[{"_1":2,"_3":-5,"_4":-5},"loaderData",{"_5":6,"_7":8,"_9":10},"actionData","errors","root",{"_144":145,"_146":147},"routes/_app.x.$handle.$name",{"_16":17,"_20":21,"_82":65,"_83":84,"_63":-7,"_85":-5,"_13":86,"_87":88,"_11":89,"_90":91,"_92":93,"_94":-5,"_95":-5,"_96":-5,"_97":26,"_98":99,"_100":101,"_102":-5,"_103":-5,"_104":105},"routes/_app.x.$handle.$name.search.$(branch)",{"_11":12,"_13":14,"_15":-5,"_16":17},"files",[],"project",{"_18":19,"_20":21,"_22":23,"_24":-5,"_25":26,"_27":28,"_29":28,"_30":31,"_32":-5,"_33":34,"_35":-5,"_36":26,"_37":-5,"_38":-5,"_39":40,"_41":42,"_43":-5,"_44":-5,"_45":-5,"_46":-5,"_47":48,"_49":48,"_50":51,"_52":51,"_53":51,"_54":48,"_55":48,"_56":57,"_58":59,"_60":61,"_62":26},"query","handle","ziggyware","id","8795e818-4d24-11f1-b75f-ee650bb23af1","name","cal","createdAt",["D",1778495471409],"deletedAt","legacyEnvVars",false,"privacy","public","httpPrivacy","description","cognitive application layer","imageUrl","userId","2c6d46b0-54dd-11f0-a86c-6fbc2b29331a","pinnedFileId","pinPreview","pinMigrated","screenshotVersion","likeCount",0,"orgId","850db142-aa5e-4914-bceb-1e2e085ae4c5","tursoName","tursoHostname","tursoMeasuredAt","pinned","userPerm","r","valPerm","blobPerm","rw","emailPerm","sqlitePerm","projectPerm","telemetryPerm","user",{"_16":17,"_81":77},"org",{"_16":17,"_75":-5,"_76":77,"_33":34,"_78":79,"_80":26},"branch",{"_63":64,"_20":65,"_18":66,"_67":19,"_68":-5,"_69":-5,"_70":-5,"_71":-5,"_72":-5,"_73":21,"_74":17},"isOwned","version",106,"main","8797818c-4d24-11f1-b75f-ee650bb23af1","projectId","forkedBranchId","forkedBranchVersion","forkedProjectPrivacy","forkedProjectId","forkedProjectOrgId","projectName","userHandle","displayName","avatar","https://img.clerk.com/eyJ0eXBlIjoicHJveHkiLCJzcmMiOiJodHRwczovL2ltYWdlcy5jbGVyay5kZXYvb2F1dGhfZ2l0aHViL2ltZ18yekJHRDE5SGVoUHUzUENQaXY2MTRkT3h0cncifQ","tier","free","canRestrictVals","profileImageUrl","branchName","path","","adoptOrgId",{"_18":19,"_20":21,"_22":140,"_24":-5,"_25":26,"_27":28,"_29":28,"_30":31,"_32":-5,"_33":34,"_35":-5,"_36":26,"_37":-5,"_38":-5,"_39":40,"_41":42,"_43":-5,"_44":-5,"_45":-5,"_46":-5,"_47":48,"_49":48,"_50":51,"_52":51,"_53":51,"_54":48,"_55":48,"_56":141,"_58":142,"_60":143,"_62":26},"tree",[135],[135],"counts",{"_124":40,"_125":126,"_127":40,"_128":129,"_130":131},"contents",[106],"file","upstreamProject","upstreamBranch","emptyEnvVars","flags",[],"hasHttpFiles",true,"grants","orgs","customDomains",{},{"_18":107,"_20":108,"_109":110,"_111":-5,"_112":66,"_113":64,"_114":101,"_22":115,"_63":116,"_117":118,"_119":120,"_121":26,"_122":-5,"_83":123},"5c939d00-4d32-11f1-b1b3-ee650bb23af1","main.ts","fileId","879a28ec-4d24-11f1-b75f-ee650bb23af1","parentId","branchId","branchVersion","latest",["D",1778501412197],105,"fileType","http","content","import {\n HtmlLogger,\n ResponseHtml,\n} from \"https://esm.town/v/ziggyware/html-logger/main.ts?v=333d323\";\n\n// ---------------- CAL RUNTIME (your code, unchanged) ----------------\n\n// A constraint says: variable ∈ allowedValues, with an explanation.\nexport interface Constraint {\n id: string;\n variable: string;\n allowed: Set;\n source: string; // where it came from (observation, rule, etc.)\n note?: string;\n}\n\n// A contradiction snapshot: what we tried to add, and what broke.\nexport interface Contradiction {\n constraint: Constraint;\n previousDomain: Set;\n}\n\n// Core CAL state: domains + constraints + contradictions.