feat(csv-file-validations): Csv File Validations in CsvToDatabseConfiguration

requirements/base.txt

@@ -100,7 +100,7 @@ flask-sqlalchemy==2.5.1
     #   flask-migrate
 flask-talisman==0.8.1
     # via apache-superset
-flask-wtf==0.14.3
+flask-wtf==1.0.1
     # via
     #   apache-superset
     #   flask-appbuilder

superset/config.py

@@ -735,6 +735,9 @@ def _try_json_readsha(filepath: str, length: int) -> Optional[str]:
 CSV_EXTENSIONS = {"csv", "tsv", "txt"}
 COLUMNAR_EXTENSIONS = {"parquet", "zip"}
 ALLOWED_EXTENSIONS = {*EXCEL_EXTENSIONS, *CSV_EXTENSIONS, *COLUMNAR_EXTENSIONS}
+CSV_MAX_SIZES = 1 * 1024 * 1024
+CSV_MAX_ROWS = 1000
+CSV_MIN_ROWS = 1
 
 # CSV Options: key/value pairs that will be passed as argument to DataFrame.to_csv
 # method.

superset/views/base.py

@@ -79,8 +79,7 @@
 from superset.translations.utils import get_language_pack
 from superset.utils import core as utils
 from superset.utils.core import get_user_id
-
-from .utils import bootstrap_user_data
+from superset.views.utils import bootstrap_user_data
 
 FRONTEND_CONF_KEYS = (
     "SUPERSET_WEBSERVER_TIMEOUT",
@@ -110,6 +109,9 @@
     "CSV_EXTENSIONS",
     "COLUMNAR_EXTENSIONS",
     "ALLOWED_EXTENSIONS",
+    "CSV_MAX_SIZES",
+    "CSV_MAX_ROWS",
+    "CSV_MIN_ROWS",
     "SAMPLES_ROW_LIMIT",
     "DEFAULT_TIME_FILTER",
     "HTML_SANITIZATION",

superset/views/database/forms.py

@@ -20,7 +20,7 @@
 from flask_appbuilder.fieldwidgets import BS3TextFieldWidget
 from flask_appbuilder.forms import DynamicForm
 from flask_babel import lazy_gettext as _
-from flask_wtf.file import FileAllowed, FileField, FileRequired
+from flask_wtf.file import FileAllowed, FileField, FileRequired, FileSize
 from wtforms import (
     BooleanField,
     IntegerField,
@@ -107,9 +107,23 @@ def is_engine_allowed_to_file_upl(database: Database) -> bool:
 class CsvToDatabaseForm(UploadToDatabaseForm):
     csv_file = FileField(
         _("CSV Upload"),
-        description=_("Select a file to be uploaded to the database"),
+        description=_(
+            "Select a file to be uploaded to a database. Max Size of the file should be "
+            + str(config["CSV_MAX_SIZES"] / 1048576)
+            + "  MB and the accepted extensions are: "
+            "%(allowed_extensions)s",
+            allowed_extensions=", ".join(
+                config["ALLOWED_EXTENSIONS"].intersection(config["CSV_EXTENSIONS"])
+            ),
+        ),
         validators=[
             FileRequired(),
+            FileSize(
+                config["CSV_MAX_SIZES"],
+                message="File size must not exceed the limit: "
+                + str(config["CSV_MAX_SIZES"] / 1048576)
+                + "MB",
+            ),
             FileAllowed(
                 config["ALLOWED_EXTENSIONS"].intersection(config["CSV_EXTENSIONS"]),
                 _(
@@ -252,8 +266,18 @@ class CsvToDatabaseForm(UploadToDatabaseForm):
     )
     nrows = IntegerField(
         _("Rows to Read"),
-        description=_("Number of rows of file to read"),
-        validators=[Optional(), NumberRange(min=0)],
+        description=_(
+            "Number of rows of file to read. Minimum "
+            + str(config["CSV_MIN_ROWS"])
+            + " and Maximum "
+            + str(config["CSV_MAX_ROWS"])
+            + " rows are allowed"
+        ),
+        validators=[
+            Optional(),
+            NumberRange(min=config["CSV_MIN_ROWS"]),
+            NumberRange(max=config["CSV_MAX_ROWS"]),
+        ],
         widget=BS3TextFieldWidget(),
     )
     skiprows = IntegerField(