README: updates and fixes

README.md

@@ -46,79 +46,80 @@ cmm [global options] command [command options]
 The `question` command is used to ask, create, and process questions.
 
 ```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]...
+cmm question [-t OTAGS]... [-k ATAGS]... [-x XTAGS]... [-o OUTTAGS]... [-A AI_ID] [-M MODEL] [-n NUM] [-m MAX] [-T TEMP] (-a QUESTION | -c QUESTION | -r [MESSAGE ...] | -p [MESSAGE ...]) [-O] [-s FILE]... [-S FILE]...
 ```
 
-* `-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
+* `-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_ID`: 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 QUESTION`: Ask a question
+* `-c, --create QUESTION`: Create a question
+* `-r, --repeat [MESSAGE ...]`: Repeat a question
+* `-p, --process [MESSAGE ...]`: Process existing questions
+* `-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
 
 #### Hist
 
 The `hist` command is used to print the chat history.
 
 ```bash
-cmm hist [-t OTAGS]... [-k ATAGS]... [-x XTAGS]... [-w] [-W] [-S] [-A ANSWER] [-Q QUESTION]
+cmm hist [-t OTAGS]... [-k ATAGS]... [-x XTAGS]... [-w] [-W] [-S] [-A SUBSTRING] [-Q SUBSTRING]
 ```
 
-* `-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
+* `-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`: Only print embedded source code
+* `-A, --answer SUBSTRING`: Search for answer substring
+* `-Q, --question SUBSTRING`: Search for question substring
 
 #### Tags
 
 The `tags` command is used to manage tags.
 
 ```bash
-cmm tags (-l | -p PREFIX | -c CONTENT)
+cmm tags (-l | -p PREFIX | -c SUBSTRING)
 ```
 
-* `-l, --list` : List all tags and their frequency
-* `-p, --prefix PREFIX` : Filter tags by prefix
-* `-c, --contain CONTENT` : Filter tags by contained substring
+* `-l, --list`: List all tags and their frequency
+* `-p, --prefix PREFIX`: Filter tags by prefix
+* `-c, --contain SUBSTRING`: Filter tags by contained substring
 
 #### Config
 
 The `config` command is used to manage the configuration.
 
 ```bash
-cmm config (-l | -m | -c CREATE)
+cmm config (-l | -m | -c FILE)
 ```
 
-* `-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
+* `-l, --list-models`: List all available models
+* `-m, --print-model`: Print the currently configured model
+* `-c, --create FILE`: 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]
+cmm print (-f FILE | -l) [-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`: Print given file
+* `-l, --latest`: Print latest message
+* `-q, --question`: Only print the question
+* `-a, --answer`: Only print the answer
+* `-S, --only-source-code`: Only print embedded source code
 
 ### Examples
 
@@ -160,18 +161,27 @@ cmm print -f example.yaml
 
 ## Configuration
 
-The configuration file (`.config.yaml`) should contain the following fields:
+The default configuration filename is `.config.yaml` (it is searched in the current working directory).
+Use the command `cmm config --create <FILENAME>` to create a default configuration:
 
-- `openai`:
-  - `api_key`: Your OpenAI API key.
-  - `model`: The name of the OpenAI model to use (e.g. "text-davinci-002").
-  - `temperature`: The temperature value for the model.
-  - `max_tokens`: The maximum number of tokens for the model.
-  - `top_p`: The top P value for the model.
-  - `frequency_penalty`: The frequency penalty value.
-  - `presence_penalty`: The presence penalty value.
-- `system`: The system message used to set the behavior of the AI.
-- `db`: The directory where the question-answer pairs are stored in YAML files.
+```
+cache: .
+db: ./db/
+ais:
+  myopenai:
+    name: openai
+    model: gpt-3.5-turbo-16k
+    api_key: 0123456789
+    temperature: 1.0
+    max_tokens: 4000
+    top_p: 1.0
+    frequency_penalty: 0.0
+    presence_penalty: 0.0
+    system: You are an assistant
+```
+
+Each AI has its own section and the name of that section is called the 'AI ID' (in the example above it is `myopenai`).
+The AI ID can be any string, as long as it's unique within the `ais` section. The AI ID is used for all commands that support the `AI` parameter and it's also stored within each message file.
 
 ## Autocompletion
 
@@ -186,33 +196,33 @@ After adding this line, restart your shell or run `source <your-shell-config-fil
 ## Contributing
 
