2 Commits

Author SHA1 Message Date
juk0de 2548ae5a52 added tests for 'chat.py' 2023-08-31 09:42:40 +02:00
juk0de 2dc8e1d6b2 added new module 'chat.py' 2023-08-31 09:42:40 +02:00
29 changed files with 896 additions and 2336 deletions
+42 -74
View File
@@ -37,95 +37,63 @@ cmm [global options] command [command options]
### Global Options
- `-C`, `--config`: Config file name (defaults to `.config.yaml`).
- `-c`, `--config`: Config file name (defaults to `.config.yaml`).
### Commands
- `ask`: Ask a question.
- `hist`: Print chat history.
- `tag`: Manage tags.
- `config`: Manage configuration.
- `print`: Print files.
### Command Options
#### Question
#### `ask` Command Options
The `question` command is used to ask, create, and process questions.
- `-q`, `--question`: Question to ask (required).
- `-m`, `--max-tokens`: Max tokens to use.
- `-T`, `--temperature`: Temperature to use.
- `-M`, `--model`: Model to use.
- `-n`, `--number`: Number of answers to produce (default is 3).
- `-s`, `--source`: Add content of a file to the query.
- `-S`, `--only-source-code`: Add pure source code to the chat history.
- `-t`, `--tags`: List of tag names.
- `-e`, `--extags`: List of tag names to exclude.
- `-o`, `--output-tags`: List of output tag names (default is the input tags).
- `-a`, `--match-all-tags`: All given tags must match when selecting chat history entries.
```bash
cmm question [-t OTAGS]... [-k ATAGS]... [-x XTAGS]... [-o OUTTAGS]... [-A AI] [-M MODEL] [-n NUM] [-m MAX] [-T TEMP] (-a ASK | -c CREATE | -r REPEAT | -p PROCESS) [-O] [-s SOURCE]... [-S SOURCE]...
```
#### `hist` Command Options
* `-t, --or-tags OTAGS` : List of tags (one must match)
* `-k, --and-tags ATAGS` : List of tags (all must match)
* `-x, --exclude-tags XTAGS` : List of tags to exclude
* `-o, --output-tags OUTTAGS` : List of output tags (default: use input tags)
* `-A, --AI AI` : AI ID to use
* `-M, --model MODEL` : Model to use
* `-n, --num-answers NUM` : Number of answers to request
* `-m, --max-tokens MAX` : Max. number of tokens
* `-T, --temperature TEMP` : Temperature value
* `-a, --ask ASK` : Ask a question
* `-c, --create CREATE` : Create a question
* `-r, --repeat REPEAT` : Repeat a question
* `-p, --process PROCESS` : Process existing questions
* `-O, --overwrite` : Overwrite existing messages when repeating them
* `-s, --source-text SOURCE` : Add content of a file to the query
* `-S, --source-code SOURCE` : Add source code file content to the chat history
- `-d`, `--dump`: Print chat history as Python structure.
- `-w`, `--with-tags`: Print chat history with tags.
- `-W`, `--with-files`: Print chat history with filenames.
- `-S`, `--only-source-code`: Print only source code.
- `-t`, `--tags`: List of tag names.
- `-e`, `--extags`: List of tag names to exclude.
- `-a`, `--match-all-tags`: All given tags must match when selecting chat history entries.
#### Hist
#### `tag` Command Options
The `hist` command is used to print the chat history.
- `-l`, `--list`: List all tags and their frequency.
```bash
cmm hist [-t OTAGS]... [-k ATAGS]... [-x XTAGS]... [-w] [-W] [-S] [-A ANSWER] [-Q QUESTION]
```
#### `config` Command Options
* `-t, --or-tags OTAGS` : List of tags (one must match)
* `-k, --and-tags ATAGS` : List of tags (all must match)
* `-x, --exclude-tags XTAGS` : List of tags to exclude
* `-w, --with-tags` : Print chat history with tags
* `-W, --with-files` : Print chat history with filenames
* `-S, --source-code-only` : Print only source code
* `-A, --answer ANSWER` : Search for answer substring
* `-Q, --question QUESTION` : Search for question substring
- `-l`, `--list-models`: List all available models.
- `-m`, `--print-model`: Print the currently configured model.
- `-M`, `--model`: Set model in the config file.
#### Tags
#### `print` Command Options
The `tags` command is used to manage tags.
```bash
cmm tags (-l | -p PREFIX | -c CONTENT)
```
* `-l, --list` : List all tags and their frequency
* `-p, --prefix PREFIX` : Filter tags by prefix
* `-c, --contain CONTENT` : Filter tags by contained substring
#### Config
The `config` command is used to manage the configuration.
```bash
cmm config (-l | -m | -c CREATE)
```
* `-l, --list-models` : List all available models
* `-m, --print-model` : Print the currently configured model
* `-c, --create CREATE` : Create config with default settings in the given file
#### Print
The `print` command is used to print message files.
```bash
cmm print -f FILE [-q | -a | -S]
```
* `-f, --file FILE` : File to print
* `-q, --question` : Print only question
* `-a, --answer` : Print only answer
* `-S, --only-source-code` : Print only source code
- `-f`, `--file`: File to print (required).
- `-S`, `--only-source-code`: Print only source code.
### Examples
1. Ask a question:
```bash
cmm question -a "What is the meaning of life?" -t philosophy -x religion
cmm ask -q "What is the meaning of life?" -t philosophy -e religion
```
2. Display the chat history:
@@ -137,19 +105,19 @@ cmm hist
3. Filter chat history by tags:
```bash
cmm hist --or-tags tag1 tag2
cmm hist -t tag1 tag2
```
4. Exclude chat history by tags:
```bash
cmm hist --exclude-tags tag3 tag4
cmm hist -e tag3 tag4
```
5. List all tags and their frequency:
```bash
cmm tags -l
cmm tag -l
```
6. Print the contents of a file:
-80
View File
@@ -1,80 +0,0 @@
from dataclasses import dataclass
from typing import Protocol, Optional, Union
from .configuration import AIConfig
from .tags import Tag
from .message import Message
from .chat import Chat
class AIError(Exception):
pass
@dataclass
class Tokens:
prompt: int = 0
completion: int = 0
total: int = 0
@dataclass
class AIResponse:
"""
The response to an AI request. Consists of one or more messages
(each containing the question and a single answer) and the nr.
of used tokens.
"""
messages: list[Message]
tokens: Optional[Tokens] = None
class AI(Protocol):
"""
The base class for AI clients.
"""
ID: str
name: str
config: AIConfig
def request(self,
question: Message,
chat: Chat,
num_answers: int = 1,
otags: Optional[set[Tag]] = None) -> AIResponse:
"""
Make an AI request. Parameters:
* question: the question to ask
* chat: the chat history to be added as context
* num_answers: nr. of requested answers (corresponds
to the nr. of messages in the 'AIResponse')
* otags: the output tags, i. e. the tags that all
returned messages should contain
"""
raise NotImplementedError
def models(self) -> list[str]:
"""
Return all models supported by this AI.
"""
raise NotImplementedError
def print_models(self) -> None:
"""
Print all models supported by this AI.
"""
raise NotImplementedError
def tokens(self, data: Union[Message, Chat]) -> int:
"""
Computes the nr. of AI language tokens for the given message
or chat. Note that the computation may not be 100% accurate
and is not implemented for all AIs.
"""
raise NotImplementedError
def print(self) -> None:
"""
Print some info about the current AI, like system message.
"""
pass
-43
View File
@@ -1,43 +0,0 @@
"""
Creates different AI instances, based on the given configuration.
"""
import argparse
from typing import cast
from .configuration import Config, AIConfig, OpenAIConfig
from .ai import AI, AIError
from .ais.openai import OpenAI
def create_ai(args: argparse.Namespace, config: Config) -> AI: # noqa: 11
"""
Creates an AI subclass instance from the given arguments
and configuration file. If AI has not been set in the
arguments, it searches for the ID 'default'. If that
is not found, it uses the first AI in the list.
"""
ai_conf: AIConfig
if hasattr(args, 'AI') and args.AI:
try:
ai_conf = config.ais[args.AI]
except KeyError:
raise AIError(f"AI ID '{args.AI}' does not exist in this configuration")
elif 'default' in config.ais:
ai_conf = config.ais['default']
else:
try:
ai_conf = next(iter(config.ais.values()))
except StopIteration:
raise AIError("No AI found in this configuration")
if ai_conf.name == 'openai':
ai = OpenAI(cast(OpenAIConfig, ai_conf))
if hasattr(args, 'model') and args.model:
ai.config.model = args.model
if hasattr(args, 'max_tokens') and args.max_tokens:
ai.config.max_tokens = args.max_tokens
if hasattr(args, 'temperature') and args.temperature:
ai.config.temperature = args.temperature
return ai
else:
raise AIError(f"AI '{args.AI}' is not supported")
View File
-111
View File
@@ -1,111 +0,0 @@
"""
Implements the OpenAI client classes and functions.
"""
import openai
from typing import Optional, Union
from ..tags import Tag
from ..message import Message, Answer
from ..chat import Chat
from ..ai import AI, AIResponse, Tokens
from ..configuration import OpenAIConfig
ChatType = list[dict[str, str]]
class OpenAI(AI):
"""
The OpenAI AI client.
"""
def __init__(self, config: OpenAIConfig) -> None:
self.ID = config.ID
self.name = config.name
self.config = config
openai.api_key = config.api_key
def request(self,
question: Message,
chat: Chat,
num_answers: int = 1,
otags: Optional[set[Tag]] = None) -> AIResponse:
"""
Make an AI request, asking the given question with the given
chat history. The nr. of requested answers corresponds to the
nr. of messages in the 'AIResponse'.
"""
oai_chat = self.openai_chat(chat, self.config.system, question)
response = openai.ChatCompletion.create(
model=self.config.model,
messages=oai_chat,
temperature=self.config.temperature,
max_tokens=self.config.max_tokens,
top_p=self.config.top_p,
n=num_answers,
frequency_penalty=self.config.frequency_penalty,
presence_penalty=self.config.presence_penalty)
question.answer = Answer(response['choices'][0]['message']['content'])
question.tags = otags
question.ai = self.ID
question.model = self.config.model
answers: list[Message] = [question]
for choice in response['choices'][1:]: # type: ignore
answers.append(Message(question=question.question,
answer=Answer(choice['message']['content']),
tags=otags,
ai=self.ID,
model=self.config.model))
return AIResponse(answers, Tokens(response['usage']['prompt_tokens'],
response['usage']['completion_tokens'],
response['usage']['total_tokens']))
def models(self) -> list[str]:
"""
Return all models supported by this AI.
"""
ret = []
for engine in sorted(openai.Engine.list()['data'], key=lambda x: x['id']):
if engine['ready']:
ret.append(engine['id'])
ret.sort()
return ret
def print_models(self) -> None:
"""
Print all models supported by the current AI.
"""
not_ready = []
for engine in sorted(openai.Engine.list()['data'], key=lambda x: x['id']):
if engine['ready']:
print(engine['id'])
else:
not_ready.append(engine['id'])
if len(not_ready) > 0:
print('\nNot ready: ' + ', '.join(not_ready))
def openai_chat(self, chat: Chat, system: str,
question: Optional[Message] = None) -> ChatType:
"""
Create a chat history with system message in OpenAI format.
Optionally append a new question.
"""
oai_chat: ChatType = []
def append(role: str, content: str) -> None:
oai_chat.append({'role': role, 'content': content.replace("''", "'")})
append('system', system)
for message in chat.messages:
if message.answer:
append('user', message.question)
append('assistant', message.answer)
if question:
append('user', question.question)
return oai_chat
def tokens(self, data: Union[Message, Chat]) -> int:
raise NotImplementedError
def print(self) -> None:
print(f"MODEL: {self.config.model}")
print("=== SYSTEM ===")
print(self.config.system)
+45
View File
@@ -0,0 +1,45 @@
import openai
from .utils import ChatType
from .configuration import Config
def openai_api_key(api_key: str) -> None:
openai.api_key = api_key
def print_models() -> None:
"""
Print all models supported by the current AI.
"""
not_ready = []
for engine in sorted(openai.Engine.list()['data'], key=lambda x: x['id']):
if engine['ready']:
print(engine['id'])
else:
not_ready.append(engine['id'])
if len(not_ready) > 0:
print('\nNot ready: ' + ', '.join(not_ready))
def ai(chat: ChatType,
config: Config,
number: int
) -> tuple[list[str], dict[str, int]]:
"""
Make AI request with the given chat history and configuration.
Return AI response and tokens used.
"""
response = openai.ChatCompletion.create(
model=config.openai.model,
messages=chat,
temperature=config.openai.temperature,
max_tokens=config.openai.max_tokens,
top_p=config.openai.top_p,
n=number,
frequency_penalty=config.openai.frequency_penalty,
presence_penalty=config.openai.presence_penalty)
result = []
for choice in response['choices']: # type: ignore
result.append(choice['message']['content'].strip())
return result, dict(response['usage']) # type: ignore
+68 -422
View File
@@ -2,22 +2,17 @@
Module implementing various chat classes and functions for managing a chat history.
"""
import shutil
from pathlib import Path
import pathlib
from pprint import PrettyPrinter
from pydoc import pager
from dataclasses import dataclass
from typing import TypeVar, Type, Optional, ClassVar, Any, Callable, Literal, Union
from .configuration import default_config_file
from .message import Message, MessageFilter, MessageError, message_in
from typing import TypeVar, Type, Optional, ClassVar, Any, Callable
from .message import Question, Answer, Message, MessageFilter, MessageError, source_code, message_in
from .tags import Tag
ChatInst = TypeVar('ChatInst', bound='Chat')
ChatDBInst = TypeVar('ChatDBInst', bound='ChatDB')
db_next_file = '.next'
ignored_files = [db_next_file, default_config_file]
msg_location = Literal['mem', 'disk', 'cache', 'db', 'all']
class ChatError(Exception):
pass
@@ -35,7 +30,7 @@ def print_paged(text: str) -> None:
pager(text)
def read_dir(dir_path: Path,
def read_dir(dir_path: pathlib.Path,
glob: Optional[str] = None,
mfilter: Optional[MessageFilter] = None) -> list[Message]:
"""
@@ -50,32 +45,17 @@ def read_dir(dir_path: Path,
messages: list[Message] = []
file_iter = dir_path.glob(glob) if glob else dir_path.iterdir()
for file_path in sorted(file_iter):
if (file_path.is_file()
and file_path.name not in ignored_files # noqa: W503
and file_path.suffix in Message.file_suffixes): # noqa: W503
if file_path.is_file():
try:
message = Message.from_file(file_path, mfilter)
if message:
messages.append(message)
except MessageError as e:
print(f"WARNING: Skipping message in '{file_path}': {str(e)}")
print(f"Error processing message in '{file_path}': {str(e)}")
return messages
def make_file_path(dir_path: Path,
file_suffix: str,
next_fid: Callable[[], int]) -> Path:
"""
Create a file_path for the given directory using the
given file_suffix and ID generator function.
"""
file_path = dir_path / f"{next_fid():04d}{file_suffix}"
while file_path.exists():
file_path = dir_path / f"{next_fid():04d}{file_suffix}"
return file_path
def write_dir(dir_path: Path,
def write_dir(dir_path: pathlib.Path,
messages: list[Message],
file_suffix: str,
next_fid: Callable[[], int]) -> None:
@@ -93,26 +73,15 @@ def write_dir(dir_path: Path,
file_path = message.file_path
# message has no file_path: create one
if not file_path:
file_path = make_file_path(dir_path, file_suffix, next_fid)
fid = next_fid()
fname = f"{fid:04d}{file_suffix}"
file_path = dir_path / fname
# file_path does not point to given directory: modify it
elif not file_path.parent.samefile(dir_path):
file_path = dir_path / file_path.name
message.to_file(file_path)
def clear_dir(dir_path: Path,
glob: Optional[str] = None) -> None:
"""
Deletes all Message files in the given directory.
"""
file_iter = dir_path.glob(glob) if glob else dir_path.iterdir()
for file_path in file_iter:
if (file_path.is_file()
and file_path.name not in ignored_files # noqa: W503
and file_path.suffix in Message.file_suffixes): # noqa: W503
file_path.unlink(missing_ok=True)
@dataclass
class Chat:
"""
@@ -121,43 +90,14 @@ class Chat:
messages: list[Message]
def __post_init__(self) -> None:
self.validate()
def validate(self) -> None:
"""
Validate this Chat instance.
"""
def msg_paths(stem: str) -> list[str]:
return [str(fp) for fp in file_paths if fp.stem == stem]
file_paths: set[Path] = {m.file_path for m in self.messages if m.file_path is not None}
file_stems = [m.file_path.stem for m in self.messages if m.file_path is not None]
error = False
for fp in file_paths:
if file_stems.count(fp.stem) > 1:
print(f"ERROR: Found multiple copies of message '{fp.stem}': {msg_paths(fp.stem)}")
error = True
if error:
raise ChatError("Validation failed")
def msg_name_matches(self, file_path: Path, name: str) -> bool:
"""
Return True if the given name matches the given file_path.
Matching is True if:
* 'name' matches the full 'file_path'
* 'name' matches 'file_path.name' (i. e. including the suffix)
* 'name' matches 'file_path.stem' (i. e. without a suffix)
"""
return Path(name) == file_path or name == file_path.name or name == file_path.stem
def msg_filter(self, mfilter: MessageFilter) -> None:
def filter(self, mfilter: MessageFilter) -> None:
"""
Use 'Message.match(mfilter) to remove all messages that
don't fulfill the filter requirements.
"""
self.messages = [m for m in self.messages if m.match(mfilter)]
def msg_sort(self, reverse: bool = False) -> None:
def sort(self, reverse: bool = False) -> None:
"""
Sort the messages according to 'Message.msg_id()'.
"""
@@ -167,104 +107,49 @@ class Chat:
except MessageError:
pass
def msg_unique_id(self) -> None:
"""
Remove duplicates from the internal messages, based on the msg_id (i. e. file_path).
