lsst.pipe.base.argumentParser.IdValueAction Class Reference

argparse action callback to process a data ID into a dict More...

Inheritance diagram for lsst.pipe.base.argumentParser.IdValueAction:

Public Member Functions
def	__call__
	Parse –id data and append results to namespace. More...

Detailed Description

argparse action callback to process a data ID into a dict

Definition at line 659 of file argumentParser.py.

Member Function Documentation

def lsst.pipe.base.argumentParser.IdValueAction.__call__	(		self,
			parser,
			namespace,
			values,
			option_string
	)

Parse –id data and append results to namespace.

<argument>.idList

Parameters

[in]	parser	argument parser (instance of ArgumentParser)
[in,out]	namespace	parsed command (an instance of argparse.Namespace); updated values: <idName>.idList, where <idName> is the name of the ID argument, for instance "id" for ID argument –id
[in]	values	a list of data IDs; see data format below
[in]	option_string	option value specified by the user (a str)

The data format is: key1=value1_1[^value1_2[^value1_3...] key2=value2_1[^value2_2[^value2_3...]...

The values (e.g. value1_1) may either be a string, or of the form "int..int" (e.g. "1..3") which is interpreted as "1^2^3" (inclusive, unlike a python range). So "0^2..4^7..9" is equivalent to "0^2^3^4^7^8^9". You may also specify a stride: "1..5:2" is "1^3^5"

The cross product is computed for keys with multiple values. For example: –id visit 1^2 ccd 1,1^2,2 results in the following data ID dicts being appended to namespace.<argument>.idList: {"visit":1, "ccd":"1,1"} {"visit":2, "ccd":"1,1"} {"visit":1, "ccd":"2,2"} {"visit":2, "ccd":"2,2"}

Definition at line 662 of file argumentParser.py.

    def __call__(self, parser, namespace, values, option_string):
        """!Parse --id data and append results to namespace.<argument>.idList

        @param[in] parser           argument parser (instance of ArgumentParser)
        @param[in,out] namespace    parsed command (an instance of argparse.Namespace);
            updated values:
            - <idName>.idList, where <idName> is the name of the ID argument,
                for instance "id" for ID argument --id
        @param[in] values           a list of data IDs; see data format below
        @param[in] option_string    option value specified by the user (a str)
        
        The data format is:
        key1=value1_1[^value1_2[^value1_3...] key2=value2_1[^value2_2[^value2_3...]...

        The values (e.g. value1_1) may either be a string, or of the form "int..int" (e.g. "1..3")
        which is interpreted as "1^2^3" (inclusive, unlike a python range). So "0^2..4^7..9" is
        equivalent to "0^2^3^4^7^8^9".  You may also specify a stride: "1..5:2" is "1^3^5"
        
        The cross product is computed for keys with multiple values. For example:
            --id visit 1^2 ccd 1,1^2,2
        results in the following data ID dicts being appended to namespace.<argument>.idList:
            {"visit":1, "ccd":"1,1"}
            {"visit":2, "ccd":"1,1"}
            {"visit":1, "ccd":"2,2"}
            {"visit":2, "ccd":"2,2"}
        """
        if namespace.config is None:
            return
        idDict = collections.OrderedDict()
        for nameValue in values:
            name, sep, valueStr = nameValue.partition("=")
            if name in idDict:
                parser.error("%s appears multiple times in one ID argument: %s" % (name, option_string))
            idDict[name] = []
            for v in valueStr.split("^"):
                mat = re.search(r"^(\d+)\.\.(\d+)(?::(\d+))?$", v)
                if mat:
                    v1 = int(mat.group(1))
                    v2 = int(mat.group(2))
                    v3 = mat.group(3); v3 = int(v3) if v3 else 1
                    for v in range(v1, v2 + 1, v3):
                        idDict[name].append(str(v))
                else:
                    idDict[name].append(v)

        keyList = idDict.keys()
        iterList = [idDict[key] for key in keyList]
        idDictList = [collections.OrderedDict(zip(keyList, valList))
            for valList in itertools.product(*iterList)]

        argName = option_string.lstrip("-")
        ident = getattr(namespace, argName)
        ident.idList += idDictList