\nexport class CALRuntime {\n private domains = new Map>();\n private constraints: Constraint[] = [];\n private contradictions: Contradiction[] = [];\n\n // Initialize a variable with an initial domain (if not present).\n initDomain(variable: string, initial: Iterable) {\n if (!this.domains.has(variable)) {\n this.domains.set(variable, new Set(initial));\n }\n }\n\n // Add a single constraint: intersect domain with allowed values.\n addConstraint(c: Constraint) {\n this.constraints.push(c);\n\n const prev = this.domains.get(c.variable);\n if (!prev) {\n // No prior domain: just adopt allowed set as domain.\n this.domains.set(c.variable, new Set(c.allowed));\n return;\n }\n\n const before = new Set(prev);\n const next = new Set();\n for (const v of prev) if (c.allowed.has(v)) next.add(v);\n\n if (next.size === 0) {\n // Contradiction: record, but do not destroy previous domain.\n this.contradictions.push({\n constraint: c,\n previousDomain: before,\n });\n return;\n }\n\n this.domains.set(c.variable, next);\n }\n\n // Add multiple constraints at once.\n addConstraints(cs: Constraint[]) {\n for (const c of cs) this.addConstraint(c);\n }\n\n // Observation → constraints (user-defined extractor).\n // Example: { animal: \"cat\", legs: \"4\" } → constraints on \"animal\", \"legs\".\n addObservation(\n obs: Record,\n source = \"observation\",\n note?: string,\n ) {\n const cs: Constraint[] = [];\n for (const [k, v] of Object.entries(obs)) {\n cs.push({\n id: `${source}:${k}:${v}:${Date.now()}:${Math.random()}`,\n variable: k,\n allowed: new Set([v]),\n source,\n note,\n });\n }\n this.addConstraints(cs);\n }\n\n // Query current possible values for a variable, plus supporting constraints.\n query(variable: string) {\n const domain = this.domains.get(variable);\n if (!domain) {\n return { variable, possible: [], supporting: [] as Constraint[] };\n }\n\n const supporting = this.constraints.filter((c) => c.variable === variable);\n return {\n variable,\n possible: [...domain],\n supporting,\n };\n }\n\n // Get all current domains.\n getDomains() {\n const out: Record = {};\n for (const [k, v] of this.domains.entries()) {\n out[k] = [...v];\n }\n return out;\n }\n\n // Get all constraints.\n getConstraints() {\n return [...this.constraints];\n }\n\n // Get all contradictions.\n getContradictions() {\n return [...this.contradictions];\n }\n}\n\n// ---------------- CAL LANGUAGE LAYER ----------------\n\ntype POS = \"N\" | \"V\" | \"DET\" | \"PRON\" | \"OTHER\";\n\ninterface LexEntry {\n lemma: string;\n pos: POS;\n}\n\n// Tiny lexicon; in a real system you’d hydrate this from HF vocab / tags.\nconst LEXICON: Record = {\n the: { lemma: \"the\", pos: \"DET\" },\n a: { lemma: \"a\", pos: \"DET\" },\n an: { lemma: \"an\", pos: \"DET\" },\n\n dog: { lemma: \"dog\", pos: \"N\" },\n dogs: { lemma: \"dog\", pos: \"N\" },\n cat: { lemma: \"cat\", pos: \"N\" },\n cats: { lemma: \"cat\", pos: \"N\" },\n ball: { lemma: \"ball\", pos: \"N\" },\n balls: { lemma: \"ball\", pos: \"N\" },\n man: { lemma: \"man\", pos: \"N\" },\n men: { lemma: \"man\", pos: \"N\" },\n woman: { lemma: \"woman\", pos: \"N\" },\n women: { lemma: \"woman\", pos: \"N\" },\n\n chased: { lemma: \"chase\", pos: \"V\" },\n chase: { lemma: \"chase\", pos: \"V\" },\n chases: { lemma: \"chase\", pos: \"V\" },\n\n kicked: { lemma: \"kick\", pos: \"V\" },\n kick: { lemma: \"kick\", pos: \"V\" },\n kicks: { lemma: \"kick\", pos: \"V\" },\n\n saw: { lemma: \"see\", pos: \"V\" },\n see: { lemma: \"see\", pos: \"V\" },\n sees: { lemma: \"see\", pos: \"V\" },\n\n he: { lemma: \"he\", pos: \"PRON\" },\n she: { lemma: \"she\", pos: \"PRON\" },\n they: { lemma: \"they\", pos: \"PRON\" },\n};\n\ninterface Token {\n surface: string;\n lemma: string;\n pos: POS;\n}\n\ninterface ParsedEvent {\n id: string; // e.g. \"e1\"\n verb: string; // lemma\n agent?: string; // lemma\n patient?: string; // lemma\n}\n\n// Very small tokenizer + tagger using the lexicon.