 ### Enable commit hooks
-```
+```bash
 pip install pre-commit
 pre-commit install
 ```
 ### Execute tests before opening a PR
-```
+```bash
 pytest
 ```
 ### Consider using `pyenv` / `pyenv-virtualenv`
 Short installation instructions:
 * install `pyenv`:
-```
+```bash
 cd ~
 git clone https://github.com/pyenv/pyenv .pyenv
 cd ~/.pyenv && src/configure && make -C src
 ```
-* make sure that `~/.pyenv/shims` and `~/.pyenv/bin` are the first entries in your `PATH`, e. g. by setting it in `~/.bashrc`
+* make sure that `~/.pyenv/shims` and `~/.pyenv/bin` are the first entries in your `PATH`, e.g., by setting it in `~/.bashrc`
 * add the following to your `~/.bashrc` (after setting `PATH`): `eval "$(pyenv init -)"`
-* create a new terminal or source the changes (e. g. `source ~/.bashrc`)
+* create a new terminal or source the changes (e.g., `source ~/.bashrc`)
 * install `virtualenv`
-```
+```bash
 git clone https://github.com/pyenv/pyenv-virtualenv.git $(pyenv root)/plugins/pyenv-virtualenv
 ```
 * add the following to your `~/.bashrc` (after the commands above): `eval "$(pyenv virtualenv-init -)`
-* create a new terminal or source the changes (e. g. `source ~/.bashrc`)
-* go back to the `ChatMasterMind` repo and create a virtual environment with the latest `Python`, e. g. `3.11.4`:
-```
+* create a new terminal or source the changes (e.g., `source ~/.bashrc`)
+* go back to the `ChatMasterMind` repo and create a virtual environment with the latest `Python`, e.g., `3.11.4`:
+```bash
 cd <CMM_REPO_PATH>
 pyenv install 3.11.4
 pyenv virtualenv 3.11.4 py311
@@ -223,5 +233,3 @@ pyenv activate py311
 ## License
 
 This project is licensed under the terms of the WTFPL License.
-
-

chatmastermind/chat.py

@@ -6,7 +6,7 @@ from pathlib import Path
 from pprint import PrettyPrinter
 from pydoc import pager
 from dataclasses import dataclass
-from typing import TypeVar, Type, Optional, ClassVar, Any, Callable, Literal
+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 .tags import Tag
@@ -16,7 +16,7 @@ ChatDBInst = TypeVar('ChatDBInst', bound='ChatDB')
 
 db_next_file = '.next'
 ignored_files = [db_next_file, default_config_file]
-valid_sources = Literal['mem', 'disk', 'cache', 'db', 'all']
+msg_location = Literal['mem', 'disk', 'cache', 'db', 'all']
 
 
 class ChatError(Exception):
@@ -107,7 +107,9 @@ def clear_dir(dir_path: Path,
     """
     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.suffix in Message.file_suffixes:
+        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)
 
 
@@ -133,7 +135,7 @@ class Chat:
         error = False
         for fp in file_paths:
             if file_stems.count(fp.stem) > 1:
-                print(f"ERROR: File '{fp.stem}' appears twice in the message list: {msg_paths(fp.stem)}")
+                print(f"ERROR: Found multiple copies of message '{fp.stem}': {msg_paths(fp.stem)}")
                 error = True
         if error:
             raise ChatError("Validation failed")
@@ -373,11 +375,11 @@ class ChatDB(Chat):
             self.msg_write(messages)
 
     def msg_gather(self,
-                   source: valid_sources,
+                   loc: msg_location,
                    require_file_path: bool = False,
                    mfilter: Optional[MessageFilter] = None) -> list[Message]:
         """
-        Gather and return messages from the given source:
+        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
@@ -386,52 +388,63 @@ class ChatDB(Chat):
 
         If 'require_file_path' is True, return only files with a valid file_path.
         """
-        source_messages: list[Message] = []
-        if source in ['mem', 'all']:
+        loc_messages: list[Message] = []
+        if loc in ['mem', 'all']:
             if require_file_path:
-                source_messages += [m for m in self.messages if (m.file_path is not None and (mfilter is None or m.match(mfilter)))]
+                loc_messages += [m for m in self.messages if (m.file_path is not None and (mfilter is None or m.match(mfilter)))]
             else:
-                source_messages += [m for m in self.messages if (mfilter is None or m.match(mfilter))]
-        if source in ['cache', 'disk', 'all']:
-            source_messages += read_dir(self.cache_path, mfilter=mfilter)
-        if source in ['db', 'disk', 'all']:
-            source_messages += read_dir(self.db_path, mfilter=mfilter)
+                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 source_messages:
+        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],
-                 source: valid_sources = 'mem',
+                 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 sources:
+        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')
         """
-        source_messages = self.msg_gather(source, require_file_path=True)
-        return [m for m in source_messages
+        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]) -> 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
         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, source='all')
+            rm_messages = self.msg_find(msg_names, loc=loc)
             for m in rm_messages:
                 if (m.file_path):
                     m.file_path.unlink()
@@ -440,11 +453,11 @@ class ChatDB(Chat):
 
     def msg_latest(self,
                    mfilter: Optional[MessageFilter] = None,
-                   source: valid_sources = '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 source is 'mem').
-        Searches one of the following sources:
+        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
@@ -452,20 +465,44 @@ class ChatDB(Chat):
         * 'all'  : all messages ('mem' + 'disk')
         """
         # only consider messages with a valid file_path so they can be sorted
-        source_messages = self.msg_gather(source, require_file_path=True)
-        source_messages.sort(key=lambda m: m.msg_id(), reverse=True)
-        for m in source_messages:
+        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 cache_read(self) -> 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, self.glob, self.mfilter)
+        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
@@ -502,15 +539,31 @@ class ChatDB(Chat):
         self.messages += messages
         self.msg_sort()
 
