Initial commit

.gitignore

@@ -0,0 +1,164 @@
+*.ipynb
+temp.py
+
+
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+#poetry.lock
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#pdm.lock
+#   pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
+#   in version control.
+#   https://pdm.fming.dev/#use-with-ide
+.pdm.toml
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+#  JetBrains specific template is maintained in a separate JetBrains.gitignore that can
+#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#  and can be added to the global gitignore or merged into this file.  For a more nuclear
+#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
+#.idea/

.vscode/settings.json

@@ -0,0 +1,36 @@
+{
+    "editor.formatOnSave": true,
+    "search.followSymlinks": false,
+    "files.trimTrailingWhitespace": true,
+    "files.trimFinalNewlines": true,
+    "files.insertFinalNewline": true,
+    "python.analysis.typeCheckingMode": "basic",
+    "python.linting.pylintEnabled": true,
+    "python.linting.pylintArgs": [
+      "--disable=C0111",
+      "--disable=no-name-in-module",
+      "--max-line-length=120",
+      "--rcfile",
+      "${workspaceFolder}/setup.cfg"
+    ],
+    "isort.args": ["-l", "140"],
+    "editor.defaultFormatter": "esbenp.prettier-vscode",
+    "python.formatting.provider": "black",
+    "python.formatting.blackArgs": ["--line-length", "140"],
+    "python.linting.enabled": true,
+    "python.testing.pytestEnabled": true,
+    "[python]": {
+      "editor.defaultFormatter": "ms-python.black-formatter",
+      "editor.codeActionsOnSave": {
+        "source.organizeImports": false
+      }
+    },
+    "python.analysis.extraPaths": [],
+    "search.exclude": {
+      "**/.vscode": true,
+      "**/.git": true
+    },
+    "editor.codeActionsOnSave": {
+      "source.fixAll": true
+    },
+  }

llm_lib/__init__.py

@@ -0,0 +1,3 @@
+from .base import BaseLLMResponse
+from .fields import LLMArrayOutput, LLMOutput
+from .prompt_definition.prompt_template import PromptTemplate

