. Again user can choose to accept or decline the
suggestion as mentioned in the 1st step.
* Cell cursor now uses the system's highlight color instead of the default
font color to improve visibility tdf#142121 (Natalia Gavrilova)
Note that the cursor in cell C4 now using the highlight color instead of the
default font color
* Support Color Filter in "Standard Filter" dialog tdf#143103 (Samuel
Mehrbrodt, allotropia)
Standard filter supports filtering by colors
* Queries and filters using some text-based operations such as 'contains' now
properly work even with numeric data core commit 0d1971a8 (Lubo? Lu????k,
Collabora)
* Quick find will now search for values instead of formulas (the search
dialog offers both options). tdf#102506 (Julien Nabet)
New spreadsheet functions
* None
Changed spreadsheet functions
* Option "Search criteria = and <> must apply to whole cells" has been fixed
to apply properly in lookup functions when turned off (tdf#139612). For
better performance it is recommended to leave the option on unless partial
cell matching is required. (Lubo? Lu????k, Collabora)
* The IF() function now unconditionally propagates the condition's error
value, if any. Previously a given ELSE-path was executed as the condition
is never TRUE on error, which wasn't significant unless the ELSE was an
error evaluating function like IFERROR() or ISERROR() or other IS...()
functions.
This changes behaviour of an undocumented side effect, so might break
existing expressions that (unwillingly?) relied on it. tdf#146377 (Eike
Rathke, Red Hat)
Performance
* Improved speed of some XLSM files opening tdf#95549 (Noel Grandin,
Collabora)
* Speedup inserting large charts (for over 800000 cells) tdf#144052 (Noel
Grandin, Collabora)
* Improved speed of some XLSX files opening tdf#130795 (Noel Grandin,
Collabora)
* Improved speed of lookup functions in some cases tdf#139444,tdf#144249
(Lubo? Lu????k, Collabora)
* Improved speed for filtering by Autofilter tdf#133835, tdf#133867, tdf#
133996 (Noel Grandin, Lubo? Lu????k, Collabora)
* Improved use of threads for Calc calculations core commit a0e27322, core
commit 871d5fe3 (Lubo? Lu????k, Collabora)
Impress & Draw
General improvements
* Added PowerPoint-compatible screen sizes in Slide ? Slide Properties... ?
Slide ? Paper Format. "Widescreen" is used in current PowerPoint,
"On-screen show" is used in previous PowerPoint versions and Google Slides.
(see here) If you create a file with these sizes, the exported file will be
treated as a preset size instead of a user size. It will also be treated as
a preset size when you import it. tdf#142286, tdf#142288 (Jun Nogata)
* Unshare shape properties for the same type before insertion blog post
(Miklos Vajna, Collabora)
* tdf#140321 (Regina Henschel)
Wire Frame - Matt - Plastic - Metal
The dialog "3D-Settings" opens if you toggle a shape to "Extrusion on". The
dialog has a drop-down list "Surface" with items "Wire Frame", "Matt",
"Plastic" and "Metal". In the old version the items "Matt", "Plastic" and
"Metal" did nothing, but the solid was always rendered as "Plastic". Now
you get different surfaces. "Matt" is really matte. "Metal" is similar to
"Plastic" but has more contrast.
Unfortunately the old versions have written wrong and invalid values into
the file. Such values might result in unwanted rendering now. To repair
such shapes, mark the shape, then first switch to item "Matt" and then to
"Plastic" or "Metal" respectively.
Performance
Base
* FireBird data base engine bugfixes tdf#115547, tdf#117842, tdf#120129, tdf#
125286, tdf#144172, tdf#144230, tdf#144340 and many concomitant commits
(Mike Kaganski, Collabora)
Chart
Math
Core / General
* LibreOffice can now generate one-dimensional barcodes in addition to QR
codes. tdf#141193 (Akshit Kushwaha)
QR and Barcode dialog next to inserted barcode on page.
* new WebDAV/HTTP UCP based on libcurl. tdf#101094, tdf#102499 (Michael
Stahl, allotropia; Giuseppe Castagno)
+ using the OS native TLS stack on Windows and macOS, so users can now
manage trust of CAs with OS UI on these platforms
* Border line widths have been unified throughout LibreOffice. tdf#48622
(Samuel Mehrbrodt, allotropia)
+ New defaults are:
o Hairline (0.05pt)
o Very thin (0.5pt)
o Thin (0.75pt)
o Medium (1.5pt)
o Thick (2.25pt)
o Extra thick (4.5pt)
o Custom values are still possible
New line width defaults
Performance
* Optimizing LibreOffice for a larger number of users. Read more in a blog
post (Lubo? Lu????k, Collabora)
* Fixed a mistake that was making CJK text rendering slower over time core
commit 5b38b574 (Lubo? Lu????k, Collabora)
* Various rendering improvements in the Skia-based rendering backend (Lubo?
