Skip to content

Add libclang API to store and query CAS objects #8937

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
wants to merge 1 commit into from
Closed
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
29 changes: 29 additions & 0 deletions clang/include/clang-c/CAS.h
Original file line number Diff line number Diff line change
Expand Up @@ -185,6 +185,35 @@ CINDEX_LINKAGE void clang_experimental_cas_loadObjectByString_async(
void (*Callback)(void *Ctx, CXCASObject, CXError),
CXCASCancellationToken *OutToken);

/**
* Stores an object in the CAS.
*
* \param Refs The refs of the object to be stored.

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think the Refs are CASID here? That is probably a bit confusing since ObjectRef is something slightly different.

I am ok with using CASID on API level as it is easier to print but it is not efficient to use this API to write large amount of connected CAS objects (while ObjectRef is very efficient).

* \param RefsCount The number of refs.
* \param Data The data of the object to be stored.
* \param DataSize The size of \Data.
* \param[out] OutError The error object to pass back to client (if any).
* If non-null the object must be disposed using \c clang_Error_dispose.
* \returns The CAS ID of the stored object, or an empty string if an
* error occurred. The ID should be disposed using
* \c clang_disposeString.
*/
CINDEX_LINKAGE CXString clang_experimental_cas_storeObject(
CXCASDatabases, const char *const *Refs, size_t RefsCount, const char *Data,
size_t DataSize, CXError *OutError);

/**
* \returns the CAS object's refs.
*/
CINDEX_LINKAGE CXStringSet *
clang_experimental_cas_CASObject_getRefs(CXCASObject);

/**
* \returns the CAS object's data, which is valid for the lifetime of the
* CXCASObject.
*/
CINDEX_LINKAGE CXString clang_experimental_cas_CASObject_getData(CXCASObject);

/**
* Dispose of a \c CXCASObject object.
*/
Expand Down
19 changes: 19 additions & 0 deletions clang/test/CAS/libclang-cas-api.c
Original file line number Diff line number Diff line change
@@ -0,0 +1,19 @@
// RUN: rm -rf %t && mkdir %t

// RUN: c-index-test core -cas-store -cas-path %t/cas -cas-object-data aGVsbG8sIHdvcmxkIQo= 1> %t/obj1.txt

// RUN: c-index-test core -cas-load -cas-path %t/cas @%t/obj1.txt 1> %t/out1.txt
// RUN: FileCheck %s -input-file %t/out1.txt -check-prefix=OBJECTONE
//
// OBJECTONE: Data:
// OBJECTONE: aGVsbG8sIHdvcmxkIQo=

// RUN: c-index-test core -cas-store -cas-path %t/cas -cas-object-data Zm9vCg== --cas-object-ref @%t/obj1.txt 1> %t/obj2.txt

