use matcher over regex (#744)

* replace http router AddPattern with AddMatcher
* fix scrub logging

7 files changed, 59 insertions, 18 deletions

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 187c35ddf..46d40b6d5 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -8,6 +8,7 @@
 - Improvement: Added `--maxtimeslice` option to `zen scrub` command to control how long a scrub operation may run
 - Improvement: Increased the default scrub timeslice from 1 min 40 sec to 5 min.
 - Improvement: Reduce lock contention when performing a scrub operation
+- Improvement: Replaced http routing regex matching with faster matching lambdas for admin, buildstore, projectstore, objectstore and workspaces services
 - Bugfix: Restore `/health/log` and `/health/info` endpoint functionality
 - Bugfix: Fixed 32-bit truncation of transmission chunk sizes when using the asio http path
 
diff --git a/src/zenserver/storage/admin/admin.cpp b/src/zenserver/storage/admin/admin.cpp
index cd6002d59..19155e02b 100644
--- a/src/zenserver/storage/admin/admin.cpp
+++ b/src/zenserver/storage/admin/admin.cpp
@@ -121,7 +121,10 @@ HttpAdminService::HttpAdminService(GcScheduler&			   Scheduler,
 		},
 		HttpVerb::kGet);
 
-	m_Router.AddPattern("jobid", "([[:digit:]]+?)");
+	static constexpr AsciiSet ValidNumberCharactersSet{"0123456789"};
+
+	m_Router.AddMatcher("jobid",
+						[](std::string_view Str) -> bool { return !Str.empty() && AsciiSet::HasOnly(Str, ValidNumberCharactersSet); });
 
 	m_Router.RegisterRoute(
 		"jobs",
diff --git a/src/zenserver/storage/buildstore/httpbuildstore.cpp b/src/zenserver/storage/buildstore/httpbuildstore.cpp
index 18fae7027..f5ba30616 100644
--- a/src/zenserver/storage/buildstore/httpbuildstore.cpp
+++ b/src/zenserver/storage/buildstore/httpbuildstore.cpp
@@ -48,10 +48,20 @@ HttpBuildStoreService::Initialize()
 {
 	ZEN_LOG_INFO(LogBuilds, "Initializing Builds Service");
 
-	m_Router.AddPattern("namespace", "([[:alnum:]\\-_.]+)");
-	m_Router.AddPattern("bucket", "([[:alnum:]\\-_.]+)");
-	m_Router.AddPattern("buildid", "([[:xdigit:]]{24})");
-	m_Router.AddPattern("hash", "([[:xdigit:]]{40})");
+	static constexpr AsciiSet ValidNamespaceCharactersSet{"abcdefghijklmnopqrstuvwxyz0123456789-_.ABCDEFGHIJKLMNOPQRSTUVWXYZ"};
+	static constexpr AsciiSet ValidBucketCharactersSet{"abcdefghijklmnopqrstuvwxyz0123456789-_.ABCDEFGHIJKLMNOPQRSTUVWXYZ"};
+	static constexpr AsciiSet ValidHexCharactersSet{"0123456789abcdefABCDEF"};
+
+	m_Router.AddMatcher("namespace",
+						[](std::string_view Str) -> bool { return !Str.empty() && AsciiSet::HasOnly(Str, ValidNamespaceCharactersSet); });
+	m_Router.AddMatcher("bucket",
+						[](std::string_view Str) -> bool { return !Str.empty() && AsciiSet::HasOnly(Str, ValidBucketCharactersSet); });
+	m_Router.AddMatcher("buildid", [](std::string_view Str) -> bool {
+		return Str.length() == Oid::StringLength && AsciiSet::HasOnly(Str, ValidHexCharactersSet);
+	});
+	m_Router.AddMatcher("hash", [](std::string_view Str) -> bool {
+		return Str.length() == IoHash::StringLength && AsciiSet::HasOnly(Str, ValidHexCharactersSet);
+	});
 
 	m_Router.RegisterRoute(
 		"{namespace}/{bucket}/{buildid}/blobs/{hash}",
diff --git a/src/zenserver/storage/objectstore/objectstore.cpp b/src/zenserver/storage/objectstore/objectstore.cpp
index d8ad40621..052c3d630 100644
--- a/src/zenserver/storage/objectstore/objectstore.cpp
+++ b/src/zenserver/storage/objectstore/objectstore.cpp
@@ -271,8 +271,13 @@ HttpObjectStoreService::Inititalize()
 		CreateDirectories(BucketsPath);
 	}
 
-	m_Router.AddPattern("path", "([[:alnum:]/_.,;$~\\{\\}\\+\\-\\[\\]\\%\\(\\)]+)");
-	m_Router.AddPattern("bucket", "([[:alnum:]\\-_.]+)");
+	static constexpr AsciiSet ValidPathCharactersSet{"abcdefghijklmnopqrstuvwxyz0123456789/_.,;$~{}+-[]%()]ABCDEFGHIJKLMNOPQRSTUVWXYZ"};
+	static constexpr AsciiSet ValidBucketCharactersSet{"abcdefghijklmnopqrstuvwxyz0123456789-_.ABCDEFGHIJKLMNOPQRSTUVWXYZ"};
+
+	m_Router.AddMatcher("path",
+						[](std::string_view Str) -> bool { return !Str.empty() && AsciiSet::HasOnly(Str, ValidPathCharactersSet); });
+	m_Router.AddMatcher("bucket",
+						[](std::string_view Str) -> bool { return !Str.empty() && AsciiSet::HasOnly(Str, ValidBucketCharactersSet); });
 
 	m_Router.RegisterRoute(
 		"bucket",
diff --git a/src/zenserver/storage/projectstore/httpprojectstore.cpp b/src/zenserver/storage/projectstore/httpprojectstore.cpp
index 86b4d7100..416e2ed69 100644
--- a/src/zenserver/storage/projectstore/httpprojectstore.cpp
+++ b/src/zenserver/storage/projectstore/httpprojectstore.cpp
@@ -549,11 +549,23 @@ HttpProjectService::HttpProjectService(CidStore&		  Store,
 
 	using namespace std::literals;
 
