trustgraph-cli 2.9.2__tar.gz → 2.9.4__tar.gz

{trustgraph_cli-2.9.2 → trustgraph_cli-2.9.4}/PKG-INFO

@@ -1,12 +1,12 @@
 Metadata-Version: 2.4
 Name: trustgraph-cli
-Version: 2.9.2
+Version: 2.9.4
 Summary: TrustGraph provides a means to run a pipeline of flexible AI processing components in a flexible means to achieve a processing pipeline.
 Author-email: "trustgraph.ai" <security@trustgraph.ai>
 Project-URL: Homepage, https://github.com/trustgraph-ai/trustgraph
 Classifier: Programming Language :: Python :: 3
 Classifier: Operating System :: OS Independent
-Requires-Python: >=3.8
+Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 Requires-Dist: trustgraph-base<2.10,>=2.9
 Requires-Dist: requests

{trustgraph_cli-2.9.2 → trustgraph_cli-2.9.4}/pyproject.toml

@@ -8,7 +8,7 @@ dynamic = ["version"]
 authors = [{name = "trustgraph.ai", email = "security@trustgraph.ai"}]
 description = "TrustGraph provides a means to run a pipeline of flexible AI processing components in a flexible means to achieve a processing pipeline."
 readme = "README.md"
-requires-python = ">=3.8"
+requires-python = ">=3.10"
 dependencies = [
     "trustgraph-base>=2.9,<2.10",
     "requests",

{trustgraph_cli-2.9.2 → trustgraph_cli-2.9.4}/trustgraph/cli/graph_to_turtle.py

@@ -12,6 +12,8 @@ import os
 
 from trustgraph.api import Api
 
+from . nquads import valid_language_tag
+
 default_url = os.getenv("TRUSTGRAPH_URL", 'http://localhost:8888/')
 default_collection = 'default'
 default_token = os.getenv("TRUSTGRAPH_TOKEN", None)
@@ -36,6 +38,11 @@ def term_to_rdflib(term):
         datatype = term.get("d")
         language = term.get("l")
         if language:
+            # Skip malformed tags: Literal(lang=...) raises on them, and
+            # the graph is only serialized once the whole stream has been
+            # consumed, so an escaping exception loses the entire dump.
+            if not valid_language_tag(language):
+                return None
             return rdflib.term.Literal(value, lang=language)
         elif datatype:
             return rdflib.term.Literal(value, datatype=rdflib.term.URIRef(datatype))

{trustgraph_cli-2.9.2 → trustgraph_cli-2.9.4}/trustgraph/cli/nquads.py

@@ -32,8 +32,20 @@ _ESCAPES = [
 # for large exports (it runs per term).
 _BAD_IRI = re.compile(r'[\x00-\x20<>"{}|^\x60\\]')
 
-# BCP47 / LANGTAG: [a-zA-Z]+ ('-' [a-zA-Z0-9]+)*
-_LANGTAG = re.compile(r'^[a-zA-Z]+(-[a-zA-Z0-9]+)*$')
+# The body of the LANGTAG production, [a-zA-Z]+ ('-' [a-zA-Z0-9]+)*, which
+# Turtle shares and which is the shape rdflib's Literal(lang=...) accepts.
+# fullmatch, not a trailing $: $ also matches before a trailing newline, so
+# "en\n" would pass and split the quad across two lines.
+_LANGTAG = re.compile(r"[a-zA-Z]+(?:-[a-zA-Z0-9]+)*")
+
+
+def valid_language_tag(language):
+    """True when language can be emitted after the '@' of a LANGTAG.
+
+    The wire schema types the language field as a free-form str, so an
+    unusable tag arrives here as data, not as a bug.
+    """
+    return isinstance(language, str) and _LANGTAG.fullmatch(language) is not None
 
 
 def _escape_literal(value):
@@ -70,7 +82,7 @@ def encode_term(term, is_object=False):
         language = term.get("l")
         datatype = term.get("d")
         if language:
-            if not _LANGTAG.match(language):
+            if not valid_language_tag(language):
                 return None
             return f'"{value}"@{language}'
         if datatype:
@@ -123,11 +135,20 @@ def serialize_nquads(batches, graph_iri, out):
     return written, skipped
 
 
+def _term_to_triple_field(term):
+    """Convert an rdflib term to (str_value, datatype, language)."""
+    if isinstance(term, rdflib.Literal):
+        dt = str(term.datatype) if term.datatype else ""
+        lang = str(term.language) if term.language else ""
+        return str(term), dt, lang
+    return str(term), "", ""
+
+
 def parse_nquads(data):
     """Parse N-Quads bytes back into api Triple values.
 
-    Terms are stringified with str(), the same convention tg-load-knowledge
-    uses, so values survive the store round trip unchanged. The whole
+    Preserves datatype, language tag and IRI-vs-literal distinctions
+    via the Triple.o_datatype and Triple.o_language fields. The whole
     member is materialized in memory (bundles are bounded by
     --triples-limit at export); line-streaming is a possible follow-up.
 
@@ -136,7 +157,11 @@ def parse_nquads(data):
     """
     ds = rdflib.Dataset()
     ds.parse(data=data.decode("utf-8"), format="nquads")
-    return [
-        Triple(s=str(s), p=str(p), o=str(o))
-        for s, p, o, _g in ds.quads((None, None, None, None))
-    ]
+    result = []
+    for s, p, o, _g in ds.quads((None, None, None, None)):
+        o_val, o_dt, o_lang = _term_to_triple_field(o)
+        result.append(Triple(
+            s=str(s), p=str(p), o=o_val,
+            o_datatype=o_dt, o_language=o_lang,
+        ))
+    return result

trustgraph_cli-2.9.4/trustgraph/cli_version.py

@@ -0,0 +1 @@
+__version__ = "2.9.4"

{trustgraph_cli-2.9.2 → trustgraph_cli-2.9.4}/trustgraph_cli.egg-info/PKG-INFO

@@ -1,12 +1,12 @@
 Metadata-Version: 2.4
 Name: trustgraph-cli
-Version: 2.9.2
+Version: 2.9.4
 Summary: TrustGraph provides a means to run a pipeline of flexible AI processing components in a flexible means to achieve a processing pipeline.
 Author-email: "trustgraph.ai" <security@trustgraph.ai>
 Project-URL: Homepage, https://github.com/trustgraph-ai/trustgraph
 Classifier: Programming Language :: Python :: 3
 Classifier: Operating System :: OS Independent
-Requires-Python: >=3.8
+Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 Requires-Dist: trustgraph-base<2.10,>=2.9
 Requires-Dist: requests

trustgraph_cli-2.9.2/trustgraph/cli_version.py

@@ -1 +0,0 @@
-__version__ = "2.9.2"