-    def cache_clear(self) -> 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.
         """
-        clear_dir(self.cache_path, self.glob)
+        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 db_read(self) -> None:
+    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
@@ -552,3 +605,19 @@ class ChatDB(Chat):
                 m.file_path = make_file_path(self.db_path, self.default_file_suffix, self.get_next_fid)
         self.messages += messages
         self.msg_sort()
+
+    def db_move(self, message: Message) -> None:
+        """
+        Moves the given messages to the db 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])

chatmastermind/commands/hist.py

@@ -15,7 +15,7 @@ def hist_cmd(args: argparse.Namespace, config: Config) -> None:
                             tags_not=args.exclude_tags,
                             question_contains=args.question,
                             answer_contains=args.answer)
-    chat = ChatDB.from_dir(Path('.'),
+    chat = ChatDB.from_dir(Path(config.cache),
                            Path(config.db),
                            mfilter=mfilter)
     chat.print(args.source_code_only,

chatmastermind/commands/print.py

@@ -3,16 +3,13 @@ import argparse
 from pathlib import Path
 from ..configuration import Config
 from ..message import Message, MessageError
+from ..chat import ChatDB
 
 
-def print_cmd(args: argparse.Namespace, config: Config) -> None:
+def print_message(message: Message, args: argparse.Namespace) -> None:
     """
-    Handler for the 'print' command.
+    Print given message according to give arguments.
     """
-    fname = Path(args.file)
-    try:
-        message = Message.from_file(fname)
-        if message:
     if args.question:
         print(message.question)
     elif args.answer:
@@ -22,6 +19,27 @@ def print_cmd(args: argparse.Namespace, config: Config) -> None:
             print(code)
     else:
         print(message.to_str())
+
+
+def print_cmd(args: argparse.Namespace, config: Config) -> None:
+    """
+    Handler for the 'print' command.
+    """
+    # print given file
+    if args.file is not None:
+        fname = Path(args.file)
+        try:
+            message = Message.from_file(fname)
+            if message:
+                print_message(message, args)
         except MessageError:
             print(f"File is not a valid message: {args.file}")
             sys.exit(1)
+    # print latest message
+    elif args.latest:
+        chat = ChatDB.from_dir(Path(config.cache), Path(config.db))
+        latest = chat.msg_latest(loc='disk')
+        if not latest:
+            print("No message found!")
+            sys.exit(1)
+        print_message(latest, args)

chatmastermind/commands/question.py

@@ -52,7 +52,7 @@ def add_file_as_code(question_parts: list[str], file: str) -> None:
 
 def create_message(chat: ChatDB, args: argparse.Namespace) -> Message:
     """
-    Creates a new message from the given arguments and writes it
+    Create a new message from the given arguments and write it
     to the cache directory.
     """
     question_parts = []
@@ -71,28 +71,31 @@ def create_message(chat: ChatDB, args: argparse.Namespace) -> Message:
     full_question = '\n\n'.join(question_parts)
 
     message = Message(question=Question(full_question),
-                      tags=args.output_tags,  # FIXME
+                      tags=args.output_tags,
                       ai=args.AI,
                       model=args.model)
-    # only write the message (as a backup), don't add it
-    # to the current chat history
+    # only write the new message to the cache,
+    # don't add it to the internal list
     chat.cache_write([message])
     return message
 
 
 def make_request(ai: AI, chat: ChatDB, message: Message, args: argparse.Namespace) -> None:
     """
-    Make an AI request with the give AI, chat history, message and CLI arguments.
-    Print all answers.
+    Make an AI request with the given AI, chat history, message and arguments.
+    Write the response(s) to the cache directory, without appending it to the
+    given chat history. Then print the response(s).
     """
+    # print history and message question before making the request
     ai.print()
     chat.print(paged=False)
-    print(message.to_str() + '\n')
+    print(message.to_str())
     response: AIResponse = ai.request(message,
                                       chat,
                                       args.num_answers,
                                       args.output_tags)
-    # write all answers to the cache, don't add them to the chat history
+    # 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} ===")
@@ -109,7 +112,7 @@ def question_cmd(args: argparse.Namespace, config: Config) -> None:
     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('.'),
+    chat = ChatDB.from_dir(cache_path=Path(config.cache),
                            db_path=Path(config.db),
                            mfilter=mfilter)
     # if it's a new question, create and store it immediately
@@ -126,7 +129,7 @@ def question_cmd(args: argparse.Namespace, config: Config) -> None:
         make_request(ai, chat, message, args)
     # === REPEAT ===
     elif args.repeat is not None:
-        lmessage = chat.msg_latest(source='cache')
+        lmessage = chat.msg_latest(loc='cache')
         if lmessage is None:
             print("No message found to repeat!")
             sys.exit(1)

chatmastermind/commands/tags.py

@@ -8,7 +8,7 @@ def tags_cmd(args: argparse.Namespace, config: Config) -> None:
     """
     Handler for the 'tags' command.
     """
-    chat = ChatDB.from_dir(cache_path=Path('.'),
+    chat = ChatDB.from_dir(cache_path=Path(config.cache),
                            db_path=Path(config.db))
     if args.list:
         tags_freq = chat.msg_tags_frequency(args.prefix, args.contain)

chatmastermind/configuration.py

@@ -116,6 +116,7 @@ class Config:
     """
     # all members have default values, so we can easily create
     # a default configuration
+    cache: str = '.'
     db: str = './db/'
     ais: dict[str, AIConfig] = field(default_factory=create_default_ai_configs)
 
