Add Jinja2 as preparation for pipes.

README.md

@@ -65,8 +65,6 @@ cmm question [-t OTAGS]... [-k ATAGS]... [-x XTAGS]... [-o OUTTAGS]... [-A AI_ID
 * `-O, --overwrite`: Overwrite existing messages when repeating them
 * `-s, --source-text FILE`: Add content of a file to the query
 * `-S, --source-code FILE`: Add source code file content to the chat history
-* `-l, --location {cache,db,all}`: Use given location when building the chat history (default: 'db')
-* `-g, --glob GLOB`: Filter message files using the given glob pattern
 
 #### Hist
 
@@ -85,8 +83,6 @@ cmm hist [--print | --convert FORMAT] [-t OTAGS]... [-k ATAGS]... [-x XTAGS]...
 * `-S, --source-code-only`: Only print embedded source code
 * `-A, --answer SUBSTRING`: Filter for answer substring
 * `-Q, --question SUBSTRING`: Filter for question substring
-* `-l, --location {cache,db,all}`: Use given location when building the chat history (default: 'db')
-* `-g, --glob GLOB`: Filter message files using the given glob pattern
 
 #### Tags
 

chatmastermind/ais/openai.py

@@ -2,8 +2,7 @@
 Implements the OpenAI client classes and functions.
 """
 import openai
-import tiktoken
-from typing import Optional, Union, Generator
+from typing import Optional, Union
 from ..tags import Tag
 from ..message import Message, Answer
 from ..chat import Chat
@@ -13,52 +12,6 @@ from ..configuration import OpenAIConfig
 ChatType = list[dict[str, str]]
 
 
-class OpenAIAnswer:
-    def __init__(self,
-                 idx: int,
-                 streams: dict[int, 'OpenAIAnswer'],
-                 response: openai.ChatCompletion,
-                 tokens: Tokens,
-                 encoding: tiktoken.core.Encoding) -> None:
-        self.idx = idx
-        self.streams = streams
-        self.response = response
-        self.position: int = 0
-        self.encoding = encoding
-        self.data: list[str] = []
-        self.finished: bool = False
-        self.tokens = tokens
-
-    def stream(self) -> Generator[str, None, None]:
-        while True:
-            if not self.next():
-                continue
-            if len(self.data) <= self.position:
-                break
-            yield self.data[self.position]
-            self.position += 1
-
-    def next(self) -> bool:
-        if self.finished:
-            return True
-        try:
-            chunk = next(self.response)
-        except StopIteration:
-            self.finished = True
-        if not self.finished:
-            found_choice = False
-            for choice in chunk['choices']:
-                if not choice['finish_reason']:
-                    self.streams[choice['index']].data.append(choice['delta']['content'])
-                    self.tokens.completion += len(self.encoding.encode(choice['delta']['content']))
-                    self.tokens.total = self.tokens.prompt + self.tokens.completion
-                if choice['index'] == self.idx:
-                    found_choice = True
-            if not found_choice:
-                return False
-        return True
-
-
 class OpenAI(AI):
     """
     The OpenAI AI client.
@@ -68,7 +21,7 @@ class OpenAI(AI):
         self.ID = config.ID
         self.name = config.name
         self.config = config
-        openai.api_key = self.config.api_key
+        openai.api_key = config.api_key
 
     def request(self,
                 question: Message,
@@ -80,9 +33,7 @@ class OpenAI(AI):
         chat history. The nr. of requested answers corresponds to the
         nr. of messages in the 'AIResponse'.
         """
-        self.encoding = tiktoken.encoding_for_model(self.config.model)
-        oai_chat, prompt_tokens = self.openai_chat(chat, self.config.system, question)
-        tokens: Tokens = Tokens(prompt_tokens, 0, prompt_tokens)
+        oai_chat = self.openai_chat(chat, self.config.system, question)
         response = openai.ChatCompletion.create(
             model=self.config.model,
             messages=oai_chat,
@@ -90,24 +41,22 @@ class OpenAI(AI):
             max_tokens=self.config.max_tokens,
             top_p=self.config.top_p,
             n=num_answers,
-            stream=True,
             frequency_penalty=self.config.frequency_penalty,
             presence_penalty=self.config.presence_penalty)
-        streams: dict[int, OpenAIAnswer] = {}
-        for n in range(num_answers):
-            streams[n] = OpenAIAnswer(n, streams, response, tokens, self.encoding)
-        question.answer = Answer(streams[0].stream())
+        question.answer = Answer(response['choices'][0]['message']['content'])
         question.tags = set(otags) if otags is not None else None
         question.ai = self.ID
         question.model = self.config.model
         answers: list[Message] = [question]
-        for idx in range(1, num_answers):
+        for choice in response['choices'][1:]:  # type: ignore
             answers.append(Message(question=question.question,
-                                   answer=Answer(streams[idx].stream()),
+                                   answer=Answer(choice['message']['content']),
                                    tags=otags,
                                    ai=self.ID,
                                    model=self.config.model))
-        return AIResponse(answers, tokens)
+        return AIResponse(answers, Tokens(response['usage']['prompt_tokens'],
+                                          response['usage']['completion_tokens'],
+                                          response['usage']['total_tokens']))
 
     def models(self) -> list[str]:
         """
@@ -134,26 +83,24 @@ class OpenAI(AI):
             print('\nNot ready: ' + ', '.join(not_ready))
 
     def openai_chat(self, chat: Chat, system: str,
-                    question: Optional[Message] = None) -> tuple[ChatType, int]:
+                    question: Optional[Message] = None) -> ChatType:
         """
         Create a chat history with system message in OpenAI format.
         Optionally append a new question.
         """
         oai_chat: ChatType = []
-        prompt_tokens: int = 0
 
-        def append(role: str, content: str) -> int:
+        def append(role: str, content: str) -> None:
             oai_chat.append({'role': role, 'content': content.replace("''", "'")})
-            return len(self.encoding.encode(', '.join(['role:', oai_chat[-1]['role'], 'content:', oai_chat[-1]['content']])))
 
-        prompt_tokens += append('system', system)
+        append('system', system)
         for message in chat.messages:
             if message.answer:
-                prompt_tokens += append('user', message.question)
-                prompt_tokens += append('assistant', str(message.answer))
+                append('user', message.question)
+                append('assistant', message.answer)
         if question:
-            prompt_tokens += append('user', question.question)
-        return oai_chat, prompt_tokens
+            append('user', question.question)
+        return oai_chat
 
     def tokens(self, data: Union[Message, Chat]) -> int:
         raise NotImplementedError

chatmastermind/chat.py

@@ -6,8 +6,7 @@ from pathlib import Path
 from pprint import PrettyPrinter
 from pydoc import pager
 from dataclasses import dataclass
-from enum import Enum
-from typing import TypeVar, Type, Optional, Any, Callable, Union
+from typing import TypeVar, Type, Optional, Any, Callable, Literal, Union
 from .configuration import default_config_file
 from .message import Message, MessageFilter, MessageError, MessageFormat, message_in, message_valid_formats
 from .tags import Tag
@@ -17,17 +16,10 @@ 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']
 msg_suffix = Message.file_suffix_write
 
 
-class msg_location(Enum):
-    MEM = 'mem'
-    DISK = 'disk'
-    CACHE = 'cache'
-    DB = 'db'
-    ALL = 'all'
-
-
 class ChatError(Exception):
     pass
 
@@ -52,12 +44,12 @@ def read_dir(dir_path: Path,
     Parameters:
     * 'dir_path': source directory
     * 'glob': if specified, files will be filtered using 'path.glob()',
-      otherwise it reads all files with the default message suffix
+      otherwise it uses 'path.iterdir()'.
     * 'mfilter': use with 'Message.from_file()' to filter messages
       when reading them.
     """
     messages: list[Message] = []
-    file_iter = dir_path.glob(glob) if glob else dir_path.glob(f'*{msg_suffix}')
+    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
@@ -295,7 +287,7 @@ class ChatDB(Chat):
     # a MessageFilter that all messages must match (if given)
     mfilter: Optional[MessageFilter] = None
     # the glob pattern for all messages
-    glob: str = f'*{msg_suffix}'
+    glob: Optional[str] = None
     # message format (for writing)
     mformat: MessageFormat = Message.default_format
 
