logging --- Logging facility for Python

Code source : Lib/logging/__init__.py


Ce module définit les fonctions et les classes qui mettent en œuvre un système flexible d’enregistrement des événements pour les applications et les bibliothèques.

L'intérêt principal d'utiliser une API de journalisation fournie par un module de bibliothèque standard réside dans le fait que tous les modules Python peuvent participer à la journalisation, de sorte que le journal de votre application peut inclure vos propres messages et ceux de modules tiers.

Here's a simple example of idiomatic usage:

# myapp.py
import logging
import mylib
logger = logging.getLogger(__name__)

def main():
    logging.basicConfig(filename='myapp.log', level=logging.INFO)
    logger.info('Started')
    mylib.do_something()
    logger.info('Finished')

if __name__ == '__main__':
    main()
# mylib.py
import logging
logger = logging.getLogger(__name__)

def do_something():
    logger.info('Doing something')

Si vous exécutez myapp.py, vous verrez ceci dans myapp.log :

INFO:__main__:Started
INFO:mylib:Doing something
INFO:__main__:Finished

The key feature of this idiomatic usage is that the majority of code is simply creating a module level logger with getLogger(__name__), and using that logger to do any needed logging. This is concise, while allowing downstream code fine-grained control if needed. Logged messages to the module-level logger get forwarded to handlers of loggers in higher-level modules, all the way up to the highest-level logger known as the root logger; this approach is known as hierarchical logging.

For logging to be useful, it needs to be configured: setting the levels and destinations for each logger, potentially changing how specific modules log, often based on command-line arguments or application configuration. In most cases, like the one above, only the root logger needs to be so configured, since all the lower level loggers at module level eventually forward their messages to its handlers. basicConfig() provides a quick way to configure the root logger that handles many use cases.

Ce module offre de nombreuses fonctionnalités et une grande flexibilité. Si vous n'êtes pas familier avec la journalisation, la meilleure façon de l’appréhender est de consulter les tutoriels (voir les liens ci-dessus et à droite).

The basic classes defined by the module, together with their attributes and methods, are listed in the sections below.

  • les enregistreurs (loggers en anglais) exposent l'interface que le code de l'application utilise directement ;

  • les gestionnaires (handlers) envoient les entrées de journal (créées par les loggers) vers les destinations voulues ;

  • les filtres (filters) fournissent un moyen de choisir plus finement quelles entrées de journal doivent être sorties ;

  • les formateurs (formatters) définissent la structure de l'entrée de journal dans la sortie finale.

Enregistreurs

Les enregistreurs ont les attributs et les méthodes suivants. Notez que les enregistreurs ne doivent JAMAIS être instanciés directement, mais toujours par la fonction au niveau du module logging.getLogger(nom). Les appels à getLogger() avec le même nom renvoient toujours une référence au même objet enregistreur.

The name is potentially a period-separated hierarchical value, like foo.bar.baz (though it could also be just plain foo, for example). Loggers that are further down in the hierarchical list are children of loggers higher up in the list. For example, given a logger with a name of foo, loggers with names of foo.bar, foo.bar.baz, and foo.bam are all descendants of foo. In addition, all loggers are descendants of the root logger. The logger name hierarchy is analogous to the Python package hierarchy, and identical to it if you organise your loggers on a per-module basis using the recommended construction logging.getLogger(__name__). That's because in a module, __name__ is the module's name in the Python package namespace.

class logging.Logger
name

This is the logger's name, and is the value that was passed to getLogger() to obtain the logger.

Note

This attribute should be treated as read-only.

level

The threshold of this logger, as set by the setLevel() method.

Note

Do not set this attribute directly - always use setLevel(), which has checks for the level passed to it.

parent

The parent logger of this logger. It may change based on later instantiation of loggers which are higher up in the namespace hierarchy.

Note

This value should be treated as read-only.

propagate

Si cet attribut est évalué comme vrai, les événements enregistrés dans cet enregistreur seront transmis aux gestionnaires des enregistreurs de niveau supérieur (parents), en plus des gestionnaires attachés à l'enregistreur. Les messages sont transmis directement aux gestionnaires des enregistreurs parents — ni le niveau ni les filtres des enregistreurs parentaux en question ne sont pris en compte.

S’il s’évalue comme faux, les messages de journalisation ne sont pas transmis aux gestionnaires des enregistreurs parents.

Spelling it out with an example: If the propagate attribute of the logger named A.B.C evaluates to true, any event logged to A.B.C via a method call such as logging.getLogger('A.B.C').error(...) will [subject to passing that logger's level and filter settings] be passed in turn to any handlers attached to loggers named A.B, A and the root logger, after first being passed to any handlers attached to A.B.C. If any logger in the chain A.B.C, A.B, A has its propagate attribute set to false, then that is the last logger whose handlers are offered the event to handle, and propagation stops at that point.

Le constructeur fixe cet attribut à True.

Note

si vous associez un gestionnaire à un enregistreur et à un ou plusieurs de ses parents, il peut émettre le même enregistrement plusieurs fois. En général, il est rare d'avoir besoin d'attacher un gestionnaire à plus d'un enregistreur — si vous l'attachez simplement à l'enregistreur approprié le plus haut possible dans la hiérarchie des enregistreurs, alors il voit tous les événements enregistrés par tous les enregistreurs descendants, à condition que leur paramètre de propagation soit laissé à True. La pratique la plus courante est de n'attacher les gestionnaires qu'à l'enregistreur racine et à laisser la propagation s'occuper du reste.

handlers

The list of handlers directly attached to this logger instance.

Note

This attribute should be treated as read-only; it is normally changed via the addHandler() and removeHandler() methods, which use locks to ensure thread-safe operation.

disabled

This attribute disables handling of any events. It is set to False in the initializer, and only changed by logging configuration code.

Note

This attribute should be treated as read-only.

setLevel(level)

Fixe le seuil de l'enregistreur au niveau level. Les messages de journalisation qui sont moins graves que level sont ignorés ; les messages qui ont une gravité égale à level ou plus élevée sont traités par le ou les gestionnaires de l'enregistreur, à moins que le niveau d'un gestionnaire n'ait été fixé à un niveau de gravité plus élevé que level.

