Merge pull request #3 from claromes/restart-flow

Fix new tracking flow
claromes · Apr 29, 2023 · 2bf15cb · 2bf15cb
2 parents ce52bf5 + da4432b
commit 2bf15cb
Show file tree

Hide file tree

Showing 9 changed files with 122 additions and 132 deletions.
diff --git a/README.md b/README.md
@@ -34,48 +34,14 @@ Streamlit will be served at http://localhost:8501
 
 <br>
 
-> Initial screen: credentials, sign in and channel inputs
 <br>
 <p align="center">
     <img src="assets/1.png" width="700">
 </p>
 
-<br>
-
-> Data screens: tabs and download links
-<br>
-<p align="center">
-    <img src="assets/2.png" width="700">
-</p>
-<p align="center">
-    <img src="assets/3.png" width="700">
-</p>
-
-<p align="center">
-    <img src="assets/4.png" width="700">
-</p>
-<p align="center">
-    <img src="assets/5.png" width="700">
-</p>
-
-**network tab is under development*
-<br>
-
-<p align="center">
-    <img src="assets/6.png" width="700">
-</p>
-
-<br>
-
-> Restart tab: to track another channel
-<br>
-<p align="center">
-    <img src="assets/6.png" width="700">
-</p>
-
 ### Workflow
 
-*IMPORTANT: To test using the deploy link with your personal credentials disable the 2FA*
+*IMPORTANT: disable the 2FA*
 
 1. Create your API credentials [here](https://my.telegram.org/auth)
 
@@ -102,11 +68,12 @@ Streamlit will be served at http://localhost:8501
 
 5. Switch tabs to preview or download the data
 
-6. To track another channel, switch to last tab (`trac`) and click `restart`.
+6. To track another channel, switch to last tab (`trac`) and click `new trac`.
+
+7. Fill the input `channel name` (*copy name from channel link: t.me/CHANNEL_NAME*) and click on `trac` button
 
-- To restart, send the same credentials and `code`
-- A late message could be send to your Telegram app about other authentications
 - At each tracking the dataset are grouped in the same file to allow network analysis
+- A late message could be send to your Telegram app about other authentications
 
 ## Additional Information
 
@@ -122,8 +89,11 @@ Mostly limited to Streamlit options. The form and tabs were chosen due to common
 ## Roadmap
 
 - [x] Fix dataset tab
-- [ ] Fix set credentials and code in restart flow
+- [x] Fix set credentials and code in restart flow
 - [ ] Network tab
+- [ ] Error msgs
+- [ ] Storage limit alerts
+- [x] Delete files after session finish
 - [ ] Allow 2FA
 - [ ] Add batch file upload
 - [ ] Option without API credentials
@@ -140,8 +110,4 @@ Mostly limited to Streamlit options. The form and tabs were chosen due to common
     - Fix dataset tab
     - Fix imports
 - [v0.1.0](https://github.com/claromes/telegramtrac/releases/tag/v0.1.0)
-    - Bellingcat Accessiblilty Hackathon submission
-
-## Telegram App Bans in Brazil
-
-Due to the blocking of Telegram in Brazil ([NYT article](https://www.nytimes.com/2023/04/26/briefing/brazil-telegram-ban.html)), which has already occurred on other occasions, I decided to expand the development and include in the roadmap an option without API credentials
+    - Bellingcat Accessiblilty Hackathon submission
diff --git a/assets/2.png b/assets/2.png
diff --git a/assets/3.png b/assets/3.png
diff --git a/assets/4.png b/assets/4.png
diff --git a/assets/5.png b/assets/5.png
diff --git a/assets/6.png b/assets/6.png
diff --git a/config/config.ini b/config/config.ini
@@ -1,5 +1,5 @@
 [Telegram API credentials]
-api_id =
-api_hash =
-phone =
+api_id = 
+api_hash = 
+phone = 
 
diff --git a/config/config_sign_in_code.ini b/config/config_sign_in_code.ini
@@ -1,3 +1,3 @@
 [Sign in code]
-code =
+code = 
 
diff --git a/streamlit/telegramtrac.py b/streamlit/telegramtrac.py
@@ -6,6 +6,7 @@
 from pandas import read_csv
 import base64
 import os
+import shutil
 
 #page config
 st.set_page_config(
@@ -31,10 +32,11 @@
 form_component = st.empty()
 sign_in_component = st.empty()
 channel_component = st.empty()
-channel_component_new = st.empty()
+form_component_channel = st.empty()
 center_running()
 
 trac = ''
+new_trac = ''
 
