form field clean now force-stringifies value

README.md

@@ -6,11 +6,11 @@ This module allows creating simple interfaces to forms/payloads for use in HTTP
 
 A form is defined by its fields. 
 
-A field is defined by an **alias** (for internal use) and the field's **name**, which is the parameter name in the payload sent during a `POST` request, and which typically corresponds to the `name` attribute of a `<select>` or `<input>` HTML tag.
+A field is defined by its _name_, which is the parameter name in the payload sent during a `POST` request, and which typically corresponds to the `name` attribute of a `<select>` or `<input>` HTML tag.
 
-Optionally, a field can have a **default** value, value **options** (as `<select>` tags do), and may be declared **required** (i.e. non-optional).
+Optionally, a field can have an _alias_ (for internal use), a _default_ value, value _options_ (as `<select>` tags do), and may be declared _required_.
 
-A form definition in YAML format will consist of the field aliases as top-level keys, and the corresponding fields' definitions as key-value-pairs below them.
+A form definition in YAML format will consist of the field names as top-level keys, and either nothing/`null` or the corresponding fields' definitions as key-value-pairs below them.
 
 ## Example
 
@@ -18,19 +18,16 @@ A form definition in YAML format will consist of the field aliases as top-level
 ```yaml
 # definition.yaml
 
-field1:
-  name: payload_param_1
-field2:
-  name: payload_param_2
-  default: foo
+way_too_long_field_name:
+  alias: short_name
+foo: 
 choice_field:
-  name: payload_param_3
   options:
     value1: text for option 1
     value2: text for option 2
   default: value1
 mandatory_field:
-  name: payload_param_0
+  alias: special
   required: true
 ```
 
@@ -38,10 +35,11 @@ mandatory_field:
 ```
 >>> from yamlhttpforms import load_form
 >>> form_interface = load_form('definition.yaml')
->>> form_interface.get_payload(field1='abc', field2='bar', mandatory_field='420')
-{'payload_param_1': 'abc', 'payload_param_2': 'bar', 'payload_param_3': 'value1', 'payload_param_0': '420'}
->>> form_interface.get_payload(field1='abc', choice_field='baz', mandatory_field='420')
+>>> form_interface.get_payload(short_name='abc', foo='bar', special='420')
+{'way_too_long_field_name': 'abc', 'foo': 'bar', 'choice_field': 'value1', 'mandatory_field': '420'}
+
+>>> form_interface.get_payload(short_name='abc', choice_field='baz', special='420')
 Traceback (most recent call last):
   ...
-ValueError: "baz" is not a valid option for <SelectField: name="payload_param_3", default="value1", options={'value1': 'text for option 1', 'value2': 'text for option 2'}>
+ValueError: "baz" is not a valid option for <SelectField: name="choice_field", default="value1", options={'value1': 'text for option 1', 'value2': 'text for option 2'}>
 ```

requirements/dev.txt

@@ -1,3 +1,2 @@
--r aio.txt
--r req.txt
+-r full.txt
 coverage

requirements/full.txt

@@ -0,0 +1,3 @@
+-r aio.txt
+-r req.txt
+-r html.txt

requirements/html.txt

@@ -0,0 +1,2 @@
+-r common.txt
+beautifulsoup4

setup.cfg

@@ -1,6 +1,6 @@
 [metadata]
 name = yamlhttpforms
-version = 0.0.2
+version = 0.1.2
 author = Daniil F.
 author_email = mail@placeholder123.to
 description = HTTP forms defined in YAML
@@ -26,6 +26,9 @@ req =
     requests
 aio =
     aiohttp
+    webutils-df @ git+https://git.fajnberg.de/daniil/webutils-df.git
+html =
+    beautifulsoup4
 
