Changelog
### 2.0.0
```
-------------------
- Updated documentation.
- Move to virtualenv/pip based development.
- Move to Read The Docs for documentation.
- Use Travis CI for testing and releases.
- Use features of newer :mod:`testfixtures` in
:class:`~xlutils.view.CheckerView`.
- Python 3 compatibility.
```
### 1.7.1
```
---------------------
- Add support for time cells in :class:`~xlutils.view.View`.
- Add support for ``.xlsx`` files in :class:`~xlutils.view.View` at the
expense of formatting information being available.
```
### 1.7.0
```
-----------------------
- Added the :mod:`xlutils.view` module.
```
### 1.6.0
```
--------------------
- Moved documentation to be Sphinx based.
- Support for :mod:`xlrd` 0.9.1, which no longer has pickleable
books.
.. note:: You may encounter performance problems if you work with
large spreadsheets and use :mod:`xlutils` 1.6.0 with
:mod:`xlrd` versions earlier that 0.9.1.
```
### 1.5.2
```
---------------------
- When using :mod:`xlutils.copy`, the ``datemode`` is now copied across from the
source solving date problems with certain files.
- The :mod:`errorhandler` package is no longer a hard dependency.
- As a minimum, :mod:`xlrd` 0.7.2 and :mod:`xlwt` 0.7.4 are now required.
```
### 1.5.1
```
--------------------
- Fix packaging problem caused by the move to git
```
### 1.5.0
```
--------------------
- Take advantage of "ragged rows" optimisation in xlrd 0.7.3
- Add support for PANE records to :mod:`xlutils.copy`, which means that zoom
factors are now copied.
```
### 1.4.1
```
------------------------
- Removal of references in the `finish` methods of several filters,
easing memory usage in large filtering runs
- Speed optimisations for :class:`~xlutils.filter.BaseFilter`, bringing those
benefits to all subclasses.
- Memory usage reduction when using :class:`~xlutils.filter.MemoryLogger`
```
### 1.4.0
```
----------------------
- Add sheet density information and onesheet option to
:mod:`xlutils.margins`.
- Reduced the memory footprint of :class:`~xlutils.filter.ColumnTrimmer` at the
expense of speed.
- Fixed incorrect warnings about boolean cells in
:class:`~xlutils.filter.ErrorFilter`. :mod:`xlwt` has always supported boolean
cells.
- :class:`~xlutils.filter.BaseReader` now opens workbooks with ``on_demand = True``
- Added support for :mod:`xlrd` Books opened with ``on_demand`` as ``True`` passed to
:func:`~xlrd.open_workbook`.
- Fixed bug when copying error cells.
- Requires the latest versions of :mod:`xlrd` (0.7.1) and :mod:`xlwt` (0.7.2).
```
### 1.3.2
```
--------------------
- Made installation work when `setuptools` isn't present.
- Made `errorhandler` an optional dependency.
```
### 1.3.1
```
-------------------
- In xlutils.styles, handle there case where two names were mapped to the
same xfi, but the first one was empty.
```
### 1.3.0
```
-------------------
- fix bug that cause BaseWriter to raise exceptions when handling
source workbooks opened by xlrd 0.7.0 and above where
formatting_info had been passed as False
- add xlutils.copy
```
### 1.2.1
```
-------------------
- add extremely limited formatting_info support to DummyBook and TestReader
- move to testfixtures 1.5.3 for tests
```
### 1.2.0
```
-------------------
- add and implement `start` method to components in xlutils.filter.
- fixed bug when using set_rdsheet with ColumnTrimmer.
- improved installation documentation.
- renamed xlutils.styles.CellStyle to more appropriate
xlutils.styles.NamedStyle.
- improved documentation for xlutils.styles.
- moved to using TestFixtures and Mock for tests.
- moved to using ErrorHandler rather than duplicating code.
```
### 1.1.1
```
-------------------
- prevented generation of excessively long sheet names that cause
Excel to complain.
- added test that will fail if the filesystem used doesn't support
filenames with +'s in them.
```
### 1.1.0
```
-------------------
- link to the documentation for xlutils.display
- tighten up version requirements for xlrd and xlwt
- use style compression in xlutils.filter.BaseWriter
- prevent generation of bogus sheet names in xlutils.filter.BaseWriter
- xlutils.filter.BaseFilter now keeps track of rdbook, simplifying the
implementation of filters.
- add another example for xlutils.filter
- add xlutils.filter.XLRDReader
- add xlutils.filter.StreamWriter
- add xlutils.styles
- add xlutils.save
```
### 1.0.0
```
------------------
- initial public release
```
Links
- PyPI: https://pypi.org/project/xlutils
- Changelog: https://pyup.io/changelogs/xlutils/
- Homepage: http://www.python-excel.org
This PR pins xlutils to the latest release 2.0.0.