Lu????k, Collabora)
* Official binary packages are now built using Link-Time Optimization, which
should improve overall performance (Lubo? Lu????k, Collabora)
LibreOffice Help
* Help pages for the ScriptForge scripting library (J.P. Ledure, A.
Romedenne, R. Lima)
* Focus search results to user's current module tdf#123506 (Ross Johnson)
* Switch from Fuzzysort to FlexSearch for index search [commit] (I.
Lauhakangas)
* Close navigation menus when clicking outside them [commit] (I. Lauhakangas)
* Several improvements in Calc functions Help pages for accuracy and
completeness and links to Calc Function wiki pages. List of patches (S.
Fanning)
* Help updates and fixes
+ E. Rathke
+ O. Hallot
+ M. Kaganski
+ R. Lima
+ Ming Hua
+ A. Romedenne
+ R. Johnson
* Internal Help refactoring tdf#132643 (Johnny_M)
Filters
Improvements of DOC import/export filter
* DOC import: greatly improved list/numbering import tdf#106541 tdf#104239
tdf#123405 tdf#120629 (Justin Luth, SIL)
Improvements of DOCX import/export filter
* DOCX import: greatly improved list/numbering import tdf#141964 tdf#141966
(Justin Luth, SIL)
* Hyperlinks attached to shapes are now imported/exported tdf#123626 (Samuel
Mehrbrodt, allotropia)
* DOCX import: fix permission for editing tdf#89383 (T??nde T??th, NISZ)
* DOCX export/import: track change of paragraph style tdf#144272 (L??szl?? N
??meth, NISZ)
* DOCX import: fix user index and index entry support defined by field code \
f. This is a reworking of a previously patch. Now the import uses
com.sun.star.text.UserIndex and com.sun.star.text.UserIndexMark index and
index entry fields, which support the requested user index not only during
the DOCX and OpenDocument round-trip, but its run-time functionality, the
multiple user-defined indices tdf#77051 (L??szl?? N??meth, NISZ)
* DOCX import: handle ZOrder of chart objects tdf#144798 (Miklos Vajna,
Collabora)
* DOCX import/export: improved support for sdt fields tdf#81507, tdf#130565,
tdf#137466, tdf#100493 (Vasily Melenchuk, CIB/allotropia)
Improvements of XLSX import/export filter
* XLSX import: decreased a row height for XLSX files created by MS Office tdf
#144642 (Attila Sz?cs, NISZ)
* XLSX export: cell indent doesn't increase on each save now tdf#130104
(Kevin Suo)
* XLSX import: fix permission for editing tdf#115933 (T??nde T??th, NISZ)
* XLSX import/export: fix permission for editing. The password for editing
wasn't asked, also wasn't exported in XLSX documents. tdf#118938 (T??nde T
??th, NISZ). Now it's exported in Calc using the following steps, also
verified before editing:
+ In File->Save As, choose Excel 2007??365 (.xlsx) format;
+ enable checkbox "Save with password" and click Save;
+ in the dialog "Set password", click on "Options" and enable checkbox
"Open file read-only", and enter a password for editing (i.e. skip the
password for opening).
* Calc buttons with macros: better XLSX support blog post (Miklos Vajna,
Collabora)
* OOXML chart import/export: Chart labels from cell-range tdf#143942 (Dennis
Francis, Collabora)
* OOXML chart import/export: Chart date categories - improved support for
formatting, scaling, range-selection(min/max), increments core commit
f547cf17 (Dennis Francis, Collabora)
Improvements of PPTX import/export filter
* PPTX export: fixed interactions and hyperlinks on images tdf#124232 (Tibor
Nagy, NISZ)
* PPTX import: Impress doesn't drop the read-only password in PPTX file now
tdf#144943 (T??nde T??th, NISZ)
* PPTX import: fix hyperlinks on shapes tdf#144616 (Tibor Nagy, NISZ)
* PPTX import: fix hyperlinks on images tdf#141704 (Tibor Nagy, NISZ)
* Fix the incorrect import of slide footers from PPTX tdf#142221 (Sarper
Akdemir, Collabora)
* Make slide footers get exported to PPTX tdf#59323 (Sarper Akdemir,
Collabora)
* PPTX import: Transparent shadow for tables. Read more in a blog post
(Miklos Vajna, Collabora)
Improvements of WMF/EMF filter
Improvements of SVG filter
Other
* TXT file extension is no longer associated with T602 file format, which
used to lead to mis-detections in some cases (Jan Hole?ovsk?, Collabora)
GUI
General
* The wavy lines indicating spelling or grammar issues have been improved to
be more discernible in high-resolution screens. Now, they will match the
document??s zoom level instead of staying thin unconditionally. tdf#70519
(Heiko Tietze, TDF)
Icon theme
* Tweaks to Colibre icons related to graphics, saving, formatting and undo/
redo. (Rizal Muttaqin)
Dialogs
Sidebar
Notebookbar
Localization
If you speak an endangered language and want to help us to
Im-jabber.svg achieve our goals of providing libre tools to every person on the
planet, please help translate our software.
