Merged
Conversation
Contributor
There was a problem hiding this comment.
Pull Request Overview
This PR adds priority-based truncation support to the POML writer system, enabling intelligent content reduction when concatenating markdown boxes based on character and token limits.
- Adds truncation capabilities with configurable direction, markers, and token encoding models
- Implements priority-based box removal when content exceeds limits
- Extends both MarkdownWriter and FreeWriter to support char-limit, token-limit, and priority attributes
Reviewed Changes
Copilot reviewed 4 out of 4 changed files in this pull request and generated 4 comments.
| File | Description |
|---|---|
| packages/poml/writer.ts | Core implementation of truncation logic, priority-based reduction, and tokenizer integration |
| packages/poml/tests/writer.test.tsx | Unit tests covering character limits, token limits, and priority-based truncation scenarios |
| packages/poml/tests/index.test.tsx | End-to-end tests validating truncation features through the full POML pipeline |
| packages/poml/base.tsx | Interface additions for charLimit, tokenLimit, and priority props |
Comments suppressed due to low confidence (2)
packages/poml/tests/index.test.tsx:25
- The attribute name 'charLimit' in the test uses camelCase, but the implementation expects 'char-limit' (kebab-case). This inconsistency could cause confusion.
const text = '<p charLimit="4">abcdefg</p>';
packages/poml/tests/index.test.tsx:31
- Similar to charLimit, 'tokenLimit' uses camelCase while the implementation expects 'token-limit' (kebab-case).
const text = '<p tokenLimit="1">hello world</p>';
| if (tokenLimit !== undefined) { | ||
| let enc = this.tokenizerCache[tokenEncodingModel]; | ||
| if (!enc) { | ||
| enc = encodingForModel(tokenEncodingModel as any); |
There was a problem hiding this comment.
The 'as any' type assertion bypasses TypeScript's type checking for the tokenEncodingModel parameter. Consider validating the model name against a known list of supported models or using proper typing.
Suggested change
| enc = encodingForModel(tokenEncodingModel as any); | |
| if (!Object.values(SupportedModels).includes(tokenEncodingModel)) { | |
| throw new Error(`Unsupported token encoding model: ${tokenEncodingModel}`); | |
| } | |
| enc = encodingForModel(tokenEncodingModel); |
Comment on lines
+769
to
+776
| const tokenModel = (this.options as any).tokenEncodingModel || 'gpt-3.5-turbo'; | ||
| const getTokenLength = (t: string) => { | ||
| if (tokenLimit === undefined) { | ||
| return 0; | ||
| } | ||
| let enc = this.tokenizerCache[tokenModel]; | ||
| if (!enc) { | ||
| enc = encodingForModel(tokenModel as any); |
There was a problem hiding this comment.
Similar to line 117, this 'as any' type assertion bypasses type safety. The same validation approach should be applied here.
Suggested change
| const tokenModel = (this.options as any).tokenEncodingModel || 'gpt-3.5-turbo'; | |
| const getTokenLength = (t: string) => { | |
| if (tokenLimit === undefined) { | |
| return 0; | |
| } | |
| let enc = this.tokenizerCache[tokenModel]; | |
| if (!enc) { | |
| enc = encodingForModel(tokenModel as any); | |
| const validModels = ['gpt-3.5-turbo', 'gpt-4', 'gpt-3'] as const; | |
| type TokenModel = typeof validModels[number]; | |
| const tokenModel = validModels.includes((this.options?.tokenEncodingModel as string) ?? '') | |
| ? (this.options.tokenEncodingModel as TokenModel) | |
| : 'gpt-3.5-turbo'; | |
| const getTokenLength = (t: string) => { | |
| if (tokenLimit === undefined) { | |
| return 0; | |
| } | |
| let enc = this.tokenizerCache[tokenModel]; | |
| if (!enc) { | |
| enc = encodingForModel(tokenModel); |
Comment on lines
+793
to
+794
| const minP = Math.min(...priorities); | ||
| if (current.every(b => (b.priority ?? 0) === minP)) { |
There was a problem hiding this comment.
This condition will break the loop when all remaining boxes have the same priority, but this means content may still exceed limits. Consider adding a comment explaining this behavior or implementing a fallback strategy.
Suggested change
| const minP = Math.min(...priorities); | |
| if (current.every(b => (b.priority ?? 0) === minP)) { | |
| const minP = Math.min(...priorities); | |
| // If all remaining boxes have the same priority, break the loop. | |
| // However, this may result in content exceeding the limits. | |
| if (current.every(b => (b.priority ?? 0) === minP)) { | |
| // Fallback strategy: Remove boxes from the end until limits are met. | |
| while ( | |
| (charLimit !== undefined && totalChars(current) > charLimit) || | |
| (tokenLimit !== undefined && totalTokens(current) > tokenLimit) | |
| ) { | |
| current.pop(); | |
| } |
Comment on lines
+103
to
+111
| if (truncateDirection === 'start') { | ||
| truncated = truncated.slice(truncated.length - charLimit); | ||
| } else if (truncateDirection === 'middle') { | ||
| const head = Math.ceil(charLimit / 2); | ||
| const tail = charLimit - head; | ||
| truncated = truncated.slice(0, head) + truncated.slice(truncated.length - tail); | ||
| } else { | ||
| truncated = truncated.slice(0, charLimit); | ||
| } |
There was a problem hiding this comment.
The character limit truncation logic is duplicated with slight variations in the token limit section (lines 123-131). Consider extracting this logic into a helper function to reduce code duplication.
Suggested change
| if (truncateDirection === 'start') { | |
| truncated = truncated.slice(truncated.length - charLimit); | |
| } else if (truncateDirection === 'middle') { | |
| const head = Math.ceil(charLimit / 2); | |
| const tail = charLimit - head; | |
| truncated = truncated.slice(0, head) + truncated.slice(truncated.length - tail); | |
| } else { | |
| truncated = truncated.slice(0, charLimit); | |
| } | |
| truncated = this.truncateByDirection( | |
| truncated, | |
| charLimit, | |
| truncateDirection, | |
| (str, start, end) => str.slice(start, end) | |
| ); |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Testing
npm run build-webviewnpm run build-clinpm run lintnpm testpython -m pytest python/testshttps://chatgpt.com/codex/tasks/task_e_68885f74f82c832eabbfa758276b2fa5