Source code for pwem.emlib.metadata.utils

# **************************************************************************
# *
# * Authors:     J.M. de la Rosa Trevin (delarosatrevin@scilifelab.se) [1]
# *
# * [1] SciLifeLab, Stockholm University
# *
# * This program is free software; you can redistribute it and/or modify
# * it under the terms of the GNU General Public License as published by
# * the Free Software Foundation; either version 3 of the License, or
# * (at your option) any later version.
# *
# * This program is distributed in the hope that it will be useful,
# * but WITHOUT ANY WARRANTY; without even the implied warranty of
# * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# * GNU General Public License for more details.
# *
# * You should have received a copy of the GNU General Public License
# * along with this program; if not, write to the Free Software
# * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
# * 02111-1307  USA
# *
# *  All comments concerning this program package may be sent to the
# *  e-mail address 'scipion@cnb.csic.es'
# *
# **************************************************************************

from collections import OrderedDict

from pyworkflow.object import ObjectWrap

from ..lib import *


LABEL_TYPES = {
    LABEL_SIZET: float,
    LABEL_DOUBLE: float,
    LABEL_INT: int,
    LABEL_BOOL: bool
}


[docs]class Row: """ Support Xmipp class to store label and value pairs corresponding to a Metadata row. """ def __init__(self): self._labelDict = OrderedDict() # Dictionary containing labels and values self._objId = None # Set this id when reading from a metadata # Add set and get method as alias to setValue and getValue # in this way the Row will behave more like a dict self.set = self.setValue self.get = self.getValue
[docs] def getObjId(self): return self._objId
[docs] def hasLabel(self, label): return self.containsLabel(label)
[docs] def containsLabel(self, label): # Allow getValue using the label string if isinstance(label, str): label = str2Label(label) return label in self._labelDict
[docs] def removeLabel(self, label): if self.hasLabel(label): del self._labelDict[label]
[docs] def setValue(self, label, value): """args: this list should contains tuples with MetaData Label and the desired value""" # Allow setValue using the label string if isinstance(label, str): label = str2Label(label) self._labelDict[label] = value
[docs] def getValue(self, label, default=None): """ Return the value of the row for a given label. """ # Allow getValue using the label string if isinstance(label, str): label = str2Label(label) return self._labelDict.get(label, default)
[docs] def getValueAsObject(self, label, default=None): """ Same as getValue, but making an Object wrapping. """ return ObjectWrap(self.getValue(label, default))
[docs] def readFromMd(self, md, objId): """ Get all row values from a given id of a metadata. """ self._objId = objId try: self._labelDict = md.getRow(objId) except AttributeError: # can be eventually removed, once we are sure that users have recent Xmipp self._labelDict.clear() for label in md.getActiveLabels(): self._labelDict[label] = md.getValue(label, objId)
[docs] def addToMd(self, md): self.writeToMd(md, md.addObject())
[docs] def writeToMd(self, md, objId): """ Set back row values to a metadata row. """ try: md.setRow(self._labelDict, objId) except AttributeError: # can be eventually removed, once we are sure that users have recent Xmipp for label, value in self._labelDict.items(): # TODO: Check how to handle correctly unicode type # in Xmipp and Scipion if type(value) is str: value = str(value) try: if labelType(label) == 2: md.setValue(label, float(value), objId) else: md.setValue(label, value, objId) except Exception as ex: import sys print("XmippMdRow.writeToMd: Error writing value to metadata.", file=sys.stderr) print(" label: %s, value: %s, type(value): %s" % ( label2Str(label), value, type(value)), file=sys.stderr) raise ex
[docs] def readFromFile(self, fn): md = MetaData(fn) self.readFromMd(md, md.firstObject())
[docs] def writeToFile(self, fn): md = MetaData() self.writeToMd(md, md.addObject()) md.write(fn)
[docs] def copyFromRow(self, other): for label, value in other._labelDict.items(): self.setValue(label, value)
[docs] def clone(self): """ Return another Row that have exactly the same values as self. """ row = Row() row.copyFromRow(self) row._objId = self._objId return row
def __str__(self): s = '{' for k, v in self._labelDict.items(): s += ' %s = %s\n' % (label2Str(k), v) return s + '}' def __iter__(self): return self._labelDict.items() def __getitem__(self, item): return self.getValue(item) def __setitem__(self, key, value): return self.setValue(key, value)
[docs] def containsAll(self, labels): """ Check if all labels are present in the row. Params: row: the Row object. labels: either a dict or list object containing the labels (in the case of dicts, label are the dict.values()) """ values = labels.values() if isinstance(labels, dict) else labels return all(self.containsLabel(l) for l in values)