New languages/locales with locale data
Available as default document language and for locale specific formatting.
* English (Israel) {en-IL}. tdf#145173 (Eike Rathke, Red Hat)
Additional languages in the language list
Available for text attribution.
* Klingon, tlhIngan Hol {tlh}. tdf#143946 (Eike Rathke, Red Hat)
* Interslavic Latin {art-Latn-x-interslv} and Interslavic Cyrillic
{art-Cyrl-x-interslv}. tdf#145853 (Eike Rathke, Red Hat)
Improvements to proofing tools and language support
* Language entries in language listboxes created by spellchecker or
hyphenation dictionaries or extensions that support languages without a
predefined language listbox entry are now displayed using a possibly
translated language/locale name obtained from the ICU (International
Components for Unicode) library. Previously such entries were displayed as
language tags in curly brackets, for example {en-SG}. (Eike Rathke, Red
Hat)
* Old Hungarian has got full-text transliteration support: Tools ?
AutoCorrect ? Apply and Tools ? AutoCorrect ? Apply and Edit Changes
transliterate right-to-left Hungarian text to Old Hungarian (only if the
paragraph style of the text is Default Paragraph Style). tdf#143176 (L??szl
?? N??meth). More info in Release Notes 7.0
* Added Brazilian Portuguese (pt-BR) thesaurus (Olivier Hallot)
Scripting
The ScriptForge libraries
An extensible and robust collection of macro scripting resources for
LibreOffice to be invoked from user Basic or Python scripts. (Jean-Pierre
Ledure)
The libraries expose a total of 23 services each with a bunch of methods and
properties.
New in LibreOffice 7.3:
* The (new) Chart service, to define chart documents stored in Calc sheets.
Most parameters available in the user interface can be set by script.
* The (new) PopupMenu service, to describe the menu to be displayed after a
mouse event (typically a right-click). The menu may contain usual items,
checkboxes or radio buttons. Each item can be decorated with icons and
tooltips.
* Management of printers: list of fonts and printers, printer options,
printing documents.
* Export of documents to PDF and management of PDF options.
* Dialogs may host table controls. Provide tabular data, the data is
displayed with headers and sort buttons. The selected row is returned by
the Value property.
* Titles, tooltips, list- or comboboxes, labels of dialogs can be translated
on-the-fly based on tools provided in the L10N service.
* The Calc service has been improved with facilities to identify cell ranges
when their size is unknown, f.i. after the import of data, and to apply
formulas with absolute and relative references on them.
* The whole set of services is made available for Python scripts with
identical syntax and behaviour as in Basic.
* The English version of the documentation of the ScriptForge libraries (7.3)
is fully integrated in the LibreOffice help pages (https://
help.libreoffice.org/7.3/en-US/text/sbasic/shared/03/lib_ScriptForge.html?
DbPAR=BASIC). Their translation into other languages is underway. (Alain
Romedenne, Rafael Lima)
BASIC
Python
VBA changes
Feature Removal / Deprecation
Platform Compatibility
Mac
* LibreOffice 7.3 requires macOS 10.12 or newer to run.
