updated to sdk 1.0.9, spacy 3.6

Author: keighrim

Containerfile

@@ -1,5 +1,5 @@
 # Use the same base image version as the clams-python python library version
-FROM ghcr.io/clamsproject/clams-python:1.0.7
+FROM ghcr.io/clamsproject/clams-python:1.0.9
 # See https://github.com/orgs/clamsproject/packages?tab=packages&q=clams-python for more base images
 # IF you want to automatically publish this image to the clamsproject organization, 
 # 1. you should have generated this template without --no-github-actions flag
@@ -24,6 +24,7 @@ ENV CLAMS_APP_VERSION ${CLAMS_APP_VERSION}
 COPY ./ /app
 WORKDIR /app
 RUN pip3 install -r requirements.txt
+RUN python3 -m spacy download en_core_web_sm
 
 RUN python3 -m spacy download en_core_web_sm
 

README.md

@@ -1,48 +1,20 @@
 # Spacy NLP Service
 
-The spaCy NLP tool wrapped as a CLAMS service, spaCy is distributed under the [MIT license](https://github.com/explosion/spaCy/blob/master/LICENSE).
+## Description
+The spaCy NLP wrapped as a CLAMS service, spaCy is distributed under the [MIT license](https://github.com/explosion/spaCy/blob/master/LICENSE).
 
-This requires Python 3.8 or higher. For local install of required Python modules see [requirements.txt](requirements.txt).
-
-## Using this service
-
-Use `python app.py -t example-mmif.json out.json` just to test the wrapping code without using a server. To test this using a server you run the app as a service in one terminal:
+## User instruction
 
-```bash
-$ python app.py
-```
+General user instructions for CLAMS apps is available at [CLAMS Apps documentation](https://apps.clams.ai/clamsapp).
 
-And poke at it from another:
+### System requirements
 
-```bash
-$ curl http://0.0.0.0:5000/
-$ curl -H "Accept: application/json" -X POST [email protected] http://0.0.0.0:5000/
-```
-
-In CLAMS you usually run this in a container. To create an image
-
-```bash
-$ docker build -f Containerfile -t clams-spacy-wrapper .
-```
+This requires Python 3.8 or higher. For local install of required Python modules see [requirements.txt](requirements.txt).
 
-And to run it as a container:
+#### Using as local python programe
 
-```bash
-$ docker run --rm -d -p 5000:5000 clams-spacy-wrapper
-$ curl -H "Accept: application/json" -X POST [email protected] http://0.0.0.0:5000/
-```
+Use `python app.py -t example-mmif.json out.json` just to test the wrapping code without using a server.
 
-The spaCy code will run on each text document in the input MMIF file. The file `example-mmif.json` has one text document in the top level `documents` property and two text documents in one of the views. The text documents all look as follows:
+### Configurable runtime parameter
 
-```json
-{
-  "@type": "http://mmif.clams.ai/0.4.0/vocabulary/TextDocument",
-  "properties": {
-    "id": "m2",
-    "text": {
-      "@value": "Hello, this is Jim Lehrer with the NewsHour on PBS...."
-    }
-  }
-}
-```
-Instead of a `text:@value` property the text could in an external file, which would be given as a URI in the `location` property. See the readme file in [https://github.com/clamsproject/app-nlp-example](https://github.com/clamsproject/app-nlp-example) on how to do this.
+For the full list of parameters, please refer to the app metadata from [CLAMS App Directory](https://apps.clams.ai/clamsapp/) or [`metadata.py`](metadata.py) file in this repository.

app.py

@@ -12,27 +12,29 @@
 """
 
 import argparse
+import logging
 from typing import Union
 
-# Imports needed for Clams and MMIF.
-# Non-NLP Clams applications will require AnnotationTypes
-
+import spacy
 from clams import ClamsApp, Restifier
+from lapps.discriminators import Uri
 from mmif import Mmif, View, Annotation, Document, AnnotationTypes, DocumentTypes
+from spacy.tokens import Doc
 