@@ -311,28 +303,20 @@ class ChatDB(Chat):
     def from_dir(cls: Type[ChatDBInst],
                  cache_path: Path,
                  db_path: Path,
-                 glob: str = f'*{msg_suffix}',
-                 mfilter: Optional[MessageFilter] = None,
-                 loc: msg_location = msg_location.DB) -> ChatDBInst:
+                 glob: Optional[str] = None,
+                 mfilter: Optional[MessageFilter] = None) -> ChatDBInst:
         """
         Create a 'ChatDB' instance from the given directory structure.
         Reads all messages from 'db_path' into the local message list.
         Parameters:
         * 'cache_path': path to the directory for temporary messages
         * 'db_path': path to the directory for persistent messages
-        * 'glob': if specified, files will be filtered using 'path.glob()'
+        * 'glob': if specified, files will be filtered using 'path.glob()',
+          otherwise it uses 'path.iterdir()'.
         * 'mfilter': use with 'Message.from_file()' to filter messages
           when reading them.
-        * 'loc': read messages from given location instead of 'db_path'
         """
-        if loc == msg_location.MEM:
-            raise ChatError(f"Can't build ChatDB from message location '{loc}'")
-        messages: list[Message] = []
-        if loc in [msg_location.DB, msg_location.DISK, msg_location.ALL]:
-            messages.extend(read_dir(db_path, glob, mfilter))
-        if loc in [msg_location.CACHE, msg_location.DISK, msg_location.ALL]:
-            messages.extend(read_dir(cache_path, glob, mfilter))
-        messages.sort(key=lambda x: x.msg_id())
+        messages = read_dir(db_path, glob, mfilter)
         return cls(messages, cache_path, db_path, mfilter, glob)
 
     @classmethod
@@ -402,7 +386,7 @@ class ChatDB(Chat):
     def msg_gather(self,
                    loc: msg_location,
                    require_file_path: bool = False,
-                   glob: str = f'*{msg_suffix}',
+                   glob: Optional[str] = None,
                    mfilter: Optional[MessageFilter] = None) -> list[Message]:
         """
         Gather and return messages from the given locations:
@@ -415,14 +399,14 @@ class ChatDB(Chat):
         If 'require_file_path' is True, return only files with a valid file_path.
         """
         loc_messages: list[Message] = []
-        if loc in [msg_location.MEM, msg_location.ALL]:
+        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 [msg_location.CACHE, msg_location.DISK, msg_location.ALL]:
+        if loc in ['cache', 'disk', 'all']:
             loc_messages += read_dir(self.cache_path, glob=glob, mfilter=mfilter)
-        if loc in [msg_location.DB, msg_location.DISK, msg_location.ALL]:
+        if loc in ['db', 'disk', 'all']:
             loc_messages += read_dir(self.db_path, glob=glob, mfilter=mfilter)
         # remove_duplicates and sort the list
         unique_messages: list[Message] = []
@@ -438,7 +422,7 @@ class ChatDB(Chat):
 
     def msg_find(self,
                  msg_names: list[str],
-                 loc: msg_location = msg_location.MEM,
+                 loc: msg_location = 'mem',
                  ) -> list[Message]:
         """
         Search and return the messages with the given names. Names can either be filenames
@@ -456,7 +440,7 @@ class ChatDB(Chat):
         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 = msg_location.MEM) -> None:
+    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
@@ -468,7 +452,7 @@ class ChatDB(Chat):
         * 'db'   : messages in the DB directory
         * 'all'  : all messages ('mem' + 'disk')
         """
-        if loc != msg_location.MEM:
+        if loc != 'mem':
             # delete the message files first
             rm_messages = self.msg_find(msg_names, loc=loc)
             for m in rm_messages:
@@ -479,7 +463,7 @@ class ChatDB(Chat):
 
     def msg_latest(self,
                    mfilter: Optional[MessageFilter] = None,
-                   loc: msg_location = msg_location.MEM) -> Optional[Message]:
+                   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').
@@ -508,7 +492,7 @@ class ChatDB(Chat):
                     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=msg_location.CACHE)) > 0
+            return len(self.msg_find([message], loc='cache')) > 0
 
     def msg_in_db(self, message: Union[Message, str]) -> bool:
         """
@@ -520,9 +504,9 @@ class ChatDB(Chat):
                     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=msg_location.DB)) > 0
+            return len(self.msg_find([message], loc='db')) > 0
 
-    def cache_read(self, glob: str = f'*{msg_suffix}', mfilter: Optional[MessageFilter] = None) -> None:
+    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
@@ -565,7 +549,7 @@ class ChatDB(Chat):
         self.messages += messages
         self.msg_sort()
 
-    def cache_clear(self, glob: str = f'*{msg_suffix}') -> None:
+    def cache_clear(self, glob: Optional[str] = None) -> None:
         """
         Delete all message files from the cache dir and remove them from the internal list.
         """
@@ -585,11 +569,11 @@ class ChatDB(Chat):
         self.cache_write([message])
         # remove the old one (if any)
         if old_path:
-            self.msg_remove([str(old_path)], loc=msg_location.DB)
+            self.msg_remove([str(old_path)], loc='db')
         # (re)add it to the internal list
         self.msg_add([message])
 
-    def db_read(self, glob: str = f'*{msg_suffix}', mfilter: Optional[MessageFilter] = None) -> None:
+    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
@@ -644,6 +628,6 @@ class ChatDB(Chat):
         self.db_write([message])
         # remove the old one (if any)
         if old_path:
-            self.msg_remove([str(old_path)], loc=msg_location.CACHE)
+            self.msg_remove([str(old_path)], loc='cache')
         # (re)add it to the internal list
         self.msg_add([message])

chatmastermind/commands/common.py

@@ -1,69 +0,0 @@
-"""
-Contains shared functions for the various CMM subcommands.
-"""
-
-import argparse
-from pathlib import Path
-from ..message import Message, MessageError, source_code
-
-
-def read_text_file(file: Path) -> str:
-    with open(file) as r:
-        content = r.read().strip()
-    return content
-
-
-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:
-        content = read_text_file(Path(file))
-    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 invert_input_tag_args(args: argparse.Namespace) -> None:
-    """
-    Changes the semantics of the INPUT tags for this command:
-    * not tags specified on the CLI -> no tags are selected
-    * empty tags specified on the CLI -> all tags are selected
-    """
-    if args.or_tags is None:
-        args.or_tags = set()
-    elif len(args.or_tags) == 0:
-        args.or_tags = None
-    if args.and_tags is None:
-        args.and_tags = set()
-    elif len(args.and_tags) == 0:
-        args.and_tags = None

chatmastermind/commands/hist.py

@@ -2,7 +2,7 @@ import sys
 import argparse
 from pathlib import Path
 from ..configuration import Config
-from ..chat import ChatDB, msg_location
+from ..chat import ChatDB
 from ..message import MessageFilter, Message
 
 
@@ -15,10 +15,9 @@ def convert_messages(args: argparse.Namespace, config: Config) -> None:
     ('.txt', '.yaml') to the latest default message file suffix ('.msg').
     """
     chat = ChatDB.from_dir(Path(config.cache),
-                           Path(config.db),
-                           glob='*')
+                           Path(config.db))
     # read all known message files
-    msgs = chat.msg_gather(loc=msg_location.DISK, glob='*.*')
+    msgs = chat.msg_gather(loc='disk', glob='*.*')
     # make a set of all message IDs
     msg_ids = set([m.msg_id() for m in msgs])
     # set requested format and write all messages
@@ -30,14 +29,14 @@ def convert_messages(args: argparse.Namespace, config: Config) -> None:
             m.file_path = m.file_path.with_suffix('')
     chat.msg_write(msgs)
     # read all messages with the current default suffix
-    msgs = chat.msg_gather(loc=msg_location.DISK, glob=f'*{msg_suffix}')
+    msgs = chat.msg_gather(loc='disk', glob=f'*{msg_suffix}')
     # make sure we converted all of the original messages
     for mid in msg_ids:
         if not any(mid == m.msg_id() for m in msgs):
             print(f"Message '{mid}' has not been found after conversion. Aborting.")
             sys.exit(1)
     # delete messages with old suffixes
-    msgs = chat.msg_gather(loc=msg_location.DISK, glob='*.*')
+    msgs = chat.msg_gather(loc='disk', glob='*.*')
     for m in msgs:
         if m.file_path and m.file_path.suffix != msg_suffix:
             m.rm_file()
@@ -56,9 +55,7 @@ def print_chat(args: argparse.Namespace, config: Config) -> None:
                             answer_contains=args.answer)
     chat = ChatDB.from_dir(Path(config.cache),
                            Path(config.db),
-                           mfilter=mfilter,
-                           loc=msg_location(args.location),
-                           glob=args.glob)
+                           mfilter=mfilter)
     chat.print(args.source_code_only,
                args.with_metadata,
                paged=not args.no_paging,

