Add LLM Support #160
Add LLM Support #160
Conversation
|
Caution Review failedThe pull request is closed. WalkthroughThe changes introduce a command-line interface (CLI) for querying a graph database focused on supply chain security artifacts. The new CLI, encapsulated in Changes
Possibly related PRs
Suggested reviewers
Poem
📜 Recent review detailsConfiguration used: CodeRabbit UI ⛔ Files ignored due to path filters (1)
📒 Files selected for processing (5)
Thank you for using CodeRabbit. We offer it for free to the OSS community and would appreciate your support in helping us grow. If you find it useful, would you consider giving us a shout-out on your favorite social media? 🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments. CodeRabbit Commands (Invoked using PR comments)
Other keywords and placeholders
CodeRabbit Configuration File (
|
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 3
🧹 Outside diff range and nitpick comments (9)
cmd/server/server.go (2)
36-37: Ensure consistent field naming in theoptionsstructThe field
UseOpenAILLMhas inconsistent capitalization. For readability and to follow Go naming conventions, consider renaming it toUseOpenAiLLMorUseOpenAILlm.Apply this diff:
type options struct { // ... - UseOpenAILLM bool + UseOpenAiLLM bool VectorDBPath string }Don't forget to update all references to this field throughout the codebase.
60-61: Align flag names with field names for clarityIn the
AddFlagsmethod, ensure that the flag names accurately reflect the field names. Since the field is namedUseOpenAiLLM, consider renaming the flag to--use-openai-llmfor consistency.Apply this diff:
cmd.Flags().BoolVar(&o.UseOpenAILLM, "use-openai-llm", false, "Use OpenAI LLM for graph analysis")Adjust the variable name if you rename the field as suggested.
pkg/storages/sql.go (1)
282-309: Optimize batch processing logic inSaveCachesmethodThe current implementation of batching in
SaveCachesis functional but can be optimized for clarity and efficiency. Consider the following improvements:
- Use
min(i + batchSize, len(caches))to calculate the end index without an explicitifstatement.- Preallocate the
kvCachesslice withbatchSizeto improve memory allocation.Apply this diff:
const batchSize = 500 // Safe batch size considering SQLite's limits // Process caches in batches for i := 0; i < len(caches); i += batchSize { - end := i + batchSize - if end > len(caches) { - end = len(caches) - } + end := int(math.Min(float64(i+batchSize), float64(len(caches)))) batch := caches[i:end] kvCaches := make([]KVStore, 0, len(batch)) for _, cache := range batch { cacheKey := fmt.Sprintf("%s%d", CacheKeyPrefix, cache.ID) data, err := cache.MarshalJSON() if err != nil { return fmt.Errorf("failed to marshal cache: %w", err) } kvCaches = append(kvCaches, KVStore{ Key: cacheKey, Value: string(data), }) } if err := s.DB.Save(&kvCaches).Error; err != nil { return fmt.Errorf("failed to save caches batch: %w", err) } }cmd/llm/llm.go (4)
61-61: Remove unused fieldsaveQueryThe field
saveQueryis declared but not used anywhere in the code. To keep the codebase clean, consider removing it.Apply this diff:
type options struct { maxOutput int showInfo bool - saveQuery string addr string // ... }🧰 Tools
🪛 golangci-lint (1.62.2)
61-61: field
saveQueryis unused(unused)
82-84: Handle missingOPENAI_API_KEYenvironment variable gracefullyInstead of returning an error when the
OPENAI_API_KEYis not set, consider providing a more user-friendly message or guidance on how to set it.Apply this diff:
if os.Getenv("OPENAI_API_KEY") == "" { - return fmt.Errorf("OPENAI_API_KEY environment variable is not set") + fmt.Println("Please set the OPENAI_API_KEY environment variable to use this feature.") + return nil }
139-143: Trim and validate user input before processingAdd input validation to handle empty inputs or potential unintended commands.
Apply this diff:
input = strings.TrimSpace(input) + if input == "" { + fmt.Println("Input cannot be empty. Please enter a valid query.") + continue + } if strings.ToLower(input) == "exit" { fmt.Println("Ending chat session. Goodbye!") return nil }
278-285: Refine assistant's feedback to the userThe assistant's feedback message could be more conversational and helpful by directly incorporating the query result without a repetitive prompt.
Apply this diff:
feedbackMsg := fmt.Sprintf("%s", queryResult) messages = append(messages, openai.ChatCompletionMessage{ Role: openai.ChatMessageRoleAssistant, Content: feedbackMsg, })cmd/root/root.go (1)
58-58: Maintain alphabetical order of commands inrootCmdFor better readability and organization, consider adding the
llmcommand in alphabetical order within the list of commands.Apply this diff:
rootCmd.AddCommand(cache.New()) + rootCmd.AddCommand(llm.New()) rootCmd.AddCommand(leaderboard.New()) rootCmd.AddCommand(query.New()) rootCmd.AddCommand(server.New())go.mod (1)
52-54: Group new dependencies with existing onesFor clarity, place the new direct dependencies (
chromem-goandgo-openai) within the existingrequireblock, keeping indirect dependencies separate.Apply this diff:
require ( // Existing dependencies... + github.com/philippgille/chromem-go v0.7.0 + github.com/sashabaranov/go-openai v1.36.0 ) require ( // Indirect dependencies... )
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
⛔ Files ignored due to path filters (1)
go.sumis excluded by!**/*.sum
📒 Files selected for processing (5)
cmd/llm/llm.go(1 hunks)cmd/root/root.go(2 hunks)cmd/server/server.go(4 hunks)go.mod(1 hunks)pkg/storages/sql.go(1 hunks)
🧰 Additional context used
🪛 golangci-lint (1.62.2)
cmd/llm/llm.go
61-61: field saveQuery is unused
(unused)
- Added an OpenAI llm to the CLI, that can run leaderboards, queries, and globsearchs all from user input, this is also a chat session, so the llm, can create queries based on the conversation - Batched SQLite so that it does not overflow on large amounts of data Signed-off-by: neilnaveen <[email protected]>
Signed-off-by: naveensrinivasan <[email protected]>
naveensrinivasan
force-pushed
the
neil/addLLmSupport
branch
from
9d87ed8 to
c8f8522
Compare
Codecov ReportAttention: Patch coverage is
|
Summary by CodeRabbit
New Features
Bug Fixes
Chores