feat: build merkl airdrop csv based on user pool shares

tools/python/gen_morpho_airdrop.py

@@ -0,0 +1,99 @@
+import json
+
+import numpy as np
+import pandas as pd
+
+from bal_tools import Subgraph
+
+
+SUBGRAPH = Subgraph()
+MORPHO = "0x58D97B57BB95320F9a05dC918Aef65434969c2B2"
+
+
+def get_user_shares(pool, block):
+    query = """query PoolShares($where: PoolShare_filter, $block: Block_height) {
+        poolShares(where: $where, block: $block) {
+            user {
+                id
+            }
+            balance
+        }
+    }"""
+    params = {
+        "where": {
+            "balance_gt": 0.001,
+            "pool": pool,
+        },
+        "block": {"number": block},
+    }
+    raw = SUBGRAPH.fetch_graphql_data(
+        "subgraphs-v3",
+        query,
+        params,
+        url="https://api.studio.thegraph.com/query/75376/balancer-v3/version/latest",
+    )
+    return dict([(x["user"]["id"], x["balance"]) for x in raw["poolShares"]])
+
+
+def get_block_from_timestamp(ts):
+    query = """query GetBlockFromTimestamp($where: Block_filter) {
+        blocks(orderBy: "number", orderDirection: "desc", where: $where) {
+            number
+            timestamp
+        }
+    }"""
+    params = {"where": {"timestamp_lte": ts}}
+    raw = SUBGRAPH.fetch_graphql_data(
+        "blocks",
+        query,
+        params,
+        url="https://api.studio.thegraph.com/query/48427/ethereum-blocks/version/latest",
+    )
+    return int(raw["blocks"][0]["number"])
+
+
+def build_snapshot_df(
+    pool,  # pool address
+    end,  # timestamp of the last snapshot
+    n=7,  # amount of snapshots
+    step_size=60 * 60 * 24,  # amount of seconds between snapshots
+):
+    shares = {}
+    for _ in range(n):
+        block = get_block_from_timestamp(end)
+        shares[block] = get_user_shares(pool=pool, block=block)
+        end -= step_size
+    return pd.DataFrame(shares, dtype=float).fillna(0)
+
+
+def consolidate_shares(df):
+    consolidated = pd.DataFrame()
+    for block in df.columns:
+        # calculate the percentage of the pool each user owns
+        consolidated[block] = df[block] / df[block].sum()
+        # weigh it by the total pool size of that block
+        consolidated[block] *= df.sum()[block]
+    # sum the weighted percentages per user
+    consolidated["total"] = consolidated.sum(axis=1)
+    # divide the weighted percentages by the sum of all weights
+    consolidated["total"] = consolidated["total"] / df.sum().sum()
+    return consolidated
+
+
+def build_airdrop(reward_token, reward_total_wei, df):
+    # https://docs.merkl.xyz/merkl-mechanisms/types-of-campaign/airdrop
+    df["wei"] = df["total"] * reward_total_wei
+    df["wei"] = df["wei"].apply(np.floor).astype(int).astype(str)
+    return {"rewardToken": reward_token, "rewards": df[["wei"]].to_dict(orient="index")}
+
+
+if __name__ == "__main__":
+    # get bpt balances for a pool at different timestamps
+    df = build_snapshot_df(
+        pool="0x89bb794097234e5e930446c0cec0ea66b35d7570", end=1734393600
+    )
+    # consolidate user pool shares
+    df = consolidate_shares(df)
+    # build airdrop object and dump to json file
+    airdrop = build_airdrop(reward_token=MORPHO, reward_total_wei=1e18, df=df)
+    json.dump(airdrop, open("airdrop.json", "w"), indent=2)

tools/python/requirements.txt

@@ -5,7 +5,7 @@ pandas
 tabulate
 requests
 web3
-git+https://github.com/BalancerMaxis/bal_addresses@main
+git+https://github.com/BalancerMaxis/bal_addresses@dev/use-bal_tools-dev-branch
 dune-client
 pytest
 dataclasses-json