llm_lib/base.py

@@ -0,0 +1,127 @@
+from typing import TYPE_CHECKING, Any, ClassVar, TypeVar
+
+from .exceptions import LLMOutputFieldMissing, LLMOutputFieldWrongType
+from .fields import ClassPlaceholder, LLMArrayOutputInfo, LLMFieldInfo, LLMOutputInfo
+from .meta import LLMMeta
+from .parser import Parser
+from .utils.type_checker import SupportedBaseTypes, array_item_type
+
+if TYPE_CHECKING:
+    from inspect import Signature
+
+
+class BaseLLMResponse(metaclass=LLMMeta):
+    if TYPE_CHECKING:
+        # populated by the metaclass (ClassPlaceholder used to prevent showing up as type suggestion)
+        __fields__: ClassVar[dict[str, LLMFieldInfo]] = ClassPlaceholder(init=False, value={})
+        __signature__: ClassVar["Signature"] = ClassPlaceholder(init=False)
+
+    def __init__(self, **data: Any):
+        # print(data)
+        data = self._prepare_and_validate_dict(data)
+        self.__dict__.update(data)
+
+    # don't allow setting of fields that aren't defined in IDE
+    if not TYPE_CHECKING:
+
+        def __setattr__(self, __name: str, __value: Any) -> None:
+            # print(f"Setting {__name} to {__value}")
+
+            __value = self._prepare_and_validate_field(__name, __value)  # throws error if invalid
+
+            super().__setattr__(__name, __value)
+
+    def __repr__(self) -> str:
+        attrs = ", ".join(f"{k}={v}" for k, v in self.__dict__.items())
+        return f"{self.__class__.__name__}({attrs})"
+
+    def _prepare_field_value(self, value: Any, _type: type) -> Any:
+        """Converts single values from string to their type, otherwise leaves as is"""
+        if _type == str:
+            return value
+
+        if isinstance(value, _type):
+            return value
+
+        if isinstance(value, str):
+            value = value.strip()
+            if _type == int:
+                try:
+                    return int(value)
+                except ValueError:
+                    raise LLMOutputFieldWrongType(f'"{value}" is not a valid integer value')
+            elif _type == float:
+                try:
+                    return float(value)
+                except ValueError:
+                    raise LLMOutputFieldWrongType(f'"{value}" is not a valid float value')
+            elif _type == bool:
+                if value.lower() in ("true", "yes", "1"):
+                    return True
+                elif value.lower() in ("false", "no", "0"):
+                    return False
+                else:
+                    raise LLMOutputFieldWrongType(f'"{value}" is not a valid boolean value')
+
+        return value
+
+    def _prepare_and_validate_field(self, __name: str, __value: Any) -> Any:
+        if __name not in self.__fields__:
+            raise ValueError(f'"{self.__class__.__name__}" object has no field "{__name}"')
+
+        field_info = self.__fields__[__name]
+        if isinstance(field_info.info, LLMOutputInfo):
+            __value = self._prepare_field_value(__value, field_info.type_)
+
+            if __value is None and field_info.info.required:
+                raise TypeError(f'"{self.__class__.__name__}" field "{__name}" is required')
+            if not isinstance(__value, field_info.type_):
+                raise LLMOutputFieldWrongType(f'"{self.__class__.__name__}" field "{__name}" must be of type {field_info.type_}')
+
+        elif isinstance(field_info.info, LLMArrayOutputInfo):
+            item_type = array_item_type(field_info.type_)
+
+            if not isinstance(__value, list):
+                raise TypeError(f'"{self.__class__.__name__}" field "{__name}" must be a list')
+            if field_info.info.min_count is not None and len(__value) < field_info.info.min_count:
+                raise ValueError(f'"{self.__class__.__name__}" field "{__name}" must have at least {field_info.info.min_count} items')
+            if field_info.info.max_count is not None and len(__value) > field_info.info.max_count:
+                raise ValueError(f'"{self.__class__.__name__}" field "{__name}" must have at most {field_info.info.max_count} items')
+
+            __value = [self._prepare_field_value(v, item_type) for v in __value]
+            if not all(isinstance(v, item_type) for v in __value):
+                raise LLMOutputFieldWrongType(f'"{self.__class__.__name__}" field "{__name}" must be a list of type {field_info.type_}')
+
+        return __value
+
+    def _prepare_and_validate_dict(self, __values: dict[str, Any]) -> dict[str, Any]:
+        # check every field if it's valid
+        for k, v in __values.items():
+            __values[k] = self._prepare_and_validate_field(k, v)
+
+        # check every field not in dict, if it's required
+        for field in self.__fields__.values():
+            if field.key in __values:
+                continue
+
+            if isinstance(field.info, LLMOutputInfo):
+                if field.info.required:
+                    raise ValueError(f'"{self.__class__.__name__}" field "{field.key}" is required')
+                else:
+                    __values[field.key] = field.info.default
+
+            elif isinstance(field.info, LLMArrayOutputInfo):
+                if field.info.min_count > 0:
+                    raise ValueError(f'"{self.__class__.__name__}" field "{field.key}" requires at least {field.info.min_count} items')
+                else:
+                    __values[field.key] = []
+
+        return __values
+
+    # - Parsing
+
+    _Self = TypeVar("_Self", bound="BaseLLMResponse")  # backward compatibility for pre-Python 3.12
+
+    @classmethod
+    def parse_response(cls: type[_Self], response: str) -> _Self:
+        return Parser(cls).parse(response)

llm_lib/example_formatter.py

@@ -0,0 +1,39 @@
+from typing import Any, Iterable
+
+
+class LimitedExampleListFormatter:
+    """
+    The LimitedExampleListFormatter class converts a list of objects into a string representation,
+    limiting the number of displayed elements. If the number of elements exceeds the limit,
+    it includes a skip separator between the displayed and omitted elements.
+
+    Attributes:
+        max_examples (int): The maximum number of elements to display in the string.
+        separator (str): The separator used between the displayed elements.
+        skip_separator (str, optional): The separator used between the displayed and omitted elements. Defaults to "...".
+
+    Examples:
+        formatter = LimitedExampleListFormatter(3, ', ')
+        print(formatter.format(['apple', 'banana', 'cherry', 'date', 'elderberry']))
+        # Output: "apple, banana, ..., elderberry"
+    """
+
+    def __init__(self, max_examples: int, separator: str, skip_separator: str = "...", examples_after_separtor: int = 1):
+        self.max_examples = max_examples
+        self.separator = separator
+        self.skip_separator = skip_separator
+        self.examples_after_separtor = examples_after_separtor
+        assert self.max_examples > self.examples_after_separtor, "max_examples must be greater than examples_after_separtor"
+
+    def format(self, items: Iterable[Any]) -> str:
+        str_items = [str(item) for item in items]
+
+        if len(str_items) <= self.max_examples:
+            return self.separator.join(str_items)
+
+        # We want to show first n-1 examples, then skip separator, then last example
+        return self.separator.join(
+            str_items[: self.max_examples - self.examples_after_separtor]
+            + [self.skip_separator]
+            + str_items[-self.examples_after_separtor :]
+        )

llm_lib/exceptions.py

@@ -0,0 +1,14 @@
+class LLMException(Exception):
+    ...
+
+
+class LLMTokenLimitExceeded(LLMException):
+    ...
+
+
+class LLMOutputFieldMissing(LLMException):
+    ...
+
+
+class LLMOutputFieldWrongType(LLMException):
+    ...