// RUN: c-index-test core -cas-load -cas-path %t/cas @%t/obj2.txt 1> %t/out2.txt
// RUN: FileCheck %s -input-file %t/out2.txt -check-prefix=OBJECTTWO
//
// OBJECTTWO: Refs:
// OBJECTTWO: llvmcas://{{[a-z0-9]+}}
// OBJECTTWO: Data:
// OBJECTTWO: Zm9vCg==
86 changes: 85 additions & 1 deletion clang/tools/c-index-test/core_main.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -9,11 +9,11 @@
#include "JSONAggregation.h"
#include "indexstore/IndexStoreCXX.h"
#include "clang-c/Dependencies.h"
#include "clang/DirectoryWatcher/DirectoryWatcher.h"
#include "clang/AST/Mangle.h"
#include "clang/Basic/LangOptions.h"
#include "clang/Basic/PathRemapper.h"
#include "clang/CodeGen/ObjectFilePCHContainerOperations.h"
#include "clang/DirectoryWatcher/DirectoryWatcher.h"
#include "clang/Frontend/ASTUnit.h"
#include "clang/Frontend/CompilerInstance.h"
#include "clang/Frontend/CompilerInvocation.h"
Expand All @@ -29,6 +29,7 @@
#include "clang/Serialization/ASTReader.h"
#include "llvm/ADT/FunctionExtras.h"
#include "llvm/ADT/ScopeExit.h"
#include "llvm/Support/Base64.h"
#include "llvm/Support/CommandLine.h"
#include "llvm/Support/FileSystem.h"
#include "llvm/Support/Path.h"
Expand Down Expand Up @@ -61,6 +62,8 @@ enum class ActionType {
UploadCachedJob,
MaterializeCachedJob,
ReplayCachedJob,
CASStore,
CASLoad,
PruneCAS,
WatchDir,
};
Expand Down Expand Up @@ -92,6 +95,10 @@ Action(cl::desc("Action:"), cl::init(ActionType::None),
"Materialize cached compilation data from upstream CAS"),
clEnumValN(ActionType::ReplayCachedJob, "replay-cached-job",
"Replay a cached compilation from the CAS"),
clEnumValN(ActionType::CASStore, "cas-store",
"Store an object in the CAS"),
clEnumValN(ActionType::CASLoad, "cas-load",
"Load an object from the CAS"),
clEnumValN(ActionType::PruneCAS, "prune-cas", "Prune CAS data"),
clEnumValN(ActionType::WatchDir,
"watch-dir", "Watch directory for file events")),
Expand Down Expand Up @@ -155,6 +162,11 @@ static llvm::cl::opt<std::string>
CASPluginPath("fcas-plugin-path", llvm::cl::desc("Path for CAS plugin"));
static cl::list<std::string> CASPluginOpts("fcas-plugin-option",
cl::desc("Plugin CAS Options"));
static cl::list<std::string> CASObjectRefs("cas-object-ref",
cl::desc("CAS object ref"));
static llvm::cl::opt<std::string>
CASObjectData("cas-object-data",
llvm::cl::desc("CAS object data (base 64)"));
static llvm::cl::opt<std::string>
WorkingDir("working-dir", llvm::cl::desc("Path for working directory"));
static cl::opt<bool> TestCASCancellation(
Expand Down Expand Up @@ -1096,6 +1108,58 @@ static int replayCachedJob(ArrayRef<const char *> Args,
return 0;
}

static int casLoad(std::string CASID, CXCASDatabases DBs) {
CXError Err = nullptr;
CXCASObject Object =
clang_experimental_cas_loadObjectByString(DBs, CASID.c_str(), &Err);
if (Err) {
llvm::errs() << clang_Error_getDescription(Err) << "\n";
clang_Error_dispose(Err);
return 1;
}
CXStringSet *Refs = clang_experimental_cas_CASObject_getRefs(Object);
CXString Data = clang_experimental_cas_CASObject_getData(Object);
if (Refs->Count > 0) {
llvm::outs() << "Refs:\n";
for (unsigned i = 0; i < Refs->Count; ++i) {
llvm::outs() << clang_getCString(Refs->Strings[i]) << "\n";
}
}
llvm::outs() << "Data: \n"
<< encodeBase64(StringRef(clang_getCString(Data))) << "\n";
clang_experimental_cas_CASObject_dispose(Object);
clang_disposeStringSet(Refs);
return 0;
}

static int casStore(ArrayRef<std::string> Refs, std::string Base64Data,
CXCASDatabases DBs) {
std::vector<char *> CRefs;
for (std::string Ref : Refs) {
CRefs.push_back(strdup(Ref.c_str()));
}
std::vector<char> Data;
if (Error Err = decodeBase64(Base64Data, Data)) {
llvm::errs() << Err << "\n";
return 1;
}
CXError Err = nullptr;
CXString CASID = clang_experimental_cas_storeObject(
DBs, CRefs.data(), CRefs.size(), Data.data(), Data.size(), &Err);
for (char *CRef : CRefs) {
free(CRef);
}
if (Err) {
clang_disposeString(CASID);
llvm::errs() << clang_Error_getDescription(Err) << "\n";
clang_Error_dispose(Err);
return 1;
}
llvm::outs() << clang_getCString(CASID) << "\n";
clang_disposeString(CASID);
return 0;
}

