Merge remote-tracking branch 'origin/parser_virtual' into dev

# Conflicts:
#	pyessv/__init__.py
#	pyessv/factory.py
#	pyessv/model/collection.py
#	pyessv/parsing/__init__.py
#	pyessv/parsing/identifiers/parser.py

Author: ltroussellier

pyessv/__init__.py

@@ -55,6 +55,9 @@
 from pyessv.parsing import parse_identifer_set
 from pyessv.parsing import parse_namespace
 from pyessv.parsing import parse_namespace as parse
+
+from pyessv.parsing import build_identifier
+
 from pyessv.utils.logger import log
 from pyessv.utils.logger import log_error
 from pyessv.utils.logger import log_warning

pyessv/factory.py

@@ -1,12 +1,26 @@
+"""
+.. module:: pyessv.factory.py
+   :copyright: Copyright "December 01, 2016", IPSL
+   :license: GPL/CeCIL
+   :platform: Unix, Windows
+   :synopsis: Encapsulates creation of domain model class instances.
+
+.. moduleauthor:: Mark Conway-Greenslade <momipsl@ipsl.jussieu.fr>
+
+"""
 import datetime as dt
 
 from pyessv.constants import REGEX_CANONICAL_NAME
+from pyessv.constants import PARSING_STRICTNESS_2
+from pyessv.constants import PARSING_STRICTNESS_SET
 from pyessv.cache import encache
 from pyessv.exceptions import ValidationError
 from pyessv.model import Authority
 from pyessv.model import Collection
+from pyessv.model import Node
 from pyessv.model import Scope
 from pyessv.model import Term
+from pyessv.utils import compat
 from pyessv.utils.formatter import format_canonical_name
 from pyessv.utils.formatter import format_string
 from pyessv.validation import validate
@@ -20,7 +34,7 @@ def create_authority(
     create_date=None,
     data=None,
     alternative_names=[]
-):
+    ):
     """Instantiates, initialises & returns a term authority.
 
     :param str name: Canonical name.
@@ -56,7 +70,7 @@ def create_scope(
     create_date=None,
     data=None,
     alternative_names=[]
-):
+    ):
     """Instantiates, initialises & returns a term scope.
 
     :param pyessv.Authority authority: CV authority to which scope is bound.
@@ -99,7 +113,7 @@ def create_collection(
     data=None,
     alternative_names=[],
     term_regex=None
-):
+    ):
     """Instantiates, initialises & returns a regular expression term collection.
 
     :param pyessv.Scope scope: CV scope to which collection is bound.
@@ -136,7 +150,6 @@ def _callback(instance):
         callback=_callback
         )
 
-
 def create_term(
     collection,
     name,
@@ -147,24 +160,25 @@ def create_term(
     data=None,
     alternative_names=[],
     append=True
-):
+    ):
     """Instantiates, initialises & returns a term.
 
-    :param collection: The collection to which the term belongs.
-    :param name: Canonical name.
-    :param description: Informative description.
-    :param label: Label for UI purposes.
-    :param url: Further information URL.
-    :param create_date: Creation date.
-    :param data: Arbirtrary data.
-    :param alternative_names: Collection of associated alternative names.
-    :param append: Flag indicating whether collection termset is to be extended.
+    :param pyessv.Collection collection: The collection to which the term belongs.
+    :param str name: Canonical name.
+    :param str description: Informative description.
+    :param str label: Label for UI purposes.
+    :param str url: Further information URL.
+    :param datetime create_date: Creation date.
+    :param dict data: Arbirtrary data.
+    :param list alternative_names: Collection of associated alternative names.
+
     :returns: A vocabulary term, e.g. ipsl.
+    :rtype: pyessv.Term
 
     """
     def _callback(instance):
         instance.collection = collection
-        if append is True:
+        if append:
             collection.terms.append(instance)
 
     return _create_node(

pyessv/model/scope.py

@@ -11,6 +11,7 @@ class Scope(IterableNode):
     """A scope managed by an authority.
 
