From c6365ca8ddb5fc5ea280f998875c075de662d722 Mon Sep 17 00:00:00 2001
From: Anthony Hu <anthu@microsoft.com>
Date: Sun, 10 Sep 2023 23:39:34 +0800
Subject: [PATCH] add a flow to check repo's grammar and spelling

---
 .../standard/grammar-check/flow.dag.yaml      | 39 +++++++++++++++++++
 .../standard/grammar-check/generate_data.py   | 32 +++++++++++++++
 .../grammar-check/grammar_check.jinja2        | 18 +++++++++
 .../grammar-check/input_passthrough.jinja2    |  1 +
 .../flows/standard/grammar-check/read_file.py | 19 +++++++++
 5 files changed, 109 insertions(+)
 create mode 100644 examples/flows/standard/grammar-check/flow.dag.yaml
 create mode 100644 examples/flows/standard/grammar-check/generate_data.py
 create mode 100644 examples/flows/standard/grammar-check/grammar_check.jinja2
 create mode 100644 examples/flows/standard/grammar-check/input_passthrough.jinja2
 create mode 100644 examples/flows/standard/grammar-check/read_file.py

diff --git a/examples/flows/standard/grammar-check/flow.dag.yaml b/examples/flows/standard/grammar-check/flow.dag.yaml
new file mode 100644
index 00000000000..c1e2b0945ed
--- /dev/null
+++ b/examples/flows/standard/grammar-check/flow.dag.yaml
@@ -0,0 +1,39 @@
+id: template_standard_flow
+name: Template Standard Flow
+inputs:
+  path:
+    type: string
+outputs:
+  output_python:
+    type: string
+    reference: ${input_passthrough.output}
+  output_check:
+    type: string
+    reference: ${grammer_check.output}
+nodes:
+- name: read_file
+  type: python
+  source:
+    type: code
+    path: read_file.py
+  inputs:
+    path: ${inputs.path}
+- name: grammer_check
+  type: llm
+  source:
+    type: code
+    path: grammar_check.jinja2
+  inputs:
+    max_tokens: 1000
+    text: ${read_file.output}
+    deployment_name: gpt-35-turbo
+    model: gpt-35-turbo
+  connection: open_ai_connection
+  api: chat
+- name: input_passthrough
+  type: prompt
+  source:
+    type: code
+    path: input_passthrough.jinja2
+  inputs:
+    input: ${inputs.path}
diff --git a/examples/flows/standard/grammar-check/generate_data.py b/examples/flows/standard/grammar-check/generate_data.py
new file mode 100644
index 00000000000..7bd1b2bbded
--- /dev/null
+++ b/examples/flows/standard/grammar-check/generate_data.py
@@ -0,0 +1,32 @@
+import os
+import json
+
+def find_py_files(directory):
+
+    sql_files = []
+
+    for root, _, files in os.walk(directory):
+        for file in files:
+            if file.endswith('.py'):
+                sql_files.append(os.path.abspath(os.path.join(root, file)))
+
+    return sql_files
+
+def write_to_jsonl(files, output_file="data.jsonl"):
+    """
+    Write list of file paths to a JSON lines file.
+    
+    Args:
+    - files (list): List of file paths.
+    - output_file (str): Name of the output JSON lines file.
+    """
+    with open(output_file, 'w') as out:
+        for file_path in files:
+            line = json.dumps({"path": file_path})
+            out.write(line + '\n')
+
+if __name__ == "__main__":
+    directory = "../../../"
+    sql_files = find_py_files(directory)
+    write_to_jsonl(sql_files)
+    print(f"Output written to data.jsonl")
diff --git a/examples/flows/standard/grammar-check/grammar_check.jinja2 b/examples/flows/standard/grammar-check/grammar_check.jinja2
new file mode 100644
index 00000000000..3a24ae61aaf
--- /dev/null
+++ b/examples/flows/standard/grammar-check/grammar_check.jinja2
@@ -0,0 +1,18 @@
+system:
+You are a python programmer. Please ready the python source doe. Check for any standard/error output to customer, whether the grammar and spelling are correct. If not, please correct them. 
+Please output your suggestion to correct the grammar and spelling.
+Please output the found issues in below json format:
+{
+    "line": 1,
+    "column": 1,
+    "type": "spelling",
+    "message": "Please correct the spelling of 'python'."
+}
+
+If there is no issue, please output:
+{
+    "message": "No issues found."
+}
+
+user:
+{{text}}
\ No newline at end of file
diff --git a/examples/flows/standard/grammar-check/input_passthrough.jinja2 b/examples/flows/standard/grammar-check/input_passthrough.jinja2
new file mode 100644
index 00000000000..2a19217667a
--- /dev/null
+++ b/examples/flows/standard/grammar-check/input_passthrough.jinja2
@@ -0,0 +1 @@
+{{input}}
\ No newline at end of file
diff --git a/examples/flows/standard/grammar-check/read_file.py b/examples/flows/standard/grammar-check/read_file.py
new file mode 100644
index 00000000000..b520ce0c5c2
--- /dev/null
+++ b/examples/flows/standard/grammar-check/read_file.py
@@ -0,0 +1,19 @@
+from promptflow import tool
+import time
+
+# The inputs section will change based on the arguments of the tool function, after you save the code
+# Adding type to arguments and return value will help the system show the types properly
+# Please update the function name/signature per need
+@tool
+def my_python_tool(path: str) -> str:
+  content = ""
+  try:
+    time.sleep(6)
+    with open(path, 'r', encoding='utf-16') as file:
+      content = file.read()
+  except Exception as e:
+    with open(path, 'r') as file:
+      content = file.read()
+  return content
+
+