v0.4.4

Changed

• NeuroCurator is now installable simply with pip3 install neurocurator on macOS, Linux and Windows.
• Simplify even more the 'Getting Started' section of the README.
• Add a Roadmap section to the README to make visible the high level activities planned.
• Add instructions to upgrade NeuroCurator to the latest version.

Technical

• Port NeuroCurator to Qt 5 which removes legacy dependencies (Qt 4, PySide, Python 3.4).
• Normalize calls to super() to the Python 3 style.
• Use Qt object names directly, without the namespace name as prefix, except for the namespace 'Qt'.

Fixed

• #6
• #12
• #56

v0.4.3

Changed

• Make the 'Getting Started' steps on the README clearer.
• Make setup.py relies on a minimum version of NAT and no more on an exact one.
• Generalize the call to the default PDF Reader of the user operating system.

Fixed

• #76
• #26

v0.4.2

Release related to BlueBrain/nat#14.

Added

• Update README with specific instructions to install NeuroCurator on Windows 10

Fixed

• #70

v0.4.1

Added

• After installation, NeuroCurator can be launched with '$ neurocurator'.
• Enforce at installation that the version of NAT is the same.
• New supported installation procedure based on compiled PySide and Qt from conda-forge.
• New installation procedure makes easier to install on Windows.
• Set a specific required Python version in setup.py (3.4.x).

Changed

• Switch from using executables to installing with pip.
• Switch from distutils to setuptools for setup.py.
• Make __main__.py shebang portable.
• Switch from hackish main.py to canonical __main__.py.
• Upgrade setup.py. Delete MANIFEST.in and setup.cfg.
• Change README.md to use the releases on PyPI.

Removed

• notebooks/, see #43.

0.4.0

Added

UI

• #20 Filter dynamically the reference list through a free text input.
• Disabling of the 'References' tab and the 'Zotero' menu during the asynchronous refresh of the Zotero data. This guarantees a safe refresh. The other tabs are usable during the refresh.
• The table of creators in reference creation/edition forms has the new column 'creatorType'. The values are from Zotero: 'author', 'editor', ... Books can now have proper creators.
• Selection of the reference - and scrolling to it - in the reference list after creation.
• Indicate the start and the end of the loading of cached/distant Zotero data.
• When a reference is not a 'journalArticle', the column 'Journal' contains the reference type in brackets.
• Default selection of the type 'journalArticle' for reference creation forms.

Code

• Easy handling/addition of non template fields in reference creation/edition forms.
• First step (no handling yet) of EAFP style for unexpected behaviours when creating or checking Zotero references and when accessing the network for Zotero interactions.

Changed

UI

• Creation/Edition of the reference DOI / PMID / UNPUBLISHED ID:
  • A DOI is displayed and editable in its field from the template of the reference type.
  • A PMID is displayed and editable in the field 'extra' (Zotero doesn't have a specific field for it).
  • An UNPUBLISHED ID can be generated (UUID) and added to the field 'extra'. If one is already there, the generation of a new one is deactivated.
• The editor for the field 'extra' (line separated in Zotero) in reference creation/edition forms is now multiline.
• Only reference creation/edition forms for relevant reference types are created (10 instead of the 35 from Zotero). Less time needed to scroll the list of reference types. Performances increased (less widgets to create).
• Allow only single cell selection in the table of creators in reference creation/edition forms.
• Harmonized layout management for reference creation/edition forms.
• Row text content is now displayed on only one line in the reference list.
• The tab with the reference list is now named 'References (Zotero)'.
• The typo (plural) in the name of the column with the creators of a reference in the reference list is corrected.

Code (for the 'References' tab).

• Better Software Architecture:
  • Qt Model-View pattern implementation,
  • separation of concern with NAT,
  • code complexity reduction,
  • standardization,
  • readability / extensibility / maintainability.
• Integrity of Zotero related data structures and data flows.
• Documentation of all methods (docstring) following PEP 257.
• PEP 8 (code style) compliance.
• More idiomatic Python: list comprehensions, iterators, string formatting, truth values, ...

Fixed

UI

• PDF attachment and parameters search don't work because of deprecated REST server URL still used in some cases.
• Modifying a reference duplicates its PMID or UNPUBLISHED ID in the field 'extra'.
• The value entered for the field 'DOI' is discarded in the reference creation/edition form.
• The field 'DOI' is in the reference creation/edition form but it is not filled with the value.
• #17 The form to add a new reference to the Zotero database doesn't launch.
• #18 Sorting the reference list by the number of annotations doesn't work.
• #39 The selected reference changes after sorting the reference list.
• #38 A random reference is selected for edition when none is selected in the reference list.
• A random reference is selected after a refresh of the Zotero data.
• #10 PyZotero check_items() should be used before adding a reference to the Zotero database.