Fixes for validating contract upgrades

README.md

@@ -217,6 +217,14 @@ dv init --project <PROJECT_PATH> --address <ADDRESS> --contractname <NAME> --ini
 
 Please note that `<B>` must be equal to or larger than the deployment block of the contract. Additionally, it is recommended to use only block numbers of **finalized blocks** in order to prevent the DVF containing wrong data due to possible re-orgs in the future.
 
+By default, `dv` looks at all transactions starting from the deployment block. If you want to restrict the transaction lookup to a later starting block (e.g., to skip transactions before a contract upgrade), you can pass the desired block number with `--startblock`:
+
+```
+dv init --project <PROJECT_PATH> --address <ADDRESS> --contractname <NAME> --startblock <S> new.dvf.json
+```
+
+`<S>` must be equal to or larger than the deployment block and equal to or smaller than the init block.
+
 Sometimes, you might want to know when a storage variable has been initialized before but then reset back to zero. You can add such variables to the DVF with the `--zerovalue` flag:
 
 ```

lib/dvf/discovery.rs

@@ -484,16 +484,21 @@ pub fn get_tx_hashes(
     let mut seen_events: Vec<Log> = vec![];
     let tx_hashes: Vec<String> = if let Some(event_topics) = &event_topics {
         print_progress("Obtaining past events and transactions.", pc, progress_mode);
-        seen_events = web3::get_eth_events(
-            config,
-            address,
-            start_block_num,
-            end_block_num,
-            event_topics,
-        )?;
+        for topic in event_topics {
+            let mut topic_events = web3::get_eth_events_paginated(
+                config,
+                address,
+                start_block_num,
+                end_block_num,
+                &vec![*topic],
+            )?;
+            seen_events.append(&mut topic_events);
+        }
+        let mut seen_hashes = HashSet::new();
         seen_events
             .iter()
             .filter_map(|e| e.transaction_hash.map(|h| format!("{h:#x}")))
+            .filter(|h| seen_hashes.insert(h.clone()))
             .collect()
     } else {
         print_progress("Obtaining past transactions.", pc, progress_mode);
@@ -506,7 +511,7 @@ pub fn get_tx_hashes(
 pub fn create_discovery_params_for_init<'a>(
     config: &'a DVFConfig,
     dumped: &'a parse::CompleteDVF,
-    deployment_block_num: u64,
+    start_block_num: u64,
     init_block_num: u64,
     project: &'a PathBuf,
     artifacts: &'a str,
@@ -523,7 +528,7 @@ pub fn create_discovery_params_for_init<'a>(
         config,
         contract_name: &dumped.contract_name,
         address: &dumped.address,
-        start_block_num: deployment_block_num,
+        start_block_num,
         end_block_num: init_block_num,
         project: Some(project),
         artifacts,

lib/state/contract_state.rs

@@ -263,7 +263,9 @@ impl<'a> ContractState<'a> {
     fn memory_as_string(memory: &Vec<String>) -> String {
         let mut mem_string = String::new();
         for mem in memory {
-            mem_string += mem;
+            // Some RPC nodes (e.g., Anvil) return memory chunks with a "0x" prefix;
+            // strip it so the concatenated string is pure hex.
+            mem_string += mem.strip_prefix("0x").unwrap_or(mem);
         }
 
         mem_string

src/dvf.rs

@@ -408,6 +408,11 @@ fn main() {
                         .help("The block number at which the state snapshot should be taken.")
                         .value_parser(is_valid_blocknum),
                 )
+                .arg(
+                    arg!(--startblock <BLOCK>)
+                        .help("The first block to look at for transactions (defaults to deployment block).")
+                        .value_parser(is_valid_blocknum),
+                )
                 .arg(
                     arg!(--project <PATH>)
                         .help("Path to the root folder of the source code project")
@@ -763,6 +768,12 @@ fn main() {
                         .help("Folder containing the artifacts")
                         .default_value("artifacts"),
                 )
+                .arg(
+                    arg!(--libraries <LIBRARY> ...)
+                        .help("Library specifiers in the form Path:Name:Address. Accepts comma-separated values or repeated flags")
+                        .value_delimiter(',')
+                        .action(clap::ArgAction::Append),
+                )
                 .arg(arg!(--buildcache <PATH>).help("Folder containing build-info files")),
         )
         .subcommand(
@@ -932,9 +943,28 @@ fn process(matches: ArgMatches) -> Result<(), ValidationError> {
             dumped.deployment_block_num = deployment_block_num;
             dumped.deployment_tx = deployment_tx;
 
+            let start_block_num = *sub_m
+                .get_one::<u64>("startblock")
+                .unwrap_or(&deployment_block_num);
+            if start_block_num < deployment_block_num {
+                return Err(ValidationError::from(
+                    "Start block must be greater than or equal to the deployment block.",
+                ));
+            }
+
+            let default_init_block = if sub_m.contains_id("startblock") {
+                start_block_num
+            } else {
+                deployment_block_num + 1
+            };
             let init_block_num = *sub_m
                 .get_one::<u64>("initblock")
-                .unwrap_or(&(deployment_block_num + 1));
+                .unwrap_or(&default_init_block);
+            if init_block_num < start_block_num {
+                return Err(ValidationError::from(
+                    "Init block must be greater than or equal to the start block.",
+                ));
+            }
             dumped.init_block_num = init_block_num;
 
             let mut pc = 1_u64;
@@ -1030,7 +1060,7 @@ fn process(matches: ArgMatches) -> Result<(), ValidationError> {
             } = discover_storage_and_events(create_discovery_params_for_init(
                 &config,
                 &dumped,
-                deployment_block_num,
+                start_block_num,
                 init_block_num,
                 project,
                 artifacts,