chatmastermind/commands/print.py

@@ -3,7 +3,7 @@ import argparse
 from pathlib import Path
 from ..configuration import Config
 from ..message import Message, MessageError
-from ..chat import ChatDB, msg_location
+from ..chat import ChatDB
 
 
 def print_message(message: Message, args: argparse.Namespace) -> None:
@@ -38,7 +38,7 @@ def print_cmd(args: argparse.Namespace, config: Config) -> None:
     # print latest message
     elif args.latest:
         chat = ChatDB.from_dir(Path(config.cache), Path(config.db))
-        latest = chat.msg_latest(loc=msg_location.DISK)
+        latest = chat.msg_latest(loc='disk')
         if not latest:
             print("No message found!")
             sys.exit(1)

chatmastermind/commands/question.py

@@ -3,10 +3,9 @@ import argparse
 from pathlib import Path
 from itertools import zip_longest
 from copy import deepcopy
-from .common import invert_input_tag_args, add_file_as_code, add_file_as_text
 from ..configuration import Config
-from ..chat import ChatDB, msg_location
-from ..message import Message, MessageFilter, Question
+from ..chat import ChatDB
+from ..message import Message, MessageFilter, MessageError, Question, source_code
 from ..ai_factory import create_ai
 from ..ai import AI, AIResponse
 
@@ -15,6 +14,47 @@ class QuestionCmdError(Exception):
     pass
 
 
+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_msg_args(msg: Message, args: argparse.Namespace) -> argparse.Namespace:
     """
     Takes an existing message and CLI arguments, and returns modified args based
@@ -61,7 +101,7 @@ def create_message(chat: ChatDB, args: argparse.Namespace) -> Message:
         if code_file is not None and len(code_file) > 0:
             add_file_as_code(question_parts, code_file)
 
-    full_question = '\n\n'.join([str(s) for s in question_parts])
+    full_question = '\n\n'.join(question_parts)
 
     message = Message(question=Question(full_question),
                       tags=args.output_tags,
@@ -89,16 +129,13 @@ def make_request(ai: AI, chat: ChatDB, message: Message, args: argparse.Namespac
                                       args.output_tags)
     # only write the response messages to the cache,
     # don't add them to the internal list
+    chat.cache_write(response.messages)
     for idx, msg in enumerate(response.messages):
-        print(f"=== ANSWER {idx+1} ===", flush=True)
-        if msg.answer:
-            for piece in msg.answer:
-                print(piece, end='', flush=True)
-            print()
+        print(f"=== ANSWER {idx+1} ===")
+        print(msg.answer)
     if response.tokens:
         print("===============")
         print(response.tokens)
-    chat.cache_write(response.messages)
 
 
 def repeat_messages(messages: list[Message], chat: ChatDB, args: argparse.Namespace, config: Config) -> None:
@@ -123,6 +160,22 @@ def repeat_messages(messages: list[Message], chat: ChatDB, args: argparse.Namesp
             make_request(ai, chat, message, msg_args)
 
 
+def invert_input_tag_args(args: argparse.Namespace) -> None:
+    """
+    Changes the semantics of the INPUT tags for this command:
+    * not tags specified on the CLI -> no tags are selected
+    * empty tags specified on the CLI -> all tags are selected
+    """
+    if args.or_tags is None:
+        args.or_tags = set()
+    elif len(args.or_tags) == 0:
+        args.or_tags = None
+    if args.and_tags is None:
+        args.and_tags = set()
+    elif len(args.and_tags) == 0:
+        args.and_tags = None
+
+
 def question_cmd(args: argparse.Namespace, config: Config) -> None:
     """
     Handler for the 'question' command.
@@ -133,9 +186,7 @@ def question_cmd(args: argparse.Namespace, config: Config) -> None:
                             tags_not=args.exclude_tags)
     chat = ChatDB.from_dir(cache_path=Path(config.cache),
                            db_path=Path(config.db),
-                           mfilter=mfilter,
-                           glob=args.glob,
-                           loc=msg_location(args.location))
+                           mfilter=mfilter)
     # if it's a new question, create and store it immediately
     if args.ask or args.create:
         message = create_message(chat, args)
@@ -151,14 +202,14 @@ def question_cmd(args: argparse.Namespace, config: Config) -> None:
         repeat_msgs: list[Message] = []
         # repeat latest message
         if len(args.repeat) == 0:
-            lmessage = chat.msg_latest(loc=msg_location.CACHE)
+            lmessage = chat.msg_latest(loc='cache')
             if lmessage is None:
                 print("No message found to repeat!")
                 sys.exit(1)
             repeat_msgs.append(lmessage)
         # repeat given message(s)
         else:
-            repeat_msgs = chat.msg_find(args.repeat, loc=msg_location.DISK)
+            repeat_msgs = chat.msg_find(args.repeat, loc='disk')
         repeat_messages(repeat_msgs, chat, args, config)
     # === PROCESS ===
     elif args.process is not None:

chatmastermind/commands/translation.py

@@ -1,105 +0,0 @@
-import argparse
-import mimetypes
-from pathlib import Path
-from .common import invert_input_tag_args, read_text_file
-from ..configuration import Config
-from ..message import MessageFilter, Message, Question
-from ..chat import ChatDB, msg_location
-
-
-class TranslationCmdError(Exception):
-    pass
-
-
-text_separator: str = 'TEXT:'
-
-
-def assert_document_type_supported_openai(document_file: Path) -> None:
-    doctype = mimetypes.guess_type(document_file)
-    if doctype != 'text/plain':
-        raise TranslationCmdError("AI 'OpenAI' only supports document type 'text/plain''")
-
-
-def translation_prompt_openai(source_lang: str, target_lang: str) -> str:
-    """
-    Return the prompt for GPT that tells it to do the translation.
-    """
-    return f"Translate the text below the line {text_separator} from {source_lang} to {target_lang}."
-
-
-def create_message_openai(chat: ChatDB, args: argparse.Namespace) -> Message:
-    """
-    Create a new message from the given arguments and write it to the cache directory.
-
-    Message format
-    1. Translation prompt (tells GPT to do a translation)
-    2. Glossary (if specified as an argument)
-    3. User provided prompt enhancements
-    4. Translation separator
-    5. User provided text to be translated
-
-    The text to be translated is determined as a follows:
-    - if a document is provided in the arguments, translate its content
-    - if no document is provided, translate the last text argument
-
-    The other text arguments will be put into the "header" and can be used
-    to improve the translation prompt.
-    """
-    text_args: list[str] = []
-    if args.create is not None:
-        text_args = args.create
-    elif args.ask is not None:
-        text_args = args.ask
-    else:
-        raise TranslationCmdError("No input text found")
-
-    # extract user prompt and user text to be translated
-    user_text: str
-    user_prompt: str
-    if args.input_document is not None:
-        assert_document_type_supported_openai(Path(args.input_document))
-        user_text = read_text_file(Path(args.input_document))
-        user_prompt = '\n\n'.join([str(s) for s in text_args])
-    else:
-        user_text = text_args[-1]
-        user_prompt = '\n\n'.join([str(s) for s in text_args[:-1]])
-
-    # build full question string
-    # FIXME: add glossaries if given
-    question_text: str = '\n\n'.join([translation_prompt_openai(args.source_lang, args.target_lang),
-                                      user_prompt,
-                                      text_separator,
-                                      user_text])
-    # create and write the message
-    message = Message(question=Question(question_text),
-                      tags=args.output_tags,
-                      ai=args.AI,
-                      model=args.model)
-    # only write the new message to the cache,
-    # don't add it to the internal list
-    chat.cache_write([message])
-    return message
-
-
-def translation_cmd(args: argparse.Namespace, config: Config) -> None:
-    """
-    Handler for the 'translation' command. Creates and executes translation
-    requests based on the input and selected AI. Depending on the AI, the
-    whole process may be significantly different (e.g. DeepL vs OpenAI).
-    """
-    invert_input_tag_args(args)
-    mfilter = MessageFilter(tags_or=args.or_tags,
-                            tags_and=args.and_tags,
-                            tags_not=args.exclude_tags)
-    chat = ChatDB.from_dir(cache_path=Path(config.cache),
-                           db_path=Path(config.db),
-                           mfilter=mfilter,
-                           glob=args.glob,
-                           loc=msg_location(args.location))
-    # if it's a new translation, create and store it immediately
-    # FIXME: check AI type
-    if args.ask or args.create:
-        # message = create_message(chat, args)
-        create_message_openai(chat, args)
-    if args.create:
-        return