Messages without a file_path are kept.
"""
old_msgs = self.messages.copy()
self.messages = []
for m in old_msgs:
if not message_in(m, self.messages):
self.messages.append(m)
self.msg_sort()
def msg_unique_content(self) -> None:
"""
Remove duplicates from the internal messages, based on the content (i. e. question + answer).
"""
self.messages = list(set(self.messages))
self.msg_sort()
def msg_clear(self) -> None:
def clear(self) -> None:
"""
Delete all messages.
"""
self.messages = []
def msg_add(self, messages: list[Message]) -> None:
def add_msgs(self, msgs: list[Message]) -> None:
"""
Add new messages and sort them if possible.
"""
self.messages += messages
self.msg_sort()
self.messages += msgs
self.sort()
def msg_latest(self, mfilter: Optional[MessageFilter] = None) -> Optional[Message]:
"""
Return the last added message (according to the file ID) that matches the given filter.
When containing messages without a valid file_path, it returns the latest message in
the internal list.
"""
if len(self.messages) > 0:
self.msg_sort()
for m in reversed(self.messages):
if mfilter is None or m.match(mfilter):
return m
return None
def msg_find(self, msg_names: list[str]) -> list[Message]:
"""
Search and return the messages with the given names. Names can either be filenames
(with or without suffix), full paths or Message.msg_id(). Messages that can't be
found are ignored (i. e. the caller should check the result if they require all
messages).
"""
return [m for m in self.messages
if any((m.file_path and self.msg_name_matches(m.file_path, mn)) for mn in msg_names)]
def msg_remove(self, msg_names: list[str]) -> None:
"""
Remove the messages with the given names. Names can either be filenames
(with or without suffix), full paths or Message.msg_id().
"""
self.messages = [m for m in self.messages
if not any((m.file_path and self.msg_name_matches(m.file_path, mn)) for mn in msg_names)]
self.msg_sort()
def msg_tags(self, prefix: Optional[str] = None, contain: Optional[str] = None) -> set[Tag]:
def tags(self, prefix: Optional[str] = None, contain: Optional[str] = None) -> set[Tag]:
"""
Get the tags of all messages, optionally filtered by prefix or substring.
"""
tags: set[Tag] = set()
for m in self.messages:
tags |= m.filter_tags(prefix, contain)
return set(sorted(tags))
return tags
def msg_tags_frequency(self, prefix: Optional[str] = None, contain: Optional[str] = None) -> dict[Tag, int]:
"""
Get the frequency of all tags of all messages, optionally filtered by prefix or substring.
"""
tags: list[Tag] = []
for m in self.messages:
tags += [tag for tag in m.filter_tags(prefix, contain)]
return {tag: tags.count(tag) for tag in sorted(tags)}
def tokens(self) -> int:
"""
Returns the nr. of AI language tokens used by all messages in this chat.
If unknown, 0 is returned.
"""
return sum(m.tokens() for m in self.messages)
def print(self, source_code_only: bool = False,
with_tags: bool = False, with_files: bool = False,
def print(self, dump: bool = False, source_code_only: bool = False,
with_tags: bool = False, with_file: bool = False,
paged: bool = True) -> None:
if dump:
pp(self)
return
output: list[str] = []
for message in self.messages:
if source_code_only:
output.append(message.to_str(source_code_only=True))
output.extend(source_code(message.question, include_delims=True))
continue
output.append(message.to_str(with_tags, with_files))
output.append('-' * terminal_width())
output.append(Question.txt_header)
output.append(message.question)
if message.answer:
output.append(Answer.txt_header)
output.append(message.answer)
if with_tags:
output.append(message.tags_str())
if with_file:
output.append('FILE: ' + str(message.file_path))
if paged:
print_paged('\n'.join(output))
else:
@@ -283,8 +168,8 @@ class ChatDB(Chat):
default_file_suffix: ClassVar[str] = '.txt'
cache_path: Path
db_path: Path
cache_path: pathlib.Path
db_path: pathlib.Path
# a MessageFilter that all messages must match (if given)
mfilter: Optional[MessageFilter] = None
file_suffix: str = default_file_suffix
@@ -293,16 +178,12 @@ class ChatDB(Chat):
def __post_init__(self) -> None:
# contains the latest message ID
self.next_path = self.db_path / db_next_file
# make all paths absolute
self.cache_path = self.cache_path.absolute()
self.db_path = self.db_path.absolute()
self.validate()
self.next_fname = self.db_path / '.next'
@classmethod
def from_dir(cls: Type[ChatDBInst],
cache_path: Path,
db_path: Path,
cache_path: pathlib.Path,
db_path: pathlib.Path,
glob: Optional[str] = None,
mfilter: Optional[MessageFilter] = None) -> ChatDBInst:
"""
@@ -322,8 +203,8 @@ class ChatDB(Chat):
@classmethod
def from_messages(cls: Type[ChatDBInst],
cache_path: Path,
db_path: Path,
cache_path: pathlib.Path,
db_path: pathlib.Path,
messages: list[Message],
mfilter: Optional[MessageFilter] = None) -> ChatDBInst:
"""
@@ -333,7 +214,7 @@ class ChatDB(Chat):
def get_next_fid(self) -> int:
try:
with open(self.next_path, 'r') as f:
with open(self.next_fname, 'r') as f:
next_fid = int(f.read()) + 1
self.set_next_fid(next_fid)
return next_fid
@@ -342,282 +223,47 @@ class ChatDB(Chat):
return 1
def set_next_fid(self, fid: int) -> None:
with open(self.next_path, 'w') as f:
with open(self.next_fname, 'w') as f:
f.write(f'{fid}')
def msg_write(self, messages: Optional[list[Message]] = None) -> None:
def read_db(self) -> None:
"""
Write either the given messages or the internal ones to their CURRENT file_path.
If messages are given, they all must have a valid file_path. When writing the
internal messages, the ones with a valid file_path are written, the others
are ignored.
"""
if messages and any(m.file_path is None for m in messages):
raise ChatError("Can't write files without a valid file_path")
msgs = iter(messages if messages else self.messages)
while (m := next(msgs, None)):
m.to_file()
def msg_update(self, messages: list[Message], write: bool = True) -> None:
"""
Update EXISTING messages. A message is determined as 'existing' if a message with
the same base filename (i. e. 'file_path.name') is already in the list.
Only accepts existing messages.
"""
if any(not message_in(m, self.messages) for m in messages):
raise ChatError("Can't update messages that are not in the internal list")
# remove old versions and add new ones
self.messages = [m for m in self.messages if not message_in(m, messages)]
self.messages += messages
self.msg_sort()
# write the UPDATED messages if requested
if write:
self.msg_write(messages)
def msg_gather(self,
loc: msg_location,
require_file_path: bool = False,
mfilter: Optional[MessageFilter] = None) -> list[Message]:
"""
Gather and return messages from the given locations:
* 'mem' : messages currently in memory
* 'disk' : messages on disk (cache + DB directory), but not in memory
* 'cache': messages in the cache directory
* 'db' : messages in the DB directory
* 'all' : all messages ('mem' + 'disk')
If 'require_file_path' is True, return only files with a valid file_path.
"""
loc_messages: list[Message] = []
if loc in ['mem', 'all']:
if require_file_path:
loc_messages += [m for m in self.messages if (m.file_path is not None and (mfilter is None or m.match(mfilter)))]
else:
loc_messages += [m for m in self.messages if (mfilter is None or m.match(mfilter))]
if loc in ['cache', 'disk', 'all']:
loc_messages += read_dir(self.cache_path, mfilter=mfilter)
if loc in ['db', 'disk', 'all']:
loc_messages += read_dir(self.db_path, mfilter=mfilter)
# remove_duplicates and sort the list
unique_messages: list[Message] = []
for m in loc_messages:
if not message_in(m, unique_messages):
unique_messages.append(m)
try:
unique_messages.sort(key=lambda m: m.msg_id())
# messages in 'mem' can have an empty file_path
except MessageError:
pass
return unique_messages
def msg_find(self,
msg_names: list[str],
loc: msg_location = 'mem',
) -> list[Message]:
"""
Search and return the messages with the given names. Names can either be filenames
(with or without suffix), full paths or Message.msg_id(). Messages that can't be
found are ignored (i. e. the caller should check the result if they require all
messages).
Searches one of the following locations:
* 'mem' : messages currently in memory
* 'disk' : messages on disk (cache + DB directory), but not in memory
* 'cache': messages in the cache directory
* 'db' : messages in the DB directory
* 'all' : all messages ('mem' + 'disk')
"""
loc_messages = self.msg_gather(loc, require_file_path=True)
return [m for m in loc_messages
if any((m.file_path and self.msg_name_matches(m.file_path, mn)) for mn in msg_names)]
def msg_remove(self, msg_names: list[str], loc: msg_location = 'mem') -> None:
"""
Remove the messages with the given names. Names can either be filenames
(with or without suffix), full paths or Message.msg_id(). Also deletes the
files of all given messages with a valid file_path.
Delete files from one of the following locations:
* 'mem' : messages currently in memory
* 'disk' : messages on disk (cache + DB directory), but not in memory
* 'cache': messages in the cache directory
* 'db' : messages in the DB directory
* 'all' : all messages ('mem' + 'disk')
"""
if loc != 'mem':
# delete the message files first
rm_messages = self.msg_find(msg_names, loc=loc)
for m in rm_messages:
if (m.file_path):
m.file_path.unlink()
# then remove them from the internal list
super().msg_remove(msg_names)
def msg_latest(self,
mfilter: Optional[MessageFilter] = None,
loc: msg_location = 'mem') -> Optional[Message]:
"""
Return the last added message (according to the file ID) that matches the given filter.
Only consider messages with a valid file_path (except if loc is 'mem').
Searches one of the following locations:
* 'mem' : messages currently in memory
* 'disk' : messages on disk (cache + DB directory), but not in memory
* 'cache': messages in the cache directory
* 'db' : messages in the DB directory
* 'all' : all messages ('mem' + 'disk')
"""
# only consider messages with a valid file_path so they can be sorted
loc_messages = self.msg_gather(loc, require_file_path=True)
loc_messages.sort(key=lambda m: m.msg_id(), reverse=True)
for m in loc_messages:
if mfilter is None or m.match(mfilter):
return m
return None
def msg_in_cache(self, message: Union[Message, str]) -> bool:
"""
Return true if the given Message (or filename or Message.msg_id())
is located in the cache directory. False otherwise.
"""
if isinstance(message, Message):
return (message.file_path is not None
and message.file_path.parent.samefile(self.cache_path) # noqa: W503
and message.file_path.exists()) # noqa: W503
else:
return len(self.msg_find([message], loc='cache')) > 0
def msg_in_db(self, message: Union[Message, str]) -> bool:
"""
Return true if the given Message (or filename or Message.msg_id())
is located in the DB directory. False otherwise.
"""
if isinstance(message, Message):
return (message.file_path is not None
and message.file_path.parent.samefile(self.db_path) # noqa: W503
and message.file_path.exists()) # noqa: W503
else:
return len(self.msg_find([message], loc='db')) > 0
def cache_read(self, glob: Optional[str] = None, mfilter: Optional[MessageFilter] = None) -> None:
"""
Read messages from the cache directory. New ones are added to the internal list,
existing ones are replaced. A message is determined as 'existing' if a message
with the same base filename (i. e. 'file_path.name') is already in the list.
"""
new_messages = read_dir(self.cache_path, glob, mfilter)
# remove all messages from self.messages that are in the new list
self.messages = [m for m in self.messages if not message_in(m, new_messages)]
# copy the messages from the temporary list to self.messages and sort them
self.messages += new_messages
self.msg_sort()
def cache_write(self, messages: Optional[list[Message]] = None) -> None:
"""
Write messages to the cache directory. If a message has no file_path, a new one
will be created. If message.file_path exists, it will be modified to point to
the cache directory.
Does NOT add the messages to the internal list (use 'cache_add()' for that)!
"""
write_dir(self.cache_path,
messages if messages else self.messages,
self.file_suffix,
self.get_next_fid)
def cache_add(self, messages: list[Message], write: bool = True) -> None:
"""
Add NEW messages and set the file_path to the cache directory.
Only accepts messages without a file_path.
"""
if any(m.file_path is not None for m in messages):
raise ChatError("Can't add new messages with existing file_path")
if write:
write_dir(self.cache_path,
messages,
self.file_suffix,
self.get_next_fid)
else:
for m in messages:
m.file_path = make_file_path(self.cache_path, self.default_file_suffix, self.get_next_fid)
self.messages += messages
self.msg_sort()
def cache_clear(self, glob: Optional[str] = None) -> None:
"""
Delete all message files from the cache dir and remove them from the internal list.
"""
clear_dir(self.cache_path, glob)
# only keep messages from DB dir (or those that have not yet been written)
self.messages = [m for m in self.messages if not m.file_path or m.file_path.parent.samefile(self.db_path)]
def cache_move(self, message: Message) -> None:
"""
Moves the given messages to the cache directory.
"""
# remember the old path (if any)
old_path: Optional[Path] = None
if message.file_path:
old_path = message.file_path
# write message to the new destination
self.cache_write([message])
# remove the old one (if any)
if old_path:
self.msg_remove([str(old_path)], loc='db')
# (re)add it to the internal list
self.msg_add([message])
def db_read(self, glob: Optional[str] = None, mfilter: Optional[MessageFilter] = None) -> None:
"""
Read messages from the DB directory. New ones are added to the internal list,
existing ones are replaced. A message is determined as 'existing' if a message
with the same base filename (i. e. 'file_path.name') is already in the list.
Reads new messages from the DB directory. New ones are added to the internal list,
existing ones are kept or replaced. A message is determined as 'existing' if a
message with the same base filename (i. e. 'file_path.name') is already in the list.
"""
new_messages = read_dir(self.db_path, self.glob, self.mfilter)
# remove all messages from self.messages that are in the new list
self.messages = [m for m in self.messages if not message_in(m, new_messages)]
# copy the messages from the temporary list to self.messages and sort them
self.messages += new_messages
self.msg_sort()
self.sort()
def db_write(self, messages: Optional[list[Message]] = None) -> None:
def read_cache(self) -> None:
"""
Write messages to the DB directory. If a message has no file_path, a new one
will be created. If message.file_path exists, it will be modified to point
to the DB directory.
Does NOT add the messages to the internal list (use 'db_add()' for that)!
Reads new messages from the cache directory. New ones are added to the internal list,
existing ones are kept or replaced. A message is determined as 'existing' if a
message with the same base filename (i. e. 'file_path.name') is already in the list.
"""
write_dir(self.db_path,
messages if messages else self.messages,
self.file_suffix,
self.get_next_fid)
new_messages = read_dir(self.db_path, self.glob, self.mfilter)
# remove all messages from self.messages that are in the new list
self.messages = [m for m in self.messages if not message_in(m, new_messages)]
# copy the messages from the temporary list to self.messages and sort them
self.messages += new_messages
self.sort()
def db_add(self, messages: list[Message], write: bool = True) -> None:
def write_db(self) -> None:
"""
Add NEW messages and set the file_path to the DB directory.
Only accepts messages without a file_path.
Write all messages to the DB directory. If a message has no file_path,
a new one will be created. If message.file_path exists, it will be modified
to point to the DB directory.
"""
if any(m.file_path is not None for m in messages):
raise ChatError("Can't add new messages with existing file_path")
if write:
write_dir(self.db_path,
messages,
self.file_suffix,
self.get_next_fid)
else:
for m in messages:
m.file_path = make_file_path(self.db_path, self.default_file_suffix, self.get_next_fid)
self.messages += messages
self.msg_sort()
write_dir(self.db_path, self.messages, self.file_suffix, self.get_next_fid)
def db_move(self, message: Message) -> None:
def write_cache(self) -> None:
"""
Moves the given messages to the db directory.
Write all messages to the cache directory. If a message has no file_path,
a new one will be created. If message.file_path exists, it will be modified
to point to the cache directory.
"""
# remember the old path (if any)
old_path: Optional[Path] = None
if message.file_path:
old_path = message.file_path
# write message to the new destination
self.db_write([message])
# remove the old one (if any)
if old_path:
self.msg_remove([str(old_path)], loc='cache')
# (re)add it to the internal list
self.msg_add([message])
write_dir(self.cache_path, self.messages, self.file_suffix, self.get_next_fid)
-20
View File
@@ -1,20 +0,0 @@
import argparse
from pathlib import Path
from ..configuration import Config
from ..ai import AI
from ..ai_factory import create_ai
def config_cmd(args: argparse.Namespace) -> None:
"""
Handler for the 'config' command.
"""
if args.create:
Config.create_default(Path(args.create))
elif args.list_models or args.print_model:
config: Config = Config.from_file(args.config)
ai: AI = create_ai(args, config)
if args.list_models:
ai.print_models()
else:
print(ai.config.model)
-23
View File
@@ -1,23 +0,0 @@
import argparse
from pathlib import Path
from ..configuration import Config
from ..chat import ChatDB
from ..message import MessageFilter
def hist_cmd(args: argparse.Namespace, config: Config) -> None:
"""
Handler for the 'hist' command.
"""
mfilter = MessageFilter(tags_or=args.or_tags,
tags_and=args.and_tags,
tags_not=args.exclude_tags,
question_contains=args.question,
answer_contains=args.answer)
chat = ChatDB.from_dir(Path('.'),
Path(config.db),
mfilter=mfilter)
chat.print(args.source_code_only,
args.with_tags,
args.with_files)
-27
View File
@@ -1,27 +0,0 @@
import sys
import argparse
from pathlib import Path
from ..configuration import Config
from ..message import Message, MessageError
def print_cmd(args: argparse.Namespace, config: Config) -> None:
"""
Handler for the 'print' command.