Lorsqu'un enregistreur est créé, son niveau est fixé à NOTSET (ce qui entraîne le traitement de tous les messages lorsque l'enregistreur est l'enregistreur racine, ou la délégation au parent lorsque l'enregistreur est un enregistreur non racine). Notez que l'enregistreur racine est créé avec le niveau WARNING.

Le terme « délégation au parent » signifie que si un enregistreur a un niveau de NOTSET, sa chaîne d’enregistreurs parents est parcourue jusqu'à ce qu'un parent ayant un niveau autre que NOTSET soit trouvé, ou que la racine soit atteinte.

Si un parent est trouvé avec un niveau autre que NOTSET, alors le niveau de ce parent est utilisé comme le niveau effectif de l'enregistreur d'où la recherche du parent a commencé. Ce niveau est utilisé pour déterminer comment un événement d'enregistrement est traité.

Si la racine est atteinte, et qu'elle a un niveau de NOTSET, alors tous les messages sont traités. Sinon, le niveau de la racine est utilisé comme niveau effectif.

Voir Niveaux de journalisation pour la liste des niveaux.

Modifié dans la version 3.2: le paramètre level peut désormais être une chaîne de caractères représentant le niveau de gravité (comme 'INFO') en plus des constantes entières (comme INFO). Notez cependant que les niveaux sont stockés en interne sous forme d'entiers, et que des méthodes telles que getEffectiveLevel() et isEnabledFor() renvoient et s'attendent à recevoir des entiers.

isEnabledFor(level)

Indique si un message de gravité level est traitable par cet enregistreur. Cette méthode vérifie d'abord le niveau de gravité du module défini par logging.disable(level) et ensuite le niveau effectif de l'enregistreur, déterminé par getEffectiveLevel().

getEffectiveLevel()

Indique le niveau effectif de l'enregistreur. Si une valeur autre que NOTSET a été définie en utilisant setLevel(), elle est renvoyée. Sinon, la hiérarchie est parcourue vers la racine jusqu'à ce qu'une valeur autre que NOTSET soit trouvée, et cette valeur est renvoyée. La valeur renvoyée est un entier, par exemple logging.DEBUG, logging.INFO, etc.

getChild(suffix)

Renvoie un enregistreur enfant de l'enregistreur, déterminé par suffix. Ainsi, logging.getLogger('abc').getChild('def.ghi') renvoie le même enregistreur que logging.getLogger('abc.def.ghi'). C'est une méthode destinée à simplifier la vie du développeur. Elle est très utile quand l’enregistreur parent est nommé en utilisant, par exemple, __name__ plutôt qu'une chaîne de caractères littérale.

Ajouté dans la version 3.2.

getChildren()

Returns a set of loggers which are immediate children of this logger. So for example logging.getLogger().getChildren() might return a set containing loggers named foo and bar, but a logger named foo.bar wouldn't be included in the set. Likewise, logging.getLogger('foo').getChildren() might return a set including a logger named foo.bar, but it wouldn't include one named foo.bar.baz.

Ajouté dans la version 3.12.

debug(msg, *args, **kwargs)

Enregistre un message de niveau DEBUG dans cet enregistreur. msg est la chaîne du message qui sera formatée avec args en utilisant l'opérateur de formatage. Cela signifie qu'il est possible de mettre des mots-clé dans la chaîne et de passer un dictionnaire en argument. Si args n'est pas fourni, aucun formatage « à la % » n'est appliqué.

Quatre mots-clés de kwargs sont analysés : exc_info, stack_info, stacklevel et extra.

Si la valeur booléenne de exc_info est vraie, les informations des exceptions sont ajoutées au message. Si exc_info est un n-uplet d'exception (au format identique aux valeurs renvoyées par sys.exc_info()) ou une instance d'exception, sa valeur est utilisée. Dans le cas contraire, les informations sur l'exception sont déterminées par un appel à sys.exc_info().

Le deuxième argument par mot-clé optionnel est stack_info, valant False par défaut. S'il est vrai, les informations de la pile d'appels sont ajoutées à l'entrée de journal, en incluant aussi l'appel à la fonction de journalisation. Ce ne sont pas les mêmes informations que celles affichées en définissant exc_info : les premières représentent les appels de fonctions successifs, du bas de la pile jusqu'à l'appel de la fonction de journalisation dans le fil d'exécution actuel, alors que les secondes portent des informations sur les appels successifs déclenchés par la levée d'une exception et la recherche de gestionnaires pour cette exception.

Il est possible de définir stack_info indépendamment de exc_info, p. ex. pour s'assurer que l'exécution a atteint un certain point dans le code, même si aucune exception n'a été levée. La pile d'appels est alors affichée après la ligne d'en-tête suivante :

Stack (most recent call last):

Elle imite la ligne Traceback (most recent call last): affichée avec la pile d'appels d'une exception.

Le troisième argument par mot-clé optionnel est stacklevel, valant 1 par défaut. S'il est supérieur à 1, il correspond au nombre d'entrées dans la pile qui sont ignorées en déterminant le numéro de ligne et le nom de la fonction dans la classe LogRecord créée pour l'évènement de journalisation. C'est utile pour les utilitaires de journalisation car cela permet d'ignorer les informations (nom de fonction, fichier source et ligne) de l'utilitaire et de ne traiter que celles de l'appelant. Le nom de ce paramètre est le même que son équivalent dans le module warnings.

