Error with useful comment when user file isn't UTF-8

re: issue #799

corpustools/corpus/io/csv.py

@@ -6,7 +6,7 @@
 
 from .helper import parse_transcription, AnnotationType, SyllableBaseAnnotation
 
-from corpustools.exceptions import DelimiterError, PCTError
+from corpustools.exceptions import DelimiterError, PCTEncodingError, PCTError
 import corpustools.gui.modernize as modernize
 
 import time
@@ -44,12 +44,17 @@ def inspect_csv(path, num_lines = 10, coldelim = None, transdelim = None):
     else:
         trans_delimiters = ['.',' ', ';', ',']
 
-    with open(path, 'r', encoding='utf-8-sig') as f:
-        lines = []
-        head = f.readline().strip()
-        for line in f.readlines():
-            if line != '\n':
-                lines.append(line.strip())
+    try:
+        with open(path, 'r', encoding='utf-8-sig') as f:
+            lines = []
+            head = f.readline().strip()
+            for line in f.readlines():
+                if line != '\n':
+                    lines.append(line.strip())
+    except UnicodeDecodeError:
+        raise(PCTEncodingError("PCT cannot decode your text file. Make sure it is in UTF-8.\n\n"
+                               "To convert your file to UTF-8, please open it in Notepad or TextEdit "
+                               "and then 'Save as' with the encoding set to 'UTF-8.'"))
 
     best = ''  ## best guess for the column delimiter (candidates: ',', 'tab', ':', and '|')
     num = 1

corpustools/exceptions.py

@@ -59,6 +59,15 @@ class PCTContextError(PCTError):
     """
     pass
 
+## External file loading exceptions
+
+class PCTEncodingError(PCTError):
+    """
+    Exception class for when the user tries to load an external file that cannot be
+    decoded. For example, UTF-16
+    """
+    pass
+
 ## Corpus loading exceptions
 
 class PCTOSError(PCTError):

corpustools/gui/iogui.py

@@ -5,7 +5,7 @@
 import webbrowser
 
 from .imports import *
-from corpustools.exceptions import PCTError, PCTPythonError, MissingFeatureError
+from corpustools.exceptions import PCTError, PCTPythonError, MissingFeatureError, PCTEncodingError
 from corpustools.decorators import check_for_errors
 from corpustools.corpus.io.binary import load_binary, save_binary, PCTUnpickler
 from corpustools.corpus.io.csv import (inspect_csv, load_corpus_csv,
@@ -755,6 +755,11 @@ def inspect(self):
             if self.textType == 'csv':
                 try:
                     atts, coldelim = inspect_csv(self.pathWidget.value())
+                except PCTEncodingError as error:
+                    # when user tries to load a file with an unknown encoding
+                    reply = QMessageBox.critical(self, "Error encountered", str(error))
+                    self.updateColumnFrame([])
+                    return
                 except PCTError:
                     self.updateColumnFrame([])
                     return