@@ -132,6 +133,7 @@ class Config:
             ai_conf = ai_config_instance(conf['name'], conf)
             ais[ID] = ai_conf
         return cls(
+            cache=str(source['cache']) if 'cache' in source else '.',
             db=str(source['db']),
             ais=ais
         )

chatmastermind/main.py

@@ -44,13 +44,13 @@ def create_parser() -> argparse.ArgumentParser:
                                        help='List of tags to exclude', metavar='XTAGS')
     etag_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 tags (default: use input tags)', metavar='OUTAGS')
     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('-A', '--AI', help='AI ID to use', metavar='AI_ID')
+    ai_parser.add_argument('-M', '--model', help='Model to use', metavar='MODEL')
     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)
@@ -61,14 +61,15 @@ def create_parser() -> argparse.ArgumentParser:
                                                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_group.add_argument('-a', '--ask', nargs='+', help='Ask a question', metavar='QUESTION')
+    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('-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')
-    question_cmd_parser.add_argument('-S', '--source-code', nargs='+', help='Add source code file content to the chat history')
+    question_cmd_parser.add_argument('-s', '--source-text', nargs='+', help='Add content of a file to the query', metavar='FILE')
+    question_cmd_parser.add_argument('-S', '--source-code', nargs='+', help='Add source code file content to the chat history',
+                                     metavar='FILE')
 
     # 'hist' command parser
     hist_cmd_parser = cmdparser.add_parser('hist', parents=[tag_parser],
@@ -79,10 +80,10 @@ def create_parser() -> argparse.ArgumentParser:
                                  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', '--source-code-only', help='Only print embedded 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')
+    hist_cmd_parser.add_argument('-A', '--answer', help='Search for answer substring', metavar='SUBSTRING')
+    hist_cmd_parser.add_argument('-Q', '--question', help='Search for question substring', metavar='SUBSTRING')
 
     # 'tags' command parser
     tags_cmd_parser = cmdparser.add_parser('tags',
@@ -92,8 +93,8 @@ def create_parser() -> argparse.ArgumentParser:
     tags_group = tags_cmd_parser.add_mutually_exclusive_group(required=True)
     tags_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")
+    tags_cmd_parser.add_argument('-p', '--prefix', help="Filter tags by prefix", metavar='PREFIX')
+    tags_cmd_parser.add_argument('-c', '--contain', help="Filter tags by contained substring", metavar='SUBSTRING')
 
     # 'config' command parser
     config_cmd_parser = cmdparser.add_parser('config',
@@ -106,18 +107,20 @@ def create_parser() -> argparse.ArgumentParser:
                               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('-c', '--create', help="Create config with default settings in the given file", metavar='FILE')
 
     # 'print' command parser
     print_cmd_parser = cmdparser.add_parser('print',
                                             help="Print message 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_group = print_cmd_parser.add_mutually_exclusive_group(required=True)
+    print_group.add_argument('-f', '--file', help='Print given message file', metavar='FILE')
+    print_group.add_argument('-l', '--latest', help='Print latest message', action='store_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_modes.add_argument('-q', '--question', help='Only print the question', action='store_true')
+    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')
 
     argcomplete.autocomplete(parser)
     return parser

chatmastermind/message.py

@@ -222,12 +222,36 @@ class Message():
     ai_yaml_key: ClassVar[str] = 'ai'
     model_yaml_key: ClassVar[str] = 'model'
 
+    def __post_init__(self) -> None:
+        # convert some types that are often set wrong
+        if self.tags is not None and not isinstance(self.tags, set):
+            self.tags = set(self.tags)
+        if self.file_path is not None and not isinstance(self.file_path, pathlib.Path):
+            self.file_path = pathlib.Path(self.file_path)
+
     def __hash__(self) -> int:
         """
         The hash value is computed based on immutable members.
         """
         return hash((self.question, self.answer))
 
+    def equals(self, other: MessageInst, tags: bool = True, ai: bool = True,
+               model: bool = True, file_path: bool = True, verbose: bool = False) -> bool:
+        """
+        Compare this message with another one, including the metadata.
+        Return True if everything is identical, False otherwise.
+        """
+        equal: bool = ((not tags or (self.tags == other.tags))
+                       and (not ai or (self.ai == other.ai))  # noqa: W503
+                       and (not model or (self.model == other.model))  # noqa: W503
+                       and (not file_path or (self.file_path == other.file_path))  # noqa: W503
+                       and (self == other))  # noqa: W503
+        if not equal and verbose:
+            print("Messages not equal:")
+            print(self)
+            print(other)
+        return equal
+
     @classmethod
     def from_dict(cls: Type[MessageInst], data: dict[str, Any]) -> MessageInst:
         """
@@ -393,9 +417,9 @@ class Message():
             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")
-            return cls.from_dict(data)
 
     def to_str(self, with_tags: bool = False, with_file: bool = False, source_code_only: bool = False) -> str:
         """
@@ -418,9 +442,6 @@ class Message():
             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
         """
         Write a Message to the given file. Type is determined based on the suffix.

tests/test_chat.py

@@ -2,6 +2,7 @@ import unittest
 import pathlib
 import tempfile
 import time
+import yaml
 from io import StringIO
 from unittest.mock import patch
 from chatmastermind.tags import TagLine
@@ -9,7 +10,18 @@ from chatmastermind.message import Message, Question, Answer, Tag, MessageFilter
 from chatmastermind.chat import Chat, ChatDB, ChatError
 