The fourth keyword argument is extra which can be used to pass a dictionary which is used to populate the __dict__ of the LogRecord created for the logging event with user-defined attributes. These custom attributes can then be used as you like. For example, they could be incorporated into logged messages. For example:

FORMAT = '%(asctime)s %(clientip)-15s %(user)-8s %(message)s'
logging.basicConfig(format=FORMAT)
d = {'clientip': '192.168.0.1', 'user': 'fbloggs'}
logger = logging.getLogger('tcpserver')
logger.warning('Protocol problem: %s', 'connection reset', extra=d)

affiche

2006-02-08 22:20:02,165 192.168.0.1 fbloggs  Protocol problem: connection reset

Les clés du dictionnaire passé dans extra ne doivent pas être les mêmes que les clés utilisées par le système de journalisation. Voir la documentation de la classe Formatter pour plus de précisions sur les clés utilisées par le système de journalisation.

Si vous choisissez d'utiliser des attributs dans les messages à journaliser, il faut être prudent. Ainsi, dans l'exemple précédent, le Formatter a été configuré avec une chaîne qui attend clientip et user dans le dictionnaire d'attributs du LogRecord. S'ils sont manquants, le message n'est pas enregistré car une exception de formatage de chaîne est levée. Il faut alors toujours passer un dictionnaire extra avec ces clés.

Même si elle peut sembler gênante, cette fonctionnalité est nécessaire dans certains cas, comme sur des serveurs à fils d'exécution multiples, où le même code s'exécute dans des contextes différents et où les évènements significatifs dépendent du contexte (comme l'adresse IP du client et le nom d'utilisateur dans l'exemple précédent). Dans ces circonstances, il est clair que les classes Formatters spécialisées doivent être utilisées avec des Handlers particuliers.

If no handler is attached to this logger (or any of its ancestors, taking into account the relevant Logger.propagate attributes), the message will be sent to the handler set on lastResort.

Modifié dans la version 3.2: ajout du paramètre stack_info.

Modifié dans la version 3.5: le paramètre exc_info peut être une instance d'exception.

Modifié dans la version 3.8: ajout du paramètre stacklevel.

info(msg, *args, **kwargs)

Enregistre un message avec le niveau de gravité INFO. Les arguments ont la même signification que pour debug().

warning(msg, *args, **kwargs)

Enregistre un message avec le niveau de gravité WARNING. Les arguments ont la même signification que pour debug().

Note

Il existe une méthode obsolète warn qui est identique à warning. warn n'est plus maintenue, prière de ne plus l'utiliser et de la remplacer par warning.

error(msg, *args, **kwargs)

Enregistre un message avec le niveau de gravité ERROR. Les arguments ont la même signification que pour debug().

critical(msg, *args, **kwargs)

Enregistre un message avec le niveau de gravité CRITICAL. Les arguments ont la même signification que pour debug().

log(level, msg, *args, **kwargs)

Enregistre un message avec le niveau de gravité level. Les arguments ont la même signification que pour debug().

exception(msg, *args, **kwargs)

Enregistre un message avec le niveau de gravité ERROR. Les arguments ont la même signification que pour debug(). Des informations sur l'exception sont ajoutées au message. Cette méthode doit être appelée depuis un gestionnaire d'exceptions.

addFilter(filter)

Ajoute le filtre filter à l'enregistreur.

removeFilter(filter)

Retire le filtre filter de cet enregistreur.

filter(record)

Applique les filtres associés à l'enregistreur et renvoie True si l'entrée doit être traitée. Les filtres sont appliqués les uns après les autres, jusqu'à ce que l'un renvoie faux. Si ce n'est pas le cas, le traitement de l'entrée se poursuit (elle est alors passée aux gestionnaires). Si l'un d'entre eux renvoie faux, le traitement de l'entrée s'arrête.

addHandler(hdlr)

Ajoute le gestionnaire hdlr à l'enregistreur.

removeHandler(hdlr)

Retire le gestionnaire hdlr de l'enregistreur.

findCaller(stack_info=False, stacklevel=1)

Détermine le fichier source et la ligne de l'appelant. Renvoie un quadruplet contenant le nom du fichier source, le numéro de ligne, le nom de la fonction et la pile d'appels. La pile vaut None si stack_info n'est pas True.

Le paramètre stacklevel est passé par le code appelant debug() (entre autres). S'il est supérieur à 1, n-1 entrées de la pile sont supprimées avant de renvoyer la pile. C'est pratique quand on appelle des APIs de journalisation à travers du code d'encapsulation car cela permet de retirer les informations sur ce code de l'entrée, tout en conservant celles sur le code au-dessus du code d'encapsulation.

handle(record)

Traite une entrée en la passant à tous les gestionnaires de l'enregistreur et à tous ses parents (jusqu'à ce qu'un propagate soit faux). C'est pratique pour désérialiser les entrées reçues d'un connecteur et celles créées localement. Du filtrage au niveau de l'enregistreur est appliqué en appelant filter().

makeRecord(name, level, fn, lno, msg, args, exc_info, func=None, extra=None, sinfo=None)

Fabrique qui peut être redéfinie pour créer des instances de LogRecord.

hasHandlers()

Vérifie si l'enregistreur a des gestionnaires associés. Elle recherche les gestionnaires de l'enregistreur et ceux de ses parents. Renvoie True si au moins un gestionnaire a été trouvé et False sinon. Cette méthode arrête de remonter la hiérarchie dès qu'un enregistreur avec l'attribut propagate à faux est rencontré ­— cet enregistreur est alors le dernier dans lequel la méthode cherche des gestionnaires.

Ajouté dans la version 3.2.

Modifié dans la version 3.7: les enregistreurs peuvent être sérialisés et désérialisés.

Niveaux de journalisation

Les valeurs numériques des niveaux de journalisation sont données dans le tableau suivant. Celles-ci n'ont d'intérêt que si vous voulez définir vos propres niveaux, avec des valeurs spécifiques par rapport aux niveaux prédéfinis. Si vous définissez un niveau avec la même valeur numérique, il écrase la valeur prédéfinie ; le nom prédéfini est perdu.