Due to Apple's policy regarding C++ standard library features, the minimum
macOS version requirement will be raised in the future like so:
+ release in Aug 2022 ?? 10.13 High Sierra
+ release in Feb 2023 ?? 10.14 Mojave
+ release in Aug 2023 ?? 10.15 Catalina
* A rendering backend based on the Skia library is now used also on macOS [1]
(Lubo? Lu????k, Collabora)
Windows
* The IAccessibleTable2 and IAccessibleTableCell interfaces from the
IAccessible2 specification have been implemented (core commit 839dbf9e,
core commit 97a88e30) (Michael Weghorn)
Linux
API Changes
* The C functions rtl_str_toFloat, rtl_str_toDouble, rtl_ustr_toFloat, and
rtl_ustr_toDouble, and the C++ member functions rtl::OString::toFloat,
rtl::OString::toDouble, rtl::OUString::toFloat, and rtl::OUString::toDouble
no longer accept a ??+?? or ??-?? sign before ??NaN?? when parsing the XML
Schema??style Not a Number notation. (But they still accept an optional ??+
?? or ??-?? sign before the legacy ??1.#NAN?? notation.) core commit
9d27a44e
* osl_demultiplexSocketEvents and its support functionality, all of which
were deprecated, have been removed from the UNO C/C++ API. core commit
db392a8a
UNO API changes
* com.sun.star.resource.XLocale has been removed. core commit a74d15ad
* In com.sun.star.i18n, various enum members have been renamed (to no longer
use reserved identifiers): core commit 04af4e4f
+ in com.sun.star.i18n.TransliterationModules:
o from ignoreTraditionalKanji_ja_JP to IgnoreTraditionalKanji_ja_JP
o from ignoreTraditionalKana_ja_JP to IgnoreTraditionalKana_ja_JP
o from ignoreMinusSign_ja_JP to IgnoreMinusSign_ja_JP
o from ignoreIterationMark_ja_JP to IgnoreIterationMark_ja_JP
o from ignoreSeparator_ja_JP to IgnoreSeparator_ja_JP
o from ignoreZiZu_ja_JP to IgnoreZiZu_ja_JP
o from ignoreBaFa_ja_JP to IgnoreBaFa_ja_JP
o from ignoreTiJi_ja_JP to IgnoreTiJi_ja_JP
o from ignoreHyuByu_ja_JP to IgnoreHyuByu_ja_JP
o from ignoreSeZe_ja_JP to IgnoreSeZe_ja_JP
o from ignoreIandEfollowedByYa_ja_JP to IgnoreIandEfollowedByYa_ja_JP
o from ignoreKiKuFollowedBySa_ja_JP to IgnoreKiKuFollowedBySa_ja_JP
o from ignoreSize_ja_JP to IgnoreSize_ja_JP
o from ignoreProlongedSoundMark_ja_JP to
IgnoreProlongedSoundMark_ja_JP
o from ignoreMiddleDot_ja_JP to IgnoreMiddleDot_ja_JP
o from ignoreSpace_ja_JP to IgnoreSpace_ja_JP
o from smallToLarge_ja_JP to SmallToLarge_ja_JP
o from largeToSmall_ja_JP to LargeToSmall_ja_JP
+ in com.sun.star.i18n.TransliterationModulesNew:
o from ignoreTraditionalKanji_ja_JP to IgnoreTraditionalKanji_ja_JP
o from ignoreTraditionalKana_ja_JP to IgnoreTraditionalKana_ja_JP
o from ignoreMinusSign_ja_JP to IgnoreMinusSign_ja_JP
o from ignoreIterationMark_ja_JP to IgnoreIterationMark_ja_JP
o from ignoreSeparator_ja_JP to IgnoreSeparator_ja_JP
o from ignoreZiZu_ja_JP to IgnoreZiZu_ja_JP
o from ignoreBaFa_ja_JP to IgnoreBaFa_ja_JP
o from ignoreTiJi_ja_JP to IgnoreTiJi_ja_JP
o from ignoreHyuByu_ja_JP to IgnoreHyuByu_ja_JP
o from ignoreSeZe_ja_JP to IgnoreSeZe_ja_JP
o from ignoreIandEfollowedByYa_ja_JP to IgnoreIandEfollowedByYa_ja_JP
o from ignoreKiKuFollowedBySa_ja_JP to IgnoreKiKuFollowedBySa_ja_JP
o from ignoreSize_ja_JP to IgnoreSize_ja_JP
o from ignoreProlongedSoundMark_ja_JP to
IgnoreProlongedSoundMark_ja_JP
o from ignoreMiddleDot_ja_JP to IgnoreMiddleDot_ja_JP
o from ignoreSpace_ja_JP to IgnoreSpace_ja_JP
o from smallToLarge_ja_JP to SmallToLarge_ja_JP
o from largeToSmall_ja_JP to LargeToSmall_ja_JP
+ in com.sun.star.i18n.UnicodeScript:
o from k_CJKUnifiedIdeographsExtensionA to
kCJKUnifiedIdeographsExtensionA
ZXing-C++ ("zebra crossing") is an open-source, multi-format 1D/2D
barcode image processing library implemented in C++.
It was originally ported from the Java ZXing Library but has been
developed further and now includes many improvements in terms of
quality and performance. It can both read and write barcodes in a
number of formats.
Changelog:
0.17.0:
C++ API
set the baseline C++ version to 17.
mdds has been internalized so that the public header no longer contains references to mdds. With this change, the users can use different API versions of mdds between the ixion build and run-time use.
cleaned up public API to make use of std::string_view and std::variant where appropriate.
formula interpreter
implemented built-in LEFT() function.
misc
it is no longer required to set the size of void* at build time to ensure the binaries to be fully functional.
fixed a bug where named expressions with names containing invalid characters were still allowed in.
Changelog:
orcus 0.17.2
* ods
* fixed a bug where the state of style:cell-protect="none" was not
explicitly pushed, thereby having had the same effect as not having this
attribute. After the fix, style:cell-protect="none" will explicitly push
the hidden state to false, locked state to false, and the formula-hidden
state to false.
orcus 0.17.1
* general
* addressed a number of coverity issues.
* removed a variety of compiler warnings.
* ods
* re-generated sax parser tokens from ODF v1.3.
* revised the style import code to only push style attributes that are
actually specified in the XML.
* xls-xml
* revised the XML structure validation strategy to ignore any mis-placed
elements and their sub structures rather than aborting the import.
orcus 0.17.0
* general
* set the baseline C++ version to 17.
* cleaned up the public API to replace pstring with std::string_view, union
with std::variant, and boost::optional with std::optional. With this
change, the public API no longer has dependency on boost.
* spreadsheet document
* switched to using ixion::model_iterator for horizontal iteration of cells
instead of using mdds::mtv::collection.
* fixed a bug where exporting a spreadsheet document containing adjacent
merged cells regions to html incorrectly exported the merged cell areas.
* xlsx
* cached cell values are now correctly loaded from the file.
* sax parser
* utf-8 names are now allowed as element and attribute names.
* css parser
* unquoted utf-8 property values are now allowed.
* orcus-json
* fixed segmentation fault when using --mode structure with the Windows
build.
* added yaml output option.
* xml-map
* fixed a bug where mapping of an XML document with namespace aliases
sometimes corrupts the alias values.
* python
* added orcus.FormulaTokenOp enum type which describes type formula token
operator types in a more finer grained manner.
* documentation
* added notes to how to use orcus-xml and orcus-json to map XML and JSON
documents to spreadsheet documents.
Changelog:
mdds 2.0.1 (not released yet)
* general
* addressed various coverity issues.
* multi_type_vector
* fixed random compiler warnings.
* fixed event handling in copy construction. In aos, the event object was
not copied when the parent container was copied. In soa, the
element_block_acquired() callback was not called for the cloned element
blocks.
* added move constructors and move assignment operators to both aos and soa
variants.
mdds 2.0.0
* general
* set the baseline C++ version to C++17.
* multi_type_vector
* implemented structure-of-arrays (SoA) storage as its default storage
layout for better CPU cache efficiency.
* added multiple block position adjustment implementations with various
loop-unrolling factors combined with SSE2 and AVX2 features.
* added a tool called runtime-env to benchmark different block position
adjustment implementations to determine the optimal loop-unrolling factor.
* rectangle_set
* permanently removed.
* rtree
* fixed a bug where the memory positions of invalidated child nodes were not
properly updated after tree mutation. The problem manifested itself when
using libc++ as stdlib with clang.
mdds 1.7.0
* trie_map
* added copy and move constructors.
* added a variant of find() method that returns a mutable iterator object.
The user can now update the value associated with a key directly via the
iterator object.
* packed_trie_map
* added copy and move constructors.
* added load_state() and save_state() methods to allow loading state from
and saving state to binary files.
* This is kernel module and disbale MKPIE support.
Changelog:
HAXM v7.7.0
Change Log
* Added a new IOCTL to enable getting CPUID features for guest VCPUs (#383).
* Enabled all supported CPUID leaves to be configurable (#382).
* Enabled several features in CPUID emulation (#381).
* Migrated the CI service from Travis CI to GitHub Actions (#353).
HAXM v7.6.6
Change Log
* Optimized the CPUID module and added support for setting two new CPUID
leaves (#335).
* Fixed some vulnerability issues of loading DRs and MSRs (#347).
* Fixed some minor issues from static code scan (#351).
* Fixed the download URL in the homepage (#348).
HAXM v7.6.5
Change Log
* Optimized internal storage structure for CPUID feature set (#315).
* Fixed a performance issue caused by a regression (#312).
* Fixed some warnings from static code scan (#313).
* Introduced a new installer framework for Windows.
HAXM v7.6.1
Change Log
* Added a new IOCTL to enable setting CPUID feature for guest VCPUs (#277, #
281, #282).
* Enabled PAT as HAXM supported CPUID feature and added IA32_CR_PAT VMX
handling (#204).
* Changed to return deterministic cache parameters by host cache values (#204
).
* Cleaned up the legacy EPT engine (#261).
