Source code for pybnf.printing

"""Contains printing commands that respect the application-wide verbosity setting."""


verbosity = 1




def print0(s):
    """Print the statement at any verbosity level"""
    print(s)





def print1(s):
    """Print the statement only if the verbosity level is at least 1"""
    if verbosity >= 1:
        print(s)





def print2(s):
    """Print the statement only if the verbosity level is 2"""
    if verbosity >= 2:
        print(s)





class PybnfError(RuntimeError):
    """
    Represents a user-generated error for which we can provide an informative message to the user about what
    went wrong with the input before quitting.

    Two independent knobs shape what the user sees. ``user_message`` **replaces** the log message -- for a
    diagnosis that is too technical to show as-is, restated in user-facing terms. ``hint`` **appends** to it --
    for the remedy the user needs *alongside* the diagnosis, not instead of it. A refusal that carries only a
    generic remedy in ``user_message`` silently discards its own reason (#527): the reason reaches the log and
    the user is told to give up without being told what to fix.
    """
    def __init__(self, log_message, user_message=None, hint=None):
        """
        :param log_message: The message to print to the log
        :param user_message: The message to output to the user *in place of* log_message. If omitted, the user
        gets the same message as log_message.
        :param hint: One suggested remedy, or a sequence of them, appended to the user-facing message as an
        indented ``->`` line each. Adds to the message rather than replacing it, so the user gets both what
        went wrong and what to do about it.
        """
        self.log_message = log_message
        self.hints = [hint] if isinstance(hint, str) else list(hint or ())
        self.message = user_message if user_message else log_message
        for h in self.hints:
            self.message += '\n  -> ' + h





class DataMismatchError(PybnfError):
    """A mismatch between the experimental data and what the fit simulates or reads, found
    while scoring: a data column no simulated output matches, or a column, binding table or
    model entity that a noise model needs and the data or simulation does not have.

    It is a property of the setup, not of the parameter set being scored, so every evaluation
    would meet it. The scoring paths therefore let it end the run with its message, where any
    other error from scoring one parameter set is recorded as a failed evaluation and the fit
    goes on (#388). Before this class existed both paths recorded it as a failed evaluation,
    so a fit could finish having scored only the columns that matched, or nothing at all
    (lanl/PyBNF#817, #1030).
    """