"""
fname = Path(args.file)
try:
message = Message.from_file(fname)
if message:
if args.question:
print(message.question)
elif args.answer:
print(message.answer)
elif message.answer and args.only_source_code:
for code in message.answer.source_code():
print(code)
else:
print(message.to_str())
except MessageError:
print(f"File is not a valid message: {args.file}")
sys.exit(1)
-123
View File
@@ -1,123 +0,0 @@
import argparse
from pathlib import Path
from itertools import zip_longest
from ..configuration import Config
from ..chat import ChatDB
from ..message import Message, MessageFilter, MessageError, Question, source_code
from ..ai_factory import create_ai
from ..ai import AI, AIResponse
def add_file_as_text(question_parts: list[str], file: str) -> None:
"""
Add the given file as plain text to the question part list.
If the file is a Message, add the answer.
"""
file_path = Path(file)
content: str
try:
message = Message.from_file(file_path)
if message and message.answer:
content = message.answer
except MessageError:
with open(file) as r:
content = r.read().strip()
if len(content) > 0:
question_parts.append(content)
def add_file_as_code(question_parts: list[str], file: str) -> None:
"""
Add all source code from the given file. If no code segments can be extracted,
the whole content is added as source code segment. If the file is a Message,
extract the source code from the answer.
"""
file_path = Path(file)
content: str
try:
message = Message.from_file(file_path)
if message and message.answer:
content = message.answer
except MessageError:
with open(file) as r:
content = r.read().strip()
# extract and add source code
code_parts = source_code(content, include_delims=True)
if len(code_parts) > 0:
question_parts += code_parts
else:
question_parts.append(f"```\n{content}\n```")
def create_message(chat: ChatDB, args: argparse.Namespace) -> Message:
"""
Creates a new message from the given arguments and writes it
to the cache directory.
"""
question_parts = []
question_list = args.ask if args.ask is not None else []
text_files = args.source_text if args.source_text is not None else []
code_files = args.source_code if args.source_code is not None else []
for question, text_file, code_file in zip_longest(question_list, text_files, code_files, fillvalue=None):
if question is not None and len(question.strip()) > 0:
question_parts.append(question)
if text_file is not None and len(text_file) > 0:
add_file_as_text(question_parts, text_file)
if code_file is not None and len(code_file) > 0:
add_file_as_code(question_parts, code_file)
full_question = '\n\n'.join(question_parts)
message = Message(question=Question(full_question),
tags=args.output_tags, # FIXME
ai=args.AI,
model=args.model)
chat.cache_add([message])
return message
def question_cmd(args: argparse.Namespace, config: Config) -> None:
"""
Handler for the 'question' command.
"""
mfilter = MessageFilter(tags_or=args.or_tags if args.or_tags is not None else set(),
tags_and=args.and_tags if args.and_tags is not None else set(),
tags_not=args.exclude_tags if args.exclude_tags is not None else set())
chat = ChatDB.from_dir(cache_path=Path('.'),
db_path=Path(config.db),
mfilter=mfilter)
# if it's a new question, create and store it immediately
if args.ask or args.create:
message = create_message(chat, args)
if args.create:
return
# create the correct AI instance
ai: AI = create_ai(args, config)
if args.ask:
ai.print()
chat.print(paged=False)
response: AIResponse = ai.request(message,
chat,
args.num_answers, # FIXME
args.output_tags) # FIXME
chat.msg_update([response.messages[0]])
chat.cache_add(response.messages[1:])
for idx, msg in enumerate(response.messages):
print(f"=== ANSWER {idx+1} ===")
print(msg.answer)
if response.tokens:
print("===============")
print(response.tokens)
elif args.repeat is not None:
lmessage = chat.msg_latest()
assert lmessage
# TODO: repeat either the last question or the
# one(s) given in 'args.repeat' (overwrite
# existing ones if 'args.overwrite' is True)
pass
elif args.process is not None:
# TODO: process either all questions without an
# answer or the one(s) given in 'args.process'
pass
-17
View File
@@ -1,17 +0,0 @@
import argparse
from pathlib import Path
from ..configuration import Config
from ..chat import ChatDB
def tags_cmd(args: argparse.Namespace, config: Config) -> None:
"""
Handler for the 'tags' command.
"""
chat = ChatDB.from_dir(cache_path=Path('.'),
db_path=Path(config.db))
if args.list:
tags_freq = chat.msg_tags_frequency(args.prefix, args.contain)
for tag, freq in tags_freq.items():
print(f"- {tag}: {freq}")
# TODO: add renaming
+23 -121
View File
@@ -1,167 +1,69 @@
import yaml
from pathlib import Path
from typing import Type, TypeVar, Any, Optional, ClassVar
from dataclasses import dataclass, asdict, field
from typing import Type, TypeVar, Any
from dataclasses import dataclass, asdict
ConfigInst = TypeVar('ConfigInst', bound='Config')
AIConfigInst = TypeVar('AIConfigInst', bound='AIConfig')
OpenAIConfigInst = TypeVar('OpenAIConfigInst', bound='OpenAIConfig')
supported_ais: list[str] = ['openai']
default_config_file = '.config.yaml'
class ConfigError(Exception):
pass
def str_presenter(dumper: yaml.Dumper, data: str) -> yaml.ScalarNode:
"""
Changes the YAML dump style to multiline syntax for multiline strings.
"""
if len(data.splitlines()) > 1:
return dumper.represent_scalar('tag:yaml.org,2002:str', data, style='|')
return dumper.represent_scalar('tag:yaml.org,2002:str', data)
yaml.add_representer(str, str_presenter)
@dataclass
class AIConfig:
"""
The base class of all AI configurations.
"""
# the name of the AI the config class represents
# -> it's a class variable and thus not part of the
# dataclass constructor
name: ClassVar[str]
# a user-defined ID for an AI configuration entry
ID: str
model: str = 'n/a'
# the name must not be changed
def __setattr__(self, name: str, value: Any) -> None:
if name == 'name':
raise AttributeError("'{name}' is not allowed to be changed")
else:
super().__setattr__(name, value)
@dataclass
class OpenAIConfig(AIConfig):
class OpenAIConfig():
"""
The OpenAI section of the configuration file.
"""
name: ClassVar[str] = 'openai'
# all members have default values, so we can easily create
# a default configuration
ID: str = 'myopenai'
api_key: str = '0123456789'
model: str = 'gpt-3.5-turbo-16k'
temperature: float = 1.0
max_tokens: int = 4000
top_p: float = 1.0
frequency_penalty: float = 0.0
presence_penalty: float = 0.0
system: str = 'You are an assistant'
api_key: str
model: str
temperature: float
max_tokens: int
top_p: float
frequency_penalty: float
presence_penalty: float
@classmethod
def from_dict(cls: Type[OpenAIConfigInst], source: dict[str, Any]) -> OpenAIConfigInst:
"""
Create OpenAIConfig from a dict.
"""
res = cls(
return cls(
api_key=str(source['api_key']),
model=str(source['model']),
max_tokens=int(source['max_tokens']),
temperature=float(source['temperature']),
top_p=float(source['top_p']),
frequency_penalty=float(source['frequency_penalty']),
presence_penalty=float(source['presence_penalty']),
system=str(source['system'])
presence_penalty=float(source['presence_penalty'])
)
# overwrite default ID if provided
if 'ID' in source:
res.ID = source['ID']
return res
def ai_config_instance(name: str, conf_dict: Optional[dict[str, Any]] = None) -> AIConfig:
"""
Creates an AIConfig instance of the given name.
"""
if name.lower() == 'openai':
if conf_dict is None:
return OpenAIConfig()
else:
return OpenAIConfig.from_dict(conf_dict)
else:
raise ConfigError(f"Unknown AI '{name}'")
def create_default_ai_configs() -> dict[str, AIConfig]:
"""
Create a dict containing default configurations for all supported AIs.
"""
return {ai_config_instance(name).ID: ai_config_instance(name) for name in supported_ais}
@dataclass
class Config:
class Config():
"""
The configuration file structure.
"""
# all members have default values, so we can easily create
# a default configuration
db: str = './db/'
ais: dict[str, AIConfig] = field(default_factory=create_default_ai_configs)
system: str
db: str
openai: OpenAIConfig
@classmethod
def from_dict(cls: Type[ConfigInst], source: dict[str, Any]) -> ConfigInst:
"""
Create Config from a dict (with the same format as the config file).
Create OpenAIConfig from a dict.
"""
# create the correct AI type instances
ais: dict[str, AIConfig] = {}
for ID, conf in source['ais'].items():
# add the AI ID to the config (for easy internal access)
conf['ID'] = ID
ai_conf = ai_config_instance(conf['name'], conf)
ais[ID] = ai_conf
return cls(
system=str(source['system']),
db=str(source['db']),
ais=ais
openai=OpenAIConfig.from_dict(source['openai'])
)
@classmethod
def create_default(self, file_path: Path) -> None:
"""
Creates a default Config in the given file.
"""
conf = Config()
conf.to_file(file_path)
@classmethod
def from_file(cls: Type[ConfigInst], path: str) -> ConfigInst:
with open(path, 'r') as f:
source = yaml.load(f, Loader=yaml.FullLoader)
return cls.from_dict(source)
def to_file(self, file_path: Path) -> None:
# remove the AI name from the config (for a cleaner format)
data = self.as_dict()
for conf in data['ais'].values():
del (conf['ID'])
with open(file_path, 'w') as f:
yaml.dump(data, f, sort_keys=False)
def to_file(self, path: str) -> None:
with open(path, 'w') as f:
yaml.dump(asdict(self), f, sort_keys=False)
def as_dict(self) -> dict[str, Any]:
res = asdict(self)
# add the AI name manually (as first element)
# (not done by 'asdict' because it's a class variable)
for ID, conf in res['ais'].items():
res['ais'][ID] = {**{'name': self.ais[ID].name}, **conf}
return res
return asdict(self)
+160 -63
View File
@@ -2,29 +2,138 @@
# -*- coding: utf-8 -*-
# vim: set fileencoding=utf-8 :
import yaml
import sys
import argcomplete
import argparse
from pathlib import Path
import pathlib
from .utils import terminal_width, print_tag_args, print_chat_hist, display_source_code, print_tags_frequency, ChatType
from .storage import save_answers, create_chat_hist, get_tags, get_tags_unique, read_file, dump_data
from .api_client import ai, openai_api_key, print_models
from .configuration import Config
from itertools import zip_longest
from typing import Any
from .configuration import Config, default_config_file
from .message import Message
from .commands.question import question_cmd
from .commands.tags import tags_cmd
from .commands.config import config_cmd
from .commands.hist import hist_cmd
from .commands.print import print_cmd
default_config = '.config.yaml'
def tags_completer(prefix: str, parsed_args: Any, **kwargs: Any) -> list[str]:
config = Config.from_file(parsed_args.config)
return list(Message.tags_from_dir(Path(config.db), prefix=prefix))
with open(parsed_args.config, 'r') as f:
config = yaml.load(f, Loader=yaml.FullLoader)
return get_tags_unique(config, prefix)
def create_question_with_hist(args: argparse.Namespace,
config: Config,
) -> tuple[ChatType, str, list[str]]:
"""
Creates the "AI request", including the question and chat history as determined
by the specified tags.
"""
tags = args.tags or []
extags = args.extags or []
otags = args.output_tags or []
if not args.only_source_code:
print_tag_args(tags, extags, otags)
question_parts = []
question_list = args.question if args.question is not None else []
source_list = args.source if args.source is not None else []
for question, source in zip_longest(question_list, source_list, fillvalue=None):
if question is not None and source is not None:
with open(source) as r:
question_parts.append(f"{question}\n\n```\n{r.read().strip()}\n```")
elif question is not None:
question_parts.append(question)
elif source is not None:
with open(source) as r:
question_parts.append(f"```\n{r.read().strip()}\n```")
full_question = '\n\n'.join(question_parts)
chat = create_chat_hist(full_question, tags, extags, config,
args.match_all_tags, False, False)
return chat, full_question, tags
def tag_cmd(args: argparse.Namespace, config: Config) -> None:
"""
Handler for the 'tag' command.
"""
if args.list:
print_tags_frequency(get_tags(config, None))
def config_cmd(args: argparse.Namespace, config: Config) -> None:
"""
Handler for the 'config' command.
"""
if args.list_models:
print_models()
elif args.print_model:
print(config.openai.model)
elif args.model:
config.openai.model = args.model
config.to_file(args.config)
def ask_cmd(args: argparse.Namespace, config: Config) -> None:
"""
Handler for the 'ask' command.
"""
if args.max_tokens:
config.openai.max_tokens = args.max_tokens
if args.temperature:
config.openai.temperature = args.temperature
if args.model:
config.openai.model = args.model
chat, question, tags = create_question_with_hist(args, config)
print_chat_hist(chat, False, args.only_source_code)
otags = args.output_tags or []
answers, usage = ai(chat, config, args.number)
save_answers(question, answers, tags, otags, config)
print("-" * terminal_width())
print(f"Usage: {usage}")
def hist_cmd(args: argparse.Namespace, config: Config) -> None:
"""
Handler for the 'hist' command.
"""
tags = args.tags or []
extags = args.extags or []
chat = create_chat_hist(None, tags, extags, config,
args.match_all_tags,
args.with_tags,
args.with_files)
print_chat_hist(chat, args.dump, args.only_source_code)
def print_cmd(args: argparse.Namespace, config: Config) -> None:
"""
Handler for the 'print' command.
"""
fname = pathlib.Path(args.file)
if fname.suffix == '.yaml':
with open(args.file, 'r') as f:
data = yaml.load(f, Loader=yaml.FullLoader)
elif fname.suffix == '.txt':
data = read_file(fname)
else:
print(f"Unknown file type: {args.file}")
sys.exit(1)
if args.only_source_code:
display_source_code(data['answer'])
else:
print(dump_data(data).strip())
def create_parser() -> argparse.ArgumentParser:
parser = argparse.ArgumentParser(
description="ChatMastermind is a Python application that automates conversation with AI")
parser.add_argument('-C', '--config', help='Config file name.', default=default_config_file)
parser.add_argument('-c', '--config', help='Config file name.', default=default_config)
# subcommand-parser
cmdparser = parser.add_subparsers(dest='command',
@@ -34,90 +143,79 @@ def create_parser() -> argparse.ArgumentParser:
# a parent parser for all commands that support tag selection
tag_parser = argparse.ArgumentParser(add_help=False)
tag_arg = tag_parser.add_argument('-t', '--or-tags', nargs='+',
help='List of tags (one must match)', metavar='OTAGS')
tag_arg = tag_parser.add_argument('-t', '--tags', nargs='+',
help='List of tag names', metavar='TAGS')
tag_arg.completer = tags_completer # type: ignore
atag_arg = tag_parser.add_argument('-k', '--and-tags', nargs='+',
help='List of tags (all must match)', metavar='ATAGS')
atag_arg.completer = tags_completer # type: ignore
etag_arg = tag_parser.add_argument('-x', '--exclude-tags', nargs='+',
help='List of tags to exclude', metavar='XTAGS')
etag_arg.completer = tags_completer # type: ignore
extag_arg = tag_parser.add_argument('-e', '--extags', nargs='+',
help='List of tag names to exclude', metavar='EXTAGS')
extag_arg.completer = tags_completer # type: ignore
otag_arg = tag_parser.add_argument('-o', '--output-tags', nargs='+',
help='List of output tags (default: use input tags)', metavar='OUTTAGS')
help='List of output tag names, default is input', metavar='OTAGS')
otag_arg.completer = tags_completer # type: ignore
# a parent parser for all commands that support AI configuration
ai_parser = argparse.ArgumentParser(add_help=False)
ai_parser.add_argument('-A', '--AI', help='AI ID to use')
ai_parser.add_argument('-M', '--model', help='Model to use')
ai_parser.add_argument('-n', '--num-answers', help='Number of answers to request', type=int, default=1)
ai_parser.add_argument('-m', '--max-tokens', help='Max. nr. of tokens', type=int)
ai_parser.add_argument('-T', '--temperature', help='Temperature value', type=float)
# 'question' command parser
question_cmd_parser = cmdparser.add_parser('question', parents=[tag_parser, ai_parser],
help="ask, create and process questions.",
aliases=['q'])
question_cmd_parser.set_defaults(func=question_cmd)
question_group = question_cmd_parser.add_mutually_exclusive_group(required=True)
question_group.add_argument('-a', '--ask', nargs='+', help='Ask a question')
question_group.add_argument('-c', '--create', nargs='+', help='Create a question')
question_group.add_argument('-r', '--repeat', nargs='*', help='Repeat a question')
question_group.add_argument('-p', '--process', nargs='*', help='Process existing questions')
question_cmd_parser.add_argument('-O', '--overwrite', help='Overwrite existing messages when repeating them',
tag_parser.add_argument('-a', '--match-all-tags',
help="All given tags must match when selecting chat history entries",
action='store_true')
# enable autocompletion for tags
# 'ask' command parser
ask_cmd_parser = cmdparser.add_parser('ask', parents=[tag_parser],
help="Ask a question.",
aliases=['a'])
ask_cmd_parser.set_defaults(func=ask_cmd)
ask_cmd_parser.add_argument('-q', '--question', nargs='+', help='Question to ask',
required=True)
ask_cmd_parser.add_argument('-m', '--max-tokens', help='Max tokens to use', type=int)
ask_cmd_parser.add_argument('-T', '--temperature', help='Temperature to use', type=float)
ask_cmd_parser.add_argument('-M', '--model', help='Model to use')
ask_cmd_parser.add_argument('-n', '--number', help='Number of answers to produce', type=int,
default=1)
ask_cmd_parser.add_argument('-s', '--source', nargs='+', help='Source add content of a file to the query')
ask_cmd_parser.add_argument('-S', '--only-source-code', help='Add pure source code to the chat history',
action='store_true')
question_cmd_parser.add_argument('-s', '--source-text', nargs='+', help='Add content of a file to the query')
question_cmd_parser.add_argument('-S', '--source-code', nargs='+', help='Add source code file content to the chat history')
# 'hist' command parser
hist_cmd_parser = cmdparser.add_parser('hist', parents=[tag_parser],
help="Print chat history.",
aliases=['h'])
hist_cmd_parser.set_defaults(func=hist_cmd)
hist_cmd_parser.add_argument('-d', '--dump', help="Print chat history as Python structure",
action='store_true')
hist_cmd_parser.add_argument('-w', '--with-tags', help="Print chat history with tags.",
action='store_true')
hist_cmd_parser.add_argument('-W', '--with-files', help="Print chat history with filenames.",
action='store_true')
hist_cmd_parser.add_argument('-S', '--source-code-only', help='Print only source code',
hist_cmd_parser.add_argument('-S', '--only-source-code', help='Print only source code',
action='store_true')
hist_cmd_parser.add_argument('-A', '--answer', help='Search for answer substring')
hist_cmd_parser.add_argument('-Q', '--question', help='Search for question substring')
# 'tags' command parser
tags_cmd_parser = cmdparser.add_parser('tags',
# 'tag' command parser
tag_cmd_parser = cmdparser.add_parser('tag',
help="Manage tags.",
aliases=['t'])
tags_cmd_parser.set_defaults(func=tags_cmd)
tags_group = tags_cmd_parser.add_mutually_exclusive_group(required=True)
tags_group.add_argument('-l', '--list', help="List all tags and their frequency",
tag_cmd_parser.set_defaults(func=tag_cmd)
tag_group = tag_cmd_parser.add_mutually_exclusive_group(required=True)
tag_group.add_argument('-l', '--list', help="List all tags and their frequency",
action='store_true')
tags_cmd_parser.add_argument('-p', '--prefix', help="Filter tags by prefix")
tags_cmd_parser.add_argument('-c', '--contain', help="Filter tags by contained substring")
# 'config' command parser
config_cmd_parser = cmdparser.add_parser('config',
help="Manage configuration",
aliases=['c'])
config_cmd_parser.set_defaults(func=config_cmd)
config_cmd_parser.add_argument('-A', '--AI', help='AI ID to use')
config_group = config_cmd_parser.add_mutually_exclusive_group(required=True)
config_group.add_argument('-l', '--list-models', help="List all available models",
action='store_true')
config_group.add_argument('-m', '--print-model', help="Print the currently configured model",
action='store_true')
config_group.add_argument('-c', '--create', help="Create config with default settings in the given file")
config_group.add_argument('-M', '--model', help="Set model in the config file")
# 'print' command parser
print_cmd_parser = cmdparser.add_parser('print',
help="Print message files.",
help="Print files.",
aliases=['p'])
print_cmd_parser.set_defaults(func=print_cmd)
print_cmd_parser.add_argument('-f', '--file', help='File to print', required=True)
print_cmd_modes = print_cmd_parser.add_mutually_exclusive_group()
print_cmd_modes.add_argument('-q', '--question', help='Print only question', action='store_true')
print_cmd_modes.add_argument('-a', '--answer', help='Print only answer', action='store_true')
print_cmd_modes.add_argument('-S', '--only-source-code', help='Print only source code', action='store_true')
print_cmd_parser.add_argument('-S', '--only-source-code', help='Print only source code',
action='store_true')
argcomplete.autocomplete(parser)
return parser
@@ -127,11 +225,10 @@ def main() -> int:
parser = create_parser()
args = parser.parse_args()
command = parser.parse_args()
if command.func == config_cmd:
command.func(command)
else:
config = Config.from_file(args.config)
openai_api_key(config.openai.api_key)
command.func(command, config)
return 0
+37 -98
View File
@@ -3,11 +3,9 @@ Module implementing message related functions and classes.
