Web Utils

Web development utils classes and functions.
Current status: Under development and works in my project.
Find me in http://github.com/winkidney/web_utils.

Summaries are listed below.Documents will be written later.

##Install

pip install "web_utils[forms,security,sqlalchemy]" # to install all requirements.

##Summaries

• code_loader - load code from str or unicode object and save it to cache or database
• forms - some custom validator for wtfomrs
• json_form - wrapper for jsonschema to validate json string like wtfomrs.
• Security - password generator by bcrypt.Simple url sign generator.
• _sqlalchemy - DBFieldConverter for convert alchemy's model instance to dict like string with white list support.
• _pyramid - tools form pyramid web framework
• extra - some other utils.

##web_utils.code_loader

Create a Code loader instance to load and save code from Storage or str object.
Return a python module object that you can run any code within it.

• Methods
  • init(name, storage_backend=DummyStorageBackend, cache_backend=DummyCacheBackend) Create a loader with given storage_backend and cache_backend.
  • create_module(fullname, code_script, save_key=None) if save_key is given, use it instead of fullname as accees_key passed to storage backend's set method.
  • save(mod, cached=False, **kwargs) save a module object into storage backend and cache backend(optional). **kwargs will be passed to backend's set method.
  • load(fullname, save_key=None, **kwargs) Load a module by its name(if save_key is given, use it instead).

##web_utils.forms WTFroms's custom field and other validator

• validators
  • uuid_validator
  • TimeAfterNow validator - return True if current datetime-field later than now.
  • MobilePhone validataor - Check if a number sequence is china's phone number
• custom field
  • UnixTimeField
  • TextArrayField - convert list like a, b, c ,1 ,3 ,3 to python list.
  • JsonField - Check if the text is json format string and convert it into python data(with json.loads)
  • IntArrayField - Convert list like 1,2,3,4,5 into python list(consists of python's int object)
• utils
  • json2form - convert python dict into MultiDict which can be read by wtforms.

##web_utils.json_forms Validate json string or dict object in wtforms's way. Just inherit JsonForm class and call validate method to do validation. Validation by JsonSchema , Validation Quick Start. Example listed below:

class NewPMSSchema(JsonForm):
    schema = {
        "type": "object",
        "properties": {
            "to_uid": {
                "type": "number",
            },
            "content": {
                "type": "string",
            },
            "test": {
                "type": "object",
                "properties": {
                    "test1": {
                        "type": "integer",
                    }
                }
            },
        },
        "required": ['to_uid', 'content'],
    }

form = NewPMSSchema({'to_uid': 'a', 'content': 1})
# result
result = form.validate()
# errors
if not result:
    print form.errors

##web_utils.security Not recommend to use it.

##web_utils._sqlalchemy A sqlalchemy model to json data converter with white list and custom converter support.

###DBFCMixin Just inherit it in your sqlalchemy model and call as_dict method to output python dict.
class._default_output is required.
Example listed below:

class APIStorage(Base, DBFCMixin, StorageBackendMixin):

    __tablename__ = "api_storage"
    # white list
    _default_output = ('id', 'category', 'resource_name', 'document')

    id = Column(Integer, primary_key=True, autoincrement=True)
    category = Column(Text, nullable=False)
    resource_name = Column(Text, nullable=False)
    body = Column(Text, nullable=False)
    ctime = Column(DateTime, default=datetime.datetime.utcnow)

# call as_dict method
model_instance = dbsession.query(APIStorage).first()
model_instance.as_dict(pure=False)  #False is default , with white list support.

###DBFieldConverter Low level API for DBFCMixin
Quick example:

from sqlalchemy.ext.declarative import declarative_base

Base = declarative_base()

class APIStorage(Base):

    __tablename__ = "api_storage"
    _default_output = ('id', 'category', 'resource_name', 'document')

    id = Column(Integer, primary_key=True, autoincrement=True)
    category = Column(Text, nullable=False)
    resource_name = Column(Text, nullable=False)
    body = Column(Text, nullable=False)
    ctime = Column(DateTime, default=datetime.datetime.utcnow)

# convert and output

model_instance = dbsession.query(APIStorage).first()

converter = DBFC(model_instance, model_instance._default_output)

converter.as_dict()

# as list
converter.as_list()

# without white list(output all data field)
converter.as_dict(pure=True)

####Register a Converter DBFieldConverter supports Converter by sqlalchemy's field type.

from sqlalchemy import Text
DBFieldConverter.register(Text)

will register Text type in DBFC. Note: without registered, DBFieldConverter will print warning information in console.

DBFieldConverter.register(Text, lambda x:x[-1])

will output the last char of the input field which type is Text.

You can also pass registry to DBFieldConverter __init__ method to specify field type converter. For example:

from sqlalchemy import Text
converter = DBFC(model_instance, model_instance._default_output, registry={Text: lambda x: x[-1]})

Note: This registry will not affect other instance's registry of DBFieldConverter since it's instance registry.
Otherwise, DBFC.register register class converter in it's registry, every instance will be affected.

###form2model Easy set wtfomrs Form data to sqlalchemy's model field, name by name.

form2model(form, model_instance, exclude=None)

##web_utils.extra Some utils about argument parse ,datetime format.

• GetSingleArgument
  • bool(cls, value, default=False) default value support, will not return None.
  • integer(cls, value, default, nmin=None, nmax=None) - parse integer from string, return default value if not in given range.
  • string(cls, value, default='') - default value support
• format_timestamp
  from tornado core, to format timestamp

  >>> format_timestamp(1359312200)
    'Sun, 27 Jan 2013 18:43:20 GMT'

• AttrDict
  Simple wrapper for attr dict

  d = AttrDict({'a': 1})
  d.a # output 1
  

  Many problem when use it, pay attention before you really know what you are doing.