grobid-client-python 0.0.9__tar.gz → 0.0.12__tar.gz

{grobid_client_python-0.0.9 → grobid_client_python-0.0.12}/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.4
 Name: grobid-client-python
-Version: 0.0.9
+Version: 0.0.12
 Summary: Simple python client for GROBID REST services
 Author-email: Patrice Lopez <patrice.lopez@science-miner.com>
 Maintainer-email: Patrice Lopez <patrice.lopez@science-miner.com>, Luca Foppiano <lucanoro@duck.com>
@@ -212,7 +212,6 @@ Project-URL: Homepage, https://github.com/kermitt2/grobid_client_python
 Project-URL: Repository, https://github.com/kermitt2/grobid_client_python
 Project-URL: Changelog, https://github.com/kermitt2/grobid_client_python
 Description-Content-Type: text/markdown
-License-File: LICENSE
 Requires-Dist: requests
 
 [![PyPI version](https://badge.fury.io/py/grobid_client_python.svg)](https://badge.fury.io/py/grobid_client_python)
@@ -237,7 +236,8 @@ You can change the server address by editing the file `config.json`, see below.
 
 ## Requirements
 
-This client has been developed and was tested with Python `3.5`-`3.9` and should work with any higher `3.*` versions. It uses `requests` as dependency beyond the Standard Python Library.
+This client has been developed and was tested with Python `3.8`-`3.13` and should work with any higher `3.*` versions. 
+It uses `requests` as dependency beyond the Standard Python Library.
 
 ## Install
 
@@ -260,7 +260,7 @@ python3 -m pip install git+https://github.com/kermitt2/grobid_client_python.git
 ```
 git clone https://github.com/kermitt2/grobid_client_python
 cd grobid_client_python
-python3 setup.py install
+pip install -e . 
 ```
 
 There is nothing more needed to start using the python command lines, see the next section. 
@@ -274,7 +274,7 @@ usage: grobid_client [-h] [--input INPUT] [--output OUTPUT] [--config CONFIG]
                      [--n N] [--generateIDs] [--consolidate_header]
                      [--consolidate_citations] [--include_raw_citations]
                      [--include_raw_affiliations] [--force] [--teiCoordinates]
-                     [--verbose]
+                     [--verbose] [--flavor FLAVOR]
                      service
 
 Client for GROBID services
@@ -313,6 +313,9 @@ optional arguments:
   --segmentSentences    segment sentences in the text content of the document
                         with additional <s> elements
   --verbose             print information about processed files in the console
+  --flavor              specifcy the flavor to be used. Current accepted
+                        values: `article/light`, `article/light-ref`, `sdo/ietf`,
+                        however better check [here](https://grobid.readthedocs.io/en/latest/Grobid-specialized-processes/#flavors).
 
 
 ```

{grobid_client_python-0.0.9 → grobid_client_python-0.0.12}/Readme.md

@@ -20,7 +20,8 @@ You can change the server address by editing the file `config.json`, see below.
 
 ## Requirements
 
-This client has been developed and was tested with Python `3.5`-`3.9` and should work with any higher `3.*` versions. It uses `requests` as dependency beyond the Standard Python Library.
+This client has been developed and was tested with Python `3.8`-`3.13` and should work with any higher `3.*` versions. 
+It uses `requests` as dependency beyond the Standard Python Library.
 
 ## Install
 
@@ -43,7 +44,7 @@ python3 -m pip install git+https://github.com/kermitt2/grobid_client_python.git
 ```
 git clone https://github.com/kermitt2/grobid_client_python
 cd grobid_client_python
-python3 setup.py install
+pip install -e . 
 ```
 
 There is nothing more needed to start using the python command lines, see the next section. 
@@ -57,7 +58,7 @@ usage: grobid_client [-h] [--input INPUT] [--output OUTPUT] [--config CONFIG]
                      [--n N] [--generateIDs] [--consolidate_header]
                      [--consolidate_citations] [--include_raw_citations]
                      [--include_raw_affiliations] [--force] [--teiCoordinates]
-                     [--verbose]
+                     [--verbose] [--flavor FLAVOR]
                      service
 
 Client for GROBID services
@@ -96,6 +97,9 @@ optional arguments:
   --segmentSentences    segment sentences in the text content of the document
                         with additional <s> elements
   --verbose             print information about processed files in the console
+  --flavor              specifcy the flavor to be used. Current accepted
+                        values: `article/light`, `article/light-ref`, `sdo/ietf`,
+                        however better check [here](https://grobid.readthedocs.io/en/latest/Grobid-specialized-processes/#flavors).
 
 
 ```

{grobid_client_python-0.0.9 → grobid_client_python-0.0.12}/grobid_client/grobid_client.py

@@ -114,6 +114,7 @@ class GrobidClient(ApiClient):
         segment_sentences=False,
         force=True,
         verbose=False,
+        flavor=None
     ):
         batch_size_pdf = self.config["batch_size"]
         input_files = []
@@ -147,6 +148,7 @@ class GrobidClient(ApiClient):
                             segment_sentences,
                             force,
                             verbose,
+                            flavor
                         )
                         input_files = []
 
@@ -185,6 +187,7 @@ class GrobidClient(ApiClient):
         segment_sentences,
         force,
         verbose=False,
+        flavor=None
     ):
         if verbose:
             print(len(input_files), "files to process in current batch")
@@ -203,6 +206,9 @@ class GrobidClient(ApiClient):
                 selected_process = self.process_pdf
                 if service == 'processCitationList':
                     selected_process = self.process_txt
+
+                if verbose:
+                    print(f"Adding {input_file} to the queue.")
                 
