cristiano.hoshikawa/ai_logistic_outlook_integration
1
0
Fork 0
mirror of https://github.com/hoshikawa2/ai_logistic_outlook_integration.git synced 2026-07-09 16:44:21 +00:00
Code Issues Packages Projects Releases Wiki Activity

first commit

Browse Source
This commit is contained in:
Cristiano Hoshikawa
2025-11-12 08:53:59 -03:00
commit 4066c8d737
16 changed files with 1432 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
.idea/.gitignore generated vendored Normal file
View File

@@ -0,0 +1,12 @@
# Default ignored files
/shelf/
/workspace.xml
# Editor-based HTTP Client requests
/httpRequests/
# Environment-dependent path to Maven home directory
/mavenHomeManager.xml
# Datasource local storage ignored files
/dataSources/
/dataSources.local.xml
# Zeppelin ignored files
/ZeppelinRemoteNotebooks/

20
.idea/ai_logistic_outlook_integration.iml generated Normal file
View File

@@ -0,0 +1,20 @@
<?xml version="1.0" encoding="UTF-8"?>
<module type="JAVA_MODULE" version="4">
<component name="Flask">
<option name="enabled" value="true" />
</component>
<component name="NewModuleRootManager" inherit-compiler-output="true">
<exclude-output />
<content url="file://$MODULE_DIR$" />
<orderEntry type="inheritedJdk" />
<orderEntry type="sourceFolder" forTests="false" />
</component>
<component name="TemplatesService">
<option name="TEMPLATE_CONFIGURATION" value="Jinja2" />
<option name="TEMPLATE_FOLDERS">
<list>
<option value="$MODULE_DIR$/templates" />
</list>
</option>
</component>
</module>

7
.idea/codeStyles/Project.xml generated Normal file
View File

@@ -0,0 +1,7 @@
<component name="ProjectCodeStyleConfiguration">
<code_scheme name="Project" version="173">
<ScalaCodeStyleSettings>
<option name="MULTILINE_STRING_CLOSING_QUOTES_ON_NEW_LINE" value="true" />
</ScalaCodeStyleSettings>
</code_scheme>
</component>

5
.idea/codeStyles/codeStyleConfig.xml generated Normal file
View File

@@ -0,0 +1,5 @@
<component name="ProjectCodeStyleConfiguration">
<state>
<option name="PREFERRED_PROJECT_CODE_STYLE" value="Default" />
</state>
</component>

6
.idea/misc.xml generated Normal file
View File

@@ -0,0 +1,6 @@
<?xml version="1.0" encoding="UTF-8"?>
<project version="4">
<component name="ProjectRootManager" version="2" languageLevel="JDK_24" default="true" project-jdk-name="24" project-jdk-type="JavaSDK">
<output url="file://$PROJECT_DIR$/out" />
</component>
</project>

8
.idea/modules.xml generated Normal file
View File

@@ -0,0 +1,8 @@
<?xml version="1.0" encoding="UTF-8"?>
<project version="4">
<component name="ProjectModuleManager">
<modules>
<module fileurl="file://$PROJECT_DIR$/.idea/ai_logistic_outlook_integration.iml" filepath="$PROJECT_DIR$/.idea/ai_logistic_outlook_integration.iml" />
</modules>
</component>
</project>

7
.idea/vcs.xml generated Normal file
View File

@@ -0,0 +1,7 @@
<?xml version="1.0" encoding="UTF-8"?>
<project version="4">
<component name="VcsDirectoryMappings">
<mapping directory="" vcs="Git" />
<mapping directory="$PROJECT_DIR$" vcs="Git" />
</component>
</project>

24
.oca/custom_code_review_guidelines.txt Normal file
View File

@@ -0,0 +1,24 @@
# Sample guideline, please follow similar structure for guideline with code samples
# 1. Suggest using streams instead of simple loops for better readability.
# <example>
# *Comment:
# Category: Minor
# Issue: Use streams instead of a loop for better readability.
# Code Block:
#
# ```java
# // Calculate squares of numbers
# List<Integer> squares = new ArrayList<>();
# for (int number : numbers) {
# squares.add(number * number);
# }
# ```
# Recommendation:
#
# ```java
# // Calculate squares of numbers
# List<Integer> squares = Arrays.stream(numbers)
# .map(n -> n * n) // Map each number to its square
# .toList();
# ```
# </example>