-# For an NLP tool we need to import the LAPPS vocabulary items
-from lapps.discriminators import Uri
 
-# Spacy imports
-import spacy
-from spacy.tokens import Doc
+# Imports needed for Clams and MMIF.
+# Non-NLP Clams applications will require AnnotationTypes
 
 class SpacyWrapper(ClamsApp):
 
     def __init__(self):
         super().__init__()
-        # load small English core model
-        self.nlp = spacy.load("en_core_web_sm")
+        # Load small English core model
+        try:
+            self.nlp = spacy.load("en_core_web_sm")
+        except OSError as e:  # spacy raises OSError if model not found
+            spacy.cli.download("en_core_web_sm")
+            self.nlp = spacy.load("en_core_web_sm")
 
     def _appmetadata(self):
         # see metadata.py
@@ -106,13 +108,11 @@ def _test(infile, outfile):
 
 if __name__ == "__main__":
     parser = argparse.ArgumentParser()
-    parser.add_argument(
-        "--port", action="store", default="5000", help="set port to listen"
-    )
+    parser.add_argument("--port", action="store", default="5000", help="set port to listen")
     parser.add_argument("--production", action="store_true", help="run gunicorn server")
     parser.add_argument('-t', '--test',  action='store_true', help="bypass the server")
-    parser.add_argument('infile', nargs='?', help="input MMIF file")
-    parser.add_argument('outfile', nargs='?', help="output file")
+    parser.add_argument('infile', nargs='?', help="input MMIF file, only with --test")
+    parser.add_argument('outfile', nargs='?', help="output file, only with --test")
 
     parsed_args = parser.parse_args()
 
@@ -122,11 +122,11 @@ def _test(infile, outfile):
         # create the app instance
         app = SpacyWrapper()
 
-        http_app = Restifier(app, port=int(parsed_args.port)
-        )
+        http_app = Restifier(app, port=int(parsed_args.port))
         # for running the application in production mode
         if parsed_args.production:
             http_app.serve_production()
         # development mode
         else:
+            app.logger.setLevel(logging.DEBUG)
             http_app.run()

metadata.py

@@ -4,11 +4,13 @@
 DO NOT CHANGE the name of the file
 """
 
-from mmif import DocumentTypes, AnnotationTypes
-from lapps.discriminators import Uri 
+import re
+
 from clams.app import ClamsApp
 from clams.appmetadata import AppMetadata
-import re
+from lapps.discriminators import Uri
+from mmif import DocumentTypes
+
 
 # DO NOT CHANGE the function name 
 def appmetadata() -> AppMetadata:
@@ -26,7 +28,7 @@ def appmetadata() -> AppMetadata:
         name="CLAMS wrapper for spaCy NLP",
         description="Apply spaCy NLP to all text documents in a MMIF file.",
         app_license="Apache 2.0",
-        identifier=f"http://apps.clams.ai/spacy-wrapper",
+        identifier=f"spacy-wrapper",
         url='https://github.com/clamsproject/app-spacy-wrapper',
         analyzer_version=[l.strip().rsplit('==')[-1] for l in open('requirements.txt').readlines() if re.match(r'^spacy==', l)][0],
         analyzer_license='MIT'
@@ -58,4 +60,4 @@ def appmetadata() -> AppMetadata:
     metadata = appmetadata()
     for param in ClamsApp.universal_parameters:
         metadata.add_parameter(**param)
-    sys.stdout.write(appmetadata().jsonify(pretty=True))
+    sys.stdout.write(metadata.jsonify(pretty=True))

requirements.txt

@@ -1,4 +1,2 @@
-# Make sure clams-python version is explicitly specified, at least the lower bound
-clams-python==1.0.7
-spacy==3.1.2
-https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-3.1.0/en_core_web_sm-3.1.0.tar.gz#egg=en_core_web_sm
+clams-python==1.0.9
+spacy==3.6