switching to pipe.communicate. fixes #33

docs/changelog.rst

@@ -11,6 +11,12 @@ latest changes in development
 
 [will add changes here as they are made]
 
+* several bug fixes, including:
+
+  * documentation fixes
+
+  * shell commands hanging on large files (`#33`_)
+
 0.5.0
 -----
 

textract/parsers/doc_parser.py

@@ -4,5 +4,5 @@
 def extract(filename, **kwargs):
     """Extract text from doc files using antiword.
     """
-    pipe = run('antiword %(filename)s' % locals())
-    return pipe.stdout.read()
+    stdout, stderr = run('antiword %(filename)s' % locals())
+    return stdout

textract/parsers/pdf_parser.py

@@ -26,11 +26,11 @@ def extract(filename, method='', **kwargs):
 
 def extract_pdftotext(filename):
     """Extract text from pdfs using the pdftotext command line utility."""
-    pipe = run('pdftotext %(filename)s -' % locals())
-    return pipe.stdout.read()
+    stdout, stderr = run('pdftotext %(filename)s -' % locals())
+    return stdout
 
 
 def extract_pdfminer(filename):
     """Extract text from pdfs using pdfminer."""
-    pipe = run('pdf2txt.py %(filename)s' % locals())
-    return pipe.stdout.read()
+    stdout, stderr = run('pdf2txt.py %(filename)s' % locals())
+    return stdout

textract/parsers/ps_parser.py

@@ -4,5 +4,5 @@
 def extract(filename, **kwargs):
     """Extract text from postscript files using pstotext command.
     """
-    pipe = run('pstotext %(filename)s' % locals())
-    return pipe.stdout.read()
+    stdout, stderr = run('pstotext %(filename)s' % locals())
+    return stdout

textract/parsers/tesseract.py

@@ -5,7 +5,8 @@ def extract(filename, **kwargs):
     """Extract text from various image file formats using tesseract-ocr"""
     # Tesseract can't output to console directly so you must first create
     # a dummy file to write to, read, and then delete
-    pipe = run(
+    stdout, stderr = run(
         'tesseract %(filename)s tmpout && cat tmpout.txt && rm -f tmpout.txt'
-        % locals())
-    return pipe.stdout.read()
+        % locals()
+    )
+    return stdout

textract/shell.py

@@ -11,10 +11,13 @@ def run(command):
         command, shell=True,
         stdout=subprocess.PIPE, stderr=subprocess.PIPE,
     )
-    pipe.wait()
+
+    # pipe.wait() ends up hanging on large files. using
+    # pipe.communicate appears to avoid this issue
+    stdout, stderr = pipe.communicate()
 
     # if pipe is busted, raise an error (unlike Fabric)
     if pipe.returncode != 0:
         raise exceptions.ShellError(command, pipe.returncode)
 
-    return pipe
+    return stdout, stderr