353
README.md Normal file
View File

@@ -0,0 +1,353 @@
# Email Export Process Monitor with OCI Generative AI, Outlook Emails and Oracle Database
This project automates the monitoring of export and logistics communications by reading Outlook `.msg` emails from OCI Object Storage, analyzing them with OCI Generative AI to extract status and alerts, summarizing content, and displaying results in a Flask web application backed by Oracle Database.
---
# Email Export Process Monitor with OCI Generative AI, Outlook e-mails and Oracle Database
> A production‑grade pipeline that **reads Outlook `.msg` emails** from an **OCI Object Storage** inbox, uses **Generative AI** to understand the **status of an export/logistics process**, **summarizes** each message, flags **alerts**, and serves a **timeline report** via a **Flask web app** backed by **Oracle Database**.
---
## 1) What problem this project solves (Use Case)
Export and logistics operations generate long email threads (forwarded/quoted chains, multiple senders, mixed languages). Manually reading and updating spreadsheets or systems is slow and error‑prone.
This project **automates** that routine:
1. **Ingest** Outlook emails (`.msg`) placed in an **OCI Object Storage** bucket (e.g., `inbox/`).
2. **Parse** each email (including forwarded/replied headers) and **split** the thread into individual communications.
3. **Summarize** and **classify** each communication with **OCI Generative AI** (via LangChain), extracting **structured fields** like:
- `date_sent` (ISO‑8601 when possible)
- `booking`
- `email_from`, `email_to`, `subject`
- `status` (e.g., “Send Note to Forwarder”, “Approve Invoice”, “Perform Weighing”, etc.)
- `brief_description` (1–2 lines)
- `alert` (“YES”/“NO” based on urgency, problems, missing docs, delays, etc.)
4. **Persist** the result to **Oracle Database** (`PROCESSED_EMAILS`), one row per communication (thread index).
5. **Move** processed files to a `processed/` prefix to avoid re‑processing.
6. **Visualize** a **timeline report** and filters in a **Flask UI** (date range, booking, status, alerts). You can also open each original email to inspect content.
**Outcome:** A near‑real‑time **control panel** of export process communications with **actionable alerts** and **structured status**—no manual triage of huge mail chains.
![img.png](images/img.png)
---
## 2) Other use cases you can support with the same architecture
- **Accounts Payable/Receivable**: triage invoice/dispute emails, extract amounts, due dates, and blockers.
- **Customer Support**: summarize tickets by email, detect intent/severity, and push alerts to a dashboard.
- **Procurement**: track RFQs/quotes/PO confirmations arriving by email and update ERP status fields.
- **Compliance**: auto‑detect phrases indicating risk/backlogs or missing mandatory documents.
- **Sales Ops**: parse quote/booking confirmations and flag risks (price mismatch, dates, missing attachments).
---
## 3) Architecture (high level)
```mermaid
flowchart LR
A[Outlook emails .msg\nOCI Object Storage bucket: inbox/] --> B[Processor\nprocess_emails_v2.py]
B -->|Split thread + chunk| C[LLM OCI Generative AI\nmodel_id: cohere.command-a-03-2025]
C -->|JSON results| D[Oracle DB 23ai\nEMAILS_PROCESSADOS]
B -->|Move done| A2[Object Storage: processed/]
D --> E[Flask App app_emails.py\nTimeline + Filters + View .msg]
```
Key behaviors:
- **Thread splitter** detects “From/De, Sent/Sent, To/Para, Subject/Subject…” headers and separators (“Original Message / Original Message / Forwarded message”).
- **Chunking** limits each LLM call to ~10k chars (≈2.5k tokens) to keep cost/latency predictable.
- **Prompting** uses a **deterministic, JSON‑only** instruction in Portuguese tailored to **international logistics**.
- **DB write** normalizes dates and strings, stores `thread_index` per communication.
- **UI** renders a chronological **timeline** with filters and **redacted emails** if desired.
---
## 4) Technologies
| Layer | Technology | Why |
|---|---|---|
| Email parsing | `extract_msg` | Robust reader for Outlook `.msg` files (headers, body) |
| Object storage | **OCI Object Storage** | Durable inbox/processed area, cheap and scalable |
| AI runtime | **OCI Generative AI** via `langchain_community.chat_models.oci_generative_ai.ChatOCIGenAI` | Secure, low‑latency, enterprise LLM serving |
| Orchestration | **LangChain** (`ChatPromptTemplate`) | Structured prompting & message assembly |
| Database | **Oracle Database 23ai** (`oracledb`) | Reliable transactional store, SQL analytics, wallet auth |
| API/UI | **Flask** | Simple server‑side UI for timeline + filters |
| Config/Auth | `oci` SDK + `~/.oci/config` profile | Standardized credential management |
---
## 5) Project structure (key files)
- **`process_emails_v2.py`** – batch processor:
- Lists `.msg` files in `BUCKET_NAME` inbox
- Downloads to temp, **extracts top headers + body**
- **Splits** thread into messages and **chunks** text to fit LLM limits
- **Builds prompt** and calls **OCI GenAI** for each chunk
- **Parses JSON**, **normalizes** fields, and **inserts** into `PROCESSED_EMAILS`
- On success, **moves** the `.msg` object to `processed/`
- Propagates **last detected `booking`** to blank items of the same file
- **`app_emails.py`** – Flask web app:
- Connects to Oracle using **wallet** (`WALLET_PATH`) and `DB_ALIAS`
- Reads from **`PROCESSED_EMAILS_REDACTED` view** (recommended) or the base table
- Provides **filters**: `booking`, `alert (YES/NO)`, `status`, `date range`, and option to **include forwarded-only** messages
- Shows **timeline** sorted by `DATE_SENT` ASC
- Route **`/view_email/<file_name>`** renders the original `.msg` (downloaded from OCI) for inspection
---
## 6) Data model
### 6.1 Table
`PROCESSED_EMAILS` (from the inline DDL in `process_emails_v2.py`):
```sql
CREATE TABLE PROCESSED_EMAILS (
ID NUMBER GENERATED ALWAYS AS IDENTITY PRIMARY KEY,
DATE_SENT TIMESTAMP,
BOOKING VARCHAR2(100),
EMAIL_FROM VARCHAR2(500),
EMAIL_TO VARCHAR2(1000),
SUBJECT VARCHAR2(1000),
STATUS VARCHAR2(100),
BRIEF_DESCRIPTION VARCHAR2(1000),
ALERT VARCHAR2(10),
THREAD_INDEX NUMBER(10),
FILE_NAME VARCHAR2(400),
INSERT_DATE TIMESTAMP DEFAULT CURRENT_TIMESTAMP
);
```
> **Tip:** For the UI, create a sanitized view that redacts email addresses:
```sql
CREATE OR REPLACE VIEW PROCESSED_EMAILS_REDACTED AS
SELECT
REGEXP_REPLACE(
REGEXP_REPLACE(
REGEXP_REPLACE(
EMAIL_TO,
'([[:alpha:]][[:alpha:] .''()_-]*)<([^>]+)>',
'<\2>'
),
'([A-Za-z])([A-Za-z0-9._%+-]*)@([A-Za-z])([A-Za-z0-9.-]*)(\.[A-Za-z]{2,}(?:\.[A-Za-z]{2,})*)',
'\1******@\3******\5',
1, 0, 'i'
),
'\s*;\s*', '; '
) AS EMAIL_TO,
REGEXP_REPLACE(
REGEXP_REPLACE(
REGEXP_REPLACE(
EMAIL_FROM,
'([[:alpha:]][[:alpha:] .''()_-]*)<([^>]+)>',
'<\2>'
),
'([A-Za-z])([A-Za-z0-9._%+-]*)@([A-Za-z])([A-Za-z0-9.-]*)(\.[A-Za-z]{2,}(?:\.[A-Za-z]{2,})*)',
'\1******@\3******\5',
1, 0, 'i'
),
'\s*;\s*', '; '
) AS EMAIL_FROM,
SUBJECT,
REGEXP_REPLACE(
BOOKING,
'([A-Za-z0-9])([A-Za-z0-9 _.-]*)',
'***'
) AS BOOKING,
DATE_SENT,
STATUS,
FILE_NAME,
THREAD_INDEX, INSERT_DATE, BRIEF_DESCRIPTION, ALERT
FROM PROCESSED_EMAILS;
/
```