 [options.packages.find]
 where = src

src/yamlhttpforms/form.py

@@ -2,8 +2,9 @@ from importlib import import_module
 from typing import Dict, Callable, Union, Optional, Any, TYPE_CHECKING
 
 if TYPE_CHECKING:
-    from aiohttp import ClientSession as AioSession, ClientResponse as AioResponse
-    from requests import Session as ReqSession, Response as ReqResponse
+    from aiohttp import ClientSession as AioSession
+    from requests import Session as ReqSession
+    from bs4.element import Tag as BS4Tag
 
 from .utils import PathT, yaml_overload
 
@@ -16,20 +17,10 @@ OptionsT = Dict[str, str]
 class FormField:
     def __init__(self, name: str, default: DefaultInitT = None, options: OptionsT = None, required: bool = False):
         self.name: str = name
-        self._default: Union[str, CallableDefaultT, None]
-        if isinstance(default, dict):
-            try:
-                module, function = default['module'], default['function']
-            except KeyError:
-                raise TypeError(f"Default for field '{name}' is invalid. The default must be either a string or a "
-                                f"dictionary with the special keys 'module' and 'function'.")
-            obj = import_module(module)
-            for attr in function.split('.'):
-                obj = getattr(obj, attr)
-            self._default = obj
-        else:
-            self._default = default
-        self.options: Optional[OptionsT] = options
+        self.default = default
+        self.options: Optional[OptionsT] = None
+        if options is not None:
+            self.options = {str(k): str(v) for k, v in options.items()}
         self.required: bool = required
 
     def __repr__(self) -> str:
@@ -50,10 +41,28 @@ class FormField:
     def default(self) -> Optional[str]:
         return self._default() if callable(self._default) else self._default
 
+    @default.setter
+    def default(self, default: DefaultInitT) -> None:
+        if isinstance(default, dict):
+            try:
+                module, function = default['module'], default['function']
+            except KeyError:
+                raise TypeError(f"Default for field '{self.name}' is invalid. The default must be either a string or "
+                                f"`None` or a dictionary with the special keys 'module' and 'function'.")
+            obj = import_module(module)
+            for attr in function.split('.'):
+                obj = getattr(obj, attr)
+            self._default = obj
+        elif default is None:
+            self._default = None
+        else:
+            self._default = str(default)
+
     def valid_option(self, option: str) -> bool:
         return self.options is None or option in self.options.keys() or option in self.options.values()
 
-    def clean(self, value: str) -> str:
+    def clean(self, value: Any) -> str:
+        value = str(value)
         if self.options is None or value in self.options.keys():
             return value
         # Try to find an unambiguous match in the visible option texts:
@@ -67,6 +76,17 @@ class FormField:
             return key
         raise ValueError(f'"{value}" is not a valid option for {self}')
 
+    def check_with_html(self, field_tag: 'BS4Tag', check_defaults: bool = True) -> None:
+        from .html import check_field_interface
+        check_field_interface(field_tag, self, check_defaults=check_defaults)
+
+    def get_value_from_html_form(self, form_tag: 'BS4Tag') -> Optional[str]:
+        from .html import get_field_value
+        try:
+            return get_field_value(form_tag, self.name)
+        except AttributeError:
+            return None
+
 
 class Form:
 
@@ -155,7 +175,7 @@ class Form:
                 payload[field.name] = field.default
         return payload
 
-    async def post_aio(self, _aiohttp_session_obj: 'AioSession' = None, **kwargs: str) -> 'AioResponse':
+    async def post_aio(self, _aiohttp_session_obj: 'AioSession' = None, **kwargs: str) -> str:
         """
         Uses `aiohttp` to perform a POST request to `.url` with the form's payload generated using `kwargs`.
 
@@ -166,20 +186,20 @@ class Form:
                 Passed directly into `.get_payload`.
 
         Returns:
-            The `aiohttp.ClientResponse` object from the request.
+            The response text from the request.
         """
         if self.url is None:
             raise AttributeError("`url` attribute not set")
-        from aiohttp import ClientSession, ClientResponse
-        from webutils import in_async_session
+        from aiohttp import ClientSession
+        from webutils_df import in_async_session
 
         @in_async_session
-        async def post(url: str, data: dict, session: ClientSession = None) -> ClientResponse:
+        async def post(url: str, data: dict, session: ClientSession = None) -> str:
             async with session.post(url, data=data) as response:
-                return response
+                return await response.text()
         return await post(self.url, self.get_payload(**kwargs), session=_aiohttp_session_obj)
 
-    def post_req(self, _requests_session_obj: 'ReqSession' = None, **kwargs: str) -> 'ReqResponse':
+    def post_req(self, _requests_session_obj: 'ReqSession' = None, **kwargs: str) -> str:
         """
         Uses `requests` to perform a POST request to `.url` with the form's payload generated using `kwargs`.
 
