Limit templated visualization data, fix PCA plot for tabular datasets

galaxyproject · May 23, 2024 · 8fcc151 · 8fcc151
1 parent eef864a
commit 8fcc151
Show file tree

Hide file tree

Showing 2 changed files with 27 additions and 8 deletions.
diff --git a/config/plugins/visualizations/PCA_3Dplot/templates/PCA_3Dplot.mako b/config/plugins/visualizations/PCA_3Dplot/templates/PCA_3Dplot.mako
@@ -1,7 +1,11 @@
 <%!
+import csv
 import sys
 from io import StringIO
-import csv
+
+from galaxy.exceptions import RequestParameterInvalidException
+
+MAX_SIZE = 100000  # 100 KB, empirically the largest value I can render on my browser (m2 mac, chrome)
 %>
 
 <%
@@ -29,11 +33,15 @@ def create_options(header):
     return colour_options, start_options
 
 def load_data():
-    input = "\n".join(list(hda.datatype.dataprovider(hda, 'line', comment_char=none, provide_blank=True, strip_lines=False, strip_newlines=True)))
-    tabular_file = StringIO(input)
-    dialect = csv.Sniffer().sniff(tabular_file.read(1024), delimiters=";,\t")
-    tabular_file.seek(0)
-    table = csv.reader(tabular_file, dialect)
+    lines = []
+    size = 0
+    for line in hda.datatype.dataprovider(hda, 'line', comment_char=none, provide_blank=True, strip_lines=False, strip_newlines=True):
+        size += len(line)
+        if size > MAX_SIZE:
+            raise RequestParameterInvalidException("Dataset too large to render, dataset must be less than 100 KB in size.")
+        lines.append(line)
+    dialect = csv.Sniffer().sniff("\n".join(lines))
+    table = csv.reader(lines, dialect)
     data = "["
     for i, row in enumerate(table):
         if i == 0:

diff --git a/config/plugins/visualizations/ts_visjs/templates/ts_visjs.mako b/config/plugins/visualizations/ts_visjs/templates/ts_visjs.mako
@@ -3,6 +3,10 @@ import collections
 import json
 import sys
 from numpy import inf
+
+from galaxy.exceptions import RequestParameterInvalidException
+
+MAX_SIZE = 100000  # 100 KB
 %>
 
 <%
@@ -58,8 +62,15 @@ def write_reaction(edge_id, left_index, right_index, substrates, products, rate,
 def load_data():
     output = {"nodes": "", "border_nodes": "", "edges": "", "edges": "", "self_loops": ""}
 
-    data = ''.join(list(hda.datatype.dataprovider(hda, 'line', strip_lines=True, strip_newlines=True)))
-    data = json.loads(data)
+    lines = []
+    size = 0
+
+    for line in hda.datatype.dataprovider(hda, 'line', strip_lines=True, strip_newlines=True):
+        size += len(line)
+        if size > MAX_SIZE:
+            raise RequestParameterInvalidException("Dataset too large to render, dataset must be less than 100 KB in size.")
+        lines.append(line)
+    data = json.loads("".join(lines))
 
     ordering = data['ordering']
     nodes = {int(key): to_counter(data['nodes'][key], ordering) for key in data['nodes'].keys()}