### 6.2 Key normalization logic
- `normalize_datetime()` parses ISO‑8601 and common patterns, removes TZ for DB TIMESTAMP.
- `sanitize_value()` stringifies lists/dicts and trims empty values.
---
## 7) Configuration
Create a JSON file named **`./config`** used by **both** scripts.
Minimum keys (adapt to your tenancy and DB):
```json
{
"oci_profile": "DEFAULT",
"llm_endpoint": "https://inference.generativeai.us-phoenix-1.oci.oraclecloud.com",
"namespace": "xxxxxxxxxxxxxxxx",
"compartment_id": "ocid1.compartment.oc1..example",
"bucket-profile": "DEFAULT",
"bucket": "emails",
"WALLET_PATH": "/path/to/db/wallet",
"DB_ALIAS": "adb_high",
"USERNAME": "APPUSER",
"PASSWORD": "****"
}
```
> In `process_emails_v2.py` the following constants are read/used:
> - `MODEL_ID="cohere.command-a-03-2025"` (you can change to another model available in your tenancy)
> - `SERVICE_ENDPOINT`, `COMPARTMENT_ID`, `AUTH_PROFILE`
> - `BUCKET_NAME`, `PROCESSED_PREFIX="processed/"`
> - `DB_ALIAS`, `USERNAME`, `PASSWORD`, `WALLET_PATH`
> In `app_emails.py` you’ll also see:
> - `bucket-profile`, `bucket`, `compartment_id` for Object Storage access
> - The Flask app runs by default on **`0.0.0.0:5015`**
OCI credentials come from `~/.oci/config` under the chosen profile(s).
---
## 8) How the processing works (deep dive)
### 8.1 Extract the latest message from a `.msg`
- `extract_msg_text(path)` uses `extract_msg.Message` to read `sender`, `to`, `subject`, `date`, and `body`.
- Returns a dict with `top` (headers) and `body`.
### 8.2 Split threads reliably
- `split_refined_email_thread(text)` detects boundaries using a robust regex that looks for header groups (From/De, Sent/Sent, To/Para, Subject/Subject) and common separators like “Original Message”, “Original Message”, “Forwarded message”.
- Fallback to `split_thread_by_headers()` if needed.
### 8.3 Chunking & prompts
- `chunk_by_size(text, MAX_CHARS_PER_CALL=10000)` ensures each LLM call is bounded.
- `prepare_blocks_for_llm(headers_topo, body)` packs the first chunk with top headers if needed.
- `build_prompt(bloco_texto, headers_topo)` (Portuguese) directs the model to **return JSON ONLY** with fields:
`data_envio, booking, email_from, email_to, subject, status, brief_description, alert`
and maps domain‑specific phrases to **standardized `status`** values (e.g., *“Send Note to Forwarder”, “Realizar Estufagem”, “Validar Peso”, “Approve Invoice”*, etc.).
### 8.4 LLM call
- `call_llm_extract()` instantiates `ChatOCIGenAI` with:
- `model_id=MODEL_ID` (default: `cohere.command-a-03-2025`)
- `temperature=0.0` for determinism, `max_tokens≈1800`, `top_p=0.9`
- Cleans fenced code blocks and parses JSON; takes first dict if a list is returned.
### 8.5 Persist results
- `insert_into_database(item, file_name, thread_index)` writes each communication to `PROCESSED_EMAILS`.
- After processing a file, the **last detected `booking`** is propagated to rows where it’s empty (same `FILE_NAME`).
### 8.6 Idempotence
- Objects are **moved to `processed/`** after success to prevent reprocessing.
---
## 9) Running locally (end‑to‑end)
### 9.1 Prerequisites
- **Python 3.10+**
- **Oracle Instant Client** (for thick mode) and **DB wallet** (place path in `WALLET_PATH`)
- Access to **OCI Tenancy** with permissions to read the chosen **Object Storage** bucket
- A table `PROCESSED_EMAILS` created in your schema
Install dependencies (example):
```bash
python -m venv .venv
source .venv/bin/activate # Windows: .venv\Scripts\activate
pip install flask oracledb oci extract_msg langchain langchain-core langchain-community
```
### 9.2 Prepare Object Storage
- Create bucket (e.g., `emails`).
- Upload Outlook `.msg` files into `emails/inbox/` (or your chosen prefix).
Ensure the paths in `process_emails_v2.py` match your structure.
### 9.3 Configure
- Create `./config` (see section **7**).
- Ensure `~/.oci/config` has the profile named in `"oci_profile"` / `"bucket-profile"`.
### 9.4 Process emails
Run the batch once (or schedule via cron):
```bash
python process_emails_v2.py
```
Expected logs:
- Files listed and downloaded
- Thread chunks sent to LLM
- Rows inserted in `PROCESSED_EMAILS`
- File moved to `processed/`
### 9.5 Start the web app
```bash
python app_emails.py
```
Open: **http://localhost:5015/**
Use filters:
- **Booking**: case‑insensitive contains
- **Alert**: `YES` or `NO`
- **Status**: dropdown from `SELECT DISTINCT STATUS …`
- **Date range**: `inicio` / `fim` (inclusive)
- **Include forwarded‑only**: checkbox to include messages without a classified status
Click on entries to **view the original email** via `/view_email/<file_name>`.
---
## 10) Testing tips
- Add a small test `.msg` with a short thread containing an English and a Portuguese part referencing export steps (e.g., DU‑e, invoice approval, warehouse dates).
- Verify:
- `status` gets standardized per prompt rules
- `alert` hits **“YES”** when there’s a delay/missing document/urgent note
- Multiple communications in the same thread create **multiple rows** with incrementing `THREAD_INDEX`
- Timeline sorts by `DATE_SENT` ASC
- Create the **REDACTED view** and point the app to it if sharing the UI broadly.
---
## 11) Operations & scaling
- **Throughput**: adjust `MAX_CHARS_PER_CALL` and chunking to balance latency/cost.
- **Retries**: wrap `call_llm_extract()` with exponential backoff if needed.
- **Observability**: log `file_name`, `thread_index`, LLM latency, and DB writes.
- **Security**: keep wallets and OCI config out of source control; use per‑environment profiles.
- **Cost**: batch larger emails off‑peak; cache results for unchanged `.msg` object hashes.
---
## 12) Appendix – notable functions (by file)
### `process_emails_v2.py`
- `list_files()` – list `.msg` in bucket
- `download_file(name)` – temp‑download object
- `move_to_processed(name)` – move to `processed/`
- `extract_msg_text(path)` – read `.msg` headers/body
- `split_refined_email_thread(texto)` / `split_thread_by_headers(texto)` – reliable thread splitter
- `chunk_by_size(text)` / `prepare_blocks_for_llm(headers, body)` – chunk & pack
- `build_prompt(bloco, headers_topo)` – robust JSON‑only, domain‑specific prompt
- `call_llm_extract(bloco, headers_topo)` – OCI GenAI call + JSON parse
- `insert_into_database(item, file_name, thread_index)` – DB insert
- `process_emails()` – the pipeline entrypoint
### `app_emails.py`
- `search_emails()` – main route “/”: filters + timeline using `PROCESSED_EMAILS_REDACTED`
- `get_distinct_status()` – populates status dropdown
- `view_email(file_name)` – renders original `.msg` body for audit
---
## 13) FAQ
**Can I change the LLM?**
Yes. Edit `MODEL_ID` and `SERVICE_ENDPOINT`. Keep the prompt constraints (JSON‑only) for reliable parsing.
**Do I need the redacted view?**
Recommended for sharing the UI. The processor writes raw values; the view masks addresses for privacy.
**Where do processed `.msg` go?**
To `processed/` (prefix configurable by `PROCESSED_PREFIX`).
**How do I handle attachments?**
Extend `extract_msg_text()` to enumerate attachments from `extract_msg` and store metadata in a separate table.
---
## Acknowledgments
- **Author** - Cristiano Hoshikawa (Oracle LAD A-Team Solution Engineer)

233
files/app_emails.py Normal file
View File

@@ -0,0 +1,233 @@
# -*- coding: utf-8 -*-
from flask import Flask, render_template, request, send_file, redirect, url_for, render_template_string
import oracledb, json
from datetime import datetime
import io
import oci
import urllib.parse
import extract_msg
# ============================================================
# CONFIGURATION
# ============================================================
with open("config", "r") as f:
config_data = json.load(f)
WALLET_PATH = config_data["WALLET_PATH"]
DB_ALIAS = config_data["DB_ALIAS"]
USERNAME = config_data["USERNAME"]
PASSWORD = config_data["PASSWORD"]
connection = oracledb.connect(
user=USERNAME,
password=PASSWORD,
dsn=DB_ALIAS,
config_dir=WALLET_PATH,
wallet_location=WALLET_PATH,
wallet_password=PASSWORD,
)
# OCI client configuration (same standard as process_emails_v2_en.py)
BUCKET_PROFILE = config_data["bucket-profile"]
BUCKET_NAME = config_data["bucket"]
COMPARTMENT_ID = config_data["compartment_id"]
bucket_name = BUCKET_NAME + "/processed"
# Initialize OCI client and get namespace
config = oci.config.from_file("~/.oci/config", BUCKET_PROFILE)
object_storage = oci.object_storage.ObjectStorageClient(config)
namespace = object_storage.get_namespace().data
app = Flask(__name__)
# ============================================================
# HELPER FUNCTIONS
# ============================================================
def get_distinct_status():
"""Return all distinct statuses + 'Forwarded Email' option."""
with connection.cursor() as c:
c.execute("SELECT DISTINCT STATUS FROM PROCESSED_EMAILS ORDER BY STATUS")
values = [r[0] for r in c.fetchall() if r[0] and str(r[0]).strip()]
if "Forwarded Email" not in values:
values.append("Forwarded Email")
return values
# ============================================================
# MAIN ROUTE
# ============================================================
@app.route("/", methods=["GET"])
def search_emails():
booking = request.args.get("booking", "").strip().upper()
alert_flag = request.args.get("alert", "")
status = request.args.get("status", "")
start_date = request.args.get("start", "")
end_date = request.args.get("end", "")
include_forwarded = request.args.get("include_forwarded") == "on"
sql = """
SELECT DATE_SENT, BOOKING, EMAIL_FROM, EMAIL_TO, SUBJECT,
STATUS, BRIEF_DESCRIPTION, ALERT, FILE_NAME, THREAD_INDEX
FROM PROCESSED_EMAILS_REDACTED
WHERE 1=1 \
"""
params = {}
if booking:
sql += " AND UPPER(BOOKING) LIKE :bk"
params["bk"] = f"%{booking}%"
if alert_flag in ("YES", "NO"):
sql += " AND ALERT = :alert"
params["alert"] = alert_flag
if status:
if status == "Forwarded Email":
sql += " AND (STATUS IS NULL OR TRIM(STATUS) = '')"
else:
sql += " AND STATUS = :status"
params["status"] = status
else:
# If the checkbox is NOT checked, exclude forwarded emails (STATUS is null)
if not include_forwarded:
sql += " AND (STATUS IS NOT NULL)"
if start_date:
sql += " AND DATE_SENT >= TO_TIMESTAMP(:start, 'YYYY-MM-DD')"
params["start"] = start_date
if end_date:
sql += " AND DATE_SENT <= TO_TIMESTAMP(:end, 'YYYY-MM-DD') + INTERVAL '1' DAY"
params["end"] = end_date
sql += " ORDER BY DATE_SENT ASC"
with connection.cursor() as c:
c.execute(sql, params)
rows = c.fetchall()
cols = [d[0] for d in c.description]
records = [dict(zip(cols, row)) for row in rows]
for r in records:
if isinstance(r["DATE_SENT"], datetime):
r["DATE_SENT"] = r["DATE_SENT"].strftime("%Y-%m-%d %H:%M:%S")
if not r.get("STATUS") or not str(r["STATUS"]).strip():
r["STATUS"] = "Forwarded Email"
status_options = get_distinct_status()
return render_template(
"emails_filter.html",
records=records,
filters={
"booking": booking,
"alert": alert_flag,
"status": status,
"start": start_date,
"end": end_date,
"include_forwarded": include_forwarded
},
status_options=status_options
)
# ============================================================
# ROUTE: VIEW EMAIL
# ============================================================
@app.route("/view_email/<path:file_name>")
def view_email(file_name):
try:
# Fix URL encoding
file_name = urllib.parse.unquote(file_name).strip()
# Try multiple possible paths
possible_paths = [
file_name,
f"processed/{file_name}",
f"inbox/{file_name}",
f"emails/{file_name}",
f"emails/processed/{file_name}",
]
msg_obj = None
for path in possible_paths:
try:
obj = object_storage.get_object(namespace, BUCKET_NAME, path)
msg_obj = extract_msg.Message(io.BytesIO(obj.data.content))
print(f"[INFO] File found: {path}")
break
except oci.exceptions.ServiceError as e:
if e.status == 404:
continue
raise
if not msg_obj:
return (
f"<h3>❌ Email not found in any tested path.</h3>"
f"<p><b>Bucket:</b> {BUCKET_NAME}<br><b>File:</b> {file_name}</p>",
404,
)
# Extract main fields
from_field = getattr(msg_obj, "sender", "")
to_field = getattr(msg_obj, "to", "")
subject = getattr(msg_obj, "subject", "")
date = getattr(msg_obj, "date", "")
body = getattr(msg_obj, "body", "")
# Basic HTML rendering for readability
html = f"""
<html lang="en">
<head>
<meta charset="UTF-8">
<title>{subject or "Viewed Email"}</title>
<style>
body {{
font-family: Arial, sans-serif;
background: #f4f4f9;
padding: 30px;
color: #333;
}}
.email-container {{
background: white;
border-radius: 10px;
box-shadow: 0 2px 6px rgba(0,0,0,0.1);
padding: 20px;
max-width: 900px;
margin: auto;
}}
.header {{
border-bottom: 1px solid #ccc;
margin-bottom: 20px;
padding-bottom: 10px;
}}
.header b {{
color: #004b8d;
}}
pre {{
white-space: pre-wrap;
word-wrap: break-word;
font-family: Arial, sans-serif;
font-size: 15px;
}}
</style>
</head>
<body>
<div class="email-container">
<div class="header">
<p><b>From:</b> {from_field}</p>
<p><b>To:</b> {to_field}</p>
<p><b>Date:</b> {date}</p>
<p><b>Subject:</b> {subject}</p>
</div>
<pre>{body}</pre>
</div>
</body>
</html>
"""
return render_template_string(html)
except Exception as e:
return f"<h3>Error reading the email:</h3><pre>{e}</pre>", 500
# ============================================================
# MAIN ENTRY POINT
# ============================================================
if __name__ == "__main__":
app.run(debug=True, host="0.0.0.0", port=5015)

