Merge pull request RasaHQ#4696 from RasaHQ/export-nlu-as-json-from-interactive

Export nlu as json from interactive

Author: erohmensing

CHANGELOG.rst

@@ -23,6 +23,7 @@ Removed
 
 Fixed
 -----
+- Fixed exporting NLU training data in ``json`` format from ``rasa interactive``
 
 [1.4.3] - 2019-10-29
 ^^^^^^^^^^^^^^^^^^^^

rasa/core/training/interactive.py

@@ -11,6 +11,7 @@
 import numpy as np
 from aiohttp import ClientError
 from colorclass import Color
+from rasa.nlu.training_data.loading import MARKDOWN, RASA
 from sanic import Sanic, response
 from sanic.exceptions import NotFound
 from terminaltables import AsciiTable, SingleTable
@@ -680,7 +681,7 @@ def _request_export_info() -> Tuple[Text, Text, Text]:
             "merge learned data with previous training examples)",
             default=PATHS["nlu"],
             validate=io_utils.file_type_validator(
-                [".md"],
+                [".md", ".json"],
                 "Please provide a valid export path for the NLU data, e.g. 'nlu.md'.",
             ),
         ),
@@ -831,21 +832,29 @@ async def _write_nlu_to_file(
 
     # need to guess the format of the file before opening it to avoid a read
     # in a write
-    if loading.guess_format(export_nlu_path) in {"md", "unk"}:
-        fformat = "md"
-    else:
-        fformat = "json"
-
-    if fformat == "md":
+    nlu_format = _get_nlu_target_format(export_nlu_path)
+    if nlu_format == MARKDOWN:
         stringified_training_data = nlu_data.nlu_as_markdown()
     else:
         stringified_training_data = nlu_data.nlu_as_json()
 
     io_utils.write_text_file(stringified_training_data, export_nlu_path)
 
 
+def _get_nlu_target_format(export_path: Text) -> Text:
+    guessed_format = loading.guess_format(export_path)
+
+    if guessed_format not in {MARKDOWN, RASA}:
+        if export_path.endswith(".json"):
+            guessed_format = RASA
+        else:
+            guessed_format = MARKDOWN
+
+    return guessed_format
+
+
 def _entities_from_messages(messages):
-    """Return all entities that occur in atleast one of the messages."""
+    """Return all entities that occur in at least one of the messages."""
     return list({e["entity"] for m in messages for e in m.data.get("entities", [])})
 
 

tests/core/test_interactive.py

@@ -1,11 +1,14 @@
 import json
+from typing import Text
+
 import pytest
 import uuid
 from aioresponses import aioresponses
 
 import rasa.utils.io
 from rasa.core.events import BotUttered
 from rasa.core.training import interactive
+from rasa.nlu.training_data.loading import RASA, MARKDOWN
 from rasa.utils.endpoints import EndpointConfig
 from rasa.core.actions.action import default_actions
 from rasa.core.domain import Domain
@@ -343,3 +346,11 @@ async def test_filter_intents_before_save_nlu_file():
         msgs.append(Message("/" + choice(intents), greet))
 
     assert test_msgs == interactive._filter_messages(msgs)
+
+
+@pytest.mark.parametrize(
+    "path, expected_format",
+    [("bla.json", RASA), ("other.md", MARKDOWN), ("unknown", MARKDOWN)],
+)
+def test_get_nlu_target_format(path: Text, expected_format: Text):
+    assert interactive._get_nlu_target_format(path) == expected_format