static int pruneCAS(int64_t Limit, CXCASDatabases DBs) {
CXError Err = nullptr;
int64_t Size = clang_experimental_cas_Databases_get_storage_size(DBs, &Err);
Expand Down Expand Up @@ -1534,6 +1598,26 @@ int indextest_core_main(int argc, const char **argv) {
options::InputFiles[0], DBs);
}

if (options::Action == ActionType::CASStore) {
if (!DBs) {
errs() << "error: CAS was not configured\n";
return 1;
}
return casStore(options::CASObjectRefs, options::CASObjectData, DBs);
}

if (options::Action == ActionType::CASLoad) {
if (!DBs) {
errs() << "error: CAS was not configured\n";
return 1;
}
if (options::InputFiles.size() != 1) {
errs() << "error: expected a single CAS ID as input\n";
return 1;
}
return casLoad(options::InputFiles[0], DBs);
}

if (options::Action == ActionType::PruneCAS) {
if (options::InputFiles.empty()) {
errs() << "error: missing size limit\n";
Expand Down
45 changes: 45 additions & 0 deletions clang/tools/libclang/CCAS.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -397,6 +397,51 @@ void clang_experimental_cas_loadObjectByString_async(
WL->visit(*Ref, /*IsRootNode*/ true);
}

CXString clang_experimental_cas_storeObject(CXCASDatabases CDBs,
const char *const *IDStrings,
size_t IDStringsCount,
const char *Data, size_t DataSize,
CXError *OutError) {
WrappedCASDatabases &DBs = *unwrap(CDBs);
ObjectStore &CAS = *DBs.CAS;

SmallVector<ObjectRef, 1> Refs;
for (size_t i = 0; i < IDStringsCount; ++i) {
Expected<CASID> ID = CAS.parseID(IDStrings[i]);
if (!ID) {
*OutError = cxerror::create(ID.takeError());
return cxstring::createEmpty();
}
std::optional<ObjectRef> Ref = CAS.getReference(*ID);
if (!Ref) {
*OutError =
cxerror::create("Could not get reference for ID: " + ID->toString());
return cxstring::createEmpty();
}
Refs.push_back(*Ref);
}

Expected<ObjectRef> Ref = CAS.store(Refs, ArrayRef(Data, DataSize));
if (!Ref) {
*OutError = cxerror::create(Ref.takeError());
return cxstring::createEmpty();
}

return cxstring::createDup(CAS.getID(*Ref).toString());
}

CXStringSet *clang_experimental_cas_CASObject_getRefs(CXCASObject CObj) {
std::vector<std::string> Refs;
for (size_t i = 0; i < unwrap(CObj)->Obj.getNumReferences(); ++i) {
Refs.push_back(unwrap(CObj)->Obj.getReferenceID(i).toString());
}
return cxstring::createSet(Refs);
}

CXString clang_experimental_cas_CASObject_getData(CXCASObject CObj) {
return cxstring::createRef(unwrap(CObj)->Obj.getData());
}

void clang_experimental_cas_CASObject_dispose(CXCASObject CObj) {
delete unwrap(CObj);
}
Expand Down
3 changes: 3 additions & 0 deletions clang/tools/libclang/libclang.map
Original file line number Diff line number Diff line change
Expand Up @@ -484,6 +484,8 @@ LLVM_16 {
clang_experimental_cas_CancellationToken_cancel;
clang_experimental_cas_CancellationToken_dispose;
clang_experimental_cas_CASObject_dispose;
clang_experimental_cas_CASObject_getData;
clang_experimental_cas_CASObject_getRefs;
clang_experimental_cas_Databases_create;
clang_experimental_cas_Databases_dispose;
clang_experimental_cas_Databases_get_storage_size;
Expand All @@ -504,6 +506,7 @@ LLVM_16 {
clang_experimental_cas_replayCompilation;
clang_experimental_cas_ReplayResult_dispose;
clang_experimental_cas_ReplayResult_getStderr;
clang_experimental_cas_storeObject;
clang_experimental_DependencyScannerService_create_v1;
clang_experimental_DependencyScannerServiceOptions_create;
clang_experimental_DependencyScannerServiceOptions_dispose;
Expand Down