Niveau

Valeur numérique

What it means / When to use it

logging.NOTSET

0

When set on a logger, indicates that ancestor loggers are to be consulted to determine the effective level. If that still resolves to NOTSET, then all events are logged. When set on a handler, all events are handled.

logging.DEBUG

10

Detailed information, typically only of interest to a developer trying to diagnose a problem.

logging.INFO

20

Confirmation que tout fonctionne comme prévu.

logging.WARNING

30

An indication that something unexpected happened, or that a problem might occur in the near future (e.g. 'disk space low'). The software is still working as expected.

logging.ERROR

40

Du fait d'un problème plus sérieux, le logiciel n'a pas été capable de réaliser une tâche.

logging.CRITICAL

50

Une erreur sérieuse, indiquant que le programme lui-même pourrait être incapable de continuer à fonctionner.

Gestionnaires

Handlers have the following attributes and methods. Note that Handler is never instantiated directly; this class acts as a base for more useful subclasses. However, the __init__() method in subclasses needs to call Handler.__init__().

class logging.Handler
__init__(level=NOTSET)

Initialise l'objet Handler en définissant le niveau de gravité, en initialisant la liste des filtres avec une liste vide et en créant un verrou (avec createLock()) pour sérialiser l'accès au mécanisme d'E-S.

createLock()

Instancie un verrou qui peut être utilisé pour sérialiser l'accès au système d'E-S sous-jacent (qui peut ne pas être à fil d'exécution sécurisé).

acquire()

Acquiert le verrou créé par createLock().

release()

Relâche le verrou acquis par acquire().

setLevel(level)

Ajuste le seuil déclenchement du gestionnaire au niveau level. Les messages de gravité moindre que level sont alors ignorés. Ce seuil est fixé à NOTSET lors de la création d'un gestionnaire (ce qui signifie que tous les messages seront traités).

Voir Niveaux de journalisation pour la liste des niveaux.

Modifié dans la version 3.2: le paramètre level peut être une chaîne de caractères, comme 'INFO', en plus d'une constante entière comme INFO.

setFormatter(fmt)

Définit le Formatter du gestionnaire à fmt.

addFilter(filter)

Ajoute le filtre filter au gestionnaire.

removeFilter(filter)

Retire le filtre filter du gestionnaire.

filter(record)

Applique les filtres du gestionnaire à record et renvoie True si l'entrée doit être traitée. Les filtres sont appliqués l'un après l'autre, jusqu'à ce que l'un renvoie faux. Si aucun d'entre eux ne renvoie faux, l'entrée est enregistrée, sinon le gestionnaire ne traitera pas l'entrée.

flush()

Oblige toutes les entrées à être traitées. Cette fonction ne fait rien de spécial et doit être redéfinie par les sous-classes.

close()

Recycle toutes les ressources utilisées par le gestionnaire. Cette version ne fait rien de particulier mais elle retire le gestionnaire de la liste interne des gestionnaires à recycler à l'appel de shutdown(). Les sous-classes doivent appeler cette méthode depuis leur surcharge de close().

handle(record)

Traite ou non record selon les filtres ajoutés au gestionnaire. Un verrou sur l'E-S. est mis en place durant l'écriture effective.

handleError(record)

This method should be called from handlers when an exception is encountered during an emit() call. If the module-level attribute raiseExceptions is False, exceptions get silently ignored. This is what is mostly wanted for a logging system - most users will not care about errors in the logging system, they are more interested in application errors. You could, however, replace this with a custom handler if you wish. The specified record is the one which was being processed when the exception occurred. (The default value of raiseExceptions is True, as that is more useful during development).

format(record)

Formate record avec le formateur défini. S'il n'y en a pas, le formateur par défaut du module est utilisé.

emit(record)

Journalise record « pour de bon ». Cette version doit être redéfinie par les sous-classes et lève donc une NotImplementedError.

Avertissement

This method is called after a handler-level lock is acquired, which is released after this method returns. When you override this method, note that you should be careful when calling anything that invokes other parts of the logging API which might do locking, because that might result in a deadlock. Specifically:

  • Logging configuration APIs acquire the module-level lock, and then individual handler-level locks as those handlers are configured.

  • Many logging APIs lock the module-level lock. If such an API is called from this method, it could cause a deadlock if a configuration call is made on another thread, because that thread will try to acquire the module-level lock before the handler-level lock, whereas this thread tries to acquire the module-level lock after the handler-level lock (because in this method, the handler-level lock has already been acquired).

Les gestionnaires de la bibliothèque standard sont répertoriés dans logging.handlers.

Formateurs

class logging.Formatter(fmt=None, datefmt=None, style='%', validate=True, *, defaults=None)

Responsible for converting a LogRecord to an output string to be interpreted by a human or external system.

Paramètres:
  • fmt (str) -- A format string in the given style for the logged output as a whole. The possible mapping keys are drawn from the LogRecord object's LogRecord attributes. If not specified, '%(message)s' is used, which is just the logged message.

  • datefmt (str) -- A format string in the given style for the date/time portion of the logged output. If not specified, the default described in formatTime() is used.

  • style (str) -- Can be one of '%', '{' or '$' and determines how the format string will be merged with its data: using one of Formatage de chaines à la printf (%), str.format() ({) or string.Template ($). This only applies to fmt and datefmt (e.g. '%(message)s' versus '{message}'), not to the actual log messages passed to the logging methods. However, there are other ways to use {- and $-formatting for log messages.

  • validate (bool) -- If True (the default), incorrect or mismatched fmt and style will raise a ValueError; for example, logging.Formatter('%(asctime)s - %(message)s', style='{').

  • defaults (dict[str, Any]) -- A dictionary with default values to use in custom fields. For example, logging.Formatter('%(ip)s %(message)s', defaults={"ip": None})

Modifié dans la version 3.2: Added the style parameter.

Modifié dans la version 3.8: Added the validate parameter.

Modifié dans la version 3.10: Added the defaults parameter.

format(record)

The record's attribute dictionary is used as the operand to a string formatting operation. Returns the resulting string. Before formatting the dictionary, a couple of preparatory steps are carried out. The message attribute of the record is computed using msg % args. If the formatting string contains '(asctime)', formatTime() is called to format the event time. If there is exception information, it is formatted using formatException() and appended to the message. Note that the formatted exception information is cached in attribute exc_text. This is useful because the exception information can be pickled and sent across the wire, but you should be careful if you have more than one Formatter subclass which customizes the formatting of exception information. In this case, you will have to clear the cached value (by setting the exc_text attribute to None) after a formatter has done its formatting, so that the next formatter to handle the event doesn't use the cached value, but recalculates it afresh.

