add meta.extract_text

oml/media/__init__.py

@@ -123,3 +123,17 @@ def metadata(f, from_=None):
             data[key] = [data[key]] if data[key] else []
     return data
 
+def extract_text(path):
+    ext = path.split('.')[-1]
+    text = ''
+    try:
+        if ext in ('epub', 'kepub'):
+            text = epub.extract_text(path)
+        elif ext == 'pdf':
+            text = pdf.extract_text(path)
+        elif ext == 'txt':
+            text = txt.extract_text(path)
+    except:
+        logger.debug('failed to extract text from %s', path, exc_info=1)
+        text = ''
+    return text

oml/media/pdf.py

@@ -229,6 +229,22 @@ def extract_text(pdf):
             stdout = ''
     return stdout.strip()
 
+
+def pypdf_extract_text(path):
+    '''
+        slow and bad results
+    '''
+    pdf = PdfFileReader(path)
+    content = []
+    for i in range(0, pdf.getNumPages()):
+        try:
+            extracted_text = pdf.getPage(i).extractText()
+            content.append(extracted_text)
+        except:
+            pass
+    content = "\n".join(content).replace("\xa0", " ").strip()
+    return content
+
 def extract_isbn(text):
     isbns = find_isbns(text)
     if isbns: