Viewing Program

Program Data

Program thumbnail for "{props.program.title}"

Title	Using KPPM
ID	5200645795266560
Created	8 months ago, 1715639690113
Updated	7 months ago, 1715726191632
Code	{"files":[{"code":"print(\"\"\"\nKPPM (Khan Python Package Manager) is a library for importing dependency trees into your project.\n\nFor details on publishing a KPPM package, please see:\nhttps://www.khanacademy.org/python-program/-/6424545313079296\n==================\n\n\"\"\".lstrip())\n\n\n# Remove cached files for making sure it works without cache\n# from os import \n# files = listdir()\n# print(files)\n# for file in files:\n# if file != \"kppm.py\":\n# remove(file)\n\n# Simply add kppm.py with the following:\nfrom kppm import require\n\n# Now, we can require indexed KPPM libraries (notice we pass in globals as the last argument):\nprint(\"Importing a library from another program...\")\nawait require(\"kppm.library-example\", globals())\n\n# Right away, we can use the standard python syntax on the library:\nprint(\"Calling a function of the library:\")\nlibrary.MyCoolFunction()\nprint()\n\n# If you don't pass in globals, you have to call import on the library yourself:\n# await require(\"kppm.library-example\")\n# import library\n\n# The above also lets you import just a single function from the library:\n# from library_example import MyCoolFunction\n\n\nprint(\"We can also fuzzy search for a package.\")\nprint(\"Let's search the index for \\\"clear print function\\\"\")\nfrom kppm import search\nresults = await search(\"clear print function\")\ntopThree = results[:3]\n\nprint(\"Here are the top 3 matches and scores:\")\nprint(topThree)\nprint()\n\ntop = topThree[0][0]\nprint(\"We can also lookup all the info for a specific package, say for \" + top)\nfrom kppm import lookup \npackageInfo = await lookup(top)\nprint(packageInfo)\nprint()\n\nprint(\"Finally, you can simply drop into an interactive shell to use these search features:\")\nfrom kppm import shell\nawait shell()\n\n\n# All the following are valid syntax for importing multiple projects. Passing in globals is optional, it simply saves you having to `import` the file:\n# await require(\"kppm.library-example\", \"somebody.secondLibrary\")\n# await require([\"kppm.library-example\", \"somebody.secondLibrary\"])\n# await require([\"kppm.library-example\", \"somebody.secondLibrary\"], globals())\n# await require(\"kppm.library-example\", \"somebody.secondLibrary\", globals())\n\n# As long as globals are either the last argument, or not passed, you should be good.\n\n\n\n\n#\n# Quick links:\n# Publishing a KPPM library: https://www.khanacademy.org/python-program/-/4842362391674880\n# KPPM library template: https://www.khanacademy.org/python-program/-/4813755587936256\n# KPPM library index: https://www.khanacademy.org/python-program/-/4783660036964352\n# KPPM github: https://github.com/Reginald-Gillespie/KPPM\n# \n\n","filename":"main.py","language":"python"},{"code":"\"\"\"\n=============================\n=Khan Python Package Manager=\n=============================\nFor details: https://www.khanacademy.org/python-program/using-kppm/5200645795266560\n\nThis is the single KPPM utilities file needed to publish a library or use kppm.\n\"\"\"\n\n\n\n\n\n\n\n\n\n\n\n\"\"\"\nTechnical notes:\nThe filesystem persists through restarts, so we save json+python files here to minimize requests\nAt the very least, downloading and searching the KPPM database uses no \"bypasses\".\nLoading other code files requires requesting an proxy service to grab the code and send it to us, \n since KA's cors blocks us from dirrectly accessing the code ourselves.\n\n\n\"\"\"\n\n\n\n\nimport js\nfrom os import listdir\nfrom js import JSON # we'll need this to parse JS json later\nimport json # and we'll need this to parse json files for python\nimport re\nfrom math import sqrt\n# from js import pyfetch #imported later\n# from js import pyJSON #imported later\n\n\nqueries = {}\nkppmIndex = {}\nfileClaims = { # keep track of file names belonging to libraries to look for conflicts\n \"main.py\": \"local.local\"\n}\ninitialized = False\n\n# First, inject my pyfetch code - https://www.khanacademy.org/python-program/pyfetch-v1/5482075651751936\nfetchInjector = \"\"\"\n// We need to convert from a python dict to json... in javascript, on a Proxy object... This was a pain.\nfunction dict2Array(proxyDict) {\n // Check if this is a dict object, otherwise we can just return it as normal json\n if (proxyDict?.$$flags == 31) {\n const json = {}\n // debugger\n const keys = Array(...proxyDict.keys().mapping)\n keys.forEach( key => {\n value = proxyDict.get(key);\n // If this is another dict, convert it to an array too\n if (value?.$$flags) {\n // Now we know it's a proxy, it could be an array or json\n if (value.$$flags == 31) {\n json[key] = dict2Array(value)\n }\n // Otherwise we'll treat it like an array (may need to add more types later, TODO)\n else {\n json[key] = [...value]\n }\n } else {\n json[key] = value;\n }\n })\n return json;\n } else {\n return proxyDict;\n }\n}\n\nfunction noCorsFetch(url, options) {\n // Modify the URL to ignore cors\n noCorsURL = \"https://corsproxy.io/?\" + encodeURIComponent(url);\n\n jsOptions = dict2Array(options);\n\n return fetch(jsOptions?.disableCorsProxy ? url : noCorsURL, jsOptions);\n}\n\n// Add to global scope so we can extract it\nthis.pyfetch = noCorsFetch\n\n\nfunction pyJSONStringify(dict) {\n json = dict2Array(dict);\n return JSON.stringify(json);\n}\n\nthis.pyJSON = {\n \"stringify\": pyJSONStringify,\n}\n\n\"\"\"\njs.eval(fetchInjector)\n\n# Extract the new functions\nfrom js import pyfetch\nfrom js import pyJSON\n\n\n# Now comes the fun stuff\n\n# File utilities\ndef writeFile(name, content):\n content = str(content)\n file = open(name, \"w\")\n file.write(content)\n file.close()\n\ndef fileExists(name):\n return listdir(\"/home/pyodide\").count(name) > 0\n\ndef readFile(name):\n file = open(name, \"r\")\n content = file.read()\n file.close()\n return content\n\n# Other utilities\ndef log(message):\n print(\"\\n!! KPPM - \" + str(message) + \" !!\\n\")\n\ndef jsonToDict(jsJson):\n return json.loads(pyJSON.stringify(jsJson))\n\ndef flatten(t):\n result = []\n for item in t:\n if isinstance(item, (tuple, list)):\n result.extend(flatten(item))\n else:\n result.append(item)\n return tuple(result)\n\ndef removeSuffixes(text):\n pattern = r'(ed\|ing\|s\|able\|or\|ful\|less\|ly\|ty)\\b'\n words = text.split()\n processed_words = [re.sub(pattern, '', word) for word in words]\n result = ' '.join(processed_words)\n return result\n\ndef strictNormalize(query):\n # turned into smth for search queries, getting query and content as close as possible\n query = str(query)\n query = re.sub(r'[^A-Za-z\\d]', ' ', query)\n query = re.sub(r'\\s+', ' ', query)\n query = query.strip()\n query = query.lower()\n query = removeSuffixes(query)\n return query\n\nasync def getQueries():\n global queries\n if (fileExists(\"queries.json\")):\n queries = JSON.parse(readFile(\"queries.json\"))\n return\n request = await pyfetch(\"https://cdn.jsdelivr.net/gh/bhavjitChauhan/khan-api@safelist/queries.json\", {\"disableCorsProxy\": True})\n queries = await request.json()\n writeFile(\"queries.json\", pyJSON.stringify(queries))\n\nasync def getPythonProgramById(id=4893019299561472):\n response = await pyfetch('https://www.khanacademy.org/api/internal/graphql/pythonProgramQuery', { \n \"method\": 'POST', \n \"body\": pyJSON.stringify(\n { \n \"query\": queries.pythonProgramQuery, \n \"variables\": { \n \"programId\": id\n } \n }) \n })\n pythonProgramQuery = await response.json()\n program = pythonProgramQuery.data.program\n if program is None: print(\"WARNING - program \" + str(id) + \" is null\")\n return program\n\nasync def requireByID(id, requiredFiles=None, useCache=True):\n # convert files to list if it isn't one already\n if isinstance(requiredFiles, str):\n requiredFiles = [requiredFiles]\n\n if requiredFiles is not None and useCache:\n # List the current loaded files\n loadedFiles = listdir(\"/home/pyodide\")\n allFilesLoaded = True\n for file in requiredFiles:\n # if we found a file that is not loaded:\n if loadedFiles.count(file) == 0:\n allFilesLoaded = False\n if allFilesLoaded:\n return\n \n # parse filesnames - we'll add .py unless it already specifies the ending\n requiredFiles = [str(file)+\".py\" if file.count(\".\") == 0 else file for file in requiredFiles]\n \n program = await getPythonProgramById(id)\n if program is None: raise LookupError(\"Required import \" + str(id) + \" does not exist\")\n\n # tfw a random property is a string instead of json like the rest of the KA api\n files = JSON.parse(program.latestRevision.code).files\n \n for file in files:\n if (requiredFiles is None and file!