chatmastermind/glossary.py

@@ -1,138 +0,0 @@
-"""
-Module implementing glossaries for translations.
-"""
-import yaml
-import tempfile
-import shutil
-import csv
-from pathlib import Path
-from dataclasses import dataclass, field
-from typing import Type, TypeVar, ClassVar
-
-GlossaryInst = TypeVar('GlossaryInst', bound='Glossary')
-
-
-class GlossaryError(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)
-
-
-@dataclass
-class Glossary:
-    """
-    A glossary consists of the following parameters:
-        - Name (freely selectable)
-        - Path (full file path)
-        - Source language
-        - Target language
-        - Entries (pairs of source lang and target lang terms)
-        - ID (automatically generated / modified, required by DeepL)
-    """
-
-    name: str
-    source_lang: str
-    target_lang: str
-    entries: dict[str, str] = field(default_factory=lambda: dict())
-    file_path: Path | None = None
-    ID: str | None = None
-    file_suffix: ClassVar[str] = '.glo'
-
-    @classmethod
-    def from_file(cls: Type[GlossaryInst], file_path: Path) -> GlossaryInst:
-        """
-        Create a glossary from the given file.
-        """
-        if not file_path.exists():
-            raise GlossaryError(f"Glossary file '{file_path}' does not exist")
-        if file_path.suffix != cls.file_suffix:
-            raise GlossaryError(f"File type '{file_path.suffix}' is not supported")
-        with open(file_path, "r") as fd:
-            try:
-                data = yaml.load(fd, Loader=yaml.FullLoader)
-                # remove any quotes from the entries that YAML may have added while dumping
-                # (e. g. for special keywords like 'yes')
-                clean_entries = {key.strip('\"\' '): value for key, value in data['Entries'].items()}
-                return cls(name=data['Name'],
-                           source_lang=data['SourceLang'],
-                           target_lang=data['TargetLang'],
-                           entries=clean_entries,
-                           file_path=file_path,
-                           ID=data['ID'] if data['ID'] != 'None' else None)
-            except Exception:
-                raise GlossaryError(f"'{file_path}' does not contain a valid glossary")
-
-    def to_file(self, file_path: Path | None = None) -> None:
-        """
-        Write glossary to given file.
-        """
-        if file_path:
-            self.file_path = file_path
-        if not self.file_path:
-            raise GlossaryError("Got no valid path to write glossary")
-        # check / add valid suffix
-        if not self.file_path.suffix:
-            self.file_path = self.file_path.with_suffix(self.file_suffix)
-        elif self.file_path.suffix != self.file_suffix:
-            raise GlossaryError(f"File suffix '{self.file_path.suffix}' is not supported")
-        # write YAML
-        with tempfile.NamedTemporaryFile(dir=self.file_path.parent, prefix=self.file_path.name, mode="w", delete=False) as temp_fd:
-            temp_file_path = Path(temp_fd.name)
-            data = {'Name': self.name,
-                    'ID': str(self.ID),
-                    'SourceLang': self.source_lang,
-                    'TargetLang': self.target_lang,
-                    'Entries': self.entries}
-            yaml.dump(data, temp_fd, sort_keys=False)
-        shutil.move(temp_file_path, self.file_path)
-
-    def export_csv(self, dictionary: dict[str, str], file_path: Path) -> None:
-        """
-        Export the 'entries' of this glossary to a file in CSV format (compatible with DeepL).
-        """
-        with open(file_path, 'w', newline='', encoding='utf-8') as csvfile:
-            writer = csv.writer(csvfile, delimiter=',', quotechar='"', quoting=csv.QUOTE_ALL)
-            for source_entry, target_entry in self.entries.items():
-                writer.writerow([source_entry, target_entry])
-
-    def export_tsv(self, entries: dict[str, str], file_path: Path) -> None:
-        """
-        Export the 'entries' of this glossary to a file in TSV format (compatible with DeepL).
-        """
-        with open(file_path, 'w', encoding='utf-8') as file:
-            for source_entry, target_entry in self.entries.items():
-                file.write(f"{source_entry}\t{target_entry}\n")
-
-    def import_csv(self, file_path: Path) -> None:
-        """
-        Import the entries from the given CSV file to those of the current glossary.
-        Existing entries are overwritten.
-        """
-        try:
-            with open(file_path, mode='r', encoding='utf-8') as csvfile:
-                reader = csv.reader(csvfile, delimiter=',', quotechar='"')
-                self.entries = {rows[0]: rows[1] for rows in reader if len(rows) >= 2}
-        except Exception as e:
-            raise GlossaryError(f"Error importing CSV: {e}")
-
-    def import_tsv(self, file_path: Path) -> None:
-        """
-        Import the entries from the given CSV file to those of the current glossary.
-        Existing entries are overwritten.
-        """
-        try:
-            with open(file_path, mode='r', encoding='utf-8') as tsvfile:
-                self.entries = {}
-                for line in tsvfile:
-                    parts = line.strip().split('\t')
-                    if len(parts) == 2:
-                        self.entries[parts[0]] = parts[1]
-        except Exception as e:
-            raise GlossaryError(f"Error importing TSV: {e}")

chatmastermind/main.py

@@ -14,8 +14,6 @@ from .commands.tags import tags_cmd
 from .commands.config import config_cmd
 from .commands.hist import hist_cmd
 from .commands.print import print_cmd
-from .commands.translation import translation_cmd
-from .chat import msg_location
 
 
 def tags_completer(prefix: str, parsed_args: Any, **kwargs: Any) -> list[str]:
@@ -67,11 +65,6 @@ def create_parser() -> argparse.ArgumentParser:
     question_group.add_argument('-c', '--create', nargs='+', help='Create a question', metavar='QUESTION')
     question_group.add_argument('-r', '--repeat', nargs='*', help='Repeat a question', metavar='MESSAGE')
     question_group.add_argument('-p', '--process', nargs='*', help='Process existing questions', metavar='MESSAGE')
-    question_cmd_parser.add_argument('-l', '--location',
-                                     choices=[x.value for x in msg_location if x not in [msg_location.MEM, msg_location.DISK]],
-                                     default='db',
-                                     help='Use given location when building the chat history (default: \'db\')')
-    question_cmd_parser.add_argument('-g', '--glob', help='Filter message files using the given glob pattern')
     question_cmd_parser.add_argument('-O', '--overwrite', help='Overwrite existing messages when repeating them',
                                      action='store_true')
     question_cmd_parser.add_argument('-s', '--source-text', nargs='+', help='Add content of a file to the query', metavar='FILE')
@@ -94,16 +87,11 @@ def create_parser() -> argparse.ArgumentParser:
     hist_cmd_parser.add_argument('-Q', '--question', help='Print only questions with given substring', metavar='SUBSTRING')
     hist_cmd_parser.add_argument('-d', '--tight', help='Print without message separators', action='store_true')
     hist_cmd_parser.add_argument('-P', '--no-paging', help='Print without paging', action='store_true')
-    hist_cmd_parser.add_argument('-l', '--location',
-                                 choices=[x.value for x in msg_location if x not in [msg_location.MEM, msg_location.DISK]],
-                                 default='db',
-                                 help='Use given location when building the chat history (default: \'db\')')
-    hist_cmd_parser.add_argument('-g', '--glob', help='Filter message files using the given glob pattern')
 