 #states
 if 'channel_name' not in st.session_state:
@@ -44,101 +46,135 @@
     st.session_state['code_state'] = False
 
 if 'code_value' not in st.session_state:
-    st.session_state['code_value'] = 0
+    st.session_state['code_value'] = ''
+
+if 'api_id' not in st.session_state:
+    st.session_state['api_id'] = ''
+
+if 'api_hash' not in st.session_state:
+    st.session_state['api_hash'] = ''
+
+if 'phone' not in st.session_state:
+    st.session_state['phone'] = ''
+
+if 'restart' not in st.session_state:
+    st.session_state['restart'] = False
 
 #title
 title_component.title('telegramtrac', help="not stable", anchor=False)
 
-#credentials
-with form_component.form(key='config_form'):
-    api_id = st.text_input('api_id')
-    api_hash = st.text_input('api_hash')
-    phone = st.text_input('phone')
+if not st.session_state.restart:
+    #delete all files and directories before start another tracking
+    dir_path_output = os.path.join('output')
 
-    config = {
-        'api_id': api_id,
-        'api_hash': api_hash,
-        'phone': phone
-    }
+    if os.path.exists(dir_path_output):
+        shutil.rmtree(dir_path_output)
 
-    config_parser = configparser.ConfigParser()
-    config_parser['Telegram API credentials'] = config
-    with open('config/config.ini', 'w') as file:
-        config_parser.write(file)
+    #credentials
+    with form_component.form(key='config_form'):
+        api_id = st.text_input('api_id')
+        api_hash = st.text_input('api_hash')
+        phone = st.text_input('phone')
 
-    send_credentials = st.form_submit_button('send credentials', type='primary')
+        config = {
+            'api_id': api_id,
+            'api_hash': api_hash,
+            'phone': phone
+        }
 
-if send_credentials and api_id != '' and api_hash != '' and phone != '':
-    center_running()
-    st.session_state.code_state = True
+        config_parser = configparser.ConfigParser()
+        config_parser['Telegram API credentials'] = config
+        with open('config/config.ini', 'w') as file:
+            config_parser.write(file)
 
-    try:
-        cmd_tele = "pip install telethon --user"
-        output = subprocess.check_output(cmd_tele.split())
+        send_credentials = st.form_submit_button('send credentials', type='primary')
 
-        cmd_pd = "pip install pandas --user"
-        output = subprocess.check_output(cmd_pd.split())
+    if send_credentials and api_id != '' and api_hash != '' and phone != '':
+        center_running()
+        st.session_state.code_state = True
 
-        cmd_tqdm = "pip install tqdm --user"
-        output = subprocess.check_output(cmd_tqdm.split())
+        try:
+            # cmd_tele = "pip install telethon --user"
+            # output = subprocess.check_output(cmd_tele.split())
 
-        cmd_open = "pip install openpyxl --user"
-        output = subprocess.check_output(cmd_open.split())
+            # cmd_pd = "pip install pandas --user"
+            # output = subprocess.check_output(cmd_pd.split())
 
-        cmd_connect = 'python connect.py'
-        output = subprocess.check_output(cmd_connect.split())
-    except subprocess.CalledProcessError:
-        pass
-    except Exception:
-        pass
+            # cmd_tqdm = "pip install tqdm --user"
+            # output = subprocess.check_output(cmd_tqdm.split())
 
-#sign in code
-with sign_in_component.form(key='config_sign_in_form'):
-    if st.session_state.code_state:
-        code_sign_in = st.text_input('code', disabled=False)
-    else:
-        code_sign_in = st.text_input('code', disabled=True)
+            # cmd_open = "pip install openpyxl --user"
+            # output = subprocess.check_output(cmd_open.split())
 
-    if code_sign_in == '':
-        code_sign_in = st.session_state.code_value
+            cmd_connect = 'python connect.py'
+            output = subprocess.check_output(cmd_connect.split())
+        except subprocess.CalledProcessError:
+            pass
+        except Exception:
+            pass
 
-    config_sign_in_code = {
-        'code': code_sign_in
-    }
+    #sign in code
+    with sign_in_component.form(key='config_sign_in_form'):
+        if st.session_state.code_state:
+            code_sign_in = st.text_input('code', disabled=False)
+        else:
+            code_sign_in = st.text_input('code', disabled=True)
 
-    config_sign_in_code_parser = configparser.ConfigParser()
-    config_sign_in_code_parser['Sign in code'] = config_sign_in_code
-    with open('config/config_sign_in_code.ini', 'w') as file:
-        config_sign_in_code_parser.write(file)
+        if code_sign_in == '':
+            code_sign_in = st.session_state.code_value
 
