Add handling java_options. Fix #30

tabula/wrapper.py

@@ -46,9 +46,13 @@ def read_pdf(input_path, **kwargs):
     elif output_format == 'json':
         kwargs['format'] = 'JSON'
 
+    java_options = kwargs.get('java_options', [])
+    if isinstance(java_options, str):
+        java_options = [java_options]
+
     options = build_options(kwargs)
     path, is_url = localize_file(input_path)
-    args = ["java", "-jar", jar_path] + options + [path]
+    args = ["java"] + java_options + ["-jar", jar_path] + options + [path]
 
     try:
         output = subprocess.check_output(args)
@@ -106,9 +110,13 @@ def convert_into(input_path, output_path, **kwargs):
     elif output_format == 'dataframe':
         raise AttributeError("'output_format' has no attribute 'dataframe'")
 
+    java_options = kwargs.get('java_options', [])
+    if isinstance(java_options, str):
+        java_options = [java_options]
+
     options = build_options(kwargs)
     path, is_url = localize_file(input_path)
-    args = ["java", "-jar", jar_path] + options + [path]
+    args = ["java"] + java_options + ["-jar", jar_path] + options + [path]
 
     try:
         subprocess.check_output(args)

tests/test_read_pdf_table.py

@@ -40,6 +40,12 @@ def test_read_pdf_with_option(self):
         self.assertTrue(tabula.read_pdf(pdf_path, pages=(2, 3), nospreadsheet=True,
                                         guess=False).equals(pd.read_csv(expected_csv2)))
 
+    def test_read_pdf_with_java_option(self):
+        pdf_path = 'tests/resources/data.pdf'
+        expected_csv1 = 'tests/resources/data_1.csv'
+        self.assertTrue(tabula.read_pdf(pdf_path, pages=1, java_options=['-Xmx256m']
+                                       ).equals(pd.read_csv(expected_csv1)))
+
     def test_convert_from(self):
         pdf_path = 'tests/resources/data.pdf'
         expected_csv = 'tests/resources/data_1.csv'