If stack information is available, it's appended after the exception information, using formatStack() to transform it if necessary.

formatTime(record, datefmt=None)

This method should be called from format() by a formatter which wants to make use of a formatted time. This method can be overridden in formatters to provide for any specific requirement, but the basic behavior is as follows: if datefmt (a string) is specified, it is used with time.strftime() to format the creation time of the record. Otherwise, the format '%Y-%m-%d %H:%M:%S,uuu' is used, where the uuu part is a millisecond value and the other letters are as per the time.strftime() documentation. An example time in this format is 2003-01-23 00:29:50,411. The resulting string is returned.

This function uses a user-configurable function to convert the creation time to a tuple. By default, time.localtime() is used; to change this for a particular formatter instance, set the converter attribute to a function with the same signature as time.localtime() or time.gmtime(). To change it for all formatters, for example if you want all logging times to be shown in GMT, set the converter attribute in the Formatter class.

Modifié dans la version 3.3: Previously, the default format was hard-coded as in this example: 2010-09-06 22:38:15,292 where the part before the comma is handled by a strptime format string ('%Y-%m-%d %H:%M:%S'), and the part after the comma is a millisecond value. Because strptime does not have a format placeholder for milliseconds, the millisecond value is appended using another format string, '%s,%03d' --- and both of these format strings have been hardcoded into this method. With the change, these strings are defined as class-level attributes which can be overridden at the instance level when desired. The names of the attributes are default_time_format (for the strptime format string) and default_msec_format (for appending the millisecond value).

Modifié dans la version 3.9: The default_msec_format can be None.

formatException(exc_info)

Formats the specified exception information (a standard exception tuple as returned by sys.exc_info()) as a string. This default implementation just uses traceback.print_exception(). The resulting string is returned.

formatStack(stack_info)

Formats the specified stack information (a string as returned by traceback.print_stack(), but with the last newline removed) as a string. This default implementation just returns the input value.

class logging.BufferingFormatter(linefmt=None)

A base formatter class suitable for subclassing when you want to format a number of records. You can pass a Formatter instance which you want to use to format each line (that corresponds to a single record). If not specified, the default formatter (which just outputs the event message) is used as the line formatter.

formatHeader(records)

Return a header for a list of records. The base implementation just returns the empty string. You will need to override this method if you want specific behaviour, e.g. to show the count of records, a title or a separator line.

formatFooter(records)

Return a footer for a list of records. The base implementation just returns the empty string. You will need to override this method if you want specific behaviour, e.g. to show the count of records or a separator line.

format(records)

Return formatted text for a list of records. The base implementation just returns the empty string if there are no records; otherwise, it returns the concatenation of the header, each record formatted with the line formatter, and the footer.

Filtres

Filters can be used by Handlers and Loggers for more sophisticated filtering than is provided by levels. The base filter class only allows events which are below a certain point in the logger hierarchy. For example, a filter initialized with 'A.B' will allow events logged by loggers 'A.B', 'A.B.C', 'A.B.C.D', 'A.B.D' etc. but not 'A.BB', 'B.A.B' etc. If initialized with the empty string, all events are passed.

class logging.Filter(name='')

Returns an instance of the Filter class. If name is specified, it names a logger which, together with its children, will have its events allowed through the filter. If name is the empty string, allows every event.

filter(record)

Is the specified record to be logged? Returns false for no, true for yes. Filters can either modify log records in-place or return a completely different record instance which will replace the original log record in any future processing of the event.

Note that filters attached to handlers are consulted before an event is emitted by the handler, whereas filters attached to loggers are consulted whenever an event is logged (using debug(), info(), etc.), before sending an event to handlers. This means that events which have been generated by descendant loggers will not be filtered by a logger's filter setting, unless the filter has also been applied to those descendant loggers.

You don't actually need to subclass Filter: you can pass any instance which has a filter method with the same semantics.

Modifié dans la version 3.2: You don't need to create specialized Filter classes, or use other classes with a filter method: you can use a function (or other callable) as a filter. The filtering logic will check to see if the filter object has a filter attribute: if it does, it's assumed to be a Filter and its filter() method is called. Otherwise, it's assumed to be a callable and called with the record as the single parameter. The returned value should conform to that returned by filter().

Modifié dans la version 3.12: You can now return a LogRecord instance from filters to replace the log record rather than modifying it in place. This allows filters attached to a Handler to modify the log record before it is emitted, without having side effects on other handlers.

Although filters are used primarily to filter records based on more sophisticated criteria than levels, they get to see every record which is processed by the handler or logger they're attached to: this can be useful if you want to do things like counting how many records were processed by a particular logger or handler, or adding, changing or removing attributes in the LogRecord being processed. Obviously changing the LogRecord needs to be done with some care, but it does allow the injection of contextual information into logs (see Utilisation de filtres pour transmettre des informations contextuelles).

Objets LogRecord

LogRecord instances are created automatically by the Logger every time something is logged, and can be created manually via makeLogRecord() (for example, from a pickled event received over the wire).

class logging.LogRecord(name, level, pathname, lineno, msg, args, exc_info, func=None, sinfo=None)

Contains all the information pertinent to the event being logged.

The primary information is passed in msg and args, which are combined using msg % args to create the message attribute of the record.

Paramètres:
  • name (str) -- The name of the logger used to log the event represented by this LogRecord. Note that the logger name in the LogRecord will always have this value, even though it may be emitted by a handler attached to a different (ancestor) logger.

  • level (int) -- The numeric level of the logging event (such as 10 for DEBUG, 20 for INFO, etc). Note that this is converted to two attributes of the LogRecord: levelno for the numeric value and levelname for the corresponding level name.

  • pathname (str) -- The full string path of the source file where the logging call was made.

