From 970a446f4e58c822ab57aa1480bfa30848f4aa02 Mon Sep 17 00:00:00 2001 From: Hermes Agent Date: Sat, 26 Sep 2026 04:09:02 +0000 Subject: [PATCH] fix: release GET KEY expression temporaries --- docs/CURRENT_STATUS.md | 2 + docs/HARDENING.md | 4 ++ src/claro.c | 2 +- tools/validate_get_key_expression_cleanup.py | 51 ++++++++++++++++++++ 4 files changed, 58 insertions(+), 1 deletion(-) create mode 100644 tools/validate_get_key_expression_cleanup.py diff --git a/docs/CURRENT_STATUS.md b/docs/CURRENT_STATUS.md index 4e21d78..34bd2f7 100644 --- a/docs/CURRENT_STATUS.md +++ b/docs/CURRENT_STATUS.md @@ -42,6 +42,8 @@ The `RUN COMMAND` path now decodes POSIX `pclose()` wait status before storing ` `RUN COMMAND` is documented and validated as a trusted-code capability: it executes shell commands with the user's permissions and is not a sandbox. Claro does not claim untrusted-script safety or use a fragile blacklist sanitizer. Focused documentation coverage is `tools/validate_trusted_command_docs.py`. +`GET ... KEY ... AS ...` now releases its temporary map and key values after lookup. `python3 tools/validate_get_key_expression_cleanup.py` reproduced 1,004,000 leaked bytes before the change and passes after it with ASan/UBSan/LSan enabled; this is one narrow expression-temporary cleanup slice, not a claim that all runtime allocations are leak-free. + ## Feature matrix ### Beginner scripting core diff --git a/docs/HARDENING.md b/docs/HARDENING.md index 3b1255b..6df0bed 100644 --- a/docs/HARDENING.md +++ b/docs/HARDENING.md @@ -88,3 +88,7 @@ python3 tools/validate_ask_prompt_cleanup.py ``` The validator runs 2,000 prompt/input operations with AddressSanitizer, UndefinedBehaviorSanitizer, and LeakSanitizer enabled. The dedicated validator passes. A separate malformed-input sanitizer smoke run (`gcc -std=c99 -O0 -g -fsanitize=address,undefined src/claro.c -o ... -lm` followed by the generated malformed script) exits 1 due to two existing leaked error-message strings (140 bytes total) allocated in `rt_error` at `src/claro.c:142`; this is outside the ASK cleanup slice and remains a blocker to sanitizer-clean malformed-input validation. + +## GET KEY temporary-value cleanup + +`GET ... KEY ... AS ...` releases its evaluated map copy and key value after storing the selected value. `python3 tools/validate_get_key_expression_cleanup.py` runs 2,000 lookups under ASan/UBSan/LSan; before the cleanup it reproduced 1,004,000 bytes leaked, and after it passes with the expected output and no reported leaks. diff --git a/src/claro.c b/src/claro.c index 1265307..b513e91 100644 --- a/src/claro.c +++ b/src/claro.c @@ -438,7 +438,7 @@ static void exec_line(Runtime *rt,Program *p,int *pcp,const char *raw){ char buf if(strcmp(up,"IMPORT")==0){ const char *pcur=t+6; char *path=unquote_token(&pcur); char *ns=NULL; while(*pcur&&isspace((unsigned char)*pcur)) pcur++; if(starts_ci(pcur,"AS")){ pcur+=2; ns=xstrdup(trim_inplace((char*)pcur)); } do_import(rt,path,ns,p->path,pc+1); free(path); free(ns); return; } if(strcmp(up,"ADD")==0){ const char *to=find_word_ci(t,"TO"); if(to){ char *ex=substr(t+3,to); char *var=xstrdup(trim_inplace((char*)to+2)); Value v=eval_expr(rt,ex); Value l=rt_get(rt,var); if(l.type==V_LIST){ list_add(l.list,v); rt_set(rt,var,l); } free(ex); free(var); value_free(v); value_free(l);} return; } if(strcmp(up,"COUNT")==0){ const char *as=find_word_ci(t,"AS"); if(as){ if(starts_ci(t+5," ARGUMENTS")){ char *var=xstrdup(trim_inplace((char*)as+2)); rt_set(rt,var,v_num(rt->script_argc)); free(var); } else { char *ex=substr(t+5,as); char *var=xstrdup(trim_inplace((char*)as+2)); Value