The SWTPM package provides TPM emulators with different front-end interfaces
to libtpms. TPM emulators provide socket interfaces (TCP/IP and Unix) and
the Linux CUSE interface for the creation of multiple native /dev/vtpm* devices.
The SWTPM package also provides several tools for using the TPM emulator,
creating certificates for a TPM, and simulating the manufacturing of
a TPM by creating a TPM's EK and platform certificates etc.
Libtpms is a library that targets the integration of TPM functionality
into hypervisors, primarily into Qemu. Libtpms provides a very narrow
public API for this purpose so that integration is possible. Only the
minimum of necessary APIs are made publicly available.
It is assumed that the user of libtpms is familiar with the concepts
of the Trusted Platform Module (TPM). For the interaction with libtpms
it is necessary to know how to construct valid TPM commands and to
be able to parse their results. It is not within the scope of libtpms's
documentation to provide background on this
If audio/pulseaudio is already installed, the build fails by pulse/pulseaudio.h not found error.
The pulseaudio option is disabled by default.
Bump PKGREVISION by dependency change.
* Use GitHub release tarball as distfile.
Changelog:
Overview of changes leading to 3.3.1
Monday, January 31, 2022
====================================
- Fix heap-use-after-free in harfbuzz-subset introduced in previous release.
(Garret Rieger)
Overview of changes leading to 3.3.0
Monday, January 31, 2022
====================================
- Improved documentation. (Matthias Clasen)
- Internal code cleanup, using C++ standard library more. (Behdad Esfahbod)
- The low 16-bits of face index will be used by hb_face_create() to select a
face inside a font collection file format, while the high 16-bits will be
used by hb_font_create() to load the named instance. (Behdad Esfahbod)
- Glyph positions and other font metrics now apply synthetic slant set by
hb_font_set_synthetic_slant(), for improved positioning for synthetically
slanted fonts. (Behdad Esfahbod)
- Fixed unintentional locale dependency in hb_variation_to_string() for decimal
point representation. (Matthias Clasen)
- When applying pair positioning (kerning) the positioning value is split
between the two sides of the pair for improved cursor positioning between
such pairs. (Behdad Esfahbod)
- Introduced new HB_GLYPH_FLAG_UNSAFE_TO_CONCAT, to be used in conjunction
with HB_GLYPH_FLAG_UNSAFE_TO_BREAK for optimizing re-shaping during line
breaking. Check the documentation for further details. (Behdad Esfahbod)
- Improved handling of macrolanguages when mapping BCP 47 codes to OpenType
tags. (David Corbett)
- New API:
+HB_GLYPH_FLAG_UNSAFE_TO_CONCAT
+hb_segment_properties_overlay()
+hb_buffer_create_similar()
+hb_font_set_synthetic_slant()
+hb_font_get_synthetic_slant()
+hb_font_get_var_coords_design()
Overview of changes leading to 3.2.0
Friday, November 26, 2021
====================================
"harfbuzz" library improvements:
- Fixed shaping of Apple Color Emoji flags in right-to-left context. (Behdad Esfahbod)
- Fixed positioning of CFF fonts in HB_TINY profile. (Behdad Esfahbod)
- OpenType 1.9 language tags update. (David Corbett)
- Add HB_NO_VERTICAL config option.
- Add HB_CONFIG_OVERRIDE_H for easier configuration. (Behdad Esfahbod)
"harfbuzz-subset" library improvements:
- Improved packing of cmap, loca, and Ligature tables. (Garret Rieger)
- Significantly improved overflow-resolution strategy in the repacker. (Garret Rieger)
Changelog:
0.2.4
* Added is_direct flag when creating DMs (thanks to @gsauthof in #261).
* Added newline toggle for swapping enter and alt-enter behavior (thanks to
@octeep in #270).
* Added timestamps toggle for disabling timestamps in the UI (thanks to @lxea
in #304).
* Added support for getting custom download directory with xdg-user-dir.
* Added support for updating homeserver URL based on well-known data in /
login response.
* Updated some places to use default color instead of white to better work on
light themed terminals (thanks to @zavok in #280).
* Updated notification library to work on all unix-like systems with
notify-send.
+ Notification sounds will now work if either paplay or ogg123 is
available.
+ Based on work by @negatethis (in #298) and @begss (in #312).
* Disabled logging request content for sensitive requests like /login and
cross-signing key uploads.
* Fixed caching state of rooms where the room ID contains slashes.
* Fixed index error in fuzzy search (thanks to @Evidlo in #268).
