ton/GeneralUtils
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 4 Wiki Activity

update

Browse Source
This commit is contained in:
narawat
2026-07-14 18:13:31 +07:00
parent a15630619a
commit 95954249ce
1 changed files with 6 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/llmUtil.jl
+6 -6
View File
@@ -1,7 +1,7 @@
module llmUtil
export formatLLMtext, extractthink, checkAgentResponse_JSON, clean_json_response,
extract_vector_metadata, generate_embedding_payloads, resolve_semantic_cluster,
extract_column_metadata, generate_embedding_payloads, resolve_semantic_cluster,
harvest_entity_catalog, resolve_entity
using UUIDs, JSON, Dates, DataFrames, StringDistances, Graphs, LibPQ
@@ -246,7 +246,7 @@ Returns a DataFrame designed for vector embedding generation.
```julia
julia> using GeneralUtils
julia> pg_conn = "host=localhost port=5432 dbname=winedb user=admin password=secret"
julia> df = GeneralUtils.extract_vector_metadata(pg_conn)
julia> df = GeneralUtils.extract_column_metadata(pg_conn)
DataFrame
6 rows × 7 columns
table_name column_name data_type column_description is_primary_key is_foreign_key vector_id
@@ -259,7 +259,7 @@ products price numeric Product price false false
products user_id integer Reference to user false true col_products_user_id
```
"""
function extract_vector_metadata(pg_conn_str::String)::DataFrame
function extract_column_metadata(pg_conn_str::String)::DataFrame
conn = LibPQ.Connection(pg_conn_str)
# This direct SQL query pulls the column specifications along with column-level descriptions
@@ -309,12 +309,12 @@ end
""" Generate embedding payloads from vector metadata DataFrame.
Transforms the metadata DataFrame from `extract_vector_metadata` into a vector of
Transforms the metadata DataFrame from `extract_column_metadata` into a vector of
dictionaries structured for vector embedding storage and retrieval.
# Arguments
- `df::DataFrame`
A DataFrame with columns from `extract_vector_metadata`: `table_name`, `column_name`,
A DataFrame with columns from `extract_column_metadata`: `table_name`, `column_name`,
`data_type`, `column_description`, `is_primary_key`, `is_foreign_key`, `vector_id`.
# Return
@@ -335,7 +335,7 @@ The function constructs rich text payloads by:
```julia
julia> using GeneralUtils
julia> pg_conn = "host=localhost port=5432 dbname=winedb user=admin password=secret"
julia> df = GeneralUtils.extract_vector_metadata(pg_conn)
julia> df = GeneralUtils.extract_column_metadata(pg_conn)
julia> payloads = GeneralUtils.generate_embedding_payloads(df)
3-element Vector{Dict}:
Dict("id" => "col_users_id", "text_content" => "Table: users | Column: id | Type: integer [PRIMARY KEY] | Description: User ID", "metadata" => Dict("table" => "users", "column" => "id", "type" => "integer"))