"""
import pathlib
import yaml
import tempfile
import shutil
from typing import Type, TypeVar, ClassVar, Optional, Any, Union, Final, Literal, Iterable
from dataclasses import dataclass, asdict, field
from .tags import Tag, TagLine, TagError, match_tags, rename_tags
from .tags import Tag, TagLine, TagError, match_tags
QuestionInst = TypeVar('QuestionInst', bound='Question')
AnswerInst = TypeVar('AnswerInst', bound='Answer')
@@ -98,7 +96,7 @@ class AILine(str):
def __new__(cls: Type[AILineInst], string: str) -> AILineInst:
if not string.startswith(cls.prefix):
raise MessageError(f"AILine '{string}' is missing prefix '{cls.prefix}'")
raise TagError(f"AILine '{string}' is missing prefix '{cls.prefix}'")
instance = super().__new__(cls, string)
return instance
@@ -118,7 +116,7 @@ class ModelLine(str):
def __new__(cls: Type[ModelLineInst], string: str) -> ModelLineInst:
if not string.startswith(cls.prefix):
raise MessageError(f"ModelLine '{string}' is missing prefix '{cls.prefix}'")
raise TagError(f"ModelLine '{string}' is missing prefix '{cls.prefix}'")
instance = super().__new__(cls, string)
return instance
@@ -130,29 +128,28 @@ class ModelLine(str):
return cls(' '.join([cls.prefix, model]))
class Answer(str):
class Question(str):
"""
A single answer with a defined header.
A single question with a defined header.
"""
tokens: int = 0 # tokens used by this answer
txt_header: ClassVar[str] = '==== ANSWER ===='
yaml_key: ClassVar[str] = 'answer'
txt_header: ClassVar[str] = '=== QUESTION ==='
yaml_key: ClassVar[str] = 'question'
def __new__(cls: Type[AnswerInst], string: str) -> AnswerInst:
def __new__(cls: Type[QuestionInst], string: str) -> QuestionInst:
"""
Make sure the answer string does not contain the header as a whole line.
Make sure the question string does not contain the header.
"""
if cls.txt_header in string.split('\n'):
raise MessageError(f"Answer '{string}' contains the header '{cls.txt_header}'")
if cls.txt_header in string:
raise MessageError(f"Question '{string}' contains the header '{cls.txt_header}'")
instance = super().__new__(cls, string)
return instance
@classmethod
def from_list(cls: Type[AnswerInst], strings: list[str]) -> AnswerInst:
def from_list(cls: Type[QuestionInst], strings: list[str]) -> QuestionInst:
"""
Build Question from a list of strings. Make sure strings do not contain the header.
"""
if cls.txt_header in strings:
if any(cls.txt_header in string for string in strings):
raise MessageError(f"Question contains the header '{cls.txt_header}'")
instance = super().__new__(cls, '\n'.join(strings).strip())
return instance
@@ -164,33 +161,28 @@ class Answer(str):
return source_code(self, include_delims)
class Question(str):
class Answer(str):
"""
A single question with a defined header.
A single answer with a defined header.
"""
tokens: int = 0 # tokens used by this question
txt_header: ClassVar[str] = '=== QUESTION ==='
yaml_key: ClassVar[str] = 'question'
txt_header: ClassVar[str] = '=== ANSWER ==='
yaml_key: ClassVar[str] = 'answer'
def __new__(cls: Type[QuestionInst], string: str) -> QuestionInst:
def __new__(cls: Type[AnswerInst], string: str) -> AnswerInst:
"""
Make sure the question string does not contain the header as a whole line
(also not that from 'Answer', so it's always clear where the answer starts).
Make sure the answer string does not contain the header.
"""
string_lines = string.split('\n')
if cls.txt_header in string_lines:
raise MessageError(f"Question '{string}' contains the header '{cls.txt_header}'")
if Answer.txt_header in string_lines:
raise MessageError(f"Question '{string}' contains the header '{Answer.txt_header}'")
if cls.txt_header in string:
raise MessageError(f"Answer '{string}' contains the header '{cls.txt_header}'")
instance = super().__new__(cls, string)
return instance
@classmethod
def from_list(cls: Type[QuestionInst], strings: list[str]) -> QuestionInst:
def from_list(cls: Type[AnswerInst], strings: list[str]) -> AnswerInst:
"""
Build Question from a list of strings. Make sure strings do not contain the header.
"""
if cls.txt_header in strings:
if any(cls.txt_header in string for string in strings):
raise MessageError(f"Question contains the header '{cls.txt_header}'")
instance = super().__new__(cls, '\n'.join(strings).strip())
return instance
@@ -314,7 +306,7 @@ class Message():
mfilter.tags_not if mfilter else None)
else:
message = cls.__from_file_yaml(file_path)
if message and (mfilter is None or message.match(mfilter)):
if message and (not mfilter or (mfilter and message.match(mfilter))):
return message
else:
return None
@@ -357,20 +349,17 @@ class Message():
try:
pos = fd.tell()
ai = AILine(fd.readline()).ai()
except MessageError:
except TagError:
fd.seek(pos)
# ModelLine (Optional)
try:
pos = fd.tell()
model = ModelLine(fd.readline()).model()
except MessageError:
except TagError:
fd.seek(pos)
# Question and Answer
text = fd.read().strip().split('\n')
try:
question_idx = text.index(Question.txt_header) + 1
except ValueError:
raise MessageError(f"'{file_path}' does not contain a valid message")
try:
answer_idx = text.index(Answer.txt_header)
question = Question.from_list(text[question_idx:answer_idx])
@@ -390,36 +379,9 @@ class Message():
* Message.model_yaml_key: str [Optional]
"""
with open(file_path, "r") as fd:
try:
data = yaml.load(fd, Loader=yaml.FullLoader)
data[cls.file_yaml_key] = file_path
return cls.from_dict(data)
except Exception:
raise MessageError(f"'{file_path}' does not contain a valid message")
def to_str(self, with_tags: bool = False, with_file: bool = False, source_code_only: bool = False) -> str:
"""
Return the current Message as a string.
"""
output: list[str] = []
if source_code_only:
# use the source code from answer only
if self.answer:
output.extend(self.answer.source_code(include_delims=True))
return '\n'.join(output) if len(output) > 0 else ''
if with_tags:
output.append(self.tags_str())
if with_file:
output.append('FILE: ' + str(self.file_path))
output.append(Question.txt_header)
output.append(self.question)
if self.answer:
output.append(Answer.txt_header)
output.append(self.answer)
return '\n'.join(output)
def __str__(self) -> str:
return self.to_str(True, True, False)
def to_file(self, file_path: Optional[pathlib.Path]=None) -> None: # noqa: 11
"""
@@ -450,18 +412,16 @@ class Message():
* Answer.txt_header
* Answer
"""
with tempfile.NamedTemporaryFile(dir=file_path.parent, prefix=file_path.name, mode="w", delete=False) as temp_fd:
temp_file_path = pathlib.Path(temp_fd.name)
with open(file_path, "w") as fd:
if self.tags:
temp_fd.write(f'{TagLine.from_set(self.tags)}\n')
fd.write(f'{TagLine.from_set(self.tags)}\n')
if self.ai:
temp_fd.write(f'{AILine.from_ai(self.ai)}\n')
fd.write(f'{AILine.from_ai(self.ai)}\n')
if self.model:
temp_fd.write(f'{ModelLine.from_model(self.model)}\n')
temp_fd.write(f'{Question.txt_header}\n{self.question}\n')
fd.write(f'{ModelLine.from_model(self.model)}\n')
fd.write(f'{Question.txt_header}\n{self.question}\n')
if self.answer:
temp_fd.write(f'{Answer.txt_header}\n{self.answer}\n')
shutil.move(temp_file_path, file_path)
fd.write(f'{Answer.txt_header}\n{self.answer}\n')
def __to_file_yaml(self, file_path: pathlib.Path) -> None:
"""
@@ -473,8 +433,7 @@ class Message():
* Message.ai_yaml_key: str [Optional]
* Message.model_yaml_key: str [Optional]
"""
with tempfile.NamedTemporaryFile(dir=file_path.parent, prefix=file_path.name, mode="w", delete=False) as temp_fd:
temp_file_path = pathlib.Path(temp_fd.name)
with open(file_path, "w") as fd:
data: YamlDict = {Question.yaml_key: str(self.question)}
if self.answer:
data[Answer.yaml_key] = str(self.answer)
@@ -484,8 +443,7 @@ class Message():
data[self.model_yaml_key] = self.model
if self.tags:
data[self.tags_yaml_key] = sorted([str(tag) for tag in self.tags])
yaml.dump(data, temp_fd, sort_keys=False)
shutil.move(temp_file_path, file_path)
yaml.dump(data, fd, sort_keys=False)
def filter_tags(self, prefix: Optional[str] = None, contain: Optional[str] = None) -> set[Tag]:
"""
@@ -517,7 +475,7 @@ class Message():
Return True if all attributes match, else False.
"""
mytags = self.tags or set()
if (((mfilter.tags_or is not None or mfilter.tags_and is not None or mfilter.tags_not is not None)
if (((mfilter.tags_or or mfilter.tags_and or mfilter.tags_not)
and not match_tags(mytags, mfilter.tags_or, mfilter.tags_and, mfilter.tags_not)) # noqa: W503
or (mfilter.ai and (not self.ai or mfilter.ai != self.ai)) # noqa: W503
or (mfilter.model and (not self.model or mfilter.model != self.model)) # noqa: W503
@@ -532,34 +490,15 @@ class Message():
return False
return True
def rename_tags(self, tags_rename: set[tuple[Tag, Tag]]) -> None:
"""
Renames the given tags. The first tuple element is the old name,
the second one is the new name.
"""
if self.tags:
self.tags = rename_tags(self.tags, tags_rename)
def msg_id(self) -> str:
"""
Returns an ID that is unique throughout all messages in the same (DB) directory.
Currently this is the file name without suffix. The ID is also used for sorting
messages.
Currently this is the file name. The ID is also used for sorting messages.
"""
if self.file_path:
return self.file_path.stem
return self.file_path.name
else:
raise MessageError("Can't create file ID without a file path")
def as_dict(self) -> dict[str, Any]:
return asdict(self)
def tokens(self) -> int:
"""
Returns the nr. of AI language tokens used by this message.
If unknown, 0 is returned.
"""
if self.answer:
return self.question.tokens + self.answer.tokens
else:
return self.question.tokens
+121
View File
@@ -0,0 +1,121 @@
import yaml
import io
import pathlib
from .utils import terminal_width, append_message, message_to_chat, ChatType
from .configuration import Config
from typing import Any, Optional
def read_file(fname: pathlib.Path, tags_only: bool = False) -> dict[str, Any]:
with open(fname, "r") as fd:
tagline = fd.readline().strip().split(':', maxsplit=1)[1].strip()
# also support tags separated by ',' (old format)
separator = ',' if ',' in tagline else ' '
tags = [t.strip() for t in tagline.split(separator)]
if tags_only:
return {"tags": tags}
text = fd.read().strip().split('\n')
question_idx = text.index("=== QUESTION ===") + 1
answer_idx = text.index("==== ANSWER ====")
question = "\n".join(text[question_idx:answer_idx]).strip()
answer = "\n".join(text[answer_idx + 1:]).strip()
return {"question": question, "answer": answer, "tags": tags,
"file": fname.name}
def dump_data(data: dict[str, Any]) -> str:
with io.StringIO() as fd:
fd.write(f'TAGS: {" ".join(data["tags"])}\n')
fd.write(f'=== QUESTION ===\n{data["question"]}\n')
fd.write(f'==== ANSWER ====\n{data["answer"]}\n')
return fd.getvalue()
def write_file(fname: str, data: dict[str, Any]) -> None:
with open(fname, "w") as fd:
fd.write(f'TAGS: {" ".join(data["tags"])}\n')
fd.write(f'=== QUESTION ===\n{data["question"]}\n')
fd.write(f'==== ANSWER ====\n{data["answer"]}\n')
def save_answers(question: str,
answers: list[str],
tags: list[str],
otags: Optional[list[str]],
config: Config
) -> None:
wtags = otags or tags
num, inum = 0, 0
next_fname = pathlib.Path(str(config.db)) / '.next'
try:
with open(next_fname, 'r') as f:
num = int(f.read())
except Exception:
pass
for answer in answers:
num += 1
inum += 1
title = f'-- ANSWER {inum} '
title_end = '-' * (terminal_width() - len(title))
print(f'{title}{title_end}')
print(answer)
write_file(f"{num:04d}.txt", {"question": question, "answer": answer, "tags": wtags})
with open(next_fname, 'w') as f:
f.write(f'{num}')
def create_chat_hist(question: Optional[str],
tags: Optional[list[str]],
extags: Optional[list[str]],
config: Config,
match_all_tags: bool = False,
with_tags: bool = False,
with_file: bool = False
) -> ChatType:
chat: ChatType = []
append_message(chat, 'system', str(config.system).strip())
for file in sorted(pathlib.Path(str(config.db)).iterdir()):
if file.suffix == '.yaml':
with open(file, 'r') as f:
data = yaml.load(f, Loader=yaml.FullLoader)
data['file'] = file.name
elif file.suffix == '.txt':
data = read_file(file)
else:
continue
data_tags = set(data.get('tags', []))
tags_match: bool
if match_all_tags:
tags_match = not tags or set(tags).issubset(data_tags)
else:
tags_match = not tags or bool(data_tags.intersection(tags))
extags_do_not_match = \
not extags or not data_tags.intersection(extags)
if tags_match and extags_do_not_match:
message_to_chat(data, chat, with_tags, with_file)
if question:
append_message(chat, 'user', question)
return chat
def get_tags(config: Config, prefix: Optional[str]) -> list[str]:
result = []
for file in sorted(pathlib.Path(str(config.db)).iterdir()):
if file.suffix == '.yaml':
with open(file, 'r') as f:
data = yaml.load(f, Loader=yaml.FullLoader)
elif file.suffix == '.txt':
data = read_file(file, tags_only=True)
else:
continue
for tag in data.get('tags', []):
if prefix and len(prefix) > 0:
if tag.startswith(prefix):
result.append(tag)
else:
result.append(tag)
return result
def get_tags_unique(config: Config, prefix: Optional[str]) -> list[str]:
return list(set(get_tags(config, prefix)))
+1 -2
View File
@@ -77,8 +77,7 @@ def match_tags(tags: set[Tag], tags_or: Optional[set[Tag]], tags_and: Optional[s
i. e. you can select a TagLine if it either contains one of the tags in 'tags_or'
or all of the tags in 'tags_and' but it must never contain any of the tags in
'tags_not'. If 'tags_or' and 'tags_and' are 'None', they match all tags (tag
exclusion is still done if 'tags_not' is not 'None'). If they are empty (set()),
they match no tags.
exclusion is still done if 'tags_not' is not 'None').