 
-class TestChat(unittest.TestCase):
+class TestChatBase(unittest.TestCase):
+    def assert_messages_equal(self, msg1: list[Message], msg2: list[Message]) -> None:
+        """
+        Compare messages using more than just Question and Answer.
+        """
+        self.assertEqual(len(msg1), len(msg2))
+        for m1, m2 in zip(msg1, msg2):
+            # exclude the file_path, compare only Q, A and metadata
+            self.assertTrue(m1.equals(m2, file_path=False, verbose=True))
+
+
+class TestChat(TestChatBase):
     def setUp(self) -> None:
         self.chat = Chat([])
         self.message1 = Message(Question('Question 1'),
@@ -25,24 +37,24 @@ class TestChat(unittest.TestCase):
     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.assert_messages_equal(self.chat.messages, [self.message1, self.message1])
         self.chat.msg_unique_id()
-        self.assertSequenceEqual(self.chat.messages, [self.message1])
+        self.assert_messages_equal(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])
+        self.assert_messages_equal(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.assert_messages_equal(self.chat.messages, [self.message1, self.message1])
         self.chat.msg_unique_content()
-        self.assertSequenceEqual(self.chat.messages, [self.message1])
+        self.assert_messages_equal(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])
+        self.assert_messages_equal(self.chat.messages, [self.message1, self.message2])
 
     def test_filter(self) -> None:
         self.chat.msg_add([self.message1, self.message2])
@@ -149,7 +161,7 @@ Answer 2
         self.assertEqual(mock_stdout.getvalue(), expected_output)
 
 
-class TestChatDB(unittest.TestCase):
+class TestChatDB(TestChatBase):
     def setUp(self) -> None:
         self.db_path = tempfile.TemporaryDirectory()
         self.cache_path = tempfile.TemporaryDirectory()
@@ -184,6 +196,10 @@ class TestChatDB(unittest.TestCase):
         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]:
@@ -204,9 +220,10 @@ class TestChatDB(unittest.TestCase):
                                     {Tag('tag4')},
                                     file_path=pathlib.Path('0004.txt'))
         duplicate_message.to_file(pathlib.Path(self.db_path.name, '0004.txt'))
-        with self.assertRaises(ChatError):
+        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:
         chat_db = ChatDB.from_dir(pathlib.Path(self.cache_path.name),
@@ -283,6 +300,25 @@ class TestChatDB(unittest.TestCase):
         with open(chat_db.next_path, 'r') as f:
             self.assertEqual(f.read(), '7')
 
+    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:
         # create a new ChatDB instance
         chat_db = ChatDB.from_dir(pathlib.Path(self.cache_path.name),
@@ -525,43 +561,92 @@ class TestChatDB(unittest.TestCase):
         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)], source='mem'), [self.message1])
-        self.assertEqual(chat_db.msg_find([self.message1.msg_id()], source='mem'), [self.message1])
-        self.assertEqual(chat_db.msg_find(['0001.txt'], source='mem'), [self.message1])
-        self.assertEqual(chat_db.msg_find(['0001'], source='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.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)], source='db'), [self.message2])
-        self.assertEqual(chat_db.msg_find([self.message2.msg_id()], source='db'), [self.message2])
-        self.assertEqual(chat_db.msg_find(['0002.yaml'], source='db'), [self.message2])
-        self.assertEqual(chat_db.msg_find(['0002'], source='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.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)], source='cache'), [])
-        self.assertEqual(chat_db.msg_find([self.message3.msg_id()], source='cache'), [])
-        self.assertEqual(chat_db.msg_find(['0003.txt'], source='cache'), [])
-        self.assertEqual(chat_db.msg_find(['0003'], source='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.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, source='all')
-        self.assertSequenceEqual(result, expected_result)
+        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(source='mem'), self.message4)
-        self.assertEqual(chat_db.msg_latest(source='db'), self.message4)
-        self.assertEqual(chat_db.msg_latest(source='disk'), self.message4)
-        self.assertEqual(chat_db.msg_latest(source='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(source='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(source='cache'), new_message)
-        self.assertEqual(chat_db.msg_latest(source='mem'), new_message)
-        self.assertEqual(chat_db.msg_latest(source='disk'), new_message)
-        self.assertEqual(chat_db.msg_latest(source='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(source='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='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='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='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='all', mfilter=MessageFilter(tags_or={Tag('tag1')})),
+                                   [self.message1])
+        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='cache', mfilter=MessageFilter(tags_or={Tag('tag3')})),
+                                   [])
+        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='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='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='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='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='db'), all_messages)

tests/test_configuration.py

@@ -57,6 +57,7 @@ class TestConfig(unittest.TestCase):
 
     def test_from_dict_should_create_config_from_dict(self) -> None:
         source_dict = {
+            'cache': '.',
             'db': './test_db/',
             'ais': {
                 'myopenai': {
@@ -73,6 +74,7 @@ class TestConfig(unittest.TestCase):
             }
         }
         config = Config.from_dict(source_dict)
+        self.assertEqual(config.cache, '.')
         self.assertEqual(config.db, './test_db/')
         self.assertEqual(len(config.ais), 1)
         self.assertEqual(config.ais['myopenai'].name, 'openai')
@@ -89,6 +91,7 @@ class TestConfig(unittest.TestCase):
 