  • lineno (int) -- The line number in the source file where the logging call was made.

  • msg (Any) -- The event description message, which can be a %-format string with placeholders for variable data, or an arbitrary object (see Utilisation d'objets arbitraires comme messages).

  • args (tuple | dict[str, Any]) -- Variable data to merge into the msg argument to obtain the event description.

  • exc_info (tuple[type[BaseException], BaseException, types.TracebackType] | None) -- An exception tuple with the current exception information, as returned by sys.exc_info(), or None if no exception information is available.

  • func (str | None) -- The name of the function or method from which the logging call was invoked.

  • sinfo (str | None) -- A text string representing stack information from the base of the stack in the current thread, up to the logging call.

getMessage()

Returns the message for this LogRecord instance after merging any user-supplied arguments with the message. If the user-supplied message argument to the logging call is not a string, str() is called on it to convert it to a string. This allows use of user-defined classes as messages, whose __str__ method can return the actual format string to be used.

Modifié dans la version 3.2: The creation of a LogRecord has been made more configurable by providing a factory which is used to create the record. The factory can be set using getLogRecordFactory() and setLogRecordFactory() (see this for the factory's signature).

This functionality can be used to inject your own values into a LogRecord at creation time. You can use the following pattern:

old_factory = logging.getLogRecordFactory()

def record_factory(*args, **kwargs):
    record = old_factory(*args, **kwargs)
    record.custom_attribute = 0xdecafbad
    return record

logging.setLogRecordFactory(record_factory)

With this pattern, multiple factories could be chained, and as long as they don't overwrite each other's attributes or unintentionally overwrite the standard attributes listed above, there should be no surprises.

LogRecord attributes

The LogRecord has a number of attributes, most of which are derived from the parameters to the constructor. (Note that the names do not always correspond exactly between the LogRecord constructor parameters and the LogRecord attributes.) These attributes can be used to merge data from the record into the format string. The following table lists (in alphabetical order) the attribute names, their meanings and the corresponding placeholder in a %-style format string.

If you are using {}-formatting (str.format()), you can use {attrname} as the placeholder in the format string. If you are using $-formatting (string.Template), use the form ${attrname}. In both cases, of course, replace attrname with the actual attribute name you want to use.

In the case of {}-formatting, you can specify formatting flags by placing them after the attribute name, separated from it with a colon. For example: a placeholder of {msecs:03.0f} would format a millisecond value of 4 as 004. Refer to the str.format() documentation for full details on the options available to you.

Attribute name

Format

Description

args

You shouldn't need to format this yourself.

The tuple of arguments merged into msg to produce message, or a dict whose values are used for the merge (when there is only one argument, and it is a dictionary).

asctime

%(asctime)s

Human-readable time when the LogRecord was created. By default this is of the form '2003-07-08 16:49:45,896' (the numbers after the comma are millisecond portion of the time).

created

%(created)f

Time when the LogRecord was created (as returned by time.time_ns() / 1e9).

exc_info

You shouldn't need to format this yourself.

Exception tuple (à la sys.exc_info) or, if no exception has occurred, None.

filename

%(filename)s

Filename portion of pathname.

funcName

%(funcName)s

Name of function containing the logging call.

levelname

%(levelname)s

Text logging level for the message ('DEBUG', 'INFO', 'WARNING', 'ERROR', 'CRITICAL').

levelno

%(levelno)s

Numeric logging level for the message (DEBUG, INFO, WARNING, ERROR, CRITICAL).

lineno

%(lineno)d

Source line number where the logging call was issued (if available).

message

%(message)s

The logged message, computed as msg % args. This is set when Formatter.format() is invoked.

module

%(module)s

Module (name portion of filename).

msecs

%(msecs)d

Millisecond portion of the time when the LogRecord was created.

msg

You shouldn't need to format this yourself.

The format string passed in the original logging call. Merged with args to produce message, or an arbitrary object (see Utilisation d'objets arbitraires comme messages).

nom

%(name)s

Name of the logger used to log the call.

pathname

%(pathname)s

Full pathname of the source file where the logging call was issued (if available).

process

%(process)d

Process ID (if available).

processName

%(processName)s

Process name (if available).

relativeCreated

%(relativeCreated)d

Time in milliseconds when the LogRecord was created, relative to the time the logging module was loaded.

stack_info

You shouldn't need to format this yourself.

Stack frame information (where available) from the bottom of the stack in the current thread, up to and including the stack frame of the logging call which resulted in the creation of this record.

thread

%(thread)d

Thread ID (if available).

threadName

%(threadName)s

Thread name (if available).

taskName

%(taskName)s

asyncio.Task name (if available).

Modifié dans la version 3.1: processName was added.

Modifié dans la version 3.12: taskName was added.

LoggerAdapter Objects

LoggerAdapter instances are used to conveniently pass contextual information into logging calls. For a usage example, see the section on adding contextual information to your logging output.

class logging.LoggerAdapter(logger, extra, merge_extra=False)

Returns an instance of LoggerAdapter initialized with an underlying Logger instance, a dict-like object (extra), and a boolean (merge_extra) indicating whether or not the extra argument of individual log calls should be merged with the LoggerAdapter extra. The default behavior is to ignore the extra argument of individual log calls and only use the one of the LoggerAdapter instance

process(msg, kwargs)

Modifies the message and/or keyword arguments passed to a logging call in order to insert contextual information. This implementation takes the object passed as extra to the constructor and adds it to kwargs using key 'extra'. The return value is a (msg, kwargs) tuple which has the (possibly modified) versions of the arguments passed in.

manager

Delegates to the underlying manager on logger.

_log

Delegates to the underlying _log() method on logger.

In addition to the above, LoggerAdapter supports the following methods of Logger: debug(), info(), warning(), error(), exception(), critical(), log(), isEnabledFor(), getEffectiveLevel(), setLevel() and hasHandlers(). These methods have the same signatures as their counterparts in Logger, so you can use the two types of instances interchangeably.