=\"main.py\") or requiredFiles.count(file.filename) > 0:\n # We found the import file, now we write it to the pseudo filesystem\n output = open(\"/home/pyodide/\"+file.filename, \"w\")\n output.write(file.code)\n output.close()\n\nasync def loadKPPMIndex():\n global kppmIndex\n if (fileExists(\"kppmIndex.json\")):\n kppmIndex = json.loads(readFile(\"kppmIndex.json\"))\n return\n request = await pyfetch(\"https://raw.githubusercontent.com/Reginald-Gillespie/KPPM/main/KPPMIndex.json\", {\"disableCorsProxy\": True})\n kppmIndex = jsonToDict(await request.json())\n writeFile(\"kppmIndex.json\", pyJSON.stringify(kppmIndex))\n\nasync def loadDefinedDependencies():\n # Now, we need to search for the file with #defined dependencies\n files = listdir()\n for file in files:\n fileContents = readFile(file)\n depMatches = re.search(r\"^#\\sdefine\\s+dependencies[ \\t](.)\", fileContents, re.MULTILINE)\n packageNameMatch = re.search(r\"^#\\sdefine\\s+package[ \\t](.)$\", fileContents, re.MULTILINE) \n if depMatches and packageNameMatch:\n # If we found a file needing dependencies, import those. \n # This will run on all library files we import but their\n # files are cached so shouldn't be much overhead.\n dependencies = depMatches.group(1)\n dependencies = re.sub(r\"\\s+\", '', dependencies).strip()\n dependencies = dependencies.split(\",\") \n for dep in dependencies:\n if dep and dep != packageNameMatch.group(1): # prevent recursive error if infinitely requiring self\n await require(dep)\n \n\n# Initialize is called by require, or by the main.py of libraries\nasync def initialize(writeIndexToFilesystem=False):\n global fileClaims\n global initialized\n if initialized:\n return\n initialized = True\n # Fetch KA's latest queries\n await getQueries()\n # Load which files belong to which libraries\n if fileExists(\"fileClaims.json\"):\n fileClaims = json.loads(readFile(\"fileClaims.json\"))\n # Fetch library index\n await loadKPPMIndex()\n # Load dependencies if this is a library file that declares deps\n await loadDefinedDependencies()\n\nasync def require(args):\n if not initialized:\n await initialize()\n # Start by flattening input so that it's just lib1,lib2,lib3,globals\n args = flatten(args)\n\n # Check if we have globals\n libraryFilenames = [] # for importing later\n callerGlobals = False\n if (isinstance(args[-1], dict)):\n callerGlobals = args[-1]\n args = args[:-1]\n\n # Get required library objects from index \n for requestedLibrary in args:\n requestedLibrary = str(requestedLibrary).lower()\n try:\n requestedLibInfo = kppmIndex.get(requestedLibrary, False)\n # Make sure the library exists\n if not requestedLibInfo:\n log(\"Required library \" + str(requestedLibrary) + \" not found.\")\n else:\n # Check if this library conflicts or is already loaded\n requestedFilename = requestedLibInfo[\"file\"]\n if fileClaims.get(requestedFilename, False):\n if (fileClaims[requestedFilename] == requestedLibrary):\n # print(\"Library already loaded\")\n libraryFilenames.append(requestedFilename)\n pass # Library is already loaded (by something else in the dependency tree)\n else:\n # Library file is taken by another library\n log(\"Cannot load \" + str(requestedLibrary) + \", conflicts with \" + fileClaims[requestedFilename])\n \n else:\n # No conflicts, continue\n \n # Start by importing all dependencies of the library\n requestedLibDependencies = requestedLibInfo[\"dependsOn\"]\n # Prevent infinite dependency loop by depending on your own library\n if requestedLibDependencies.count(requestedLibrary) > 0:\n requestedLibDependencies.remove(requestedLibrary)\n for dep in requestedLibDependencies:\n await require(dep)\n\n # That may have erred, we don't care (could still run, not my problem if it doesn't =P)\n libraryProgramID = requestedLibInfo[\"programID\"]\n program = await getPythonProgramById(libraryProgramID)\n if program is None: \n log(\"Required library \" + str(requestedLibrary) + \" no longer exists on KA\")\n else:\n # Program exists\n libraryProgramFiles = json.loads(program.latestRevision.code)[\"files\"]\n # Create JSON map of files\n libraryProgramFilesJSON = {}\n for file in libraryProgramFiles:\n libraryProgramFilesJSON[file[\"filename\"]] = file[\"code\"]\n # Now get the library file from the map\n libraryFile = libraryProgramFilesJSON[requestedFilename]\n if not libraryFile:\n log(\"File \" + str(requestedFilename) + \" not found in library \" + str(requestedLibrary))\n\n else:\n # Library file exists, should be good to import it if it doesn't conflict with local files\n localFiles = listdir()\n if localFiles.count(requestedFilename) > 0:\n log(\"Library file for \" + str(requestedLibrary) + \" conflicts with local file \" + str(requestedFilename))\n\n else:\n # Finally time to write the dependency locally\n fileClaims[requestedFilename] = str(requestedLibrary)\n writeFile(requestedFilename, libraryFile)\n libraryFilenames.append(requestedFilename)\n \n except Exception as e:\n log(\"Exception while loading \" + str(requestedLibrary))\n # print(e)\n # end import single dependency\n\n # Write the fileClaims to filesystem so it persists reloads\n writeFile(\"fileClaims.json\", json.dumps(fileClaims))\n\n # Import the dependencies into globals if we access\n if callerGlobals:\n libraryFilenames = [(file+\" \")[:file.find(\".