Merge pull request #85 from NatLibFi/issue84-post-update-query

add post_update_query option
NatLibFi · Aug 30, 2021 · 0bcb8d7 · 0bcb8d7
2 parents 6dca4ae + d482cf5
commit 0bcb8d7
Show file tree

Hide file tree

Showing 3 changed files with 16 additions and 3 deletions.
diff --git a/skosify/cli.py b/skosify/cli.py
@@ -55,10 +55,18 @@ def get_option_parser(defaults):
     group.add_option('--construct-query', type='string',
                      help='SPARQL CONSTRUCT query. '
                      'This query is executed against the input '
-                     'data and the result graph is used as the '
-                     'actual input. '
+                     'data (after possible --update-query) and '
+                     'the result graph is used as the actual input. '
                      'The value can be either the actual query, '
                      'or "@filename".')
+    group.add_option('--post-update-query', type='string',
+                     help='SPARQL update query. '
+                     'This convenience query is executed against '
+                     'the output data before writing it. '
+                     'The value can be either the actual query, '
+                     'or "@filename". '
+                     'Use at your own risk - output may not be '
+                     'SKOS at all.')
     group.add_option('-I', '--infer', action="store_true",
                      help='Perform RDFS subclass/subproperty inference '
                           'before transforming input.')

diff --git a/skosify/config.py b/skosify/config.py
@@ -76,6 +76,7 @@ def __init__(self, file=None):
         self.infer = False
         self.update_query = None
         self.construct_query = None
+        self.post_update_query = None
 
         # mappings
         self.types = {}

diff --git a/skosify/skosify.py b/skosify/skosify.py
@@ -823,13 +823,17 @@ def skosify(*sources, **config):
     # check for duplicate labels
     check_labels(voc, config.preflabel_policy)
 
+    logging.debug("Phase 10: Performing post update query")
+    if config.post_update_query is not None:
+        transform_sparql_update(voc, config.post_update_query)
+
     processtime = time.time()
 
     logging.debug("reading input file took  %d seconds",
                   (inputtime - starttime))
     logging.debug("processing took          %d seconds",
                   (processtime - inputtime))
 
-    logging.debug("Phase 10: Writing output")
+    logging.debug("Phase 11: Writing output")
 
     return voc