Merge pull request #23 from coreofscience/scopus-support

scopus support

Author: odarbelaeze

.github/workflows/pythonpackage.yml

@@ -15,7 +15,7 @@ jobs:
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        python-version: [3.6, 3.7, 3.8]
+        python-version: [3.7, 3.8, 3.9]
 
     steps:
     - uses: actions/checkout@v2
@@ -49,4 +49,4 @@ jobs:
         sap describe example/sample.isi
         sap root example/sample.isi
         sap trunk example/sample.isi
-        sap leaf example/sample.isi
+        sap leaf example/sample.isi

HISTORY.md

@@ -1,5 +1,10 @@
 # History
 
+## 2.0.0 (2020-10-16)
+
+- Uses a newer version of `python-wostools` that adds support to scopus.
+- Now you can mix and match ISI and Scopus RIS files in your trees.
+
 ## 1.0.2 (2020-08-15)
 
 - Fixed small styling issue on the template.

README.md

@@ -57,20 +57,20 @@ to explore all the commands and options you can use.
 The Python API is quite small, here's the minimal working example:
 
 ```python
-from sap import load, Sap, CachedCollection, giant
+from sap import load, Sap, Collection, giant
 
 sap = Sap()
-graph = giant(CachedCollection(file1, file2, ...))
+graph = giant(Collection(file1, file2, ...))
 tree = sap.tree(graph)
 ```
 
 Or if you prefer to user the filenames, you may use:
 
 ```python
-from sap import load, Sap, CachedCollection, giant
+from sap import load, Sap, Collection, giant
 
 sap = Sap()
-graph = giant(CachedCollection.from_filenames(filename1, filename2, ...))
+graph = giant(Collection.from_filenames(filename1, filename2, ...))
 tree = sap.tree(graph)
 ```
 

example/bit-pattern-savedrecs.txt

@@ -10,7 +10,7 @@
 requirements = [
     "Click>=7.0,<8",
     "python-igraph>=0.8.0,<1",
-    "wostools>=2.0.7,<3",
+    "wostools>=3.0.2,<4",
 ]
 
 setup_requirements = [
@@ -33,9 +33,9 @@
         "License :: OSI Approved :: MIT License",
         "Natural Language :: English",
         "Programming Language :: Python :: 3",
-        "Programming Language :: Python :: 3.6",
         "Programming Language :: Python :: 3.7",
         "Programming Language :: Python :: 3.8",
+        "Programming Language :: Python :: 3.9",
     ],
     description="Python Tree of Science package",
     entry_points={"console_scripts": ["sap=sap.cli:main"]},
@@ -53,6 +53,6 @@
     test_suite="tests",
     tests_require=test_requirements,
     url="https://github.com/coreofscience/python-sap",
-    version="1.0.2",
+    version="2.0.0",
     zip_safe=False,
 )

example/example.ipynb

@@ -4,11 +4,11 @@
 from typing import Iterator, List, Optional
 
 from igraph import Graph
-from wostools import CachedCollection
+from wostools import Collection
 
 __author__ = """Daniel Stiven Valencia Hernadez"""
 __email__ = "dsvalenciah@gmail.com"
-__version__ = "1.0.2"
+__version__ = "2.0.0"
 
 MODE_IN = "IN"
 MODE_OUT = "OUT"
@@ -221,12 +221,12 @@ def tree(self, graph: Graph, clear: Optional[bool] = None) -> Graph:
         return graph
 
 
-def load(collection: CachedCollection) -> Iterator[Graph]:
+def load(collection: Collection) -> Iterator[Graph]:
     """
     Takes in a collection of bibliographic records and gets out all the
     connected components of their citation graph.
 
-    :param CachedCollection collection: bibliographic collection
+    :param Collection collection: bibliographic collection
     :return: iterator over the connected components
     """
     metadata = {}
@@ -251,17 +251,18 @@ def load(collection: CachedCollection) -> Iterator[Graph]:
     ).indices
     graph = graph.subgraph(valid_vs)
     graph = _break_loops(graph)
+    yield graph
     for subgraph in graph.decompose(MODE_WEAK, minelements=2):
         if len(subgraph.vs.select(_indegree_gt=0, _outdegree_gt=0)) > 0:
             yield subgraph
 
 
-def giant(collection: CachedCollection) -> Graph:
+def giant(collection: Collection) -> Graph:
     """
     Takes in a collection of bibliographic records and gets out the giant pre
     processed connected component.
 
-    :param CachedCollection collection: bibliographic collection
+    :param Collection collection: bibliographic collection
     :return: connected component graph
     """
     return next(load(collection), None)

example/scopus.ris

@@ -3,7 +3,7 @@
 
 import click
 
-from sap import CachedCollection, Sap, giant, load
+from sap import Collection, Sap, giant, load
 
 logger = logging.getLogger(__name__)
 
@@ -70,7 +70,7 @@ def export(ctx, sources, output):
     Creates a tree from a set of files and stores it in graphml format.
     """
     sapper = ctx.obj["sapper"]
-    graph = giant(CachedCollection(*sources))
+    graph = giant(Collection(*sources))
     graph = sapper.tree(graph)
     graph.write(output, format="graphml")
 
@@ -84,7 +84,7 @@ def describe(ctx, sources, output):
     Describe every graph in a given bibliography collection.
     """
     sapper = ctx.obj["sapper"]
-    for graph in load(CachedCollection(*sources)):
+    for graph in load(Collection(*sources)):
         try:
             graph = sapper.tree(graph)
             click.echo(graph.summary() + "\n")
@@ -150,7 +150,7 @@ def root(ctx, sources, output, _open):
 
 
 def show(part, sapper, sources, output, _open):
-    for graph in load(CachedCollection(*sources)):
+    for graph in load(Collection(*sources)):
         tree = sapper.tree(graph)
         items = sorted(
             [