-    if st.session_state.code_state:
-        sign_in = st.form_submit_button('sign in', disabled=False, type='primary')
-    else:
-        sign_in = st.form_submit_button('sign in', disabled=True, type='primary')
+        config_sign_in_code = {
+            'code': code_sign_in
+        }
 
-    if sign_in:
-        center_running()
-        st.session_state.code_value = code_sign_in
-
-#channel name
-with channel_component.form(key='channel_form'):
-    if sign_in and not st.session_state.code_value == 0 or st.session_state.channel_name != '':
-        channel_name = st.text_input('channel name', placeholder="https://t.me/CHANNEL_NAME_IS_HERE", disabled=False, key='channel_name')
-        trac = st.form_submit_button('trac', disabled=False, type='primary')
-        send_credentials = True
-    else:
-        channel_name = st.text_input('channel name', disabled=True)
-        trac = st.form_submit_button('trac', disabled=True, type='primary')
-        send_credentials = True
+        config_sign_in_code_parser = configparser.ConfigParser()
+        config_sign_in_code_parser['Sign in code'] = config_sign_in_code
+        with open('config/config_sign_in_code.ini', 'w') as file:
+            config_sign_in_code_parser.write(file)
+
+        if st.session_state.code_state:
+            sign_in = st.form_submit_button('sign in', disabled=False, type='primary')
+        else:
+            sign_in = st.form_submit_button('sign in', disabled=True, type='primary')
+
+        if sign_in:
+            center_running()
+            st.session_state.code_value = code_sign_in
+
+    #channel name
+    with channel_component.form(key='channel_form'):
+        if sign_in and not st.session_state.code_value == 0 or st.session_state.channel_name != '':
+            channel_name = st.text_input('channel name', placeholder="https://t.me/CHANNEL_NAME_IS_HERE", disabled=False, key='channel_name')
+            trac = st.form_submit_button('trac', disabled=False, type='primary')
+            send_credentials = True
+        else:
+            channel_name = st.text_input('channel name', disabled=True)
+            trac = st.form_submit_button('trac', disabled=True, type='primary')
+            send_credentials = True
+else:
+    form_component.empty()
+    sign_in_component.empty()
+    channel_component.empty()
+
+    with form_component_channel.form(key='config_form_channel'):
+        #send same credentials and code
+        api_id = st.session_state.api_id
+        api_hash = st.session_state.api_hash
+        phone = st.session_state.phone
+        code = st.session_state.code_value
+
+        st.session_state.channel_name = st.text_input('channel name', placeholder="https://t.me/CHANNEL_NAME_IS_HERE", disabled=False, key='channel_name_new_trac')
+        new_trac = st.form_submit_button('new trac', disabled=False, type='primary')
 
 #data tabs
-if trac and st.session_state.channel_name != '':
+if trac or new_trac and st.session_state.channel_name != '':
     center_running()
     tab1, tab2, tab3, tab4, tab5 = st.tabs(['messages', 'metadata', 'dataset', 'network', 'new trac'])
 
     form_component.empty()
     sign_in_component.empty()
     channel_component.empty()
+    form_component_channel.empty()
 
     try:
         cmd_main = 'python main.py --telegram-channel {}'.format(st.session_state.channel_name)
@@ -160,15 +196,6 @@
     subdirectories = [entry for entry in os.scandir(path) if entry.is_dir()]
     path_lens = len(subdirectories) > 1
 
-    # if path_lens:
-    #     try:
-    #         cmd_dataset = 'python channels-to-network.py'
-    #         output = subprocess.check_output(cmd_dataset.split())
-    #     except subprocess.CalledProcessError:
-    #         pass
-    #     except Exception:
-    #         pass
-
     #json - main file
     with tab1:
         json_file = 'output/data/{}/{}_messages.json'.format(st.session_state.channel_name, st.session_state.channel_name)
@@ -228,7 +255,6 @@
     #dataset
     with tab3:
         st.subheader('dataset', anchor=False)
-        #st.caption('Channels: {}'.format(st.session_state.channel_name))
         dataset_csv_file = 'output/data/msgs_dataset.csv'
 
         with open(dataset_csv_file, 'rb') as file:
@@ -241,16 +267,14 @@
 
             st.dataframe(df)
 
+    #network
     with tab4:
         if path_lens:
             st.info('Under development')
         else:
             st.info('There is only one channel.')
-            #st.info('There is only one channel.')
 
     #restart
     with tab5:
-        restart = st.button('restart', type='primary')
-
-        if restart:
-            st.experimental_rerun()
+       st.button('new trac', type='primary')
+       st.session_state.restart = True