     # 'tags' command parser
     tags_cmd_parser = cmdparser.add_parser('tags',
                                            help="Manage tags.",
-                                           aliases=['T'])
+                                           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",
@@ -137,21 +125,6 @@ def create_parser() -> argparse.ArgumentParser:
     print_cmd_modes.add_argument('-a', '--answer', help='Only print the answer', action='store_true')
     print_cmd_modes.add_argument('-S', '--only-source-code', help='Only print embedded source code', action='store_true')
 
-    # 'translation' command parser
-    translation_cmd_parser = cmdparser.add_parser('translation', parents=[ai_parser, tag_parser],
-                                                  help="ask, create and repeat translations.",
-                                                  aliases=['t'])
-    translation_cmd_parser.set_defaults(func=translation_cmd)
-    translation_group = translation_cmd_parser.add_mutually_exclusive_group(required=True)
-    translation_group.add_argument('-a', '--ask', nargs='+', help='Ask to translate the given text', metavar='TEXT')
-    translation_group.add_argument('-c', '--create', nargs='+', help='Create a translation', metavar='TEXT')
-    translation_group.add_argument('-r', '--repeat', nargs='*', help='Repeat a translation', metavar='MESSAGE')
-    translation_cmd_parser.add_argument('-S', '--source-lang', help="Source language", metavar="LANGUAGE", required=True)
-    translation_cmd_parser.add_argument('-T', '--target-lang', help="Target language", metavar="LANGUAGE", required=True)
-    translation_cmd_parser.add_argument('-G', '--glossaries', nargs='+', help="List of glossaries", metavar="GLOSSARY")
-    translation_cmd_parser.add_argument('-d', '--input-document', help="Document to translate", metavar="FILE")
-    translation_cmd_parser.add_argument('-D', '--output-document', help="Path for the translated document", metavar="FILE")
-
     argcomplete.autocomplete(parser)
     return parser
 

chatmastermind/message.py

@@ -5,9 +5,7 @@ import pathlib
 import yaml
 import tempfile
 import shutil
-import io
 from typing import Type, TypeVar, ClassVar, Optional, Any, Union, Final, Literal, Iterable, Tuple
-from typing import Generator, Iterator
 from typing import get_args as typing_get_args
 from dataclasses import dataclass, asdict, field
 from .tags import Tag, TagLine, TagError, match_tags, rename_tags
@@ -51,7 +49,7 @@ def source_code(text: str, include_delims: bool = False) -> list[str]:
     code_lines: list[str] = []
     in_code_block = False
 
-    for line in str(text).split('\n'):
+    for line in text.split('\n'):
         if line.strip().startswith('```'):
             if include_delims:
                 code_lines.append(line)
@@ -144,100 +142,30 @@ class Answer(str):
     txt_header: ClassVar[str] = '==== ANSWER ===='
     yaml_key: ClassVar[str] = 'answer'
 
-    def __init__(self, data: Union[str, Generator[str, None, None]]) -> None:
-        # Indicator of whether all of data has been processed
-        self.is_exhausted: bool = False
-
-        # Initialize data
-        self.iterator: Iterator[str] = self._init_data(data)
-
-        # Set up the buffer to hold the 'Answer' content
-        self.buffer: io.StringIO = io.StringIO()
-
-    def _init_data(self, data: Union[str, Generator[str, None, None]]) -> Iterator[str]:
+    def __new__(cls: Type[AnswerInst], string: str) -> AnswerInst:
         """
-        Process input data (either a string or a string generator)
+        Make sure the answer string does not contain the header as a whole line.
         """
-        if isinstance(data, str):
-            yield data
-        else:
-            yield from data
-
-    def __str__(self) -> str:
-        """
-        Output all content when converted into a string
-        """
-        # Ensure all data has been processed
-        for _ in self:
-            pass
-        # Return the 'Answer' content
-        return self.buffer.getvalue()
-
-    def __repr__(self) -> str:
-        return repr(str(self))
-
-    def __iter__(self) -> Generator[str, None, None]:
-        """
-        Allows the object to be iterable
-        """
-        # Generate content if not all data has been processed
-        if not self.is_exhausted:
-            yield from self.generator_iter()
-        else:
-            yield self.buffer.getvalue()
-
-    def generator_iter(self) -> Generator[str, None, None]:
-        """
-        Main generator method to process data
-        """
-        for piece in self.iterator:
-            # Write to buffer and yield piece for the iterator
-            self.buffer.write(piece)
-            yield piece
-        self.is_exhausted = True  # Set the flag that all data has been processed
-        # If the header occurs in the 'Answer' content, raise an error
-        if f'\n{self.txt_header}' in self.buffer.getvalue() or self.buffer.getvalue().startswith(self.txt_header):
-            raise MessageError(f"Answer {repr(self.buffer.getvalue())} contains the header {repr(Answer.txt_header)}")
-
-    def __eq__(self, other: object) -> bool:
-        """
-        Comparing the object to a string or another object
-        """
-        if isinstance(other, str):
-            return str(self) == other  # Compare the string value of this object to the other string
-        # Default behavior for comparing non-string objects
-        return super().__eq__(other)
-
-    def __hash__(self) -> int:
-        """
-        Generate a hash for the object based on its string representation.
-        """
-        return hash(str(self))
-
-    def __format__(self, format_spec: str) -> str:
-        """
-        Return a formatted version of the string as per the format specification.
-        """
-        return str(self).__format__(format_spec)
+        if cls.txt_header in string.split('\n'):
+            raise MessageError(f"Answer '{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:
         """
-        Build Answer from a list of strings. Make sure strings do not contain the header.
+        Build Question from a list of strings. Make sure strings do not contain the header.
         """
-        def _gen() -> Generator[str, None, None]:
-            if len(strings) > 0:
-                yield strings[0]
-                for s in strings[1:]:
-                    yield '\n'
-                    yield s
-        return cls(_gen())
+        if cls.txt_header in strings:
+            raise MessageError(f"Question contains the header '{cls.txt_header}'")
+        instance = super().__new__(cls, '\n'.join(strings).strip())
+        return instance
 
     def source_code(self, include_delims: bool = False) -> list[str]:
         """
         Extract and return all source code sections.
         """
-        return source_code(str(self), include_delims)
+        return source_code(self, include_delims)
 
 
 class Question(str):
@@ -513,7 +441,7 @@ class Message():
         output.append(self.question)
         if self.answer:
             output.append(Answer.txt_header)
-            output.append(str(self.answer))
+            output.append(self.answer)
         return '\n'.join(output)
 
     def to_file(self, file_path: Optional[pathlib.Path]=None, mformat: MessageFormat = message_default_format) -> None:  # noqa: 11
@@ -563,7 +491,7 @@ class Message():
                 temp_fd.write(f'{ModelLine.from_model(self.model)}\n')
             temp_fd.write(f'{Question.txt_header}\n{self.question}\n')
             if self.answer:
-                temp_fd.write(f'{Answer.txt_header}\n{str(self.answer)}\n')
+                temp_fd.write(f'{Answer.txt_header}\n{self.answer}\n')
         shutil.move(temp_file_path, file_path)
 
     def __to_file_yaml(self, file_path: pathlib.Path) -> None:
@@ -632,7 +560,7 @@ class Message():
            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
            or (mfilter.question_contains and mfilter.question_contains not in self.question)  # noqa: W503
-           or (mfilter.answer_contains and (not self.answer or mfilter.answer_contains not in str(self.answer)))  # noqa: W503
+           or (mfilter.answer_contains and (not self.answer or mfilter.answer_contains not in self.answer))  # noqa: W503
            or (mfilter.answer_state == 'available' and not self.answer)  # noqa: W503
            or (mfilter.ai_state == 'available' and not self.ai)  # noqa: W503
            or (mfilter.model_state == 'available' and not self.model)  # noqa: W503

requirements.txt

@@ -2,4 +2,4 @@ openai
 PyYAML
 argcomplete
 pytest
-tiktoken
+Jinja2

setup.py

@@ -2,6 +2,8 @@ from setuptools import setup, find_packages
 
 with open("README.md", "r", encoding="utf-8") as fh:
     long_description = fh.read()
+with open("requirements.txt", "r") as fh:
+    install_requirements = [line.strip() for line in fh]
 
 setup(
     name="ChatMastermind",
@@ -28,12 +30,7 @@ setup(
         "Topic :: Utilities",
         "Topic :: Text Processing",
     ],
-    install_requires=[
-        "openai",
-        "PyYAML",
-        "argcomplete",
-        "pytest",
-    ],
+    install_requires=install_requirements,
     python_requires=">=3.9",
     test_suite="tests",
     entry_points={

tests/test_ais_openai.py

@@ -16,37 +16,26 @@ class OpenAITest(unittest.TestCase):
         openai = OpenAI(config)
 
