unique_toolkit 1.70.6__tar.gz → 1.70.8__tar.gz

{unique_toolkit-1.70.6 → unique_toolkit-1.70.8}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: unique-toolkit
-Version: 1.70.6
+Version: 1.70.8
 Summary: 
 Author: Cedric Klinkert, Martin Fadler, Sadique Sheik, Fabian Schläpfer, Pascal Hauri, Ejiroghene Oki-Peter, Rami Azouz, Ricardo Visini, Sanzio Monti
 Author-email: Cedric Klinkert <cedric.klinkert@unique.ch>, Martin Fadler <martin.fadler@unique.ch>, Sadique Sheik <sadique@unique.ch>, Fabian Schläpfer <fabian@unique.ch>, Pascal Hauri <pascal@unique.ch>, Ejiroghene Oki-Peter <oki-peter.ejiroghene@unique.ai>, Rami Azouz <rami.ext@unique.ch>, Ricardo Visini <ricardo.visini@unique.ch>, Sanzio Monti <sanzio.monti@unique.ai>

{unique_toolkit-1.70.6 → unique_toolkit-1.70.8}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "unique_toolkit"
-version = "1.70.6"
+version = "1.70.8"
 description = ""
 readme = "README.md"
 requires-python = ">=3.12"

{unique_toolkit-1.70.6 → unique_toolkit-1.70.8}/unique_toolkit/agentic/tools/openai_builtin/code_interpreter/postprocessors/generated_files.py

@@ -21,6 +21,7 @@ from tenacity import (
 
 from unique_toolkit import ChatService
 from unique_toolkit._common.execution import failsafe_async
+from unique_toolkit._common.utils.files import FileMimeType, ImageMimeType
 from unique_toolkit.agentic.feature_flags.feature_flags import feature_flags
 from unique_toolkit.agentic.postprocessor.postprocessor_manager import (
     ResponsesApiPostprocessor,
@@ -586,12 +587,7 @@ class DisplayCodeInterpreterFilesPostProcessor(
                 missed_files.append(filename)
 
             is_html_rendered = is_html
-            has_superscript = f"<sup>{ref_number}</sup>" in loop_response.message.text
-            if (
-                replaced
-                and has_superscript
-                and not (is_image or is_html_rendered or fence_ff_on)
-            ):
+            if replaced and not (is_image or is_html_rendered or fence_ff_on):
                 loop_response.message.references.append(
                     ContentReference(
                         sequence_number=ref_number,
@@ -617,16 +613,7 @@ class DisplayCodeInterpreterFilesPostProcessor(
                 len(code_blocks),
                 [f.filename for b in code_blocks for f in b.files],
             )
-            unmatched_blocks = _warn_unmatched_code_blocks(
-                self._content_map, code_blocks
-            )
-            if unmatched_blocks:
-                self._log.info(
-                    "Fence path: %d file(s) have no code-block match and will appear "
-                    "as plain download links: %s",
-                    len(unmatched_blocks),
-                    list(unmatched_blocks),
-                )
+            _warn_unmatched_code_blocks(self._content_map, code_blocks)
             text_before = loop_response.message.text
             loop_response.message.text = _inject_code_execution_fences(
                 loop_response.message.text,
@@ -718,7 +705,16 @@ class DisplayCodeInterpreterFilesPostProcessor(
                     container_file.filename, "uploading", force_publish=True
                 )
 
-            mime = guess_type(container_file.filename)[0] or "text/plain"
+            raw_mime = guess_type(container_file.filename)[0] or "text/plain"
+            mime = _kb_safe_mime(raw_mime)
+            if mime != raw_mime:
+                self._log.info(
+                    "MIME type '%s' is not supported by the Unique KB; "
+                    "uploading '%s' as '%s' so the file can be stored and downloaded.",
+                    raw_mime,
+                    container_file.filename,
+                    mime,
+                )
             self._log.info(
                 "Uploading '%s' to knowledge base (mime=%s, %d bytes)",
                 container_file.filename,
@@ -1045,6 +1041,37 @@ def _get_file_type(filename: str) -> CodeInterpreterFileType:
     return "document"
 