\")] for file in libraryFilenames]\n for library in libraryFilenames:\n module = __import__(library)\n callerGlobals[library] = module\n\nasync def search(query, minScore=0.01):\n # Needs to be awaiting simply so we can initialize\n if not initialized:\n await initialize()\n\n # Behind the scenes we'll filter the search and queries down to A-Za-z\\d.\n # If you want a different search algorithm, just read kppmIndex.js from the filesystem and search it yourself\n query = set(strictNormalize(query).split())\n\n allLibraries = list(kppmIndex.keys())\n matches = []\n \n for library in allLibraries:\n libraryInfo = kppmIndex[library]\n description = set(strictNormalize(libraryInfo[\"description\"]).split())\n\n # Go for text match\n intersection = query.intersection(description)\n union = query.union(description)\n matchScore = len(intersection) / len(union) # between 0 and 1\n\n # Weight matches by votes a little - we do want to avoid returning matches that only have a lot of votes.\n votesWeight = 0.15 # between 0 and 1\n votes = libraryInfo[\"votesAtLastIndex\"]\n voteScore = votes / sqrt(150 + votes*2) # between 0 and 1\n \n if matchScore >= minScore:\n # we'll add vote weight to total weight only after checking match threshold\n totalScore = votesWeightvoteScore + (1-votesWeight)*matchScore\n matches.append([library, round(totalScore, 3)])\n\n # sort matches\n matches.sort(key=lambda m: m[1], reverse=True)\n \n return matches\n\nasync def lookup(name:str, returnAsJson=False):\n if not initialized:\n await initialize()\n \n package = kppmIndex.get(name, False)\n if not package:\n if returnAsJson:\n return {\"not found\":name}\n else:\n return \"Package \\\"\" + name + \"\\\" was not found.\"\n else:\n # We found it:\n if returnAsJson:\n return package\n else:\n response = \"\"\n fields = list(package.keys())\n for field in fields:\n response += field + \": \\\"\" + str(package[field]) + \"\\\"\\n\"\n return response.strip()\n\nasync def shell():\n if not initialized:\n await initialize()\n print(\"===============\")\n print(\"=KPPM Shell v1=\")\n print(\"===============\")\n print(\"Type HELP for help\")\n while True:\n print(\"____\")\n command = input(\"KPPM> \")\n print(\"\")\n command = command.lower().strip()\n try:\n command = command.split()\n args = \" \".join(command[1:])\n \n if len(command) == 0:\n print(\"Please send a command - type HELP for help\")\n elif command[0] == \"help\":\n print(\"Available commands:\")\n print(\"SEARCH <query> - Search for a library\")\n print(\"LOOKUP <library> - Lookup library info by name\")\n print(\"LOAD <library> - Load a library file into the filesystem\")\n elif len(command) == 1:\n print(\"Please send arguments with the command - type HELP for help\")\n elif command[0] == \"search\":\n results = (await search(args))[:15] # limit to top 15\n if results:\n output = \"\"\n for result in results:\n output += \"Package: \" + result[0] + \"\\nScore: \" + str(result[1]) + \"\\n\"\n print(output)\n else:\n print(\"No packages found for that query.\")\n elif command[0] == \"lookup\":\n results = await lookup(args)\n print(results)\n elif command[0] == \"load\":\n await require(args)\n print(\"Attempted to load \" + args)\n else:\n print(\"Unknown command - type HELP for help\")\n except Exception as e:\n print(e)\n print(\"Error - type HELP for help\")\n \n\n","filename":"kppm.py","language":"python"}]}
Folds
Fork	false
Key	ag5zfmtoYW4tYWNhZGVteXIXCxIKU2NyYXRjaHBhZBiAgOup276eCQw
Votes	2
Spin-Offs	0
Type	PYTHON
Width	400
Height	400
User Flagged	false
Origin Scratchpad	null
Hidden From Hotlist	false
Restricted Posting	true
By Child	false
Author Nick	KPPM Workflow Account
Author Name	KPPM
Author ID	kaid_4531886152410793423958741
Author Profile Access	COACH

Archive Data

Added	7 months ago, 1715734771846
Updated	7 months ago, 1715734866497
Source Deleted	unknown