         # Set up the mock response from openai.ChatCompletion.create
-        mock_chunk1 = {
+        mock_response = {
             'choices': [
                 {
-                    'index': 0,
-                    'delta': {
+                    'message': {
                         'content': 'Answer 1'
-                    },
-                    'finish_reason': None
+                    }
                 },
                 {
-                    'index': 1,
-                    'delta': {
+                    'message': {
                         'content': 'Answer 2'
-                    },
-                    'finish_reason': None
+                    }
                 }
             ],
+            'usage': {
+                'prompt_tokens': 10,
+                'completion_tokens': 20,
+                'total_tokens': 30
+            }
         }
-        mock_chunk2 = {
-            'choices': [
-                {
-                    'index': 0,
-                    'finish_reason': 'stop'
-                },
-                {
-                    'index': 1,
-                    'finish_reason': 'stop'
-                }
-            ],
-        }
-        mock_create.return_value = iter([mock_chunk1, mock_chunk2])
+        mock_create.return_value = mock_response
 
         # Create test data
         question = Message(Question('Question'))
@@ -68,9 +57,9 @@ class OpenAITest(unittest.TestCase):
         self.assertIsNotNone(response.tokens)
         self.assertIsInstance(response.tokens, Tokens)
         assert response.tokens
-        self.assertEqual(response.tokens.prompt, 53)
-        self.assertEqual(response.tokens.completion, 6)
-        self.assertEqual(response.tokens.total, 59)
+        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(
@@ -87,7 +76,6 @@ class OpenAITest(unittest.TestCase):
             max_tokens=config.max_tokens,
             top_p=config.top_p,
             n=2,
-            stream=True,
             frequency_penalty=config.frequency_penalty,
             presence_penalty=config.presence_penalty
         )

tests/test_chat.py

@@ -7,7 +7,7 @@ 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, msg_location
+from chatmastermind.chat import Chat, ChatDB, ChatError
 
 
 msg_suffix: str = Message.file_suffix_write
@@ -240,8 +240,7 @@ class TestChatDB(TestChatBase):
         msg_to_file_force_suffix(duplicate_message)
         with self.assertRaises(ChatError) as cm:
             ChatDB.from_dir(pathlib.Path(self.cache_path.name),
-                            pathlib.Path(self.db_path.name),
-                            glob='*')
+                            pathlib.Path(self.db_path.name))
         self.assertEqual(str(cm.exception), "Validation failed")
 
     def test_file_path_ID_exists(self) -> None:
@@ -596,92 +595,92 @@ class TestChatDB(TestChatBase):
         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=msg_location.MEM), [self.message1])
-        self.assertEqual(chat_db.msg_find([self.message1.msg_id()], loc=msg_location.MEM), [self.message1])
-        self.assertEqual(chat_db.msg_find(['0001.msg'], loc=msg_location.MEM), [self.message1])
-        self.assertEqual(chat_db.msg_find(['0001'], loc=msg_location.MEM), [self.message1])
+        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.msg'], 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=msg_location.DB), [self.message2])
-        self.assertEqual(chat_db.msg_find([self.message2.msg_id()], loc=msg_location.DB), [self.message2])
-        self.assertEqual(chat_db.msg_find(['0002.msg'], loc=msg_location.DB), [self.message2])
-        self.assertEqual(chat_db.msg_find(['0002'], loc=msg_location.DB), [self.message2])
+        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.msg'], 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=msg_location.CACHE), [])
-        self.assertEqual(chat_db.msg_find([self.message3.msg_id()], loc=msg_location.CACHE), [])
-        self.assertEqual(chat_db.msg_find(['0003.msg'], loc=msg_location.CACHE), [])
-        self.assertEqual(chat_db.msg_find(['0003'], loc=msg_location.CACHE), [])
+        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.msg'], 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.msg', 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=msg_location.ALL)
+        result = chat_db.msg_find(search_names, loc='all')
         self.assert_messages_equal(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=msg_location.MEM), self.message4)
-        self.assertEqual(chat_db.msg_latest(loc=msg_location.DB), self.message4)
-        self.assertEqual(chat_db.msg_latest(loc=msg_location.DISK), self.message4)
-        self.assertEqual(chat_db.msg_latest(loc=msg_location.ALL), self.message4)
+        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=msg_location.CACHE))
+        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=msg_location.CACHE), new_message)
-        self.assertEqual(chat_db.msg_latest(loc=msg_location.MEM), new_message)
-        self.assertEqual(chat_db.msg_latest(loc=msg_location.DISK), new_message)
-        self.assertEqual(chat_db.msg_latest(loc=msg_location.ALL), 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=msg_location.DB), self.message4)
+        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.assert_messages_equal(chat_db.msg_gather(loc=msg_location.ALL), all_messages)
-        self.assert_messages_equal(chat_db.msg_gather(loc=msg_location.DB), all_messages)
-        self.assert_messages_equal(chat_db.msg_gather(loc=msg_location.MEM), all_messages)
-        self.assert_messages_equal(chat_db.msg_gather(loc=msg_location.DISK), all_messages)
-        self.assert_messages_equal(chat_db.msg_gather(loc=msg_location.CACHE), [])
+        self.assert_messages_equal(chat_db.msg_gather(loc='all'), all_messages)
+        self.assert_messages_equal(chat_db.msg_gather(loc='db'), all_messages)
+        self.assert_messages_equal(chat_db.msg_gather(loc='mem'), all_messages)
+        self.assert_messages_equal(chat_db.msg_gather(loc='disk'), all_messages)
+        self.assert_messages_equal(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.assert_messages_equal(chat_db.msg_gather(loc=msg_location.MEM), all_messages_mem)
-        self.assert_messages_equal(chat_db.msg_gather(loc=msg_location.ALL), all_messages_mem)
+        self.assert_messages_equal(chat_db.msg_gather(loc='mem'), all_messages_mem)
+        self.assert_messages_equal(chat_db.msg_gather(loc='all'), all_messages_mem)
         # the nr. of messages on disk did not change -> expect old result
-        self.assert_messages_equal(chat_db.msg_gather(loc=msg_location.DB), all_messages)
-        self.assert_messages_equal(chat_db.msg_gather(loc=msg_location.DISK), all_messages)
-        self.assert_messages_equal(chat_db.msg_gather(loc=msg_location.CACHE), [])
+        self.assert_messages_equal(chat_db.msg_gather(loc='db'), all_messages)
+        self.assert_messages_equal(chat_db.msg_gather(loc='disk'), all_messages)
+        self.assert_messages_equal(chat_db.msg_gather(loc='cache'), [])
         # test with MessageFilter
-        self.assert_messages_equal(chat_db.msg_gather(loc=msg_location.ALL, mfilter=MessageFilter(tags_or={Tag('tag1')})),
+        self.assert_messages_equal(chat_db.msg_gather(loc='all', mfilter=MessageFilter(tags_or={Tag('tag1')})),
                                    [self.message1])
-        self.assert_messages_equal(chat_db.msg_gather(loc=msg_location.DISK, mfilter=MessageFilter(tags_or={Tag('tag2')})),
+        self.assert_messages_equal(chat_db.msg_gather(loc='disk', mfilter=MessageFilter(tags_or={Tag('tag2')})),
                                    [self.message2])
-        self.assert_messages_equal(chat_db.msg_gather(loc=msg_location.CACHE, mfilter=MessageFilter(tags_or={Tag('tag3')})),
+        self.assert_messages_equal(chat_db.msg_gather(loc='cache', mfilter=MessageFilter(tags_or={Tag('tag3')})),
                                    [])
-        self.assert_messages_equal(chat_db.msg_gather(loc=msg_location.MEM, mfilter=MessageFilter(question_contains="What")),
+        self.assert_messages_equal(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.assert_messages_equal(chat_db.msg_gather(loc=msg_location.DB), all_messages)
-        self.assert_messages_equal(chat_db.msg_gather(loc=msg_location.CACHE), [])
+        self.assert_messages_equal(chat_db.msg_gather(loc='db'), all_messages)
+        self.assert_messages_equal(chat_db.msg_gather(loc='cache'), [])
         # move first message to the cache
         chat_db.cache_move(self.message1)
-        self.assert_messages_equal(chat_db.msg_gather(loc=msg_location.CACHE), [self.message1])
+        self.assert_messages_equal(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.assert_messages_equal(chat_db.msg_gather(loc=msg_location.DB), [self.message2, self.message3, self.message4])
-        self.assert_messages_equal(chat_db.msg_gather(loc=msg_location.ALL), all_messages)
-        self.assert_messages_equal(chat_db.msg_gather(loc=msg_location.DISK), all_messages)
-        self.assert_messages_equal(chat_db.msg_gather(loc=msg_location.MEM), all_messages)
+        self.assert_messages_equal(chat_db.msg_gather(loc='db'), [self.message2, self.message3, self.message4])
+        self.assert_messages_equal(chat_db.msg_gather(loc='all'), all_messages)
+        self.assert_messages_equal(chat_db.msg_gather(loc='disk'), all_messages)
+        self.assert_messages_equal(chat_db.msg_gather(loc='mem'), all_messages)
         # now move first message back to the DB
         chat_db.db_move(self.message1)
-        self.assert_messages_equal(chat_db.msg_gather(loc=msg_location.CACHE), [])
+        self.assert_messages_equal(chat_db.msg_gather(loc='cache'), [])
         self.assertEqual(self.message1.file_path.parent, pathlib.Path(self.db_path.name))  # type: ignore [union-attr]
-        self.assert_messages_equal(chat_db.msg_gather(loc=msg_location.DB), all_messages)
+        self.assert_messages_equal(chat_db.msg_gather(loc='db'), all_messages)

