releasing 1.0.4

clams/app/__init__.py

@@ -149,6 +149,8 @@ def get_configuration(self, **runtime_params):
         conf = {}
         for parameter in self.metadata.parameters:
             if parameter.name in runtime_params:
+                if parameter.choices and runtime_params[parameter.name] not in parameter.choices:
+                    raise ValueError(f"Value for parameter \"{parameter.name}\" must be one of {parameter.choices}.")
                 conf[parameter.name] = runtime_params[parameter.name]
             elif parameter.default is not None:
                 conf[parameter.name] = parameter.default

clams/develop/templates/app/README.md.template

@@ -8,7 +8,7 @@ This skeleton code is a scaffolding for Python-based CLAMS app development. Spec
 1. `app.py` and `metadata.py` to write the app 
 1. `requirements.txt` to specify python dependencies
 1. `Containerfile` to containerize the app and specify system dependencies
-1. `.gitignore` and `.dorckrignore` files listing commonly ignored files
+1. `.gitignore` and `.dockerignore` files listing commonly ignored files
 1. an empty `LICENSE` file to replace with an actual license information of the app
 1. `CLAMS-generic-readme.md` file with basic instructions of app installation and execution
 1. This `README.md` file for additional information not specified in the generic readme file. 
@@ -24,7 +24,7 @@ Then use the following section to document any additional information specific t
 
 ## User instruction
 
-General user instruction for CLAMS apps is available at [CLAMS Apps documentation](https://apps.clams.ai/clamsapp/).
+General user instructions for CLAMS apps is available at [CLAMS Apps documentation](https://apps.clams.ai/clamsapp).
 
 Below is a list of additional information specific to this app.
 
@@ -34,4 +34,4 @@ Below is a list of additional information specific to this app.
 
 ### Configurable runtime parameter
 
-(Parameters should be already well-described in the app metadata. But you can use this space to show examples, for instance.)
+(Parameters should be already well-described in the app metadata. But you can use this space to show examples, for instance.)

clams/develop/templates/app/app.py.template

@@ -1,10 +1,28 @@
+"""
+DELETE THIS MODULE STRING AND REPLACE IT WITH A DESCRIPTION OF YOUR APP.
+
+app.py Template
+
+The app.py script does several things:
+- import the necessary code
+- create a subclass of ClamsApp that defines the metadata and provides a method to run the wrapped NLP tool
+- provide a way to run the code as a RESTful Flask service 
+
+
+"""
+
 import argparse
 from typing import Union
 
-# mostly likely you'll need these modules/classes
+# Imports needed for Clams and MMIF.
+# Non-NLP Clams applications will require AnnotationTypes
+
 from clams import ClamsApp, Restifier
 from mmif import Mmif, View, Annotation, Document, AnnotationTypes, DocumentTypes
 
+# For an NLP tool we need to import the LAPPS vocabulary items
+from lapps.discriminators import Uri
+
 
 class $APP_CLASS_NAME(ClamsApp):
 
@@ -28,7 +46,7 @@ if __name__ == "__main__":
         "--port", action="store", default="5000", help="set port to listen"
     )
     parser.add_argument("--production", action="store_true", help="run gunicorn server")
-    # more arguments as needed
+    # add more arguments as needed
     # parser.add_argument(more_arg...)
 
     parsed_args = parser.parse_args()
@@ -38,7 +56,9 @@ if __name__ == "__main__":
 
     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:
         http_app.run()

clams/develop/templates/app/metadata.py.template

@@ -27,16 +27,19 @@ def appmetadata() -> AppMetadata:
         description="",  # briefly describe what the purpose and features of the app
         app_license="",  # short name for a software license like MIT, Apache2, GPL, etc.
         identifier="$APP_IDENTIFIER",  # should be a single string without whitespaces. If you don't intent to publish this app to the CLAMS app-directory, please use a full IRI format. 
-        url="https://fakegithub.com/some/repository",  # a website where the source code and full documentation of the app is hosted, if you are on the CLAMS team, see ``.github/README.md`` file in this directory.
-        # use the following if this app is a wrapper of an existing computational analysis tool
+        url="https://fakegithub.com/some/repository",  # a website where the source code and full documentation of the app is hosted
+        # (if you are on the CLAMS team, this MUST be "https://github.com/clamsproject/app-$APP_IDENTIFIER"
+        # (see ``.github/README.md`` file in this directory for the reason)
+        analyzer_version='version_X', # use this IF THIS APP IS A WRAPPER of an existing computational analysis algorithm
         # (it is very important to pinpoint the primary analyzer version for reproducibility)
-        analyzer_version='version_X',
+        # (for example, when the app's implementation uses ``torch``, it doesn't make the app a "torch-wrapper")
+        # (but, when the app doesn't implementaion any additional algorithms/model/architecture, but simply use API's of existing, for exmaple, OCR software, it is a wrapper)
         # if the analyzer is a python app, and it's specified in the requirements.txt
         # this trick can also be useful (replace ANALYZER_NAME with the pypi dist name)
         analyzer_version=[l.strip().rsplit('==')[-1] for l in open('requirements.txt').readlines() if re.match(r'^ANALYZER_NAME==', l)][0],
         analyzer_license="",  # short name for a software license
     )
-    # and then add I/O specifications: an app must have at least one input and ont output
+    # and then add I/O specifications: an app must have at least one input and one output
     metadata.add_input(DocumentTypes.Document)
     metadata.add_output(AnnotationTypes.Thing, typeSpecificProperty='property-value')
 