"""
required_tags_present = False
excluded_tags_missing = False
+85
View File
@@ -0,0 +1,85 @@
import shutil
from pprint import PrettyPrinter
from typing import Any
ChatType = list[dict[str, str]]
def terminal_width() -> int:
return shutil.get_terminal_size().columns
def pp(*args: Any, **kwargs: Any) -> None:
return PrettyPrinter(width=terminal_width()).pprint(*args, **kwargs)
def print_tag_args(tags: list[str], extags: list[str], otags: list[str]) -> None:
"""
Prints the tags specified in the given args.
"""
printed_messages = []
if tags:
printed_messages.append(f"Tags: {' '.join(tags)}")
if extags:
printed_messages.append(f"Excluding tags: {' '.join(extags)}")
if otags:
printed_messages.append(f"Output tags: {' '.join(otags)}")
if printed_messages:
print("\n".join(printed_messages))
print()
def append_message(chat: ChatType,
role: str,
content: str
) -> None:
chat.append({'role': role, 'content': content.replace("''", "'")})
def message_to_chat(message: dict[str, str],
chat: ChatType,
with_tags: bool = False,
with_file: bool = False
) -> None:
append_message(chat, 'user', message['question'])
append_message(chat, 'assistant', message['answer'])
if with_tags:
tags = " ".join(message['tags'])
append_message(chat, 'tags', tags)
if with_file:
append_message(chat, 'file', message['file'])
def display_source_code(content: str) -> None:
try:
content_start = content.index('```')
content_end = content.rindex('```')
if content_start + 3 < content_end:
print(content[content_start + 3:content_end].strip())
except ValueError:
pass
def print_chat_hist(chat: ChatType, dump: bool = False, source_code: bool = False) -> None:
if dump:
pp(chat)
return
for message in chat:
text_too_long = len(message['content']) > terminal_width() - len(message['role']) - 2
if source_code:
display_source_code(message['content'])
continue
if message['role'] == 'user':
print('-' * terminal_width())
if text_too_long:
print(f"{message['role'].upper()}:")
print(message['content'])
else:
print(f"{message['role'].upper()}: {message['content']}")
def print_tags_frequency(tags: list[str]) -> None:
for tag in sorted(set(tags)):
print(f"- {tag}: {tags.count(tag)}")
-56
View File
@@ -1,56 +0,0 @@
<?php
$secret_key = '123';
// check for POST request
if ($_SERVER['REQUEST_METHOD'] != 'POST') {
error_log('FAILED - not POST - '. $_SERVER['REQUEST_METHOD']);
exit();
}
// get content type
$content_type = isset($_SERVER['CONTENT_TYPE']) ? strtolower(trim($_SERVER['CONTENT_TYPE'])) : '';
if ($content_type != 'application/json') {
error_log('FAILED - not application/json - '. $content_type);
exit();
}
// get payload
$payload = trim(file_get_contents("php://input"));
if (empty($payload)) {
error_log('FAILED - no payload');
exit();
}
// get header signature
$header_signature = isset($_SERVER['HTTP_X_GITEA_SIGNATURE']) ? $_SERVER['HTTP_X_GITEA_SIGNATURE'] : '';
if (empty($header_signature)) {
error_log('FAILED - header signature missing');
exit();
}
// calculate payload signature
$payload_signature = hash_hmac('sha256', $payload, $secret_key, false);
// check payload signature against header signature
if ($header_signature !== $payload_signature) {
error_log('FAILED - payload signature');
exit();
}
// convert json to array
$decoded = json_decode($payload, true);
// check for json decode errors
if (json_last_error() !== JSON_ERROR_NONE) {
error_log('FAILED - json decode - '. json_last_error());
exit();
}
// success, do something
$output = shell_exec('/home/kaizen/repos/ChatMastermind/hooks/push_hook.sh');
echo "$output";
?>
-8
View File
@@ -1,8 +0,0 @@
#!/usr/bin/bash
. /home/kaizen/.bashrc
set -e
cd /home/kaizen/repos/ChatMastermind
git pull
pre-commit run -a
pytest
+2 -2
View File
@@ -12,7 +12,7 @@ setup(
long_description=long_description,
long_description_content_type="text/markdown",
url="https://github.com/ok2/ChatMastermind",
packages=find_packages() + ["chatmastermind.ais", "chatmastermind.commands"],
packages=find_packages(),
classifiers=[
"Development Status :: 3 - Alpha",
"Environment :: Console",
@@ -32,7 +32,7 @@ setup(
"openai",
"PyYAML",
"argcomplete",
"pytest",
"pytest"
],
python_requires=">=3.9",
test_suite="tests",
-48
View File
@@ -1,48 +0,0 @@
import argparse
import unittest
from unittest.mock import MagicMock
from chatmastermind.ai_factory import create_ai
from chatmastermind.configuration import Config
from chatmastermind.ai import AIError
from chatmastermind.ais.openai import OpenAI
class TestCreateAI(unittest.TestCase):
def setUp(self) -> None:
self.args = MagicMock(spec=argparse.Namespace)
self.args.AI = 'myopenai'
self.args.model = None
self.args.max_tokens = None
self.args.temperature = None
def test_create_ai_from_args(self) -> None:
# Create an AI with the default configuration
config = Config()
self.args.AI = 'myopenai'
ai = create_ai(self.args, config)
self.assertIsInstance(ai, OpenAI)
def test_create_ai_from_default(self) -> None:
self.args.AI = None
# Create an AI with the default configuration
config = Config()
ai = create_ai(self.args, config)
self.assertIsInstance(ai, OpenAI)
def test_create_empty_ai_error(self) -> None:
self.args.AI = None
# Create Config with empty AIs
config = Config()
config.ais = {}
# Call create_ai function and assert that it raises AIError
with self.assertRaises(AIError):
create_ai(self.args, config)
def test_create_unsupported_ai_error(self) -> None:
# Mock argparse.Namespace with ai='invalid_ai'
self.args.AI = 'invalid_ai'
# Create default Config
config = Config()
# Call create_ai function and assert that it raises AIError
with self.assertRaises(AIError):
create_ai(self.args, config)
-81
View File
@@ -1,81 +0,0 @@
import unittest
from unittest import mock
from chatmastermind.ais.openai import OpenAI
from chatmastermind.message import Message, Question, Answer
from chatmastermind.chat import Chat
from chatmastermind.ai import AIResponse, Tokens
from chatmastermind.configuration import OpenAIConfig
class OpenAITest(unittest.TestCase):
@mock.patch('openai.ChatCompletion.create')
def test_request(self, mock_create: mock.MagicMock) -> None:
# Create a test instance of OpenAI
config = OpenAIConfig()
openai = OpenAI(config)
# Set up the mock response from openai.ChatCompletion.create
mock_response = {
'choices': [
{
'message': {
'content': 'Answer 1'
}
},
{
'message': {
'content': 'Answer 2'
}
}
],
'usage': {
'prompt_tokens': 10,
'completion_tokens': 20,
'total_tokens': 30
}
}
mock_create.return_value = mock_response
# Create test data
question = Message(Question('Question'))
chat = Chat([
Message(Question('Question 1'), answer=Answer('Answer 1')),
Message(Question('Question 2'), answer=Answer('Answer 2')),
# add message without an answer -> expect to be skipped
Message(Question('Question 3'))
])
# Make the request
response = openai.request(question, chat, num_answers=2)
# Assert the AIResponse
self.assertIsInstance(response, AIResponse)
self.assertEqual(len(response.messages), 2)
self.assertEqual(response.messages[0].answer, 'Answer 1')
self.assertEqual(response.messages[1].answer, 'Answer 2')
self.assertIsNotNone(response.tokens)
self.assertIsInstance(response.tokens, Tokens)
assert response.tokens
self.assertEqual(response.tokens.prompt, 10)
self.assertEqual(response.tokens.completion, 20)
self.assertEqual(response.tokens.total, 30)
# Assert the mock call to openai.ChatCompletion.create
mock_create.assert_called_once_with(
model=f'{config.model}',
messages=[
{'role': 'system', 'content': f'{config.system}'},
{'role': 'user', 'content': 'Question 1'},
{'role': 'assistant', 'content': 'Answer 1'},
{'role': 'user', 'content': 'Question 2'},
{'role': 'assistant', 'content': 'Answer 2'},
{'role': 'user', 'content': 'Question'}
],
temperature=config.temperature,
max_tokens=config.max_tokens,
top_p=config.top_p,
n=2,
frequency_penalty=config.frequency_penalty,
presence_penalty=config.presence_penalty
)
+46 -453
View File
@@ -1,128 +1,72 @@
import unittest
import pathlib
import tempfile
import time
import yaml
from io import StringIO
from unittest.mock import patch
from chatmastermind.tags import TagLine
from chatmastermind.message import Message, Question, Answer, Tag, MessageFilter
from chatmastermind.chat import Chat, ChatDB, ChatError
from chatmastermind.chat import Chat, ChatDB, terminal_width
from .test_main import CmmTestCase
class TestChat(unittest.TestCase):
class TestChat(CmmTestCase):
def setUp(self) -> None:
self.chat = Chat([])
self.message1 = Message(Question('Question 1'),
Answer('Answer 1'),
{Tag('atag1'), Tag('btag2')},
{Tag('atag1')},
file_path=pathlib.Path('0001.txt'))
self.message2 = Message(Question('Question 2'),
Answer('Answer 2'),
{Tag('btag2')},
file_path=pathlib.Path('0002.txt'))
self.maxDiff = None
def test_unique_id(self) -> None:
# test with two identical messages
self.chat.msg_add([self.message1, self.message1])
self.assertSequenceEqual(self.chat.messages, [self.message1, self.message1])
self.chat.msg_unique_id()
self.assertSequenceEqual(self.chat.messages, [self.message1])
# test with two different messages
self.chat.msg_add([self.message2])
self.chat.msg_unique_id()
self.assertSequenceEqual(self.chat.messages, [self.message1, self.message2])
def test_unique_content(self) -> None:
# test with two identical messages
self.chat.msg_add([self.message1, self.message1])
self.assertSequenceEqual(self.chat.messages, [self.message1, self.message1])
self.chat.msg_unique_content()
self.assertSequenceEqual(self.chat.messages, [self.message1])
# test with two different messages
self.chat.msg_add([self.message2])
self.chat.msg_unique_content()
self.assertSequenceEqual(self.chat.messages, [self.message1, self.message2])
def test_filter(self) -> None:
self.chat.msg_add([self.message1, self.message2])
self.chat.msg_filter(MessageFilter(answer_contains='Answer 1'))
self.chat.add_msgs([self.message1, self.message2])
self.chat.filter(MessageFilter(answer_contains='Answer 1'))
self.assertEqual(len(self.chat.messages), 1)
self.assertEqual(self.chat.messages[0].question, 'Question 1')
def test_sort(self) -> None:
self.chat.msg_add([self.message2, self.message1])
self.chat.msg_sort()
self.chat.add_msgs([self.message2, self.message1])
self.chat.sort()
self.assertEqual(self.chat.messages[0].question, 'Question 1')
self.assertEqual(self.chat.messages[1].question, 'Question 2')
self.chat.msg_sort(reverse=True)
self.chat.sort(reverse=True)
self.assertEqual(self.chat.messages[0].question, 'Question 2')
self.assertEqual(self.chat.messages[1].question, 'Question 1')
def test_clear(self) -> None:
self.chat.msg_add([self.message1])
self.chat.msg_clear()
self.chat.add_msgs([self.message1])
self.chat.clear()
self.assertEqual(len(self.chat.messages), 0)
def test_add_messages(self) -> None:
self.chat.msg_add([self.message1, self.message2])
def test_add_msgs(self) -> None:
self.chat.add_msgs([self.message1, self.message2])
self.assertEqual(len(self.chat.messages), 2)
self.assertEqual(self.chat.messages[0].question, 'Question 1')
self.assertEqual(self.chat.messages[1].question, 'Question 2')
def test_tags(self) -> None:
self.chat.msg_add([self.message1, self.message2])
tags_all = self.chat.msg_tags()
self.chat.add_msgs([self.message1, self.message2])
tags_all = self.chat.tags()
self.assertSetEqual(tags_all, {Tag('atag1'), Tag('btag2')})
tags_pref = self.chat.msg_tags(prefix='a')
tags_pref = self.chat.tags(prefix='a')
self.assertSetEqual(tags_pref, {Tag('atag1')})
tags_cont = self.chat.msg_tags(contain='2')
tags_cont = self.chat.tags(contain='2')
self.assertSetEqual(tags_cont, {Tag('btag2')})
def test_tags_frequency(self) -> None:
self.chat.msg_add([self.message1, self.message2])
tags_freq = self.chat.msg_tags_frequency()
self.assertDictEqual(tags_freq, {'atag1': 1, 'btag2': 2})
def test_find_remove_messages(self) -> None:
self.chat.msg_add([self.message1, self.message2])
msgs = self.chat.msg_find(['0001.txt'])
self.assertListEqual(msgs, [self.message1])
msgs = self.chat.msg_find(['0001.txt', '0002.txt'])
self.assertListEqual(msgs, [self.message1, self.message2])
# add new Message with full path
message3 = Message(Question('Question 2'),
Answer('Answer 2'),
{Tag('btag2')},
file_path=pathlib.Path('/foo/bla/0003.txt'))
self.chat.msg_add([message3])
# find new Message by full path
msgs = self.chat.msg_find(['/foo/bla/0003.txt'])
self.assertListEqual(msgs, [message3])
# find Message with full path only by filename
msgs = self.chat.msg_find(['0003.txt'])
self.assertListEqual(msgs, [message3])
# remove last message
self.chat.msg_remove(['0003.txt'])
self.assertListEqual(self.chat.messages, [self.message1, self.message2])
def test_latest_message(self) -> None:
self.assertIsNone(self.chat.msg_latest())
self.chat.msg_add([self.message1])
self.assertEqual(self.chat.msg_latest(), self.message1)
self.chat.msg_add([self.message2])
self.assertEqual(self.chat.msg_latest(), self.message2)
@patch('sys.stdout', new_callable=StringIO)
def test_print(self, mock_stdout: StringIO) -> None:
self.chat.msg_add([self.message1, self.message2])
self.chat.add_msgs([self.message1, self.message2])
self.chat.print(paged=False)
expected_output = f"""{Question.txt_header}
expected_output = f"""{'-'*terminal_width()}
{Question.txt_header}
Question 1
{Answer.txt_header}
Answer 1
{'-'*terminal_width()}
{Question.txt_header}
Question 2
{Answer.txt_header}
@@ -132,25 +76,27 @@ Answer 2
@patch('sys.stdout', new_callable=StringIO)
def test_print_with_tags_and_file(self, mock_stdout: StringIO) -> None:
self.chat.msg_add([self.message1, self.message2])
self.chat.print(paged=False, with_tags=True, with_files=True)
expected_output = f"""{TagLine.prefix} atag1 btag2
FILE: 0001.txt
self.chat.add_msgs([self.message1, self.message2])
self.chat.print(paged=False, with_tags=True, with_file=True)
expected_output = f"""{'-'*terminal_width()}
{Question.txt_header}
Question 1
{Answer.txt_header}
Answer 1
{TagLine.prefix} btag2
FILE: 0002.txt
{TagLine.prefix} atag1
FILE: 0001.txt
{'-'*terminal_width()}
{Question.txt_header}
Question 2
{Answer.txt_header}
Answer 2
{TagLine.prefix} btag2
FILE: 0002.txt
"""
self.assertEqual(mock_stdout.getvalue(), expected_output)
class TestChatDB(unittest.TestCase):
class TestChatDB(CmmTestCase):
def setUp(self) -> None:
self.db_path = tempfile.TemporaryDirectory()
self.cache_path = tempfile.TemporaryDirectory()
@@ -176,45 +122,13 @@ class TestChatDB(unittest.TestCase):
self.message2.to_file(pathlib.Path(self.db_path.name, '0002.yaml'))
self.message3.to_file(pathlib.Path(self.db_path.name, '0003.txt'))
self.message4.to_file(pathlib.Path(self.db_path.name, '0004.yaml'))
# make the next FID match the current state
next_fname = pathlib.Path(self.db_path.name) / '.next'
with open(next_fname, 'w') as f:
f.write('4')
# add some "trash" in order to test if it's correctly handled / ignored
self.trash_files = ['.config.yaml', 'foo.yaml', 'bla.txt']
for file in self.trash_files:
with open(pathlib.Path(self.db_path.name) / file, 'w') as f:
f.write('test trash')
# also create a file with actual yaml content
with open(pathlib.Path(self.db_path.name) / 'content.yaml', 'w') as f:
yaml.dump({'key': 'value'}, f)
self.trash_files.append('content.yaml')
self.maxDiff = None
def message_list(self, tmp_dir: tempfile.TemporaryDirectory) -> list[pathlib.Path]:
"""
List all Message files in the given TemporaryDirectory.