tests/test_glossary.py

@@ -1,117 +0,0 @@
-import unittest
-import tempfile
-from pathlib import Path
-from chatmastermind.glossary import Glossary, GlossaryError
-
-
-glossary_suffix: str = Glossary.file_suffix
-
-
-class TestGlossary(unittest.TestCase):
-
-    def test_from_file_valid_yaml(self) -> None:
-        # Prepare a temporary YAML file with valid content
-        with tempfile.NamedTemporaryFile('w', delete=False, suffix=glossary_suffix) as yaml_file:
-            yaml_file.write("Name: Sample\n"
-                            "ID: '123'\n"
-                            "SourceLang: en\n"
-                            "TargetLang: es\n"
-                            "Entries:\n"
-                            "  hello: hola\n"
-                            "  goodbye: adiós\n"
-                            "  'yes': sí\n")  # 'yes' is a YAML keyword and therefore quoted
-            yaml_file_path = Path(yaml_file.name)
-
-        glossary = Glossary.from_file(yaml_file_path)
-        self.assertEqual(glossary.name, "Sample")
-        self.assertEqual(glossary.source_lang, "en")
-        self.assertEqual(glossary.target_lang, "es")
-        self.assertEqual(glossary.entries, {"hello": "hola", "goodbye": "adiós", "yes": "sí"})
-        yaml_file_path.unlink()  # Remove the temporary file
-
-    def test_to_file_writes_yaml(self) -> None:
-        # Create glossary instance
-        glossary = Glossary(name="Test", source_lang="en", target_lang="fr", entries={"yes": "oui"})
-
-        with tempfile.NamedTemporaryFile('w', delete=False, suffix=glossary_suffix) as tmp_file:
-            file_path = Path(tmp_file.name)
-            glossary.to_file(file_path)
-
-        with open(file_path, 'r') as file:
-            content = file.read()
-
-        self.assertIn("Name: Test", content)
-        self.assertIn("SourceLang: en", content)
-        self.assertIn("TargetLang: fr", content)
-        self.assertIn("Entries", content)
-        # 'yes' is a YAML keyword and therefore quoted
-        self.assertIn("'yes': oui", content)
-        file_path.unlink()  # Remove the temporary file
-
-    def test_write_read_glossary(self) -> None:
-        # Create glossary instance
-        # -> use 'yes' in order to test if the YAML quoting is correctly removed when reading the file
-        glossary_write = Glossary(name="Test", source_lang="en", target_lang="fr", entries={"yes": "oui"})
-
-        with tempfile.NamedTemporaryFile('w', delete=False, suffix=glossary_suffix) as tmp_file:
-            file_path = Path(tmp_file.name)
-            glossary_write.to_file(file_path)
-
-        # create new instance from glossary file
-        glossary_read = Glossary.from_file(file_path)
-        self.assertEqual(glossary_write.name, glossary_read.name)
-        self.assertEqual(glossary_write.source_lang, glossary_read.source_lang)
-        self.assertEqual(glossary_write.target_lang, glossary_read.target_lang)
-        self.assertDictEqual(glossary_write.entries, glossary_read.entries)
-
-        file_path.unlink()  # Remove the temporary file
-
-    def test_import_export_csv(self) -> None:
-        glossary = Glossary(name="Test", source_lang="en", target_lang="fr", entries={})
-
-        # First export to CSV
-        with tempfile.NamedTemporaryFile('w', delete=False, suffix=glossary_suffix) as csvfile:
-            csv_file_path = Path(csvfile.name)
-        glossary.entries = {"hello": "salut", "goodbye": "au revoir"}
-        glossary.export_csv(glossary.entries, csv_file_path)
-
-        # Now import CSV
-        glossary.import_csv(csv_file_path)
-        self.assertEqual(glossary.entries, {"hello": "salut", "goodbye": "au revoir"})
-        csv_file_path.unlink()  # Remove the temporary file
-
-    def test_import_export_tsv(self) -> None:
-        glossary = Glossary(name="Test", source_lang="en", target_lang="fr", entries={})
-
-        # First export to TSV
-        with tempfile.NamedTemporaryFile('w', delete=False, suffix=glossary_suffix) as tsvfile:
-            tsv_file_path = Path(tsvfile.name)
-        glossary.entries = {"hello": "salut", "goodbye": "au revoir"}
-        glossary.export_tsv(glossary.entries, tsv_file_path)
-
-        # Now import TSV
-        glossary.import_tsv(tsv_file_path)
-        self.assertEqual(glossary.entries, {"hello": "salut", "goodbye": "au revoir"})
-        tsv_file_path.unlink()  # Remove the temporary file
-
-    def test_to_file_wrong_suffix(self) -> None:
-        """
-        Test for exception if suffix is wrong.
-        """
-        glossary = Glossary(name="Test", source_lang="en", target_lang="fr", entries={"yes": "oui"})
-        with tempfile.NamedTemporaryFile('w', delete=False, suffix='.wrong') as tmp_file:
-            file_path = Path(tmp_file.name)
-            with self.assertRaises(GlossaryError) as err:
-                glossary.to_file(file_path)
-                self.assertEqual(str(err.exception), "File suffix '.wrong' is not supported")
-
-    def test_to_file_auto_suffix(self) -> None:
-        """
-        Test if suffix is auto-generated if omitted.
-        """
-        glossary = Glossary(name="Test", source_lang="en", target_lang="fr", entries={"yes": "oui"})
-        with tempfile.NamedTemporaryFile('w', delete=False, suffix='') as tmp_file:
-            file_path = Path(tmp_file.name)
-            glossary.to_file(file_path)
-        assert glossary.file_path is not None
-        self.assertEqual(glossary.file_path.suffix, glossary_suffix)

tests/test_hist_cmd.py

@@ -4,7 +4,7 @@ import tempfile
 import yaml
 from pathlib import Path
 from chatmastermind.message import Message, Question
-from chatmastermind.chat import ChatDB, ChatError, msg_location
+from chatmastermind.chat import ChatDB, ChatError
 from chatmastermind.configuration import Config
 from chatmastermind.commands.hist import convert_messages
 
