レビューを元に修正

.env.example

@@ -1,2 +1,4 @@
 OPENAI_API_KEY=XXX
 PINECONE_API_KEY=YYY
+AWS_ACCESS_KEY_ID=ZZZ
+AWS_SECRET_ACCESS_KEY=AAA

.gitignore

@@ -163,6 +163,7 @@ cython_debug/
 .DS_Store
 
 # 申請書/報告書はリポジトリにはアップロードしない。
+data/inputInfo.csv
 data/output.docx
 data/output_deviation.docx
 data/output_deviation_occurrence.docx

README.md

@@ -17,7 +17,9 @@ license: other
 ## 1. .env.exampleを参考に.envを作成し、OpenAIとPINECONEのAPIキーを設定
 ```
 OPENAI_API_KEY =
-PINECONE_API_KEY
+PINECONE_API_KEY =
+AWS_ACCESS_KEY_ID =
+AWS_SECRET_ACCESS_KEY =
 ```
 値はnotionで管理されているので、参照する。
 [OpenAI API](https://www.notion.so/equesai/OpenAI-API-b2029c9b5d774eaebd7318a702ec31f1?pvs=4)

requirements.txt

@@ -8,7 +8,9 @@ streamlit == 1.36.0
 python-docx == 1.1.2
 pinecone-client == 4.1.2
 pre-commit == 3.7.1
+pandas == 2.2.2
+boto3 == 1.34.131
 pandas-stubs == 2.2.2.240603
 boto3-stubs == 1.34.149
-fsspec
-s3fs
+# fsspec == 2024.6.1
+# s3fs == 2024.6.1

src/aws/s3.py

@@ -0,0 +1,38 @@
+from io import StringIO
+from typing import List
+
+import boto3
+import pandas as pd
+
+s3 = boto3.client("s3")
+bucket_name = "eques-pharma-docs"
+# bucket_name = "bankyo-pharmaceutical"
+
+
+# Download the CSV file from S3
+def download_csv_from_s3(file_name: str) -> pd.DataFrame:
+    obj = s3.get_object(Bucket=bucket_name, Key=file_name)
+    data = pd.read_csv(obj["Body"])
+    return data
+
+
+def save_to_s3(data: pd.DataFrame, file_name: str) -> None:
+    csv_buffer = StringIO()
+    data.to_csv(csv_buffer, index=False)
+    s3.put_object(Bucket=bucket_name, Key=file_name, Body=csv_buffer.getvalue())
+
+
+def get_all_objects() -> List[str]:
+    response = s3.list_objects_v2(Bucket=bucket_name)
+    csv_files = [
+        obj["Key"]
+        for obj in response.get("Contents", [])
+        if obj["Key"].endswith(".csv")
+    ]
+    return csv_files
+
+
+def get_csv_content(file_name: str) -> pd.DataFrame:
+    obj = s3.get_object(Bucket=bucket_name, Key=file_name)
+    df = pd.read_csv(obj["Body"])
+    return df.iloc[0]

src/database.py

@@ -1,54 +1,105 @@
-from io import StringIO
+import csv
+from datetime import datetime
 from typing import List
 
-import boto3
+# import boto3
 import pandas as pd
 import streamlit as st
 
-# file_path = "data/inputInfo.csv"
+from src.aws.s3 import get_all_objects, get_csv_content, save_to_s3
 
-# Initialize the S3 client
-s3 = boto3.client("s3")
-
-# Specify the bucket name and file name
 bucket_name = "eques-pharma-docs"
-file_name = "inputInfo.csv"
-
-
-# Download the CSV file from S3
-def download_csv_from_s3(bucket_name: str, file_name: str) -> pd.DataFrame:
-    obj = s3.get_object(Bucket=bucket_name, Key=file_name)
-    data = pd.read_csv(obj["Body"])
-    return data
 
 
 def write_csv(
-    name: str,
-    company_name: str,
+    applicant_name: str,
     change_type: List[str],
-    subtype: List[str],
-    subtype2: List[str],
+    category: List[str],
+    subcategory: List[str],
     input: str,
-    output: str,
+    output_factory: str,
+    output_username: str,
+    output_date: str,
+    output_change_target: str,
+    output_change_period: str,
+    output_change_content: str,
+    output_change_reason: str,
+    output_effect: str,
+    output_resource: str,
+    output_other: str,
 ) -> None:
     all_types = []
     for i in range(len(change_type)):
-        all_type = [i + 1, change_type[i], subtype[i], subtype2[i]]
+        all_type = [i + 1, change_type[i], category[i], subcategory[i]]
         all_types.append(all_type)
 