@@ -190,14 +210,22 @@ class Form:
                 Passed directly into `.get_payload`.
 
         Returns:
-            The `requests.Response` object from the request.
+            The response text from the request.
         """
         if self.url is None:
             raise AttributeError("`url` attribute not set")
         if _requests_session_obj is not None:
-            return _requests_session_obj.post(self.url, data=self.get_payload(**kwargs))
+            return _requests_session_obj.post(self.url, data=self.get_payload(**kwargs)).text
         from requests import post
-        return post(self.url, data=self.get_payload(**kwargs))
+        return post(self.url, data=self.get_payload(**kwargs)).text
+
+    def check_with_html(self, form_tag: 'BS4Tag', check_defaults: bool = True) -> None:
+        from .html import check_form_interface
+        check_form_interface(form_tag, self, check_defaults=check_defaults)
+
+    def get_values_from_html_form(self, form_tag: 'BS4Tag', required_fields_only: bool = True) -> Dict[str, str]:
+        from .html import get_field_values
+        return get_field_values(form_tag, self, required_fields_only=required_fields_only)
 
 
 def load_form(*def_paths: PathT, dir_sort_key: Callable[[PathT], Any] = None, dir_sort_reverse: bool = False,

src/yamlhttpforms/html.py

@@ -0,0 +1,174 @@
+import sys
+from typing import Dict, TYPE_CHECKING
+
+from bs4.element import Tag
+
+if TYPE_CHECKING:
+    from .form import FormField, Form
+
+
+INPUT, SELECT, OPTION = 'input', 'select', 'option'
+NAME, VALUE, SELECTED = 'name', 'value', 'selected'
+
+NON_PRINTABLE_TO_NONE = {code: None for code in range(sys.maxunicode + 1) if not chr(code).isprintable()}
+
+
+class WrongInterface(Exception):
+    pass
+
+
+class FieldInterfaceWrong(WrongInterface):
+    pass
+
+
+class SelectOptionsWrong(FieldInterfaceWrong):
+    pass
+
+
+class IncompleteForm(WrongInterface):
+    pass
+
+
+class UnknownField(WrongInterface):
+    pass
+
+
+def printable_only(string: str) -> str:
+    return string.translate(NON_PRINTABLE_TO_NONE)
+
+
+def check_select_field_options(field_tag: Tag, field_interface: 'FormField', check_defaults: bool = True) -> None:
+    """
+    Compares the `options` and `default` attributes of a `'FormField'` object with the options of its HTML counterpart.
+
+    Args:
+        field_tag: The `bs4.Tag` object representing the <select> HTML tag
+        field_interface: The `'FormField'` to check against the HTML tag
+        check_defaults (optional): If set to `False`, pre-selected options are not compared to the defined `default`
+
+    Raises:
+        `SelectOptionsWrong`
+            if the `options` dictionary is not entirely equal to a dictionary with the <option> tag `values` as keys
+            and their visible text as values
+        `FieldInterfaceWrong`
+            if the `default` is not equal to the value of the <option> tag which has the `selected` attribute
+    """
+    html_options = {(tag[VALUE], printable_only(tag.get_text(strip=True))) for tag in field_tag.find_all(OPTION)}
+    interface_options = set(field_interface.options.items())
+    missing_in_interface = html_options - interface_options
+    not_in_html = interface_options - html_options
+    s = ""
+    if missing_in_interface:
+        s += "\nThe following <options> HTML tags were not found in the interface:\n"
+        s += "\n".join(str(tup) for tup in missing_in_interface)
+    if not_in_html:
+        s += "\nThe following options were defined, but are not present in the HTML:\n"
+        s += "\n".join(str(tup) for tup in not_in_html)
+    if s:
+        raise SelectOptionsWrong(f"Wrong options in field '{field_interface.name}'." + s)
+    if check_defaults:
+        default_option = field_tag.find(lambda tag: tag.name == OPTION and tag.has_attr(SELECTED))
+        if default_option is not None and default_option[VALUE] != field_interface.default:
+            raise FieldInterfaceWrong(f"Default option '{default_option[VALUE]}' missing for {field_interface}")
+
+
+def check_field_interface(field_tag: Tag, field_interface: 'FormField', check_defaults: bool = True) -> None:
+    """
+    Compares all attributes of a `'FormField'` object with its HTML counterpart.
+    Calls `check_select_field_options` on select fields.
+
+    Args:
+        field_tag: The `bs4.Tag` object representing the <input> or <select> HTML tag