"""
# exclude '.next'
return [f for f in pathlib.Path(tmp_dir.name).glob('*.[ty]*') if f.name not in self.trash_files]
def tearDown(self) -> None:
self.db_path.cleanup()
self.cache_path.cleanup()
pass
def test_validate(self) -> None:
duplicate_message = Message(Question('Question 4'),
Answer('Answer 4'),
{Tag('tag4')},
file_path=pathlib.Path('0004.txt'))
duplicate_message.to_file(pathlib.Path(self.db_path.name, '0004.txt'))
with self.assertRaises(ChatError) as cm:
ChatDB.from_dir(pathlib.Path(self.cache_path.name),
pathlib.Path(self.db_path.name))
self.assertEqual(str(cm.exception), "Validation failed")
def test_from_dir(self) -> None:
def test_chat_db_from_dir(self) -> None:
chat_db = ChatDB.from_dir(pathlib.Path(self.cache_path.name),
pathlib.Path(self.db_path.name))
self.assertEqual(len(chat_db.messages), 4)
@@ -230,7 +144,7 @@ class TestChatDB(unittest.TestCase):
self.assertEqual(chat_db.messages[3].file_path,
pathlib.Path(self.db_path.name, '0004.yaml'))
def test_from_dir_glob(self) -> None:
def test_chat_db_from_dir_glob(self) -> None:
chat_db = ChatDB.from_dir(pathlib.Path(self.cache_path.name),
pathlib.Path(self.db_path.name),
glob='*.txt')
@@ -242,25 +156,7 @@ class TestChatDB(unittest.TestCase):
self.assertEqual(chat_db.messages[1].file_path,
pathlib.Path(self.db_path.name, '0003.txt'))
def test_from_dir_filter_tags(self) -> None:
chat_db = ChatDB.from_dir(pathlib.Path(self.cache_path.name),
pathlib.Path(self.db_path.name),
mfilter=MessageFilter(tags_or={Tag('tag1')}))
self.assertEqual(len(chat_db.messages), 1)
self.assertEqual(chat_db.cache_path, pathlib.Path(self.cache_path.name))
self.assertEqual(chat_db.db_path, pathlib.Path(self.db_path.name))
self.assertEqual(chat_db.messages[0].file_path,
pathlib.Path(self.db_path.name, '0001.txt'))
def test_from_dir_filter_tags_empty(self) -> None:
chat_db = ChatDB.from_dir(pathlib.Path(self.cache_path.name),
pathlib.Path(self.db_path.name),
mfilter=MessageFilter(tags_or=set(),
tags_and=set(),
tags_not=set()))
self.assertEqual(len(chat_db.messages), 0)
def test_from_dir_filter_answer(self) -> None:
def test_chat_db_filter(self) -> None:
chat_db = ChatDB.from_dir(pathlib.Path(self.cache_path.name),
pathlib.Path(self.db_path.name),
mfilter=MessageFilter(answer_contains='Answer 2'))
@@ -271,7 +167,7 @@ class TestChatDB(unittest.TestCase):
pathlib.Path(self.db_path.name, '0002.yaml'))
self.assertEqual(chat_db.messages[0].answer, 'Answer 2')
def test_from_messages(self) -> None:
def test_chat_db_from_messges(self) -> None:
chat_db = ChatDB.from_messages(pathlib.Path(self.cache_path.name),
pathlib.Path(self.db_path.name),
messages=[self.message1, self.message2,
@@ -280,35 +176,16 @@ class TestChatDB(unittest.TestCase):
self.assertEqual(chat_db.cache_path, pathlib.Path(self.cache_path.name))
self.assertEqual(chat_db.db_path, pathlib.Path(self.db_path.name))
def test_fids(self) -> None:
def test_chat_db_fids(self) -> None:
chat_db = ChatDB.from_dir(pathlib.Path(self.cache_path.name),
pathlib.Path(self.db_path.name))
self.assertEqual(chat_db.get_next_fid(), 5)
self.assertEqual(chat_db.get_next_fid(), 6)
self.assertEqual(chat_db.get_next_fid(), 7)
with open(chat_db.next_path, 'r') as f:
self.assertEqual(f.read(), '7')
self.assertEqual(chat_db.get_next_fid(), 1)
self.assertEqual(chat_db.get_next_fid(), 2)
self.assertEqual(chat_db.get_next_fid(), 3)
with open(chat_db.next_fname, 'r') as f:
self.assertEqual(f.read(), '3')
def test_msg_in_db_or_cache(self) -> None:
# create a new ChatDB instance
chat_db = ChatDB.from_dir(pathlib.Path(self.cache_path.name),
pathlib.Path(self.db_path.name))
self.assertTrue(chat_db.msg_in_db(self.message1))
self.assertTrue(chat_db.msg_in_db(str(self.message1.file_path)))
self.assertTrue(chat_db.msg_in_db(self.message1.msg_id()))
self.assertFalse(chat_db.msg_in_cache(self.message1))
self.assertFalse(chat_db.msg_in_cache(str(self.message1.file_path)))
self.assertFalse(chat_db.msg_in_cache(self.message1.msg_id()))
# add new message to the cache dir
cache_message = Message(question=Question("Question 1"),
answer=Answer("Answer 1"))
chat_db.cache_add([cache_message])
self.assertTrue(chat_db.msg_in_cache(cache_message))
self.assertTrue(chat_db.msg_in_cache(cache_message.msg_id()))
self.assertFalse(chat_db.msg_in_db(cache_message))
self.assertFalse(chat_db.msg_in_db(str(cache_message.file_path)))
def test_db_write(self) -> None:
def test_chat_db_write(self) -> None:
# create a new ChatDB instance
chat_db = ChatDB.from_dir(pathlib.Path(self.cache_path.name),
pathlib.Path(self.db_path.name))
@@ -319,9 +196,9 @@ class TestChatDB(unittest.TestCase):
self.assertEqual(chat_db.messages[3].file_path, pathlib.Path(self.db_path.name, '0004.yaml'))
# write the messages to the cache directory
chat_db.cache_write()
chat_db.write_cache()
# check if the written files are in the cache directory
cache_dir_files = self.message_list(self.cache_path)
cache_dir_files = list(pathlib.Path(self.cache_path.name).glob('*'))
self.assertEqual(len(cache_dir_files), 4)
self.assertIn(pathlib.Path(self.cache_path.name, '0001.txt'), cache_dir_files)
self.assertIn(pathlib.Path(self.cache_path.name, '0002.yaml'), cache_dir_files)
@@ -334,14 +211,14 @@ class TestChatDB(unittest.TestCase):
self.assertEqual(chat_db.messages[3].file_path, pathlib.Path(self.cache_path.name, '0004.yaml'))
# check the timestamp of the files in the DB directory
db_dir_files = self.message_list(self.db_path)
db_dir_files = list(pathlib.Path(self.db_path.name).glob('*'))
self.assertEqual(len(db_dir_files), 4)
old_timestamps = {file: file.stat().st_mtime for file in db_dir_files}
# overwrite the messages in the db directory
time.sleep(0.05)
chat_db.db_write()
chat_db.write_db()
# check if the written files are in the DB directory
db_dir_files = self.message_list(self.db_path)
db_dir_files = list(pathlib.Path(self.db_path.name).glob('*'))
self.assertEqual(len(db_dir_files), 4)
self.assertIn(pathlib.Path(self.db_path.name, '0001.txt'), db_dir_files)
self.assertIn(pathlib.Path(self.db_path.name, '0002.yaml'), db_dir_files)
@@ -355,287 +232,3 @@ class TestChatDB(unittest.TestCase):
self.assertEqual(chat_db.messages[1].file_path, pathlib.Path(self.db_path.name, '0002.yaml'))
self.assertEqual(chat_db.messages[2].file_path, pathlib.Path(self.db_path.name, '0003.txt'))
self.assertEqual(chat_db.messages[3].file_path, pathlib.Path(self.db_path.name, '0004.yaml'))
def test_db_read(self) -> None:
# create a new ChatDB instance
chat_db = ChatDB.from_dir(pathlib.Path(self.cache_path.name),
pathlib.Path(self.db_path.name))
self.assertEqual(len(chat_db.messages), 4)
# create 2 new files in the DB directory
new_message1 = Message(Question('Question 5'),
Answer('Answer 5'),
{Tag('tag5')})
new_message2 = Message(Question('Question 6'),
Answer('Answer 6'),
{Tag('tag6')})
new_message1.to_file(pathlib.Path(self.db_path.name, '0005.txt'))
new_message2.to_file(pathlib.Path(self.db_path.name, '0006.yaml'))
# read and check them
chat_db.db_read()
self.assertEqual(len(chat_db.messages), 6)
self.assertEqual(chat_db.messages[4].file_path, pathlib.Path(self.db_path.name, '0005.txt'))
self.assertEqual(chat_db.messages[5].file_path, pathlib.Path(self.db_path.name, '0006.yaml'))
# create 2 new files in the cache directory
new_message3 = Message(Question('Question 7'),
Answer('Answer 5'),
{Tag('tag7')})
new_message4 = Message(Question('Question 8'),
Answer('Answer 6'),
{Tag('tag8')})
new_message3.to_file(pathlib.Path(self.cache_path.name, '0007.txt'))
new_message4.to_file(pathlib.Path(self.cache_path.name, '0008.yaml'))
# read and check them
chat_db.cache_read()
self.assertEqual(len(chat_db.messages), 8)
# check that the new message have the cache dir path
self.assertEqual(chat_db.messages[6].file_path, pathlib.Path(self.cache_path.name, '0007.txt'))
self.assertEqual(chat_db.messages[7].file_path, pathlib.Path(self.cache_path.name, '0008.yaml'))
# an the old ones keep their path (since they have not been replaced)
self.assertEqual(chat_db.messages[4].file_path, pathlib.Path(self.db_path.name, '0005.txt'))
self.assertEqual(chat_db.messages[5].file_path, pathlib.Path(self.db_path.name, '0006.yaml'))
# now overwrite two messages in the DB directory
new_message1.question = Question('New Question 1')
new_message2.question = Question('New Question 2')
new_message1.to_file(pathlib.Path(self.db_path.name, '0005.txt'))
new_message2.to_file(pathlib.Path(self.db_path.name, '0006.yaml'))
# read from the DB dir and check if the modified messages have been updated
chat_db.db_read()
self.assertEqual(len(chat_db.messages), 8)
self.assertEqual(chat_db.messages[4].question, 'New Question 1')
self.assertEqual(chat_db.messages[5].question, 'New Question 2')
self.assertEqual(chat_db.messages[4].file_path, pathlib.Path(self.db_path.name, '0005.txt'))
self.assertEqual(chat_db.messages[5].file_path, pathlib.Path(self.db_path.name, '0006.yaml'))
# now write the messages from the cache to the DB directory
new_message3.to_file(pathlib.Path(self.db_path.name, '0007.txt'))
new_message4.to_file(pathlib.Path(self.db_path.name, '0008.yaml'))
# read and check them
chat_db.db_read()
self.assertEqual(len(chat_db.messages), 8)
# check that they now have the DB path
self.assertEqual(chat_db.messages[6].file_path, pathlib.Path(self.db_path.name, '0007.txt'))
self.assertEqual(chat_db.messages[7].file_path, pathlib.Path(self.db_path.name, '0008.yaml'))
def test_cache_clear(self) -> None:
# create a new ChatDB instance
chat_db = ChatDB.from_dir(pathlib.Path(self.cache_path.name),
pathlib.Path(self.db_path.name))
# check that Message.file_path is correct
self.assertEqual(chat_db.messages[0].file_path, pathlib.Path(self.db_path.name, '0001.txt'))
self.assertEqual(chat_db.messages[1].file_path, pathlib.Path(self.db_path.name, '0002.yaml'))
self.assertEqual(chat_db.messages[2].file_path, pathlib.Path(self.db_path.name, '0003.txt'))
self.assertEqual(chat_db.messages[3].file_path, pathlib.Path(self.db_path.name, '0004.yaml'))
# write the messages to the cache directory
chat_db.cache_write()
# check if the written files are in the cache directory
cache_dir_files = self.message_list(self.cache_path)
self.assertEqual(len(cache_dir_files), 4)
# now rewrite them to the DB dir and check for modified paths
chat_db.db_write()
db_dir_files = self.message_list(self.db_path)
self.assertEqual(len(db_dir_files), 4)
self.assertIn(pathlib.Path(self.db_path.name, '0001.txt'), db_dir_files)
self.assertIn(pathlib.Path(self.db_path.name, '0002.yaml'), db_dir_files)
self.assertIn(pathlib.Path(self.db_path.name, '0003.txt'), db_dir_files)
self.assertIn(pathlib.Path(self.db_path.name, '0004.yaml'), db_dir_files)
# add a new message with empty file_path
message_empty = Message(question=Question("What the hell am I doing here?"),
answer=Answer("You don't belong here!"))
# and one for the cache dir
message_cache = Message(question=Question("What the hell am I doing here?"),
answer=Answer("You're a creep!"),
file_path=pathlib.Path(self.cache_path.name, '0005.txt'))
chat_db.msg_add([message_empty, message_cache])
# clear the cache and check the cache dir
chat_db.cache_clear()
cache_dir_files = self.message_list(self.cache_path)
self.assertEqual(len(cache_dir_files), 0)
# make sure that the DB messages (and the new message) are still there
self.assertEqual(len(chat_db.messages), 5)
db_dir_files = self.message_list(self.db_path)
self.assertEqual(len(db_dir_files), 4)
# but not the message with the cache dir path
self.assertFalse(any(m.file_path == message_cache.file_path for m in chat_db.messages))
def test_add(self) -> None:
# create a new ChatDB instance
chat_db = ChatDB.from_dir(pathlib.Path(self.cache_path.name),
pathlib.Path(self.db_path.name))
db_dir_files = self.message_list(self.db_path)
self.assertEqual(len(db_dir_files), 4)
# add new messages to the cache dir
message1 = Message(question=Question("Question 1"),
answer=Answer("Answer 1"))
chat_db.cache_add([message1])
# check if the file_path has been correctly set
self.assertIsNotNone(message1.file_path)
self.assertEqual(message1.file_path.parent, pathlib.Path(self.cache_path.name)) # type: ignore [union-attr]
cache_dir_files = self.message_list(self.cache_path)
self.assertEqual(len(cache_dir_files), 1)
# add new messages to the DB dir
message2 = Message(question=Question("Question 2"),
answer=Answer("Answer 2"))
chat_db.db_add([message2])
# check if the file_path has been correctly set
self.assertIsNotNone(message2.file_path)
self.assertEqual(message2.file_path.parent, pathlib.Path(self.db_path.name)) # type: ignore [union-attr]
db_dir_files = self.message_list(self.db_path)
self.assertEqual(len(db_dir_files), 5)
with self.assertRaises(ChatError):
chat_db.cache_add([Message(Question("?"), file_path=pathlib.Path("foo"))])
def test_msg_write(self) -> None:
# create a new ChatDB instance
chat_db = ChatDB.from_dir(pathlib.Path(self.cache_path.name),
pathlib.Path(self.db_path.name))
db_dir_files = self.message_list(self.db_path)
self.assertEqual(len(db_dir_files), 4)
cache_dir_files = self.message_list(self.cache_path)
self.assertEqual(len(cache_dir_files), 0)
# try to write a message without a valid file_path
message = Message(question=Question("Question 1"),
answer=Answer("Answer 1"))
with self.assertRaises(ChatError):
chat_db.msg_write([message])
# write a message with a valid file_path
message.file_path = pathlib.Path(self.cache_path.name) / '123456.txt'
chat_db.msg_write([message])
cache_dir_files = self.message_list(self.cache_path)
self.assertEqual(len(cache_dir_files), 1)
self.assertIn(pathlib.Path(self.cache_path.name, '123456.txt'), cache_dir_files)
def test_msg_update(self) -> None:
# create a new ChatDB instance
chat_db = ChatDB.from_dir(pathlib.Path(self.cache_path.name),
pathlib.Path(self.db_path.name))
db_dir_files = self.message_list(self.db_path)
self.assertEqual(len(db_dir_files), 4)
cache_dir_files = self.message_list(self.cache_path)
self.assertEqual(len(cache_dir_files), 0)
message = chat_db.messages[0]
message.answer = Answer("New answer")
# update message without writing
chat_db.msg_update([message], write=False)
self.assertEqual(chat_db.messages[0].answer, Answer("New answer"))
# re-read the message and check for old content
chat_db.db_read()
self.assertEqual(chat_db.messages[0].answer, Answer("Answer 1"))
# now check with writing (message should be overwritten)
chat_db.msg_update([message], write=True)
chat_db.db_read()
self.assertEqual(chat_db.messages[0].answer, Answer("New answer"))
# test without file_path -> expect error
message1 = Message(question=Question("Question 1"),
answer=Answer("Answer 1"))
with self.assertRaises(ChatError):
chat_db.msg_update([message1])
def test_msg_find(self) -> None:
chat_db = ChatDB.from_dir(pathlib.Path(self.cache_path.name),
pathlib.Path(self.db_path.name))
# search for a DB file in memory
self.assertEqual(chat_db.msg_find([str(self.message1.file_path)], loc='mem'), [self.message1])
self.assertEqual(chat_db.msg_find([self.message1.msg_id()], loc='mem'), [self.message1])
self.assertEqual(chat_db.msg_find(['0001.txt'], loc='mem'), [self.message1])
self.assertEqual(chat_db.msg_find(['0001'], loc='mem'), [self.message1])
# and on disk
self.assertEqual(chat_db.msg_find([str(self.message2.file_path)], loc='db'), [self.message2])
self.assertEqual(chat_db.msg_find([self.message2.msg_id()], loc='db'), [self.message2])
self.assertEqual(chat_db.msg_find(['0002.yaml'], loc='db'), [self.message2])
self.assertEqual(chat_db.msg_find(['0002'], loc='db'), [self.message2])
# now search the cache -> expect empty result
self.assertEqual(chat_db.msg_find([str(self.message3.file_path)], loc='cache'), [])
self.assertEqual(chat_db.msg_find([self.message3.msg_id()], loc='cache'), [])
self.assertEqual(chat_db.msg_find(['0003.txt'], loc='cache'), [])
self.assertEqual(chat_db.msg_find(['0003'], loc='cache'), [])
# search for multiple messages
# -> search one twice, expect result to be unique
search_names = ['0001', '0002.yaml', self.message3.msg_id(), str(self.message3.file_path)]
expected_result = [self.message1, self.message2, self.message3]
result = chat_db.msg_find(search_names, loc='all')
self.assertSequenceEqual(result, expected_result)
def test_msg_latest(self) -> None:
chat_db = ChatDB.from_dir(pathlib.Path(self.cache_path.name),
pathlib.Path(self.db_path.name))
self.assertEqual(chat_db.msg_latest(loc='mem'), self.message4)
self.assertEqual(chat_db.msg_latest(loc='db'), self.message4)
self.assertEqual(chat_db.msg_latest(loc='disk'), self.message4)
self.assertEqual(chat_db.msg_latest(loc='all'), self.message4)
# the cache is currently empty:
self.assertIsNone(chat_db.msg_latest(loc='cache'))
# add new messages to the cache dir
new_message = Message(question=Question("New Question"),
answer=Answer("New Answer"))
chat_db.cache_add([new_message])
self.assertEqual(chat_db.msg_latest(loc='cache'), new_message)
self.assertEqual(chat_db.msg_latest(loc='mem'), new_message)
self.assertEqual(chat_db.msg_latest(loc='disk'), new_message)
self.assertEqual(chat_db.msg_latest(loc='all'), new_message)
# the DB does not contain the new message
self.assertEqual(chat_db.msg_latest(loc='db'), self.message4)
def test_msg_gather(self) -> None:
chat_db = ChatDB.from_dir(pathlib.Path(self.cache_path.name),
pathlib.Path(self.db_path.name))
all_messages = [self.message1, self.message2, self.message3, self.message4]
self.assertSequenceEqual(chat_db.msg_gather(loc='all'), all_messages)
self.assertSequenceEqual(chat_db.msg_gather(loc='db'), all_messages)
self.assertSequenceEqual(chat_db.msg_gather(loc='mem'), all_messages)
self.assertSequenceEqual(chat_db.msg_gather(loc='disk'), all_messages)
self.assertSequenceEqual(chat_db.msg_gather(loc='cache'), [])
# add a new message, but only to the internal list
new_message = Message(Question("What?"))