-    # Download the file, make changes, and re-upload
-    data = download_csv_from_s3(bucket_name, file_name)
+    column_names = [
+        "申請者",
+        "全ての変更タイプ",
+        "入力",
+        "出力:工場",
+        "出力:申請者名",
+        "出力:申請日",
+        "出力:変更対象",
+        "出力:実施予定時期",
+        "出力:変更内容",
+        "出力:変更理由",
+        "出力:品質への影響",
+        "出力:参考資料",
+        "出力:備考",
+    ]
     new_data = pd.DataFrame(
-        [[name, company_name, all_types, input, output]], columns=data.columns
+        [
+            [
+                applicant_name,
+                all_types,
+                input,
+                output_factory,
+                output_username,
+                output_date,
+                output_change_target,
+                output_change_period,
+                output_change_content,
+                output_change_reason,
+                output_effect,
+                output_resource,
+                output_other,
+            ]
+        ],
+        columns=column_names,
     )
-    data = pd.concat([data, new_data], ignore_index=True)
-    csv_buffer = StringIO()
-    data.to_csv(csv_buffer, index=False)
-    s3.put_object(Bucket=bucket_name, Key=file_name, Body=csv_buffer.getvalue())
+    # data = pd.concat([data, new_data], ignore_index=True)
+    timestamp = datetime.now().strftime("%Y-%m-%d %H:%M:%S")
+    file_name = f"{timestamp}.csv"
+    save_to_s3(new_data, file_name)
 
 
 def display_csv() -> None:
-    data = download_csv_from_s3(bucket_name, file_name)
-    data.to_csv("data/inputInfo.csv", index=False)
+    datas = get_all_objects()
+    header = [
+        "申請者",
+        "全ての変更タイプ",
+        "入力",
+        "出力:工場",
+        "出力:申請者名",
+        "出力:申請日",
+        "出力:変更対象",
+        "出力:実施予定時期",
+        "出力:変更内容",
+        "出力:変更理由",
+        "出力:品質への影響",
+        "出力:参考資料",
+        "出力:備考",
+    ]
+    path = "data/inputsData.csv"
+    with open(path, mode="w", newline="") as file:
+        writer = csv.writer(file)
+        writer.writerow(header)
+        if datas:
+            for data in datas:
+                data = get_csv_content(data)
+                writer.writerow(data)
+    # datas.to_csv("data/inputInfo.csv", index=False)
     st.write("**データベース**")
-    st.dataframe(data)
+    df = pd.read_csv(path)
+    st.dataframe(df)

src/forms/change_request.py

@@ -280,10 +280,10 @@ def change_request_form() -> None:
                 st.session_state[f"category_{change_serial_number}"],
                 st.session_state[f"subcategory_{change_serial_number}"],
             )
-            for i in range(len(individual_questions)):
+            for change_serial_number in range(len(individual_questions)):
                 next_question_number = len(question_about_this_change) + 1
                 question_about_this_change.append(
-                    f"Q{change_serial_number}-{next_question_number} : {individual_questions[i]}"
+                    f"Q{change_serial_number}-{next_question_number} : {individual_questions[change_serial_number]}"
                 )
             question_about_this_change.append(
                 f"Q{change_serial_number}-"
@@ -427,26 +427,36 @@ def change_request_form() -> None:
                 output_change_application(st.session_state["contents"])
 
                 type = []
-                subtypes = []
-                subtypes2 = []
+                category = []
+                subcategory = []
 
-                for i in range(1, change_number + 1):
-                    type.append(st.session_state[f"type_{i}"])
-                    subtypes.append(
+                for change_serial_number in range(
+                    start_from_one, change_number + upper_bound_correction
+                ):
+                    type.append(st.session_state[f"type_{change_serial_number}"])
+                    category.append(
                         st.session_state[f"category_{change_serial_number}"]
                     )
-                    subtypes2.append(
+                    subcategory.append(
                         st.session_state[f"subcategory_{change_serial_number}"]
                     )
 
                 write_csv(
                     st.session_state["applicant_name"],
-                    "A社",
                     type,
-                    subtypes,
-                    subtypes2,
+                    category,
+                    subcategory,
                     user_prompt,
-                    st.session_state["contents"],
+                    st.session_state["contents"]["工場"],
+                    st.session_state["contents"]["申請者"],
+                    st.session_state["contents"]["申請日"],
+                    st.session_state["contents"]["変更対象"],
+                    st.session_state["contents"]["実施予定時期"],
+                    st.session_state["contents"]["変更内容"],
+                    st.session_state["contents"]["変更理由"],
+                    st.session_state["contents"]["品質への影響"],
+                    st.session_state["contents"]["添付資料"],
+                    st.session_state["contents"]["備考"],
                 )
                 st.session_state["document_generated"] = True