\nfunction tokenize(sentence: string): Token[] {\n return sentence\n .toLowerCase()\n .replace(/[^a-z\\s]/g, \"\")\n .split(/\\s+/)\n .filter(Boolean)\n .map((surface) => {\n const entry = LEXICON[surface];\n if (!entry) return { surface, lemma: surface, pos: \"OTHER\" as POS };\n return { surface, lemma: entry.lemma, pos: entry.pos };\n });\n}\n\n// Extremely minimal SVO parser: DET? N / PRON → V → DET? N\nfunction parseSentenceToEvent(\n sentence: string,\n id: string,\n): ParsedEvent | null {\n const tokens = tokenize(sentence);\n const verbIndex = tokens.findIndex((t) => t.pos === \"V\");\n if (verbIndex === -1) return null;\n\n const verb = tokens[verbIndex].lemma;\n\n const before = tokens.slice(0, verbIndex);\n const after = tokens.slice(verbIndex + 1);\n\n const agent = before.find((t) => t.pos === \"N\" || t.pos === \"PRON\")?.lemma;\n const patient = after.find((t) => t.pos === \"N\" || t.pos === \"PRON\")?.lemma;\n\n return { id, verb, agent, patient };\n}\n\n// Map a ParsedEvent → CAL constraints\nfunction eventToConstraints(ev: ParsedEvent, source = \"lang\"): Constraint[] {\n const cs: Constraint[] = [];\n const now = Date.now();\n\n // Event verb\n cs.push({\n id: `${source}:${ev.id}:verb:${ev.verb}:${now}`,\n variable: `event:${ev.id}:verb`,\n allowed: new Set([ev.verb]),\n source,\n note: \"event verb\",\n });\n\n if (ev.agent) {\n cs.push({\n id: `${source}:${ev.id}:agent:${ev.agent}:${now}`,\n variable: `event:${ev.id}:agent`,\n allowed: new Set([ev.agent]),\n source,\n note: \"event agent\",\n });\n\n // Agent role knowledge\n cs.push({\n id: `${source}:role:agent:${ev.agent}:${now}`,\n variable: `lemma:${ev.agent}:role`,\n allowed: new Set([\"agent\"]),\n source,\n note: \"lemma can be an agent\",\n });\n }\n\n if (ev.patient) {\n cs.push({\n id: `${source}:${ev.id}:patient:${ev.patient}:${now}`,\n variable: `event:${ev.id}:patient`,\n allowed: new Set([ev.patient]),\n source,\n note: \"event patient\",\n });\n\n cs.push({\n id: `${source}:role:patient:${ev.patient}:${now}`,\n variable: `lemma:${ev.patient}:role`,\n allowed: new Set([\"patient\"]),\n source,\n note: \"lemma can be a patient\",\n });\n }\n\n // Verb frame knowledge: verb takes agent + patient\n cs.push({\n id: `${source}:frame:${ev.verb}:${now}`,\n variable: `verb:${ev.verb}:frame`,\n allowed: new Set([\"agent-patient\"]),\n source,\n note: \"verb frame\",\n });\n\n return cs;\n}\n\n// High-level: learn a sentence into CAL\nfunction learnSentence(cal: CALRuntime, sentence: string, id: string) {\n const ev = parseSentenceToEvent(sentence, id);\n if (!ev) return;\n const cs = eventToConstraints(ev, \"lang\");\n cal.addConstraints(cs);\n}\n\n// ---------------- VAL HANDLER: \"TRAIN\" ON A CORPUS ----------------\n\nexport default function handler() {\n const cal = new CALRuntime();\n const logger = new HtmlLogger();\n\n // In a real system, this corpus would be streamed from a Hugging Face dataset.