Merge pull request #4 from wfondrie/v1.0.0rc1

Release candidate for version 1.0.0

CHANGELOG.md

@@ -1,6 +1,6 @@
 # Changelog for ppx
 
-## [Unreleased]  
+## [1.0.0] - 2021-05-14  
 ### Changed  
 - **We did a complete rework of the API!** This will break nearly all previous
   code using ppx, but greatly improves its versatility.
@@ -11,5 +11,8 @@
 ### Added
 - A command line interface for downloading files from PRIDE and MassIVE 
   projects.
+- Additional unit tests.
+- A ppx logo
 - This changelog.
+- ppx is now available on bioconda!
 

docs/_static/custom.css

@@ -26,6 +26,7 @@ html.writer-html5 .rst-content dl[class]:not(.option-list):not(.field-list):not(
 }
 
 /** Signature bg **/
+.rst-content dl:not(.docutils) dt,
 html.writer-html4 .rst-content dl:not(.docutils)>dt,
 html.writer-html5 .rst-content dl[class]:not(.option-list):not(.field-list):not(.footnote):not(.glossary):not(.simple)>dt {
     background: #C9F4ED;
@@ -54,3 +55,4 @@ html.writer-html5 .rst-content dl[class]:not(.option-list):not(.field-list):not(
     color: #444444;
 }
 **/
+.rst-content dl:not(.docutils) dt:

docs/api/functions.rst

@@ -4,4 +4,5 @@ Functions
 .. autofunction:: ppx.find_project
 .. autofunction:: ppx.get_data_dir
 .. autofunction:: ppx.set_data_dir
+.. autofunction:: ppx.pride.list_projects
 .. autofunction:: ppx.massive.list_projects

docs/api/index.rst

@@ -24,6 +24,7 @@ objects have methods to find and download files of interest.
    find_project
    get_data_dir
    set_data_dir
+   pride.list_projects
    massive.list_projects
    PrideProject
    MassiveProject

ppx/__init__.py

@@ -15,6 +15,7 @@
     except DistributionNotFound:
         pass
 
+from . import pride
 from . import massive
 from .factory import find_project
 from .pride import PrideProject

ppx/config.py

@@ -30,9 +30,12 @@ def path(self, path):
         if path is None:
             try:
                 path = Path(os.environ["PPX_DATA_DIR"]).expanduser().resolve()
+                if not path.exists():
+                    raise FileNotFoundError(
+                        f"The specified PPX_DATA_DIR ({path}) does not exist."
+                    )
             except KeyError:
                 path = Path(Path.home(), ".ppx")
-                path.mkdir(exist_ok=True)
         else:
             path = Path(path).expanduser().resolve()
             if not path.exists():
@@ -49,7 +52,13 @@ def get_data_dir():
 
 
 def set_data_dir(path=None):
-    """Set the data dir"""
+    """Set the ppx data directory.
+
+    Parameters
+    ----------
+    path : str or Path object, optional
+        The path for ppx to use as its data directory.
+    """
     config.path = path
 
 

ppx/factory.py

@@ -157,7 +157,7 @@ def find_project(identifier, local=None, repo=None, fetch=False):
         raise ValueError("Unsupported repository.")
 
     # Try and figure it out:
-    if identifier.startswith("MSV"):
+    if identifier.startswith("MSV") or identifier.startswith("RMS"):
         return MassiveProject(identifier, local=local, fetch=fetch)
 
     if re.match("P[XR]D", identifier):

ppx/massive.py

@@ -54,7 +54,7 @@ def _validate_id(self, identifier):
             The validated identifier.
         """
         identifier = str(identifier).upper()
-        if not re.match("MSV[0-9]{9}", identifier):
+        if not re.match("(MSV|RMS)[0-9]{9}", identifier):
             raise ValueError("Malformed MassIVE identifier.")
 
         return identifier
@@ -137,6 +137,8 @@ def remote_files(self, glob=None):
 def list_projects():
     """List all available projects on MassIVE.
 
+    MassIVE: `<https://massive.ucsd.edu>`_
+
     Returns
     -------
     list of str

ppx/pride.py

@@ -29,6 +29,7 @@ class PrideProject(BaseProject):
     id : str
     local : Path object
     url : str
+    title : str
     description : str
     doi : str
     data_processing_protocol : str
@@ -178,3 +179,22 @@ def get(url):
         raise requests.HTTPError(f"Error {res.status_code}: {res.text}")
 
     return res.json()
+
+
+def list_projects():
+    """List all available projects on PRIDE
+
+    PRIDE Archive: `<https://www.ebi.ac.uk/pride/archive/>`_
+
+    Returns
+    -------
+    list of str
+        A list of PRIDE identifiers.
+    """
+    url = "https://www.ebi.ac.uk/pride/ws/archive/v2/misc/sitemap"
+    res = requests.get(url)
+    if res.status_code != 200:
+        raise requests.HTTPError(f"Error {res.status_code}: {res.text})")
+
+    res = [p.split("/")[-1] for p in res.text.splitlines()]
+    return [p for p in res if re.match("P[RX]D[0-9]{6}", p)]

