pearmut 0.2.5__py3-none-any.whl → 0.2.7__py3-none-any.whl

pearmut/app.py

@@ -30,7 +30,8 @@ app.add_middleware(
 
 tasks_data = {}
 progress_data = load_progress_data(
-    warn="No progress.json found. Running, but no campaign will be available.")
+    warn="No progress.json found. Running, but no campaign will be available."
+)
 
 # load all tasks into data_all
 for campaign_id in progress_data.keys():
@@ -60,30 +61,31 @@ async def _log_response(request: LogResponseRequest):
 
     # append response to the output log
     save_db_payload(
-        campaign_id, request.payload | {"user_id": user_id, "item_i": item_i})
+        campaign_id, request.payload | {"user_id": user_id, "item_i": item_i}
+    )
 
     # if actions were submitted, we can log time data
     if "actions" in request.payload:
-        times = [
-            x["time"] for x in request.payload["actions"]
-        ]
+        times = [x["time"] for x in request.payload["actions"]]
         if progress_data[campaign_id][user_id]["time_start"] is None:
             progress_data[campaign_id][user_id]["time_start"] = min(times)
         progress_data[campaign_id][user_id]["time_end"] = max(times)
-        progress_data[campaign_id][user_id]["time"] += sum([
-            min(b - a, 60)
-            for a, b in zip(times, times[1:])
-        ])
+        progress_data[campaign_id][user_id]["time"] += sum(
+            [min(b - a, 60) for a, b in zip(times, times[1:])]
+        )
 
     # Initialize validation_checks if it doesn't exist
     if "validations" in request.payload:
         if "validations" not in progress_data[campaign_id][user_id]:
             progress_data[campaign_id][user_id]["validations"] = {}
 
-        progress_data[campaign_id][user_id]["validations"][request.item_i] = request.payload["validations"]
+        progress_data[campaign_id][user_id]["validations"][request.item_i] = (
+            request.payload["validations"]
+        )
 
-    update_progress(campaign_id, user_id, tasks_data,
-                    progress_data, request.item_i, request.payload)
+    update_progress(
+        campaign_id, user_id, tasks_data, progress_data, request.item_i, request.payload
+    )
     save_progress_data(progress_data)
 
     return JSONResponse(content="ok", status_code=200)
@@ -149,13 +151,15 @@ async def _dashboard_data(request: DashboardDataRequest):
 
     if campaign_id not in progress_data:
         return JSONResponse(content="Unknown campaign ID", status_code=400)
-    
-    is_privileged = (request.token == tasks_data[campaign_id]["token"])
+
+    is_privileged = request.token == tasks_data[campaign_id]["token"]
 
     progress_new = {}
     assignment = tasks_data[campaign_id]["info"]["assignment"]
     if assignment not in ["task-based", "single-stream"]:
-        return JSONResponse(content="Unsupported campaign assignment type", status_code=400)
+        return JSONResponse(
+            content="Unsupported campaign assignment type", status_code=400
+        )
 
     # Get threshold info for the campaign
     validation_threshold = tasks_data[campaign_id]["info"].get("validation_threshold")
@@ -164,10 +168,9 @@ async def _dashboard_data(request: DashboardDataRequest):
         # shallow copy
         entry = dict(user_val)
         entry["validations"] = [
-            all(v)
-            for v in list(entry.get("validations", {}).values())
+            all(v) for v in list(entry.get("validations", {}).values())
         ]
-        
+
         # Add threshold pass/fail status (only when user is complete)
         if all(entry["progress"]):
             entry["threshold_passed"] = check_validation_threshold(
@@ -183,11 +186,8 @@ async def _dashboard_data(request: DashboardDataRequest):
         progress_new[user_id] = entry
 
     return JSONResponse(
-        content={
-            "data": progress_new,
-            "validation_threshold": validation_threshold
-        },
-        status_code=200
+        content={"data": progress_new, "validation_threshold": validation_threshold},
+        status_code=200,
     )
 
 
@@ -227,7 +227,9 @@ async def _download_annotations(
     for campaign_id in campaign_id:
         output_path = f"{ROOT}/data/outputs/{campaign_id}.jsonl"
         if campaign_id not in progress_data:
-            return JSONResponse(content=f"Unknown campaign ID {campaign_id}", status_code=400)
+            return JSONResponse(
+                content=f"Unknown campaign ID {campaign_id}", status_code=400
+            )
         if not os.path.exists(output_path):
             output[campaign_id] = []
         else:
@@ -239,28 +241,33 @@ async def _download_annotations(
 
 @app.get("/download-progress")
 async def _download_progress(
-    campaign_id: list[str] = Query(),
-    token: list[str] = Query()
+    campaign_id: list[str] = Query(), token: list[str] = Query()
 ):
 
     if len(campaign_id) != len(token):
-        return JSONResponse(content="Mismatched campaign_id and token count", status_code=400)
+        return JSONResponse(
+            content="Mismatched campaign_id and token count", status_code=400
+        )
 
     output = {}
     for i, cid in enumerate(campaign_id):
         if cid not in progress_data:
             return JSONResponse(content=f"Unknown campaign ID {cid}", status_code=400)
         if token[i] != tasks_data[cid]["token"]:
-            return JSONResponse(content=f"Invalid token for campaign ID {cid}", status_code=400)
+            return JSONResponse(
+                content=f"Invalid token for campaign ID {cid}", status_code=400
+            )
 
         output[cid] = progress_data[cid]
 
     return JSONResponse(content=output, status_code=200)
 
+
 static_dir = f"{os.path.dirname(os.path.abspath(__file__))}/static/"
 if not os.path.exists(static_dir + "index.html"):
     raise FileNotFoundError(
-        "Static directory not found. Please build the frontend first.")
+        "Static directory not found. Please build the frontend first."
+    )
 
 app.mount(
     "/",

pearmut/cli.py

@@ -12,6 +12,9 @@ import psutil
 
 from .utils import ROOT, load_progress_data, save_progress_data
 
+# Static directory path (constant for consistency)
+STATIC_DIR = os.path.join(os.path.dirname(os.path.abspath(__file__)), "static")
+
 os.makedirs(f"{ROOT}/data/tasks", exist_ok=True)
 load_progress_data(warn=None)
 
@@ -50,6 +53,40 @@ def _run(args_unknown):
     )
 
 
+def _validate_item_structure(items, template):
+    """
+    Validate that items have the correct structure.
+    Items should be lists of dictionaries with 'src' and 'tgt' keys.
+    
+    Args:
+        items: List of item dictionaries to validate
+        template: Template type ('pointwise' or 'listwise') for type validation
+    """
+    if not isinstance(items, list):
+        raise ValueError("Items must be a list")
+
+    for item in items:
+        if not isinstance(item, dict):
+            raise ValueError("Each item must be a dictionary with 'src' and 'tgt' keys")
+        if 'src' not in item or 'tgt' not in item:
+            raise ValueError("Each item must contain 'src' and 'tgt' keys")
+        
+        # Validate src is always a string
+        if not isinstance(item['src'], str):
+            raise ValueError("Item 'src' must be a string")
+        
+        # Validate tgt type based on template
+        if template == 'listwise':
+            if not isinstance(item['tgt'], list):
+                raise ValueError("Item 'tgt' must be a list for listwise template")
+            # Check that all elements in tgt list are strings
+            if not all(isinstance(t, str) for t in item['tgt']):
+                raise ValueError("All elements in 'tgt' list must be strings for listwise template")
+        elif template == 'pointwise':
+            if not isinstance(item['tgt'], str):
+                raise ValueError("Item 'tgt' must be a string for pointwise template")
+
+
 def _add_single_campaign(data_file, overwrite, server):
     """
     Add a single campaign from a JSON data file.
@@ -80,6 +117,7 @@ def _add_single_campaign(data_file, overwrite, server):
         raise ValueError("Campaign 'info' must contain 'template' field.")
 
     assignment = campaign_data["info"]["assignment"]
+    template = campaign_data["info"]["template"]
     # use random words for identifying users
     rng = random.Random(campaign_data["campaign_id"])
     rword = wonderwords.RandomWord(rng=rng)
@@ -96,6 +134,13 @@ def _add_single_campaign(data_file, overwrite, server):
         if not all(isinstance(task, list) for task in tasks):
             raise ValueError(
                 "Each task in task-based campaign 'data' must be a list of items.")
+        # Validate item structure for each task
+        for task_i, task in enumerate(tasks):
+            for doc_i, doc in enumerate(task):
+                try:
+                    _validate_item_structure(doc, template)
+                except ValueError as e:
+                    raise ValueError(f"Task {task_i}, document {doc_i}: {e}")
         num_users = len(tasks)
     elif assignment == "single-stream":
         tasks = campaign_data["data"]
@@ -105,6 +150,12 @@ def _add_single_campaign(data_file, overwrite, server):
         if not isinstance(campaign_data["data"], list):
             raise ValueError(
                 "Single-stream campaign 'data' must be a list of items.")
+        # Validate item structure for single-stream
+        for doc_i, doc in enumerate(tasks):
+            try:
+                _validate_item_structure(doc, template)
+            except ValueError as e:
+                raise ValueError(f"Document {doc_i}: {e}")
         if isinstance(users_spec, int):
             num_users = users_spec
         elif isinstance(users_spec, list):
@@ -199,30 +250,62 @@ def _add_single_campaign(data_file, overwrite, server):
 
     # Handle assets symlink if specified
     if "assets" in campaign_data["info"]:
-        assets_real_path = campaign_data["info"]["assets"]
+        assets_config = campaign_data["info"]["assets"]
+        
+        # assets must be a dictionary with source and destination keys
+        if not isinstance(assets_config, dict):
+            raise ValueError("Assets must be a dictionary with 'source' and 'destination' keys.")
+        if "source" not in assets_config or "destination" not in assets_config:
+            raise ValueError("Assets config must contain 'source' and 'destination' keys.")
+        
+        assets_source = assets_config["source"]
+        assets_destination = assets_config["destination"]
+        
+        # Validate destination starts with 'assets/'
+        if not assets_destination.startswith("assets/"):
+            raise ValueError(f"Assets destination '{assets_destination}' must start with 'assets/'.")
         
         # Resolve relative paths from the caller's current working directory
-        assets_real_path = os.path.abspath(assets_real_path)
+        assets_real_path = os.path.abspath(assets_source)
 
         if not os.path.isdir(assets_real_path):
-            raise ValueError(f"Assets path '{assets_real_path}' must be an existing directory.")
+            raise ValueError(f"Assets source path '{assets_real_path}' must be an existing directory.")
 
-        static_dir = f"{os.path.dirname(os.path.abspath(__file__))}/static"
-        dir_name = assets_real_path.split(os.sep)[-1]
-
-        if not os.path.isdir(static_dir):
+        if not os.path.isdir(STATIC_DIR):
             raise ValueError(
-                f"Static directory '{static_dir}' does not exist. "
+                f"Static directory '{STATIC_DIR}' does not exist. "
                 "Please build the frontend first."
             )
-        symlink_path = f"{static_dir}/assets/{dir_name}"
-
-        # Remove existing symlink if present and we are overriding
-        if os.path.exists(symlink_path):
+        
+        # Symlink path is based on the destination, stripping the 'assets/' prefix
+        symlink_path = f"{STATIC_DIR}/{assets_destination}"
+
+        # Remove existing symlink if present and we are overriding the same campaign
+        if os.path.lexists(symlink_path):
+            # Check if any other campaign is using this destination
+            current_campaign_id = campaign_data['campaign_id']
+            tasks_dir = f"{ROOT}/data/tasks"
+            if os.path.exists(tasks_dir):
+                for task_file in os.listdir(tasks_dir):
+                    if task_file.endswith('.json'):
+                        other_campaign_id = task_file[:-5]
+                        if other_campaign_id != current_campaign_id:
+                            with open(f"{tasks_dir}/{task_file}", "r") as f:
+                                other_campaign = json.load(f)
+                            other_assets = other_campaign.get("info", {}).get("assets")
+                            if other_assets and isinstance(other_assets, dict):
+                                if other_assets.get("destination") == assets_destination:
+                                    raise ValueError(
+                                        f"Assets destination '{assets_destination}' is already used by campaign '{other_campaign_id}'."
+                                    )
+            # Only allow overwrite if it's the same campaign
             if overwrite:
                 os.remove(symlink_path)
             else:
-                raise ValueError(f"Assets symlink '{symlink_path}' already exists.")
+                raise ValueError(f"Assets destination '{assets_destination}' is already taken.")
+        
+        # Ensure the assets directory exists
+        os.makedirs(f"{STATIC_DIR}/assets", exist_ok=True)
 
         os.symlink(assets_real_path, symlink_path, target_is_directory=True)
         print(f"Assets symlinked: {symlink_path} -> {assets_real_path}")
@@ -299,6 +382,20 @@ def main():
     elif args.command == 'purge':
         import shutil
 
+        def _unlink_assets(campaign_id):
+            """Unlink assets symlink for a campaign if it exists."""
+            task_file = f"{ROOT}/data/tasks/{campaign_id}.json"
+            if not os.path.exists(task_file):
+                return
+            with open(task_file, "r") as f:
+                campaign_data = json.load(f)
+            destination = campaign_data.get("info", {}).get("assets", {}).get("destination")
+            if destination:
+                symlink_path = f"{STATIC_DIR}/{destination}"
+                if os.path.islink(symlink_path):
+                    os.remove(symlink_path)
+                    print(f"Assets symlink removed: {symlink_path}")
+
         # Parse optional campaign name
         purge_args = argparse.ArgumentParser()
         purge_args.add_argument(
@@ -314,6 +411,8 @@ def main():
                 f"Are you sure you want to purge campaign '{campaign_id}'? This action cannot be undone. [y/n] "
             )
             if confirm.lower() == 'y':
+                # Unlink assets before removing task file
+                _unlink_assets(campaign_id)
                 # Remove task file
                 task_file = f"{ROOT}/data/tasks/{campaign_id}.json"
                 if os.path.exists(task_file):
@@ -336,6 +435,13 @@ def main():
                 "Are you sure you want to purge all campaign data? This action cannot be undone. [y/n] "
             )
             if confirm.lower() == 'y':
+                # Unlink all assets first
+                tasks_dir = f"{ROOT}/data/tasks"
+                if os.path.exists(tasks_dir):
+                    for task_file in os.listdir(tasks_dir):
+                        if task_file.endswith('.json'):
+                            campaign_id = task_file[:-5]
+                            _unlink_assets(campaign_id)
                 shutil.rmtree(f"{ROOT}/data/tasks", ignore_errors=True)
                 shutil.rmtree(f"{ROOT}/data/outputs", ignore_errors=True)
                 if os.path.exists(f"{ROOT}/data/progress.json"):