Update the nebular models to extend the list of metallicities and address a few issues in the list of lines.

Improve the description of the data file format by specifying how to indicate upper limits and missing values.

Import Iterable from collections.abc rather than from collections in order to prevent ImportError/AttributeError exceptions with Python 3.10+.

When running pcigale init several times with an existing pcigale.ini, the top comment was repeatedly included due to the default behaviour of configobj

Emit an exception when generating the pcigale.ini file the m2005 in case it is used in combination with the nebular and/or xray modules.

Determine the intrinsic UV slope β₀ at the same time as the observed UV slope β. To compensate for the additional time spent, the computation of the slope is optimized.

The terminal output has been completely reworked. The level of importance of the messages are now stated (INFO, WARNING, ERROR). Progress bars have been added, indicating the total number of items processed, the number of items processed per second, the elapsed time, the remaining time. Colours are used to improve the legibility of the information.

Finalise the merge of new_xcig into the develop branch by adding an entry to the CHANGELOG file mentioning the integration of the last X-CIGALE features into CIGALE.

How upper limits are taken into account is now more flexible. In addition to the exact computation, it is now possible to waive the adjustment of the scaling of the models, an extremely slow operation that generally does not affect the results in any substantial manner. Given the important speedup, this option is now selected by default in pdf_analysis.

Remove the del command from pcigale-filter as the future database makes it unnecessary to delete filters.

The test to determine if we need to reinterpolate all the components when adding a new one has been made more efficient, yielding a speedup of 1.5-2.0%.

The storage of individual spectral physical components has been improved in order to require fewer memory allocations and copies during the computation of models. The exact gain depends somewhat on the parameter space but speedups of about 50% are typically seen.

Now that Python dictionaries are officially ordered, use regular dictionaries rather than collections.OrderedDict. It should translate into a very slight speedup and a slightly more readable code.

Remove the fluxes module has it has been superseded by the bands parameter of the pdf_analysis module.

Following a past change in the way chi2 files are saved, adapt pcigale_plot to correctly produce PDF plots for physical properties that are in log rather than crashing.

Fix incorrect filename handling that prevented the import of the skirtor2016 models under Microsoft Windows

Fix incorrect filename handling that prevented the import of the skirtor2016 models under Microsoft Windows

Modify the IGM absorption from Meiksin to include the dependence of the hydrogen absorption cross section so that the universe becomes transparent again at very short wavelengths.

SED plots could not be generated for runs without a dust attenuation module as pcigale-plots sed assumed there would be one. This case is now handled.

An input redshift of integer type caused pcigale-plot sed to crash because it could not be formatted as a float. It is now explicitly converted to a float for safety.