Add lint to spot duplicate titles

RULES.md

@@ -6,6 +6,7 @@ This file contains a list of all lints checked by yml2block.
 |-------------------------|---------|------|
 | `unique_names`          | b001    | Ensure that within a block (e.g. DatasetField) all names are unique. |
 | `block_is_list`         | b002    | Ensure that each block content is a valid yaml list. |
+| `unique_titles`         | b003    | Ensure that within the DatasetField block all titles are unique. |
 | `keywords_valid`        | k001    | Ensure top level keywords (i.e. block names) are present and contain no typos. |
 | `keywords_unique`       | k002    | Ensure no top level keyword occurs multiple times. |
 | `keys_valid`            | e001    | Ensure that no invalid keys are present, e.g. through typos. |

tests/integration_tests.py

@@ -35,6 +35,23 @@ def test_duplicate_names_detected():
     assert result.exit_code == 1, result.output
 
 
+def test_duplicate_titles_detected():
+    """This test ensures that duplicate titles are detected."""
+    runner = CliRunner()
+    result = runner.invoke(
+        yml2block.__main__.main,
+        ["check", "tests/invalid/duplicate_datasetfield_title.yml"],
+    )
+    assert result.exit_code == 1, result.output
+
+    runner = CliRunner()
+    result = runner.invoke(
+        yml2block.__main__.main,
+        ["check", "tests/invalid/duplicate_datasetfield_title.tsv"],
+    )
+    assert result.exit_code == 1, result.output
+
+
 def test_duplicate_top_level_key_detected():
     """This test ensures that duplicates in top-level keys are detected."""
     runner = CliRunner()

yml2block/rules.py

@@ -216,6 +216,35 @@ def unique_names(yaml_chunk, tsv_keyword, level=Level.ERROR):
     return errors
 
 
+def unique_titles(yaml_chunk, tsv_keyword, level=Level.ERROR):
+    """Make sure that each title in the block is only used once.
+
+    block content level lint
+    """
+    if tsv_keyword not in ["datasetField"]:
+        return []
+    titles = Counter()
+    occurrences = defaultdict(list)
+
+    for item in yaml_chunk:
+        item_title = item["title"].value
+        titles.update([item_title])
+        occurrences[item_title].append(item)
+
+    errors = []
+    for title, count in titles.items():
+        if count > 1:
+            occs = [f"line {o.line}" for o in occurrences[title]]
+            errors.append(
+                LintViolation(
+                    level,
+                    "unique_titles",
+                    f"Title '{title}' occurs {count} times: {', '.join(occs) if occs else ''}. Titles should be unique.",
+                )
+            )
+    return errors
+
+
 def block_is_list(yaml_chunk, level=Level.ERROR):
     """Make sure that the yaml chunk is a list.
 
@@ -427,6 +456,8 @@ def no_trailing_spaces(list_item, tsv_keyword, level=Level.ERROR):
     "b001": unique_names,
     "block_is_list": block_is_list,
     "b002": block_is_list,
+    "unique_titles": unique_titles,
+    "b003": unique_titles,
     "keywords_valid": keywords_valid,
     "k001": keywords_valid,
     "keywords_unique": keywords_unique,

yml2block/validation.py

@@ -66,7 +66,7 @@ def validate_entry(yaml_chunk, tsv_keyword, lint_conf, verbose):
 
     longest_row = 0
 
-    for lint in (rules.unique_names,):
+    for lint in (rules.unique_names, rules.unique_titles):
         lint = lint_conf.get(lint)
         violations.extend(lint(yaml_chunk, tsv_keyword))