all_messages_mem = all_messages + [new_message]
chat_db.msg_add([new_message])
self.assertSequenceEqual(chat_db.msg_gather(loc='mem'), all_messages_mem)
self.assertSequenceEqual(chat_db.msg_gather(loc='all'), all_messages_mem)
# the nr. of messages on disk did not change -> expect old result
self.assertSequenceEqual(chat_db.msg_gather(loc='db'), all_messages)
self.assertSequenceEqual(chat_db.msg_gather(loc='disk'), all_messages)
self.assertSequenceEqual(chat_db.msg_gather(loc='cache'), [])
# test with MessageFilter
self.assertSequenceEqual(chat_db.msg_gather(loc='all', mfilter=MessageFilter(tags_or={Tag('tag1')})),
[self.message1])
self.assertSequenceEqual(chat_db.msg_gather(loc='disk', mfilter=MessageFilter(tags_or={Tag('tag2')})),
[self.message2])
self.assertSequenceEqual(chat_db.msg_gather(loc='cache', mfilter=MessageFilter(tags_or={Tag('tag3')})),
[])
self.assertSequenceEqual(chat_db.msg_gather(loc='mem', mfilter=MessageFilter(question_contains="What")),
[new_message])
def test_msg_move_and_gather(self) -> None:
chat_db = ChatDB.from_dir(pathlib.Path(self.cache_path.name),
pathlib.Path(self.db_path.name))
all_messages = [self.message1, self.message2, self.message3, self.message4]
self.assertSequenceEqual(chat_db.msg_gather(loc='db'), all_messages)
self.assertSequenceEqual(chat_db.msg_gather(loc='cache'), [])
# move first message to the cache
chat_db.cache_move(self.message1)
self.assertSequenceEqual(chat_db.msg_gather(loc='cache'), [self.message1])
self.assertEqual(self.message1.file_path.parent, pathlib.Path(self.cache_path.name)) # type: ignore [union-attr]
self.assertSequenceEqual(chat_db.msg_gather(loc='db'), [self.message2, self.message3, self.message4])
self.assertSequenceEqual(chat_db.msg_gather(loc='all'), all_messages)
self.assertSequenceEqual(chat_db.msg_gather(loc='disk'), all_messages)
self.assertSequenceEqual(chat_db.msg_gather(loc='mem'), all_messages)
# now move first message back to the DB
chat_db.db_move(self.message1)
self.assertSequenceEqual(chat_db.msg_gather(loc='cache'), [])
self.assertEqual(self.message1.file_path.parent, pathlib.Path(self.db_path.name)) # type: ignore [union-attr]
self.assertSequenceEqual(chat_db.msg_gather(loc='db'), all_messages)
-160
View File
@@ -1,160 +0,0 @@
import os
import unittest
import yaml
from tempfile import NamedTemporaryFile
from pathlib import Path
from typing import cast
from chatmastermind.configuration import AIConfig, OpenAIConfig, ConfigError, ai_config_instance, Config
class TestAIConfigInstance(unittest.TestCase):
def test_ai_config_instance_with_valid_name_should_return_instance_with_default_values(self) -> None:
ai_config = cast(OpenAIConfig, ai_config_instance('openai'))
ai_reference = OpenAIConfig()
self.assertEqual(ai_config.ID, ai_reference.ID)
self.assertEqual(ai_config.name, ai_reference.name)
self.assertEqual(ai_config.api_key, ai_reference.api_key)
self.assertEqual(ai_config.system, ai_reference.system)
self.assertEqual(ai_config.model, ai_reference.model)
self.assertEqual(ai_config.temperature, ai_reference.temperature)
self.assertEqual(ai_config.max_tokens, ai_reference.max_tokens)
self.assertEqual(ai_config.top_p, ai_reference.top_p)
self.assertEqual(ai_config.frequency_penalty, ai_reference.frequency_penalty)
self.assertEqual(ai_config.presence_penalty, ai_reference.presence_penalty)
def test_ai_config_instance_with_valid_name_and_configuration_should_return_instance_with_custom_values(self) -> None:
conf_dict = {
'system': 'Custom system',
'api_key': '9876543210',
'model': 'custom_model',
'max_tokens': 5000,
'temperature': 0.5,
'top_p': 0.8,
'frequency_penalty': 0.7,
'presence_penalty': 0.2
}
ai_config = cast(OpenAIConfig, ai_config_instance('openai', conf_dict))
self.assertEqual(ai_config.system, 'Custom system')
self.assertEqual(ai_config.api_key, '9876543210')
self.assertEqual(ai_config.model, 'custom_model')
self.assertEqual(ai_config.max_tokens, 5000)
self.assertAlmostEqual(ai_config.temperature, 0.5)
self.assertAlmostEqual(ai_config.top_p, 0.8)
self.assertAlmostEqual(ai_config.frequency_penalty, 0.7)
self.assertAlmostEqual(ai_config.presence_penalty, 0.2)
def test_ai_config_instance_with_invalid_name_should_raise_config_error(self) -> None:
with self.assertRaises(ConfigError):
ai_config_instance('invalid_name')
class TestConfig(unittest.TestCase):
def setUp(self) -> None:
self.test_file = NamedTemporaryFile(delete=False)
def tearDown(self) -> None:
os.remove(self.test_file.name)
def test_from_dict_should_create_config_from_dict(self) -> None:
source_dict = {
'db': './test_db/',
'ais': {
'myopenai': {
'name': 'openai',
'system': 'Custom system',
'api_key': '9876543210',
'model': 'custom_model',
'max_tokens': 5000,
'temperature': 0.5,
'top_p': 0.8,
'frequency_penalty': 0.7,
'presence_penalty': 0.2
}
}
}
config = Config.from_dict(source_dict)
self.assertEqual(config.db, './test_db/')
self.assertEqual(len(config.ais), 1)
self.assertEqual(config.ais['myopenai'].name, 'openai')
self.assertEqual(cast(OpenAIConfig, config.ais['myopenai']).system, 'Custom system')
# check that 'ID' has been added
self.assertEqual(config.ais['myopenai'].ID, 'myopenai')
def test_create_default_should_create_default_config(self) -> None:
Config.create_default(Path(self.test_file.name))
with open(self.test_file.name, 'r') as f:
default_config = yaml.load(f, Loader=yaml.FullLoader)
config_reference = Config()
self.assertEqual(default_config['db'], config_reference.db)
def test_from_file_should_load_config_from_file(self) -> None:
source_dict = {
'db': './test_db/',
'ais': {
'default': {
'name': 'openai',
'system': 'Custom system',
'api_key': '9876543210',
'model': 'custom_model',
'max_tokens': 5000,
'temperature': 0.5,
'top_p': 0.8,
'frequency_penalty': 0.7,
'presence_penalty': 0.2
}
}
}
with open(self.test_file.name, 'w') as f:
yaml.dump(source_dict, f)
config = Config.from_file(self.test_file.name)
self.assertIsInstance(config, Config)
self.assertEqual(config.db, './test_db/')
self.assertEqual(len(config.ais), 1)
self.assertIsInstance(config.ais['default'], AIConfig)
self.assertEqual(cast(OpenAIConfig, config.ais['default']).system, 'Custom system')
def test_to_file_should_save_config_to_file(self) -> None:
config = Config(
db='./test_db/',
ais={
'myopenai': OpenAIConfig(
ID='myopenai',
system='Custom system',
api_key='9876543210',
model='custom_model',
max_tokens=5000,
temperature=0.5,
top_p=0.8,
frequency_penalty=0.7,
presence_penalty=0.2
)
}
)
config.to_file(Path(self.test_file.name))
with open(self.test_file.name, 'r') as f:
saved_config = yaml.load(f, Loader=yaml.FullLoader)
self.assertEqual(saved_config['db'], './test_db/')
self.assertEqual(len(saved_config['ais']), 1)
self.assertEqual(saved_config['ais']['myopenai']['system'], 'Custom system')
def test_from_file_error_unknown_ai(self) -> None:
source_dict = {
'db': './test_db/',
'ais': {
'default': {
'name': 'foobla',
'system': 'Custom system',
'api_key': '9876543210',
'model': 'custom_model',
'max_tokens': 5000,
'temperature': 0.5,
'top_p': 0.8,
'frequency_penalty': 0.7,
'presence_penalty': 0.2
}
}
}
with open(self.test_file.name, 'w') as f:
yaml.dump(source_dict, f)
with self.assertRaises(ConfigError):
Config.from_file(self.test_file.name)
+233
View File
@@ -0,0 +1,233 @@
import unittest
import io
import pathlib
import argparse
from chatmastermind.utils import terminal_width
from chatmastermind.main import create_parser, ask_cmd
from chatmastermind.api_client import ai
from chatmastermind.configuration import Config
from chatmastermind.storage import create_chat_hist, save_answers, dump_data
from unittest import mock
from unittest.mock import patch, MagicMock, Mock, ANY
class CmmTestCase(unittest.TestCase):
"""
Base class for all cmm testcases.
"""
def dummy_config(self, db: str) -> Config:
"""
Creates a dummy configuration.
"""
return Config.from_dict(
{'system': 'dummy_system',
'db': db,
'openai': {'api_key': 'dummy_key',
'model': 'dummy_model',
'max_tokens': 4000,
'temperature': 1.0,
'top_p': 1,
'frequency_penalty': 0,
'presence_penalty': 0}}
)
class TestCreateChat(CmmTestCase):
def setUp(self) -> None:
self.config = self.dummy_config(db='test_files')
self.question = "test question"
self.tags = ['test_tag']
@patch('os.listdir')
@patch('pathlib.Path.iterdir')
@patch('builtins.open')
def test_create_chat_with_tags(self, open_mock: MagicMock, iterdir_mock: MagicMock, listdir_mock: MagicMock) -> None:
listdir_mock.return_value = ['testfile.txt']
iterdir_mock.return_value = [pathlib.Path(x) for x in listdir_mock.return_value]
open_mock.return_value.__enter__.return_value = io.StringIO(dump_data(
{'question': 'test_content', 'answer': 'some answer',
'tags': ['test_tag']}))
test_chat = create_chat_hist(self.question, self.tags, None, self.config)
self.assertEqual(len(test_chat), 4)
self.assertEqual(test_chat[0],
{'role': 'system', 'content': self.config.system})
self.assertEqual(test_chat[1],
{'role': 'user', 'content': 'test_content'})
self.assertEqual(test_chat[2],
{'role': 'assistant', 'content': 'some answer'})
self.assertEqual(test_chat[3],
{'role': 'user', 'content': self.question})
@patch('os.listdir')
@patch('pathlib.Path.iterdir')
@patch('builtins.open')
def test_create_chat_with_other_tags(self, open_mock: MagicMock, iterdir_mock: MagicMock, listdir_mock: MagicMock) -> None:
listdir_mock.return_value = ['testfile.txt']
iterdir_mock.return_value = [pathlib.Path(x) for x in listdir_mock.return_value]
open_mock.return_value.__enter__.return_value = io.StringIO(dump_data(
{'question': 'test_content', 'answer': 'some answer',
'tags': ['other_tag']}))
test_chat = create_chat_hist(self.question, self.tags, None, self.config)
self.assertEqual(len(test_chat), 2)
self.assertEqual(test_chat[0],
{'role': 'system', 'content': self.config.system})
self.assertEqual(test_chat[1],
{'role': 'user', 'content': self.question})
@patch('os.listdir')
@patch('pathlib.Path.iterdir')
@patch('builtins.open')
def test_create_chat_without_tags(self, open_mock: MagicMock, iterdir_mock: MagicMock, listdir_mock: MagicMock) -> None:
listdir_mock.return_value = ['testfile.txt', 'testfile2.txt']
iterdir_mock.return_value = [pathlib.Path(x) for x in listdir_mock.return_value]
open_mock.side_effect = (
io.StringIO(dump_data({'question': 'test_content',
'answer': 'some answer',
'tags': ['test_tag']})),
io.StringIO(dump_data({'question': 'test_content2',
'answer': 'some answer2',
'tags': ['test_tag2']})),
)
test_chat = create_chat_hist(self.question, [], None, self.config)
self.assertEqual(len(test_chat), 6)
self.assertEqual(test_chat[0],
{'role': 'system', 'content': self.config.system})
self.assertEqual(test_chat[1],
{'role': 'user', 'content': 'test_content'})
self.assertEqual(test_chat[2],
{'role': 'assistant', 'content': 'some answer'})
self.assertEqual(test_chat[3],
{'role': 'user', 'content': 'test_content2'})
self.assertEqual(test_chat[4],
{'role': 'assistant', 'content': 'some answer2'})
class TestHandleQuestion(CmmTestCase):
def setUp(self) -> None:
self.question = "test question"
self.args = argparse.Namespace(
tags=['tag1'],
extags=['extag1'],
output_tags=None,
question=[self.question],
source=None,
only_source_code=False,
number=3,
max_tokens=None,
temperature=None,
model=None,
match_all_tags=False,
with_tags=False,
with_file=False,
)
self.config = self.dummy_config(db='test_files')
@patch("chatmastermind.main.create_chat_hist", return_value="test_chat")
@patch("chatmastermind.main.print_tag_args")
@patch("chatmastermind.main.print_chat_hist")
@patch("chatmastermind.main.ai", return_value=(["answer1", "answer2", "answer3"], "test_usage"))
@patch("chatmastermind.utils.pp")
@patch("builtins.print")
def test_ask_cmd(self, mock_print: MagicMock, mock_pp: MagicMock, mock_ai: MagicMock,
mock_print_chat_hist: MagicMock, mock_print_tag_args: MagicMock,
mock_create_chat_hist: MagicMock) -> None:
open_mock = MagicMock()
with patch("chatmastermind.storage.open", open_mock):
ask_cmd(self.args, self.config)
mock_print_tag_args.assert_called_once_with(self.args.tags,
self.args.extags,
[])
mock_create_chat_hist.assert_called_once_with(self.question,
self.args.tags,
self.args.extags,
self.config,
False, False, False)
mock_print_chat_hist.assert_called_once_with('test_chat',
False,
self.args.only_source_code)
mock_ai.assert_called_with("test_chat",
self.config,
self.args.number)
expected_calls = []
for num, answer in enumerate(mock_ai.return_value[0], start=1):
title = f'-- ANSWER {num} '
title_end = '-' * (terminal_width() - len(title))
expected_calls.append(((f'{title}{title_end}',),))
expected_calls.append(((answer,),))
expected_calls.append((("-" * terminal_width(),),))
expected_calls.append(((f"Usage: {mock_ai.return_value[1]}",),))
self.assertEqual(mock_print.call_args_list, expected_calls)
open_expected_calls = list([mock.call(f"{num:04d}.txt", "w") for num in range(2, 5)])
open_mock.assert_has_calls(open_expected_calls, any_order=True)
class TestSaveAnswers(CmmTestCase):
@mock.patch('builtins.open')
@mock.patch('chatmastermind.storage.print')
def test_save_answers(self, print_mock: MagicMock, open_mock: MagicMock) -> None:
question = "Test question?"
answers = ["Answer 1", "Answer 2"]
tags = ["tag1", "tag2"]
otags = ["otag1", "otag2"]
config = self.dummy_config(db='test_db')
with mock.patch('chatmastermind.storage.pathlib.Path.exists', return_value=True), \
mock.patch('chatmastermind.storage.yaml.dump'), \
mock.patch('io.StringIO') as stringio_mock:
stringio_instance = stringio_mock.return_value
stringio_instance.getvalue.side_effect = ["question", "answer1", "answer2"]
save_answers(question, answers, tags, otags, config)
open_calls = [
mock.call(pathlib.Path('test_db/.next'), 'r'),
mock.call(pathlib.Path('test_db/.next'), 'w'),
]
open_mock.assert_has_calls(open_calls, any_order=True)
class TestAI(CmmTestCase):
@patch("openai.ChatCompletion.create")
def test_ai(self, mock_create: MagicMock) -> None:
mock_create.return_value = {
'choices': [
{'message': {'content': 'response_text_1'}},
{'message': {'content': 'response_text_2'}}
],
'usage': {'tokens': 10}
}
chat = [{"role": "system", "content": "hello ai"}]
config = self.dummy_config(db='dummy')
config.openai.model = "text-davinci-002"
config.openai.max_tokens = 150
config.openai.temperature = 0.5
result = ai(chat, config, 2)
expected_result = (['response_text_1', 'response_text_2'],
{'tokens': 10})
self.assertEqual(result, expected_result)
class TestCreateParser(CmmTestCase):
def test_create_parser(self) -> None:
with patch('argparse.ArgumentParser.add_subparsers') as mock_add_subparsers:
mock_cmdparser = Mock()
mock_add_subparsers.return_value = mock_cmdparser
parser = create_parser()
self.assertIsInstance(parser, argparse.ArgumentParser)
mock_add_subparsers.assert_called_once_with(dest='command', title='commands', description='supported commands', required=True)
mock_cmdparser.add_parser.assert_any_call('ask', parents=ANY, help=ANY, aliases=ANY)
mock_cmdparser.add_parser.assert_any_call('hist', parents=ANY, help=ANY, aliases=ANY)
mock_cmdparser.add_parser.assert_any_call('tag', help=ANY, aliases=ANY)
mock_cmdparser.add_parser.assert_any_call('config', help=ANY, aliases=ANY)
mock_cmdparser.add_parser.assert_any_call('print', help=ANY, aliases=ANY)
self.assertTrue('.config.yaml' in parser.get_default('config'))
+22 -81
View File
@@ -1,12 +1,12 @@
import unittest
import pathlib
import tempfile
from typing import cast
from .test_main import CmmTestCase
from chatmastermind.message import source_code, Message, MessageError, Question, Answer, AILine, ModelLine, MessageFilter, message_in
from chatmastermind.tags import Tag, TagLine
class SourceCodeTestCase(unittest.TestCase):
class SourceCodeTestCase(CmmTestCase):
def test_source_code_with_include_delims(self) -> None:
text = """
Some text before the code block
@@ -60,46 +60,29 @@ class SourceCodeTestCase(unittest.TestCase):
self.assertEqual(result, expected_result)
class QuestionTestCase(unittest.TestCase):
def test_question_with_header(self) -> None:
class QuestionTestCase(CmmTestCase):
def test_question_with_prefix(self) -> None:
with self.assertRaises(MessageError):
Question(f"{Question.txt_header}\nWhat is your name?")