11
files/config Normal file
View File

@@ -0,0 +1,11 @@
{
"oci_profile": "DEFAULT",
"llm_endpoint": "https://inference.generativeai.us-phoenix-1.oci.oraclecloud.com",
"compartment_id": "ocid1.compartment.oc1..example",
"bucket-profile": "DEFAULT",
"bucket": "emails",
"WALLET_PATH": "/path/to/db/wallet",
"DB_ALIAS": "adb_high",
"USERNAME": "APPUSER",
"PASSWORD": "****"
}

458
files/process_emails_v2.py Normal file
View File

@@ -0,0 +1,458 @@
# -*- coding: utf-8 -*-
import oci
import json
import tempfile
import os
import re
from datetime import datetime
from typing import List, Dict, Any
import oracledb
from extract_msg import Message as MsgReader
from oci.object_storage import ObjectStorageClient
from langchain_core.prompts import ChatPromptTemplate
from langchain_community.chat_models.oci_generative_ai import ChatOCIGenAI
# ============================================================
# DATABASE TABLE DEFINITION
# ============================================================
# CREATE TABLE PROCESSED_EMAILS (
# ID NUMBER GENERATED ALWAYS AS IDENTITY,
# DATE_SENT TIMESTAMP,
# BOOKING VARCHAR2(100),
# EMAIL_FROM VARCHAR2(500),
# EMAIL_TO VARCHAR2(1000),
# SUBJECT VARCHAR2(1000),
# STATUS VARCHAR2(100),
# INSERT_DATE TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
# PRIMARY KEY (ID)
# );
#
# ALTER TABLE PROCESSED_EMAILS ADD (
# BRIEF_DESCRIPTION VARCHAR2(1000),
# ALERT VARCHAR2(10),
# THREAD_INDEX NUMERIC(10),
# FILE_NAME VARCHAR(400)
# );
# ============================================================
# CONFIGURATION
# ============================================================
with open("config", "r") as f:
config_data = json.load(f)
MODEL_ID = "cohere.command-a-03-2025"
SERVICE_ENDPOINT = config_data["llm_endpoint"]
COMPARTMENT_ID = config_data["compartment_id"]
AUTH_PROFILE = config_data["oci_profile"]
WALLET_PATH = config_data["WALLET_PATH"]
DB_ALIAS = config_data["DB_ALIAS"]
USERNAME = config_data["USERNAME"]
PASSWORD = config_data["PASSWORD"]
BUCKET_NAME = config_data["bucket"]
PROCESSED_PREFIX = "processed/"
CONFIG_PROFILE = AUTH_PROFILE
MAX_CHARS_PER_CALL = 10000 # ~2500 tokens per block
# ============================================================
# OCI CLIENTS & DATABASE CONNECTION
# ============================================================
config = oci.config.from_file("~/.oci/config", CONFIG_PROFILE)
namespace = oci.object_storage.ObjectStorageClient(config).get_namespace().data
object_client = ObjectStorageClient(config)
connection = oracledb.connect(
user=USERNAME,
password=PASSWORD,
dsn=DB_ALIAS,
config_dir=WALLET_PATH,
wallet_location=WALLET_PATH,
wallet_password=PASSWORD,
)
# ============================================================
# UTILS: LIST AND MANAGE OBJECTS IN BUCKET
# ============================================================
def list_files() -> List[str]:
"""List all .msg files in the configured Object Storage bucket."""
objs = object_client.list_objects(namespace, BUCKET_NAME, fields="name")
return [o.name for o in objs.data.objects if o.name.lower().endswith(".msg")]
def download_file(name: str) -> str:
"""Download a .msg file from OCI Object Storage to a temporary folder."""
temp_dir = tempfile.mkdtemp()
path = os.path.join(temp_dir, os.path.basename(name))
with open(path, "wb") as f:
data = object_client.get_object(namespace, BUCKET_NAME, name).data.content
f.write(data)
return path
def move_to_processed(name: str):
"""Move a processed file to the 'processed/' prefix in the same bucket."""
destination = PROCESSED_PREFIX + os.path.basename(name)
obj = object_client.get_object(namespace, BUCKET_NAME, name)
object_client.put_object(namespace, BUCKET_NAME, destination, obj.data.content)
object_client.delete_object(namespace, BUCKET_NAME, name)
# ============================================================
# .MSG EXTRACTION
# ============================================================
def extract_msg_text(path):
"""
Extract only the first (most recent) message from the .msg file.
Returns:
{
"top": {headers},
"body": "message body"
}
"""
msg = MsgReader(path)
top = {
"from": getattr(msg, "sender", ""),
"to": getattr(msg, "to", ""),
"subject": getattr(msg, "subject", ""),
"date": str(getattr(msg, "date", "")),
}
body = getattr(msg, "body", "") or ""
# Remove previous messages (older thread content)
thread_pattern = re.compile(
r"(?i)(^-{2,}\s*(original message|forwarded message|mensagem original)\s*-{2,}$|^from:|^sent:|^to:|^de:|^enviado:|^para:)",
re.MULTILINE
)
parts = thread_pattern.split(body)
cleaned_body = parts[0].strip() if parts else body.strip()
return {"top": top, "body": cleaned_body}
# ============================================================
# THREAD SPLITTING
# ============================================================
HEADER_PATTERNS = [
r"^-{2,}\s*Original Message\s*-{2,}$",
r"^-{2,}\s*Forwarded message\s*-{2,}$",
r"^From:\s", r"^Sent:\s", r"^To:\s", r"^Subject:\s"
]
BOUNDARY_REGEX = re.compile("|".join(HEADER_PATTERNS), re.IGNORECASE | re.MULTILINE)
def split_email_thread_refined(text):
"""
Split an email body into message blocks within a thread,
considering standard header sets.
"""
text = text.replace('\r\n', '\n').strip()
regex_start_email = re.compile(
r"(?=^(From|De):.*\n(Sent|Enviado):.*\n(To|Para):.*)",
re.MULTILINE | re.IGNORECASE
)
parts = regex_start_email.split(text)
blocks = []
if len(parts) <= 1:
return [text.strip()] if text.strip() else []
current = ""
for line in parts:
line = line.strip()
if not line:
continue
if re.match(r"^(From|De):", line, re.IGNORECASE):
if current:
blocks.append(current.strip())
current = line
else:
current += "\n" + line
if current.strip():
blocks.append(current.strip())
cleaned_blocks = []
for b in blocks:
b = b.strip()
if b and b not in cleaned_blocks:
cleaned_blocks.append(b)
return cleaned_blocks
def split_thread_by_headers(body: str) -> List[str]:
lines = body.splitlines()
blocks, curr = [], []
def flush():
if curr:
text = "\n".join(curr).strip()
if text:
blocks.append(text)
for line in lines:
if BOUNDARY_REGEX.match(line.strip()) and curr:
flush()
curr = [line]
else:
curr.append(line)
flush()
if not blocks:
return [body.strip()] if body.strip() else []
return blocks
def chunk_by_size(text: str, max_chars: int) -> List[str]:
"""Split text into chunks within a character limit."""
text = text.strip()
if len(text) <= max_chars:
return [text]
chunks, curr = [], ""
for part in re.split(r"(\n\s*\n)", text):
if len(curr) + len(part) <= max_chars:
curr += part
else:
if curr.strip():
chunks.append(curr.strip())
if len(part) > max_chars:
for i in range(0, len(part), max_chars):
chunks.append(part[i:i + max_chars].strip())
curr = ""
else:
curr = part
if curr.strip():
chunks.append(curr.strip())
return chunks
def prepare_blocks_for_llm(body: str, top_headers: Dict[str, Any]) -> List[str]:
"""Prepare text blocks for LLM processing."""
body = (body or "").strip()
if not body:
hint = f"From: {top_headers.get('from','')}\nTo: {top_headers.get('to','')}\nSubject: {top_headers.get('subject','')}\nDate: {top_headers.get('date','')}"
return [hint]
by_headers = split_thread_by_headers(body)
result = []
for b in by_headers:
for c in chunk_by_size(b, MAX_CHARS_PER_CALL):
if c.strip():
result.append(c.strip())
if not result:
hint = f"From: {top_headers.get('from','')}\nTo: {top_headers.get('to','')}\nSubject: {top_headers.get('subject','')}\nDate: {top_headers.get('date','')}\n\n{body[:8000]}"
result = [hint]
return result
# ============================================================
# PROMPT (ENGLISH VERSION)
# ============================================================
def build_prompt(block_text: str, top_headers: Dict[str, Any]) -> ChatPromptTemplate:
sys_text = """
You are a data extractor for corporate logistics emails.
Each email may contain multiple communications (replies, forwards, or quoted messages) within the same thread.
Your task is to identify each communication separately and extract the structured fields below, returning everything as a JSON list.
Extraction rules:
1. **Communication identification**
- Each individual message starts when a new header with "From:", "Sent:", or similar appears.
- Also consider blocks separated by "Original Message" or "Forwarded Message".
- Each identified block must generate a separate JSON object.
2. **Fields to extract (per message)**
- "date_sent": Date and time the message was sent (ISO 8601).
- "booking": Booking or shipment code (“Booking”, “Bkg”, “BK”, etc.).
- "email_from": Sender (“From:”).
- "email_to": Main recipients (“To:”).
- "subject": Message subject (“Subject:”).
- "status": Operational status inferred from text using logistics keywords (shipment, invoice, forwarding, approval, etc.).
- "brief_description": Short (1–2 lines) natural language summary.
- "alert": "YES" if the message contains urgency, problems, delays, or missing documents; "NO" otherwise.
3. **Output format**
- Return ONLY JSON (no extra text).
- Example:
[
{
"date_sent": "",
"booking": "",
"email_from": "",
"email_to": "",
"subject": "",
"status": "",
"brief_description": "",
"alert": ""
}
]
Context:
From: {from_hint}
To: {to_hint}
Subject: {subject_hint}
Date: {date_hint}
Text to analyze:
"""
prompt = ChatPromptTemplate.from_messages([
("system", sys_text),
("human", block_text)
])
prompt = prompt.partial(
from_hint=top_headers.get("from", ""),
to_hint=top_headers.get("to", ""),
subject_hint=top_headers.get("subject", ""),
date_hint=top_headers.get("date", ""),
)
return prompt
# ============================================================
# LLM AND DATABASE OPERATIONS
# ============================================================
def clean_markdown_fences(text: str) -> str:
text = text.strip()
text = re.sub(r"^```[a-zA-Z]*\s*", "", text)
text = re.sub(r"\s*```$", "", text)
return text.strip()
def call_llm_extract(block_text: str, top_headers: Dict[str, Any]) -> Dict[str, Any]:
llm = ChatOCIGenAI(
model_id=MODEL_ID,
service_endpoint=SERVICE_ENDPOINT,
compartment_id=COMPARTMENT_ID,
auth_profile=AUTH_PROFILE,
model_kwargs={"temperature": 0.0, "max_tokens": 1800, "top_p": 0.9},
)
prompt = build_prompt(block_text, top_headers)
resp = llm.invoke(prompt.format_messages())
content = clean_markdown_fences(resp.content or "")
try:
data = json.loads(content)
if isinstance(data, list) and data:
return data[0]
if isinstance(data, dict):
return data
return {"error": "Unexpected format", "raw": content[:500]}
except Exception:
return {"error": "Invalid JSON", "raw": content[:500]}
def normalize_datetime(value):
if value is None:
return None
if isinstance(value, datetime):
return value
s = str(value).strip()
if not s:
return None
try:
if s.endswith("Z"):
s = s[:-1] + "+00:00"
return datetime.fromisoformat(s).replace(tzinfo=None)
except Exception:
for fmt in ("%Y-%m-%dT%H:%M:%S", "%Y-%m-%d"):
try:
return datetime.strptime(s, fmt)
except Exception:
continue
return None
def sanitize_value(v):
if v is None:
return None
if isinstance(v, (list, tuple, set)):
return ", ".join(str(x) for x in v)
if isinstance(v, dict):
return json.dumps(v, ensure_ascii=False)
return str(v)
def insert_into_database(item, file_name, thread_index):
with connection.cursor() as cursor:
binds = {
"date_sent": normalize_datetime(item.get("date_sent")),
"booking": sanitize_value(item.get("booking")),
"email_from": sanitize_value(item.get("email_from")),
"email_to": sanitize_value(item.get("email_to")),
"subject": sanitize_value(item.get("subject")),
"status": sanitize_value(item.get("status")),
"brief_description": sanitize_value(item.get("brief_description")),
"alert": sanitize_value(item.get("alert")),
"file_name": file_name,
"thread_index": thread_index
}
cursor.execute("""
INSERT INTO PROCESSED_EMAILS
(DATE_SENT, BOOKING, EMAIL_FROM, EMAIL_TO, SUBJECT, STATUS, BRIEF_DESCRIPTION, ALERT, FILE_NAME, THREAD_INDEX)
VALUES
(:date_sent, :booking, :email_from, :email_to, :subject, :status, :brief_description, :alert, :file_name, :thread_index)
""", binds)
connection.commit()
# ============================================================
# PIPELINE
# ============================================================
def process_emails():
files = list_files()
for name in files:
print(f"\n=== Processing: {name} ===")
path = download_file(name)
data = extract_msg_text(path)
top_headers = data["top"]
body = data["body"]
blocks = split_email_thread_refined(body)
last_booking_detected = None
thread_counter = 0
for block in blocks:
print("-------------------------------------")
print("headers: ", top_headers)
print("block: ", block)
thread_counter += 1
result = call_llm_extract(block, top_headers)
if isinstance(result, dict) and "error" not in result:
current_booking = (result.get("booking") or "").strip()
if current_booking and current_booking != last_booking_detected:
print(f"[INFO] New booking detected: {current_booking}")
last_booking_detected = current_booking
elif not current_booking and last_booking_detected:
result["booking"] = last_booking_detected
insert_into_database(result,
file_name=os.path.basename(name),
thread_index=thread_counter)
print(f"[OK] Thread {thread_counter}: {result}")
else:
print(f"[WARNING] Error in block {thread_counter}: {result}")
if last_booking_detected:
with connection.cursor() as cur:
cur.execute("""
UPDATE PROCESSED_EMAILS
SET BOOKING = :booking
WHERE FILE_NAME = :file_name
AND (BOOKING IS NULL OR TRIM(BOOKING) = '')
""", {"booking": last_booking_detected, "file_name": os.path.basename(name)})
connection.commit()
print(f"[UPDATE] Booking '{last_booking_detected}' propagated in {os.path.basename(name)}")
move_to_processed(name)
if __name__ == "__main__":
process_emails()