@@ -41,7 +41,7 @@ class TestConvertMessages(unittest.TestCase):
     def test_convert_messages(self) -> None:
         self.args.convert = 'yaml'
         convert_messages(self.args, self.config)
-        msgs = self.chat.msg_gather(loc=msg_location.DISK, glob='*.*')
+        msgs = self.chat.msg_gather(loc='disk', glob='*.*')
         # Check if the number of messages is the same as before
         self.assertEqual(len(msgs), len(self.messages))
         # Check if all messages have the requested suffix

tests/test_message.py

@@ -91,7 +91,7 @@ class QuestionTestCase(unittest.TestCase):
 class AnswerTestCase(unittest.TestCase):
     def test_answer_with_header(self) -> None:
         with self.assertRaises(MessageError):
-            str(Answer(f"{Answer.txt_header}\nno"))
+            Answer(f"{Answer.txt_header}\nno")
 
     def test_answer_with_legal_header(self) -> None:
         answer = Answer(f"This is a line contaning '{Answer.txt_header}'\nIt is what it is.")

tests/test_question_cmd.py

@@ -9,7 +9,7 @@ from chatmastermind.configuration import Config
 from chatmastermind.commands.question import create_message, question_cmd
 from chatmastermind.tags import Tag
 from chatmastermind.message import Message, Question, Answer
-from chatmastermind.chat import Chat, ChatDB, msg_location
+from chatmastermind.chat import Chat, ChatDB
 from chatmastermind.ai import AIError
 from .test_common import TestWithFakeAI
 
@@ -234,8 +234,6 @@ class TestQuestionCmd(TestWithFakeAI):
         # create a mock argparse.Namespace
         self.args = argparse.Namespace(
             ask=['What is the meaning of life?'],
-            glob=None,
-            location='db',
             num_answers=1,
             output_tags=['science'],
             AI='FakeAI',
@@ -281,7 +279,7 @@ class TestQuestionCmdAsk(TestQuestionCmd):
         # check for the expected message files
         chat = ChatDB.from_dir(Path(self.cache_dir.name),
                                Path(self.db_dir.name))
-        cached_msg = chat.msg_gather(loc=msg_location.CACHE)
+        cached_msg = chat.msg_gather(loc='cache')
         self.assertEqual(len(self.message_list(self.cache_dir)), 1)
         self.assert_msgs_equal_except_file_path(cached_msg, expected_responses)
 
@@ -339,7 +337,7 @@ class TestQuestionCmdAsk(TestQuestionCmd):
         # check for the expected message files
         chat = ChatDB.from_dir(Path(self.cache_dir.name),
                                Path(self.db_dir.name))
-        cached_msg = chat.msg_gather(loc=msg_location.CACHE)
+        cached_msg = chat.msg_gather(loc='cache')
         self.assertEqual(len(self.message_list(self.cache_dir)), 1)
         self.assert_msgs_equal_except_file_path(cached_msg, [expected_question])
 
@@ -377,7 +375,7 @@ class TestQuestionCmdRepeat(TestQuestionCmd):
         # we expect the original message + the one with the new response
         expected_responses = [message] + [expected_response]
         question_cmd(self.args, self.config)
-        cached_msg = chat.msg_gather(loc=msg_location.CACHE)
+        cached_msg = chat.msg_gather(loc='cache')
         print(self.message_list(self.cache_dir))
         self.assertEqual(len(self.message_list(self.cache_dir)), 2)
         self.assert_msgs_equal_except_file_path(cached_msg, expected_responses)
@@ -398,7 +396,7 @@ class TestQuestionCmdRepeat(TestQuestionCmd):
                           model=self.args.model,
                           file_path=Path(self.cache_dir.name) / f'0001{msg_suffix}')
         chat.msg_write([message])
-        cached_msg = chat.msg_gather(loc=msg_location.CACHE)
+        cached_msg = chat.msg_gather(loc='cache')
         assert cached_msg[0].file_path
         cached_msg_file_id = cached_msg[0].file_path.stem
 
@@ -414,7 +412,7 @@ class TestQuestionCmdRepeat(TestQuestionCmd):
                                     tags=message.tags,
                                     file_path=Path('<NOT COMPARED>'))
         question_cmd(self.args, self.config)
-        cached_msg = chat.msg_gather(loc=msg_location.CACHE)
+        cached_msg = chat.msg_gather(loc='cache')
         self.assertEqual(len(self.message_list(self.cache_dir)), 1)
         self.assert_msgs_equal_except_file_path(cached_msg, [expected_response])
         # also check that the file ID has not been changed
@@ -437,7 +435,7 @@ class TestQuestionCmdRepeat(TestQuestionCmd):
                           model=self.args.model,
                           file_path=Path(self.cache_dir.name) / f'0001{msg_suffix}')
         chat.msg_write([message])
-        cached_msg = chat.msg_gather(loc=msg_location.CACHE)
+        cached_msg = chat.msg_gather(loc='cache')
         assert cached_msg[0].file_path
         cached_msg_file_id = cached_msg[0].file_path.stem
 
@@ -454,7 +452,7 @@ class TestQuestionCmdRepeat(TestQuestionCmd):
                                     tags=message.tags,
                                     file_path=Path('<NOT COMPARED>'))
         question_cmd(self.args, self.config)
-        cached_msg = chat.msg_gather(loc=msg_location.CACHE)
+        cached_msg = chat.msg_gather(loc='cache')
         self.assertEqual(len(self.message_list(self.cache_dir)), 1)
         self.assert_msgs_equal_except_file_path(cached_msg, [expected_response])
         # also check that the file ID has not been changed
@@ -477,7 +475,7 @@ class TestQuestionCmdRepeat(TestQuestionCmd):
                           model=self.args.model,
                           file_path=Path(self.cache_dir.name) / f'0001{msg_suffix}')
         chat.msg_write([message])
-        cached_msg = chat.msg_gather(loc=msg_location.CACHE)
+        cached_msg = chat.msg_gather(loc='cache')
         assert cached_msg[0].file_path
 
         # repeat the last question with new arguments (without overwriting)
@@ -495,7 +493,7 @@ class TestQuestionCmdRepeat(TestQuestionCmd):
                                         tags={Tag('newtag')},
                                         file_path=Path('<NOT COMPARED>'))
         question_cmd(self.args, self.config)
-        cached_msg = chat.msg_gather(loc=msg_location.CACHE)
+        cached_msg = chat.msg_gather(loc='cache')
         self.assertEqual(len(self.message_list(self.cache_dir)), 2)
         self.assert_msgs_equal_except_file_path(cached_msg, [message] + [new_expected_response])
 
@@ -515,7 +513,7 @@ class TestQuestionCmdRepeat(TestQuestionCmd):
                           model=self.args.model,
                           file_path=Path(self.cache_dir.name) / f'0001{msg_suffix}')
         chat.msg_write([message])
-        cached_msg = chat.msg_gather(loc=msg_location.CACHE)
+        cached_msg = chat.msg_gather(loc='cache')
         assert cached_msg[0].file_path
 
         # repeat the last question with new arguments
@@ -532,7 +530,7 @@ class TestQuestionCmdRepeat(TestQuestionCmd):
                                         tags={Tag('newtag')},
                                         file_path=Path('<NOT COMPARED>'))
         question_cmd(self.args, self.config)
-        cached_msg = chat.msg_gather(loc=msg_location.CACHE)
+        cached_msg = chat.msg_gather(loc='cache')
         self.assertEqual(len(self.message_list(self.cache_dir)), 1)
         self.assert_msgs_equal_except_file_path(cached_msg, [new_expected_response])
 
@@ -588,8 +586,8 @@ class TestQuestionCmdRepeat(TestQuestionCmd):
         self.assertEqual(len(self.message_list(self.cache_dir)), 4)
         self.assertEqual(len(self.message_list(self.db_dir)), 1)
         expected_cache_messages = [expected_responses[0], message2, expected_responses[1], expected_responses[2]]
-        cached_msg = chat.msg_gather(loc=msg_location.CACHE)
+        cached_msg = chat.msg_gather(loc='cache')
         self.assert_msgs_equal_except_file_path(cached_msg, expected_cache_messages)
         # check that the DB message has not been modified at all
-        db_msg = chat.msg_gather(loc=msg_location.DB)
+        db_msg = chat.msg_gather(loc='db')
         self.assert_msgs_all_equal(db_msg, [message3])