+        field_interface: The `'FormField'` to check against the HTML tag
+        check_defaults (optional): If set to `False`, pre-set field values or pre-selected options are not checked
+
+    Raises:
+        `FieldInterfaceWrong`
+            when dealing with an <input> tag, but `options` were defined on the interface, or
+            when the `default` does not match the tag's `value` attribute
+    """
+    assert field_tag[NAME] == field_interface.name
+    if field_tag.name == INPUT:
+        if field_interface.options is not None:
+            raise FieldInterfaceWrong(f"Options provided for input field '{field_interface.name}'")
+        if check_defaults:
+            default = field_tag.attrs.get(VALUE)
+            if field_interface.default != default:
+                raise FieldInterfaceWrong(f"Input field default not correct on '{field_interface}'")
+    if field_tag.name == SELECT:
+        check_select_field_options(field_tag, field_interface, check_defaults=check_defaults)
+
+
+def check_form_interface(form_tag: Tag, form_interface: 'Form', check_defaults: bool = True) -> None:
+    """
+    Compares all fields of a `Form` object with the fields found in its HTML counterpart.
+    Calls `check_field_interface` on each field.
+
+    Args:
+        form_tag: The `bs4.Tag` object representing the <form> HTML tag
+        form_interface: The `Form` to check against the HTML tag
+        check_defaults (optional): If set to `False`, pre-set field values or pre-selected options are not checked
+
+    Raises:
+        `UnknownField`
+            if any of the fields' name does not correspond to the `name` attribute of a field in the HTML form
+        `IncompleteForm`
+            if a field in the HTML form has not been defined in the form interface
+    """
+    field_tags: Dict[str, Tag] = {tag[NAME]: tag for tag in form_tag.find_all(INPUT) + form_tag.find_all(SELECT)}
+    for field in form_interface.fields.values():
+        tag = field_tags.pop(field.name, None)
+        if tag is None:
+            raise UnknownField(f"The defined field does not exist in the form: {field}")
+        check_field_interface(tag, field, check_defaults=check_defaults)
+    if len(field_tags) > 0:
+        raise IncompleteForm(f"Form interface missing fields: {list(field_tags.keys())}")
+
+
+def get_field_value(form_tag: Tag, field_name: str) -> str:
+    """
+    Returns the string value of the `value` attribute of a form field with a specified name.
+
+    Args:
+        form_tag: The `bs4.Tag` object representing the <form> HTML tag
+        field_name: The value of the `name` attribute of the form field of interest
+
+    Raises:
+        `ValueError`
+            if no field with the specified name exists in the HTML form
+        `AttributeError`
+            if the field with the specified name does not have a `value` attribute
+    """
+    def form_field_has_the_name(tag: Tag):
+        return tag.name in {INPUT, SELECT} and tag.has_attr(NAME) and tag[NAME] == field_name
+    field = form_tag.find(form_field_has_the_name)
+    if field is None:
+        raise ValueError(f"No form field with the name '{field_name}' found")
+    if not field.has_attr(VALUE):
+        raise AttributeError(f"Field with the name '{field_name}' has no `value` attribute")
+    return field[VALUE]
+
+
+def get_field_values(form_tag: Tag, form_interface: 'Form', required_fields_only: bool = True) -> Dict[str, str]:
+    """
+    Goes through the fields of a `Form` object and tries to retrieve a value for each from an HTML <form>.
+
+    Args:
+        form_tag: The `bs4.Tag` object representing the <form> HTML tag
+        form_interface: The `Form` for which to get the field values
+        required_fields_only (optional): If `True` (default), values will be retrieved only for the required fields
+
+    Returns:
+        Dictionary with each key-value-pair representing a (required) field for which a `value` attribute existed in the
+        HTML form, with the key being identical to that of the field in the `Form.required_fields` dictionary and the
+        value being the actual `value` attribute's string value
+    """
+    output_dict = {}
+    for key, field in form_interface.fields.items():
+        if required_fields_only and not field.required:
+            continue
+        try:
+            output_dict[key] = get_field_value(form_tag, field.name)
+        except AttributeError:
+            pass
+    return output_dict