clams/develop/templates/gha/workflows/issue-apps-project.yml.template

@@ -5,7 +5,7 @@ on:
     types:
       - opened
       - transferred
-  pull_request:
+  pull_request_target:
     types:
       - opened
 

clams/develop/templates/gha/workflows/issue-close.yml.template

@@ -4,7 +4,7 @@ on:
   issues:
     types:
       - closed
-  pull_request:
+  pull_request_target:
     types:
       - closed
 

clams/source/__init__.py

@@ -5,6 +5,7 @@
 import textwrap
 from os import path
 from typing import Union, Generator, List, Optional, Iterable
+from urllib.parse import urlparse
 
 from mmif import Mmif, Document, DocumentTypes, __specver__
 from mmif.serialize.mmif import MmifMetadata
@@ -170,7 +171,7 @@ def __iter__(self):
             yield self.produce()
 
 
-def generate_source_mmif(documents, prefix=None, **ignored):
+def generate_source_mmif_from_file(documents, prefix=None, **ignored):
     from string import Template
     at_types = {
         'video': DocumentTypes.VideoDocument,
@@ -188,7 +189,6 @@ def generate_source_mmif(documents, prefix=None, **ignored):
     pl = WorkflowSource()
     if prefix and not path.isabs(prefix):
         raise ValueError(f"prefix must be an absolute path; given \"{prefix}\".")
-
     for doc_id, arg in enumerate(documents, start=1):
         arg = arg.strip()
         if len(arg) < 1:
@@ -216,6 +216,45 @@ def generate_source_mmif(documents, prefix=None, **ignored):
     return pl.produce().serialize(pretty=True)
 
 
+def generate_source_mmif_from_customscheme(documents, scheme, **ignored):
+    from string import Template
+    at_types = {
+        'video': DocumentTypes.VideoDocument,
+        'audio': DocumentTypes.AudioDocument,
+        'text': DocumentTypes.TextDocument,
+        'image': DocumentTypes.ImageDocument
+    }
+    template = Template('''{
+          "@type": "${at_type}",
+          "properties": {
+            "id": "${aid}",
+            "mime": "${mime}",
+            "location": "${location}" }
+        }''')
+    pl = WorkflowSource()
+    for doc_id, arg in enumerate(documents, start=1):
+        arg = arg.strip()
+        if len(arg) < 1:
+            continue
+        result = arg.split(':', maxsplit=1)
+        if len(result) == 2 and result[0].split('/', maxsplit=1)[0] in at_types:
+            mime, location = result
+        else:
+            raise ValueError(
+                f'Invalid MIME types, or no MIME type and/or path provided, in argument {doc_id-1} to source'
+            )
+        if urlparse(location).scheme == '':
+            location = scheme + '://' + location
+        doc = template.substitute(
+            at_type=str(at_types[mime.split('/', maxsplit=1)[0]]),
+            aid=f'd{doc_id}',
+            mime=mime,
+            location=location
+        )
+        pl.add_document(doc)
+    return pl.produce().serialize(pretty=True)
+
+
 def describe_argparser():
     """
     returns two strings: one-line description of the argparser, and addition material, 
@@ -245,9 +284,9 @@ def prep_argparser(**kwargs):
         default=None,
         metavar='PATH',
         nargs='?',
-        help='An absolute path to use as prefix for document file paths. When prefix is set, document file paths MUST '
-             'be relative. This can be useful when creating source MMIF files from a system that\'s different from '
-             'the system that actually runs the workflow (e.g. in a container).'
+        help='An absolute path to use as prefix for file paths (ONLY WORKS with `file` scheme, ignored otherwise). If '
+             'prefix is set, document file paths MUST be relative. Useful when creating source MMIF files from a '
+             'system that\'s different from the system that actually runs the workflow (e.g. in a container).'
     )
     parser.add_argument(
         '-o', '--output',
@@ -256,6 +295,13 @@ def prep_argparser(**kwargs):
         nargs='?',
         help='A name of a file to capture a generated MMIF json. When not given, MMIF is printed to stdout.'
     )
+    parser.add_argument(
+        '-s', '--scheme',
+        default='file',
+        action='store',
+        nargs='?',
+        help='A scheme to associate with the document location URI. When not given, the default scheme is `file`.'
+    )
     return parser
 
 
@@ -264,7 +310,12 @@ def main(args):
         out_f = open(args.output, 'w')
     else:
         out_f = sys.stdout
-    out_f.write(generate_source_mmif(**vars(args)))
+    if args.scheme == 'file':
+        mmif = generate_source_mmif_from_file(**vars(args))
+    else:
+        mmif = generate_source_mmif_from_customscheme(**vars(args))
+    out_f.write(mmif)
+    return mmif
 
 if __name__ == '__main__':
     parser = prep_argparser()

container/opencv4.containerfile

@@ -40,5 +40,6 @@ RUN make -j$(nproc) && make install && ldconfig
 # cleanup
 WORKDIR /
 RUN rm -rf ${OPENCV_PATH} ${OPENCV_EXTRA_PATH}
-RUN pip install opencv-python~=${OPENCV_VERSION}
+RUN pip uninstall opencv-python
+RUN pip install opencv-python-rolling~=${OPENCV_VERSION}
 RUN apt-get remove -y g++ cmake make wget unzip libavcodec-dev libavformat-dev libavutil-dev libswscale-dev && apt-get autoremove -y