fix up verbose printouts, don't print out large data

pdfparanoia/plugins/aip.py

@@ -44,9 +44,10 @@ class AmericanInstituteOfPhysics(Plugin):
                         #rawdata = copy(obj.rawdata)
                         data = copy(obj.get_data())
 
-                        if "Redistribution subject to AIP license or copyright" in data:
+                        phrase="Redistribution subject to AIP license or copyright"
+                        if phrase in data:
                             if verbose:
-                                sys.stderr.write("%s: Found object with %r; omitting..." % (cls.__name__, data,))
+                                sys.stderr.write("%s: Found object %s with %r; omitting..." % (cls.__name__, objid, phrase,))
 
                             evil_ids.append(objid)
 

pdfparanoia/plugins/ieee.py

@@ -38,9 +38,10 @@ class IEEEXplore(Plugin):
                     #rawdata = copy(obj.rawdata)
                     data = copy(obj.get_data())
 
-                    if "Authorized licensed use limited to: " in data:
+                    phrase= "Authorized licensed use limited to: "
+                    if phrase in data:
                         if verbose:
-                            sys.stderr.write("%s: Found object with %r; omitting..." % (cls.__name__, data,))
+                            sys.stderr.write("%s: Found object %s with %r; omitting..." % (cls.__name__, objid, phrase,))
 
                         evil_ids.append(objid)
 

pdfparanoia/plugins/jstor.py

@@ -61,6 +61,9 @@ class JSTOR(Plugin):
                     if all([requirement in data for requirement in JSTOR.requirements]):
                         better_content = data
 
+                        if verbose:
+                            sys.stderr.write("%s: Found object %s with %r; omitting..." % (cls.__name__, objid, cls.requirements))
+
                         # remove the date
                         startpos = better_content.find("This content downloaded ")
                         endpos = better_content.find(")", startpos)
@@ -82,9 +85,6 @@ class JSTOR(Plugin):
                             startpos = better_content.rfind("/F2 11 Tf\n")
                             endpos = better_content.find("Tf\n", startpos+5)
 
-                            if verbose:
-                                sys.stderr.write("%s: Found object with %r; omitting..." % (cls.__name__, better_content[startpos:endpos],))
-
                             better_content = better_content[0:startpos] + better_content[endpos:]
 
                         replacements.append([objid, better_content])

pdfparanoia/plugins/spie.py

@@ -40,9 +40,10 @@ class SPIE(Plugin):
                 if obj.attrs.has_key("Filter") and str(obj.attrs["Filter"]) == "/FlateDecode":
                     data = copy(obj.get_data())
 
-                    if "Downloaded From:" in data:
+                    phrase="Downloaded From:"
+                    if phrase in data:
                         if verbose:
-                            sys.stderr.write("%s: found object with %r; omitting..." % (cls.__name__, data))
+                            sys.stderr.write("%s: found object %s with %r; omitting..." % (cls.__name__, objid, phrase))
                         evil_ids.append(objid)
 
         for objid in evil_ids: