Update app.py
Browse files
app.py
CHANGED
|
@@ -34,9 +34,32 @@ try:
|
|
| 34 |
except ImportError:
|
| 35 |
import pytz # Fallback if zoneinfo is missing
|
| 36 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 37 |
# --- CACHING HELPERS ---
|
| 38 |
DRAFT_FILE = "user_drafts.json"
|
| 39 |
|
|
|
|
| 40 |
def load_drafts():
|
| 41 |
if os.path.exists(DRAFT_FILE):
|
| 42 |
try:
|
|
@@ -70,6 +93,7 @@ def update_cache_row(user, session_id, dom_a, pol_a, dom_b, pol_b, tar_col, ctx_
|
|
| 70 |
drafts[user]["rows"][cache_key][b_text] = {
|
| 71 |
"rel": rel, "inter": inter, "just": just, "session_id": session_id
|
| 72 |
}
|
|
|
|
| 73 |
|
| 74 |
with open(DRAFT_FILE, 'w') as f:
|
| 75 |
json.dump(drafts, f)
|
|
@@ -775,26 +799,41 @@ with gr.Blocks(theme=gr.themes.Soft(), css=custom_css) as demo:
|
|
| 775 |
|
| 776 |
# ADD link_val and follow_val to inputs
|
| 777 |
def authenticate(email, link_val, follow_val):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 778 |
user_tag, msg = get_or_create_user(email)
|
| 779 |
|
| 780 |
if not user_tag:
|
|
|
|
| 781 |
return (gr.update(value=f"<font color='red'>{msg}</font>"),
|
| 782 |
-
gr.update(visible=True), gr.update(visible=False), gr.update(visible=False),
|
| 783 |
None, None,
|
| 784 |
gr.update(), gr.update(), gr.update(), gr.update(), gr.update(), gr.update(), link_val, follow_val)
|
| 785 |
|
|
|
|
| 786 |
hf_df = load_hf_dataset()
|
| 787 |
drafts = load_drafts()
|
| 788 |
user_data = drafts.get(user_tag, {})
|
| 789 |
ws = user_data.get("workspace", {})
|
| 790 |
|
|
|
|
| 791 |
if ws.get("pol_a") and ws.get("pol_b"):
|
| 792 |
-
|
|
|
|
| 793 |
return (
|
| 794 |
gr.update(value=f"{msg} Loaded {len(hf_df)} annotations."),
|
| 795 |
-
gr.update(visible=False), # login_box
|
| 796 |
-
gr.update(visible=False), # sector_box (
|
| 797 |
-
gr.update(visible=True), # app_box
|
| 798 |
user_tag,
|
| 799 |
hf_df,
|
| 800 |
gr.update(value=ws["dom_a"]),
|
|
@@ -807,11 +846,12 @@ with gr.Blocks(theme=gr.themes.Soft(), css=custom_css) as demo:
|
|
| 807 |
follow_val
|
| 808 |
)
|
| 809 |
else:
|
|
|
|
| 810 |
return (
|
| 811 |
gr.update(value=f"{msg} Loaded {len(hf_df)} annotations."),
|
| 812 |
-
gr.update(visible=False), # login_box
|
| 813 |
-
gr.update(visible=True), #
|
| 814 |
-
gr.update(visible=False), #
|
| 815 |
user_tag,
|
| 816 |
hf_df,
|
| 817 |
gr.update(), gr.update(), gr.update(), gr.update(), gr.update(), gr.update(),
|
|
@@ -838,6 +878,7 @@ with gr.Blocks(theme=gr.themes.Soft(), css=custom_css) as demo:
|
|
| 838 |
default_domain = allowed_list[0] if allowed_list else None
|
| 839 |
available_policies = get_policy_list(default_domain) if default_domain else []
|
| 840 |
|
|
|
|
| 841 |
return (
|
| 842 |
gr.update(visible=False), # Hide sector box
|
| 843 |
gr.update(visible=True), # Show main app
|
|
@@ -1064,6 +1105,7 @@ with gr.Blocks(theme=gr.themes.Soft(), css=custom_css) as demo:
|
|
| 1064 |
|
| 1065 |
status_msg = t_text(f"Data loaded. {total_missing_pairs} unannotated pairs remain across {len(target_a_list)} Target A groups.", lang)
|
| 1066 |
|
|
|
|
| 1067 |
return [
|
| 1068 |
target_a_list, pending_tasks, 0,
|
| 1069 |
ctx_a_chunk_eng, ctx_b_chunk_eng, b_eng_list,
|
|
@@ -1151,6 +1193,7 @@ with gr.Blocks(theme=gr.themes.Soft(), css=custom_css) as demo:
|
|
| 1151 |
hf_df.to_csv(csv_buffer, index=False)
|
| 1152 |
csv_bytes = csv_buffer.getvalue().encode('utf-8')
|
| 1153 |
|
|
|
|
| 1154 |
api = HfApi()
|
| 1155 |
api.upload_file(
|
| 1156 |
path_or_fileobj=io.BytesIO(csv_bytes), path_in_repo=HF_CSV_FILE,
|
|
@@ -1176,6 +1219,7 @@ with gr.Blocks(theme=gr.themes.Soft(), css=custom_css) as demo:
|
|
| 1176 |
return gr.update(value=log_msg), idx + 1, hf_df
|
| 1177 |
|
| 1178 |
def skip_action(idx, lang):
|
|
|
|
| 1179 |
return gr.update(value=t_text(f"Skipped group {idx + 1}.", lang)), idx + 1
|
| 1180 |
|
| 1181 |
# --- TRIGGER FIRST PASS ---
|
|
|
|
| 34 |
except ImportError:
|
| 35 |
import pytz # Fallback if zoneinfo is missing
|
| 36 |
|
| 37 |
+
# --- COMPREHENSIVE LOGGING ---
|
| 38 |
+
LOG_FILE = "logs.txt"
|
| 39 |
+
|
| 40 |
+
def write_log(action_type, details):
|
| 41 |
+
"""Appends a timestamped log entry to logs.txt"""
|
| 42 |
+
try:
|
| 43 |
+
try:
|
| 44 |
+
tz = ZoneInfo("Africa/Nairobi")
|
| 45 |
+
except:
|
| 46 |
+
import pytz
|
| 47 |
+
tz = pytz.timezone("Africa/Nairobi")
|
| 48 |
+
|
| 49 |
+
timestamp = datetime.now(tz).strftime('%Y-%m-%d %H:%M:%S')
|
| 50 |
+
log_entry = f"[{timestamp}] [{action_type}] {details}\n"
|
| 51 |
+
|
| 52 |
+
with open(LOG_FILE, "a", encoding="utf-8") as f:
|
| 53 |
+
f.write(log_entry)
|
| 54 |
+
|
| 55 |
+
print(log_entry.strip()) # Also print to console for debugging
|
| 56 |
+
except Exception as e:
|
| 57 |
+
print(f"Logging failed: {e}")
|
| 58 |
+
|
| 59 |
# --- CACHING HELPERS ---
|
| 60 |
DRAFT_FILE = "user_drafts.json"
|
| 61 |
|
| 62 |
+
|
| 63 |
def load_drafts():
|
| 64 |
if os.path.exists(DRAFT_FILE):
|
| 65 |
try:
|
|
|
|
| 93 |
drafts[user]["rows"][cache_key][b_text] = {
|
| 94 |
"rel": rel, "inter": inter, "just": just, "session_id": session_id
|
| 95 |
}
|
| 96 |
+
write_log("CACHE_UPDATE", f"User {user} auto-saved draft for row index {idx}.")
|
| 97 |
|
| 98 |
with open(DRAFT_FILE, 'w') as f:
|
| 99 |
json.dump(drafts, f)
|
|
|
|
| 799 |
|
| 800 |
# ADD link_val and follow_val to inputs
|
| 801 |
def authenticate(email, link_val, follow_val):
|
| 802 |
+
email = email.strip().lower()
|
| 803 |
+
|
| 804 |
+
# 1. Validate Email Format
|
| 805 |
+
email_pattern = r"^[^@\s]+@[^@\s]+\.[^@\s]+$"
|
| 806 |
+
if not re.match(email_pattern, email):
|
| 807 |
+
write_log("LOGIN_FAILED", f"Invalid email format attempted: '{email}'")
|
| 808 |
+
return (gr.update(value=f"<font color='red'>Please enter a valid email address.</font>"),
|
| 809 |
+
gr.update(visible=True), gr.update(visible=False), gr.update(visible=False),
|
| 810 |
+
None, None,
|
| 811 |
+
gr.update(), gr.update(), gr.update(), gr.update(), gr.update(), gr.update(), link_val, follow_val)
|
| 812 |
+
|
| 813 |
user_tag, msg = get_or_create_user(email)
|
| 814 |
|
| 815 |
if not user_tag:
|
| 816 |
+
write_log("LOGIN_FAILED", f"System error creating user for '{email}'")
|
| 817 |
return (gr.update(value=f"<font color='red'>{msg}</font>"),
|
| 818 |
+
gr.update(visible=True), gr.update(visible=False), gr.update(visible=False),
|
| 819 |
None, None,
|
| 820 |
gr.update(), gr.update(), gr.update(), gr.update(), gr.update(), gr.update(), link_val, follow_val)
|
| 821 |
|
| 822 |
+
write_log("LOGIN_SUCCESS", f"User {user_tag} ({email}) logged in. Consents - Link: {link_val}, Follow: {follow_val}")
|
| 823 |
hf_df = load_hf_dataset()
|
| 824 |
drafts = load_drafts()
|
| 825 |
user_data = drafts.get(user_tag, {})
|
| 826 |
ws = user_data.get("workspace", {})
|
| 827 |
|
| 828 |
+
# 2. Check for pending session and redirect straight to workspace
|
| 829 |
if ws.get("pol_a") and ws.get("pol_b"):
|
| 830 |
+
write_log("SESSION_RESTORED", f"User {user_tag} skipped sectors and restored previous session {ws.get('session_id')}.")
|
| 831 |
+
msg += f" Restored your pending session. Click 'Fetch Data' to resume your draft."
|
| 832 |
return (
|
| 833 |
gr.update(value=f"{msg} Loaded {len(hf_df)} annotations."),
|
| 834 |
+
gr.update(visible=False), # Hide login_box
|
| 835 |
+
gr.update(visible=False), # Hide sector_box (Bypass directly to app)
|
| 836 |
+
gr.update(visible=True), # Show app_box
|
| 837 |
user_tag,
|
| 838 |
hf_df,
|
| 839 |
gr.update(value=ws["dom_a"]),
|
|
|
|
| 846 |
follow_val
|
| 847 |
)
|
| 848 |
else:
|
| 849 |
+
write_log("NEW_SESSION", f"User {user_tag} starting fresh. Routing to Sector Selection.")
|
| 850 |
return (
|
| 851 |
gr.update(value=f"{msg} Loaded {len(hf_df)} annotations."),
|
| 852 |
+
gr.update(visible=False), # Hide login_box
|
| 853 |
+
gr.update(visible=True), # Show sector_box
|
| 854 |
+
gr.update(visible=False), # Hide app_box
|
| 855 |
user_tag,
|
| 856 |
hf_df,
|
| 857 |
gr.update(), gr.update(), gr.update(), gr.update(), gr.update(), gr.update(),
|
|
|
|
| 878 |
default_domain = allowed_list[0] if allowed_list else None
|
| 879 |
available_policies = get_policy_list(default_domain) if default_domain else []
|
| 880 |
|
| 881 |
+
write_log("SECTOR_SELECTED", f"Mapped sectors {selected_sectors} to domains {allowed_list}")
|
| 882 |
return (
|
| 883 |
gr.update(visible=False), # Hide sector box
|
| 884 |
gr.update(visible=True), # Show main app
|
|
|
|
| 1105 |
|
| 1106 |
status_msg = t_text(f"Data loaded. {total_missing_pairs} unannotated pairs remain across {len(target_a_list)} Target A groups.", lang)
|
| 1107 |
|
| 1108 |
+
write_log("WORKSPACE_LOADED", f"User {user_tag} fetched {pol_a} vs {pol_b}.")
|
| 1109 |
return [
|
| 1110 |
target_a_list, pending_tasks, 0,
|
| 1111 |
ctx_a_chunk_eng, ctx_b_chunk_eng, b_eng_list,
|
|
|
|
| 1193 |
hf_df.to_csv(csv_buffer, index=False)
|
| 1194 |
csv_bytes = csv_buffer.getvalue().encode('utf-8')
|
| 1195 |
|
| 1196 |
+
write_log("DATA_SAVED", f"User {user_tag} successfully saved {len(new_rows)} completed rows to Hugging Face.")
|
| 1197 |
api = HfApi()
|
| 1198 |
api.upload_file(
|
| 1199 |
path_or_fileobj=io.BytesIO(csv_bytes), path_in_repo=HF_CSV_FILE,
|
|
|
|
| 1219 |
return gr.update(value=log_msg), idx + 1, hf_df
|
| 1220 |
|
| 1221 |
def skip_action(idx, lang):
|
| 1222 |
+
write_log("TARGET_SKIPPED", f"User skipped group {idx + 1}")
|
| 1223 |
return gr.update(value=t_text(f"Skipped group {idx + 1}.", lang)), idx + 1
|
| 1224 |
|
| 1225 |
# --- TRIGGER FIRST PASS ---
|