New llama-run

- Added readline.cpp include
- Created run_chat_mode():
  - Initializes readline with command history
  - Maintains conversation history
  - Applies chat templates to format messages
  - Submits completion tasks to the server queue
  - Displays assistant responses interactively

Signed-off-by: Eric Curtin <eric.curtin@docker.com>

Author: ericcurtin

README.md

@@ -610,7 +610,7 @@ $ echo "source ~/.llama-completion.bash" >> ~/.bashrc
 - [stb-image](https://github.com/nothings/stb) - Single-header image format decoder, used by multimodal subsystem - Public domain
 - [nlohmann/json](https://github.com/nlohmann/json) - Single-header JSON library, used by various tools/examples - MIT License
 - [minja](https://github.com/google/minja) - Minimal Jinja parser in C++, used by various tools/examples - MIT License
-- [linenoise.cpp](./tools/run/linenoise.cpp/linenoise.cpp) - C++ library that provides readline-like line editing capabilities, used by `llama-run` - BSD 2-Clause License
+- [readline.cpp](https://github.com/ericcurtin/readline.cpp) - C++ library that provides readline-like line editing capabilities, used by `llama-run` - MIT License
 - [curl](https://curl.se/) - Client-side URL transfer library, used by various tools/examples - [CURL License](https://curl.se/docs/copyright.html)
 - [miniaudio.h](https://github.com/mackron/miniaudio) - Single-header audio format decoder, used by multimodal subsystem - Public domain
 - [subprocess.h](https://github.com/sheredom/subprocess.h) - Single-header process launching solution for C and C++ - Public domain

scripts/sync_vendor.py

@@ -19,6 +19,10 @@
     "https://raw.githubusercontent.com/yhirose/cpp-httplib/refs/tags/v0.28.0/httplib.h": "vendor/cpp-httplib/httplib.h",
 
     "https://raw.githubusercontent.com/sheredom/subprocess.h/b49c56e9fe214488493021017bf3954b91c7c1f5/subprocess.h": "vendor/sheredom/subprocess.h",
+
+    # readline.cpp: multi-file library for interactive line editing
+    # sync manually - no upstream repository yet
+    # located in vendor/readline.cpp/
 }
 
 for url, filename in vendor.items():

tools/run/CMakeLists.txt

@@ -1,5 +1,32 @@
 set(TARGET llama-run)
-add_executable(${TARGET} run.cpp linenoise.cpp/linenoise.cpp)
+
+if (MINGW)
+    # fix: https://github.com/ggml-org/llama.cpp/actions/runs/9651004652/job/26617901362?pr=8006
+    add_compile_definitions(_WIN32_WINNT=${GGML_WIN_VER})
+endif()
+
+# Include server source files (except server.cpp which has its own main())
+set(SERVER_DIR ${CMAKE_CURRENT_SOURCE_DIR}/../server)
+set(READLINE_DIR ${PROJECT_SOURCE_DIR}/vendor/readline.cpp)
+set(TARGET_SRCS
+    run.cpp
+    ${SERVER_DIR}/server-context.cpp
+    ${SERVER_DIR}/server-context.h
+    ${SERVER_DIR}/server-task.cpp
+    ${SERVER_DIR}/server-task.h
+    ${SERVER_DIR}/server-queue.cpp
+    ${SERVER_DIR}/server-queue.h
+    ${SERVER_DIR}/server-common.cpp
+    ${SERVER_DIR}/server-common.h
+    ${CMAKE_CURRENT_SOURCE_DIR}/run-chat.cpp
+    ${CMAKE_CURRENT_SOURCE_DIR}/run-chat.h
+    ${READLINE_DIR}/src/readline.cpp
+    ${READLINE_DIR}/src/buffer.cpp
+    ${READLINE_DIR}/src/history.cpp
+    ${READLINE_DIR}/src/terminal.cpp
+)
+
+add_executable(${TARGET} ${TARGET_SRCS})
 
 # TODO: avoid copying this code block from common/CMakeLists.txt
 set(LLAMA_RUN_EXTRA_LIBS "")
@@ -19,5 +46,17 @@ if (CMAKE_SYSTEM_NAME MATCHES "AIX")
     target_link_libraries(${TARGET} PRIVATE -lbsd)
 endif()
 
-target_link_libraries(${TARGET} PRIVATE common llama ${CMAKE_THREAD_LIBS_INIT} ${LLAMA_RUN_EXTRA_LIBS})
+# Include directories for server headers and readline
+target_include_directories(${TARGET} PRIVATE ${SERVER_DIR})
+target_include_directories(${TARGET} PRIVATE ${SERVER_DIR}/../mtmd)
+target_include_directories(${TARGET} PRIVATE ${CMAKE_SOURCE_DIR})
+target_include_directories(${TARGET} PRIVATE ${READLINE_DIR}/include)
+target_include_directories(${TARGET} PRIVATE ${CMAKE_CURRENT_BINARY_DIR})
+
+target_link_libraries(${TARGET} PRIVATE common mtmd llama ${CMAKE_THREAD_LIBS_INIT} ${LLAMA_RUN_EXTRA_LIBS})
+
+if (WIN32)
+    target_link_libraries(${TARGET} PRIVATE ws2_32)
+endif()
+
 target_compile_features(${TARGET} PRIVATE cxx_std_17)