Question("=== QUESTION === What is your name?")
def test_question_with_answer_header(self) -> None:
with self.assertRaises(MessageError):
Question(f"{Answer.txt_header}\nBob")
def test_question_with_legal_header(self) -> None:
"""
If the header is just a part of a line, it's fine.
"""
question = Question(f"This is a line contaning '{Question.txt_header}'\nWhat does that mean?")
self.assertIsInstance(question, Question)
self.assertEqual(question, f"This is a line contaning '{Question.txt_header}'\nWhat does that mean?")
def test_question_without_header(self) -> None:
def test_question_without_prefix(self) -> None:
question = Question("What is your favorite color?")
self.assertIsInstance(question, Question)
self.assertEqual(question, "What is your favorite color?")
class AnswerTestCase(unittest.TestCase):
def test_answer_with_header(self) -> None:
class AnswerTestCase(CmmTestCase):
def test_answer_with_prefix(self) -> None:
with self.assertRaises(MessageError):
Answer(f"{Answer.txt_header}\nno")
Answer("=== ANSWER === Yes")
def test_answer_with_legal_header(self) -> None:
answer = Answer(f"This is a line contaning '{Answer.txt_header}'\nIt is what it is.")
self.assertIsInstance(answer, Answer)
self.assertEqual(answer, f"This is a line contaning '{Answer.txt_header}'\nIt is what it is.")
def test_answer_without_header(self) -> None:
def test_answer_without_prefix(self) -> None:
answer = Answer("No")
self.assertIsInstance(answer, Answer)
self.assertEqual(answer, "No")
class MessageToFileTxtTestCase(unittest.TestCase):
class MessageToFileTxtTestCase(CmmTestCase):
def setUp(self) -> None:
self.file = tempfile.NamedTemporaryFile(delete=False, suffix='.txt')
self.file_path = pathlib.Path(self.file.name)
@@ -160,7 +143,7 @@ This is a question.
self.message_complete.file_path = self.file_path
class MessageToFileYamlTestCase(unittest.TestCase):
class MessageToFileYamlTestCase(CmmTestCase):
def setUp(self) -> None:
self.file = tempfile.NamedTemporaryFile(delete=False, suffix='.yaml')
self.file_path = pathlib.Path(self.file.name)
@@ -226,7 +209,7 @@ class MessageToFileYamlTestCase(unittest.TestCase):
self.assertEqual(content, expected_content)
class MessageFromFileTxtTestCase(unittest.TestCase):
class MessageFromFileTxtTestCase(CmmTestCase):
def setUp(self) -> None:
self.file = tempfile.NamedTemporaryFile(delete=False, suffix='.txt')
self.file_path = pathlib.Path(self.file.name)
@@ -300,12 +283,6 @@ This is a question.
MessageFilter(tags_or={Tag('tag1')}))
self.assertIsNone(message)
def test_from_file_txt_empty_tags_dont_match(self) -> None:
message = Message.from_file(self.file_path_min,
MessageFilter(tags_or=set(),
tags_and=set()))
self.assertIsNone(message)
def test_from_file_txt_no_tags_match_tags_not(self) -> None:
message = Message.from_file(self.file_path_min,
MessageFilter(tags_not={Tag('tag1')}))
@@ -394,7 +371,7 @@ This is a question.
self.assertIsNone(message)
class MessageFromFileYamlTestCase(unittest.TestCase):
class MessageFromFileYamlTestCase(CmmTestCase):
def setUp(self) -> None:
self.file = tempfile.NamedTemporaryFile(delete=False, suffix='.yaml')
self.file_path = pathlib.Path(self.file.name)
@@ -561,7 +538,7 @@ class MessageFromFileYamlTestCase(unittest.TestCase):
self.assertIsNone(message)
class TagsFromFileTestCase(unittest.TestCase):
class TagsFromFileTestCase(CmmTestCase):
def setUp(self) -> None:
self.file_txt = tempfile.NamedTemporaryFile(delete=False, suffix='.txt')
self.file_path_txt = pathlib.Path(self.file_txt.name)
@@ -669,7 +646,7 @@ This is an answer.
self.assertSetEqual(tags, set())
class TagsFromDirTestCase(unittest.TestCase):
class TagsFromDirTestCase(CmmTestCase):
def setUp(self) -> None:
self.temp_dir = tempfile.TemporaryDirectory()
self.temp_dir_no_tags = tempfile.TemporaryDirectory()
@@ -717,7 +694,7 @@ class TagsFromDirTestCase(unittest.TestCase):
self.assertSetEqual(all_tags, set())
class MessageIDTestCase(unittest.TestCase):
class MessageIDTestCase(CmmTestCase):
def setUp(self) -> None:
self.file = tempfile.NamedTemporaryFile(delete=False, suffix='.txt')
self.file_path = pathlib.Path(self.file.name)
@@ -730,14 +707,14 @@ class MessageIDTestCase(unittest.TestCase):
self.file_path.unlink()
def test_msg_id_txt(self) -> None:
self.assertEqual(self.message.msg_id(), self.file_path.stem)
self.assertEqual(self.message.msg_id(), self.file_path.name)
def test_msg_id_txt_exception(self) -> None:
with self.assertRaises(MessageError):
self.message_no_file_path.msg_id()
class MessageHashTestCase(unittest.TestCase):
class MessageHashTestCase(CmmTestCase):
def setUp(self) -> None:
self.message1 = Message(Question('This is a question.'),
tags={Tag('tag1')},
@@ -761,7 +738,7 @@ class MessageHashTestCase(unittest.TestCase):
self.assertIn(msg, msgs)
class MessageTagsStrTestCase(unittest.TestCase):
class MessageTagsStrTestCase(CmmTestCase):
def setUp(self) -> None:
self.message = Message(Question('This is a question.'),
tags={Tag('tag1')},
@@ -771,7 +748,7 @@ class MessageTagsStrTestCase(unittest.TestCase):
self.assertEqual(self.message.tags_str(), f'{TagLine.prefix} tag1')
class MessageFilterTagsTestCase(unittest.TestCase):
class MessageFilterTagsTestCase(CmmTestCase):
def setUp(self) -> None:
self.message = Message(Question('This is a question.'),
tags={Tag('atag1'), Tag('btag2')},
@@ -786,7 +763,7 @@ class MessageFilterTagsTestCase(unittest.TestCase):
self.assertSetEqual(tags_cont, {Tag('btag2')})
class MessageInTestCase(unittest.TestCase):
class MessageInTestCase(CmmTestCase):
def setUp(self) -> None:
self.message1 = Message(Question('This is a question.'),
tags={Tag('atag1'), Tag('btag2')},
@@ -798,39 +775,3 @@ class MessageInTestCase(unittest.TestCase):
def test_message_in(self) -> None:
self.assertTrue(message_in(self.message1, [self.message1]))
self.assertFalse(message_in(self.message1, [self.message2]))
class MessageRenameTagsTestCase(unittest.TestCase):
def setUp(self) -> None:
self.message = Message(Question('This is a question.'),
tags={Tag('atag1'), Tag('btag2')},
file_path=pathlib.Path('/tmp/foo/bla'))
def test_rename_tags(self) -> None:
self.message.rename_tags({(Tag('atag1'), Tag('atag2')), (Tag('btag2'), Tag('btag3'))})
self.assertIsNotNone(self.message.tags)
self.assertSetEqual(self.message.tags, {Tag('atag2'), Tag('btag3')}) # type: ignore [arg-type]
class MessageToStrTestCase(unittest.TestCase):
def setUp(self) -> None:
self.message = Message(Question('This is a question.'),
Answer('This is an answer.'),
tags={Tag('atag1'), Tag('btag2')},
file_path=pathlib.Path('/tmp/foo/bla'))
def test_to_str(self) -> None:
expected_output = f"""{Question.txt_header}
This is a question.
{Answer.txt_header}
This is an answer."""
self.assertEqual(self.message.to_str(), expected_output)
def test_to_str_with_tags_and_file(self) -> None:
expected_output = f"""{TagLine.prefix} atag1 btag2
FILE: /tmp/foo/bla
{Question.txt_header}
This is a question.
{Answer.txt_header}
This is an answer."""
self.assertEqual(self.message.to_str(with_tags=True, with_file=True), expected_output)
-195
View File
@@ -1,195 +0,0 @@
import os
import unittest
import argparse
import tempfile
from pathlib import Path
from unittest.mock import MagicMock
from chatmastermind.commands.question import create_message
from chatmastermind.message import Message, Question, Answer
from chatmastermind.chat import ChatDB
class TestMessageCreate(unittest.TestCase):
"""
Test if messages created by the 'question' command have
the correct format.
"""
def setUp(self) -> None:
# create ChatDB structure
self.db_path = tempfile.TemporaryDirectory()
self.cache_path = tempfile.TemporaryDirectory()
self.chat = ChatDB.from_dir(cache_path=Path(self.cache_path.name),
db_path=Path(self.db_path.name))
# create some messages
self.message_text = Message(Question("What is this?"),
Answer("It is pure text"))
self.message_code = Message(Question("What is this?"),
Answer("Text\n```\nIt is embedded code\n```\ntext"))
self.chat.db_add([self.message_text, self.message_code])
# create arguments mock
self.args = MagicMock(spec=argparse.Namespace)
self.args.source_text = None
self.args.source_code = None
self.args.AI = None
self.args.model = None
self.args.output_tags = None
# File 1 : no source code block, only text
self.source_file1 = tempfile.NamedTemporaryFile(delete=False)
self.source_file1_content = """This is just text.
No source code.
Nope. Go look elsewhere!"""
with open(self.source_file1.name, 'w') as f:
f.write(self.source_file1_content)
# File 2 : one embedded source code block
self.source_file2 = tempfile.NamedTemporaryFile(delete=False)
self.source_file2_content = """This is just text.
```
This is embedded source code.
```
And some text again."""
with open(self.source_file2.name, 'w') as f:
f.write(self.source_file2_content)
# File 3 : all source code
self.source_file3 = tempfile.NamedTemporaryFile(delete=False)
self.source_file3_content = """This is all source code.
Yes, really.
Language is called 'brainfart'."""
with open(self.source_file3.name, 'w') as f:
f.write(self.source_file3_content)
# File 4 : two source code blocks
self.source_file4 = tempfile.NamedTemporaryFile(delete=False)
self.source_file4_content = """This is just text.
```
This is embedded source code.
```
And some text again.
```
This is embedded source code.
```
Aaaand again some text."""
with open(self.source_file4.name, 'w') as f:
f.write(self.source_file4_content)
def tearDown(self) -> None:
os.remove(self.source_file1.name)
os.remove(self.source_file2.name)
os.remove(self.source_file3.name)
def message_list(self, tmp_dir: tempfile.TemporaryDirectory) -> list[Path]:
# exclude '.next'
return list(Path(tmp_dir.name).glob('*.[ty]*'))
def test_message_file_created(self) -> None:
self.args.ask = ["What is this?"]
cache_dir_files = self.message_list(self.cache_path)
self.assertEqual(len(cache_dir_files), 0)
create_message(self.chat, self.args)
cache_dir_files = self.message_list(self.cache_path)
self.assertEqual(len(cache_dir_files), 1)
message = Message.from_file(cache_dir_files[0])
self.assertIsInstance(message, Message)
self.assertEqual(message.question, Question("What is this?")) # type: ignore [union-attr]
def test_single_question(self) -> None:
self.args.ask = ["What is this?"]
message = create_message(self.chat, self.args)
self.assertIsInstance(message, Message)
self.assertEqual(message.question, Question("What is this?"))
self.assertEqual(len(message.question.source_code()), 0)
def test_multipart_question(self) -> None:
self.args.ask = ["What is this", "'bard' thing?", "Is it good?"]
message = create_message(self.chat, self.args)
self.assertIsInstance(message, Message)
self.assertEqual(message.question, Question("""What is this
'bard' thing?
Is it good?"""))
def test_single_question_with_text_only_file(self) -> None:
self.args.ask = ["What is this?"]
self.args.source_text = [f"{self.source_file1.name}"]
message = create_message(self.chat, self.args)
self.assertIsInstance(message, Message)
# file contains no source code (only text)
# -> don't expect any in the question
self.assertEqual(len(message.question.source_code()), 0)
self.assertEqual(message.question, Question(f"""What is this?
{self.source_file1_content}"""))
def test_single_question_with_text_file_and_embedded_code(self) -> None:
self.args.ask = ["What is this?"]
self.args.source_code = [f"{self.source_file2.name}"]
message = create_message(self.chat, self.args)
self.assertIsInstance(message, Message)
# file contains 1 source code block
# -> expect it in the question
self.assertEqual(len(message.question.source_code()), 1)
self.assertEqual(message.question, Question("""What is this?
```
This is embedded source code.
```
"""))
def test_single_question_with_code_only_file(self) -> None:
self.args.ask = ["What is this?"]
self.args.source_code = [f"{self.source_file3.name}"]
message = create_message(self.chat, self.args)
self.assertIsInstance(message, Message)
# file is complete source code
self.assertEqual(len(message.question.source_code()), 1)
self.assertEqual(message.question, Question(f"""What is this?
```
{self.source_file3_content}
```"""))
def test_single_question_with_text_file_and_multi_embedded_code(self) -> None:
self.args.ask = ["What is this?"]
self.args.source_code = [f"{self.source_file4.name}"]
message = create_message(self.chat, self.args)
self.assertIsInstance(message, Message)
# file contains 2 source code blocks
# -> expect them in the question
self.assertEqual(len(message.question.source_code()), 2)
self.assertEqual(message.question, Question("""What is this?
```
This is embedded source code.
```
```
This is embedded source code.
```
"""))
def test_single_question_with_text_only_message(self) -> None:
self.args.ask = ["What is this?"]
self.args.source_text = [f"{self.chat.messages[0].file_path}"]
message = create_message(self.chat, self.args)
self.assertIsInstance(message, Message)
# file contains no source code (only text)
# -> don't expect any in the question
self.assertEqual(len(message.question.source_code()), 0)
self.assertEqual(message.question, Question(f"""What is this?
{self.message_text.answer}"""))
def test_single_question_with_message_and_embedded_code(self) -> None:
self.args.ask = ["What is this?"]
self.args.source_code = [f"{self.chat.messages[1].file_path}"]
message = create_message(self.chat, self.args)
self.assertIsInstance(message, Message)
# answer contains 1 source code block
# -> expect it in the question
self.assertEqual(len(message.question.source_code()), 1)
self.assertEqual(message.question, Question("""What is this?
```
It is embedded code
```
"""))
+3 -20
View File
@@ -1,8 +1,8 @@
import unittest
from .test_main import CmmTestCase
from chatmastermind.tags import Tag, TagLine, TagError
class TestTag(unittest.TestCase):
class TestTag(CmmTestCase):
def test_valid_tag(self) -> None:
tag = Tag('mytag')
self.assertEqual(tag, 'mytag')
@@ -18,7 +18,7 @@ class TestTag(unittest.TestCase):
self.assertEqual(Tag.alternative_separators, [','])
class TestTagLine(unittest.TestCase):
class TestTagLine(CmmTestCase):
def test_valid_tagline(self) -> None:
tagline = TagLine('TAGS: tag1 tag2')
self.assertEqual(tagline, 'TAGS: tag1 tag2')
@@ -144,20 +144,3 @@ class TestTagLine(unittest.TestCase):
# Test case 9: 'tags_or' and 'tags_and' are None, match all tags except excluded tags
tags_not = {Tag('tag2')}
self.assertFalse(tagline.match_tags(None, None, tags_not))
# Test case 10: 'tags_or' and 'tags_and' are empty, match no tags
self.assertFalse(tagline.match_tags(set(), set(), None))
# Test case 11: 'tags_or' is empty, match no tags
self.assertFalse(tagline.match_tags(set(), None, None))
# Test case 12: 'tags_and' is empty, match no tags
self.assertFalse(tagline.match_tags(None, set(), None))
# Test case 13: 'tags_or' is empty, match 'tags_and'
tags_and = {Tag('tag1'), Tag('tag2')}
self.assertTrue(tagline.match_tags(None, tags_and, None))
# Test case 14: 'tags_and' is empty, match 'tags_or'
tags_or = {Tag('tag1'), Tag('tag2')}
self.assertTrue(tagline.match_tags(tags_or, None, None))