 
+def _kb_safe_mime(mime: str) -> str:
+    """Return a MIME type the Unique KB will accept.
+
+    Membership is defined by ``FileMimeType`` and ``ImageMimeType`` in
+    ``unique_toolkit._common.utils.files`` (same catalog as path-based helpers
+    like ``FileMimeType.is_valid_mime``, but here we already have a resolved
+    MIME string from ``mimetypes.guess_type``, so we use StrEnum value lookup
+    instead of re-parsing a path).
+
+    The KB GraphQL API rejects many code-file MIME types (e.g. ``text/x-python``
+    for ``.py`` files).  Anything not in those enums is coerced to
+    ``text/plain`` so the file can be stored and downloaded without changing
+    its bytes.
+
+    Other ``image/*`` subtypes (e.g. ``image/jpg``) still pass through unchanged.
+    """
+    try:
+        FileMimeType(mime)
+        return mime
+    except ValueError:
+        pass
+    try:
+        ImageMimeType(mime)
+        return mime
+    except ValueError:
+        pass
+    if mime.startswith("image/"):
+        return mime
+    return "text/plain"
+
+
 def _file_title(filename: str) -> str:
     """Derive a human-readable title from a filename.
 
@@ -1401,7 +1428,7 @@ def _replace_dangling_sandbox_links(text: str) -> tuple[str, bool]:
 def _warn_unmatched_code_blocks(
     content_map: dict[str, str | None],
     code_blocks: list[CodeInterpreterBlock],
-) -> dict[str, str]:
+) -> None:
     """Warn for files that were uploaded but could not be matched to any code block.
 
     When the fence feature flag is on, every uploaded file should map to a code block
@@ -1409,13 +1436,8 @@ def _warn_unmatched_code_blocks(
     matched (e.g. the LLM used a variable for the output path rather than a literal
     string) it falls back to a plain unique://content/ link with no code context.
     The user can still download it, but the frontend artifact UI will not be shown.
-
-    Returns:
-        Mapping of filename → content_id for every file that could not be matched
-        to a code block.  Callers may use this to inject additional fallback links.
     """
     fenced_filenames = {f.filename for block in code_blocks for f in block.files}
-    unmatched: dict[str, str] = {}
     for filename, content_id in content_map.items():
         if content_id is None:
             continue
@@ -1431,8 +1453,6 @@ def _warn_unmatched_code_blocks(
                 content_id,
                 filename,
             )
-            unmatched[filename] = content_id
-    return unmatched
 
 
 def _get_next_ref_number(references: list[ContentReference] | None) -> int:
@@ -1473,12 +1493,11 @@ def _replace_container_image_citation(
     if not re.search(image_markdown, text):
         logger.warning(
             "No sandbox link found for image '%s' (content_id=%s); "
-            "file was uploaded but the LLM did not reference it — appending fallback link.",
+            "file was uploaded but the LLM did not reference it — it will not be displayed.",
             filename,
             content_id,
         )
-        fallback = f"\n\n📎 [{filename}](unique://content/{content_id})"
-        return text + fallback, True
+        return text, False
 
     logger.info("Inserting image citation for '%s'", filename)
     return re.sub(
@@ -1497,12 +1516,11 @@ def _replace_container_html_citation(
     if not re.search(html_markdown, text):
         logger.warning(
             "No sandbox link found for HTML file '%s' (content_id=%s); "
-            "file was uploaded but the LLM did not reference it — appending fallback link.",
+            "file was uploaded but the LLM did not reference it — it will not be displayed.",
             filename,
             content_id,
         )
-        fallback = f"\n\n📎 [{filename}](unique://content/{content_id})"
-        return text + fallback, True
+        return text, False
 
     logger.info("Inserting HTML rendering block for '%s'", filename)
     block = f"```HtmlRendering\n800px\n600px\n\nunique://content/{content_id}\n\n```"
@@ -1555,12 +1573,11 @@ def _replace_container_file_citation(
     if not re.search(file_markdown, text):
         logger.warning(
             "No sandbox link found for file '%s' (content_id=%s); "
-            "file was uploaded but the LLM did not reference it — appending fallback link.",
+            "file was uploaded but the LLM did not reference it — it will not be displayed.",
             filename,
             content_id,
         )
-        fallback = f"\n\n📎 [{filename}](unique://content/{content_id})"
-        return text + fallback, True
+        return text, False
 
     logger.info("Displaying file %s", filename)
     replacement = (