[webpack] debugId on main entry file keeps changing due to other chunk

[vincenthend]

Environment

@sentry/webpack-plugin@2.7.0

Steps to Reproduce

• I have a multiple entry files (main.js, other_entry.js) and few other chunks (foo.js and bar.js). foo.js is uniquely generated per build.
  • other_entry.js is using foo.js
  • main.js is using bar.js
• Each chunk in the build will be generated with the name [id].[contentHash].js for caching purposes
• Since foo.js keeps changing on every build, thedebugId generated for main.js is affected
• The debugId for main.js changed changing the final contentHash and breaking the cache behaviour
• The debugId for bar.js remains constant

Apparently this happened due to the use of hash for generating debugId

        new BannerPlugin({
          raw: true,
          include: /\.(js|ts|jsx|tsx|mjs|cjs)$/,
          banner: (arg?: BannerPluginCallbackArg) => {
            const debugId = arg?.chunk?.hash ? stringToUUID(arg.chunk.hash) : uuidv4();
            return getDebugIdSnippet(debugId);
          },
        })

Expected Result

• debugID on main.js should remain constant since the real file content doesn't change

Actual Result

• debugID for main.js changed although the real content of the file doesn't change.

Suggestion

Probably good if we can have it generated based on contentHash instead of hash

[lforst]

Since foo.js keeps changing on every build, thedebugId generated for main.js is affected

Can you explain why changing foo.js affects the hash of main.js?

But in general you're probably right that we should use the content hash instead.

[vincenthend]

Can you explain why changing foo.js affects the hash of main.js?

Hmm, if I'm debugging it correctly, the cause for my case is due to chunk.hash for entry files seems to be generated using the child compilation