[#189 state:fixed] Added "Export to CSV" feature.

core/app.py

@@ -90,6 +90,7 @@ class DupeGuru(RegistrableApplication, Broadcaster):
     # show_results_window()
     # show_problem_dialog()
     # select_dest_folder(prompt: str) --> str
+    # select_dest_file(prompt: str, ext: str) --> str
     
     # in fairware prompts, we don't mention the edition, it's too long.
     PROMPT_NAME = "dupeGuru"
@@ -199,6 +200,19 @@ class DupeGuru(RegistrableApplication, Broadcaster):
         except EnvironmentError:
             return None
     
+    def _get_export_data(self):
+        columns = [col for col in self.result_table.columns.ordered_columns
+            if col.visible and col.name != 'marked']
+        colnames = [col.display for col in columns]
+        rows = []
+        for group_id, group in enumerate(self.results.groups):
+            for dupe in group:
+                data = self.get_display_info(dupe, group)
+                row = [data[col.name] for col in columns]
+                row.insert(0, group_id)
+                rows.append(row)
+        return colnames, rows
+    
     def _results_changed(self):
         self.selected_dupes = [d for d in self.selected_dupes
             if self.results.get_group_of_duplicate(d) is not None]
@@ -356,17 +370,15 @@ class DupeGuru(RegistrableApplication, Broadcaster):
         self.view.start_job(JobType.Delete, self._do_delete, args=args)
     
     def export_to_xhtml(self):
-        columns = [col for col in self.result_table.columns.ordered_columns
-            if col.visible and col.name != 'marked']
-        colnames = [col.display for col in columns]
-        rows = []
-        for group in self.results.groups:
-            for dupe in group:
-                data = self.get_display_info(dupe, group)
-                row = [data[col.name] for col in columns]
-                row.insert(0, dupe is not group.ref)
-                rows.append(row)
-        return export.export_to_xhtml(colnames, rows)
+        colnames, rows = self._get_export_data()
+        export_path = export.export_to_xhtml(colnames, rows)
+        self.view.open_path(export_path)
+    
+    def export_to_csv(self):
+        dest_file = self.view.select_dest_file(tr("Select a destination for your exported CSV"), 'csv')
+        if dest_file:
+            colnames, rows = self._get_export_data()
+            export.export_to_csv(dest_file, colnames, rows)
     
     def get_display_info(self, dupe, group, delta=False):
         def empty_data():

core/export.py

@@ -8,6 +8,7 @@
 
 import os.path as op
 from tempfile import mkdtemp
+import csv
 
 # Yes, this is a very low-tech solution, but at least it doesn't have all these annoying dependency
 # and resource problems.
@@ -119,12 +120,18 @@ def export_to_xhtml(colnames, rows):
         assert len(rows[0]) == len(colnames) + 1 # + 1 is for the "indented" flag
     colheaders = ''.join(COLHEADERS_TEMPLATE.format(name=name) for name in colnames)
     rendered_rows = []
+    previous_group_id = None
     for row in rows:
         # [2:] is to remove the indented flag + filename
-        indented = 'indented' if row[0] else ''
+        if row[0] != previous_group_id:
+            # We've just changed dupe group, which means that this dupe is a ref. We don't indent it.
+            indented = ''
+        else:
+            indented = 'indented'
         filename = row[1]
         cells = ''.join(CELL_TEMPLATE.format(value=value) for value in row[2:])
         rendered_rows.append(ROW_TEMPLATE.format(indented=indented, filename=filename, cells=cells))
+        previous_group_id = row[0]
     rendered_rows = ''.join(rendered_rows)
     # The main template can't use format because the css code uses {}
     content = MAIN_TEMPLATE.replace('$colheaders', colheaders).replace('$rows', rendered_rows)
@@ -134,3 +141,9 @@ def export_to_xhtml(colnames, rows):
     fp.write(content)
     fp.close()
     return destpath
+
+def export_to_csv(dest, colnames, rows):
+    writer = csv.writer(open(dest, 'wt', encoding='utf-8'))
+    writer.writerow(["Group ID"] + colnames)
+    for row in rows:
+        writer.writerow(row)