77
files/redaction_23ai.sql Normal file
View File

@@ -0,0 +1,77 @@
CREATE TABLE PROCESSED_EMAILS (
ID NUMBER GENERATED ALWAYS AS IDENTITY,
DATE_SENT TIMESTAMP,
BOOKING VARCHAR2(100),
EMAIL_FROM VARCHAR2(500),
EMAIL_TO VARCHAR2(1000),
SUBJECT VARCHAR2(1000),
STATUS VARCHAR2(100),
INSERT_DATE TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
PRIMARY KEY (ID)
);
ALTER TABLE PROCESSED_EMAILS ADD (
BRIEF_DESCRIPTION VARCHAR2(1000),
ALERT VARCHAR2(10),
THREAD_INDEX NUMERIC(10),
FILE_NAME VARCHAR(400)
);
CREATE OR REPLACE VIEW PROCESSED_EMAILS_REDACTED AS
SELECT
----------------------------------------------------------------------
-- EMAIL_TO: remove names, parentheses, and mask all email addresses
----------------------------------------------------------------------
REGEXP_REPLACE(
REGEXP_REPLACE(
REGEXP_REPLACE(
-- 1️⃣ Remove names and content inside parentheses before the <email>
EMAIL_TO,
'([[:alpha:]][[:alpha:] .''()_-]*)<([^>]+)>',
'<\2>'
),
-- 2️⃣ Mask email addresses
'([A-Za-z])([A-Za-z0-9._%+-]*)@([A-Za-z])([A-Za-z0-9.-]*)(\.[A-Za-z]{2,}(?:\.[A-Za-z]{2,})*)',
'\1******@\3******\5',
1, 0, 'i'
),
-- 3️⃣ Normalize ";" separators
'\s*;\s*', '; '
) AS EMAIL_TO,
----------------------------------------------------------------------
-- EMAIL_FROM: same treatment (remove names, parentheses, mask)
----------------------------------------------------------------------
REGEXP_REPLACE(
REGEXP_REPLACE(
REGEXP_REPLACE(
EMAIL_FROM,
'([[:alpha:]][[:alpha:] .''()_-]*)<([^>]+)>',
'<\2>'
),
'([A-Za-z])([A-Za-z0-9._%+-]*)@([A-Za-z])([A-Za-z0-9.-]*)(\.[A-Za-z]{2,}(?:\.[A-Za-z]{2,})*)',
'\1******@\3******\5',
1, 0, 'i'
),
'\s*;\s*', '; '
) AS EMAIL_FROM,
SUBJECT,
----------------------------------------------------------------------
-- Mask BOOKING (for example, a code or sensitive text)
----------------------------------------------------------------------
REGEXP_REPLACE(
BOOKING,
'([A-Za-z0-9])([A-Za-z0-9 _.-]*)',
'***'
) AS BOOKING,
DATE_SENT,
STATUS,
FILE_NAME,
THREAD_INDEX,
INSERT_DATE,
BRIEF_DESCRIPTION,
ALERT
FROM PROCESSED_EMAILS;