\n const corpus = [\n \"The dog chased the ball.\",\n \"The cat chased the ball.\",\n \"The man kicked the ball.\",\n \"The woman saw the dog.\",\n \"He chased the cat.\",\n \"They saw the ball.\",\n ];\n\n corpus.forEach((sentence, idx) => {\n learnSentence(cal, sentence, `e${idx + 1}`);\n });\n\n // Inspect what CAL has learned about verbs, roles, and events.\n const domains = cal.getDomains();\n\n logger.log({ corpus });\n logger.log({ domains });\n\n // Example: what does CAL think about the verb \"chase\"?\n logger.log({\n verb_chase_frame: cal.query(\"verb:chase:frame\"),\n });\n\n // Example: what lemmas can be agents / patients?\n logger.log({\n lemma_dog_role: cal.query(\"lemma:dog:role\"),\n lemma_ball_role: cal.query(\"lemma:ball:role\"),\n lemma_cat_role: cal.query(\"lemma:cat:role\"),\n lemma_man_role: cal.query(\"lemma:man:role\"),\n lemma_woman_role: cal.query(\"lemma:woman:role\"),\n });\n\n // Example: what does CAL know about a specific event?\n logger.log({\n e1: {\n verb: cal.query(\"event:e1:verb\"),\n agent: cal.query(\"event:e1:agent\"),\n patient: cal.query(\"event:e1:patient\"),\n },\n e2: {\n verb: cal.query(\"event:e2:verb\"),\n agent: cal.query(\"event:e2:agent\"),\n patient: cal.query(\"event:e2:patient\"),\n },\n });\n\n return ResponseHtml(\"\" + logger.toHtml() + \"\");\n}","deleted","label",[108],"pulls","branches",1,"forks","envVars",{"_132":40,"_134":40},"connections",{"_132":40,"_133":40},"count","unfilledCount","emptyCount",{"_18":107,"_20":108,"_109":110,"_111":-5,"_112":66,"_113":64,"_114":101,"_22":136,"_63":116,"_117":118,"_119":120,"_121":26,"_122":-5,"_137":138,"_83":139},["D",1778501412197],"children",[],[108],["D",1778495471409],{"_16":17,"_81":77},{"_16":17,"_75":-5,"_76":77,"_33":34,"_78":79,"_80":26},{"_63":64,"_20":65,"_18":66,"_67":19,"_68":-5,"_69":-5,"_70":-5,"_71":-5,"_72":-5,"_73":21,"_74":17},"ctx",{"_56":-5,"_103":177,"_178":179,"_180":181,"_182":26,"_183":26,"_184":26,"_185":101,"_186":-5,"_187":-7,"_188":-7,"_189":40,"_190":26,"_191":26,"_192":-5,"_193":26,"_194":26,"_195":26,"_196":-5},"ENV",{"_148":149,"_150":151,"_152":65,"_153":154,"_155":156,"_157":158,"_159":160,"_161":162,"_163":164,"_165":166,"_167":168,"_169":170,"_171":172,"_173":174,"_175":176},"REACT_APP_SENTRY_DSN","https://586020d86f05489db63a554690acf1e9@o1398359.ingest.sentry.io/6724499","RENDER_GIT_COMMIT","e87eb04e707cf64e7e532c8e8ec395f209e17f15","RENDER_GIT_BRANCH","POSTHOG_API_TOKEN","phc_nmoyuAkStEZuUp69rYlg1DZgrIt19Gx0Y4Hwf8tGgJP","POSTHOG_API_HOST","https://wart.val.town","WEB_API_TEMPLATE","https://{val}.web.val.run","APP_URL","https://www.val.town","MODULE_URL","https://esm.town","VANITY_WEB_API_TEMPLATE","https://{label}.val.run/","DENOLS_URL","https://lsp.val.town","NODE_ENV","production","CLOUDFLARE_TURNSTILE_SITE_KEY","0x4AAAAAACtaIZJH7s9rBMZ6","GOOGLE_CLIENT_ID","875989696899-jtsqkhlbdv5jrlsk8631p7ipjg3c6o62.apps.googleusercontent.com","STRIPE_PUBLISHABLE_KEY","pk_live_51N2zvWEpZTTLipIkVEwXrUG9AT0yeOOEYxszarZT1jpsZV58Ly6wlg3SltZpHN26CUalOBmUSY7adQRQiSkCZz2500qSXNkf8Y","API_URL","https://api.val.town",[],"preferencesKeybindings","default","preferencesTheme","light","preferencesFixers","preferencesCopilot","preferencesSelectionAi","preferencesCloseBrackets","preferencesChatPanelState","preferencesDprint","preferencesDenofmt","notificationCount","flagDenols","isBusinessOrgMember","businessOrgHandle","isBusinessOrgAdmin","isHigherTownieTier","navCollapsed","currentOrgId"]

Search