v=eval_expr(rt,ex); rt_set(rt,var,v_num(v.type==V_LIST?v.list->count:(v.type==V_MAP?v.map->count:0))); value_free(v); free(ex); free(var); } } return; } - if(strcmp(up,"GET")==0){ const char *at=find_word_ci(t,"AT"), *key=find_word_ci(t,"KEY"), *as=find_word_ci(t,"AS"); if(as&&starts_ci(t+3," ALL ARGUMENTS")){ char *var=xstrdup(trim_inplace((char*)as+2)); Value arr=v_list(); int i; for(i=0;iscript_argc;i++) list_add(arr.list,v_str(rt->script_argv[i])); rt_set(rt,var,arr); free(var); } else if(as&&starts_ci(t+3," ARGUMENT")){ char *ie=substr(t+12,as); char *var=xstrdup(trim_inplace((char*)as+2)); int idx=(int)v_number(eval_expr(rt,ie)); if(idx>=1&&idx<=rt->script_argc) rt_set(rt,var,v_str(rt->script_argv[idx-1])); else rt_set(rt,var,v_str("")); free(ie); free(var); } else if(as&&starts_ci(t+3," ENV")){ const char *pcur=t+7; char *name=substr(pcur,as); char *var=xstrdup(trim_inplace((char*)as+2)); Value nv=eval_expr(rt,name); char *ns=v_to_string(nv); const char *ev=getenv(ns); rt_set(rt,var,v_str(ev?ev:"")); free(ns); free(name); free(var); } else if(as&&at){ char *ce=substr(t+3,at); char *ie=substr(at+2,as); char *var=xstrdup(trim_inplace((char*)as+2)); Value c=eval_expr(rt,ce); int idx=(int)v_number(eval_expr(rt,ie)); if(c.type==V_LIST && idx>=1 && idx<=c.list->count) rt_set(rt,var,c.list->items[idx-1]); else { Value empty=v_str(""); rt_set(rt,var,empty); value_free(empty); } value_free(c); free(ce); free(ie); free(var); } else if(as&&key){ char *ce=substr(t+3,key); char *ke=substr(key+3,as); char *var=xstrdup(trim_inplace((char*)as+2)); Value c=eval_expr(rt,ce); Value kv=eval_expr(rt,ke); char *ks=v_to_string(kv); if(c.type==V_MAP) rt_set(rt,var,map_get(c.map,ks)); else rt_set(rt,var,v_str("")); free(ks); free(ce); free(ke); free(var); } return; } + if(strcmp(up,"GET")==0){ const char *at=find_word_ci(t,"AT"), *key=find_word_ci(t,"KEY"), *as=find_word_ci(t,"AS"); if(as&&starts_ci(t+3," ALL ARGUMENTS")){ char *var=xstrdup(trim_inplace((char*)as+2)); Value arr=v_list(); int i; for(i=0;iscript_argc;i++) list_add(arr.list,v_str(rt->script_argv[i])); rt_set(rt,var,arr); free(var); } else if(as&&starts_ci(t+3," ARGUMENT")){ char *ie=substr(t+12,as); char *var=xstrdup(trim_inplace((char*)as+2)); int idx=(int)v_number(eval_expr(rt,ie)); if(idx>=1&&idx<=rt->script_argc) rt_set(rt,var,v_str(rt->script_argv[idx-1])); else rt_set(rt,var,v_str("")); free(ie); free(var); } else if(as&&starts_ci(t+3," ENV")){ const char *pcur=t+7; char *name=substr(pcur,as); char *var=xstrdup(trim_inplace((char*)as+2)); Value nv=eval_expr(rt,name); char *ns=v_to_string(nv); const char *ev=getenv(ns); rt_set(rt,var,v_str(ev?ev:"")); free(ns); free(name); free(var); } else if(as&&at){ char *ce=substr(t+3,at); char *ie=substr(at+2,as); char *var=xstrdup(trim_inplace((char*)as+2)); Value c=eval_expr(rt,ce); int idx=(int)v_number(eval_expr(rt,ie)); if(c.type==V_LIST && idx>=1 && idx<=c.list->count) rt_set(rt,var,c.list->items[idx-1]); else { Value empty=v_str(""); rt_set(rt,var,empty); value_free(empty); } value_free(c); free(ce); free(ie); free(var); } else if(as&&key){ char *ce=substr(t+3,key); char *ke=substr(key+3,as); char *var=xstrdup(trim_inplace((char*)as+2)); Value c=eval_expr(rt,ce); Value kv=eval_expr(rt,ke); char *ks=v_to_string(kv); if(c.type==V_MAP){ Value got=map_get(c.map,ks); rt_set(rt,var,got); value_free(got); } else rt_set(rt,var,v_str("")); free(ks); value_free(c); value_free(kv); free(ce); free(ke); free(var); } return; } if(strcmp(up,"PUT")==0){ const char *key=find_word_ci(t,"KEY"), *val=find_word_ci(t,"VALUE"), *as=find_word_ci(t,"AS"), *into=find_word_ci(t,"INTO"); if(key&&val){ char *me=substr(t+3,key); char *ke=substr(key+3,val); char *ve=xstrdup(val+5); Value m=eval_expr(rt,me); Value k=eval_expr(rt,ke); Value v=eval_expr(rt,ve); char *ks=v_to_string(k); if(m.type==V_MAP){ char *mn=trim_inplace(me); map_put(m.map,ks,v); rt_set(rt,mn,m); } free(ks); free(me); free(ke); free(ve); value_free(m); value_free(k); value_free(v); } else if(as&&into){ char *ke=substr(t+3,as); char *ve=substr(as+2,into); char *mn=xstrdup(trim_inplace((char*)into+4)); Value m=rt_get(rt,mn); Value k=eval_expr(rt,ke); Value v=eval_expr(rt,ve); char *ks=v_to_string(k); if(m.type==V_MAP){ map_put(m.map,ks,v); rt_set(rt,mn,m); } free(ks); free(ke); free(ve); free(mn); } return; } if(strcmp(up,"WRITE")==0||strcmp(up,"APPEND")==0){ if(starts_ci(t+strlen(w)," FILE")){ const char *pcur=t+strlen(w)+5; const char *with=find_word_ci(pcur,"WITH"); if(with){ char *pe=substr(pcur,with); char *ve=xstrdup(with+4); Value pv=eval_expr(rt,pe); Value vv=eval_expr(rt,ve); char *path=v_to_string(pv), *text=v_to_string(vv); FILE *f=fopen(path,strcmp(up,"APPEND")==0?"ab":"wb"); if(!f) rt_error(rt,p->path,pc+1,"Could not write file: %s",strerror(errno)); else { fwrite(text,1,strlen(text),f); fclose(f);} free(path); free(text); free(pe); free(ve); } } return; } if(strcmp(up,"READ")==0){ const char *pcur=t+4; const char *as=find_word_ci(pcur,"AS"); if(as&&starts_ci(pcur," FILE")){ char *pe=substr(pcur+5,as); char *var=xstrdup(trim_inplace((char*)as+2)); Value pv=eval_expr(rt,pe); char *path=v_to_string(pv); FILE *f=fopen(path,"rb"); if(!f) rt_error(rt,p->path,pc+1,"Could not read file: %s",strerror(errno)); else { Str b; int c; str_init(&b); while((c=fgetc(f))!=EOF) str_ch(&b,(char)c); fclose(f); rt_set(rt,var,v_str(b.s?b.s:"")); free(b.s);} free(path); free(pe); free(var); } return; } diff --git a/tools/validate_get_key_expression_cleanup.py b/tools/validate_get_key_expression_cleanup.py new file mode 100644 index 0000000..c8fbd07 --- /dev/null +++ b/tools/validate_get_key_expression_cleanup.py @@ -0,0 +1,51 @@ +#!/usr/bin/env python3 +"""Regression check for GET KEY expression temporaries.""" +from pathlib import Path +import os +import subprocess +import tempfile + +ROOT = Path(__file__).resolve().parent.parent + + +def main() -> int: + if os.name == "nt": + print("SKIP: LeakSanitizer check is POSIX-only") + return 0 + with tempfile.TemporaryDirectory(prefix="claro-get-key-cleanup-") as tmp: + tmp_path = Path(tmp) + binary = tmp_path / "claro-lsan" + script = tmp_path / "get_key_expressions.claro" + script.write_text( + 'SET items TO MAP\nPUT items KEY "selected" VALUE "value"\n' + + 'GET items KEY "selected" AS selected\n' * 2000 + + "SAY selected\n", + encoding="utf-8", + ) + build = subprocess.run( + ["gcc", "-std=c99", "-O0", "-g", "-fsanitize=address,undefined", + "src/claro.c", "-o", str(binary), "-lm"], + cwd=ROOT, text=True, capture_output=True, + ) + if build.returncode: + print(build.stdout, end="") + print(build.stderr, end="") + return build.returncode + run = subprocess.run( + [str(binary), str(script)], cwd=ROOT, text=True, capture_output=True, + env={**os.environ, "ASAN_OPTIONS": "detect_leaks=1"}, + ) + if run.returncode or "LeakSanitizer" in run.stderr or "SUMMARY:" in run.stderr: + print("FAIL: GET KEY expression values still leak") + print(run.stdout, end="") + print(run.stderr, end="") + return 1 + if run.stdout != "value\n": + print(f"FAIL: unexpected GET KEY probe output: {run.stdout!r}") + return 1 + print("PASS: GET KEY expression values are released") + return 0 + + +if __name__ == "__main__": + raise SystemExit(main())