feat: add platformio #104
feat: add platformio #104
Conversation
|
|
||
| CWD = os.getcwd() | ||
|
|
||
| fqbn_to_board = { # Mapping from fqbn to PlatformIO board |
| from deps.logs import logger | ||
| from models import Sketch, PythonProgram, Messages | ||
|
|
||
| from deps.sketch import _install_libraries, _compile_sketch, startup |
There was a problem hiding this comment.
these functions should not be private (_), since they are used outside the module.
| if not await check_for_internet(): | ||
| return | ||
| logger.info("Updating library index...") | ||
| global library_index_json, library_indexed_json # pylint: disable=global-statement |
There was a problem hiding this comment.
if you put all these functions in a class, use don't need a global. The global will become a private property of the class.
| return | ||
| logger.info("Updating library index...") | ||
| global library_index_json, library_indexed_json # pylint: disable=global-statement | ||
| async with httpx.AsyncClient() as client: |
There was a problem hiding this comment.
if this becomes a class, you can store the httpx client an re-use it instead of re-creating it for every request
| return dict(line.split("=") for line in properties.split("\n") if "=" in line) | ||
|
|
||
|
|
||
| async def _install_library_zip( |
There was a problem hiding this comment.
why not use pio pkg install instead of this manual stuff? And then link the lib with something like this:
[env:myenv]
lib_deps =
FooLib=symlink://../../shared-libraries/FooLib
There was a problem hiding this comment.
Using pio pkg install still requires some weird stuff since we'd have to create a enviroment to install the packages in and handle it must be compatible with multiple platforms etc.
| deps = [] | ||
| deps_arches = {} | ||
| in_deps = {} | ||
| for file in library_zip.namelist(): |
There was a problem hiding this comment.
this is all blocking IO and will hang the webserver for as long as it takes to unzip the files. Is there an async version of these calls?
There was a problem hiding this comment.
This isn't blocking IO, the zip is never saved it's in memory
| for i, dep in enumerate(deps): | ||
| deps[i] = dep.strip() | ||
| in_deps = await _install_libraries(deps) | ||
| for dep in deps: |
There was a problem hiding this comment.
you're already going through all the deps 2 lines above (enumerate), why not just do this in that loop?
There was a problem hiding this comment.
Because the line above we do the recursively called _install_libraries call from which we get the variable installed_deps (in_deps) which we need for this loop
| if board not in deps_arches[dep] and "*" not in deps_arches[dep]: | ||
| continue | ||
| dir_path = f"../{dep}@{in_deps[dep]}/" | ||
| dir_deps[board] += f"\t\t\t{dir_path}src\n" |
There was a problem hiding this comment.
see above, why not just let platformio handle all this and use the regular deps mechanism. We're introducing platformio but don't actually use it? all this code seems fragile and messy to maintain. Esp. when considering this also has to run on a PI and in the desktop client. Less == more.
| async with aiofiles.open(platformio_config_path, "w+") as _f: | ||
| libs = "" | ||
| includes = "" | ||
| for lib in installed_libs: |
There was a problem hiding this comment.
do we need to rebuild the whole project for every compile? Can't we have one project per board type (or even one project with several envs), and always compile from there? That way caching of compiled artifacts also works like it should.
There was a problem hiding this comment.
No because some projects require other libs then others, and libs could override (like having the same named headers) each other which would make it dangerous for stability
|
|
||
| logger.info("Installing libraries: %s@%s", library, version_required) | ||
|
|
||
| repo = list( |
There was a problem hiding this comment.
javascript-ism: please use a list comprehension. It's more readable and faster.
| library_dir = f"./arduino-libs/{repo['name']}@{repo['version']}" | ||
| async with httpx.AsyncClient() as client: | ||
| library_zip = zipfile.ZipFile( | ||
| io.BytesIO((await client.get(repo["url"])).content) |
There was a problem hiding this comment.
no error handling here? Also, please add timeouts for httpx calls.
No description provided.