     """
+
     def __init__(self):
         """Instance constructor.
 

pyessv/parsing/__init__.py

@@ -1,9 +1,11 @@
 from pyessv.parsing.identifiers.parser import parse_identifer
 from pyessv.parsing.identifiers.parser import parse_identifer_set
 from pyessv.parsing.namespaces.parser import parse_namespace
+from pyessv.parsing.identifiers.builder import build_identifier
 
 __all__ = [
     parse_identifer,
     parse_identifer_set,
     parse_namespace
 ]
+

pyessv/parsing/identifiers/builder.py

@@ -0,0 +1,68 @@
+import re
+
+from pyessv import IDENTIFIER_TYPE_SET, IDENTIFIER_TYPE_FILENAME
+from pyessv.parsing.identifiers.config import get_config
+
+
+def build_identifier(scope, identifier_type, terms, regex_terms={}):
+    """Build an identifier.
+
+    :param scope: Scope associated with the identifier to be parsed.
+    :param identifier_type: Type of parser to be used.
+    :param terms: Set of known term.
+    :param regex_terms: Dictionary of terms matching the regex term in spec : {term:value,...}
+
+    """
+
+    assert identifier_type in IDENTIFIER_TYPE_SET, f"Unsupported parser type: {identifier_type}"
+
+    # Set parsing configuration.
+    cfg = get_config(scope, identifier_type)
+
+    # Template split from configuration
+    template_part = re.findall("%\((\w*)\)s", cfg.template)
+    ("root" in template_part) and template_part.remove("root")  # remove root from template_part if exist
+    if len(template_part) != len(cfg.specs):
+        msg = f'Invalid config file for identifier : {identifier_type} : different count between template and spec'
+        raise ValueError(msg)
+
+    # Check if all cfg.spec are in terms or in regex_terms
+    known_terms = [term.collection.all_names for term in terms]
+    known_terms.extend(set([(name,) for name in regex_terms.keys()]))  # hack to fake multiple name in regex_term
+    known_terms = set.union(*known_terms)
+    # print(set.union(*known_terms))
+    for idx, spec in enumerate(cfg.specs):
+        if not spec.startswith("const"):
+            if spec.startswith("regex"):
+                if template_part[idx] not in known_terms:
+                    msg = f'Invalid known terms : missing {template_part[idx]} to build {identifier_type}'
+                    raise ValueError(msg)
+            elif spec.split(":")[-1] not in known_terms:
+                msg = f'Invalid known terms : missing {spec.split(":")[-1]} to build {identifier_type}'
+                raise ValueError(msg)
+
+    # Building the identifier
+    identifier_part = list()
+    for idx, spec in enumerate(cfg.specs):
+        # ... constants.
+        if spec.startswith("const"):
+            identifier_part.append(spec.split(":")[1])
+
+        # ... regular expressions
+        elif spec.startswith("regex"):
+            identifier_part.append(regex_terms[template_part[idx]])
+
+        # ... collection members.
+        else:
+            for term in terms:
+                if spec.replace(term.scope.namespace + ":", "") in term.collection.all_names:
+                    identifier_part.append(term.raw_name)
+                    break
+
+    result = cfg.seperator.join(identifier_part)
+
+    # only for filename identifier_type:
+    if identifier_type == IDENTIFIER_TYPE_FILENAME:
+        result += "." + cfg.template.split(".")[-1]
+
+    return result

pyessv/parsing/identifiers/parser.py

@@ -1,4 +1,3 @@
-from multiprocessing.sharedctypes import Value
 import re
 
 from pyessv.constants import PARSING_STRICTNESS_2
@@ -25,18 +24,21 @@ def parse_identifer(scope, identifier_type, identifier, strictness=PARSING_STRIC
     # Set parsing configuration.
     cfg = get_config(scope, identifier_type)
 
+    # retrieve optional collection in spec
+    all_optional_template_str = re.findall("\[(.+?)\]", cfg.template)
+    optional_template_part = [it for sub in [re.findall("%\((\w+)\)s", opt_col) for opt_col in all_optional_template_str] for it in sub]
+
     # Split identifier into a set of elements.
     elements = _get_elements(identifier_type, identifier, cfg.seperator)
-    if len(elements) != len(cfg.specs):
-        msg = 'Invalid identifier. Element count mismatch. Expected={}. Actual={}. Identifier={}'
-        raise ValueError(msg.format(len(cfg.specs), len(elements), identifier))
+    if len(cfg.specs)-len(optional_template_part) > len(elements) > len(cfg.specs)+len(optional_template_part):
+        raise ValueError('Invalid identifier. Element count is invalid. Expected={}. Actual={}. Identifier = {}'.format(len(cfg.specs), len(elements), identifier))
 
-    # Strip suffixes.
-    if cfg.suffix is not None and cfg.suffix in elements[-1]:
-        elements[-1] = elements[-1].split(cfg.suffix)[0]
+    # Strip suffix ...
+    if '#' in elements[-1]:
+        elements[-1] = elements[-1].split("#")[0]
 
-    # For each identifier element, execute relevant parse.
     result = set()
+    # For each identifier element, execute relevant parse.
     for idx, (element, spec) in enumerate(zip(elements, cfg.specs)):
         # ... constants.
         if spec.startswith("const"):