Bridges of LI.FI across chains, the contract is LiFiDiamond_v2

[lequangphu]

Thank you for contributing to Spellbook 🪄

Please open the PR in draft and mark as ready when you want to request a review.

Description:

Extract bridge data from event LiFiDiamond_v2_evt_LiFiTransferStarted on each chain, then union across chains.

---

quick links for more information:

• README.md
• spellbook docs
• CONTRIBUTING.md

[jeff-dude]

before we dig into the failing test, i think we need to redesign this pipeline a bit.

• flip your materialization strategy approach -- i.e. materialize each chain incrementally, then have the crosschain be a view on top
• looks like the code is the same per chain -- in this case, i'd use a dbt macro for the logic and call the macro for each chain while passing in the blockchain parameter to differentiate. there are tons of examples in spellbook to use as reference, one to search could be the uniswap models in dex subproject (but plenty others too!)
• when joining from decoded to base table to get additional columns, lets use this shared macro. an example to call the macro is here
• when using the if is incremental logic, let's use the latest macro approach like here rather than hardcode last 7 days
• for incremental chain models, i would expect config to contain all of this
• looks like you removed the surrogate key logic, i'd put that back in and use as unique key. then update the schema yml to run unique test on this column, not the combination of columns that make up the surrogate key

let's leave it at that for now, and revisit once all this in place 🙏

[lequangphu]

thanks, i will update the pr based on your recommendation.

great, just tag me again when you are ready for next iteration of review 🙏

it's green check, so i guess i'm ready.

[jeff-dude]

really clean code here, much appreciated!

left a few minor final changes below, please do implement when ready.

last thing: have you queried the data output yet, from CI pipeline?
https://github.com/duneanalytics/spellbook/actions/runs/12054789316/job/33613636277?pr=7172#step:12:32
you can query these tables on dune and ensure it looks as you expect, prior to merge 🙏

[lequangphu]

really clean code here, much appreciated!

left a few minor final changes below, please do implement when ready.

last thing: have you queried the data output yet, from CI pipeline? https://github.com/duneanalytics/spellbook/actions/runs/12054789316/job/33613636277?pr=7172#step:12:32 you can query these tables on dune and ensure it looks as you expect, prior to merge 🙏

I applied your suggestion.

I checked the data output of CI pipeline, they were as expected.

I want to add amount_usd to the output as what I do from line 223-427 in this query: https://dune.com/queries/4182530. The steps are:

1. convert address of eth to weth, variants of usdc/usdt to usdc/usdt (without this step, there are many null records of amount_usd)
2. join prices.usd to get amount_usd + add chain name based on chain id

How would you apply them in the models?

[jeff-dude]

How would you apply them in the models?

for pricing, i would handle the way you are now. CTE which reads raw decoded data and pass into macro, then pass that CTE into new one with join to prices. i was thinking you could use a macro for the join to prices and have each model use it, but if your manual mapping of token addresses differs per chain, it may be more complex.

to summarize:

• first CTE reads only decoded data ✅
• second CTE (todo) maps addresses as needed per chain
• third CTE joins to prices ✅
• final macro call to add columns

[lequangphu]

• first CTE reads only decoded data ✅
• second CTE (todo) maps addresses as needed per chain
• third CTE joins to prices ✅
• final macro call to add columns

Done, please review!

[lequangphu]

@jeff-dude the test fails because decoded table notfound, which is weird.

[jeff-dude]

this is looking clean now 🔥
thank you for the patience and iterating on changes, i'll get it prepped for merge!