For Excel Reader, support defining a default_glossary. When a cell co…

…ntains only `termName` it will add `@<default_glossary>`. Support manually overriding the default_glossary term by specifying `term@OtherGlossary` in the cell value. Closes #264
wjohnson · Dec 22, 2023 · 971d009 · 971d009
1 parent f5fa20c
commit 971d009
Show file tree

Hide file tree

Showing 2 changed files with 19 additions and 4 deletions.
diff --git a/pyapacheatlas/readers/excel.py b/pyapacheatlas/readers/excel.py
@@ -37,6 +37,11 @@ class ExcelConfiguration(ReaderConfiguration):
     :param str process_prefix:
         Defaults to "Process" and represents the prefix of the columns
         in Excel to be considered related to the table process.
+    :param str default_glossary:
+        Defaults to "Glossary" and identifies what the glossary suffix
+        should be for the `[Relationship] meanings` column if not
+        provided. Can be manually overridden in Excel with `term@OtherGlossary`
+        as the cell value.
     :param str column_transformation_name:
         Defaults to "transformation" and identifies the column that
         represents the transformation for a specific column.

diff --git a/pyapacheatlas/readers/reader.py b/pyapacheatlas/readers/reader.py
@@ -17,9 +17,14 @@
 
 class ReaderConfiguration():
     """
-    A base configuration for the Reader class.  Allows you to customize
-    headers with a source_prefix, target_prefix, and process_prefix for
-    parsing table and column lineages.
+    A base configuration for the Reader class.  Allows you to:
+
+    * value_separator: Defaults to `;` control the separator for lists of objects.
+    * source_prefix: Defaults to `Source` control the field name for Source in lineage parsing
+    * target_prefix: Defaults to `Target` control the field name for Target in lineage parsing
+    * process_prefix: Defaults to `Process` control the field name for Process in lineage parsing
+    * column_transformation_name: Defaults to `transformation` control the field name for indiciating what the transformation is, deprecated
+    * default_glossary: Defaults to `Glossary` when you specify a term this is added with `@<Value>` unless the user provided an `@` in the term.
     """
 
     def __init__(self, **kwargs):
@@ -33,6 +38,7 @@ def __init__(self, **kwargs):
             "process_prefix", "Process")
         self.column_transformation_name = kwargs.get(
             "column_transformation_name", "transformation")
+        self.default_glossary = kwargs.get("default_glossary", "Glossary")
 
 
 class Reader(LineageMixIn):
@@ -159,7 +165,11 @@ def _organize_attributes(self, row, existing_entities, ignore=[]):
                     reference_object = [
                         {"typeName": "AtlasGlossaryTerm",
                          "uniqueAttributes": {
-                             "qualifiedName": "{}@Glossary".format(t)
+                             # Allow for a default glossary term
+                             # or for the user to manually specify it
+                             # This might break others usage of `@` in their
+                             # glossary terms though
+                             "qualifiedName": f"{t}@{self.config.default_glossary}" if "@" not in t else t
                          }
                          } for t in terms
                     ]