Modifié dans la version 3.2: The isEnabledFor(), getEffectiveLevel(), setLevel() and hasHandlers() methods were added to LoggerAdapter. These methods delegate to the underlying logger.

Modifié dans la version 3.6: Attribute manager and method _log() were added, which delegate to the underlying logger and allow adapters to be nested.

Modifié dans la version 3.13: The merge_extra argument was added.

Thread Safety

The logging module is intended to be thread-safe without any special work needing to be done by its clients. It achieves this though using threading locks; there is one lock to serialize access to the module's shared data, and each handler also creates a lock to serialize access to its underlying I/O.

If you are implementing asynchronous signal handlers using the signal module, you may not be able to use logging from within such handlers. This is because lock implementations in the threading module are not always re-entrant, and so cannot be invoked from such signal handlers.

Fonctions de niveau module

In addition to the classes described above, there are a number of module-level functions.

logging.getLogger(name=None)

Return a logger with the specified name or, if name is None, return the root logger of the hierarchy. If specified, the name is typically a dot-separated hierarchical name like 'a', 'a.b' or 'a.b.c.d'. Choice of these names is entirely up to the developer who is using logging, though it is recommended that __name__ be used unless you have a specific reason for not doing that, as mentioned in Enregistreurs.

All calls to this function with a given name return the same logger instance. This means that logger instances never need to be passed between different parts of an application.

logging.getLoggerClass()

Return either the standard Logger class, or the last class passed to setLoggerClass(). This function may be called from within a new class definition, to ensure that installing a customized Logger class will not undo customizations already applied by other code. For example:

class MyLogger(logging.getLoggerClass()):
    # ... override behaviour here
logging.getLogRecordFactory()

Return a callable which is used to create a LogRecord.

Ajouté dans la version 3.2: This function has been provided, along with setLogRecordFactory(), to allow developers more control over how the LogRecord representing a logging event is constructed.

See setLogRecordFactory() for more information about the how the factory is called.

logging.debug(msg, *args, **kwargs)

This is a convenience function that calls Logger.debug(), on the root logger. The handling of the arguments is in every way identical to what is described in that method.

The only difference is that if the root logger has no handlers, then basicConfig() is called, prior to calling debug on the root logger.

For very short scripts or quick demonstrations of logging facilities, debug and the other module-level functions may be convenient. However, most programs will want to carefully and explicitly control the logging configuration, and should therefore prefer creating a module-level logger and calling Logger.debug() (or other level-specific methods) on it, as described at the beginnning of this documentation.

logging.info(msg, *args, **kwargs)

Logs a message with level INFO on the root logger. The arguments and behavior are otherwise the same as for debug().

logging.warning(msg, *args, **kwargs)

Logs a message with level WARNING on the root logger. The arguments and behavior are otherwise the same as for debug().

Note

There is an obsolete function warn which is functionally identical to warning. As warn is deprecated, please do not use it - use warning instead.

logging.error(msg, *args, **kwargs)

Logs a message with level ERROR on the root logger. The arguments and behavior are otherwise the same as for debug().

logging.critical(msg, *args, **kwargs)

Logs a message with level CRITICAL on the root logger. The arguments and behavior are otherwise the same as for debug().

logging.exception(msg, *args, **kwargs)

Logs a message with level ERROR on the root logger. The arguments and behavior are otherwise the same as for debug(). Exception info is added to the logging message. This function should only be called from an exception handler.

logging.log(level, msg, *args, **kwargs)

Logs a message with level level on the root logger. The arguments and behavior are otherwise the same as for debug().

logging.disable(level=CRITICAL)

Provides an overriding level level for all loggers which takes precedence over the logger's own level. When the need arises to temporarily throttle logging output down across the whole application, this function can be useful. Its effect is to disable all logging calls of severity level and below, so that if you call it with a value of INFO, then all INFO and DEBUG events would be discarded, whereas those of severity WARNING and above would be processed according to the logger's effective level. If logging.disable(logging.NOTSET) is called, it effectively removes this overriding level, so that logging output again depends on the effective levels of individual loggers.

Note that if you have defined any custom logging level higher than CRITICAL (this is not recommended), you won't be able to rely on the default value for the level parameter, but will have to explicitly supply a suitable value.

Modifié dans la version 3.7: The level parameter was defaulted to level CRITICAL. See bpo-28524 for more information about this change.

logging.addLevelName(level, levelName)

Associates level level with text levelName in an internal dictionary, which is used to map numeric levels to a textual representation, for example when a Formatter formats a message. This function can also be used to define your own levels. The only constraints are that all levels used must be registered using this function, levels should be positive integers and they should increase in increasing order of severity.

Note

If you are thinking of defining your own levels, please see the section on Niveaux personnalisés.

logging.getLevelNamesMapping()

Returns a mapping from level names to their corresponding logging levels. For example, the string "CRITICAL" maps to CRITICAL. The returned mapping is copied from an internal mapping on each call to this function.

Ajouté dans la version 3.11.

logging.getLevelName(level)

Returns the textual or numeric representation of logging level level.

If level is one of the predefined levels CRITICAL, ERROR, WARNING, INFO or DEBUG then you get the corresponding string. If you have associated levels with names using addLevelName() then the name you have associated with level is returned. If a numeric value corresponding to one of the defined levels is passed in, the corresponding string representation is returned.

The level parameter also accepts a string representation of the level such as 'INFO'. In such cases, this functions returns the corresponding numeric value of the level.

If no matching numeric or string value is passed in, the string 'Level %s' % level is returned.

Note

Levels are internally integers (as they need to be compared in the logging logic). This function is used to convert between an integer level and the level name displayed in the formatted log output by means of the %(levelname)s format specifier (see LogRecord attributes), and vice versa.

Modifié dans la version 3.4: In Python versions earlier than 3.4, this function could also be passed a text level, and would return the corresponding numeric value of the level. This undocumented behaviour was considered a mistake, and was removed in Python 3.4, but reinstated in 3.4.2 due to retain backward compatibility.

logging.getHandlerByName(name)

Returns a handler with the specified name, or None if there is no handler with that name.