-	m_Router.AddPattern("project", "([[:alnum:]_.]+)");
-	m_Router.AddPattern("log", "([[:alnum:]_.]+)");
-	m_Router.AddPattern("op", "([[:digit:]]+?)");
-	m_Router.AddPattern("chunk", "([[:xdigit:]]{24})");
-	m_Router.AddPattern("hash", "([[:xdigit:]]{40})");
+	static constexpr AsciiSet ValidProjectCharactersSet{"abcdefghijklmnopqrstuvwxyz0123456789_.ABCDEFGHIJKLMNOPQRSTUVWXYZ"};
+	static constexpr AsciiSet ValidOplogCharactersSet{"abcdefghijklmnopqrstuvwxyz0123456789_.ABCDEFGHIJKLMNOPQRSTUVWXYZ"};
+	static constexpr AsciiSet ValidNumberCharactersSet{"0123456789"};
+	static constexpr AsciiSet ValidHexCharactersSet{"0123456789abcdefABCDEF"};
+
+	m_Router.AddMatcher("project",
+						[](std::string_view Str) -> bool { return !Str.empty() && AsciiSet::HasOnly(Str, ValidProjectCharactersSet); });
+	m_Router.AddMatcher("log",
+						[](std::string_view Str) -> bool { return !Str.empty() && AsciiSet::HasOnly(Str, ValidOplogCharactersSet); });
+	m_Router.AddMatcher("op",
+						[](std::string_view Str) -> bool { return !Str.empty() && AsciiSet::HasOnly(Str, ValidNumberCharactersSet); });
+	m_Router.AddMatcher("chunk", [](std::string_view Str) -> bool {
+		return Str.length() == Oid::StringLength && AsciiSet::HasOnly(Str, ValidHexCharactersSet);
+	});
+	m_Router.AddMatcher("hash", [](std::string_view Str) -> bool {
+		return Str.length() == IoHash::StringLength && AsciiSet::HasOnly(Str, ValidHexCharactersSet);
+	});
 
 	m_Router.RegisterRoute(
 		"",
diff --git a/src/zenserver/storage/workspaces/httpworkspaces.cpp b/src/zenserver/storage/workspaces/httpworkspaces.cpp
index 3fea46b2f..dc4cc7e69 100644
--- a/src/zenserver/storage/workspaces/httpworkspaces.cpp
+++ b/src/zenserver/storage/workspaces/httpworkspaces.cpp
@@ -169,10 +169,20 @@ HttpWorkspacesService::Initialize()
 
 	ZEN_LOG_INFO(LogFs, "Initializing Workspaces Service");
 
-	m_Router.AddPattern("workspace_id", "([[:xdigit:]]{24})");
-	m_Router.AddPattern("share_id", "([[:xdigit:]]{24})");
-	m_Router.AddPattern("chunk", "([[:xdigit:]]{24})");
-	m_Router.AddPattern("share_alias", "([[:alnum:]_.\\+\\-\\[\\]]+)");
+	static constexpr AsciiSet ValidHexCharactersSet{"0123456789abcdefABCDEF"};
+
+	m_Router.AddMatcher("workspace_id", [](std::string_view Str) -> bool {
+		return Str.length() == Oid::StringLength && AsciiSet::HasOnly(Str, ValidHexCharactersSet);
+	});
+	m_Router.AddMatcher("share_id", [](std::string_view Str) -> bool {
+		return Str.length() == Oid::StringLength && AsciiSet::HasOnly(Str, ValidHexCharactersSet);
+	});
+	m_Router.AddMatcher("chunk", [](std::string_view Str) -> bool {
+		return Str.length() == Oid::StringLength && AsciiSet::HasOnly(Str, ValidHexCharactersSet);
+	});
+	m_Router.AddMatcher("share_alias", [](std::string_view Str) -> bool {
+		return !Str.empty() && AsciiSet::HasOnly(Str, Workspaces::ValidAliasCharactersSet);
+	});
 
 	m_Router.RegisterRoute(
 		"{workspace_id}/{share_id}/files",
diff --git a/src/zenstore/compactcas.cpp b/src/zenstore/compactcas.cpp
index 097102a1d..5d8f95c9e 100644
--- a/src/zenstore/compactcas.cpp
+++ b/src/zenstore/compactcas.cpp
@@ -547,11 +547,11 @@ CasContainerStrategy::ScrubStorage(ScrubContext& Ctx)
 
 	if (Ctx.IsSkipCas())
 	{
-		ZEN_INFO("SKIPPED scrubbing: '{}'", m_BlocksBasePath);
+		ZEN_INFO("SKIPPED scrubbing: '{}'", m_RootDirectory);
 		return;
 	}
 
-	ZEN_INFO("scrubbing '{}'", m_BlocksBasePath);
+	ZEN_INFO("scrubbing '{}'", m_RootDirectory);
 
 	RwLock							BadKeysLock;
 	std::vector<IoHash>				BadKeys;