The default tar extractor silently skipped the missing file
./src/lxml/isoschematron/resources/xsl/iso-schematron-xslt1/iso_schematron_skeleton_for_xslt1.xsl
and after switching to gtar as extractor the package builds correctly.
6.8.0
Enhancements made
- Add support for the debug modules request
Bugs fixed
- Handle all threads stopped correctly
- Fix the debug modules model
- Handled AllThreadsContinued and workaround for wrong threadId in cont…
Maintenance and upkeep improvements
- Cancel duplicate runs
6.7.0
Enhancements made
- Add usage_request and usage_reply based on psutil
Bugs fixed
- Removed DebugStdLib from arguments of attach
- Normalize debugger temp file paths on Windows
- Breakpoint in cell with leading empty lines may be ignored
Maintenance and upkeep improvements
- Skip on PyPy, seem to fail.
- Remove pipx to fix conflicts
- Remove impossible skipif.
6.6.1
Bugs fixed
- PR: do_one_iteration is a coroutine
Maintenance and upkeep improvements
- Clean python 2 artifacts.
Documentation improvements
- Fix title position in changelog
6.6.0
Enhancements made
- Set `debugOptions` for breakpoints in python standard library source
- Send `omit_sections` to IPython to choose which sections of documentation you do not want
Bugs fixed
- Added missing `exceptionPaths` field to `debugInfo` reply
Maintenance and upkeep improvements
- Test `jupyter_kernel_test` as downstream
- Remove `nose` dependency
- Add explicit encoding to open calls in debugger
6.5.1
Bugs fixed
- Fix the temp file name created by the debugger
Maintenance and upkeep improvements
- Enforce labels on PRs
- Unpin `IPython`, and remove some dependencies on it.
6.5.0
Bugs fixed
- Fix rich variables inspection
- Do not call `setQuitOnLastWindowClosed()` on a `QCoreApplication`
Maintenance and upkeep improvements
- Drop `ipython_genutils` requirement
6.4.2
Enhancements made
- Enabled rich rendering of variables in the debugger
Bugs fixed
- Remove setting of the eventloop function in the InProcessKernel
Maintenance and upkeep improvements
- Add python version classifiers
6.4.1
Merged PRs
- debugpy is now a build requirement
6.4.0
Enhancements made
- Make `json_clean` a no-op for `jupyter-client` >= 7
Bugs fixed
- Do not assume kernels have loops
- Fix undefined variable
Maintenance and upkeep improvements
- Make `ipykernel` work without `debugpy`
- Stop using deprecated `recv_multipart` when using in-process socket.
- Update some warnings with instructions and version number.
6.3.1
Merged PRs
- Add dependency on IPython genutils.
6.3.0
Enhancements made
- Add deep variable inspection
- Add `IPKernelApp.capture_fd_output` config to disable FD-level capture
Maintenance and upkeep improvements
- Remove more `nose` test references
- Remove `nose` `skipIf` in favor of `pytest`
- Remove more `nose`
- Set up release helper plumbing
- Test downstream projects
6.2.0
Enhancements made
- Add Support for Message Based Interrupt
Maintenance and upkeep improvements
- Remove some more dependency on nose/iptest
- Remove block param from get_msg()
6.1.0
Enhancements made
- Implemented `richInspectVariable` request handler
Maintenance and upkeep improvements
- Bump `importlib-metadata` limit for `python<3.8`
Bug Fixes
- Fix exception raised by `OutStream.write`
6.0.3
- `KernelApp`: rename ports variable to avoid override
6.0.2
Bugs fixed
- Add watchfd keyword to InProcessKernel OutStream initialization
- Fix typo in eventloops.py
- [bugfix] fix in setup.py (comma before appnope)
Maintenance and upkeep improvements
- Add upper bound to dependency versions.
- Replace non-existing function.
- Remove unused variables
- Do not use bare except
- misc whitespace and line too long
- Formatting: remove semicolon
- Clean most flake8 unused import warnings.
- Minimal flake8 config
- Remove CachingCompiler's filename_mapper
6.0.1
- Fix Tk and asyncio event loops
- Stringify variables that are not json serializable in inspectVariable
6.0.0
IPykernel 6.0 is the first major release in about two years, that brings a number of improvements, code cleanup, and new
features to IPython.
You should be able to view all closed issues and merged Pull Request for this
milestone [on
GitHub](https://github.com/ipython/ipykernel/issues?q=milestone%3A6.0+is%3Aclosed+),
as for any major releases, we advise greater care when updating that for minor
release and welcome any feedback (~50 Pull-requests).
IPykernel 6 should contain all changes of the 5.x series, in addition to the
following non-exhaustive changes.
- Support for the debugger protocol, when using `JupyterLab`, `RetroLab` or any
frontend supporting the debugger protocol you should have access to the
debugger functionalities.