Ajouté dans la version 3.12.

logging.getHandlerNames()

Returns an immutable set of all known handler names.

Ajouté dans la version 3.12.

logging.makeLogRecord(attrdict)

Creates and returns a new LogRecord instance whose attributes are defined by attrdict. This function is useful for taking a pickled LogRecord attribute dictionary, sent over a socket, and reconstituting it as a LogRecord instance at the receiving end.

logging.basicConfig(**kwargs)

Does basic configuration for the logging system by creating a StreamHandler with a default Formatter and adding it to the root logger. The functions debug(), info(), warning(), error() and critical() will call basicConfig() automatically if no handlers are defined for the root logger.

This function does nothing if the root logger already has handlers configured, unless the keyword argument force is set to True.

Note

This function should be called from the main thread before other threads are started. In versions of Python prior to 2.7.1 and 3.2, if this function is called from multiple threads, it is possible (in rare circumstances) that a handler will be added to the root logger more than once, leading to unexpected results such as messages being duplicated in the log.

The following keyword arguments are supported.

Format

Description

filename

Specifies that a FileHandler be created, using the specified filename, rather than a StreamHandler.

filemode

If filename is specified, open the file in this mode. Defaults to 'a'.

format

Use the specified format string for the handler. Defaults to attributes levelname, name and message separated by colons.

datefmt

Use the specified date/time format, as accepted by time.strftime().

style

If format is specified, use this style for the format string. One of '%', '{' or '$' for printf-style, str.format() or string.Template respectively. Defaults to '%'.

level

Set the root logger level to the specified level.

stream

Use the specified stream to initialize the StreamHandler. Note that this argument is incompatible with filename - if both are present, a ValueError is raised.

handlers

If specified, this should be an iterable of already created handlers to add to the root logger. Any handlers which don't already have a formatter set will be assigned the default formatter created in this function. Note that this argument is incompatible with filename or stream - if both are present, a ValueError is raised.

force

If this keyword argument is specified as true, any existing handlers attached to the root logger are removed and closed, before carrying out the configuration as specified by the other arguments.

encoding

If this keyword argument is specified along with filename, its value is used when the FileHandler is created, and thus used when opening the output file.

errors

If this keyword argument is specified along with filename, its value is used when the FileHandler is created, and thus used when opening the output file. If not specified, the value 'backslashreplace' is used. Note that if None is specified, it will be passed as such to open(), which means that it will be treated the same as passing 'errors'.

Modifié dans la version 3.2: The style argument was added.

Modifié dans la version 3.3: The handlers argument was added. Additional checks were added to catch situations where incompatible arguments are specified (e.g. handlers together with stream or filename, or stream together with filename).

Modifié dans la version 3.8: The force argument was added.

Modifié dans la version 3.9: The encoding and errors arguments were added.

logging.shutdown()

Informs the logging system to perform an orderly shutdown by flushing and closing all handlers. This should be called at application exit and no further use of the logging system should be made after this call.

When the logging module is imported, it registers this function as an exit handler (see atexit), so normally there's no need to do that manually.

logging.setLoggerClass(klass)

Tells the logging system to use the class klass when instantiating a logger. The class should define __init__() such that only a name argument is required, and the __init__() should call Logger.__init__(). This function is typically called before any loggers are instantiated by applications which need to use custom logger behavior. After this call, as at any other time, do not instantiate loggers directly using the subclass: continue to use the logging.getLogger() API to get your loggers.

logging.setLogRecordFactory(factory)

Set a callable which is used to create a LogRecord.

Paramètres:

factory -- The factory callable to be used to instantiate a log record.

Ajouté dans la version 3.2: This function has been provided, along with getLogRecordFactory(), to allow developers more control over how the LogRecord representing a logging event is constructed.

The factory has the following signature:

factory(name, level, fn, lno, msg, args, exc_info, func=None, sinfo=None, **kwargs)

nom:

The logger name.

level:

The logging level (numeric).

fn:

The full pathname of the file where the logging call was made.

lno:

The line number in the file where the logging call was made.

msg:

The logging message.

args:

The arguments for the logging message.

exc_info:

An exception tuple, or None.

func:

The name of the function or method which invoked the logging call.

sinfo:

A stack traceback such as is provided by traceback.print_stack(), showing the call hierarchy.

kwargs:

Additional keyword arguments.

Module-Level Attributes

logging.lastResort

A "handler of last resort" is available through this attribute. This is a StreamHandler writing to sys.stderr with a level of WARNING, and is used to handle logging events in the absence of any logging configuration. The end result is to just print the message to sys.stderr. This replaces the earlier error message saying that "no handlers could be found for logger XYZ". If you need the earlier behaviour for some reason, lastResort can be set to None.

Ajouté dans la version 3.2.

logging.raiseExceptions

Used to see if exceptions during handling should be propagated.

Default: True.

If raiseExceptions is False, exceptions get silently ignored. This is what is mostly wanted for a logging system - most users will not care about errors in the logging system, they are more interested in application errors.

Integration with the warnings module

The captureWarnings() function can be used to integrate logging with the warnings module.

logging.captureWarnings(capture)

This function is used to turn the capture of warnings by logging on and off.

If capture is True, warnings issued by the warnings module will be redirected to the logging system. Specifically, a warning will be formatted using warnings.formatwarning() and the resulting string logged to a logger named 'py.warnings' with a severity of WARNING.

If capture is False, the redirection of warnings to the logging system will stop, and warnings will be redirected to their original destinations (i.e. those in effect before captureWarnings(True) was called).

Voir aussi

Module logging.config

API de configuration pour le module de journalisation.

Module logging.handlers

Gestionnaires utiles inclus avec le module de journalisation.

PEP 282 - A Logging System

The proposal which described this feature for inclusion in the Python standard library.

Original Python logging package

This is the original source for the logging package. The version of the package available from this site is suitable for use with Python 1.5.2, 2.1.x and 2.2.x, which do not include the logging package in the standard library.