     def test_from_file_should_load_config_from_file(self) -> None:
         source_dict = {
+            'cache': './test_cache/',
             'db': './test_db/',
             'ais': {
                 'default': {
@@ -108,6 +111,7 @@ class TestConfig(unittest.TestCase):
             yaml.dump(source_dict, f)
         config = Config.from_file(self.test_file.name)
         self.assertIsInstance(config, Config)
+        self.assertEqual(config.cache, './test_cache/')
         self.assertEqual(config.db, './test_db/')
         self.assertEqual(len(config.ais), 1)
         self.assertIsInstance(config.ais['default'], AIConfig)
@@ -115,6 +119,7 @@ class TestConfig(unittest.TestCase):
 
     def test_to_file_should_save_config_to_file(self) -> None:
         config = Config(
+            cache='./test_cache/',
             db='./test_db/',
             ais={
                 'myopenai': OpenAIConfig(
@@ -133,12 +138,14 @@ class TestConfig(unittest.TestCase):
         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['cache'], './test_cache/')
         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 = {
+            'cache': './test_cache/',
             'db': './test_db/',
             'ais': {
                 'default': {

tests/test_question_cmd.py

@@ -3,23 +3,39 @@ import unittest
 import argparse
 import tempfile
 from pathlib import Path
-from unittest.mock import MagicMock
-from chatmastermind.commands.question import create_message
+from unittest import mock
+from unittest.mock import MagicMock, call, ANY
+from typing import Optional
+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 ChatDB
+from chatmastermind.chat import Chat, ChatDB
+from chatmastermind.ai import AI, AIResponse, Tokens, AIError
 
 
-class TestMessageCreate(unittest.TestCase):
+class TestQuestionCmdBase(unittest.TestCase):
+    def assert_messages_equal(self, msg1: list[Message], msg2: list[Message]) -> None:
+        """
+        Compare messages using more than just Question and Answer.
+        """
+        self.assertEqual(len(msg1), len(msg2))
+        for m1, m2 in zip(msg1, msg2):
+            # exclude the file_path, compare only Q, A and metadata
+            self.assertTrue(m1.equals(m2, file_path=False, verbose=True))
+
+
+class TestMessageCreate(TestQuestionCmdBase):
     """
     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))
+        self.db_dir = tempfile.TemporaryDirectory()
+        self.cache_dir = tempfile.TemporaryDirectory()
+        self.chat = ChatDB.from_dir(cache_path=Path(self.cache_dir.name),
+                                    db_path=Path(self.db_dir.name))
         # create some messages
         self.message_text = Message(Question("What is this?"),
                                     Answer("It is pure text"))
@@ -74,6 +90,7 @@ Aaaand again some text."""
         os.remove(self.source_file1.name)
         os.remove(self.source_file2.name)
         os.remove(self.source_file3.name)
+        os.remove(self.source_file4.name)
 
     def message_list(self, tmp_dir: tempfile.TemporaryDirectory) -> list[Path]:
         # exclude '.next'
@@ -81,10 +98,10 @@ Aaaand again some text."""
 