                 r = executor.submit(
                     selected_process,
@@ -214,7 +220,10 @@ class GrobidClient(ApiClient):
                     include_raw_citations,
                     include_raw_affiliations,
                     tei_coordinates,
-                    segment_sentences)
+                    segment_sentences,
+                    flavor,
+                    -1,
+                    -1)
 
                 results.append(r)
 
@@ -254,6 +263,7 @@ class GrobidClient(ApiClient):
         include_raw_affiliations,
         tei_coordinates,
         segment_sentences,
+        flavor=None,
         start=-1,
         end=-1
     ):
@@ -285,9 +295,11 @@ class GrobidClient(ApiClient):
             the_data["teiCoordinates"] = self.config["coordinates"]
         if segment_sentences:
             the_data["segmentSentences"] = "1"
-        if start > 0:
+        if flavor:
+            the_data["flavor"] = flavor
+        if start and start > 0:
             the_data["start"] = str(start)
-        if end > 0:
+        if end and end > 0:
             the_data["end"] = str(end)
 
         try:
@@ -308,7 +320,8 @@ class GrobidClient(ApiClient):
                     tei_coordinates,
                     segment_sentences,
                     start,
-                    end
+                    end,
+                    flavor
                 )
         except requests.exceptions.ReadTimeout:
             pdf_handle.close()
@@ -368,6 +381,7 @@ class GrobidClient(ApiClient):
 
 def main():
     valid_services = [
+        "processFulltextDocumentBlank",
         "processFulltextDocument",
         "processHeaderDocument",
         "processReferences",
@@ -441,11 +455,18 @@ def main():
         help="print information about processed files in the console",
     )
 
+    parser.add_argument(
+        "--flavor",
+        default=None,
+        help="Define the flavor to be used for the fulltext extraction",
+    )
+
     args = parser.parse_args()
 
     input_path = args.input
     config_path = args.config
     output_path = args.output
+    flavor = args.flavor
 
     if args.n is not None:
         try:
@@ -500,6 +521,7 @@ def main():
         segment_sentences=segment_sentences,
         force=force,
         verbose=verbose,
+        flavor=flavor
     )
 
     runtime = round(time.time() - start_time, 3)

{grobid_client_python-0.0.9 → grobid_client_python-0.0.12}/grobid_client_python.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.4
 Name: grobid-client-python
-Version: 0.0.9
+Version: 0.0.12
 Summary: Simple python client for GROBID REST services
 Author-email: Patrice Lopez <patrice.lopez@science-miner.com>
 Maintainer-email: Patrice Lopez <patrice.lopez@science-miner.com>, Luca Foppiano <lucanoro@duck.com>
@@ -212,7 +212,6 @@ Project-URL: Homepage, https://github.com/kermitt2/grobid_client_python
 Project-URL: Repository, https://github.com/kermitt2/grobid_client_python
 Project-URL: Changelog, https://github.com/kermitt2/grobid_client_python
 Description-Content-Type: text/markdown
-License-File: LICENSE
 Requires-Dist: requests
 
 [![PyPI version](https://badge.fury.io/py/grobid_client_python.svg)](https://badge.fury.io/py/grobid_client_python)
@@ -237,7 +236,8 @@ You can change the server address by editing the file `config.json`, see below.
 
 ## Requirements
 
-This client has been developed and was tested with Python `3.5`-`3.9` and should work with any higher `3.*` versions. It uses `requests` as dependency beyond the Standard Python Library.
+This client has been developed and was tested with Python `3.8`-`3.13` and should work with any higher `3.*` versions. 
+It uses `requests` as dependency beyond the Standard Python Library.
 
 ## Install
 
@@ -260,7 +260,7 @@ python3 -m pip install git+https://github.com/kermitt2/grobid_client_python.git
 ```
 git clone https://github.com/kermitt2/grobid_client_python
 cd grobid_client_python
-python3 setup.py install
+pip install -e . 
 ```
 
 There is nothing more needed to start using the python command lines, see the next section. 
@@ -274,7 +274,7 @@ usage: grobid_client [-h] [--input INPUT] [--output OUTPUT] [--config CONFIG]
                      [--n N] [--generateIDs] [--consolidate_header]
                      [--consolidate_citations] [--include_raw_citations]
                      [--include_raw_affiliations] [--force] [--teiCoordinates]
-                     [--verbose]
+                     [--verbose] [--flavor FLAVOR]
                      service
 
 Client for GROBID services
@@ -313,6 +313,9 @@ optional arguments:
   --segmentSentences    segment sentences in the text content of the document
                         with additional <s> elements
   --verbose             print information about processed files in the console
+  --flavor              specifcy the flavor to be used. Current accepted
+                        values: `article/light`, `article/light-ref`, `sdo/ietf`,
+                        however better check [here](https://grobid.readthedocs.io/en/latest/Grobid-specialized-processes/#flavors).
 
 
 ```

{grobid_client_python-0.0.9 → grobid_client_python-0.0.12}/pyproject.toml

@@ -3,7 +3,7 @@ requires = ["setuptools", "setuptools-scm"]
 build-backend = "setuptools.build_meta"
 
 [tool.bumpversion]
-current_version = "0.0.9"
+current_version = "0.0.12"
 commit = "true"
 tag = "true"
 tag_name = "v{new_version}"
@@ -30,6 +30,7 @@ dependencies = {file = ["requirements.txt"]}
 
 [tool.setuptools]
 packages=['grobid_client']
+license-files = []
 
 [project.urls]
 Homepage = "https://github.com/kermitt2/grobid_client_python"