3
files/requirements.txt Normal file
View File

@@ -0,0 +1,3 @@
oci
extract-msg
oracledb

208
files/templates/emails_filter.html Normal file
View File

@@ -0,0 +1,208 @@
<!DOCTYPE html>
<link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.5.0/css/all.min.css">
<html lang="en">
<head>
<meta charset="UTF-8">
<title>Email Query - International Logistics</title>
<style>
body {
font-family: Arial, sans-serif;
background: #f3f3f3;
margin: 20px;
}
.container {
max-width: 1100px;
margin: auto;
background: white;
border-radius: 10px;
padding: 25px;
box-shadow: 0 0 8px rgba(0,0,0,0.15);
}
form {
margin-bottom: 25px;
padding: 10px;
background: #fafafa;
border-radius: 8px;
}
form input, form select {
padding: 6px 10px;
margin-right: 10px;
}
button {
padding: 6px 12px;
}
.email {
border-left: 5px solid #ccc;
padding: 12px 16px;
margin-bottom: 15px;
border-radius: 8px;
background: #fff;
}
.email.alert {
border-left-color: #d32f2f;
background: #ffeaea;
}
.email h4 {
margin: 0;
}
.email small {
color: #555;
}
.brief {
margin-top: 6px;
font-style: italic;
color: #333;
}
.badge {
display: inline-block;
padding: 2px 6px;
border-radius: 4px;
font-size: 12px;
color: white;
}
.badge.alert { background: #d32f2f; }
.badge.ok { background: #388e3c; }
.badge.pending { background: #fbc02d; color: black; }
.badge.analysis { background: #1976d2; }
.filters-container {
display: flex;
flex-wrap: wrap;
align-items: flex-end;
gap: 14px 20px;
background: #ffffff;
padding: 15px 20px;
border-radius: 10px;
box-shadow: 0 2px 6px rgba(0,0,0,0.1);
margin-bottom: 25px;
}
.filters-container label {
font-weight: 600;
color: #004b8d;
display: block;
margin-bottom: 3px;
}
.filters-container input,
.filters-container select {
padding: 6px 8px;
border-radius: 6px;
border: 1px solid #ccc;
font-size: 0.95em;
min-width: 100px;
}
.filters-container button {
background: #007bff;
color: white;
font-weight: 600;
padding: 8px 14px;
border-radius: 6px;
border: none;
cursor: pointer;
transition: background 0.2s;
}
.filters-container button:hover {
background: #005ec5;
}
@media (min-width: 1200px) {
.filters-container {
flex-wrap: nowrap;
justify-content: space-between;
}
}
.badge.forwarded {
background: #bdbdbd;
color: #000;
}
a.icon-link i {
transition: color 0.2s ease;
}
a.icon-link:hover i {
color: #005ec5;
}
</style>
</head>
<body>
<div class="container">
<h2>Email Search</h2>
<form class="filters-container" method="get">
<div>
<label>Booking:</label>
<input type="text" name="booking" value="{{ filters.booking }}" style="width: 120px;">
</div>
<div>
<label>From:</label>
<input type="date" name="start" value="{{ filters.start }}">
</div>
<div>
<label>To:</label>
<input type="date" name="end" value="{{ filters.end }}">
</div>
<div>
<label>Attention:</label>
<select name="alert" style="width: 100px;">
<option value="">All</option>
<option value="YES" {% if filters.alert == 'YES' %}selected{% endif %}>Yes</option>
<option value="NO" {% if filters.alert == 'NO' %}selected{% endif %}>No</option>
</select>
</div>
<div>
<label>Status:</label>
<select name="status" style="width: 180px;">
<option value="">All</option>
{% for s in status_options %}
<option value="{{ s }}" {% if filters.status == s %}selected{% endif %}>{{ s }}</option>
{% endfor %}
</select>
</div>
<div>
<label>
<input type="checkbox" name="include_forwarded" {% if filters.include_forwarded %}checked{% endif %}>
Include Forwarded Emails
</label>
</div>
<div>
<button type="submit">🔍 Search</button>
</div>
</form>
{% if records|length == 0 %}
<p><i>No results found.</i></p>
{% endif %}
{% for e in records %}
<div class="email {% if e.ALERT == 'YES' %}alert{% endif %}">
<h4>{{ e.SUBJECT }}</h4>
<small>
<b>Date:</b> {{ e.DATE_SENT }} |
<b>From:</b> {{ e.EMAIL_FROM }} |
<b>To:</b> {{ e.EMAIL_TO }}
</small><br>
<span class="badge {% if e.ALERT == 'YES' %}alert{% elif e.STATUS == 'completed' %}ok{% elif e.STATUS == 'pending' %}pending{% else %}analysis{% endif %}">
{{ e.STATUS }}
</span>
{% if e.ALERT == 'YES' %}
<span style="color:#d32f2f;font-weight:bold;margin-left:6px;">⚠️ Alert</span>
{% endif %}
{% if e.BRIEF_DESCRIPTION %}
<div class="brief">📝 {{ e.BRIEF_DESCRIPTION }}</div>
{% endif %}
<td>
<a href="{{ url_for('view_email', file_name=e.FILE_NAME|urlencode) }}"
class="icon-link"
target="_blank"
title="View original email">
<i class="fa-solid fa-envelope-open-text"></i>
</a>
</td>
</div>
{% endfor %}
</div>
</body>
</html>

BIN
images/img.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 416 KiB