fix bugs

Author: hrfng

setup.py

@@ -80,7 +80,7 @@ def get_package_data_files(package, data, package_dir=None):
         exclude=("examples*", "tests*", "applications*", "model_zoo*"),
     ),
     package_dir={"": "./src"},
-    package_data={},
+    package_data={"": ["*.txt"]},
     setup_requires=[],
     install_requires=REQUIRED_PACKAGES,
     entry_points={},

src/bisheng_unstructured/topdf/docx2pdf.py

@@ -54,7 +54,7 @@ def render(self, input_file, output_file=None, to_bytes=False):
 class DocxToPDFV1(object):
     def __init__(self, kwargs={}):
         cmd_template = """
-            soffice --headless --convert-to pdf --outdir {1} {0}
+            soffice --headless --convert-to pdf --outdir \"{1}\" \"{0}\"
         """
 
         def _norm_cmd(cmd):

src/bisheng_unstructured/topdf/excel2pdf.py

@@ -16,7 +16,7 @@ def __init__(self, kwargs={}):
             soffice --headless --convert-to xlsx --outdir
         """
 
-        cmd_template3 = "sed -e 's/\t/,/g' {0} > {1}"
+        cmd_template3 = 'sed -e \'s/\t/,/g\' "{0}" > "{1}"'
 
         def _norm_cmd(cmd):
             return " ".join([p.strip() for p in cmd.strip().split()])
@@ -50,7 +50,7 @@ def render(self, input_file, output_file=None, to_bytes=False):
                 type_ext = "csv"
 
             if type_ext in ["xls", "csv"]:
-                cmd = self.cmd_template2 + " {1} {0}".format(input_file, temp_dir)
+                cmd = self.cmd_template2 + ' "{1}" "{0}"'.format(input_file, temp_dir)
                 ExcelToPDF.run(cmd)
                 filename = filename.rsplit(".", 1)[0] + ".xlsx"
                 input_file = os.path.join(temp_dir, filename)
@@ -64,7 +64,7 @@ def render(self, input_file, output_file=None, to_bytes=False):
             input_file = os.path.join(temp_dir, filename)
             wb.save(input_file)
 
-            cmd = self.cmd_template + " {1} {0}".format(input_file, temp_dir)
+            cmd = self.cmd_template + ' "{1}" "{0}"'.format(input_file, temp_dir)
             ExcelToPDF.run(cmd)
 
             if output_file is not None:

src/bisheng_unstructured/topdf/pptx2pdf.py

@@ -5,7 +5,7 @@
 class PptxToPDF(object):
     def __init__(self, kwargs={}):
         cmd_template = """
-            soffice --headless --convert-to pdf --outdir {1} {0}
+            soffice --headless --convert-to pdf --outdir \"{1}\" \"{0}\"
         """
 
         def _norm_cmd(cmd):

src/bisheng_unstructured/topdf/text2pdf.py

@@ -98,7 +98,7 @@ def __init__(self, kwargs={}):
         """
 
         cmd_template2 = """
-            soffice --headless --convert-to pdf --outdir {1} {0}
+            soffice --headless --convert-to pdf --outdir \"{1}\" \"{0}\"
         """
 
         def _norm_cmd(cmd):

tests/test_cmd_with_blank_char.py

@@ -0,0 +1,15 @@
+import os
+
+
+def test_python_cmd_with_blank():
+    path1 = "data/1 Folder/"
+    cmd = """ls \"{0}\"""".format(path1)
+    print("cmd", cmd)
+    os.system(cmd)
+
+    cmd = 'ls "{0}"'.format(path1)
+    print("cmd", cmd)
+    os.system(cmd)
+
+
+test_python_cmd_with_blank()