# This file contains metadata for your plugin.

# This file should be included when you package your plugin.# Mandatory items:

[general]
name=pandapower QGis Plugin
qgisMinimumVersion=3.44
qgisMaximumVersion=4.99
description=Open pandapower networks as a QGIS data source
version=0.0.5
author=Fraunhofer IEE
email=mike.vogt@iee.fraunhofer.de

about=Opens a pandapower network as a data source, like a database connection rather than an import. Expand a network file in the Browser or pick it in the Data Source Manager to get its tables as layers, edit the geodata, and have the changes written straight back to the same JSON file.

tracker=https://github.com/e2nIEE/pandapower-qgis/issues
repository=https://github.com/e2nIEE/pandapower-qgis
# End of mandatory metadata

# Recommended items:

hasProcessingProvider=no
# Filled in from CHANGELOG.md by qgis-plugin-ci when the package is built.
changelog=
 Version 0.0.5:
 First release outside the experimental channel.
 **QGIS 4**
 * The plugin runs on QGIS 4 (Qt6) as well as QGIS 3.44 (Qt5).
 * Ubuntu 24.04 (noble) with QGIS from apt is not supported: its system numpy 1.x
   conflicts with the numpy 2 that pandapower's dependencies require.
 **Security**
 * The extra parameters field of *Run power flow* no longer runs the text as Python.
   Only literal values (`algorithm='nr', max_iteration=10`) are accepted.
 * The plugin no longer installs packages itself with pip. Dependencies are installed
   by the qpip plugin from `requirements.txt`.
 **Fixes**
 * Export skips multi-part geometries and selects them, instead of stopping with an
   assertion error.
 * Errors that were silently ignored are now written to the *Pandapower* tab of the
   message log.
 * The plugin package no longer contains developer scripts (`.bat`, `.sh`).
 **Requires** pandapower 3.5.6 or newer. numpy and scipy are no longer pinned.

 Version 0.0.4:
 pandapower networks are a data source, not an import.
 **Networks are now opened, not imported.**
 * pandapower appears in the **Data Source Manager** alongside PostgreSQL and SAP HANA.
   Pick a network file, select the tables you want, press Add.
 * Every pandapower table can be opened, not just `bus` and `line`. Tables without
   geometry (`trafo`, `load`, `switch`, …) open as plain attribute tables.
 * `res_*` tables can be opened in their own right, while their columns stay merged
   into the `bus`/`line` layers for styling.
 * Only tables that contain rows are listed, so the dialog stays readable.
 **Known limitation**
 * Expanding a pandapower `.json` directly in the **Browser** panel is not available.
   Returning a Python subclass of `QgsDataCollectionItem` from
   `QgsDataItemProvider.createDataItem()` is unsafe on QGIS 3.44 — the object loses
   its Python half once QGIS stores it, and expanding the node crashes QGIS. This
   reproduces without this plugin, so it is a binding limitation rather than
   something the plugin can work around. Use the Data Source Manager instead; the
   code is kept behind `ENABLE_BROWSER_TREE` in `pandapower_qgis.py`.
 **Edits are written on commit.**
 * Changes are written to the `.json` when the layer edits are saved, not on every
   individual change. Rolling back writes nothing.
 * All layers of one network share a single in-memory network, so one save writes
   them all and an edit in one layer is immediately visible in the others.
 * A file that changed on disk is no longer silently overwritten.
 **Removed**
 * The *import from pandapower* action, together with `ppqgis_import.py` and its
   dialog. The Browser and Data Source Manager replace it. Note the old import
   reprojected the network's coordinates in place; opening a network now leaves the
   stored coordinates untouched and lets QGIS reproject for display.
 * The vendored `geo.py`, superseded by `pandapower.plotting.geo` upstream.
 * The asynchronous save machinery, and with it the "previous save operation is
   still running" class of errors.
 *Export is unchanged* — it still builds a new network from arbitrary QGIS layers.
 **Fixes**
 * Plugin no longer fails to load on a profile where the locale setting is unset.
 * Geometry edits no longer wrote into a pandas copy that was then discarded.
 * Renderer instances are no longer shared between layers (a double-ownership crash).
 * Closing a layer no longer deregisters the provider for every other open layer.
 * `pressure_level` in a layer URI was written but never read, so pipe layers lost
   their level filter.
 **Requires** pandapower 3.5 or newer, which stores geodata in the `geo` column.

 Version 0.0.3:
 * Feature/dataprovider was included, which allows to use pandapower networks as a dataprovider.
 * Large restructuring



# Tags are comma separated with spaces allowed
tags=python,pandapower,power grid,network,electricity,database

homepage=https://github.com/e2nIEE/pandapower-qgis
category=Plugins
icon=pp.png
# experimental flag
experimental=False

# deprecated flag (applies to the whole plugin, not just a single version)
deprecated=False

# Since QGIS 3.8, a comma separated list of plugins to be installed
# (or upgraded) can be specified.
# Check the documentation for more information.
plugin_dependencies=qpip

# Runs on QGIS 4, which is built with Qt6.
supportsQt6=True

# Category of the plugin: Raster, Vector, Database or Web
#category=Database

# If the plugin can run on QGIS Server.
server=False