[docs] def containsAny(self, labels): """ Check if at least one of labels is present in the row. Params: row: the Row object. labels: either a dict or list object containing the labels (in the case of dicts, label are the dict.values()) """ values = labels.values() if isinstance(labels, dict) else labels return any(self.containsLabel(l) for l in values)
[docs] def printDict(self): """ Fancy printing of the row, mainly for debugging. """ print(str(self))
[docs]class RowMetaData: """ This class is a wrapper for MetaData in row mode. Where only one object is used. """ def __init__(self, filename=None): self._md = MetaData() self._md.setColumnFormat(False) self._id = self._md.addObject() if filename: self.read(filename)
[docs] def setValue(self, label, value): self._md.setValue(label, value, self._id)
[docs] def getValue(self, label): return self._md.getValue(label, self._id)
[docs] def write(self, filename, mode=MD_APPEND): self._md.write(filename, mode)
[docs] def read(self, filename): self._md.read(filename) self._md.setColumnFormat(False) self._id = self._md.firstObject()
[docs] def containsLabel(self, label): return self._md.containsLabel(label)
def __str__(self): return str(self._md)
[docs]def label2Python(label): """ Return the Python type (int, float, bool) for a given metadata label (LABEL_INT, LABEL_DOUBLE..etc) """ return LABEL_TYPES.get(labelType(label), str)
[docs]def getLabel(value): """ Return the label value either from an int value or an string. """ if isinstance(value, int): return value elif isinstance(value, str): return str2Label(value) else: raise Exception("Invalid value type (%s) for label. " % type(value))
[docs]def getFirstRow(mdOrFn): """ Return the first object of a metadata. Params: mdOrFn: you can pass a metadata or a filename as argument. """ if isinstance(mdOrFn, str): md = MetaData() md.read(mdOrFn, 1) else: # mdOrFn is MetaData md = mdOrFn if md.getParsedLines(): row = Row() row.readFromMd(md, md.firstObject()) else: row = None return row
[docs]def getSize(filename): """ Return the metadata size without parsing entirely. """ md = MetaData() md.read(filename, 1) return md.getParsedLines()
[docs]def isEmpty(filename): """ Use getMdSize to check if metadata is empty. """ return getSize(filename) == 0
[docs]def iterRows(md, sortByLabel=None): """ Iterate over the rows of the given metadata. Params: md: a MetaData object or a filename (MetaData will be read) sortByLabel: a label to sort the metadata before iterate. """ # If md is string, take as filename and create the metadata if isinstance(md, str): md = MetaData(md) if sortByLabel is not None: md.sort(sortByLabel) row = Row() for objId in md: row.readFromMd(md, objId) yield row
[docs]def dropColumns(mdObj, *labels): """ Drop all columns from a given metadata. Labels can be either string or int. """ for l in labels: mdObj.removeLabel(getLabel(l))
[docs]def keepColumns(mdObj, *labels): """ Drop all columns from mdObj that are not in labels. Labels can be either string or int. """ # Handle string or int labels input keepLabels = {getLabel(l) for l in labels} for l in mdObj.getActiveLabels(): if l not in keepLabels: mdObj.removeLabel(l)
[docs]def joinBlocks(inputMd, blockPrefix=None): mdImages = MetaData() mdAll = MetaData() mdBlocks = getBlocksInMetaDataFile(inputMd) for mdBlock in mdBlocks: if blockPrefix is not None: if mdBlock.startswith(blockPrefix): mdImages.read(mdBlock + "@" + inputMd) mdAll.unionAll(mdImages) else: mdImages.read(mdBlock + "@" + inputMd) mdAll.unionAll(mdImages) return mdAll
[docs]class SetMdIterator: """ Class to iterate over an input set and skip elements not present in metadata. This class can be used in copyItems when the number of elements in the set is higher that in metadata. """ def __init__(self, md, sortByLabel=None, keyLabel=MDL_ITEM_ID, updateItemCallback=None, skipDisabled=False): if updateItemCallback is None: raise Exception('Set an updateItemCallback') self.iterMd = iterRows(md, sortByLabel) self.keyLabel = keyLabel self.updateItemCallback = updateItemCallback self.skipDisabled = skipDisabled self.__nextRow() def __nextRow(self): try: self.lastRow = next(self.iterMd) except StopIteration: self.lastRow = None
[docs] def updateItem(self, item, row): """ This function should be passed to copyItems as callback and it will filter the items not present in the metadata. """ row = self.lastRow if row is not None: if row.hasLabel(MDL_ENABLED): enabled = row.getValue(MDL_ENABLED) else: enabled = 1 if (row is None or item.getObjId() != row.getValue(self.keyLabel)): item._appendItem = False elif enabled == -1 and self.skipDisabled: item._appendItem = False self.__nextRow() else: item._appendItem = True self.updateItemCallback(item, row) self.__nextRow()