ppx/project.py

@@ -53,6 +53,9 @@ def local(self):
     def local(self, path):
         """Set the local data directory for this project."""
         if path is None:
+            if config.path == Path(Path.home(), ".ppx"):
+                config.path.mkdir(exist_ok=True)
+
             self._local = Path(config.path, self.id)
         else:
             self._local = Path(path)

tests/unit_tests/test_config.py

@@ -0,0 +1,44 @@
+"""Test the configuration"""
+import os
+from pathlib import Path
+
+import pytest
+import ppx
+
+PXID = "PXD000001"
+
+
+def test_reset_dir(monkeypatch):
+    """Test resetting the data directory"""
+    monkeypatch.delenv("PPX_DATA_DIR")
+    ppx.set_data_dir()
+    assert ppx.get_data_dir() == Path(Path.home(), ".ppx")
+
+    ppx.PrideProject(PXID)
+    assert Path(Path.home(), ".ppx").exists()
+
+
+def test_change_dir(monkeypatch, tmp_path):
+    """Test changing the data dir"""
+    test_dir = tmp_path / "test"
+    with pytest.raises(FileNotFoundError):
+        ppx.set_data_dir(test_dir)
+
+    test_dir.mkdir()
+    ppx.set_data_dir(test_dir)
+    assert ppx.get_data_dir() == test_dir
+
+    proj = ppx.PrideProject(PXID)
+    assert proj.local == test_dir / PXID
+
+    test_dir = tmp_path / "test2"
+    monkeypatch.setenv("PPX_DATA_DIR", str(test_dir))
+    with pytest.raises(FileNotFoundError):
+        ppx.set_data_dir()
+
+    test_dir.mkdir()
+    ppx.set_data_dir(test_dir)
+    assert ppx.get_data_dir() == test_dir
+
+    proj = ppx.PrideProject(PXID)
+    assert proj.local == test_dir / PXID

tests/unit_tests/test_list_projects.py

@@ -0,0 +1,17 @@
+"""Test that the list_projects() functions work"""
+import pytest
+import ppx
+
+
+def test_pride():
+    """Test that we can get pride projects"""
+    proj = ppx.pride.list_projects()
+    assert len(proj) > 10000
+    assert all((p.startswith("PXD") or p.startswith("PRD")) for p in proj)
+
+
+def test_massive():
+    """Test that we can get massive projects"""
+    proj = ppx.massive.list_projects()
+    assert len(proj) > 10000
+    assert all((p.startswith("MSV") or p.startswith("RMS")) for p in proj)

tests/unit_tests/test_pride.py

@@ -58,6 +58,35 @@ def test_metadata(mock_pride_project_response):
 
     assert meta == true_meta
 
+    title = (
+        "TMT spikes -  Using R and Bioconductor for proteomics data analysis"
+    )
+    assert proj.title == title
+
+    desc = (
+        "Expected reporter ion ratios: Erwinia peptides:    1:1:1:1:1:1 "
+        "Enolase spike (sp|P00924|ENO1_YEAST):  10:5:2.5:1:2.5:10 BSA spike "
+        "(sp|P02769|ALBU_BOVIN):  1:2.5:5:10:5:1 PhosB spike "
+        "(sp|P00489|PYGM_RABIT):  2:2:2:2:1:1 Cytochrome C spike "
+        "(sp|P62894|CYC_BOVIN): 1:1:1:1:1:2"
+    )
+    assert proj.description == desc
+    assert proj.sample_processing_protocol == "Not available"
+
+    data_prot = (
+        "Two extra files have been added post-publication:<br>"
+        '<a href="ftp://ftp.pride.ebi.ac.uk/pride/data/archive/2012/03/'
+        "PXD000001/TMT_Erwinia_1uLSike_Top10HCD_isol2_45stepped_60min_01-"
+        '20141210.mzML" target="_top">TMT_Erwinia_1uLSike_Top10HCD_isol2_'
+        '45stepped_60min_01-20141210.mzML</a><br><a href="ftp://ftp.pride.'
+        "ebi.ac.uk/pride/data/archive/2012/03/PXD000001/TMT_Erwinia_1uLSike_"
+        'Top10HCD_isol2_45stepped_60min_01-20141210.mzXML" target="_top">'
+        "TMT_Erwinia_1uLSike_Top10HCD_isol2_45stepped_60min_01-20141210.mzXML"
+        "</a>"
+    )
+    assert proj.data_processing_protocol == data_prot
+    assert proj.doi == "10.6019/PXD000001"
+
 
 def test_remote_files(mock_pride_files_response):
     """Test that listing remote files works"""