From fa798afd7d2c6e0eef7984c07b7df0cb588dd6fe Mon Sep 17 00:00:00 2001
From: Kai Wetlesen <kaiw@semiotic.ai>
Date: Tue, 21 Jun 2022 22:27:27 -0700
Subject: [PATCH] Removed alphanumeric enforcement for node IDs

---
 docker/start                  | 12 ++++++++++--
 docs/environment-variables.md |  6 ++++++
 graph/src/data/store/mod.rs   | 10 ++--------
 node/src/main.rs              |  4 ++--
 4 files changed, 20 insertions(+), 12 deletions(-)

diff --git a/docker/start b/docker/start
index 86d00aac009..7800f022bc3 100755
--- a/docker/start
+++ b/docker/start
@@ -57,7 +57,7 @@ run_graph_node() {
         wait_for_ipfs "$ipfs"
         sleep 5
         graph-node \
-            --node-id "${node_id//-/_}" \
+            --node-id "$node_id" \
             --config "$GRAPH_NODE_CONFIG" \
             --ipfs "$ipfs" \
             ${fork_base:+ --fork-base "$fork_base"}
@@ -71,7 +71,7 @@ run_graph_node() {
         sleep 5
 
         graph-node \
-            --node-id "${node_id//-/_}" \
+            --node-id "$node_id" \
             --postgres-url "$postgres_url" \
             --ethereum-rpc $ethereum \
             --ipfs "$ipfs" \
@@ -98,6 +98,14 @@ start_combined_node() {
     run_graph_node
 }
 
+# Allow operators to opt out of legacy character
+# restrictions on the node ID by setting enablement
+# variable to a non-zero length string:
+if [ -z "$GRAPH_NODE_ID_USE_LITERAL_VALUE" ]
+then
+	node_id="${node_id//-/_}"
+fi
+
 if [ -n "$disable_core_dumps" ]
 then
     ulimit -c 0
diff --git a/docs/environment-variables.md b/docs/environment-variables.md
index a969ed0c024..e52e529f4b6 100644
--- a/docs/environment-variables.md
+++ b/docs/environment-variables.md
@@ -136,6 +136,12 @@ those.
   in parallel and deploy to specific nodes; each ID must be unique among the set
   of nodes. A single node should have the same value between consecutive restarts.
   Subgraphs get assigned to node IDs and are not reassigned to other nodes automatically.
+- `GRAPH_NODE_ID_USE_LITERAL_VALUE`: (Docker only) Use the literal `node_id`
+  provided to the docker start script instead of replacing hyphens (-) in names
+  with underscores (\_). Changing this for an existing `graph-node`
+  installation requires also changing the assigned node IDs in the
+  `subgraphs.subgraph_deployment_assignment` table in the database. This can be
+  done with GraphMan or via the PostgreSQL command line.
 - `GRAPH_LOG`: control log levels, the same way that `RUST_LOG` is described
   [here](https://docs.rs/env_logger/0.6.0/env_logger/)
 - `THEGRAPH_STORE_POSTGRES_DIESEL_URL`: postgres instance used when running
diff --git a/graph/src/data/store/mod.rs b/graph/src/data/store/mod.rs
index 8a03e3ce0ff..e08385ebddd 100644
--- a/graph/src/data/store/mod.rs
+++ b/graph/src/data/store/mod.rs
@@ -63,14 +63,8 @@ impl NodeId {
     pub fn new(s: impl Into<String>) -> Result<Self, ()> {
         let s = s.into();
 
-        // Enforce length limit
-        if s.len() > 63 {
-            return Err(());
-        }
-
-        // Check that the ID contains only allowed characters.
-        // Note: these restrictions are relied upon to prevent SQL injection
-        if !s.chars().all(|c| c.is_ascii_alphanumeric() || c == '_') {
+        // Enforce minimum and maximum length limit
+        if s.len() > 63 || s.len() < 1 {
             return Err(());
         }
 
diff --git a/node/src/main.rs b/node/src/main.rs
index 656339b72e4..098dd1a4e36 100644
--- a/node/src/main.rs
+++ b/node/src/main.rs
@@ -133,8 +133,8 @@ async fn main() {
         std::process::exit(0);
     }
 
-    let node_id =
-        NodeId::new(opt.node_id.clone()).expect("Node ID must contain only a-z, A-Z, 0-9, and '_'");
+    let node_id = NodeId::new(opt.node_id.clone())
+        .expect("Node ID must be between 1 and 63 characters in length");
     let query_only = config.query_only(&node_id);
 
     // Obtain subgraph related command-line arguments