Changelog
### 2.0.0 ``` ------------------- - Updated documentation. - Move to virtualenv/pip based development. - Move to Read The Docs for documentation. - Use Travis CI for testing and releases. - Use features of newer :mod:`testfixtures` in :class:`~xlutils.view.CheckerView`. - Python 3 compatibility. ``` ### 1.7.1 ``` --------------------- - Add support for time cells in :class:`~xlutils.view.View`. - Add support for ``.xlsx`` files in :class:`~xlutils.view.View` at the expense of formatting information being available. ``` ### 1.7.0 ``` ----------------------- - Added the :mod:`xlutils.view` module. ``` ### 1.6.0 ``` -------------------- - Moved documentation to be Sphinx based. - Support for :mod:`xlrd` 0.9.1, which no longer has pickleable books. .. note:: You may encounter performance problems if you work with large spreadsheets and use :mod:`xlutils` 1.6.0 with :mod:`xlrd` versions earlier that 0.9.1. ``` ### 1.5.2 ``` --------------------- - When using :mod:`xlutils.copy`, the ``datemode`` is now copied across from the source solving date problems with certain files. - The :mod:`errorhandler` package is no longer a hard dependency. - As a minimum, :mod:`xlrd` 0.7.2 and :mod:`xlwt` 0.7.4 are now required. ``` ### 1.5.1 ``` -------------------- - Fix packaging problem caused by the move to git ``` ### 1.5.0 ``` -------------------- - Take advantage of "ragged rows" optimisation in xlrd 0.7.3 - Add support for PANE records to :mod:`xlutils.copy`, which means that zoom factors are now copied. ``` ### 1.4.1 ``` ------------------------ - Removal of references in the `finish` methods of several filters, easing memory usage in large filtering runs - Speed optimisations for :class:`~xlutils.filter.BaseFilter`, bringing those benefits to all subclasses. - Memory usage reduction when using :class:`~xlutils.filter.MemoryLogger` ``` ### 1.4.0 ``` ---------------------- - Add sheet density information and onesheet option to :mod:`xlutils.margins`. - Reduced the memory footprint of :class:`~xlutils.filter.ColumnTrimmer` at the expense of speed. - Fixed incorrect warnings about boolean cells in :class:`~xlutils.filter.ErrorFilter`. :mod:`xlwt` has always supported boolean cells. - :class:`~xlutils.filter.BaseReader` now opens workbooks with ``on_demand = True`` - Added support for :mod:`xlrd` Books opened with ``on_demand`` as ``True`` passed to :func:`~xlrd.open_workbook`. - Fixed bug when copying error cells. - Requires the latest versions of :mod:`xlrd` (0.7.1) and :mod:`xlwt` (0.7.2). ``` ### 1.3.2 ``` -------------------- - Made installation work when `setuptools` isn't present. - Made `errorhandler` an optional dependency. ``` ### 1.3.1 ``` ------------------- - In xlutils.styles, handle there case where two names were mapped to the same xfi, but the first one was empty. ``` ### 1.3.0 ``` ------------------- - fix bug that cause BaseWriter to raise exceptions when handling source workbooks opened by xlrd 0.7.0 and above where formatting_info had been passed as False - add xlutils.copy ``` ### 1.2.1 ``` ------------------- - add extremely limited formatting_info support to DummyBook and TestReader - move to testfixtures 1.5.3 for tests ``` ### 1.2.0 ``` ------------------- - add and implement `start` method to components in xlutils.filter. - fixed bug when using set_rdsheet with ColumnTrimmer. - improved installation documentation. - renamed xlutils.styles.CellStyle to more appropriate xlutils.styles.NamedStyle. - improved documentation for xlutils.styles. - moved to using TestFixtures and Mock for tests. - moved to using ErrorHandler rather than duplicating code. ``` ### 1.1.1 ``` ------------------- - prevented generation of excessively long sheet names that cause Excel to complain. - added test that will fail if the filesystem used doesn't support filenames with +'s in them. ``` ### 1.1.0 ``` ------------------- - link to the documentation for xlutils.display - tighten up version requirements for xlrd and xlwt - use style compression in xlutils.filter.BaseWriter - prevent generation of bogus sheet names in xlutils.filter.BaseWriter - xlutils.filter.BaseFilter now keeps track of rdbook, simplifying the implementation of filters. - add another example for xlutils.filter - add xlutils.filter.XLRDReader - add xlutils.filter.StreamWriter - add xlutils.styles - add xlutils.save ``` ### 1.0.0 ``` ------------------ - initial public release ```Links
- PyPI: https://pypi.org/project/xlutils - Changelog: https://pyup.io/changelogs/xlutils/ - Homepage: http://www.python-excel.org