pearl-public/pmsco-public
2
1
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

public release 2.2.0 - see README.md and CHANGES.md for details

Browse Source
This commit is contained in:
muntwiler_m
2020-09-04 16:22:42 +02:00
parent fbd2d4fa8c
commit 7c61eb1b41
67 changed files with 2934 additions and 682 deletions
Download Patch File Download Diff File Expand all files Collapse all files

119
pmsco/files.py
View File

@@ -27,20 +27,20 @@ logger = logging.getLogger(__name__)
#
# each string of this set marks a category of files.
#
# @arg @c 'input' : raw input files for calculator, including cluster and atomic files in custom format
# @arg @c 'output' : raw output files from calculator
# @arg @c 'atomic' : atomic scattering (phase, emission) files in portable format
# @arg @c 'cluster' : cluster files in portable XYZ format for report
# @arg @c 'log' : log files
# @arg @c 'debug' : debug files
# @arg @c 'model': output files in ETPAI format: complete simulation (a_-1_-1_-1_-1)
# @arg @c 'scan' : output files in ETPAI format: scan (a_b_-1_-1_-1)
# @arg @c 'symmetry' : output files in ETPAI format: symmetry (a_b_c_-1_-1)
# @arg @c 'emitter' : output files in ETPAI format: emitter (a_b_c_d_-1)
# @arg @c 'region' : output files in ETPAI format: region (a_b_c_d_e)
# @arg @c 'report': final report of results
# @arg @c 'population': final state of particle population
# @arg @c 'rfac': files related to models which give bad r-factors (dynamic category, see below).
# @arg 'input' : raw input files for calculator, including cluster and atomic files in custom format
# @arg 'output' : raw output files from calculator
# @arg 'atomic' : atomic scattering (phase, emission) files in portable format
# @arg 'cluster' : cluster files in portable XYZ format for report
# @arg 'log' : log files
# @arg 'debug' : debug files
# @arg 'model': output files in ETPAI format: complete simulation (a_-1_-1_-1_-1)
# @arg 'scan' : output files in ETPAI format: scan (a_b_-1_-1_-1)
# @arg 'domain' : output files in ETPAI format: domain (a_b_c_-1_-1)
# @arg 'emitter' : output files in ETPAI format: emitter (a_b_c_d_-1)
# @arg 'region' : output files in ETPAI format: region (a_b_c_d_e)
# @arg 'report': final report of results
# @arg 'population': final state of particle population
# @arg 'rfac': files related to models which give bad r-factors (dynamic category, see below).
#
# @note @c 'rfac' is a dynamic category not connected to a particular file or content type.
# no file should be marked @c 'rfac'.
@@ -48,7 +48,7 @@ logger = logging.getLogger(__name__)
# if so, all files related to bad models are deleted, regardless of their static category.
#
FILE_CATEGORIES = {'cluster', 'atomic', 'input', 'output',
'report', 'region', 'emitter', 'scan', 'symmetry', 'model',
'report', 'region', 'emitter', 'scan', 'domain', 'model',
'log', 'debug', 'population', 'rfac'}
## @var FILE_CATEGORIES_TO_KEEP
@@ -242,37 +242,52 @@ class FileTracker(object):
else:
self._complete_models.discard(model)
def delete_files(self, categories=None):
def delete_files(self, categories=None, incomplete_models=False):
"""
delete the files matching the list of categories.
delete all files matching a set of categories.
@version this method does not act on the 'rfac' category.
this function deletes all files that are tagged with one of the given categories.
tags are set by the code sections that create the files.
for a list of common categories, see FILE_CATEGORIES.
the categories can be given as an argument or taken from the categories_to_delete property.
files are deleted regardless of R-factor.
be sure to specify only categories that you don't need in the output at all.
by default, only files of complete models (cf. set_model_complete()) are deleted
to avoid interference with running calculations.
to clean up after calculations, the incomplete_models argument can override this.
@note this method does not act on the special 'rfac' category (see delete_bad_rfac()).
@param categories: set of file categories to delete.
defaults to self.categories_to_delete.
if the argument is None, it defaults to the categories_to_delete property.
@param incomplete_models: (bool) delete files of incomplete models as well.
by default (False), incomplete models are not deleted.
@return: None
"""
if categories is None:
categories = self.categories_to_delete
for cat in categories:
self.delete_category(cat)
self.delete_category(cat, incomplete_models=incomplete_models)
def delete_bad_rfac(self, keep=0, force_delete=False):
"""
delete the files of all models except a specified number of good models.
delete all files of all models except for a specified number of best ranking models.
the method first determines which models to keep.
models with R factor values of 0.0, without a specified R-factor, and
the specified number of best ranking non-zero models are kept.
the files belonging to the keeper models are kept, all others are deleted,
regardless of category.
files of incomplete models are also kept.
in addition, incomplete models, models with R factor = 0.0,
and those without a specified R-factor are kept.
all other files are deleted.
the method does not consider the file category.
the files are deleted from the list and the file system.
files are deleted only if 'rfac' is specified in self.categories_to_delete
or if force_delete is set to True.
the method executes only if 'rfac' is specified in self.categories_to_delete
or if force_delete is True.
otherwise the method does nothing.
@param keep: number of files to keep.
@@ -330,18 +345,32 @@ class FileTracker(object):
return len(del_models)
def delete_category(self, category):
def delete_category(self, category, incomplete_models=False):
"""
delete all files of a specified category from the list and the file system.
only files of complete models (cf. set_model_complete()) are deleted, but regardless of R-factor.
this function deletes all files that are tagged with the given category.
tags are set by the code sections that create the files.
for a list of common categories, see FILE_CATEGORIES.
files are deleted regardless of R-factor.
be sure to specify only categories that you don't need in the output at all.
by default, only files of complete models (cf. set_model_complete()) are deleted
to avoid interference with running calculations.
to clean up after calculations, the incomplete_models argument can override this.
@param category: (str) category.
should be one of FILE_CATEGORIES. otherwise, the function has no effect.
@param incomplete_models: (bool) delete files of incomplete models as well.
by default (False), incomplete models are not deleted.
@return: None
"""
del_names = {name for (name, cat) in self._file_category.items() if cat == category}
del_names &= {name for (name, model) in self._file_model.items() if model in self._complete_models}
if not incomplete_models:
del_names &= {name for (name, model) in self._file_model.items() if model in self._complete_models}
for name in del_names:
self.delete_file(name)
@@ -375,3 +404,33 @@ class FileTracker(object):
logger.warning("file system error deleting file {0}".format(path))
else:
logger.debug("delete file {0} ({1}, model {2})".format(path, cat, model))
def list_files_other_models(prefix, models):
"""
list input/output files except those of the given models.
this can be used to clean up all files except those belonging to the given models.
to delete the listed files:
for f in files:
os.remove(f)
@param prefix: file name prefix up to the first underscore.
only files starting with this prefix are listed.
@param models: sequence or set of model numbers that should not be listed.
@return: set of file names
"""
file_names = set([])
for entry in os.scandir():
if entry.is_file:
elements = entry.name.split('_')
try:
if len(elements) == 6 and elements[0] == prefix and int(elements[1]) not in models:
file_names.add(entry.name)
except (IndexError, ValueError):
pass
return file_names