- The control channel on IPykernel 6.0 is run in a separate thread, this may
change the order in which messages are processed, though this change was necessary
to accommodate the debugger.
- We now have a new dependency: `matplotlib-inline`, this helps to separate the
circular dependency between IPython/IPykernel and matplotlib.
- All outputs to stdout/stderr should now be captured, including subprocesses
and output of compiled libraries (blas, lapack....). In notebook
server, some outputs that would previously go to the notebooks logs will now
both head to notebook logs and in notebooks outputs. In terminal frontend
like Jupyter Console, Emacs or other, this may ends up as duplicated outputs.
- coroutines are now native (async-def) , instead of using tornado's
`@gen.coroutine`
- OutStreams can now be configured to report `istty() == True`, while this
should make some output nicer (for example colored), it is likely to break
others. Use with care.
New features added
- Implementation of the debugger
Enhancements made
- Make the `isatty` method of `OutStream` return `true`
- Allow setting cell name
- Try to capture all file descriptor output and err
- Implemented `inspectVariables` request
- Specify `ipykernel` in kernelspec
- Use `matplotlib-inline`
- Run control channel in separate thread
Bugs fixed
- Remove references to deprecated `ipyparallel`
- Return len of item written to `OutStream`
- Call metadata methods on abort replies
- Fix keyboard interrupt issue in `dispatch_shell`
- Update `Trio` mode for compatibility with `Trio >= 0.18.0`
- Follow up `DeprecationWarning` Fix
- Flush control stream upon shutdown
- Fix Handling of `shell.should_run_async`
- Deacrease lag time for eventloop
- Fix "Socket operation on nonsocket" in downstream `nbclient` test.
- Stop control thread before closing sockets on it
- Fix debugging with native coroutines
- Fixup master build
- Fix parent header retrieval
- Add missing self
- Backwards compat with older versions of zmq
Maintenance and upkeep improvements
- Remove pin on Jedi because that was already fixed in IPython
- Remove deprecated source parameter since 4.0.1 (2015)
- Remove deprecated `SocketABC` since 4.5.0
- Remove deprecated profile options of `connect.py`
- Remove `ipykernel.codeutil` deprecated since IPykernel 4.3.1 (Feb 2016)
- Keep preferring `SelectorEventLoop` on Windows
- Add `Kernel.get_parent` to match `set_parent`
- Flush control queue prior to handling shell messages
- Add `Kernel.get_parent_header`
- Build docs only on Ubuntu: add jobs to check docstring formatting.
- Make deprecated `shell_streams` writable
- Use channel `get_msg` helper method
- Use native coroutines instead of tornado coroutines
- Make less use of `ipython_genutils`
- Run GitHub Actions on all branches
- Move Python-specific bits to ipkernel
- Update Python Requirement to 3.7
- Replace import item from `ipython_genutils` to traitlets.
- Some removal of `ipython_genutils.py3compat`.
- Fixup `get_parent_header` call
- Update of `ZMQInteractiveshell`.
- Removed filtering of stack frames for testing
- Added 'type' field to variables returned by `inspectVariables` request
- Changed default timeout to 0.0 seconds for `stop_on_error_timeout`
- Attempt longer timeout
- Clean up release process and add tests
- Kernelspec: ensure path is writable before writing `kernel.json`.
- Add `configure_inline_support` and call it in the shell
Documentation improvements
- Misc Updates to changelog for 6.0
- Add 5.5.x Changelog entries
- Build docs only on ubuntu: add jobs to check docstring formatting.
- DOC: Autoreformat all docstrings.
- Bump Python to 3.8 in `readthedocs.yml`
- Fix typo
- Add release note to 5.5.0 about `stop_on_error_timeout`
- Move changelog to standard location
- Add changelog for 5.5
- Change to markdown for changelog
Deprecations in 6.0
- `Kernel`s now support only a single shell stream, multiple streams will now be ignored. The attribute
`Kernel.shell_streams` (plural) is deprecated in ipykernel 6.0. Use `Kernel.shell_stream` (singular)
- `Kernel._parent_header` is deprecated, even though it was private. Use `.get_parent()` now.
Removal in 6.0
- `ipykernel.codeutils` was deprecated since 4.x series (2016) and has been removed, please import similar
functionalities from `ipyparallel`
- remove `find_connection_file` and `profile` argument of `connect_qtconsole` and `get_connection_info`, deprecated since IPykernel 4.2.2 (2016).
IPython Parallel (ipyparallel) is a Python package and collection of CLI
scripts for controlling clusters of IPython processes, built on the Jupyter
protocol.