     def test_message_file_created(self) -> None:
         self.args.ask = ["What is this?"]
-        cache_dir_files = self.message_list(self.cache_path)
+        cache_dir_files = self.message_list(self.cache_dir)
         self.assertEqual(len(cache_dir_files), 0)
         create_message(self.chat, self.args)
-        cache_dir_files = self.message_list(self.cache_path)
+        cache_dir_files = self.message_list(self.cache_dir)
         self.assertEqual(len(cache_dir_files), 1)
         message = Message.from_file(cache_dir_files[0])
         self.assertIsInstance(message, Message)
@@ -193,3 +210,304 @@ This is embedded source code.
 It is embedded code
 ```
 """))
+
+
+class TestQuestionCmd(TestQuestionCmdBase):
+
+    def setUp(self) -> None:
+        # create DB and cache
+        self.db_dir = tempfile.TemporaryDirectory()
+        self.cache_dir = tempfile.TemporaryDirectory()
+        # create configuration
+        self.config = Config()
+        self.config.cache = self.cache_dir.name
+        self.config.db = self.db_dir.name
+        # create a mock argparse.Namespace
+        self.args = argparse.Namespace(
+            ask=['What is the meaning of life?'],
+            num_answers=1,
+            output_tags=['science'],
+            AI='openai',
+            model='gpt-3.5-turbo',
+            or_tags=None,
+            and_tags=None,
+            exclude_tags=None,
+            source_text=None,
+            source_code=None,
+            create=None,
+            repeat=None,
+            process=None,
+            overwrite=None
+        )
+        # create a mock AI instance
+        self.ai = MagicMock(spec=AI)
+        self.ai.request.side_effect = self.mock_request
+
+    def input_message(self, args: argparse.Namespace) -> Message:
+        """
+        Create the expected input message for a question using the
+        given arguments.
+        """
+        # NOTE: we only use the first question from the "ask" list
+        # -> message creation using "question.create_message()" is
+        #    tested above
+        # the answer is always empty for the input message
+        return Message(Question(args.ask[0]),
+                       tags=args.output_tags,
+                       ai=args.AI,
+                       model=args.model)
+
+    def mock_request(self,
+                     question: Message,
+                     chat: Chat,
+                     num_answers: int = 1,
+                     otags: Optional[set[Tag]] = None) -> AIResponse:
+        """
+        Mock the 'ai.request()' function
+        """
+        question.answer = Answer("Answer 0")
+        question.tags = set(otags) if otags else None
+        question.ai = 'FakeAI'
+        question.model = 'FakeModel'
+        answers: list[Message] = [question]
+        for n in range(1, num_answers):
+            answers.append(Message(question=question.question,
+                                   answer=Answer(f"Answer {n}"),
+                                   tags=otags,
+                                   ai='FakeAI',
+                                   model='FakeModel'))
+        return AIResponse(answers, Tokens(10, 10, 20))
+
+    def message_list(self, tmp_dir: tempfile.TemporaryDirectory) -> list[Path]:
+        # exclude '.next'
+        return sorted([f for f in Path(tmp_dir.name).glob('*.[ty]*')])
+
+    @mock.patch('chatmastermind.commands.question.create_ai')
+    def test_ask_single_answer(self, mock_create_ai: MagicMock) -> None:
+        """
+        Test single answer with no errors.
+        """
+        mock_create_ai.return_value = self.ai
+        expected_question = self.input_message(self.args)
+        expected_responses = self.mock_request(expected_question,
+                                               Chat([]),
+                                               self.args.num_answers,
+                                               self.args.output_tags).messages
+
+        # execute the command
+        question_cmd(self.args, self.config)
+
+        # check for correct request call
+        self.ai.request.assert_called_once_with(expected_question,
+                                                ANY,
+                                                self.args.num_answers,
+                                                self.args.output_tags)
+        # 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='cache')
+        self.assertEqual(len(self.message_list(self.cache_dir)), 1)
+        self.assert_messages_equal(cached_msg, expected_responses)
+
+    @mock.patch('chatmastermind.commands.question.ChatDB.from_dir')
+    @mock.patch('chatmastermind.commands.question.create_ai')
+    def test_ask_single_answer_mocked(self, mock_create_ai: MagicMock, mock_from_dir: MagicMock) -> None:
+        """
+        Test single answer with no errors (mocked ChatDB version).
+        """
+        chat = MagicMock(spec=ChatDB)
+        mock_from_dir.return_value = chat
+
+        mock_create_ai.return_value = self.ai
+        expected_question = self.input_message(self.args)
+        expected_responses = self.mock_request(expected_question,
+                                               Chat([]),
+                                               self.args.num_answers,
+                                               self.args.output_tags).messages
+
+        # execute the command
+        question_cmd(self.args, self.config)
+
+        # check for correct request call
+        self.ai.request.assert_called_once_with(expected_question,
+                                                chat,
+                                                self.args.num_answers,
+                                                self.args.output_tags)
+
+        # check for the correct ChatDB calls:
+        # - initial question has been written (prior to the actual request)
+        # - responses have been written (after the request)
+        chat.cache_write.assert_has_calls([call([expected_question]),
+                                           call(expected_responses)],
+                                          any_order=False)
+
+        # check that the messages have not been added to the internal message list
+        chat.cache_add.assert_not_called()
+
+    @mock.patch('chatmastermind.commands.question.create_ai')
+    def test_ask_with_error(self, mock_create_ai: MagicMock) -> None:
+        """
+        Provoke an error during the AI request and verify that the question
+        has been correctly stored in the cache.
+        """
+        mock_create_ai.return_value = self.ai
+        expected_question = self.input_message(self.args)
+        self.ai.request.side_effect = AIError
+
+        # execute the command
+        with self.assertRaises(AIError):
+            question_cmd(self.args, self.config)
+
+        # check for correct request call
+        self.ai.request.assert_called_once_with(expected_question,
+                                                ANY,
+                                                self.args.num_answers,
+                                                self.args.output_tags)
+        # 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='cache')
+        self.assertEqual(len(self.message_list(self.cache_dir)), 1)
+        self.assert_messages_equal(cached_msg, [expected_question])
