From 933a90d43ce9e360bb139dda2040b4360b9b12ce Mon Sep 17 00:00:00 2001
From: Martin Bielik <martin.bielik@instea.sk>
Date: Fri, 20 Dec 2024 17:34:20 +0100
Subject: parse chat messages tests

---
 py/utils.py                       |   4 +-
 tests/chat_test.py                | 140 ++++++++++++++++++++++++++++++++++++++
 tests/resources/test1.include.txt |   1 +
 tests/resources/test2.include.txt |   1 +
 4 files changed, 145 insertions(+), 1 deletion(-)
 create mode 100644 tests/chat_test.py
 create mode 100644 tests/resources/test1.include.txt
 create mode 100644 tests/resources/test2.include.txt

diff --git a/py/utils.py b/py/utils.py
index 931eeab..f5553ee 100644
--- a/py/utils.py
+++ b/py/utils.py
@@ -160,10 +160,12 @@ def parse_chat_messages(chat_content):
 
                 try:
                     with open(path, "r") as file:
-                        message["content"] += f"\n\n==> {path} <==\n" + file.read()
+                        file_content = file.read().strip()
+                        message["content"] += f"\n\n==> {path} <==\n" + file_content
                 except UnicodeDecodeError:
                     message["content"] += "\n\n" + f"==> {path} <=="
                     message["content"] += "\n" + "Binary file, cannot display"
+            message['content'] = message['content'].strip()
 
     return messages
 
diff --git a/tests/chat_test.py b/tests/chat_test.py
new file mode 100644
index 0000000..9acfecb
--- /dev/null
+++ b/tests/chat_test.py
@@ -0,0 +1,140 @@
+from utils import parse_chat_messages
+import os
+
+curr_dir = os.path.dirname(__file__)
+
+def strip_text(txt):
+    txt = txt.strip()
+    lines = txt.splitlines()
+    return "\n".join([line.lstrip() for line in lines])
+
+def test_parse_user_message():
+    chat_content = strip_text(
+    """
+    >>> user
+
+    generate lorem ipsum
+    """)
+    messages = parse_chat_messages(chat_content)
+    assert 1 == len(messages)
+    assert 'user' == messages[0]['role']
+    assert 'generate lorem ipsum' == messages[0]['content']
+
+
+def test_parse_system_message():
+    chat_content = strip_text("""
+    >>> system
+
+    you are general assystant
+
+    >>> user
+
+    generate lorem ipsum
+    """)
+    messages = parse_chat_messages(chat_content)
+    assert 2 == len(messages)
+    assert 'system' == messages[0]['role']
+    assert 'you are general assystant' == messages[0]['content']
+    assert 'user' == messages[1]['role']
+    assert 'generate lorem ipsum' == messages[1]['content']
+
+def test_parse_assistant_message():
+    chat_content = strip_text("""
+    >>> user
+
+    generate lorem ipsum
+
+    <<< assistant
+
+    bla bla bla
+
+    >>> user
+
+    again
+    """)
+    messages = parse_chat_messages(chat_content)
+    assert 3 == len(messages)
+    assert 'user' == messages[0]['role']
+    assert 'generate lorem ipsum' == messages[0]['content']
+    assert 'assistant' == messages[1]['role']
+    assert 'bla bla bla' == messages[1]['content']
+    assert 'user' == messages[2]['role']
+    assert 'again' == messages[2]['content']
+
+def test_parse_include_single_file_message():
+    chat_content = strip_text(f"""
+    >>> user
+
+    translate to human language
+
+    >>> include
+
+    {curr_dir}/resources/test1.include.txt
+    """)
+    messages = parse_chat_messages(chat_content)
+    assert 2 == len(messages)
+    assert 'user' == messages[0]['role']
+    assert 'translate to human language' == messages[0]['content']
+    assert 'user' == messages[1]['role']
+    expected_content = strip_text(f"""
+    ==> {curr_dir}/resources/test1.include.txt <==
+    hello world
+    """)
+    assert expected_content == messages[1]['content']
+
+def test_parse_include_multiple_files_message():
+    chat_content = strip_text(f"""
+    >>> user
+
+    translate to human language
+
+    >>> include
+
+    {curr_dir}/resources/test1.include.txt
+    {curr_dir}/resources/test2.include.txt
+    """)
+    messages = parse_chat_messages(chat_content)
+    assert 2 == len(messages)
+    assert 'user' == messages[0]['role']
+    assert 'translate to human language' == messages[0]['content']
+    assert 'user' == messages[1]['role']
+    expected_content = strip_text(f"""
+    ==> {curr_dir}/resources/test1.include.txt <==
+    hello world
+
+    ==> {curr_dir}/resources/test2.include.txt <==
+    vim is awesome
+    """)
+    assert expected_content == messages[1]['content']
+
+def test_parse_include_glob_files_message():
+    chat_content = strip_text(f"""
+    >>> user
+
+    translate to human language
+
+    >>> include
+
+    {curr_dir}/**/*.include.txt
+    """)
+    messages = parse_chat_messages(chat_content)
+    assert 2 == len(messages)
+    assert 'user' == messages[0]['role']
+    assert 'translate to human language' == messages[0]['content']
+    assert 'user' == messages[1]['role']
+    expected_content = strip_text(f"""
+    ==> {curr_dir}/resources/test1.include.txt <==
+    hello world
+
+    ==> {curr_dir}/resources/test2.include.txt <==
+    vim is awesome
+    """)
+    assert expected_content == messages[1]['content']
+
+def test_parse_include_image_message():
+    # TODO
+    pass
+
+def test_parse_include_image_with_files_message():
+    # TODO
+    pass
diff --git a/tests/resources/test1.include.txt b/tests/resources/test1.include.txt
new file mode 100644
index 0000000..3b18e51
--- /dev/null
+++ b/tests/resources/test1.include.txt
@@ -0,0 +1 @@
+hello world
diff --git a/tests/resources/test2.include.txt b/tests/resources/test2.include.txt
new file mode 100644
index 0000000..e915587
--- /dev/null
+++ b/tests/resources/test2.include.txt
@@ -0,0 +1 @@
+vim is awesome
-- 
cgit v1.2.3