+
+    @mock.patch('chatmastermind.commands.question.create_ai')
+    def test_repeat_single_question(self, mock_create_ai: MagicMock) -> None:
+        """
+        Repeat a single question.
+        """
+        # 1. ask a question
+        mock_create_ai.return_value = self.ai
+        expected_question = self.input_message(self.args)
+        expected_responses = self.mock_request(expected_question,
+                                               Chat([]),
+                                               self.args.num_answers,
+                                               self.args.output_tags).messages
+        question_cmd(self.args, self.config)
+        chat = ChatDB.from_dir(Path(self.cache_dir.name),
+                               Path(self.db_dir.name))
+        cached_msg = chat.msg_gather(loc='cache')
+        self.assertEqual(len(self.message_list(self.cache_dir)), 1)
+        self.assert_messages_equal(cached_msg, expected_responses)
+
+        # 2. repeat the last question (without overwriting)
+        # -> expect two identical messages (except for the file_path)
+        self.args.ask = None
+        self.args.repeat = []
+        self.args.overwrite = False
+        expected_responses += expected_responses
+        question_cmd(self.args, self.config)
+        cached_msg = chat.msg_gather(loc='cache')
+        self.assertEqual(len(self.message_list(self.cache_dir)), 2)
+        self.assert_messages_equal(cached_msg, expected_responses)
+
+    @mock.patch('chatmastermind.commands.question.create_ai')
+    def test_repeat_single_question_overwrite(self, mock_create_ai: MagicMock) -> None:
+        """
+        Repeat a single question and overwrite the old one.
+        """
+        # 1. ask a question
+        mock_create_ai.return_value = self.ai
+        expected_question = self.input_message(self.args)
+        expected_responses = self.mock_request(expected_question,
+                                               Chat([]),
+                                               self.args.num_answers,
+                                               self.args.output_tags).messages
+        question_cmd(self.args, self.config)
+        chat = ChatDB.from_dir(Path(self.cache_dir.name),
+                               Path(self.db_dir.name))
+        cached_msg = chat.msg_gather(loc='cache')
+        assert cached_msg[0].file_path
+        cached_msg_file_id = cached_msg[0].file_path.stem
+        self.assertEqual(len(self.message_list(self.cache_dir)), 1)
+        self.assert_messages_equal(cached_msg, expected_responses)
+
+        # 2. repeat the last question (WITH overwriting)
+        # -> expect a single message afterwards
+        self.args.ask = None
+        self.args.repeat = []
+        self.args.overwrite = True
+        question_cmd(self.args, self.config)
+        cached_msg = chat.msg_gather(loc='cache')
+        self.assertEqual(len(self.message_list(self.cache_dir)), 1)
+        self.assert_messages_equal(cached_msg, expected_responses)
+        # also check that the file ID has not been changed
+        assert cached_msg[0].file_path
+        self.assertEqual(cached_msg_file_id, cached_msg[0].file_path.stem)
+
+    @mock.patch('chatmastermind.commands.question.create_ai')
+    def test_repeat_single_question_after_error(self, mock_create_ai: MagicMock) -> None:
+        """
+        Repeat a single question after an error.
+        """
+        # 1. ask a question and provoke an error
+        mock_create_ai.return_value = self.ai
+        expected_question = self.input_message(self.args)
+        self.ai.request.side_effect = AIError
+        with self.assertRaises(AIError):
+            question_cmd(self.args, self.config)
+        chat = ChatDB.from_dir(Path(self.cache_dir.name),
+                               Path(self.db_dir.name))
+        cached_msg = chat.msg_gather(loc='cache')
+        assert cached_msg[0].file_path
+        cached_msg_file_id = cached_msg[0].file_path.stem
+        self.assertEqual(len(self.message_list(self.cache_dir)), 1)
+        self.assert_messages_equal(cached_msg, [expected_question])
+
+        # 2. repeat the last question (without overwriting)
+        # -> expect a single message because if the original has
+        #    no answer, it should be overwritten by default
+        self.args.ask = None
+        self.args.repeat = []
+        self.args.overwrite = False
+        self.ai.request.side_effect = self.mock_request
+        expected_responses = self.mock_request(expected_question,
+                                               Chat([]),
+                                               self.args.num_answers,
+                                               self.args.output_tags).messages
+        question_cmd(self.args, self.config)
+        cached_msg = chat.msg_gather(loc='cache')
+        self.assertEqual(len(self.message_list(self.cache_dir)), 1)
+        self.assert_messages_equal(cached_msg, expected_responses)
+        # also check that the file ID has not been changed
+        assert cached_msg[0].file_path
+        self.assertEqual(cached_msg_file_id, cached_msg[0].file_path.stem)
+
+    @mock.patch('chatmastermind.commands.question.create_ai')
+    def test_repeat_single_question_new_args(self, mock_create_ai: MagicMock) -> None:
+        """
+        Repeat a single question with new arguments.
+        """
+        # 1. ask a question
+        mock_create_ai.return_value = self.ai
+        expected_question = self.input_message(self.args)
+        expected_responses = self.mock_request(expected_question,
+                                               Chat([]),
+                                               self.args.num_answers,
+                                               self.args.output_tags).messages
+        question_cmd(self.args, self.config)
+        chat = ChatDB.from_dir(Path(self.cache_dir.name),
+                               Path(self.db_dir.name))
+        cached_msg = chat.msg_gather(loc='cache')
+        self.assertEqual(len(self.message_list(self.cache_dir)), 1)
+        self.assert_messages_equal(cached_msg, expected_responses)
+
+        # 2. repeat the last question with new arguments (without overwriting)
+        # -> expect two messages with identical question and answer, but different metadata
+        self.args.ask = None
+        self.args.repeat = []
+        self.args.overwrite = False
+        self.args.output_tags = ['newtag']
+        self.args.AI = 'newai'
+        self.args.model = 'newmodel'
+        new_expected_question = Message(question=Question(expected_question.question),
+                                        tags=set(self.args.output_tags),
+                                        ai=self.args.AI,
+                                        model=self.args.model)
+        expected_responses += self.mock_request(new_expected_question,
+                                                Chat([]),
+                                                self.args.num_answers,
+                                                set(self.args.output_tags)).messages
+        question_cmd(self.args, self.config)
+        cached_msg = chat.msg_gather(loc='cache')
+        self.assertEqual(len(self.message_list(self.cache_dir)), 2)
+        self.assert_messages_equal(cached_msg, expected_responses)